{
  "best_metric": 1.95067668,
  "best_model_checkpoint": "/scratch/ms-swift-chatas/exp_output_qwen2_vl_imagechat/v1-20250529-055538/checkpoint-103000",
  "epoch": 4.412835782528598,
  "eval_steps": 500,
  "global_step": 103000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 4.284306584979221e-05,
      "grad_norm": 8.035392761230469,
      "learning_rate": 9.999999998188407e-05,
      "loss": 4.592377185821533,
      "memory(GiB)": 10.09,
      "step": 1,
      "token_acc": 0.33783783783783783,
      "train_speed(iter/s)": 0.352947
    },
    {
      "epoch": 0.00021421532924896104,
      "grad_norm": 7.244219779968262,
      "learning_rate": 9.999999954710156e-05,
      "loss": 4.315703392028809,
      "memory(GiB)": 10.09,
      "step": 5,
      "token_acc": 0.31620553359683795,
      "train_speed(iter/s)": 0.883762
    },
    {
      "epoch": 0.0004284306584979221,
      "grad_norm": 5.759903907775879,
      "learning_rate": 9.99999981884062e-05,
      "loss": 3.7349918365478514,
      "memory(GiB)": 17.02,
      "step": 10,
      "token_acc": 0.3154121863799283,
      "train_speed(iter/s)": 1.092766
    },
    {
      "epoch": 0.0006426459877468832,
      "grad_norm": 6.463886737823486,
      "learning_rate": 9.999999592391398e-05,
      "loss": 3.3617221832275392,
      "memory(GiB)": 21.59,
      "step": 15,
      "token_acc": 0.3447204968944099,
      "train_speed(iter/s)": 1.177054
    },
    {
      "epoch": 0.0008568613169958442,
      "grad_norm": 5.3920979499816895,
      "learning_rate": 9.999999275362494e-05,
      "loss": 3.489911651611328,
      "memory(GiB)": 21.59,
      "step": 20,
      "token_acc": 0.4070175438596491,
      "train_speed(iter/s)": 1.254055
    },
    {
      "epoch": 0.0010710766462448053,
      "grad_norm": 4.974893093109131,
      "learning_rate": 9.999998867753912e-05,
      "loss": 3.31513671875,
      "memory(GiB)": 26.72,
      "step": 25,
      "token_acc": 0.3643410852713178,
      "train_speed(iter/s)": 1.238349
    },
    {
      "epoch": 0.0012852919754937663,
      "grad_norm": 4.026622772216797,
      "learning_rate": 9.999998369565659e-05,
      "loss": 3.0826137542724608,
      "memory(GiB)": 26.72,
      "step": 30,
      "token_acc": 0.39664804469273746,
      "train_speed(iter/s)": 1.233881
    },
    {
      "epoch": 0.0014995073047427273,
      "grad_norm": 4.793406009674072,
      "learning_rate": 9.999997780797748e-05,
      "loss": 3.117991638183594,
      "memory(GiB)": 26.72,
      "step": 35,
      "token_acc": 0.4230769230769231,
      "train_speed(iter/s)": 1.280209
    },
    {
      "epoch": 0.0017137226339916883,
      "grad_norm": 7.271947860717773,
      "learning_rate": 9.999997101450185e-05,
      "loss": 3.013927459716797,
      "memory(GiB)": 26.72,
      "step": 40,
      "token_acc": 0.4099722991689751,
      "train_speed(iter/s)": 1.286531
    },
    {
      "epoch": 0.0019279379632406496,
      "grad_norm": 4.337677478790283,
      "learning_rate": 9.999996331522983e-05,
      "loss": 2.9280925750732423,
      "memory(GiB)": 26.72,
      "step": 45,
      "token_acc": 0.4026315789473684,
      "train_speed(iter/s)": 1.323395
    },
    {
      "epoch": 0.0021421532924896106,
      "grad_norm": 7.95403528213501,
      "learning_rate": 9.99999547101616e-05,
      "loss": 3.090596008300781,
      "memory(GiB)": 26.72,
      "step": 50,
      "token_acc": 0.38127090301003347,
      "train_speed(iter/s)": 1.337204
    },
    {
      "epoch": 0.002356368621738572,
      "grad_norm": 11.402766227722168,
      "learning_rate": 9.999994519929725e-05,
      "loss": 3.0341163635253907,
      "memory(GiB)": 26.72,
      "step": 55,
      "token_acc": 0.38175675675675674,
      "train_speed(iter/s)": 1.372759
    },
    {
      "epoch": 0.0025705839509875326,
      "grad_norm": 6.415196418762207,
      "learning_rate": 9.9999934782637e-05,
      "loss": 2.797837829589844,
      "memory(GiB)": 26.72,
      "step": 60,
      "token_acc": 0.4051094890510949,
      "train_speed(iter/s)": 1.397544
    },
    {
      "epoch": 0.002784799280236494,
      "grad_norm": 5.0220208168029785,
      "learning_rate": 9.999992346018105e-05,
      "loss": 2.9029520034790037,
      "memory(GiB)": 26.72,
      "step": 65,
      "token_acc": 0.3992932862190813,
      "train_speed(iter/s)": 1.402898
    },
    {
      "epoch": 0.0029990146094854547,
      "grad_norm": 4.8612518310546875,
      "learning_rate": 9.999991123192957e-05,
      "loss": 2.9941619873046874,
      "memory(GiB)": 26.72,
      "step": 70,
      "token_acc": 0.4043887147335423,
      "train_speed(iter/s)": 1.425915
    },
    {
      "epoch": 0.003213229938734416,
      "grad_norm": 4.054269313812256,
      "learning_rate": 9.99998980978828e-05,
      "loss": 2.784201812744141,
      "memory(GiB)": 26.72,
      "step": 75,
      "token_acc": 0.4263322884012539,
      "train_speed(iter/s)": 1.454161
    },
    {
      "epoch": 0.0034274452679833767,
      "grad_norm": 22.715557098388672,
      "learning_rate": 9.999988405804095e-05,
      "loss": 3.0202789306640625,
      "memory(GiB)": 26.72,
      "step": 80,
      "token_acc": 0.4314516129032258,
      "train_speed(iter/s)": 1.47951
    },
    {
      "epoch": 0.003641660597232338,
      "grad_norm": 4.0691986083984375,
      "learning_rate": 9.999986911240431e-05,
      "loss": 2.7469051361083983,
      "memory(GiB)": 26.72,
      "step": 85,
      "token_acc": 0.4430379746835443,
      "train_speed(iter/s)": 1.485103
    },
    {
      "epoch": 0.003855875926481299,
      "grad_norm": 5.723708629608154,
      "learning_rate": 9.999985326097314e-05,
      "loss": 2.6862131118774415,
      "memory(GiB)": 26.72,
      "step": 90,
      "token_acc": 0.4379310344827586,
      "train_speed(iter/s)": 1.502945
    },
    {
      "epoch": 0.00407009125573026,
      "grad_norm": 5.090478420257568,
      "learning_rate": 9.999983650374773e-05,
      "loss": 2.78324031829834,
      "memory(GiB)": 26.72,
      "step": 95,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.490174
    },
    {
      "epoch": 0.004284306584979221,
      "grad_norm": 5.822147369384766,
      "learning_rate": 9.999981884072838e-05,
      "loss": 2.932465934753418,
      "memory(GiB)": 26.72,
      "step": 100,
      "token_acc": 0.42120343839541546,
      "train_speed(iter/s)": 1.494382
    },
    {
      "epoch": 0.004498521914228182,
      "grad_norm": 6.790576457977295,
      "learning_rate": 9.999980027191539e-05,
      "loss": 2.8091917037963867,
      "memory(GiB)": 26.72,
      "step": 105,
      "token_acc": 0.43190661478599224,
      "train_speed(iter/s)": 1.494894
    },
    {
      "epoch": 0.004712737243477144,
      "grad_norm": 4.94995641708374,
      "learning_rate": 9.999978079730912e-05,
      "loss": 2.6159130096435548,
      "memory(GiB)": 26.72,
      "step": 110,
      "token_acc": 0.43911439114391143,
      "train_speed(iter/s)": 1.499744
    },
    {
      "epoch": 0.0049269525727261045,
      "grad_norm": 4.683931827545166,
      "learning_rate": 9.999976041690993e-05,
      "loss": 2.5537393569946287,
      "memory(GiB)": 26.72,
      "step": 115,
      "token_acc": 0.49823321554770317,
      "train_speed(iter/s)": 1.495688
    },
    {
      "epoch": 0.005141167901975065,
      "grad_norm": 4.683681488037109,
      "learning_rate": 9.999973913071817e-05,
      "loss": 2.5687942504882812,
      "memory(GiB)": 26.72,
      "step": 120,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.498772
    },
    {
      "epoch": 0.005355383231224026,
      "grad_norm": 6.400060653686523,
      "learning_rate": 9.999971693873423e-05,
      "loss": 2.461386489868164,
      "memory(GiB)": 26.72,
      "step": 125,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.500086
    },
    {
      "epoch": 0.005569598560472988,
      "grad_norm": 5.713675498962402,
      "learning_rate": 9.999969384095851e-05,
      "loss": 2.952980613708496,
      "memory(GiB)": 26.72,
      "step": 130,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.496834
    },
    {
      "epoch": 0.0057838138897219485,
      "grad_norm": 5.028921127319336,
      "learning_rate": 9.999966983739143e-05,
      "loss": 2.865002250671387,
      "memory(GiB)": 26.72,
      "step": 135,
      "token_acc": 0.4670846394984326,
      "train_speed(iter/s)": 1.508888
    },
    {
      "epoch": 0.005998029218970909,
      "grad_norm": 4.72570276260376,
      "learning_rate": 9.999964492803344e-05,
      "loss": 2.7058420181274414,
      "memory(GiB)": 26.72,
      "step": 140,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.515212
    },
    {
      "epoch": 0.006212244548219871,
      "grad_norm": 3.9119255542755127,
      "learning_rate": 9.999961911288497e-05,
      "loss": 2.6955331802368163,
      "memory(GiB)": 26.72,
      "step": 145,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.521761
    },
    {
      "epoch": 0.006426459877468832,
      "grad_norm": 4.968887805938721,
      "learning_rate": 9.99995923919465e-05,
      "loss": 2.752030944824219,
      "memory(GiB)": 26.72,
      "step": 150,
      "token_acc": 0.4365079365079365,
      "train_speed(iter/s)": 1.518681
    },
    {
      "epoch": 0.006640675206717793,
      "grad_norm": 4.824211597442627,
      "learning_rate": 9.99995647652185e-05,
      "loss": 3.1181013107299806,
      "memory(GiB)": 26.72,
      "step": 155,
      "token_acc": 0.3935860058309038,
      "train_speed(iter/s)": 1.513063
    },
    {
      "epoch": 0.006854890535966753,
      "grad_norm": 8.688904762268066,
      "learning_rate": 9.99995362327015e-05,
      "loss": 2.7499887466430666,
      "memory(GiB)": 26.72,
      "step": 160,
      "token_acc": 0.45733788395904434,
      "train_speed(iter/s)": 1.522206
    },
    {
      "epoch": 0.007069105865215715,
      "grad_norm": 4.412156105041504,
      "learning_rate": 9.999950679439598e-05,
      "loss": 3.1137237548828125,
      "memory(GiB)": 32.37,
      "step": 165,
      "token_acc": 0.3924050632911392,
      "train_speed(iter/s)": 1.511342
    },
    {
      "epoch": 0.007283321194464676,
      "grad_norm": 6.1439971923828125,
      "learning_rate": 9.99994764503025e-05,
      "loss": 2.9106130599975586,
      "memory(GiB)": 32.37,
      "step": 170,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.513431
    },
    {
      "epoch": 0.007497536523713637,
      "grad_norm": 5.294766426086426,
      "learning_rate": 9.99994452004216e-05,
      "loss": 2.8748470306396485,
      "memory(GiB)": 32.37,
      "step": 175,
      "token_acc": 0.45819397993311034,
      "train_speed(iter/s)": 1.51648
    },
    {
      "epoch": 0.007711751852962598,
      "grad_norm": 5.394874572753906,
      "learning_rate": 9.999941304475385e-05,
      "loss": 2.5859235763549804,
      "memory(GiB)": 32.37,
      "step": 180,
      "token_acc": 0.43670886075949367,
      "train_speed(iter/s)": 1.500775
    },
    {
      "epoch": 0.007925967182211559,
      "grad_norm": 5.008523464202881,
      "learning_rate": 9.999937998329982e-05,
      "loss": 2.9563148498535154,
      "memory(GiB)": 32.37,
      "step": 185,
      "token_acc": 0.3745928338762215,
      "train_speed(iter/s)": 1.500915
    },
    {
      "epoch": 0.00814018251146052,
      "grad_norm": 4.091294288635254,
      "learning_rate": 9.999934601606014e-05,
      "loss": 2.705361557006836,
      "memory(GiB)": 32.37,
      "step": 190,
      "token_acc": 0.4192546583850932,
      "train_speed(iter/s)": 1.504781
    },
    {
      "epoch": 0.00835439784070948,
      "grad_norm": 4.025635242462158,
      "learning_rate": 9.999931114303538e-05,
      "loss": 2.8227870941162108,
      "memory(GiB)": 32.37,
      "step": 195,
      "token_acc": 0.41975308641975306,
      "train_speed(iter/s)": 1.509104
    },
    {
      "epoch": 0.008568613169958442,
      "grad_norm": 4.693427562713623,
      "learning_rate": 9.99992753642262e-05,
      "loss": 2.8427942276000975,
      "memory(GiB)": 32.37,
      "step": 200,
      "token_acc": 0.44787644787644787,
      "train_speed(iter/s)": 1.512413
    },
    {
      "epoch": 0.008782828499207404,
      "grad_norm": 6.930469512939453,
      "learning_rate": 9.999923867963326e-05,
      "loss": 3.0147377014160157,
      "memory(GiB)": 32.37,
      "step": 205,
      "token_acc": 0.3939393939393939,
      "train_speed(iter/s)": 1.510234
    },
    {
      "epoch": 0.008997043828456364,
      "grad_norm": 7.443196773529053,
      "learning_rate": 9.999920108925719e-05,
      "loss": 2.4351734161376952,
      "memory(GiB)": 32.37,
      "step": 210,
      "token_acc": 0.5021459227467812,
      "train_speed(iter/s)": 1.510754
    },
    {
      "epoch": 0.009211259157705326,
      "grad_norm": 7.709830284118652,
      "learning_rate": 9.99991625930987e-05,
      "loss": 2.7562721252441404,
      "memory(GiB)": 32.37,
      "step": 215,
      "token_acc": 0.41603053435114506,
      "train_speed(iter/s)": 1.507848
    },
    {
      "epoch": 0.009425474486954287,
      "grad_norm": 4.83250093460083,
      "learning_rate": 9.999912319115848e-05,
      "loss": 2.6840383529663088,
      "memory(GiB)": 32.37,
      "step": 220,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.510214
    },
    {
      "epoch": 0.009639689816203247,
      "grad_norm": 5.140326499938965,
      "learning_rate": 9.999908288343722e-05,
      "loss": 2.8931657791137697,
      "memory(GiB)": 32.37,
      "step": 225,
      "token_acc": 0.40540540540540543,
      "train_speed(iter/s)": 1.518391
    },
    {
      "epoch": 0.009853905145452209,
      "grad_norm": 4.432214736938477,
      "learning_rate": 9.999904166993568e-05,
      "loss": 2.5513858795166016,
      "memory(GiB)": 32.37,
      "step": 230,
      "token_acc": 0.47435897435897434,
      "train_speed(iter/s)": 1.521624
    },
    {
      "epoch": 0.010068120474701169,
      "grad_norm": 5.439079761505127,
      "learning_rate": 9.999899955065461e-05,
      "loss": 2.964206314086914,
      "memory(GiB)": 32.37,
      "step": 235,
      "token_acc": 0.4262295081967213,
      "train_speed(iter/s)": 1.52316
    },
    {
      "epoch": 0.01028233580395013,
      "grad_norm": 9.50673770904541,
      "learning_rate": 9.999895652559475e-05,
      "loss": 2.587946128845215,
      "memory(GiB)": 32.37,
      "step": 240,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.522841
    },
    {
      "epoch": 0.010496551133199092,
      "grad_norm": 5.015991687774658,
      "learning_rate": 9.999891259475688e-05,
      "loss": 2.7738439559936525,
      "memory(GiB)": 32.37,
      "step": 245,
      "token_acc": 0.4020979020979021,
      "train_speed(iter/s)": 1.526058
    },
    {
      "epoch": 0.010710766462448052,
      "grad_norm": 5.400449752807617,
      "learning_rate": 9.999886775814182e-05,
      "loss": 2.7554553985595702,
      "memory(GiB)": 32.37,
      "step": 250,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.522284
    },
    {
      "epoch": 0.010924981791697014,
      "grad_norm": 4.275041103363037,
      "learning_rate": 9.999882201575036e-05,
      "loss": 2.767551803588867,
      "memory(GiB)": 32.37,
      "step": 255,
      "token_acc": 0.4542483660130719,
      "train_speed(iter/s)": 1.521842
    },
    {
      "epoch": 0.011139197120945975,
      "grad_norm": 4.915047645568848,
      "learning_rate": 9.999877536758334e-05,
      "loss": 2.9720163345336914,
      "memory(GiB)": 32.37,
      "step": 260,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.524295
    },
    {
      "epoch": 0.011353412450194935,
      "grad_norm": 5.237390041351318,
      "learning_rate": 9.99987278136416e-05,
      "loss": 2.8887733459472655,
      "memory(GiB)": 32.37,
      "step": 265,
      "token_acc": 0.4246575342465753,
      "train_speed(iter/s)": 1.519953
    },
    {
      "epoch": 0.011567627779443897,
      "grad_norm": 6.189242839813232,
      "learning_rate": 9.9998679353926e-05,
      "loss": 2.7354158401489257,
      "memory(GiB)": 32.37,
      "step": 270,
      "token_acc": 0.4826086956521739,
      "train_speed(iter/s)": 1.521764
    },
    {
      "epoch": 0.011781843108692859,
      "grad_norm": 4.483863353729248,
      "learning_rate": 9.999862998843743e-05,
      "loss": 2.5961429595947267,
      "memory(GiB)": 38.45,
      "step": 275,
      "token_acc": 0.44171779141104295,
      "train_speed(iter/s)": 1.514263
    },
    {
      "epoch": 0.011996058437941819,
      "grad_norm": 4.702669143676758,
      "learning_rate": 9.999857971717678e-05,
      "loss": 2.832667350769043,
      "memory(GiB)": 38.45,
      "step": 280,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.509821
    },
    {
      "epoch": 0.01221027376719078,
      "grad_norm": 4.098536014556885,
      "learning_rate": 9.999852854014495e-05,
      "loss": 2.7756668090820313,
      "memory(GiB)": 38.45,
      "step": 285,
      "token_acc": 0.42333333333333334,
      "train_speed(iter/s)": 1.511148
    },
    {
      "epoch": 0.012424489096439742,
      "grad_norm": 4.666366100311279,
      "learning_rate": 9.999847645734286e-05,
      "loss": 2.5693937301635743,
      "memory(GiB)": 38.45,
      "step": 290,
      "token_acc": 0.44649446494464945,
      "train_speed(iter/s)": 1.518436
    },
    {
      "epoch": 0.012638704425688702,
      "grad_norm": 3.2439024448394775,
      "learning_rate": 9.999842346877148e-05,
      "loss": 2.6344221115112303,
      "memory(GiB)": 38.45,
      "step": 295,
      "token_acc": 0.458955223880597,
      "train_speed(iter/s)": 1.524232
    },
    {
      "epoch": 0.012852919754937664,
      "grad_norm": 4.860950946807861,
      "learning_rate": 9.999836957443175e-05,
      "loss": 2.9313159942626954,
      "memory(GiB)": 38.45,
      "step": 300,
      "token_acc": 0.3890909090909091,
      "train_speed(iter/s)": 1.528104
    },
    {
      "epoch": 0.013067135084186625,
      "grad_norm": 4.235132694244385,
      "learning_rate": 9.999831477432466e-05,
      "loss": 2.4759666442871096,
      "memory(GiB)": 38.45,
      "step": 305,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.520383
    },
    {
      "epoch": 0.013281350413435585,
      "grad_norm": 5.354401111602783,
      "learning_rate": 9.99982590684512e-05,
      "loss": 2.9464181900024413,
      "memory(GiB)": 38.45,
      "step": 310,
      "token_acc": 0.39655172413793105,
      "train_speed(iter/s)": 1.520972
    },
    {
      "epoch": 0.013495565742684547,
      "grad_norm": 4.568800449371338,
      "learning_rate": 9.999820245681236e-05,
      "loss": 2.7896701812744142,
      "memory(GiB)": 38.45,
      "step": 315,
      "token_acc": 0.412568306010929,
      "train_speed(iter/s)": 1.515008
    },
    {
      "epoch": 0.013709781071933507,
      "grad_norm": 4.526827812194824,
      "learning_rate": 9.999814493940918e-05,
      "loss": 2.6762229919433596,
      "memory(GiB)": 38.45,
      "step": 320,
      "token_acc": 0.4098939929328622,
      "train_speed(iter/s)": 1.516438
    },
    {
      "epoch": 0.013923996401182468,
      "grad_norm": 4.059617042541504,
      "learning_rate": 9.999808651624271e-05,
      "loss": 2.3909589767456056,
      "memory(GiB)": 38.45,
      "step": 325,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.51252
    },
    {
      "epoch": 0.01413821173043143,
      "grad_norm": 8.303412437438965,
      "learning_rate": 9.9998027187314e-05,
      "loss": 2.690586280822754,
      "memory(GiB)": 38.45,
      "step": 330,
      "token_acc": 0.47345132743362833,
      "train_speed(iter/s)": 1.515164
    },
    {
      "epoch": 0.01435242705968039,
      "grad_norm": 3.9448540210723877,
      "learning_rate": 9.999796695262413e-05,
      "loss": 2.6628677368164064,
      "memory(GiB)": 38.45,
      "step": 335,
      "token_acc": 0.43209876543209874,
      "train_speed(iter/s)": 1.514875
    },
    {
      "epoch": 0.014566642388929352,
      "grad_norm": 4.270360469818115,
      "learning_rate": 9.999790581217418e-05,
      "loss": 2.5807426452636717,
      "memory(GiB)": 38.45,
      "step": 340,
      "token_acc": 0.460431654676259,
      "train_speed(iter/s)": 1.512271
    },
    {
      "epoch": 0.014780857718178313,
      "grad_norm": 3.328791379928589,
      "learning_rate": 9.999784376596526e-05,
      "loss": 2.8379131317138673,
      "memory(GiB)": 38.45,
      "step": 345,
      "token_acc": 0.40963855421686746,
      "train_speed(iter/s)": 1.516468
    },
    {
      "epoch": 0.014995073047427273,
      "grad_norm": 4.7893385887146,
      "learning_rate": 9.999778081399851e-05,
      "loss": 2.5286256790161135,
      "memory(GiB)": 38.45,
      "step": 350,
      "token_acc": 0.4641509433962264,
      "train_speed(iter/s)": 1.517933
    },
    {
      "epoch": 0.015209288376676235,
      "grad_norm": 6.374311447143555,
      "learning_rate": 9.999771695627505e-05,
      "loss": 2.6312910079956056,
      "memory(GiB)": 38.45,
      "step": 355,
      "token_acc": 0.44404332129963897,
      "train_speed(iter/s)": 1.521818
    },
    {
      "epoch": 0.015423503705925197,
      "grad_norm": 4.904713153839111,
      "learning_rate": 9.999765219279605e-05,
      "loss": 2.7969123840332033,
      "memory(GiB)": 38.45,
      "step": 360,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.526488
    },
    {
      "epoch": 0.015637719035174157,
      "grad_norm": 3.51206111907959,
      "learning_rate": 9.999758652356268e-05,
      "loss": 2.7190811157226564,
      "memory(GiB)": 38.45,
      "step": 365,
      "token_acc": 0.43205574912891986,
      "train_speed(iter/s)": 1.529405
    },
    {
      "epoch": 0.015851934364423118,
      "grad_norm": 4.942959785461426,
      "learning_rate": 9.999751994857614e-05,
      "loss": 2.495261573791504,
      "memory(GiB)": 38.45,
      "step": 370,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.530248
    },
    {
      "epoch": 0.01606614969367208,
      "grad_norm": 6.035073280334473,
      "learning_rate": 9.999745246783761e-05,
      "loss": 2.8744131088256837,
      "memory(GiB)": 38.45,
      "step": 375,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.52934
    },
    {
      "epoch": 0.01628036502292104,
      "grad_norm": 4.9291462898254395,
      "learning_rate": 9.999738408134832e-05,
      "loss": 2.8079748153686523,
      "memory(GiB)": 38.45,
      "step": 380,
      "token_acc": 0.4146341463414634,
      "train_speed(iter/s)": 1.528471
    },
    {
      "epoch": 0.01649458035217,
      "grad_norm": 3.258988380432129,
      "learning_rate": 9.999731478910952e-05,
      "loss": 2.445315933227539,
      "memory(GiB)": 38.45,
      "step": 385,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.529275
    },
    {
      "epoch": 0.01670879568141896,
      "grad_norm": 12.798467636108398,
      "learning_rate": 9.999724459112247e-05,
      "loss": 2.7420465469360353,
      "memory(GiB)": 38.45,
      "step": 390,
      "token_acc": 0.43162393162393164,
      "train_speed(iter/s)": 1.524791
    },
    {
      "epoch": 0.016923011010667923,
      "grad_norm": 7.007190704345703,
      "learning_rate": 9.999717348738843e-05,
      "loss": 2.942500114440918,
      "memory(GiB)": 38.45,
      "step": 395,
      "token_acc": 0.40942028985507245,
      "train_speed(iter/s)": 1.527629
    },
    {
      "epoch": 0.017137226339916885,
      "grad_norm": 6.345561981201172,
      "learning_rate": 9.999710147790869e-05,
      "loss": 2.7135833740234374,
      "memory(GiB)": 38.45,
      "step": 400,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.529885
    },
    {
      "epoch": 0.017351441669165846,
      "grad_norm": 5.034018039703369,
      "learning_rate": 9.999702856268456e-05,
      "loss": 2.9034446716308593,
      "memory(GiB)": 38.45,
      "step": 405,
      "token_acc": 0.42435424354243545,
      "train_speed(iter/s)": 1.526357
    },
    {
      "epoch": 0.017565656998414808,
      "grad_norm": 4.46183967590332,
      "learning_rate": 9.999695474171735e-05,
      "loss": 2.5323673248291017,
      "memory(GiB)": 38.45,
      "step": 410,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.524539
    },
    {
      "epoch": 0.017779872327663766,
      "grad_norm": 4.101302623748779,
      "learning_rate": 9.99968800150084e-05,
      "loss": 2.836459732055664,
      "memory(GiB)": 38.45,
      "step": 415,
      "token_acc": 0.4169611307420495,
      "train_speed(iter/s)": 1.525782
    },
    {
      "epoch": 0.017994087656912728,
      "grad_norm": 5.307928085327148,
      "learning_rate": 9.999680438255908e-05,
      "loss": 2.5662986755371096,
      "memory(GiB)": 38.45,
      "step": 420,
      "token_acc": 0.4855072463768116,
      "train_speed(iter/s)": 1.526494
    },
    {
      "epoch": 0.01820830298616169,
      "grad_norm": 4.104459285736084,
      "learning_rate": 9.999672784437075e-05,
      "loss": 2.9653785705566404,
      "memory(GiB)": 38.45,
      "step": 425,
      "token_acc": 0.42379182156133827,
      "train_speed(iter/s)": 1.523517
    },
    {
      "epoch": 0.01842251831541065,
      "grad_norm": 4.8199462890625,
      "learning_rate": 9.999665040044477e-05,
      "loss": 3.1615955352783205,
      "memory(GiB)": 38.45,
      "step": 430,
      "token_acc": 0.41312741312741313,
      "train_speed(iter/s)": 1.520442
    },
    {
      "epoch": 0.018636733644659613,
      "grad_norm": 5.425066947937012,
      "learning_rate": 9.99965720507826e-05,
      "loss": 2.6454307556152346,
      "memory(GiB)": 38.45,
      "step": 435,
      "token_acc": 0.4384057971014493,
      "train_speed(iter/s)": 1.519672
    },
    {
      "epoch": 0.018850948973908575,
      "grad_norm": 7.106417179107666,
      "learning_rate": 9.999649279538562e-05,
      "loss": 2.5446660995483397,
      "memory(GiB)": 38.45,
      "step": 440,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.519192
    },
    {
      "epoch": 0.019065164303157533,
      "grad_norm": 3.9185760021209717,
      "learning_rate": 9.999641263425524e-05,
      "loss": 2.862480354309082,
      "memory(GiB)": 38.45,
      "step": 445,
      "token_acc": 0.43026706231454004,
      "train_speed(iter/s)": 1.512258
    },
    {
      "epoch": 0.019279379632406494,
      "grad_norm": 4.255904197692871,
      "learning_rate": 9.999633156739298e-05,
      "loss": 2.7113834381103517,
      "memory(GiB)": 38.45,
      "step": 450,
      "token_acc": 0.46360153256704983,
      "train_speed(iter/s)": 1.512535
    },
    {
      "epoch": 0.019493594961655456,
      "grad_norm": 3.6170578002929688,
      "learning_rate": 9.999624959480025e-05,
      "loss": 2.712862014770508,
      "memory(GiB)": 38.45,
      "step": 455,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.51161
    },
    {
      "epoch": 0.019707810290904418,
      "grad_norm": 4.32484245300293,
      "learning_rate": 9.999616671647858e-05,
      "loss": 2.6673866271972657,
      "memory(GiB)": 38.45,
      "step": 460,
      "token_acc": 0.44398340248962653,
      "train_speed(iter/s)": 1.511981
    },
    {
      "epoch": 0.01992202562015338,
      "grad_norm": 4.658557891845703,
      "learning_rate": 9.999608293242943e-05,
      "loss": 2.3850494384765626,
      "memory(GiB)": 38.45,
      "step": 465,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.51372
    },
    {
      "epoch": 0.020136240949402338,
      "grad_norm": 5.076476573944092,
      "learning_rate": 9.999599824265436e-05,
      "loss": 3.014224624633789,
      "memory(GiB)": 38.45,
      "step": 470,
      "token_acc": 0.4306569343065693,
      "train_speed(iter/s)": 1.515909
    },
    {
      "epoch": 0.0203504562786513,
      "grad_norm": 3.9983785152435303,
      "learning_rate": 9.999591264715485e-05,
      "loss": 2.65859375,
      "memory(GiB)": 38.45,
      "step": 475,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.512097
    },
    {
      "epoch": 0.02056467160790026,
      "grad_norm": 4.255746841430664,
      "learning_rate": 9.99958261459325e-05,
      "loss": 2.274767303466797,
      "memory(GiB)": 38.45,
      "step": 480,
      "token_acc": 0.5085470085470085,
      "train_speed(iter/s)": 1.510742
    },
    {
      "epoch": 0.020778886937149223,
      "grad_norm": 3.8533356189727783,
      "learning_rate": 9.999573873898887e-05,
      "loss": 2.840310287475586,
      "memory(GiB)": 38.45,
      "step": 485,
      "token_acc": 0.44635193133047213,
      "train_speed(iter/s)": 1.514164
    },
    {
      "epoch": 0.020993102266398184,
      "grad_norm": 3.6588633060455322,
      "learning_rate": 9.999565042632553e-05,
      "loss": 2.7850326538085937,
      "memory(GiB)": 38.45,
      "step": 490,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.514103
    },
    {
      "epoch": 0.021207317595647146,
      "grad_norm": 5.175492286682129,
      "learning_rate": 9.999556120794407e-05,
      "loss": 2.610948181152344,
      "memory(GiB)": 38.45,
      "step": 495,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.511599
    },
    {
      "epoch": 0.021421532924896104,
      "grad_norm": 4.59173059463501,
      "learning_rate": 9.999547108384612e-05,
      "loss": 2.926106262207031,
      "memory(GiB)": 38.45,
      "step": 500,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.510686
    },
    {
      "epoch": 0.021421532924896104,
      "eval_loss": 2.28562331199646,
      "eval_runtime": 14.014,
      "eval_samples_per_second": 7.136,
      "eval_steps_per_second": 7.136,
      "eval_token_acc": 0.46879150066401065,
      "step": 500
    },
    {
      "epoch": 0.021635748254145066,
      "grad_norm": 10.294635772705078,
      "learning_rate": 9.999538005403332e-05,
      "loss": 2.579853820800781,
      "memory(GiB)": 38.45,
      "step": 505,
      "token_acc": 0.46791707798617965,
      "train_speed(iter/s)": 1.447716
    },
    {
      "epoch": 0.021849963583394028,
      "grad_norm": 4.110366344451904,
      "learning_rate": 9.999528811850732e-05,
      "loss": 2.8328296661376955,
      "memory(GiB)": 38.45,
      "step": 510,
      "token_acc": 0.4592833876221498,
      "train_speed(iter/s)": 1.45066
    },
    {
      "epoch": 0.02206417891264299,
      "grad_norm": 3.967574119567871,
      "learning_rate": 9.999519527726974e-05,
      "loss": 2.645603370666504,
      "memory(GiB)": 38.45,
      "step": 515,
      "token_acc": 0.45126353790613716,
      "train_speed(iter/s)": 1.4524
    },
    {
      "epoch": 0.02227839424189195,
      "grad_norm": 4.127718925476074,
      "learning_rate": 9.999510153032234e-05,
      "loss": 2.7630361557006835,
      "memory(GiB)": 38.45,
      "step": 520,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.449849
    },
    {
      "epoch": 0.022492609571140913,
      "grad_norm": 6.0096845626831055,
      "learning_rate": 9.999500687766675e-05,
      "loss": 2.773067665100098,
      "memory(GiB)": 38.45,
      "step": 525,
      "token_acc": 0.4152823920265781,
      "train_speed(iter/s)": 1.452213
    },
    {
      "epoch": 0.02270682490038987,
      "grad_norm": 3.1059420108795166,
      "learning_rate": 9.999491131930472e-05,
      "loss": 2.55417537689209,
      "memory(GiB)": 38.45,
      "step": 530,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.454439
    },
    {
      "epoch": 0.022921040229638832,
      "grad_norm": 4.492700576782227,
      "learning_rate": 9.999481485523798e-05,
      "loss": 2.4823116302490233,
      "memory(GiB)": 38.45,
      "step": 535,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.454122
    },
    {
      "epoch": 0.023135255558887794,
      "grad_norm": 3.9863710403442383,
      "learning_rate": 9.999471748546826e-05,
      "loss": 2.539709281921387,
      "memory(GiB)": 38.45,
      "step": 540,
      "token_acc": 0.48091603053435117,
      "train_speed(iter/s)": 1.453699
    },
    {
      "epoch": 0.023349470888136756,
      "grad_norm": 13.657026290893555,
      "learning_rate": 9.999461920999733e-05,
      "loss": 2.495493698120117,
      "memory(GiB)": 38.45,
      "step": 545,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.455843
    },
    {
      "epoch": 0.023563686217385717,
      "grad_norm": 5.109379291534424,
      "learning_rate": 9.999452002882699e-05,
      "loss": 2.5587100982666016,
      "memory(GiB)": 38.45,
      "step": 550,
      "token_acc": 0.4467455621301775,
      "train_speed(iter/s)": 1.455301
    },
    {
      "epoch": 0.023777901546634676,
      "grad_norm": 9.8173828125,
      "learning_rate": 9.999441994195902e-05,
      "loss": 2.557139778137207,
      "memory(GiB)": 38.45,
      "step": 555,
      "token_acc": 0.4781021897810219,
      "train_speed(iter/s)": 1.452023
    },
    {
      "epoch": 0.023992116875883637,
      "grad_norm": 3.308816432952881,
      "learning_rate": 9.999431894939522e-05,
      "loss": 2.7724546432495116,
      "memory(GiB)": 38.45,
      "step": 560,
      "token_acc": 0.44404332129963897,
      "train_speed(iter/s)": 1.454533
    },
    {
      "epoch": 0.0242063322051326,
      "grad_norm": 8.385897636413574,
      "learning_rate": 9.999421705113743e-05,
      "loss": 2.816848564147949,
      "memory(GiB)": 38.45,
      "step": 565,
      "token_acc": 0.43034055727554177,
      "train_speed(iter/s)": 1.457619
    },
    {
      "epoch": 0.02442054753438156,
      "grad_norm": 4.684794902801514,
      "learning_rate": 9.999411424718752e-05,
      "loss": 2.410172462463379,
      "memory(GiB)": 38.45,
      "step": 570,
      "token_acc": 0.4807121661721068,
      "train_speed(iter/s)": 1.459184
    },
    {
      "epoch": 0.024634762863630522,
      "grad_norm": 5.972314357757568,
      "learning_rate": 9.999401053754731e-05,
      "loss": 2.4901058197021486,
      "memory(GiB)": 38.45,
      "step": 575,
      "token_acc": 0.4607142857142857,
      "train_speed(iter/s)": 1.459399
    },
    {
      "epoch": 0.024848978192879484,
      "grad_norm": 4.343634128570557,
      "learning_rate": 9.999390592221871e-05,
      "loss": 2.8735374450683593,
      "memory(GiB)": 38.45,
      "step": 580,
      "token_acc": 0.45132743362831856,
      "train_speed(iter/s)": 1.460147
    },
    {
      "epoch": 0.025063193522128442,
      "grad_norm": 4.469681739807129,
      "learning_rate": 9.999380040120362e-05,
      "loss": 2.8024024963378906,
      "memory(GiB)": 38.45,
      "step": 585,
      "token_acc": 0.4223433242506812,
      "train_speed(iter/s)": 1.462626
    },
    {
      "epoch": 0.025277408851377404,
      "grad_norm": 4.772650718688965,
      "learning_rate": 9.999369397450392e-05,
      "loss": 2.6395135879516602,
      "memory(GiB)": 38.45,
      "step": 590,
      "token_acc": 0.43853820598006643,
      "train_speed(iter/s)": 1.465825
    },
    {
      "epoch": 0.025491624180626365,
      "grad_norm": 4.2591633796691895,
      "learning_rate": 9.999358664212158e-05,
      "loss": 2.9030269622802733,
      "memory(GiB)": 38.45,
      "step": 595,
      "token_acc": 0.44921875,
      "train_speed(iter/s)": 1.467251
    },
    {
      "epoch": 0.025705839509875327,
      "grad_norm": 4.614182949066162,
      "learning_rate": 9.999347840405848e-05,
      "loss": 2.5291936874389647,
      "memory(GiB)": 38.45,
      "step": 600,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.466106
    },
    {
      "epoch": 0.02592005483912429,
      "grad_norm": 3.163677215576172,
      "learning_rate": 9.999336926031666e-05,
      "loss": 2.676190948486328,
      "memory(GiB)": 38.45,
      "step": 605,
      "token_acc": 0.4517133956386293,
      "train_speed(iter/s)": 1.465845
    },
    {
      "epoch": 0.02613427016837325,
      "grad_norm": 4.9240498542785645,
      "learning_rate": 9.999325921089804e-05,
      "loss": 2.810549736022949,
      "memory(GiB)": 38.45,
      "step": 610,
      "token_acc": 0.4483985765124555,
      "train_speed(iter/s)": 1.46277
    },
    {
      "epoch": 0.02634848549762221,
      "grad_norm": 4.851043701171875,
      "learning_rate": 9.999314825580464e-05,
      "loss": 3.121914100646973,
      "memory(GiB)": 38.45,
      "step": 615,
      "token_acc": 0.36271186440677966,
      "train_speed(iter/s)": 1.464436
    },
    {
      "epoch": 0.02656270082687117,
      "grad_norm": 3.2191321849823,
      "learning_rate": 9.999303639503845e-05,
      "loss": 2.8040241241455077,
      "memory(GiB)": 38.45,
      "step": 620,
      "token_acc": 0.4169381107491857,
      "train_speed(iter/s)": 1.467151
    },
    {
      "epoch": 0.026776916156120132,
      "grad_norm": 6.146385192871094,
      "learning_rate": 9.999292362860151e-05,
      "loss": 2.6678306579589846,
      "memory(GiB)": 38.45,
      "step": 625,
      "token_acc": 0.4452296819787986,
      "train_speed(iter/s)": 1.465956
    },
    {
      "epoch": 0.026991131485369094,
      "grad_norm": 4.547609329223633,
      "learning_rate": 9.999280995649588e-05,
      "loss": 2.668691635131836,
      "memory(GiB)": 38.45,
      "step": 630,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.466443
    },
    {
      "epoch": 0.027205346814618055,
      "grad_norm": 3.406956195831299,
      "learning_rate": 9.999269537872357e-05,
      "loss": 2.705013656616211,
      "memory(GiB)": 38.45,
      "step": 635,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.469634
    },
    {
      "epoch": 0.027419562143867014,
      "grad_norm": 4.823546409606934,
      "learning_rate": 9.99925798952867e-05,
      "loss": 2.1641578674316406,
      "memory(GiB)": 38.45,
      "step": 640,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.469394
    },
    {
      "epoch": 0.027633777473115975,
      "grad_norm": 6.831277847290039,
      "learning_rate": 9.999246350618734e-05,
      "loss": 2.6867881774902345,
      "memory(GiB)": 38.45,
      "step": 645,
      "token_acc": 0.45394736842105265,
      "train_speed(iter/s)": 1.470678
    },
    {
      "epoch": 0.027847992802364937,
      "grad_norm": 4.051003932952881,
      "learning_rate": 9.999234621142764e-05,
      "loss": 2.5303443908691405,
      "memory(GiB)": 38.45,
      "step": 650,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.472711
    },
    {
      "epoch": 0.0280622081316139,
      "grad_norm": 5.49873161315918,
      "learning_rate": 9.999222801100966e-05,
      "loss": 2.797929573059082,
      "memory(GiB)": 38.45,
      "step": 655,
      "token_acc": 0.41754385964912283,
      "train_speed(iter/s)": 1.471859
    },
    {
      "epoch": 0.02827642346086286,
      "grad_norm": 4.199821949005127,
      "learning_rate": 9.999210890493558e-05,
      "loss": 2.677899932861328,
      "memory(GiB)": 38.45,
      "step": 660,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.47234
    },
    {
      "epoch": 0.028490638790111822,
      "grad_norm": 6.619091987609863,
      "learning_rate": 9.999198889320755e-05,
      "loss": 2.7923709869384767,
      "memory(GiB)": 38.45,
      "step": 665,
      "token_acc": 0.40483383685800606,
      "train_speed(iter/s)": 1.474378
    },
    {
      "epoch": 0.02870485411936078,
      "grad_norm": 2.97367525100708,
      "learning_rate": 9.999186797582775e-05,
      "loss": 2.836111068725586,
      "memory(GiB)": 38.45,
      "step": 670,
      "token_acc": 0.43260188087774293,
      "train_speed(iter/s)": 1.472144
    },
    {
      "epoch": 0.028919069448609742,
      "grad_norm": 3.7363877296447754,
      "learning_rate": 9.999174615279836e-05,
      "loss": 2.5940717697143554,
      "memory(GiB)": 38.45,
      "step": 675,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.470481
    },
    {
      "epoch": 0.029133284777858703,
      "grad_norm": 4.364439964294434,
      "learning_rate": 9.999162342412161e-05,
      "loss": 2.753948783874512,
      "memory(GiB)": 38.45,
      "step": 680,
      "token_acc": 0.4229390681003584,
      "train_speed(iter/s)": 1.470741
    },
    {
      "epoch": 0.029347500107107665,
      "grad_norm": 3.7315828800201416,
      "learning_rate": 9.999149978979969e-05,
      "loss": 2.5765497207641603,
      "memory(GiB)": 38.45,
      "step": 685,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.47357
    },
    {
      "epoch": 0.029561715436356627,
      "grad_norm": 5.6918487548828125,
      "learning_rate": 9.999137524983485e-05,
      "loss": 2.78399543762207,
      "memory(GiB)": 38.45,
      "step": 690,
      "token_acc": 0.45357142857142857,
      "train_speed(iter/s)": 1.476662
    },
    {
      "epoch": 0.02977593076560559,
      "grad_norm": 2.9097728729248047,
      "learning_rate": 9.999124980422936e-05,
      "loss": 2.7044647216796873,
      "memory(GiB)": 38.45,
      "step": 695,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.47552
    },
    {
      "epoch": 0.029990146094854547,
      "grad_norm": 4.9471306800842285,
      "learning_rate": 9.999112345298549e-05,
      "loss": 2.4053537368774416,
      "memory(GiB)": 38.45,
      "step": 700,
      "token_acc": 0.4664179104477612,
      "train_speed(iter/s)": 1.472989
    },
    {
      "epoch": 0.03020436142410351,
      "grad_norm": 4.085890293121338,
      "learning_rate": 9.999099619610552e-05,
      "loss": 2.660585403442383,
      "memory(GiB)": 38.45,
      "step": 705,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.474779
    },
    {
      "epoch": 0.03041857675335247,
      "grad_norm": 6.359280109405518,
      "learning_rate": 9.999086803359174e-05,
      "loss": 2.5837047576904295,
      "memory(GiB)": 38.45,
      "step": 710,
      "token_acc": 0.48014440433212996,
      "train_speed(iter/s)": 1.476224
    },
    {
      "epoch": 0.03063279208260143,
      "grad_norm": 5.505381107330322,
      "learning_rate": 9.99907389654465e-05,
      "loss": 2.4027435302734377,
      "memory(GiB)": 38.45,
      "step": 715,
      "token_acc": 0.4811715481171548,
      "train_speed(iter/s)": 1.478193
    },
    {
      "epoch": 0.030847007411850393,
      "grad_norm": 3.7724270820617676,
      "learning_rate": 9.999060899167214e-05,
      "loss": 2.5833389282226564,
      "memory(GiB)": 38.45,
      "step": 720,
      "token_acc": 0.48520710059171596,
      "train_speed(iter/s)": 1.480148
    },
    {
      "epoch": 0.03106122274109935,
      "grad_norm": 4.620935916900635,
      "learning_rate": 9.9990478112271e-05,
      "loss": 2.739794921875,
      "memory(GiB)": 38.45,
      "step": 725,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.47927
    },
    {
      "epoch": 0.03127543807034831,
      "grad_norm": 3.3408830165863037,
      "learning_rate": 9.999034632724546e-05,
      "loss": 2.862911605834961,
      "memory(GiB)": 38.45,
      "step": 730,
      "token_acc": 0.4508670520231214,
      "train_speed(iter/s)": 1.477498
    },
    {
      "epoch": 0.03148965339959728,
      "grad_norm": 6.213047027587891,
      "learning_rate": 9.999021363659788e-05,
      "loss": 2.7660221099853515,
      "memory(GiB)": 38.45,
      "step": 735,
      "token_acc": 0.44074074074074077,
      "train_speed(iter/s)": 1.478877
    },
    {
      "epoch": 0.031703868728846236,
      "grad_norm": 7.579422473907471,
      "learning_rate": 9.99900800403307e-05,
      "loss": 2.365354537963867,
      "memory(GiB)": 38.45,
      "step": 740,
      "token_acc": 0.48134328358208955,
      "train_speed(iter/s)": 1.47808
    },
    {
      "epoch": 0.031918084058095195,
      "grad_norm": 5.464479446411133,
      "learning_rate": 9.998994553844632e-05,
      "loss": 2.5875072479248047,
      "memory(GiB)": 38.45,
      "step": 745,
      "token_acc": 0.49407114624505927,
      "train_speed(iter/s)": 1.478492
    },
    {
      "epoch": 0.03213229938734416,
      "grad_norm": 2.7492892742156982,
      "learning_rate": 9.99898101309472e-05,
      "loss": 3.0725284576416017,
      "memory(GiB)": 38.45,
      "step": 750,
      "token_acc": 0.3860182370820669,
      "train_speed(iter/s)": 1.479731
    },
    {
      "epoch": 0.03234651471659312,
      "grad_norm": 4.940596103668213,
      "learning_rate": 9.998967381783575e-05,
      "loss": 2.602863311767578,
      "memory(GiB)": 38.45,
      "step": 755,
      "token_acc": 0.4409722222222222,
      "train_speed(iter/s)": 1.480387
    },
    {
      "epoch": 0.03256073004584208,
      "grad_norm": 4.953486442565918,
      "learning_rate": 9.998953659911447e-05,
      "loss": 2.981593132019043,
      "memory(GiB)": 38.45,
      "step": 760,
      "token_acc": 0.3782051282051282,
      "train_speed(iter/s)": 1.47926
    },
    {
      "epoch": 0.03277494537509104,
      "grad_norm": 3.4166150093078613,
      "learning_rate": 9.998939847478585e-05,
      "loss": 2.607655143737793,
      "memory(GiB)": 38.45,
      "step": 765,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.47998
    },
    {
      "epoch": 0.03298916070434,
      "grad_norm": 3.688446283340454,
      "learning_rate": 9.998925944485238e-05,
      "loss": 2.6573068618774416,
      "memory(GiB)": 38.45,
      "step": 770,
      "token_acc": 0.4421768707482993,
      "train_speed(iter/s)": 1.481473
    },
    {
      "epoch": 0.033203376033588965,
      "grad_norm": 3.088634490966797,
      "learning_rate": 9.998911950931658e-05,
      "loss": 2.8627756118774412,
      "memory(GiB)": 38.45,
      "step": 775,
      "token_acc": 0.4155844155844156,
      "train_speed(iter/s)": 1.482194
    },
    {
      "epoch": 0.03341759136283792,
      "grad_norm": 4.363468170166016,
      "learning_rate": 9.998897866818099e-05,
      "loss": 2.7546947479248045,
      "memory(GiB)": 38.45,
      "step": 780,
      "token_acc": 0.41403508771929826,
      "train_speed(iter/s)": 1.483031
    },
    {
      "epoch": 0.03363180669208689,
      "grad_norm": 5.336426258087158,
      "learning_rate": 9.998883692144817e-05,
      "loss": 2.5782140731811523,
      "memory(GiB)": 38.45,
      "step": 785,
      "token_acc": 0.496875,
      "train_speed(iter/s)": 1.484279
    },
    {
      "epoch": 0.033846022021335846,
      "grad_norm": 3.5997202396392822,
      "learning_rate": 9.998869426912065e-05,
      "loss": 2.6509056091308594,
      "memory(GiB)": 38.45,
      "step": 790,
      "token_acc": 0.44755244755244755,
      "train_speed(iter/s)": 1.484196
    },
    {
      "epoch": 0.034060237350584804,
      "grad_norm": 3.448155403137207,
      "learning_rate": 9.998855071120104e-05,
      "loss": 2.881861114501953,
      "memory(GiB)": 38.45,
      "step": 795,
      "token_acc": 0.4403409090909091,
      "train_speed(iter/s)": 1.485208
    },
    {
      "epoch": 0.03427445267983377,
      "grad_norm": 3.7866625785827637,
      "learning_rate": 9.998840624769196e-05,
      "loss": 2.8026958465576173,
      "memory(GiB)": 38.45,
      "step": 800,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.486111
    },
    {
      "epoch": 0.03448866800908273,
      "grad_norm": 3.824439525604248,
      "learning_rate": 9.9988260878596e-05,
      "loss": 2.645696449279785,
      "memory(GiB)": 38.45,
      "step": 805,
      "token_acc": 0.44370860927152317,
      "train_speed(iter/s)": 1.488295
    },
    {
      "epoch": 0.03470288333833169,
      "grad_norm": 4.691617965698242,
      "learning_rate": 9.99881146039158e-05,
      "loss": 2.7255084991455076,
      "memory(GiB)": 38.45,
      "step": 810,
      "token_acc": 0.42323651452282157,
      "train_speed(iter/s)": 1.48988
    },
    {
      "epoch": 0.03491709866758065,
      "grad_norm": 3.136521100997925,
      "learning_rate": 9.998796742365402e-05,
      "loss": 2.609482192993164,
      "memory(GiB)": 38.45,
      "step": 815,
      "token_acc": 0.4309859154929577,
      "train_speed(iter/s)": 1.489195
    },
    {
      "epoch": 0.035131313996829616,
      "grad_norm": 3.6892218589782715,
      "learning_rate": 9.99878193378133e-05,
      "loss": 2.4301197052001955,
      "memory(GiB)": 38.45,
      "step": 820,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.491319
    },
    {
      "epoch": 0.035345529326078574,
      "grad_norm": 3.0272536277770996,
      "learning_rate": 9.998767034639636e-05,
      "loss": 2.8461444854736326,
      "memory(GiB)": 38.45,
      "step": 825,
      "token_acc": 0.42902208201892744,
      "train_speed(iter/s)": 1.492437
    },
    {
      "epoch": 0.03555974465532753,
      "grad_norm": 3.5021369457244873,
      "learning_rate": 9.998752044940587e-05,
      "loss": 2.7193660736083984,
      "memory(GiB)": 38.45,
      "step": 830,
      "token_acc": 0.4416403785488959,
      "train_speed(iter/s)": 1.493741
    },
    {
      "epoch": 0.0357739599845765,
      "grad_norm": 3.633901596069336,
      "learning_rate": 9.998736964684454e-05,
      "loss": 2.5642627716064452,
      "memory(GiB)": 38.45,
      "step": 835,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.493056
    },
    {
      "epoch": 0.035988175313825456,
      "grad_norm": 4.662364482879639,
      "learning_rate": 9.998721793871513e-05,
      "loss": 2.5761337280273438,
      "memory(GiB)": 38.45,
      "step": 840,
      "token_acc": 0.45054945054945056,
      "train_speed(iter/s)": 1.494073
    },
    {
      "epoch": 0.03620239064307442,
      "grad_norm": 5.195523262023926,
      "learning_rate": 9.998706532502038e-05,
      "loss": 2.638579177856445,
      "memory(GiB)": 38.45,
      "step": 845,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.492949
    },
    {
      "epoch": 0.03641660597232338,
      "grad_norm": 4.824570655822754,
      "learning_rate": 9.998691180576306e-05,
      "loss": 2.672596549987793,
      "memory(GiB)": 38.45,
      "step": 850,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.492738
    },
    {
      "epoch": 0.03663082130157234,
      "grad_norm": 2.6324217319488525,
      "learning_rate": 9.998675738094591e-05,
      "loss": 2.4606231689453124,
      "memory(GiB)": 38.45,
      "step": 855,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.492514
    },
    {
      "epoch": 0.0368450366308213,
      "grad_norm": 6.1279191970825195,
      "learning_rate": 9.998660205057179e-05,
      "loss": 2.696563148498535,
      "memory(GiB)": 38.45,
      "step": 860,
      "token_acc": 0.47435897435897434,
      "train_speed(iter/s)": 1.494138
    },
    {
      "epoch": 0.03705925196007026,
      "grad_norm": 3.731215238571167,
      "learning_rate": 9.998644581464348e-05,
      "loss": 2.98821964263916,
      "memory(GiB)": 38.45,
      "step": 865,
      "token_acc": 0.429042904290429,
      "train_speed(iter/s)": 1.494777
    },
    {
      "epoch": 0.037273467289319226,
      "grad_norm": 4.00960111618042,
      "learning_rate": 9.998628867316378e-05,
      "loss": 2.7611104965209963,
      "memory(GiB)": 38.45,
      "step": 870,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.493999
    },
    {
      "epoch": 0.037487682618568184,
      "grad_norm": 3.2089624404907227,
      "learning_rate": 9.99861306261356e-05,
      "loss": 2.7675397872924803,
      "memory(GiB)": 38.45,
      "step": 875,
      "token_acc": 0.4125,
      "train_speed(iter/s)": 1.494044
    },
    {
      "epoch": 0.03770189794781715,
      "grad_norm": 5.143516540527344,
      "learning_rate": 9.998597167356176e-05,
      "loss": 2.595576858520508,
      "memory(GiB)": 38.45,
      "step": 880,
      "token_acc": 0.459546925566343,
      "train_speed(iter/s)": 1.494389
    },
    {
      "epoch": 0.03791611327706611,
      "grad_norm": 4.214605808258057,
      "learning_rate": 9.998581181544516e-05,
      "loss": 2.522385025024414,
      "memory(GiB)": 38.45,
      "step": 885,
      "token_acc": 0.4668769716088328,
      "train_speed(iter/s)": 1.494258
    },
    {
      "epoch": 0.038130328606315066,
      "grad_norm": 3.4676084518432617,
      "learning_rate": 9.998565105178869e-05,
      "loss": 2.440280532836914,
      "memory(GiB)": 38.45,
      "step": 890,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.496183
    },
    {
      "epoch": 0.03834454393556403,
      "grad_norm": 4.099660873413086,
      "learning_rate": 9.998548938259525e-05,
      "loss": 2.7564018249511717,
      "memory(GiB)": 38.45,
      "step": 895,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.496428
    },
    {
      "epoch": 0.03855875926481299,
      "grad_norm": 3.6905577182769775,
      "learning_rate": 9.998532680786778e-05,
      "loss": 2.4071306228637694,
      "memory(GiB)": 38.81,
      "step": 900,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.496277
    },
    {
      "epoch": 0.038772974594061954,
      "grad_norm": 3.0314619541168213,
      "learning_rate": 9.998516332760924e-05,
      "loss": 2.5806015014648436,
      "memory(GiB)": 38.81,
      "step": 905,
      "token_acc": 0.4557377049180328,
      "train_speed(iter/s)": 1.496542
    },
    {
      "epoch": 0.03898718992331091,
      "grad_norm": 4.86159086227417,
      "learning_rate": 9.998499894182255e-05,
      "loss": 2.5048938751220704,
      "memory(GiB)": 38.81,
      "step": 910,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.49598
    },
    {
      "epoch": 0.03920140525255987,
      "grad_norm": 3.315594434738159,
      "learning_rate": 9.998483365051075e-05,
      "loss": 2.897009086608887,
      "memory(GiB)": 38.81,
      "step": 915,
      "token_acc": 0.43018867924528303,
      "train_speed(iter/s)": 1.493181
    },
    {
      "epoch": 0.039415620581808836,
      "grad_norm": 3.834003210067749,
      "learning_rate": 9.998466745367678e-05,
      "loss": 2.5996976852416993,
      "memory(GiB)": 38.81,
      "step": 920,
      "token_acc": 0.4365781710914454,
      "train_speed(iter/s)": 1.492773
    },
    {
      "epoch": 0.039629835911057794,
      "grad_norm": 3.0740156173706055,
      "learning_rate": 9.998450035132367e-05,
      "loss": 2.842950439453125,
      "memory(GiB)": 38.81,
      "step": 925,
      "token_acc": 0.43506493506493504,
      "train_speed(iter/s)": 1.492662
    },
    {
      "epoch": 0.03984405124030676,
      "grad_norm": 4.596339225769043,
      "learning_rate": 9.998433234345446e-05,
      "loss": 2.635214424133301,
      "memory(GiB)": 38.81,
      "step": 930,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.493597
    },
    {
      "epoch": 0.04005826656955572,
      "grad_norm": 3.8838858604431152,
      "learning_rate": 9.998416343007215e-05,
      "loss": 2.54049072265625,
      "memory(GiB)": 38.81,
      "step": 935,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.493119
    },
    {
      "epoch": 0.040272481898804675,
      "grad_norm": 3.087055206298828,
      "learning_rate": 9.998399361117985e-05,
      "loss": 2.6478448867797852,
      "memory(GiB)": 38.81,
      "step": 940,
      "token_acc": 0.43444730077120824,
      "train_speed(iter/s)": 1.493678
    },
    {
      "epoch": 0.04048669722805364,
      "grad_norm": 3.6623730659484863,
      "learning_rate": 9.998382288678062e-05,
      "loss": 3.016253662109375,
      "memory(GiB)": 38.81,
      "step": 945,
      "token_acc": 0.4161290322580645,
      "train_speed(iter/s)": 1.495231
    },
    {
      "epoch": 0.0407009125573026,
      "grad_norm": 5.71432638168335,
      "learning_rate": 9.998365125687754e-05,
      "loss": 2.4765270233154295,
      "memory(GiB)": 38.81,
      "step": 950,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.494298
    },
    {
      "epoch": 0.040915127886551564,
      "grad_norm": 3.9099810123443604,
      "learning_rate": 9.998347872147374e-05,
      "loss": 2.8657649993896483,
      "memory(GiB)": 38.81,
      "step": 955,
      "token_acc": 0.43050847457627117,
      "train_speed(iter/s)": 1.494705
    },
    {
      "epoch": 0.04112934321580052,
      "grad_norm": 3.657832384109497,
      "learning_rate": 9.998330528057234e-05,
      "loss": 2.5781782150268553,
      "memory(GiB)": 38.81,
      "step": 960,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.495667
    },
    {
      "epoch": 0.04134355854504949,
      "grad_norm": 4.406288146972656,
      "learning_rate": 9.998313093417646e-05,
      "loss": 2.432485008239746,
      "memory(GiB)": 44.8,
      "step": 965,
      "token_acc": 0.4844961240310077,
      "train_speed(iter/s)": 1.494695
    },
    {
      "epoch": 0.041557773874298445,
      "grad_norm": 3.894352912902832,
      "learning_rate": 9.998295568228928e-05,
      "loss": 2.6198606491088867,
      "memory(GiB)": 44.8,
      "step": 970,
      "token_acc": 0.4541984732824427,
      "train_speed(iter/s)": 1.495261
    },
    {
      "epoch": 0.041771989203547404,
      "grad_norm": 2.8862714767456055,
      "learning_rate": 9.998277952491397e-05,
      "loss": 2.6841056823730467,
      "memory(GiB)": 44.8,
      "step": 975,
      "token_acc": 0.45592705167173253,
      "train_speed(iter/s)": 1.493921
    },
    {
      "epoch": 0.04198620453279637,
      "grad_norm": 3.875049352645874,
      "learning_rate": 9.998260246205372e-05,
      "loss": 2.9591068267822265,
      "memory(GiB)": 44.8,
      "step": 980,
      "token_acc": 0.4228395061728395,
      "train_speed(iter/s)": 1.495308
    },
    {
      "epoch": 0.04220041986204533,
      "grad_norm": 3.3060050010681152,
      "learning_rate": 9.998242449371174e-05,
      "loss": 2.8508010864257813,
      "memory(GiB)": 44.8,
      "step": 985,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.49519
    },
    {
      "epoch": 0.04241463519129429,
      "grad_norm": 6.291374206542969,
      "learning_rate": 9.998224561989124e-05,
      "loss": 2.6170028686523437,
      "memory(GiB)": 44.8,
      "step": 990,
      "token_acc": 0.4527687296416938,
      "train_speed(iter/s)": 1.495631
    },
    {
      "epoch": 0.04262885052054325,
      "grad_norm": 3.2309229373931885,
      "learning_rate": 9.99820658405955e-05,
      "loss": 2.529778480529785,
      "memory(GiB)": 44.8,
      "step": 995,
      "token_acc": 0.4478114478114478,
      "train_speed(iter/s)": 1.493933
    },
    {
      "epoch": 0.04284306584979221,
      "grad_norm": 7.27476167678833,
      "learning_rate": 9.998188515582773e-05,
      "loss": 2.746327781677246,
      "memory(GiB)": 44.8,
      "step": 1000,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.49489
    },
    {
      "epoch": 0.04284306584979221,
      "eval_loss": 2.4384655952453613,
      "eval_runtime": 14.749,
      "eval_samples_per_second": 6.78,
      "eval_steps_per_second": 6.78,
      "eval_token_acc": 0.44221698113207547,
      "step": 1000
    },
    {
      "epoch": 0.043057281179041174,
      "grad_norm": 3.783703565597534,
      "learning_rate": 9.998170356559125e-05,
      "loss": 2.604329299926758,
      "memory(GiB)": 44.8,
      "step": 1005,
      "token_acc": 0.4456993918331885,
      "train_speed(iter/s)": 1.460458
    },
    {
      "epoch": 0.04327149650829013,
      "grad_norm": 3.809075355529785,
      "learning_rate": 9.99815210698893e-05,
      "loss": 2.831036376953125,
      "memory(GiB)": 44.8,
      "step": 1010,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.461448
    },
    {
      "epoch": 0.0434857118375391,
      "grad_norm": 3.267028570175171,
      "learning_rate": 9.99813376687252e-05,
      "loss": 2.627566909790039,
      "memory(GiB)": 44.8,
      "step": 1015,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.461377
    },
    {
      "epoch": 0.043699927166788055,
      "grad_norm": 4.049720764160156,
      "learning_rate": 9.998115336210229e-05,
      "loss": 2.621623229980469,
      "memory(GiB)": 44.8,
      "step": 1020,
      "token_acc": 0.44609665427509293,
      "train_speed(iter/s)": 1.463755
    },
    {
      "epoch": 0.04391414249603701,
      "grad_norm": 4.8473429679870605,
      "learning_rate": 9.998096815002391e-05,
      "loss": 2.7243886947631837,
      "memory(GiB)": 44.8,
      "step": 1025,
      "token_acc": 0.4290657439446367,
      "train_speed(iter/s)": 1.464354
    },
    {
      "epoch": 0.04412835782528598,
      "grad_norm": 7.37783670425415,
      "learning_rate": 9.998078203249341e-05,
      "loss": 2.462293243408203,
      "memory(GiB)": 44.8,
      "step": 1030,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.465813
    },
    {
      "epoch": 0.04434257315453494,
      "grad_norm": 3.511589288711548,
      "learning_rate": 9.998059500951415e-05,
      "loss": 2.864263725280762,
      "memory(GiB)": 44.8,
      "step": 1035,
      "token_acc": 0.4065573770491803,
      "train_speed(iter/s)": 1.465889
    },
    {
      "epoch": 0.0445567884837839,
      "grad_norm": 3.4809505939483643,
      "learning_rate": 9.998040708108953e-05,
      "loss": 2.7204833984375,
      "memory(GiB)": 44.8,
      "step": 1040,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.466553
    },
    {
      "epoch": 0.04477100381303286,
      "grad_norm": 4.293155670166016,
      "learning_rate": 9.998021824722295e-05,
      "loss": 2.4542926788330077,
      "memory(GiB)": 44.8,
      "step": 1045,
      "token_acc": 0.49606299212598426,
      "train_speed(iter/s)": 1.467031
    },
    {
      "epoch": 0.044985219142281825,
      "grad_norm": 3.556633949279785,
      "learning_rate": 9.998002850791782e-05,
      "loss": 2.9304813385009765,
      "memory(GiB)": 44.8,
      "step": 1050,
      "token_acc": 0.41754385964912283,
      "train_speed(iter/s)": 1.46766
    },
    {
      "epoch": 0.04519943447153078,
      "grad_norm": 6.549159526824951,
      "learning_rate": 9.99798378631776e-05,
      "loss": 2.503632354736328,
      "memory(GiB)": 44.8,
      "step": 1055,
      "token_acc": 0.4559386973180077,
      "train_speed(iter/s)": 1.467155
    },
    {
      "epoch": 0.04541364980077974,
      "grad_norm": 4.304167747497559,
      "learning_rate": 9.997964631300573e-05,
      "loss": 2.888266372680664,
      "memory(GiB)": 44.8,
      "step": 1060,
      "token_acc": 0.41694915254237286,
      "train_speed(iter/s)": 1.466603
    },
    {
      "epoch": 0.04562786513002871,
      "grad_norm": 4.314018726348877,
      "learning_rate": 9.997945385740568e-05,
      "loss": 2.6884815216064455,
      "memory(GiB)": 44.8,
      "step": 1065,
      "token_acc": 0.4276729559748428,
      "train_speed(iter/s)": 1.466345
    },
    {
      "epoch": 0.045842080459277665,
      "grad_norm": 3.1511995792388916,
      "learning_rate": 9.997926049638094e-05,
      "loss": 2.5820171356201174,
      "memory(GiB)": 44.8,
      "step": 1070,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.468005
    },
    {
      "epoch": 0.04605629578852663,
      "grad_norm": 5.352568626403809,
      "learning_rate": 9.9979066229935e-05,
      "loss": 2.7415964126586916,
      "memory(GiB)": 44.8,
      "step": 1075,
      "token_acc": 0.4830188679245283,
      "train_speed(iter/s)": 1.467664
    },
    {
      "epoch": 0.04627051111777559,
      "grad_norm": 2.98688006401062,
      "learning_rate": 9.997887105807141e-05,
      "loss": 2.7664222717285156,
      "memory(GiB)": 44.8,
      "step": 1080,
      "token_acc": 0.4382716049382716,
      "train_speed(iter/s)": 1.466714
    },
    {
      "epoch": 0.046484726447024546,
      "grad_norm": 4.0362114906311035,
      "learning_rate": 9.997867498079368e-05,
      "loss": 2.9460811614990234,
      "memory(GiB)": 44.8,
      "step": 1085,
      "token_acc": 0.43018867924528303,
      "train_speed(iter/s)": 1.466689
    },
    {
      "epoch": 0.04669894177627351,
      "grad_norm": 3.121070623397827,
      "learning_rate": 9.997847799810537e-05,
      "loss": 2.4676792144775392,
      "memory(GiB)": 44.8,
      "step": 1090,
      "token_acc": 0.4887459807073955,
      "train_speed(iter/s)": 1.466383
    },
    {
      "epoch": 0.04691315710552247,
      "grad_norm": 3.5065741539001465,
      "learning_rate": 9.997828011001005e-05,
      "loss": 2.639570617675781,
      "memory(GiB)": 44.8,
      "step": 1095,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.465902
    },
    {
      "epoch": 0.047127372434771435,
      "grad_norm": 3.773871898651123,
      "learning_rate": 9.99780813165113e-05,
      "loss": 2.541779899597168,
      "memory(GiB)": 44.8,
      "step": 1100,
      "token_acc": 0.456973293768546,
      "train_speed(iter/s)": 1.46657
    },
    {
      "epoch": 0.04734158776402039,
      "grad_norm": 4.274632453918457,
      "learning_rate": 9.997788161761272e-05,
      "loss": 2.544404220581055,
      "memory(GiB)": 44.8,
      "step": 1105,
      "token_acc": 0.4872881355932203,
      "train_speed(iter/s)": 1.466749
    },
    {
      "epoch": 0.04755580309326935,
      "grad_norm": 5.100984573364258,
      "learning_rate": 9.997768101331793e-05,
      "loss": 2.5767072677612304,
      "memory(GiB)": 44.8,
      "step": 1110,
      "token_acc": 0.43653250773993807,
      "train_speed(iter/s)": 1.465494
    },
    {
      "epoch": 0.047770018422518316,
      "grad_norm": 3.7199249267578125,
      "learning_rate": 9.997747950363056e-05,
      "loss": 2.7794525146484377,
      "memory(GiB)": 44.8,
      "step": 1115,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.466184
    },
    {
      "epoch": 0.047984233751767275,
      "grad_norm": 3.925203323364258,
      "learning_rate": 9.997727708855429e-05,
      "loss": 2.4455743789672852,
      "memory(GiB)": 44.8,
      "step": 1120,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.466064
    },
    {
      "epoch": 0.04819844908101624,
      "grad_norm": 5.212133407592773,
      "learning_rate": 9.997707376809273e-05,
      "loss": 2.6166486740112305,
      "memory(GiB)": 44.8,
      "step": 1125,
      "token_acc": 0.5150375939849624,
      "train_speed(iter/s)": 1.467475
    },
    {
      "epoch": 0.0484126644102652,
      "grad_norm": 5.15431022644043,
      "learning_rate": 9.997686954224963e-05,
      "loss": 2.6525360107421876,
      "memory(GiB)": 44.8,
      "step": 1130,
      "token_acc": 0.4527687296416938,
      "train_speed(iter/s)": 1.466977
    },
    {
      "epoch": 0.04862687973951416,
      "grad_norm": 3.462268829345703,
      "learning_rate": 9.997666441102864e-05,
      "loss": 2.814067077636719,
      "memory(GiB)": 44.8,
      "step": 1135,
      "token_acc": 0.3853211009174312,
      "train_speed(iter/s)": 1.467258
    },
    {
      "epoch": 0.04884109506876312,
      "grad_norm": 3.796671152114868,
      "learning_rate": 9.99764583744335e-05,
      "loss": 2.741273307800293,
      "memory(GiB)": 44.8,
      "step": 1140,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.466789
    },
    {
      "epoch": 0.04905531039801208,
      "grad_norm": 4.840392112731934,
      "learning_rate": 9.997625143246791e-05,
      "loss": 2.6720537185668944,
      "memory(GiB)": 44.8,
      "step": 1145,
      "token_acc": 0.46774193548387094,
      "train_speed(iter/s)": 1.466716
    },
    {
      "epoch": 0.049269525727261045,
      "grad_norm": 3.183471202850342,
      "learning_rate": 9.997604358513566e-05,
      "loss": 2.748935317993164,
      "memory(GiB)": 44.8,
      "step": 1150,
      "token_acc": 0.4290322580645161,
      "train_speed(iter/s)": 1.465586
    },
    {
      "epoch": 0.04948374105651,
      "grad_norm": 3.9429311752319336,
      "learning_rate": 9.997583483244051e-05,
      "loss": 2.6724550247192385,
      "memory(GiB)": 44.8,
      "step": 1155,
      "token_acc": 0.43661971830985913,
      "train_speed(iter/s)": 1.465914
    },
    {
      "epoch": 0.04969795638575897,
      "grad_norm": 4.010347366333008,
      "learning_rate": 9.997562517438621e-05,
      "loss": 2.78275146484375,
      "memory(GiB)": 44.8,
      "step": 1160,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.467317
    },
    {
      "epoch": 0.049912171715007926,
      "grad_norm": 3.6683743000030518,
      "learning_rate": 9.997541461097659e-05,
      "loss": 2.859244537353516,
      "memory(GiB)": 44.8,
      "step": 1165,
      "token_acc": 0.39473684210526316,
      "train_speed(iter/s)": 1.468769
    },
    {
      "epoch": 0.050126387044256884,
      "grad_norm": 3.7796099185943604,
      "learning_rate": 9.997520314221547e-05,
      "loss": 2.8719547271728514,
      "memory(GiB)": 44.8,
      "step": 1170,
      "token_acc": 0.42813455657492355,
      "train_speed(iter/s)": 1.467851
    },
    {
      "epoch": 0.05034060237350585,
      "grad_norm": 3.327528715133667,
      "learning_rate": 9.997499076810664e-05,
      "loss": 2.911891555786133,
      "memory(GiB)": 44.8,
      "step": 1175,
      "token_acc": 0.44694533762057875,
      "train_speed(iter/s)": 1.468474
    },
    {
      "epoch": 0.05055481770275481,
      "grad_norm": 4.597551345825195,
      "learning_rate": 9.997477748865399e-05,
      "loss": 2.5749902725219727,
      "memory(GiB)": 44.8,
      "step": 1180,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.468247
    },
    {
      "epoch": 0.05076903303200377,
      "grad_norm": 4.473481178283691,
      "learning_rate": 9.997456330386135e-05,
      "loss": 2.702045440673828,
      "memory(GiB)": 44.8,
      "step": 1185,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.46898
    },
    {
      "epoch": 0.05098324836125273,
      "grad_norm": 3.2365341186523438,
      "learning_rate": 9.997434821373262e-05,
      "loss": 2.5826753616333007,
      "memory(GiB)": 44.8,
      "step": 1190,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.469398
    },
    {
      "epoch": 0.05119746369050169,
      "grad_norm": 5.502979278564453,
      "learning_rate": 9.997413221827169e-05,
      "loss": 2.5995899200439454,
      "memory(GiB)": 44.8,
      "step": 1195,
      "token_acc": 0.48264984227129337,
      "train_speed(iter/s)": 1.470224
    },
    {
      "epoch": 0.051411679019750654,
      "grad_norm": 5.472434997558594,
      "learning_rate": 9.997391531748248e-05,
      "loss": 2.81494026184082,
      "memory(GiB)": 44.8,
      "step": 1200,
      "token_acc": 0.5036764705882353,
      "train_speed(iter/s)": 1.471308
    },
    {
      "epoch": 0.05162589434899961,
      "grad_norm": 3.0724525451660156,
      "learning_rate": 9.997369751136891e-05,
      "loss": 2.6393428802490235,
      "memory(GiB)": 44.8,
      "step": 1205,
      "token_acc": 0.4662576687116564,
      "train_speed(iter/s)": 1.470597
    },
    {
      "epoch": 0.05184010967824858,
      "grad_norm": 5.035545349121094,
      "learning_rate": 9.997347879993495e-05,
      "loss": 2.508587646484375,
      "memory(GiB)": 44.8,
      "step": 1210,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.469467
    },
    {
      "epoch": 0.052054325007497536,
      "grad_norm": 3.6228578090667725,
      "learning_rate": 9.997325918318452e-05,
      "loss": 2.4395151138305664,
      "memory(GiB)": 44.8,
      "step": 1215,
      "token_acc": 0.4875,
      "train_speed(iter/s)": 1.468092
    },
    {
      "epoch": 0.0522685403367465,
      "grad_norm": 4.701809406280518,
      "learning_rate": 9.997303866112163e-05,
      "loss": 2.589558982849121,
      "memory(GiB)": 44.8,
      "step": 1220,
      "token_acc": 0.4472843450479233,
      "train_speed(iter/s)": 1.466086
    },
    {
      "epoch": 0.05248275566599546,
      "grad_norm": 4.492598533630371,
      "learning_rate": 9.997281723375027e-05,
      "loss": 2.6755788803100584,
      "memory(GiB)": 44.8,
      "step": 1225,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.467089
    },
    {
      "epoch": 0.05269697099524442,
      "grad_norm": 3.3833775520324707,
      "learning_rate": 9.997259490107444e-05,
      "loss": 2.5095600128173827,
      "memory(GiB)": 44.8,
      "step": 1230,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.468978
    },
    {
      "epoch": 0.05291118632449338,
      "grad_norm": 5.671452045440674,
      "learning_rate": 9.997237166309818e-05,
      "loss": 2.6694406509399413,
      "memory(GiB)": 44.8,
      "step": 1235,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.469868
    },
    {
      "epoch": 0.05312540165374234,
      "grad_norm": 8.992464065551758,
      "learning_rate": 9.997214751982552e-05,
      "loss": 2.5071449279785156,
      "memory(GiB)": 44.8,
      "step": 1240,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.469788
    },
    {
      "epoch": 0.053339616982991306,
      "grad_norm": 3.487832546234131,
      "learning_rate": 9.997192247126055e-05,
      "loss": 2.59873046875,
      "memory(GiB)": 44.8,
      "step": 1245,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.469325
    },
    {
      "epoch": 0.053553832312240264,
      "grad_norm": 3.6097612380981445,
      "learning_rate": 9.99716965174073e-05,
      "loss": 2.4442211151123048,
      "memory(GiB)": 44.8,
      "step": 1250,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.471082
    },
    {
      "epoch": 0.05376804764148922,
      "grad_norm": 4.641023635864258,
      "learning_rate": 9.997146965826991e-05,
      "loss": 2.5057844161987304,
      "memory(GiB)": 44.8,
      "step": 1255,
      "token_acc": 0.45348837209302323,
      "train_speed(iter/s)": 1.471497
    },
    {
      "epoch": 0.05398226297073819,
      "grad_norm": 3.097318172454834,
      "learning_rate": 9.997124189385246e-05,
      "loss": 2.6078752517700194,
      "memory(GiB)": 44.8,
      "step": 1260,
      "token_acc": 0.46006389776357826,
      "train_speed(iter/s)": 1.47263
    },
    {
      "epoch": 0.054196478299987146,
      "grad_norm": 3.847569465637207,
      "learning_rate": 9.99710132241591e-05,
      "loss": 2.4449111938476564,
      "memory(GiB)": 44.8,
      "step": 1265,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.472272
    },
    {
      "epoch": 0.05441069362923611,
      "grad_norm": 4.654448509216309,
      "learning_rate": 9.997078364919395e-05,
      "loss": 2.8197586059570314,
      "memory(GiB)": 44.8,
      "step": 1270,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.473333
    },
    {
      "epoch": 0.05462490895848507,
      "grad_norm": 3.8379368782043457,
      "learning_rate": 9.997055316896118e-05,
      "loss": 2.7199775695800783,
      "memory(GiB)": 44.8,
      "step": 1275,
      "token_acc": 0.45724907063197023,
      "train_speed(iter/s)": 1.473807
    },
    {
      "epoch": 0.05483912428773403,
      "grad_norm": 4.283139705657959,
      "learning_rate": 9.997032178346495e-05,
      "loss": 2.676108741760254,
      "memory(GiB)": 44.8,
      "step": 1280,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.475485
    },
    {
      "epoch": 0.05505333961698299,
      "grad_norm": 4.635933876037598,
      "learning_rate": 9.997008949270947e-05,
      "loss": 2.517548179626465,
      "memory(GiB)": 44.8,
      "step": 1285,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.476331
    },
    {
      "epoch": 0.05526755494623195,
      "grad_norm": 8.383788108825684,
      "learning_rate": 9.996985629669894e-05,
      "loss": 2.239846420288086,
      "memory(GiB)": 44.8,
      "step": 1290,
      "token_acc": 0.48372093023255813,
      "train_speed(iter/s)": 1.475744
    },
    {
      "epoch": 0.055481770275480916,
      "grad_norm": 4.16169548034668,
      "learning_rate": 9.996962219543762e-05,
      "loss": 2.500482177734375,
      "memory(GiB)": 44.8,
      "step": 1295,
      "token_acc": 0.45482866043613707,
      "train_speed(iter/s)": 1.474096
    },
    {
      "epoch": 0.055695985604729874,
      "grad_norm": 3.961717367172241,
      "learning_rate": 9.996938718892969e-05,
      "loss": 2.8550065994262694,
      "memory(GiB)": 44.8,
      "step": 1300,
      "token_acc": 0.423728813559322,
      "train_speed(iter/s)": 1.473889
    },
    {
      "epoch": 0.05591020093397884,
      "grad_norm": 4.51633358001709,
      "learning_rate": 9.996915127717944e-05,
      "loss": 2.781209182739258,
      "memory(GiB)": 44.8,
      "step": 1305,
      "token_acc": 0.37992831541218636,
      "train_speed(iter/s)": 1.472603
    },
    {
      "epoch": 0.0561244162632278,
      "grad_norm": 2.7895588874816895,
      "learning_rate": 9.996891446019114e-05,
      "loss": 2.6287181854248045,
      "memory(GiB)": 44.8,
      "step": 1310,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.473199
    },
    {
      "epoch": 0.056338631592476755,
      "grad_norm": 3.305253028869629,
      "learning_rate": 9.996867673796908e-05,
      "loss": 2.6148349761962892,
      "memory(GiB)": 44.8,
      "step": 1315,
      "token_acc": 0.4626334519572954,
      "train_speed(iter/s)": 1.472924
    },
    {
      "epoch": 0.05655284692172572,
      "grad_norm": 3.111687421798706,
      "learning_rate": 9.996843811051757e-05,
      "loss": 2.5816940307617187,
      "memory(GiB)": 44.8,
      "step": 1320,
      "token_acc": 0.45741324921135645,
      "train_speed(iter/s)": 1.47242
    },
    {
      "epoch": 0.05676706225097468,
      "grad_norm": 3.986253499984741,
      "learning_rate": 9.996819857784093e-05,
      "loss": 2.53792667388916,
      "memory(GiB)": 44.8,
      "step": 1325,
      "token_acc": 0.47038327526132406,
      "train_speed(iter/s)": 1.472465
    },
    {
      "epoch": 0.056981277580223644,
      "grad_norm": 3.378695487976074,
      "learning_rate": 9.996795813994348e-05,
      "loss": 2.870656967163086,
      "memory(GiB)": 44.8,
      "step": 1330,
      "token_acc": 0.41721854304635764,
      "train_speed(iter/s)": 1.472482
    },
    {
      "epoch": 0.0571954929094726,
      "grad_norm": 5.106979846954346,
      "learning_rate": 9.996771679682961e-05,
      "loss": 2.8496253967285154,
      "memory(GiB)": 44.8,
      "step": 1335,
      "token_acc": 0.41284403669724773,
      "train_speed(iter/s)": 1.473367
    },
    {
      "epoch": 0.05740970823872156,
      "grad_norm": 5.238814353942871,
      "learning_rate": 9.996747454850368e-05,
      "loss": 2.7841476440429687,
      "memory(GiB)": 44.8,
      "step": 1340,
      "token_acc": 0.44571428571428573,
      "train_speed(iter/s)": 1.473836
    },
    {
      "epoch": 0.057623923567970525,
      "grad_norm": 3.4206931591033936,
      "learning_rate": 9.996723139497008e-05,
      "loss": 2.5446672439575195,
      "memory(GiB)": 44.8,
      "step": 1345,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.472902
    },
    {
      "epoch": 0.057838138897219483,
      "grad_norm": 3.4484548568725586,
      "learning_rate": 9.99669873362332e-05,
      "loss": 2.4125125885009764,
      "memory(GiB)": 44.8,
      "step": 1350,
      "token_acc": 0.4813664596273292,
      "train_speed(iter/s)": 1.47399
    },
    {
      "epoch": 0.05805235422646845,
      "grad_norm": 4.250040054321289,
      "learning_rate": 9.996674237229748e-05,
      "loss": 2.986496162414551,
      "memory(GiB)": 44.8,
      "step": 1355,
      "token_acc": 0.4309623430962343,
      "train_speed(iter/s)": 1.474815
    },
    {
      "epoch": 0.05826656955571741,
      "grad_norm": 3.5987865924835205,
      "learning_rate": 9.996649650316734e-05,
      "loss": 2.768035888671875,
      "memory(GiB)": 44.8,
      "step": 1360,
      "token_acc": 0.44086021505376344,
      "train_speed(iter/s)": 1.475955
    },
    {
      "epoch": 0.058480784884966365,
      "grad_norm": 4.841388702392578,
      "learning_rate": 9.996624972884725e-05,
      "loss": 2.8981708526611327,
      "memory(GiB)": 44.8,
      "step": 1365,
      "token_acc": 0.44368600682593856,
      "train_speed(iter/s)": 1.477139
    },
    {
      "epoch": 0.05869500021421533,
      "grad_norm": 5.3648295402526855,
      "learning_rate": 9.996600204934165e-05,
      "loss": 2.895512008666992,
      "memory(GiB)": 44.8,
      "step": 1370,
      "token_acc": 0.41434262948207173,
      "train_speed(iter/s)": 1.478515
    },
    {
      "epoch": 0.05890921554346429,
      "grad_norm": 5.131505012512207,
      "learning_rate": 9.996575346465508e-05,
      "loss": 2.605813980102539,
      "memory(GiB)": 44.8,
      "step": 1375,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.478754
    },
    {
      "epoch": 0.059123430872713253,
      "grad_norm": 2.7781922817230225,
      "learning_rate": 9.9965503974792e-05,
      "loss": 2.5423892974853515,
      "memory(GiB)": 44.8,
      "step": 1380,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.478682
    },
    {
      "epoch": 0.05933764620196221,
      "grad_norm": 4.505220413208008,
      "learning_rate": 9.996525357975697e-05,
      "loss": 2.81160888671875,
      "memory(GiB)": 44.8,
      "step": 1385,
      "token_acc": 0.4107744107744108,
      "train_speed(iter/s)": 1.47827
    },
    {
      "epoch": 0.05955186153121118,
      "grad_norm": 3.481152296066284,
      "learning_rate": 9.996500227955447e-05,
      "loss": 2.6756229400634766,
      "memory(GiB)": 44.8,
      "step": 1390,
      "token_acc": 0.44947735191637633,
      "train_speed(iter/s)": 1.479092
    },
    {
      "epoch": 0.059766076860460135,
      "grad_norm": 28.509784698486328,
      "learning_rate": 9.99647500741891e-05,
      "loss": 2.8012458801269533,
      "memory(GiB)": 44.8,
      "step": 1395,
      "token_acc": 0.4092526690391459,
      "train_speed(iter/s)": 1.479515
    },
    {
      "epoch": 0.05998029218970909,
      "grad_norm": 4.454799652099609,
      "learning_rate": 9.99644969636654e-05,
      "loss": 2.4655921936035154,
      "memory(GiB)": 44.8,
      "step": 1400,
      "token_acc": 0.49603174603174605,
      "train_speed(iter/s)": 1.479241
    },
    {
      "epoch": 0.06019450751895806,
      "grad_norm": 4.337578773498535,
      "learning_rate": 9.996424294798796e-05,
      "loss": 2.85654296875,
      "memory(GiB)": 44.8,
      "step": 1405,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.480932
    },
    {
      "epoch": 0.06040872284820702,
      "grad_norm": 4.483986854553223,
      "learning_rate": 9.99639880271614e-05,
      "loss": 2.564306640625,
      "memory(GiB)": 44.8,
      "step": 1410,
      "token_acc": 0.5022421524663677,
      "train_speed(iter/s)": 1.481714
    },
    {
      "epoch": 0.06062293817745598,
      "grad_norm": 3.3506863117218018,
      "learning_rate": 9.996373220119034e-05,
      "loss": 2.8136690139770506,
      "memory(GiB)": 44.8,
      "step": 1415,
      "token_acc": 0.4037854889589905,
      "train_speed(iter/s)": 1.481858
    },
    {
      "epoch": 0.06083715350670494,
      "grad_norm": 3.624328374862671,
      "learning_rate": 9.996347547007938e-05,
      "loss": 2.490763473510742,
      "memory(GiB)": 44.8,
      "step": 1420,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.480731
    },
    {
      "epoch": 0.0610513688359539,
      "grad_norm": 3.642944574356079,
      "learning_rate": 9.99632178338332e-05,
      "loss": 2.7100624084472655,
      "memory(GiB)": 44.8,
      "step": 1425,
      "token_acc": 0.45535714285714285,
      "train_speed(iter/s)": 1.482317
    },
    {
      "epoch": 0.06126558416520286,
      "grad_norm": 4.454731464385986,
      "learning_rate": 9.996295929245647e-05,
      "loss": 2.6536014556884764,
      "memory(GiB)": 44.8,
      "step": 1430,
      "token_acc": 0.4139344262295082,
      "train_speed(iter/s)": 1.483121
    },
    {
      "epoch": 0.06147979949445182,
      "grad_norm": 4.322632312774658,
      "learning_rate": 9.996269984595386e-05,
      "loss": 2.7533239364624023,
      "memory(GiB)": 44.8,
      "step": 1435,
      "token_acc": 0.43776824034334766,
      "train_speed(iter/s)": 1.483148
    },
    {
      "epoch": 0.06169401482370079,
      "grad_norm": 3.861013412475586,
      "learning_rate": 9.996243949433006e-05,
      "loss": 2.5387073516845704,
      "memory(GiB)": 44.8,
      "step": 1440,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.483337
    },
    {
      "epoch": 0.061908230152949745,
      "grad_norm": 5.236702919006348,
      "learning_rate": 9.996217823758981e-05,
      "loss": 2.83597469329834,
      "memory(GiB)": 44.8,
      "step": 1445,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.482374
    },
    {
      "epoch": 0.0621224454821987,
      "grad_norm": 2.9103434085845947,
      "learning_rate": 9.996191607573782e-05,
      "loss": 2.809426689147949,
      "memory(GiB)": 44.8,
      "step": 1450,
      "token_acc": 0.4393939393939394,
      "train_speed(iter/s)": 1.481722
    },
    {
      "epoch": 0.06233666081144767,
      "grad_norm": 4.294338703155518,
      "learning_rate": 9.996165300877886e-05,
      "loss": 2.5830265045166017,
      "memory(GiB)": 44.8,
      "step": 1455,
      "token_acc": 0.43617021276595747,
      "train_speed(iter/s)": 1.481991
    },
    {
      "epoch": 0.06255087614069663,
      "grad_norm": 5.332017421722412,
      "learning_rate": 9.99613890367177e-05,
      "loss": 2.322412109375,
      "memory(GiB)": 44.8,
      "step": 1460,
      "token_acc": 0.5394190871369294,
      "train_speed(iter/s)": 1.480375
    },
    {
      "epoch": 0.06276509146994559,
      "grad_norm": 3.3627238273620605,
      "learning_rate": 9.99611241595591e-05,
      "loss": 2.8908124923706056,
      "memory(GiB)": 44.8,
      "step": 1465,
      "token_acc": 0.4270833333333333,
      "train_speed(iter/s)": 1.481475
    },
    {
      "epoch": 0.06297930679919456,
      "grad_norm": 6.100260257720947,
      "learning_rate": 9.996085837730787e-05,
      "loss": 2.6980457305908203,
      "memory(GiB)": 44.8,
      "step": 1470,
      "token_acc": 0.4414715719063545,
      "train_speed(iter/s)": 1.48171
    },
    {
      "epoch": 0.06319352212844351,
      "grad_norm": 3.991400957107544,
      "learning_rate": 9.996059168996883e-05,
      "loss": 2.712911033630371,
      "memory(GiB)": 44.8,
      "step": 1475,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.483066
    },
    {
      "epoch": 0.06340773745769247,
      "grad_norm": 4.39341926574707,
      "learning_rate": 9.996032409754679e-05,
      "loss": 2.7493879318237306,
      "memory(GiB)": 44.8,
      "step": 1480,
      "token_acc": 0.4074074074074074,
      "train_speed(iter/s)": 1.483578
    },
    {
      "epoch": 0.06362195278694144,
      "grad_norm": 4.587990760803223,
      "learning_rate": 9.996005560004662e-05,
      "loss": 2.8541988372802733,
      "memory(GiB)": 44.8,
      "step": 1485,
      "token_acc": 0.445578231292517,
      "train_speed(iter/s)": 1.484213
    },
    {
      "epoch": 0.06383616811619039,
      "grad_norm": 3.9578187465667725,
      "learning_rate": 9.995978619747317e-05,
      "loss": 2.6278173446655275,
      "memory(GiB)": 44.8,
      "step": 1490,
      "token_acc": 0.4,
      "train_speed(iter/s)": 1.4848
    },
    {
      "epoch": 0.06405038344543935,
      "grad_norm": 6.137836933135986,
      "learning_rate": 9.995951588983135e-05,
      "loss": 2.4550449371337892,
      "memory(GiB)": 44.8,
      "step": 1495,
      "token_acc": 0.5147679324894515,
      "train_speed(iter/s)": 1.484099
    },
    {
      "epoch": 0.06426459877468832,
      "grad_norm": 4.868353366851807,
      "learning_rate": 9.995924467712601e-05,
      "loss": 2.4769081115722655,
      "memory(GiB)": 44.8,
      "step": 1500,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.48376
    },
    {
      "epoch": 0.06426459877468832,
      "eval_loss": 2.4918837547302246,
      "eval_runtime": 13.488,
      "eval_samples_per_second": 7.414,
      "eval_steps_per_second": 7.414,
      "eval_token_acc": 0.4443155452436195,
      "step": 1500
    },
    {
      "epoch": 0.06447881410393727,
      "grad_norm": 3.260181188583374,
      "learning_rate": 9.995897255936209e-05,
      "loss": 2.8356361389160156,
      "memory(GiB)": 44.8,
      "step": 1505,
      "token_acc": 0.43099068585944117,
      "train_speed(iter/s)": 1.464067
    },
    {
      "epoch": 0.06469302943318624,
      "grad_norm": 4.076349258422852,
      "learning_rate": 9.995869953654452e-05,
      "loss": 2.3611581802368162,
      "memory(GiB)": 44.8,
      "step": 1510,
      "token_acc": 0.5037878787878788,
      "train_speed(iter/s)": 1.4648
    },
    {
      "epoch": 0.0649072447624352,
      "grad_norm": 4.669004440307617,
      "learning_rate": 9.995842560867826e-05,
      "loss": 2.7666147232055662,
      "memory(GiB)": 44.8,
      "step": 1515,
      "token_acc": 0.45907473309608543,
      "train_speed(iter/s)": 1.465446
    },
    {
      "epoch": 0.06512146009168417,
      "grad_norm": 4.889930725097656,
      "learning_rate": 9.995815077576822e-05,
      "loss": 2.6024490356445313,
      "memory(GiB)": 44.8,
      "step": 1520,
      "token_acc": 0.46706586826347307,
      "train_speed(iter/s)": 1.465824
    },
    {
      "epoch": 0.06533567542093312,
      "grad_norm": 4.200222492218018,
      "learning_rate": 9.995787503781944e-05,
      "loss": 2.8650585174560548,
      "memory(GiB)": 44.8,
      "step": 1525,
      "token_acc": 0.4419475655430712,
      "train_speed(iter/s)": 1.466276
    },
    {
      "epoch": 0.06554989075018208,
      "grad_norm": 3.2476115226745605,
      "learning_rate": 9.995759839483687e-05,
      "loss": 2.609589767456055,
      "memory(GiB)": 44.8,
      "step": 1530,
      "token_acc": 0.4434250764525994,
      "train_speed(iter/s)": 1.467032
    },
    {
      "epoch": 0.06576410607943105,
      "grad_norm": 5.784073352813721,
      "learning_rate": 9.995732084682557e-05,
      "loss": 2.655089569091797,
      "memory(GiB)": 44.8,
      "step": 1535,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.4679
    },
    {
      "epoch": 0.06597832140868,
      "grad_norm": 4.483177661895752,
      "learning_rate": 9.995704239379051e-05,
      "loss": 2.618147850036621,
      "memory(GiB)": 44.8,
      "step": 1540,
      "token_acc": 0.4559748427672956,
      "train_speed(iter/s)": 1.468061
    },
    {
      "epoch": 0.06619253673792896,
      "grad_norm": 3.7003676891326904,
      "learning_rate": 9.995676303573678e-05,
      "loss": 2.632536697387695,
      "memory(GiB)": 44.8,
      "step": 1545,
      "token_acc": 0.5162337662337663,
      "train_speed(iter/s)": 1.469355
    },
    {
      "epoch": 0.06640675206717793,
      "grad_norm": 5.321671485900879,
      "learning_rate": 9.995648277266942e-05,
      "loss": 2.593501663208008,
      "memory(GiB)": 44.8,
      "step": 1550,
      "token_acc": 0.4828897338403042,
      "train_speed(iter/s)": 1.469549
    },
    {
      "epoch": 0.0666209673964269,
      "grad_norm": 4.74328088760376,
      "learning_rate": 9.995620160459351e-05,
      "loss": 2.894466209411621,
      "memory(GiB)": 44.8,
      "step": 1555,
      "token_acc": 0.41901408450704225,
      "train_speed(iter/s)": 1.469672
    },
    {
      "epoch": 0.06683518272567585,
      "grad_norm": 3.247278928756714,
      "learning_rate": 9.995591953151415e-05,
      "loss": 2.749079704284668,
      "memory(GiB)": 44.8,
      "step": 1560,
      "token_acc": 0.44904458598726116,
      "train_speed(iter/s)": 1.469661
    },
    {
      "epoch": 0.06704939805492481,
      "grad_norm": 4.189208030700684,
      "learning_rate": 9.995563655343645e-05,
      "loss": 2.3000814437866213,
      "memory(GiB)": 44.8,
      "step": 1565,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.468532
    },
    {
      "epoch": 0.06726361338417378,
      "grad_norm": 3.8709604740142822,
      "learning_rate": 9.995535267036551e-05,
      "loss": 2.960202217102051,
      "memory(GiB)": 44.8,
      "step": 1570,
      "token_acc": 0.4084967320261438,
      "train_speed(iter/s)": 1.468213
    },
    {
      "epoch": 0.06747782871342273,
      "grad_norm": 3.0151495933532715,
      "learning_rate": 9.995506788230652e-05,
      "loss": 2.9425230026245117,
      "memory(GiB)": 44.8,
      "step": 1575,
      "token_acc": 0.421875,
      "train_speed(iter/s)": 1.467536
    },
    {
      "epoch": 0.06769204404267169,
      "grad_norm": 4.0558857917785645,
      "learning_rate": 9.99547821892646e-05,
      "loss": 2.702001762390137,
      "memory(GiB)": 44.8,
      "step": 1580,
      "token_acc": 0.4326923076923077,
      "train_speed(iter/s)": 1.467546
    },
    {
      "epoch": 0.06790625937192066,
      "grad_norm": 4.547561168670654,
      "learning_rate": 9.995449559124495e-05,
      "loss": 2.6235677719116213,
      "memory(GiB)": 44.8,
      "step": 1585,
      "token_acc": 0.4921259842519685,
      "train_speed(iter/s)": 1.468384
    },
    {
      "epoch": 0.06812047470116961,
      "grad_norm": 3.4789013862609863,
      "learning_rate": 9.995420808825274e-05,
      "loss": 2.6372425079345705,
      "memory(GiB)": 44.8,
      "step": 1590,
      "token_acc": 0.4419475655430712,
      "train_speed(iter/s)": 1.468392
    },
    {
      "epoch": 0.06833469003041857,
      "grad_norm": 3.6526811122894287,
      "learning_rate": 9.995391968029318e-05,
      "loss": 2.6607057571411135,
      "memory(GiB)": 44.8,
      "step": 1595,
      "token_acc": 0.44765342960288806,
      "train_speed(iter/s)": 1.465144
    },
    {
      "epoch": 0.06854890535966754,
      "grad_norm": 3.4350223541259766,
      "learning_rate": 9.995363036737152e-05,
      "loss": 2.662113380432129,
      "memory(GiB)": 44.8,
      "step": 1600,
      "token_acc": 0.4306569343065693,
      "train_speed(iter/s)": 1.465347
    },
    {
      "epoch": 0.0687631206889165,
      "grad_norm": 10.293490409851074,
      "learning_rate": 9.995334014949297e-05,
      "loss": 2.257463073730469,
      "memory(GiB)": 44.8,
      "step": 1605,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.464816
    },
    {
      "epoch": 0.06897733601816546,
      "grad_norm": 3.9372994899749756,
      "learning_rate": 9.995304902666283e-05,
      "loss": 2.976861763000488,
      "memory(GiB)": 44.8,
      "step": 1610,
      "token_acc": 0.4072948328267477,
      "train_speed(iter/s)": 1.464416
    },
    {
      "epoch": 0.06919155134741442,
      "grad_norm": 5.508297920227051,
      "learning_rate": 9.995275699888633e-05,
      "loss": 2.9259031295776365,
      "memory(GiB)": 44.8,
      "step": 1615,
      "token_acc": 0.42028985507246375,
      "train_speed(iter/s)": 1.464444
    },
    {
      "epoch": 0.06940576667666339,
      "grad_norm": 5.092051982879639,
      "learning_rate": 9.995246406616879e-05,
      "loss": 2.6340478897094726,
      "memory(GiB)": 44.8,
      "step": 1620,
      "token_acc": 0.49096385542168675,
      "train_speed(iter/s)": 1.465172
    },
    {
      "epoch": 0.06961998200591234,
      "grad_norm": 4.231055736541748,
      "learning_rate": 9.995217022851548e-05,
      "loss": 2.37537841796875,
      "memory(GiB)": 44.8,
      "step": 1625,
      "token_acc": 0.5107296137339056,
      "train_speed(iter/s)": 1.464972
    },
    {
      "epoch": 0.0698341973351613,
      "grad_norm": 4.58602237701416,
      "learning_rate": 9.995187548593176e-05,
      "loss": 2.751350975036621,
      "memory(GiB)": 44.8,
      "step": 1630,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.465782
    },
    {
      "epoch": 0.07004841266441027,
      "grad_norm": 3.5853564739227295,
      "learning_rate": 9.995157983842297e-05,
      "loss": 2.858659362792969,
      "memory(GiB)": 44.8,
      "step": 1635,
      "token_acc": 0.3962765957446808,
      "train_speed(iter/s)": 1.46704
    },
    {
      "epoch": 0.07026262799365923,
      "grad_norm": 3.7685813903808594,
      "learning_rate": 9.995128328599444e-05,
      "loss": 2.536533546447754,
      "memory(GiB)": 44.8,
      "step": 1640,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.467404
    },
    {
      "epoch": 0.07047684332290818,
      "grad_norm": 2.9384233951568604,
      "learning_rate": 9.995098582865157e-05,
      "loss": 2.468268394470215,
      "memory(GiB)": 44.8,
      "step": 1645,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.467752
    },
    {
      "epoch": 0.07069105865215715,
      "grad_norm": 6.0222039222717285,
      "learning_rate": 9.99506874663997e-05,
      "loss": 2.4725214004516602,
      "memory(GiB)": 44.8,
      "step": 1650,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.467767
    },
    {
      "epoch": 0.07090527398140611,
      "grad_norm": 3.1422464847564697,
      "learning_rate": 9.995038819924429e-05,
      "loss": 2.6452556610107423,
      "memory(GiB)": 44.8,
      "step": 1655,
      "token_acc": 0.46920821114369504,
      "train_speed(iter/s)": 1.468471
    },
    {
      "epoch": 0.07111948931065507,
      "grad_norm": 4.771244525909424,
      "learning_rate": 9.995008802719075e-05,
      "loss": 2.6549068450927735,
      "memory(GiB)": 44.8,
      "step": 1660,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.469027
    },
    {
      "epoch": 0.07133370463990403,
      "grad_norm": 3.4673192501068115,
      "learning_rate": 9.99497869502445e-05,
      "loss": 2.8690052032470703,
      "memory(GiB)": 44.8,
      "step": 1665,
      "token_acc": 0.3874538745387454,
      "train_speed(iter/s)": 1.469623
    },
    {
      "epoch": 0.071547919969153,
      "grad_norm": 6.550461292266846,
      "learning_rate": 9.9949484968411e-05,
      "loss": 2.624686431884766,
      "memory(GiB)": 44.8,
      "step": 1670,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.469485
    },
    {
      "epoch": 0.07176213529840195,
      "grad_norm": 4.609001159667969,
      "learning_rate": 9.994918208169572e-05,
      "loss": 2.849748229980469,
      "memory(GiB)": 44.8,
      "step": 1675,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.46835
    },
    {
      "epoch": 0.07197635062765091,
      "grad_norm": 3.287458896636963,
      "learning_rate": 9.994887829010416e-05,
      "loss": 2.5628742218017577,
      "memory(GiB)": 44.8,
      "step": 1680,
      "token_acc": 0.46124031007751937,
      "train_speed(iter/s)": 1.468356
    },
    {
      "epoch": 0.07219056595689988,
      "grad_norm": 4.0205078125,
      "learning_rate": 9.994857359364181e-05,
      "loss": 2.4077875137329103,
      "memory(GiB)": 44.8,
      "step": 1685,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.466875
    },
    {
      "epoch": 0.07240478128614884,
      "grad_norm": 4.003814697265625,
      "learning_rate": 9.99482679923142e-05,
      "loss": 2.4937423706054687,
      "memory(GiB)": 44.8,
      "step": 1690,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.467723
    },
    {
      "epoch": 0.0726189966153978,
      "grad_norm": 6.048070430755615,
      "learning_rate": 9.994796148612684e-05,
      "loss": 2.6457901000976562,
      "memory(GiB)": 44.8,
      "step": 1695,
      "token_acc": 0.48253968253968255,
      "train_speed(iter/s)": 1.467999
    },
    {
      "epoch": 0.07283321194464676,
      "grad_norm": 3.582834482192993,
      "learning_rate": 9.994765407508532e-05,
      "loss": 2.455926704406738,
      "memory(GiB)": 44.8,
      "step": 1700,
      "token_acc": 0.45525291828793774,
      "train_speed(iter/s)": 1.467177
    },
    {
      "epoch": 0.07304742727389572,
      "grad_norm": 4.16580867767334,
      "learning_rate": 9.99473457591952e-05,
      "loss": 2.518654441833496,
      "memory(GiB)": 44.8,
      "step": 1705,
      "token_acc": 0.4222873900293255,
      "train_speed(iter/s)": 1.466489
    },
    {
      "epoch": 0.07326164260314467,
      "grad_norm": 3.9179389476776123,
      "learning_rate": 9.994703653846203e-05,
      "loss": 2.7111934661865233,
      "memory(GiB)": 44.8,
      "step": 1710,
      "token_acc": 0.44370860927152317,
      "train_speed(iter/s)": 1.467215
    },
    {
      "epoch": 0.07347585793239364,
      "grad_norm": 3.6408209800720215,
      "learning_rate": 9.994672641289146e-05,
      "loss": 3.0178693771362304,
      "memory(GiB)": 44.8,
      "step": 1715,
      "token_acc": 0.4272151898734177,
      "train_speed(iter/s)": 1.46759
    },
    {
      "epoch": 0.0736900732616426,
      "grad_norm": 3.802363634109497,
      "learning_rate": 9.994641538248907e-05,
      "loss": 2.4939434051513674,
      "memory(GiB)": 44.8,
      "step": 1720,
      "token_acc": 0.5148514851485149,
      "train_speed(iter/s)": 1.467842
    },
    {
      "epoch": 0.07390428859089157,
      "grad_norm": 3.98805570602417,
      "learning_rate": 9.994610344726055e-05,
      "loss": 2.795812225341797,
      "memory(GiB)": 44.8,
      "step": 1725,
      "token_acc": 0.4476987447698745,
      "train_speed(iter/s)": 1.467882
    },
    {
      "epoch": 0.07411850392014052,
      "grad_norm": 4.293564796447754,
      "learning_rate": 9.994579060721147e-05,
      "loss": 2.7436058044433596,
      "memory(GiB)": 44.8,
      "step": 1730,
      "token_acc": 0.44857142857142857,
      "train_speed(iter/s)": 1.467884
    },
    {
      "epoch": 0.07433271924938949,
      "grad_norm": 4.725361347198486,
      "learning_rate": 9.994547686234755e-05,
      "loss": 2.7824867248535154,
      "memory(GiB)": 44.8,
      "step": 1735,
      "token_acc": 0.422680412371134,
      "train_speed(iter/s)": 1.468395
    },
    {
      "epoch": 0.07454693457863845,
      "grad_norm": 3.632896661758423,
      "learning_rate": 9.994516221267446e-05,
      "loss": 2.6807174682617188,
      "memory(GiB)": 44.8,
      "step": 1740,
      "token_acc": 0.46075085324232085,
      "train_speed(iter/s)": 1.468945
    },
    {
      "epoch": 0.0747611499078874,
      "grad_norm": 5.27315616607666,
      "learning_rate": 9.994484665819793e-05,
      "loss": 2.6903053283691407,
      "memory(GiB)": 44.8,
      "step": 1745,
      "token_acc": 0.44404332129963897,
      "train_speed(iter/s)": 1.468373
    },
    {
      "epoch": 0.07497536523713637,
      "grad_norm": 4.695183277130127,
      "learning_rate": 9.994453019892364e-05,
      "loss": 2.579717445373535,
      "memory(GiB)": 44.8,
      "step": 1750,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.469647
    },
    {
      "epoch": 0.07518958056638533,
      "grad_norm": 3.082949161529541,
      "learning_rate": 9.994421283485733e-05,
      "loss": 2.7353631973266603,
      "memory(GiB)": 44.8,
      "step": 1755,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.470082
    },
    {
      "epoch": 0.0754037958956343,
      "grad_norm": 4.967978000640869,
      "learning_rate": 9.994389456600474e-05,
      "loss": 2.6519132614135743,
      "memory(GiB)": 44.8,
      "step": 1760,
      "token_acc": 0.44761904761904764,
      "train_speed(iter/s)": 1.46967
    },
    {
      "epoch": 0.07561801122488325,
      "grad_norm": 4.676754474639893,
      "learning_rate": 9.994357539237166e-05,
      "loss": 2.588191604614258,
      "memory(GiB)": 44.8,
      "step": 1765,
      "token_acc": 0.46494464944649444,
      "train_speed(iter/s)": 1.46891
    },
    {
      "epoch": 0.07583222655413221,
      "grad_norm": 3.819511651992798,
      "learning_rate": 9.994325531396387e-05,
      "loss": 2.554705047607422,
      "memory(GiB)": 44.8,
      "step": 1770,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.468704
    },
    {
      "epoch": 0.07604644188338118,
      "grad_norm": 4.374409198760986,
      "learning_rate": 9.994293433078714e-05,
      "loss": 2.24488525390625,
      "memory(GiB)": 44.8,
      "step": 1775,
      "token_acc": 0.4554794520547945,
      "train_speed(iter/s)": 1.468235
    },
    {
      "epoch": 0.07626065721263013,
      "grad_norm": 3.7358691692352295,
      "learning_rate": 9.994261244284733e-05,
      "loss": 2.6257869720458986,
      "memory(GiB)": 44.8,
      "step": 1780,
      "token_acc": 0.45517241379310347,
      "train_speed(iter/s)": 1.468495
    },
    {
      "epoch": 0.0764748725418791,
      "grad_norm": 3.8033885955810547,
      "learning_rate": 9.994228965015022e-05,
      "loss": 2.5802871704101564,
      "memory(GiB)": 44.8,
      "step": 1785,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.467199
    },
    {
      "epoch": 0.07668908787112806,
      "grad_norm": 5.154775619506836,
      "learning_rate": 9.99419659527017e-05,
      "loss": 2.960002899169922,
      "memory(GiB)": 44.8,
      "step": 1790,
      "token_acc": 0.43416370106761565,
      "train_speed(iter/s)": 1.46801
    },
    {
      "epoch": 0.07690330320037701,
      "grad_norm": 2.979706048965454,
      "learning_rate": 9.994164135050761e-05,
      "loss": 2.6417831420898437,
      "memory(GiB)": 44.8,
      "step": 1795,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.469445
    },
    {
      "epoch": 0.07711751852962598,
      "grad_norm": 5.776839256286621,
      "learning_rate": 9.994131584357384e-05,
      "loss": 2.742685317993164,
      "memory(GiB)": 44.8,
      "step": 1800,
      "token_acc": 0.461038961038961,
      "train_speed(iter/s)": 1.469615
    },
    {
      "epoch": 0.07733173385887494,
      "grad_norm": 3.6369216442108154,
      "learning_rate": 9.994098943190629e-05,
      "loss": 2.7681421279907226,
      "memory(GiB)": 44.8,
      "step": 1805,
      "token_acc": 0.437125748502994,
      "train_speed(iter/s)": 1.469898
    },
    {
      "epoch": 0.07754594918812391,
      "grad_norm": 3.295574903488159,
      "learning_rate": 9.994066211551085e-05,
      "loss": 2.7584259033203127,
      "memory(GiB)": 44.8,
      "step": 1810,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.470905
    },
    {
      "epoch": 0.07776016451737286,
      "grad_norm": 4.260218143463135,
      "learning_rate": 9.994033389439348e-05,
      "loss": 2.7495498657226562,
      "memory(GiB)": 44.8,
      "step": 1815,
      "token_acc": 0.4413793103448276,
      "train_speed(iter/s)": 1.470433
    },
    {
      "epoch": 0.07797437984662182,
      "grad_norm": 4.7456374168396,
      "learning_rate": 9.994000476856011e-05,
      "loss": 2.791260528564453,
      "memory(GiB)": 44.8,
      "step": 1820,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.470185
    },
    {
      "epoch": 0.07818859517587079,
      "grad_norm": 3.4733006954193115,
      "learning_rate": 9.993967473801671e-05,
      "loss": 2.4547632217407225,
      "memory(GiB)": 44.8,
      "step": 1825,
      "token_acc": 0.49429657794676807,
      "train_speed(iter/s)": 1.47054
    },
    {
      "epoch": 0.07840281050511974,
      "grad_norm": 4.259070873260498,
      "learning_rate": 9.993934380276926e-05,
      "loss": 2.639309310913086,
      "memory(GiB)": 44.8,
      "step": 1830,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.471465
    },
    {
      "epoch": 0.0786170258343687,
      "grad_norm": 3.966062545776367,
      "learning_rate": 9.993901196282374e-05,
      "loss": 2.3673242568969726,
      "memory(GiB)": 44.8,
      "step": 1835,
      "token_acc": 0.5023474178403756,
      "train_speed(iter/s)": 1.47104
    },
    {
      "epoch": 0.07883124116361767,
      "grad_norm": 5.579440116882324,
      "learning_rate": 9.993867921818619e-05,
      "loss": 2.858770179748535,
      "memory(GiB)": 44.8,
      "step": 1840,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.470998
    },
    {
      "epoch": 0.07904545649286664,
      "grad_norm": 3.153552532196045,
      "learning_rate": 9.993834556886259e-05,
      "loss": 2.3209178924560545,
      "memory(GiB)": 44.8,
      "step": 1845,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.47136
    },
    {
      "epoch": 0.07925967182211559,
      "grad_norm": 2.9614076614379883,
      "learning_rate": 9.993801101485903e-05,
      "loss": 2.7119884490966797,
      "memory(GiB)": 44.8,
      "step": 1850,
      "token_acc": 0.44984802431610943,
      "train_speed(iter/s)": 1.471664
    },
    {
      "epoch": 0.07947388715136455,
      "grad_norm": 4.343069553375244,
      "learning_rate": 9.993767555618157e-05,
      "loss": 2.759410858154297,
      "memory(GiB)": 44.8,
      "step": 1855,
      "token_acc": 0.4502923976608187,
      "train_speed(iter/s)": 1.471036
    },
    {
      "epoch": 0.07968810248061352,
      "grad_norm": 3.1520256996154785,
      "learning_rate": 9.993733919283624e-05,
      "loss": 3.0197704315185545,
      "memory(GiB)": 44.8,
      "step": 1860,
      "token_acc": 0.4014336917562724,
      "train_speed(iter/s)": 1.471265
    },
    {
      "epoch": 0.07990231780986247,
      "grad_norm": 3.6374616622924805,
      "learning_rate": 9.993700192482918e-05,
      "loss": 2.8920530319213866,
      "memory(GiB)": 44.8,
      "step": 1865,
      "token_acc": 0.4244791666666667,
      "train_speed(iter/s)": 1.471729
    },
    {
      "epoch": 0.08011653313911143,
      "grad_norm": 2.874727964401245,
      "learning_rate": 9.99366637521665e-05,
      "loss": 2.738356018066406,
      "memory(GiB)": 44.8,
      "step": 1870,
      "token_acc": 0.4391691394658754,
      "train_speed(iter/s)": 1.472679
    },
    {
      "epoch": 0.0803307484683604,
      "grad_norm": 2.983149290084839,
      "learning_rate": 9.99363246748543e-05,
      "loss": 2.816168785095215,
      "memory(GiB)": 44.8,
      "step": 1875,
      "token_acc": 0.4426229508196721,
      "train_speed(iter/s)": 1.472957
    },
    {
      "epoch": 0.08054496379760935,
      "grad_norm": 2.8804805278778076,
      "learning_rate": 9.993598469289874e-05,
      "loss": 2.5867034912109377,
      "memory(GiB)": 44.8,
      "step": 1880,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.47399
    },
    {
      "epoch": 0.08075917912685832,
      "grad_norm": 3.374390125274658,
      "learning_rate": 9.993564380630595e-05,
      "loss": 2.8110849380493166,
      "memory(GiB)": 44.8,
      "step": 1885,
      "token_acc": 0.45110410094637227,
      "train_speed(iter/s)": 1.474029
    },
    {
      "epoch": 0.08097339445610728,
      "grad_norm": 3.384381055831909,
      "learning_rate": 9.993530201508216e-05,
      "loss": 2.699570083618164,
      "memory(GiB)": 44.8,
      "step": 1890,
      "token_acc": 0.48125,
      "train_speed(iter/s)": 1.473377
    },
    {
      "epoch": 0.08118760978535625,
      "grad_norm": 6.725127696990967,
      "learning_rate": 9.993495931923352e-05,
      "loss": 2.5107748031616213,
      "memory(GiB)": 44.8,
      "step": 1895,
      "token_acc": 0.46037735849056605,
      "train_speed(iter/s)": 1.472904
    },
    {
      "epoch": 0.0814018251146052,
      "grad_norm": 3.8275153636932373,
      "learning_rate": 9.993461571876624e-05,
      "loss": 2.467954635620117,
      "memory(GiB)": 44.8,
      "step": 1900,
      "token_acc": 0.44594594594594594,
      "train_speed(iter/s)": 1.472298
    },
    {
      "epoch": 0.08161604044385416,
      "grad_norm": 4.472103118896484,
      "learning_rate": 9.993427121368656e-05,
      "loss": 2.788699913024902,
      "memory(GiB)": 44.8,
      "step": 1905,
      "token_acc": 0.4506578947368421,
      "train_speed(iter/s)": 1.471735
    },
    {
      "epoch": 0.08183025577310313,
      "grad_norm": 4.863548278808594,
      "learning_rate": 9.993392580400072e-05,
      "loss": 2.5571044921875,
      "memory(GiB)": 44.8,
      "step": 1910,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.471922
    },
    {
      "epoch": 0.08204447110235208,
      "grad_norm": 5.535677909851074,
      "learning_rate": 9.993357948971496e-05,
      "loss": 2.714681053161621,
      "memory(GiB)": 44.8,
      "step": 1915,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.472022
    },
    {
      "epoch": 0.08225868643160104,
      "grad_norm": 3.3805017471313477,
      "learning_rate": 9.993323227083557e-05,
      "loss": 2.60046443939209,
      "memory(GiB)": 44.8,
      "step": 1920,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.473066
    },
    {
      "epoch": 0.08247290176085001,
      "grad_norm": 3.1437110900878906,
      "learning_rate": 9.993288414736885e-05,
      "loss": 2.686944770812988,
      "memory(GiB)": 44.8,
      "step": 1925,
      "token_acc": 0.48742138364779874,
      "train_speed(iter/s)": 1.472615
    },
    {
      "epoch": 0.08268711709009897,
      "grad_norm": 6.048345565795898,
      "learning_rate": 9.993253511932108e-05,
      "loss": 2.567666244506836,
      "memory(GiB)": 44.8,
      "step": 1930,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.473231
    },
    {
      "epoch": 0.08290133241934793,
      "grad_norm": 4.505509853363037,
      "learning_rate": 9.99321851866986e-05,
      "loss": 2.66304931640625,
      "memory(GiB)": 44.8,
      "step": 1935,
      "token_acc": 0.4354243542435424,
      "train_speed(iter/s)": 1.473824
    },
    {
      "epoch": 0.08311554774859689,
      "grad_norm": 3.6984739303588867,
      "learning_rate": 9.993183434950774e-05,
      "loss": 2.5534067153930664,
      "memory(GiB)": 44.8,
      "step": 1940,
      "token_acc": 0.4983922829581994,
      "train_speed(iter/s)": 1.473403
    },
    {
      "epoch": 0.08332976307784586,
      "grad_norm": 3.472925901412964,
      "learning_rate": 9.993148260775487e-05,
      "loss": 2.303940200805664,
      "memory(GiB)": 44.8,
      "step": 1945,
      "token_acc": 0.516728624535316,
      "train_speed(iter/s)": 1.473391
    },
    {
      "epoch": 0.08354397840709481,
      "grad_norm": 3.518807888031006,
      "learning_rate": 9.993112996144634e-05,
      "loss": 2.582549285888672,
      "memory(GiB)": 44.8,
      "step": 1950,
      "token_acc": 0.47079037800687284,
      "train_speed(iter/s)": 1.473647
    },
    {
      "epoch": 0.08375819373634377,
      "grad_norm": 2.8743302822113037,
      "learning_rate": 9.993077641058856e-05,
      "loss": 2.4542316436767577,
      "memory(GiB)": 44.8,
      "step": 1955,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.473027
    },
    {
      "epoch": 0.08397240906559274,
      "grad_norm": 5.637447357177734,
      "learning_rate": 9.993042195518793e-05,
      "loss": 2.5404361724853515,
      "memory(GiB)": 44.8,
      "step": 1960,
      "token_acc": 0.45774647887323944,
      "train_speed(iter/s)": 1.473449
    },
    {
      "epoch": 0.08418662439484169,
      "grad_norm": 3.8969013690948486,
      "learning_rate": 9.993006659525087e-05,
      "loss": 2.9124237060546876,
      "memory(GiB)": 44.8,
      "step": 1965,
      "token_acc": 0.4014336917562724,
      "train_speed(iter/s)": 1.473957
    },
    {
      "epoch": 0.08440083972409065,
      "grad_norm": 4.842186450958252,
      "learning_rate": 9.99297103307838e-05,
      "loss": 2.7313570022583007,
      "memory(GiB)": 44.8,
      "step": 1970,
      "token_acc": 0.46175637393767704,
      "train_speed(iter/s)": 1.47275
    },
    {
      "epoch": 0.08461505505333962,
      "grad_norm": 3.2330164909362793,
      "learning_rate": 9.99293531617932e-05,
      "loss": 2.4571990966796875,
      "memory(GiB)": 44.8,
      "step": 1975,
      "token_acc": 0.4872611464968153,
      "train_speed(iter/s)": 1.47132
    },
    {
      "epoch": 0.08482927038258858,
      "grad_norm": 3.849078416824341,
      "learning_rate": 9.992899508828553e-05,
      "loss": 2.63256721496582,
      "memory(GiB)": 44.8,
      "step": 1980,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.471193
    },
    {
      "epoch": 0.08504348571183754,
      "grad_norm": 5.2319183349609375,
      "learning_rate": 9.992863611026725e-05,
      "loss": 2.579254722595215,
      "memory(GiB)": 44.8,
      "step": 1985,
      "token_acc": 0.45084745762711864,
      "train_speed(iter/s)": 1.470934
    },
    {
      "epoch": 0.0852577010410865,
      "grad_norm": 4.7803120613098145,
      "learning_rate": 9.992827622774494e-05,
      "loss": 2.7317478179931642,
      "memory(GiB)": 44.8,
      "step": 1990,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.471228
    },
    {
      "epoch": 0.08547191637033547,
      "grad_norm": 3.165527105331421,
      "learning_rate": 9.992791544072504e-05,
      "loss": 2.8859058380126954,
      "memory(GiB)": 44.8,
      "step": 1995,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.471087
    },
    {
      "epoch": 0.08568613169958442,
      "grad_norm": 4.092401027679443,
      "learning_rate": 9.992755374921412e-05,
      "loss": 2.788203811645508,
      "memory(GiB)": 44.8,
      "step": 2000,
      "token_acc": 0.4258064516129032,
      "train_speed(iter/s)": 1.471473
    },
    {
      "epoch": 0.08568613169958442,
      "eval_loss": 2.2102930545806885,
      "eval_runtime": 14.1978,
      "eval_samples_per_second": 7.043,
      "eval_steps_per_second": 7.043,
      "eval_token_acc": 0.4676753782668501,
      "step": 2000
    },
    {
      "epoch": 0.08590034702883338,
      "grad_norm": 5.016899585723877,
      "learning_rate": 9.992719115321872e-05,
      "loss": 2.5726345062255858,
      "memory(GiB)": 44.8,
      "step": 2005,
      "token_acc": 0.4597918637653737,
      "train_speed(iter/s)": 1.455839
    },
    {
      "epoch": 0.08611456235808235,
      "grad_norm": 3.140343427658081,
      "learning_rate": 9.992682765274543e-05,
      "loss": 3.0107297897338867,
      "memory(GiB)": 44.8,
      "step": 2010,
      "token_acc": 0.4467455621301775,
      "train_speed(iter/s)": 1.455615
    },
    {
      "epoch": 0.08632877768733131,
      "grad_norm": 5.080636024475098,
      "learning_rate": 9.992646324780082e-05,
      "loss": 2.307963752746582,
      "memory(GiB)": 44.8,
      "step": 2015,
      "token_acc": 0.5245283018867924,
      "train_speed(iter/s)": 1.455992
    },
    {
      "epoch": 0.08654299301658026,
      "grad_norm": 3.719008684158325,
      "learning_rate": 9.992609793839148e-05,
      "loss": 2.5544200897216798,
      "memory(GiB)": 44.8,
      "step": 2020,
      "token_acc": 0.458955223880597,
      "train_speed(iter/s)": 1.455796
    },
    {
      "epoch": 0.08675720834582923,
      "grad_norm": 4.171456813812256,
      "learning_rate": 9.992573172452406e-05,
      "loss": 2.617792510986328,
      "memory(GiB)": 44.8,
      "step": 2025,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.456087
    },
    {
      "epoch": 0.0869714236750782,
      "grad_norm": 4.259059429168701,
      "learning_rate": 9.992536460620516e-05,
      "loss": 2.66180477142334,
      "memory(GiB)": 44.8,
      "step": 2030,
      "token_acc": 0.4826254826254826,
      "train_speed(iter/s)": 1.45579
    },
    {
      "epoch": 0.08718563900432715,
      "grad_norm": 3.160526990890503,
      "learning_rate": 9.992499658344145e-05,
      "loss": 2.69201545715332,
      "memory(GiB)": 44.8,
      "step": 2035,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.456549
    },
    {
      "epoch": 0.08739985433357611,
      "grad_norm": 4.474733352661133,
      "learning_rate": 9.99246276562396e-05,
      "loss": 2.8203947067260744,
      "memory(GiB)": 44.8,
      "step": 2040,
      "token_acc": 0.40217391304347827,
      "train_speed(iter/s)": 1.457013
    },
    {
      "epoch": 0.08761406966282508,
      "grad_norm": 3.573944330215454,
      "learning_rate": 9.992425782460628e-05,
      "loss": 2.5302507400512697,
      "memory(GiB)": 44.8,
      "step": 2045,
      "token_acc": 0.43666666666666665,
      "train_speed(iter/s)": 1.457445
    },
    {
      "epoch": 0.08782828499207403,
      "grad_norm": 5.393106460571289,
      "learning_rate": 9.99238870885482e-05,
      "loss": 2.9964015960693358,
      "memory(GiB)": 44.8,
      "step": 2050,
      "token_acc": 0.4199288256227758,
      "train_speed(iter/s)": 1.458832
    },
    {
      "epoch": 0.08804250032132299,
      "grad_norm": 2.755122423171997,
      "learning_rate": 9.992351544807208e-05,
      "loss": 2.91259708404541,
      "memory(GiB)": 44.8,
      "step": 2055,
      "token_acc": 0.4180064308681672,
      "train_speed(iter/s)": 1.459263
    },
    {
      "epoch": 0.08825671565057196,
      "grad_norm": 3.6033883094787598,
      "learning_rate": 9.992314290318465e-05,
      "loss": 2.6889392852783205,
      "memory(GiB)": 44.8,
      "step": 2060,
      "token_acc": 0.41776315789473684,
      "train_speed(iter/s)": 1.459457
    },
    {
      "epoch": 0.08847093097982092,
      "grad_norm": 3.053887128829956,
      "learning_rate": 9.992276945389263e-05,
      "loss": 2.7946678161621095,
      "memory(GiB)": 44.8,
      "step": 2065,
      "token_acc": 0.4340175953079179,
      "train_speed(iter/s)": 1.459748
    },
    {
      "epoch": 0.08868514630906987,
      "grad_norm": 6.401619911193848,
      "learning_rate": 9.992239510020282e-05,
      "loss": 2.6602827072143556,
      "memory(GiB)": 44.8,
      "step": 2070,
      "token_acc": 0.45038167938931295,
      "train_speed(iter/s)": 1.4601
    },
    {
      "epoch": 0.08889936163831884,
      "grad_norm": 4.626623630523682,
      "learning_rate": 9.9922019842122e-05,
      "loss": 2.8021312713623048,
      "memory(GiB)": 44.8,
      "step": 2075,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.460213
    },
    {
      "epoch": 0.0891135769675678,
      "grad_norm": 3.489022970199585,
      "learning_rate": 9.992164367965696e-05,
      "loss": 2.7191648483276367,
      "memory(GiB)": 44.8,
      "step": 2080,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.460497
    },
    {
      "epoch": 0.08932779229681675,
      "grad_norm": 3.8433265686035156,
      "learning_rate": 9.992126661281452e-05,
      "loss": 2.543224906921387,
      "memory(GiB)": 44.8,
      "step": 2085,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.460068
    },
    {
      "epoch": 0.08954200762606572,
      "grad_norm": 3.1771106719970703,
      "learning_rate": 9.99208886416015e-05,
      "loss": 2.6399940490722655,
      "memory(GiB)": 44.8,
      "step": 2090,
      "token_acc": 0.428125,
      "train_speed(iter/s)": 1.460713
    },
    {
      "epoch": 0.08975622295531469,
      "grad_norm": 3.5798282623291016,
      "learning_rate": 9.992050976602474e-05,
      "loss": 2.520054244995117,
      "memory(GiB)": 44.8,
      "step": 2095,
      "token_acc": 0.4448051948051948,
      "train_speed(iter/s)": 1.460751
    },
    {
      "epoch": 0.08997043828456365,
      "grad_norm": 3.968090772628784,
      "learning_rate": 9.992012998609112e-05,
      "loss": 2.8136629104614257,
      "memory(GiB)": 44.8,
      "step": 2100,
      "token_acc": 0.4470198675496689,
      "train_speed(iter/s)": 1.460699
    },
    {
      "epoch": 0.0901846536138126,
      "grad_norm": 3.531860828399658,
      "learning_rate": 9.991974930180752e-05,
      "loss": 2.687778091430664,
      "memory(GiB)": 44.8,
      "step": 2105,
      "token_acc": 0.46303501945525294,
      "train_speed(iter/s)": 1.461212
    },
    {
      "epoch": 0.09039886894306157,
      "grad_norm": 2.9376018047332764,
      "learning_rate": 9.991936771318083e-05,
      "loss": 2.5948768615722657,
      "memory(GiB)": 44.8,
      "step": 2110,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.46008
    },
    {
      "epoch": 0.09061308427231053,
      "grad_norm": 5.6689229011535645,
      "learning_rate": 9.991898522021798e-05,
      "loss": 2.7493515014648438,
      "memory(GiB)": 44.8,
      "step": 2115,
      "token_acc": 0.45390070921985815,
      "train_speed(iter/s)": 1.460445
    },
    {
      "epoch": 0.09082729960155948,
      "grad_norm": 3.887681245803833,
      "learning_rate": 9.991860182292587e-05,
      "loss": 2.595868682861328,
      "memory(GiB)": 44.8,
      "step": 2120,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.460787
    },
    {
      "epoch": 0.09104151493080845,
      "grad_norm": 3.7782750129699707,
      "learning_rate": 9.991821752131146e-05,
      "loss": 2.73706169128418,
      "memory(GiB)": 44.8,
      "step": 2125,
      "token_acc": 0.4485049833887043,
      "train_speed(iter/s)": 1.460954
    },
    {
      "epoch": 0.09125573026005741,
      "grad_norm": 4.064642906188965,
      "learning_rate": 9.991783231538172e-05,
      "loss": 2.8414638519287108,
      "memory(GiB)": 44.8,
      "step": 2130,
      "token_acc": 0.41946308724832215,
      "train_speed(iter/s)": 1.460693
    },
    {
      "epoch": 0.09146994558930636,
      "grad_norm": 4.133368015289307,
      "learning_rate": 9.991744620514363e-05,
      "loss": 2.9908676147460938,
      "memory(GiB)": 44.8,
      "step": 2135,
      "token_acc": 0.392,
      "train_speed(iter/s)": 1.459536
    },
    {
      "epoch": 0.09168416091855533,
      "grad_norm": 4.057086944580078,
      "learning_rate": 9.991705919060416e-05,
      "loss": 2.6876079559326174,
      "memory(GiB)": 44.8,
      "step": 2140,
      "token_acc": 0.4453125,
      "train_speed(iter/s)": 1.459825
    },
    {
      "epoch": 0.0918983762478043,
      "grad_norm": 5.1124043464660645,
      "learning_rate": 9.991667127177033e-05,
      "loss": 2.4993526458740236,
      "memory(GiB)": 44.8,
      "step": 2145,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.459294
    },
    {
      "epoch": 0.09211259157705326,
      "grad_norm": 4.4299397468566895,
      "learning_rate": 9.99162824486492e-05,
      "loss": 2.6979110717773436,
      "memory(GiB)": 44.8,
      "step": 2150,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.459595
    },
    {
      "epoch": 0.09232680690630221,
      "grad_norm": 4.406503677368164,
      "learning_rate": 9.991589272124778e-05,
      "loss": 2.6013330459594726,
      "memory(GiB)": 44.8,
      "step": 2155,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.458888
    },
    {
      "epoch": 0.09254102223555118,
      "grad_norm": 5.777492523193359,
      "learning_rate": 9.991550208957312e-05,
      "loss": 2.6056400299072267,
      "memory(GiB)": 44.8,
      "step": 2160,
      "token_acc": 0.42024539877300615,
      "train_speed(iter/s)": 1.459472
    },
    {
      "epoch": 0.09275523756480014,
      "grad_norm": 4.015589237213135,
      "learning_rate": 9.991511055363232e-05,
      "loss": 2.3596290588378905,
      "memory(GiB)": 44.8,
      "step": 2165,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.459388
    },
    {
      "epoch": 0.09296945289404909,
      "grad_norm": 3.3259499073028564,
      "learning_rate": 9.991471811343248e-05,
      "loss": 2.6802696228027343,
      "memory(GiB)": 44.8,
      "step": 2170,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.459627
    },
    {
      "epoch": 0.09318366822329806,
      "grad_norm": 3.664127826690674,
      "learning_rate": 9.991432476898069e-05,
      "loss": 2.6583927154541014,
      "memory(GiB)": 44.8,
      "step": 2175,
      "token_acc": 0.48534201954397393,
      "train_speed(iter/s)": 1.459292
    },
    {
      "epoch": 0.09339788355254702,
      "grad_norm": 6.050678253173828,
      "learning_rate": 9.991393052028408e-05,
      "loss": 2.6330432891845703,
      "memory(GiB)": 44.8,
      "step": 2180,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.458408
    },
    {
      "epoch": 0.09361209888179599,
      "grad_norm": 5.467209815979004,
      "learning_rate": 9.991353536734981e-05,
      "loss": 2.928238868713379,
      "memory(GiB)": 44.8,
      "step": 2185,
      "token_acc": 0.39935064935064934,
      "train_speed(iter/s)": 1.457615
    },
    {
      "epoch": 0.09382631421104494,
      "grad_norm": 3.2603909969329834,
      "learning_rate": 9.991313931018503e-05,
      "loss": 2.7299537658691406,
      "memory(GiB)": 44.8,
      "step": 2190,
      "token_acc": 0.4431137724550898,
      "train_speed(iter/s)": 1.457003
    },
    {
      "epoch": 0.0940405295402939,
      "grad_norm": 3.839003086090088,
      "learning_rate": 9.991274234879688e-05,
      "loss": 2.3807676315307615,
      "memory(GiB)": 44.8,
      "step": 2195,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.457335
    },
    {
      "epoch": 0.09425474486954287,
      "grad_norm": 3.2841320037841797,
      "learning_rate": 9.99123444831926e-05,
      "loss": 2.8565706253051757,
      "memory(GiB)": 44.8,
      "step": 2200,
      "token_acc": 0.4325153374233129,
      "train_speed(iter/s)": 1.457065
    },
    {
      "epoch": 0.09446896019879182,
      "grad_norm": 4.429266929626465,
      "learning_rate": 9.991194571337937e-05,
      "loss": 2.6315616607666015,
      "memory(GiB)": 44.8,
      "step": 2205,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.457396
    },
    {
      "epoch": 0.09468317552804079,
      "grad_norm": 3.7662506103515625,
      "learning_rate": 9.991154603936443e-05,
      "loss": 2.3513725280761717,
      "memory(GiB)": 44.8,
      "step": 2210,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.457128
    },
    {
      "epoch": 0.09489739085728975,
      "grad_norm": 4.712777137756348,
      "learning_rate": 9.991114546115502e-05,
      "loss": 2.537898635864258,
      "memory(GiB)": 44.8,
      "step": 2215,
      "token_acc": 0.44805194805194803,
      "train_speed(iter/s)": 1.45697
    },
    {
      "epoch": 0.0951116061865387,
      "grad_norm": 4.573584079742432,
      "learning_rate": 9.991074397875836e-05,
      "loss": 2.9043529510498045,
      "memory(GiB)": 44.8,
      "step": 2220,
      "token_acc": 0.4107142857142857,
      "train_speed(iter/s)": 1.456899
    },
    {
      "epoch": 0.09532582151578767,
      "grad_norm": 5.321566581726074,
      "learning_rate": 9.991034159218178e-05,
      "loss": 2.7422548294067384,
      "memory(GiB)": 44.8,
      "step": 2225,
      "token_acc": 0.4697674418604651,
      "train_speed(iter/s)": 1.456959
    },
    {
      "epoch": 0.09554003684503663,
      "grad_norm": 3.833298683166504,
      "learning_rate": 9.990993830143253e-05,
      "loss": 2.548801040649414,
      "memory(GiB)": 44.8,
      "step": 2230,
      "token_acc": 0.5061224489795918,
      "train_speed(iter/s)": 1.456937
    },
    {
      "epoch": 0.0957542521742856,
      "grad_norm": 5.392580986022949,
      "learning_rate": 9.990953410651793e-05,
      "loss": 3.013517379760742,
      "memory(GiB)": 44.8,
      "step": 2235,
      "token_acc": 0.38515901060070673,
      "train_speed(iter/s)": 1.456928
    },
    {
      "epoch": 0.09596846750353455,
      "grad_norm": 3.42513370513916,
      "learning_rate": 9.990912900744529e-05,
      "loss": 2.3812801361083986,
      "memory(GiB)": 44.8,
      "step": 2240,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.45778
    },
    {
      "epoch": 0.09618268283278351,
      "grad_norm": 4.226132869720459,
      "learning_rate": 9.990872300422198e-05,
      "loss": 2.447560119628906,
      "memory(GiB)": 44.8,
      "step": 2245,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.457581
    },
    {
      "epoch": 0.09639689816203248,
      "grad_norm": 4.196174621582031,
      "learning_rate": 9.990831609685532e-05,
      "loss": 2.4533769607543947,
      "memory(GiB)": 44.8,
      "step": 2250,
      "token_acc": 0.47278911564625853,
      "train_speed(iter/s)": 1.457717
    },
    {
      "epoch": 0.09661111349128143,
      "grad_norm": 3.42431640625,
      "learning_rate": 9.990790828535271e-05,
      "loss": 2.4951969146728517,
      "memory(GiB)": 44.8,
      "step": 2255,
      "token_acc": 0.474025974025974,
      "train_speed(iter/s)": 1.456446
    },
    {
      "epoch": 0.0968253288205304,
      "grad_norm": 4.205296039581299,
      "learning_rate": 9.990749956972152e-05,
      "loss": 2.945000648498535,
      "memory(GiB)": 44.8,
      "step": 2260,
      "token_acc": 0.4,
      "train_speed(iter/s)": 1.456896
    },
    {
      "epoch": 0.09703954414977936,
      "grad_norm": 3.1933600902557373,
      "learning_rate": 9.990708994996916e-05,
      "loss": 2.9696237564086916,
      "memory(GiB)": 44.8,
      "step": 2265,
      "token_acc": 0.43455497382198954,
      "train_speed(iter/s)": 1.457204
    },
    {
      "epoch": 0.09725375947902833,
      "grad_norm": 2.845933437347412,
      "learning_rate": 9.990667942610303e-05,
      "loss": 2.838313674926758,
      "memory(GiB)": 44.8,
      "step": 2270,
      "token_acc": 0.45244956772334294,
      "train_speed(iter/s)": 1.456595
    },
    {
      "epoch": 0.09746797480827728,
      "grad_norm": 3.597959518432617,
      "learning_rate": 9.990626799813061e-05,
      "loss": 2.954367446899414,
      "memory(GiB)": 44.8,
      "step": 2275,
      "token_acc": 0.40808823529411764,
      "train_speed(iter/s)": 1.456928
    },
    {
      "epoch": 0.09768219013752624,
      "grad_norm": 4.257412910461426,
      "learning_rate": 9.990585566605932e-05,
      "loss": 2.70975341796875,
      "memory(GiB)": 44.8,
      "step": 2280,
      "token_acc": 0.48639455782312924,
      "train_speed(iter/s)": 1.456894
    },
    {
      "epoch": 0.09789640546677521,
      "grad_norm": 5.009130001068115,
      "learning_rate": 9.990544242989663e-05,
      "loss": 2.705342483520508,
      "memory(GiB)": 44.8,
      "step": 2285,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.457066
    },
    {
      "epoch": 0.09811062079602416,
      "grad_norm": 3.4967539310455322,
      "learning_rate": 9.990502828965005e-05,
      "loss": 2.5337432861328124,
      "memory(GiB)": 44.8,
      "step": 2290,
      "token_acc": 0.4910394265232975,
      "train_speed(iter/s)": 1.457007
    },
    {
      "epoch": 0.09832483612527312,
      "grad_norm": 3.3518471717834473,
      "learning_rate": 9.990461324532705e-05,
      "loss": 2.6714311599731446,
      "memory(GiB)": 44.8,
      "step": 2295,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.456389
    },
    {
      "epoch": 0.09853905145452209,
      "grad_norm": 4.614869594573975,
      "learning_rate": 9.990419729693519e-05,
      "loss": 2.8144582748413085,
      "memory(GiB)": 44.8,
      "step": 2300,
      "token_acc": 0.4266666666666667,
      "train_speed(iter/s)": 1.456147
    },
    {
      "epoch": 0.09875326678377104,
      "grad_norm": 3.9386801719665527,
      "learning_rate": 9.990378044448197e-05,
      "loss": 2.8945823669433595,
      "memory(GiB)": 44.8,
      "step": 2305,
      "token_acc": 0.445578231292517,
      "train_speed(iter/s)": 1.456448
    },
    {
      "epoch": 0.09896748211302,
      "grad_norm": 4.9402570724487305,
      "learning_rate": 9.990336268797494e-05,
      "loss": 2.9529052734375,
      "memory(GiB)": 44.8,
      "step": 2310,
      "token_acc": 0.4432624113475177,
      "train_speed(iter/s)": 1.456636
    },
    {
      "epoch": 0.09918169744226897,
      "grad_norm": 2.6104836463928223,
      "learning_rate": 9.990294402742171e-05,
      "loss": 2.6046127319335937,
      "memory(GiB)": 44.8,
      "step": 2315,
      "token_acc": 0.47592067988668557,
      "train_speed(iter/s)": 1.456901
    },
    {
      "epoch": 0.09939591277151794,
      "grad_norm": 5.195140361785889,
      "learning_rate": 9.990252446282982e-05,
      "loss": 2.6561588287353515,
      "memory(GiB)": 44.8,
      "step": 2320,
      "token_acc": 0.45136186770428016,
      "train_speed(iter/s)": 1.457078
    },
    {
      "epoch": 0.09961012810076689,
      "grad_norm": 3.7514607906341553,
      "learning_rate": 9.990210399420688e-05,
      "loss": 2.476618766784668,
      "memory(GiB)": 44.8,
      "step": 2325,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.457343
    },
    {
      "epoch": 0.09982434343001585,
      "grad_norm": 3.4020187854766846,
      "learning_rate": 9.990168262156052e-05,
      "loss": 2.8265119552612306,
      "memory(GiB)": 44.8,
      "step": 2330,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.457231
    },
    {
      "epoch": 0.10003855875926482,
      "grad_norm": 2.6397881507873535,
      "learning_rate": 9.990126034489837e-05,
      "loss": 2.339085006713867,
      "memory(GiB)": 44.8,
      "step": 2335,
      "token_acc": 0.5032258064516129,
      "train_speed(iter/s)": 1.45785
    },
    {
      "epoch": 0.10025277408851377,
      "grad_norm": 3.193979024887085,
      "learning_rate": 9.990083716422808e-05,
      "loss": 2.9215423583984377,
      "memory(GiB)": 44.8,
      "step": 2340,
      "token_acc": 0.4077669902912621,
      "train_speed(iter/s)": 1.457972
    },
    {
      "epoch": 0.10046698941776273,
      "grad_norm": 3.46121883392334,
      "learning_rate": 9.990041307955731e-05,
      "loss": 2.833090400695801,
      "memory(GiB)": 44.8,
      "step": 2345,
      "token_acc": 0.43962848297213625,
      "train_speed(iter/s)": 1.457646
    },
    {
      "epoch": 0.1006812047470117,
      "grad_norm": 3.982710599899292,
      "learning_rate": 9.989998809089376e-05,
      "loss": 2.6022457122802733,
      "memory(GiB)": 44.8,
      "step": 2350,
      "token_acc": 0.4844961240310077,
      "train_speed(iter/s)": 1.457811
    },
    {
      "epoch": 0.10089542007626066,
      "grad_norm": 4.487769603729248,
      "learning_rate": 9.98995621982451e-05,
      "loss": 2.5113431930541994,
      "memory(GiB)": 44.8,
      "step": 2355,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.457838
    },
    {
      "epoch": 0.10110963540550962,
      "grad_norm": 4.391627788543701,
      "learning_rate": 9.989913540161906e-05,
      "loss": 2.76812801361084,
      "memory(GiB)": 44.8,
      "step": 2360,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.458005
    },
    {
      "epoch": 0.10132385073475858,
      "grad_norm": 4.283302307128906,
      "learning_rate": 9.989870770102339e-05,
      "loss": 2.2327449798583983,
      "memory(GiB)": 44.8,
      "step": 2365,
      "token_acc": 0.5320754716981132,
      "train_speed(iter/s)": 1.457033
    },
    {
      "epoch": 0.10153806606400755,
      "grad_norm": 2.9694247245788574,
      "learning_rate": 9.989827909646581e-05,
      "loss": 2.6974130630493165,
      "memory(GiB)": 44.8,
      "step": 2370,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.456564
    },
    {
      "epoch": 0.1017522813932565,
      "grad_norm": 3.814884662628174,
      "learning_rate": 9.989784958795409e-05,
      "loss": 2.5132705688476564,
      "memory(GiB)": 44.8,
      "step": 2375,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.457108
    },
    {
      "epoch": 0.10196649672250546,
      "grad_norm": 6.999202251434326,
      "learning_rate": 9.989741917549603e-05,
      "loss": 2.3899309158325197,
      "memory(GiB)": 44.8,
      "step": 2380,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.457502
    },
    {
      "epoch": 0.10218071205175443,
      "grad_norm": 3.5750794410705566,
      "learning_rate": 9.98969878590994e-05,
      "loss": 2.8229116439819335,
      "memory(GiB)": 44.8,
      "step": 2385,
      "token_acc": 0.45819397993311034,
      "train_speed(iter/s)": 1.457884
    },
    {
      "epoch": 0.10239492738100338,
      "grad_norm": 4.229066848754883,
      "learning_rate": 9.989655563877203e-05,
      "loss": 2.663188934326172,
      "memory(GiB)": 44.8,
      "step": 2390,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.458034
    },
    {
      "epoch": 0.10260914271025234,
      "grad_norm": 4.6534857749938965,
      "learning_rate": 9.989612251452176e-05,
      "loss": 2.7501869201660156,
      "memory(GiB)": 44.8,
      "step": 2395,
      "token_acc": 0.4290322580645161,
      "train_speed(iter/s)": 1.458171
    },
    {
      "epoch": 0.10282335803950131,
      "grad_norm": 4.1199822425842285,
      "learning_rate": 9.98956884863564e-05,
      "loss": 2.644135665893555,
      "memory(GiB)": 44.8,
      "step": 2400,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.458413
    },
    {
      "epoch": 0.10303757336875027,
      "grad_norm": 4.341559410095215,
      "learning_rate": 9.989525355428386e-05,
      "loss": 2.3370315551757814,
      "memory(GiB)": 44.8,
      "step": 2405,
      "token_acc": 0.4797687861271676,
      "train_speed(iter/s)": 1.458611
    },
    {
      "epoch": 0.10325178869799922,
      "grad_norm": 2.8268990516662598,
      "learning_rate": 9.9894817718312e-05,
      "loss": 2.4584678649902343,
      "memory(GiB)": 44.8,
      "step": 2410,
      "token_acc": 0.5292207792207793,
      "train_speed(iter/s)": 1.458368
    },
    {
      "epoch": 0.10346600402724819,
      "grad_norm": 3.776341199874878,
      "learning_rate": 9.98943809784487e-05,
      "loss": 2.5440006256103516,
      "memory(GiB)": 44.8,
      "step": 2415,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.458954
    },
    {
      "epoch": 0.10368021935649716,
      "grad_norm": 5.095711708068848,
      "learning_rate": 9.989394333470188e-05,
      "loss": 3.028711700439453,
      "memory(GiB)": 44.8,
      "step": 2420,
      "token_acc": 0.4163568773234201,
      "train_speed(iter/s)": 1.459408
    },
    {
      "epoch": 0.1038944346857461,
      "grad_norm": 3.311483144760132,
      "learning_rate": 9.989350478707949e-05,
      "loss": 2.6387834548950195,
      "memory(GiB)": 44.8,
      "step": 2425,
      "token_acc": 0.4664804469273743,
      "train_speed(iter/s)": 1.460114
    },
    {
      "epoch": 0.10410865001499507,
      "grad_norm": 2.6038498878479004,
      "learning_rate": 9.989306533558944e-05,
      "loss": 2.5682758331298827,
      "memory(GiB)": 44.8,
      "step": 2430,
      "token_acc": 0.45671641791044776,
      "train_speed(iter/s)": 1.460074
    },
    {
      "epoch": 0.10432286534424404,
      "grad_norm": 3.486302614212036,
      "learning_rate": 9.98926249802397e-05,
      "loss": 2.5280033111572267,
      "memory(GiB)": 44.8,
      "step": 2435,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.459851
    },
    {
      "epoch": 0.104537080673493,
      "grad_norm": 4.014797210693359,
      "learning_rate": 9.989218372103829e-05,
      "loss": 2.842324066162109,
      "memory(GiB)": 44.8,
      "step": 2440,
      "token_acc": 0.40431266846361186,
      "train_speed(iter/s)": 1.460128
    },
    {
      "epoch": 0.10475129600274195,
      "grad_norm": 4.540597915649414,
      "learning_rate": 9.989174155799314e-05,
      "loss": 2.459308624267578,
      "memory(GiB)": 44.8,
      "step": 2445,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.459452
    },
    {
      "epoch": 0.10496551133199092,
      "grad_norm": 2.8123247623443604,
      "learning_rate": 9.989129849111229e-05,
      "loss": 2.5809600830078123,
      "memory(GiB)": 44.8,
      "step": 2450,
      "token_acc": 0.48589341692789967,
      "train_speed(iter/s)": 1.459326
    },
    {
      "epoch": 0.10517972666123988,
      "grad_norm": 3.3037538528442383,
      "learning_rate": 9.989085452040377e-05,
      "loss": 2.6184814453125,
      "memory(GiB)": 44.8,
      "step": 2455,
      "token_acc": 0.47706422018348627,
      "train_speed(iter/s)": 1.458921
    },
    {
      "epoch": 0.10539394199048883,
      "grad_norm": 3.4188737869262695,
      "learning_rate": 9.989040964587563e-05,
      "loss": 2.8132484436035154,
      "memory(GiB)": 44.8,
      "step": 2460,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.459192
    },
    {
      "epoch": 0.1056081573197378,
      "grad_norm": 5.026494979858398,
      "learning_rate": 9.988996386753591e-05,
      "loss": 2.631730651855469,
      "memory(GiB)": 44.8,
      "step": 2465,
      "token_acc": 0.44625407166123776,
      "train_speed(iter/s)": 1.459446
    },
    {
      "epoch": 0.10582237264898676,
      "grad_norm": 7.5906782150268555,
      "learning_rate": 9.988951718539269e-05,
      "loss": 2.9284337997436523,
      "memory(GiB)": 44.8,
      "step": 2470,
      "token_acc": 0.4609053497942387,
      "train_speed(iter/s)": 1.459675
    },
    {
      "epoch": 0.10603658797823572,
      "grad_norm": 3.620561122894287,
      "learning_rate": 9.988906959945407e-05,
      "loss": 2.4706584930419924,
      "memory(GiB)": 44.8,
      "step": 2475,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.460131
    },
    {
      "epoch": 0.10625080330748468,
      "grad_norm": 3.532071828842163,
      "learning_rate": 9.988862110972816e-05,
      "loss": 2.6927810668945313,
      "memory(GiB)": 44.8,
      "step": 2480,
      "token_acc": 0.4246575342465753,
      "train_speed(iter/s)": 1.460031
    },
    {
      "epoch": 0.10646501863673365,
      "grad_norm": 6.2653021812438965,
      "learning_rate": 9.988817171622307e-05,
      "loss": 2.4296083450317383,
      "memory(GiB)": 44.8,
      "step": 2485,
      "token_acc": 0.44813278008298757,
      "train_speed(iter/s)": 1.46073
    },
    {
      "epoch": 0.10667923396598261,
      "grad_norm": 4.349013328552246,
      "learning_rate": 9.988772141894696e-05,
      "loss": 2.7962924957275392,
      "memory(GiB)": 44.8,
      "step": 2490,
      "token_acc": 0.43956043956043955,
      "train_speed(iter/s)": 1.461129
    },
    {
      "epoch": 0.10689344929523156,
      "grad_norm": 3.796917200088501,
      "learning_rate": 9.988727021790796e-05,
      "loss": 2.7859775543212892,
      "memory(GiB)": 44.8,
      "step": 2495,
      "token_acc": 0.4421364985163205,
      "train_speed(iter/s)": 1.461341
    },
    {
      "epoch": 0.10710766462448053,
      "grad_norm": 3.583979845046997,
      "learning_rate": 9.988681811311428e-05,
      "loss": 2.744350624084473,
      "memory(GiB)": 44.8,
      "step": 2500,
      "token_acc": 0.41924398625429554,
      "train_speed(iter/s)": 1.461723
    },
    {
      "epoch": 0.10710766462448053,
      "eval_loss": 2.315615177154541,
      "eval_runtime": 13.9767,
      "eval_samples_per_second": 7.155,
      "eval_steps_per_second": 7.155,
      "eval_token_acc": 0.45179856115107914,
      "step": 2500
    },
    {
      "epoch": 0.1073218799537295,
      "grad_norm": 3.135530710220337,
      "learning_rate": 9.988636510457408e-05,
      "loss": 2.7376922607421874,
      "memory(GiB)": 44.8,
      "step": 2505,
      "token_acc": 0.4548076923076923,
      "train_speed(iter/s)": 1.449332
    },
    {
      "epoch": 0.10753609528297844,
      "grad_norm": 4.523245811462402,
      "learning_rate": 9.988591119229559e-05,
      "loss": 2.723159980773926,
      "memory(GiB)": 44.8,
      "step": 2510,
      "token_acc": 0.43508771929824563,
      "train_speed(iter/s)": 1.448981
    },
    {
      "epoch": 0.10775031061222741,
      "grad_norm": 4.681279182434082,
      "learning_rate": 9.988545637628702e-05,
      "loss": 2.5212112426757813,
      "memory(GiB)": 44.8,
      "step": 2515,
      "token_acc": 0.46176470588235297,
      "train_speed(iter/s)": 1.448885
    },
    {
      "epoch": 0.10796452594147637,
      "grad_norm": 4.707463264465332,
      "learning_rate": 9.98850006565566e-05,
      "loss": 2.535818862915039,
      "memory(GiB)": 44.8,
      "step": 2520,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.448804
    },
    {
      "epoch": 0.10817874127072534,
      "grad_norm": 4.106057167053223,
      "learning_rate": 9.988454403311262e-05,
      "loss": 2.681136703491211,
      "memory(GiB)": 44.8,
      "step": 2525,
      "token_acc": 0.4090909090909091,
      "train_speed(iter/s)": 1.448967
    },
    {
      "epoch": 0.10839295659997429,
      "grad_norm": 3.8454689979553223,
      "learning_rate": 9.98840865059633e-05,
      "loss": 2.5826454162597656,
      "memory(GiB)": 44.8,
      "step": 2530,
      "token_acc": 0.4728682170542636,
      "train_speed(iter/s)": 1.449246
    },
    {
      "epoch": 0.10860717192922326,
      "grad_norm": 3.1298320293426514,
      "learning_rate": 9.988362807511697e-05,
      "loss": 2.8767068862915037,
      "memory(GiB)": 44.8,
      "step": 2535,
      "token_acc": 0.43086816720257237,
      "train_speed(iter/s)": 1.449185
    },
    {
      "epoch": 0.10882138725847222,
      "grad_norm": 3.548053741455078,
      "learning_rate": 9.988316874058195e-05,
      "loss": 2.7723020553588866,
      "memory(GiB)": 44.8,
      "step": 2540,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.449884
    },
    {
      "epoch": 0.10903560258772117,
      "grad_norm": 3.266469717025757,
      "learning_rate": 9.988270850236649e-05,
      "loss": 2.901295280456543,
      "memory(GiB)": 44.8,
      "step": 2545,
      "token_acc": 0.4296577946768061,
      "train_speed(iter/s)": 1.449428
    },
    {
      "epoch": 0.10924981791697014,
      "grad_norm": 4.520561218261719,
      "learning_rate": 9.9882247360479e-05,
      "loss": 2.9140830993652345,
      "memory(GiB)": 44.8,
      "step": 2550,
      "token_acc": 0.41947565543071164,
      "train_speed(iter/s)": 1.449681
    },
    {
      "epoch": 0.1094640332462191,
      "grad_norm": 6.197085380554199,
      "learning_rate": 9.98817853149278e-05,
      "loss": 2.654117965698242,
      "memory(GiB)": 44.8,
      "step": 2555,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.449702
    },
    {
      "epoch": 0.10967824857546805,
      "grad_norm": 3.7634689807891846,
      "learning_rate": 9.988132236572126e-05,
      "loss": 2.733693313598633,
      "memory(GiB)": 44.8,
      "step": 2560,
      "token_acc": 0.43174603174603177,
      "train_speed(iter/s)": 1.449961
    },
    {
      "epoch": 0.10989246390471702,
      "grad_norm": 3.0915908813476562,
      "learning_rate": 9.988085851286778e-05,
      "loss": 2.782105255126953,
      "memory(GiB)": 44.8,
      "step": 2565,
      "token_acc": 0.4415954415954416,
      "train_speed(iter/s)": 1.450478
    },
    {
      "epoch": 0.11010667923396598,
      "grad_norm": 3.5323173999786377,
      "learning_rate": 9.988039375637574e-05,
      "loss": 2.514674758911133,
      "memory(GiB)": 44.8,
      "step": 2570,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.450904
    },
    {
      "epoch": 0.11032089456321495,
      "grad_norm": 3.008054733276367,
      "learning_rate": 9.98799280962536e-05,
      "loss": 2.7011260986328125,
      "memory(GiB)": 44.8,
      "step": 2575,
      "token_acc": 0.43137254901960786,
      "train_speed(iter/s)": 1.451432
    },
    {
      "epoch": 0.1105351098924639,
      "grad_norm": 3.8846633434295654,
      "learning_rate": 9.987946153250976e-05,
      "loss": 2.735409164428711,
      "memory(GiB)": 44.8,
      "step": 2580,
      "token_acc": 0.4318181818181818,
      "train_speed(iter/s)": 1.451935
    },
    {
      "epoch": 0.11074932522171287,
      "grad_norm": 3.538858413696289,
      "learning_rate": 9.987899406515268e-05,
      "loss": 2.938157844543457,
      "memory(GiB)": 44.8,
      "step": 2585,
      "token_acc": 0.3815789473684211,
      "train_speed(iter/s)": 1.452487
    },
    {
      "epoch": 0.11096354055096183,
      "grad_norm": 8.99933910369873,
      "learning_rate": 9.987852569419084e-05,
      "loss": 2.7111330032348633,
      "memory(GiB)": 44.8,
      "step": 2590,
      "token_acc": 0.38436482084690554,
      "train_speed(iter/s)": 1.452583
    },
    {
      "epoch": 0.11117775588021078,
      "grad_norm": 4.776661396026611,
      "learning_rate": 9.98780564196327e-05,
      "loss": 2.924737548828125,
      "memory(GiB)": 44.8,
      "step": 2595,
      "token_acc": 0.40327868852459015,
      "train_speed(iter/s)": 1.45278
    },
    {
      "epoch": 0.11139197120945975,
      "grad_norm": 3.6866297721862793,
      "learning_rate": 9.98775862414868e-05,
      "loss": 2.336515426635742,
      "memory(GiB)": 44.8,
      "step": 2600,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.453238
    },
    {
      "epoch": 0.11160618653870871,
      "grad_norm": 4.3828816413879395,
      "learning_rate": 9.987711515976164e-05,
      "loss": 2.57989387512207,
      "memory(GiB)": 44.8,
      "step": 2605,
      "token_acc": 0.46096654275092935,
      "train_speed(iter/s)": 1.453247
    },
    {
      "epoch": 0.11182040186795768,
      "grad_norm": 3.051657199859619,
      "learning_rate": 9.987664317446572e-05,
      "loss": 2.899135971069336,
      "memory(GiB)": 44.8,
      "step": 2610,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.453688
    },
    {
      "epoch": 0.11203461719720663,
      "grad_norm": 5.278879165649414,
      "learning_rate": 9.987617028560765e-05,
      "loss": 2.8429651260375977,
      "memory(GiB)": 44.8,
      "step": 2615,
      "token_acc": 0.4174174174174174,
      "train_speed(iter/s)": 1.453461
    },
    {
      "epoch": 0.1122488325264556,
      "grad_norm": 3.091171979904175,
      "learning_rate": 9.987569649319595e-05,
      "loss": 2.5073850631713865,
      "memory(GiB)": 44.8,
      "step": 2620,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.454114
    },
    {
      "epoch": 0.11246304785570456,
      "grad_norm": 3.2750463485717773,
      "learning_rate": 9.987522179723923e-05,
      "loss": 2.5251392364501952,
      "memory(GiB)": 44.8,
      "step": 2625,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.454021
    },
    {
      "epoch": 0.11267726318495351,
      "grad_norm": 5.543854236602783,
      "learning_rate": 9.987474619774609e-05,
      "loss": 2.63586368560791,
      "memory(GiB)": 44.8,
      "step": 2630,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.45397
    },
    {
      "epoch": 0.11289147851420248,
      "grad_norm": 3.494727849960327,
      "learning_rate": 9.98742696947251e-05,
      "loss": 2.8177515029907227,
      "memory(GiB)": 44.8,
      "step": 2635,
      "token_acc": 0.44785276073619634,
      "train_speed(iter/s)": 1.453574
    },
    {
      "epoch": 0.11310569384345144,
      "grad_norm": 5.487606525421143,
      "learning_rate": 9.987379228818497e-05,
      "loss": 2.436424446105957,
      "memory(GiB)": 44.8,
      "step": 2640,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.45333
    },
    {
      "epoch": 0.11331990917270039,
      "grad_norm": 3.1628775596618652,
      "learning_rate": 9.987331397813429e-05,
      "loss": 2.429340362548828,
      "memory(GiB)": 44.8,
      "step": 2645,
      "token_acc": 0.4819672131147541,
      "train_speed(iter/s)": 1.453584
    },
    {
      "epoch": 0.11353412450194936,
      "grad_norm": 3.776961326599121,
      "learning_rate": 9.987283476458174e-05,
      "loss": 2.6166324615478516,
      "memory(GiB)": 44.8,
      "step": 2650,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.454593
    },
    {
      "epoch": 0.11374833983119832,
      "grad_norm": 3.17450213432312,
      "learning_rate": 9.9872354647536e-05,
      "loss": 2.866488456726074,
      "memory(GiB)": 44.8,
      "step": 2655,
      "token_acc": 0.4386503067484663,
      "train_speed(iter/s)": 1.454631
    },
    {
      "epoch": 0.11396255516044729,
      "grad_norm": 4.373410701751709,
      "learning_rate": 9.987187362700579e-05,
      "loss": 2.6585012435913087,
      "memory(GiB)": 44.8,
      "step": 2660,
      "token_acc": 0.458041958041958,
      "train_speed(iter/s)": 1.454704
    },
    {
      "epoch": 0.11417677048969624,
      "grad_norm": 4.524393081665039,
      "learning_rate": 9.987139170299978e-05,
      "loss": 2.6862470626831056,
      "memory(GiB)": 44.8,
      "step": 2665,
      "token_acc": 0.41935483870967744,
      "train_speed(iter/s)": 1.454475
    },
    {
      "epoch": 0.1143909858189452,
      "grad_norm": 4.258168697357178,
      "learning_rate": 9.987090887552675e-05,
      "loss": 2.646759605407715,
      "memory(GiB)": 44.8,
      "step": 2670,
      "token_acc": 0.4407894736842105,
      "train_speed(iter/s)": 1.455336
    },
    {
      "epoch": 0.11460520114819417,
      "grad_norm": 3.578068256378174,
      "learning_rate": 9.987042514459541e-05,
      "loss": 2.4732179641723633,
      "memory(GiB)": 44.8,
      "step": 2675,
      "token_acc": 0.4734848484848485,
      "train_speed(iter/s)": 1.455521
    },
    {
      "epoch": 0.11481941647744312,
      "grad_norm": 5.073055267333984,
      "learning_rate": 9.986994051021454e-05,
      "loss": 2.4216739654541017,
      "memory(GiB)": 44.8,
      "step": 2680,
      "token_acc": 0.47416413373860183,
      "train_speed(iter/s)": 1.455442
    },
    {
      "epoch": 0.11503363180669209,
      "grad_norm": 3.739480495452881,
      "learning_rate": 9.98694549723929e-05,
      "loss": 2.73177547454834,
      "memory(GiB)": 44.8,
      "step": 2685,
      "token_acc": 0.46089385474860334,
      "train_speed(iter/s)": 1.455095
    },
    {
      "epoch": 0.11524784713594105,
      "grad_norm": 4.096671104431152,
      "learning_rate": 9.986896853113932e-05,
      "loss": 2.700012969970703,
      "memory(GiB)": 44.8,
      "step": 2690,
      "token_acc": 0.44871794871794873,
      "train_speed(iter/s)": 1.454966
    },
    {
      "epoch": 0.11546206246519002,
      "grad_norm": 4.17447566986084,
      "learning_rate": 9.98684811864626e-05,
      "loss": 2.514265251159668,
      "memory(GiB)": 44.8,
      "step": 2695,
      "token_acc": 0.48627450980392156,
      "train_speed(iter/s)": 1.455102
    },
    {
      "epoch": 0.11567627779443897,
      "grad_norm": 4.056727886199951,
      "learning_rate": 9.986799293837155e-05,
      "loss": 2.7692144393920897,
      "memory(GiB)": 44.8,
      "step": 2700,
      "token_acc": 0.45977011494252873,
      "train_speed(iter/s)": 1.455007
    },
    {
      "epoch": 0.11589049312368793,
      "grad_norm": 3.6206116676330566,
      "learning_rate": 9.986750378687502e-05,
      "loss": 2.3905899047851564,
      "memory(GiB)": 44.8,
      "step": 2705,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.454789
    },
    {
      "epoch": 0.1161047084529369,
      "grad_norm": 5.165829181671143,
      "learning_rate": 9.98670137319819e-05,
      "loss": 2.703978729248047,
      "memory(GiB)": 44.8,
      "step": 2710,
      "token_acc": 0.4453125,
      "train_speed(iter/s)": 1.454077
    },
    {
      "epoch": 0.11631892378218585,
      "grad_norm": 3.4856631755828857,
      "learning_rate": 9.986652277370103e-05,
      "loss": 2.891558074951172,
      "memory(GiB)": 44.8,
      "step": 2715,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.454489
    },
    {
      "epoch": 0.11653313911143481,
      "grad_norm": 3.9351048469543457,
      "learning_rate": 9.986603091204134e-05,
      "loss": 2.5380075454711912,
      "memory(GiB)": 44.8,
      "step": 2720,
      "token_acc": 0.4588235294117647,
      "train_speed(iter/s)": 1.454791
    },
    {
      "epoch": 0.11674735444068378,
      "grad_norm": 5.181811809539795,
      "learning_rate": 9.98655381470117e-05,
      "loss": 2.854144287109375,
      "memory(GiB)": 44.8,
      "step": 2725,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.454509
    },
    {
      "epoch": 0.11696156976993273,
      "grad_norm": 4.077805042266846,
      "learning_rate": 9.986504447862108e-05,
      "loss": 2.4561676025390624,
      "memory(GiB)": 44.8,
      "step": 2730,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.454177
    },
    {
      "epoch": 0.1171757850991817,
      "grad_norm": 3.2708146572113037,
      "learning_rate": 9.986454990687839e-05,
      "loss": 2.478154182434082,
      "memory(GiB)": 44.8,
      "step": 2735,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.453891
    },
    {
      "epoch": 0.11739000042843066,
      "grad_norm": 4.113687992095947,
      "learning_rate": 9.986405443179261e-05,
      "loss": 2.59047908782959,
      "memory(GiB)": 44.8,
      "step": 2740,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.454461
    },
    {
      "epoch": 0.11760421575767963,
      "grad_norm": 3.1191680431365967,
      "learning_rate": 9.986355805337268e-05,
      "loss": 2.495977783203125,
      "memory(GiB)": 44.8,
      "step": 2745,
      "token_acc": 0.4227941176470588,
      "train_speed(iter/s)": 1.454518
    },
    {
      "epoch": 0.11781843108692858,
      "grad_norm": 4.838729381561279,
      "learning_rate": 9.986306077162766e-05,
      "loss": 2.7230384826660154,
      "memory(GiB)": 44.8,
      "step": 2750,
      "token_acc": 0.4392156862745098,
      "train_speed(iter/s)": 1.455454
    },
    {
      "epoch": 0.11803264641617754,
      "grad_norm": 4.422349452972412,
      "learning_rate": 9.98625625865665e-05,
      "loss": 2.635031509399414,
      "memory(GiB)": 44.8,
      "step": 2755,
      "token_acc": 0.43317972350230416,
      "train_speed(iter/s)": 1.455255
    },
    {
      "epoch": 0.11824686174542651,
      "grad_norm": 3.5057785511016846,
      "learning_rate": 9.986206349819825e-05,
      "loss": 2.7041976928710936,
      "memory(GiB)": 44.8,
      "step": 2760,
      "token_acc": 0.4669421487603306,
      "train_speed(iter/s)": 1.455021
    },
    {
      "epoch": 0.11846107707467546,
      "grad_norm": 4.519093990325928,
      "learning_rate": 9.986156350653193e-05,
      "loss": 2.80269775390625,
      "memory(GiB)": 44.8,
      "step": 2765,
      "token_acc": 0.43795620437956206,
      "train_speed(iter/s)": 1.455074
    },
    {
      "epoch": 0.11867529240392442,
      "grad_norm": 3.121180295944214,
      "learning_rate": 9.986106261157662e-05,
      "loss": 2.543857955932617,
      "memory(GiB)": 44.8,
      "step": 2770,
      "token_acc": 0.5115511551155115,
      "train_speed(iter/s)": 1.45498
    },
    {
      "epoch": 0.11888950773317339,
      "grad_norm": 3.253769636154175,
      "learning_rate": 9.986056081334139e-05,
      "loss": 2.355792236328125,
      "memory(GiB)": 44.8,
      "step": 2775,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.455323
    },
    {
      "epoch": 0.11910372306242235,
      "grad_norm": 4.304771900177002,
      "learning_rate": 9.986005811183533e-05,
      "loss": 2.3715681076049804,
      "memory(GiB)": 44.8,
      "step": 2780,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.455675
    },
    {
      "epoch": 0.1193179383916713,
      "grad_norm": 4.088959217071533,
      "learning_rate": 9.985955450706755e-05,
      "loss": 2.558426284790039,
      "memory(GiB)": 44.8,
      "step": 2785,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.456214
    },
    {
      "epoch": 0.11953215372092027,
      "grad_norm": 3.599224328994751,
      "learning_rate": 9.985904999904715e-05,
      "loss": 2.6276081085205076,
      "memory(GiB)": 44.8,
      "step": 2790,
      "token_acc": 0.44765342960288806,
      "train_speed(iter/s)": 1.455336
    },
    {
      "epoch": 0.11974636905016924,
      "grad_norm": 3.8726751804351807,
      "learning_rate": 9.98585445877833e-05,
      "loss": 3.040993499755859,
      "memory(GiB)": 44.8,
      "step": 2795,
      "token_acc": 0.4266666666666667,
      "train_speed(iter/s)": 1.455832
    },
    {
      "epoch": 0.11996058437941819,
      "grad_norm": 5.547581672668457,
      "learning_rate": 9.985803827328515e-05,
      "loss": 2.7757171630859374,
      "memory(GiB)": 44.8,
      "step": 2800,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.455775
    },
    {
      "epoch": 0.12017479970866715,
      "grad_norm": 4.747446060180664,
      "learning_rate": 9.985753105556185e-05,
      "loss": 2.4627155303955077,
      "memory(GiB)": 44.8,
      "step": 2805,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.456337
    },
    {
      "epoch": 0.12038901503791612,
      "grad_norm": 5.369457721710205,
      "learning_rate": 9.985702293462261e-05,
      "loss": 2.4614301681518556,
      "memory(GiB)": 44.8,
      "step": 2810,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.456445
    },
    {
      "epoch": 0.12060323036716507,
      "grad_norm": 3.3231163024902344,
      "learning_rate": 9.985651391047663e-05,
      "loss": 2.8215492248535154,
      "memory(GiB)": 44.8,
      "step": 2815,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.456758
    },
    {
      "epoch": 0.12081744569641403,
      "grad_norm": 2.909749984741211,
      "learning_rate": 9.985600398313313e-05,
      "loss": 2.706439399719238,
      "memory(GiB)": 44.8,
      "step": 2820,
      "token_acc": 0.46075085324232085,
      "train_speed(iter/s)": 1.456957
    },
    {
      "epoch": 0.121031661025663,
      "grad_norm": 6.230549335479736,
      "learning_rate": 9.985549315260137e-05,
      "loss": 2.837756538391113,
      "memory(GiB)": 44.8,
      "step": 2825,
      "token_acc": 0.4246575342465753,
      "train_speed(iter/s)": 1.457146
    },
    {
      "epoch": 0.12124587635491196,
      "grad_norm": 3.530992269515991,
      "learning_rate": 9.985498141889056e-05,
      "loss": 2.4590328216552733,
      "memory(GiB)": 44.8,
      "step": 2830,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.457587
    },
    {
      "epoch": 0.12146009168416091,
      "grad_norm": 4.051492214202881,
      "learning_rate": 9.985446878201e-05,
      "loss": 2.934111404418945,
      "memory(GiB)": 44.8,
      "step": 2835,
      "token_acc": 0.40509915014164305,
      "train_speed(iter/s)": 1.458241
    },
    {
      "epoch": 0.12167430701340988,
      "grad_norm": 5.797842025756836,
      "learning_rate": 9.985395524196896e-05,
      "loss": 2.639509391784668,
      "memory(GiB)": 44.8,
      "step": 2840,
      "token_acc": 0.44525547445255476,
      "train_speed(iter/s)": 1.45777
    },
    {
      "epoch": 0.12188852234265884,
      "grad_norm": 3.338931083679199,
      "learning_rate": 9.985344079877677e-05,
      "loss": 2.498863410949707,
      "memory(GiB)": 44.8,
      "step": 2845,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.457512
    },
    {
      "epoch": 0.1221027376719078,
      "grad_norm": 3.9982433319091797,
      "learning_rate": 9.985292545244274e-05,
      "loss": 2.7938840866088865,
      "memory(GiB)": 44.8,
      "step": 2850,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.457688
    },
    {
      "epoch": 0.12231695300115676,
      "grad_norm": 3.668144941329956,
      "learning_rate": 9.985240920297618e-05,
      "loss": 2.768280029296875,
      "memory(GiB)": 44.8,
      "step": 2855,
      "token_acc": 0.43944636678200694,
      "train_speed(iter/s)": 1.457747
    },
    {
      "epoch": 0.12253116833040573,
      "grad_norm": 3.793884038925171,
      "learning_rate": 9.985189205038647e-05,
      "loss": 2.6596202850341797,
      "memory(GiB)": 44.8,
      "step": 2860,
      "token_acc": 0.46099290780141844,
      "train_speed(iter/s)": 1.458169
    },
    {
      "epoch": 0.12274538365965469,
      "grad_norm": 5.051962852478027,
      "learning_rate": 9.985137399468298e-05,
      "loss": 2.7195808410644533,
      "memory(GiB)": 44.8,
      "step": 2865,
      "token_acc": 0.4492753623188406,
      "train_speed(iter/s)": 1.458557
    },
    {
      "epoch": 0.12295959898890364,
      "grad_norm": 4.148073673248291,
      "learning_rate": 9.985085503587507e-05,
      "loss": 2.688954162597656,
      "memory(GiB)": 44.8,
      "step": 2870,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.458639
    },
    {
      "epoch": 0.12317381431815261,
      "grad_norm": 9.053473472595215,
      "learning_rate": 9.985033517397217e-05,
      "loss": 2.5241180419921876,
      "memory(GiB)": 44.8,
      "step": 2875,
      "token_acc": 0.4575645756457565,
      "train_speed(iter/s)": 1.458511
    },
    {
      "epoch": 0.12338802964740157,
      "grad_norm": 4.768235206604004,
      "learning_rate": 9.984981440898366e-05,
      "loss": 2.556084632873535,
      "memory(GiB)": 44.8,
      "step": 2880,
      "token_acc": 0.4724137931034483,
      "train_speed(iter/s)": 1.458841
    },
    {
      "epoch": 0.12360224497665052,
      "grad_norm": 3.2739498615264893,
      "learning_rate": 9.984929274091903e-05,
      "loss": 2.971639633178711,
      "memory(GiB)": 44.8,
      "step": 2885,
      "token_acc": 0.3942307692307692,
      "train_speed(iter/s)": 1.459376
    },
    {
      "epoch": 0.12381646030589949,
      "grad_norm": 3.549626111984253,
      "learning_rate": 9.984877016978768e-05,
      "loss": 2.8099727630615234,
      "memory(GiB)": 44.8,
      "step": 2890,
      "token_acc": 0.45272206303724927,
      "train_speed(iter/s)": 1.460056
    },
    {
      "epoch": 0.12403067563514845,
      "grad_norm": 2.765273094177246,
      "learning_rate": 9.984824669559911e-05,
      "loss": 2.5519046783447266,
      "memory(GiB)": 44.8,
      "step": 2895,
      "token_acc": 0.463855421686747,
      "train_speed(iter/s)": 1.460508
    },
    {
      "epoch": 0.1242448909643974,
      "grad_norm": 3.7635700702667236,
      "learning_rate": 9.984772231836279e-05,
      "loss": 2.78131217956543,
      "memory(GiB)": 44.8,
      "step": 2900,
      "token_acc": 0.4398496240601504,
      "train_speed(iter/s)": 1.460391
    },
    {
      "epoch": 0.12445910629364637,
      "grad_norm": 4.400283336639404,
      "learning_rate": 9.98471970380882e-05,
      "loss": 2.8672739028930665,
      "memory(GiB)": 47.63,
      "step": 2905,
      "token_acc": 0.4222222222222222,
      "train_speed(iter/s)": 1.45939
    },
    {
      "epoch": 0.12467332162289534,
      "grad_norm": 3.5733256340026855,
      "learning_rate": 9.98466708547849e-05,
      "loss": 2.6260372161865235,
      "memory(GiB)": 47.63,
      "step": 2910,
      "token_acc": 0.4370860927152318,
      "train_speed(iter/s)": 1.459296
    },
    {
      "epoch": 0.1248875369521443,
      "grad_norm": 2.9659810066223145,
      "learning_rate": 9.984614376846238e-05,
      "loss": 2.588872528076172,
      "memory(GiB)": 47.63,
      "step": 2915,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.459633
    },
    {
      "epoch": 0.12510175228139325,
      "grad_norm": 2.75057053565979,
      "learning_rate": 9.98456157791302e-05,
      "loss": 2.133049201965332,
      "memory(GiB)": 47.63,
      "step": 2920,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.459971
    },
    {
      "epoch": 0.1253159676106422,
      "grad_norm": 3.801823139190674,
      "learning_rate": 9.984508688679796e-05,
      "loss": 2.666956329345703,
      "memory(GiB)": 47.63,
      "step": 2925,
      "token_acc": 0.4978165938864629,
      "train_speed(iter/s)": 1.460505
    },
    {
      "epoch": 0.12553018293989118,
      "grad_norm": 2.849738597869873,
      "learning_rate": 9.984455709147519e-05,
      "loss": 2.7858362197875977,
      "memory(GiB)": 47.63,
      "step": 2930,
      "token_acc": 0.4401294498381877,
      "train_speed(iter/s)": 1.460717
    },
    {
      "epoch": 0.12574439826914013,
      "grad_norm": 3.5369927883148193,
      "learning_rate": 9.984402639317152e-05,
      "loss": 2.5744321823120115,
      "memory(GiB)": 47.63,
      "step": 2935,
      "token_acc": 0.46229508196721314,
      "train_speed(iter/s)": 1.46155
    },
    {
      "epoch": 0.1259586135983891,
      "grad_norm": 5.130538463592529,
      "learning_rate": 9.984349479189654e-05,
      "loss": 2.6871822357177733,
      "memory(GiB)": 47.63,
      "step": 2940,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.462424
    },
    {
      "epoch": 0.12617282892763806,
      "grad_norm": 4.449698448181152,
      "learning_rate": 9.984296228765991e-05,
      "loss": 2.8014768600463866,
      "memory(GiB)": 47.63,
      "step": 2945,
      "token_acc": 0.4135593220338983,
      "train_speed(iter/s)": 1.462565
    },
    {
      "epoch": 0.12638704425688702,
      "grad_norm": 3.4198319911956787,
      "learning_rate": 9.984242888047127e-05,
      "loss": 2.7933071136474608,
      "memory(GiB)": 47.63,
      "step": 2950,
      "token_acc": 0.4576271186440678,
      "train_speed(iter/s)": 1.462375
    },
    {
      "epoch": 0.126601259586136,
      "grad_norm": 5.142146587371826,
      "learning_rate": 9.984189457034026e-05,
      "loss": 2.7760799407958983,
      "memory(GiB)": 47.63,
      "step": 2955,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.461615
    },
    {
      "epoch": 0.12681547491538495,
      "grad_norm": 4.261962413787842,
      "learning_rate": 9.984135935727657e-05,
      "loss": 2.888034629821777,
      "memory(GiB)": 47.63,
      "step": 2960,
      "token_acc": 0.41916167664670656,
      "train_speed(iter/s)": 1.461648
    },
    {
      "epoch": 0.1270296902446339,
      "grad_norm": 3.322274684906006,
      "learning_rate": 9.984082324128991e-05,
      "loss": 2.7487483978271485,
      "memory(GiB)": 47.63,
      "step": 2965,
      "token_acc": 0.41694915254237286,
      "train_speed(iter/s)": 1.461641
    },
    {
      "epoch": 0.12724390557388288,
      "grad_norm": 6.232876300811768,
      "learning_rate": 9.984028622238997e-05,
      "loss": 2.972730255126953,
      "memory(GiB)": 47.63,
      "step": 2970,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.461888
    },
    {
      "epoch": 0.12745812090313183,
      "grad_norm": 4.244607448577881,
      "learning_rate": 9.98397483005865e-05,
      "loss": 2.574996757507324,
      "memory(GiB)": 47.63,
      "step": 2975,
      "token_acc": 0.5221238938053098,
      "train_speed(iter/s)": 1.462386
    },
    {
      "epoch": 0.12767233623238078,
      "grad_norm": 3.4390854835510254,
      "learning_rate": 9.983920947588923e-05,
      "loss": 3.0913532257080076,
      "memory(GiB)": 47.63,
      "step": 2980,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.462509
    },
    {
      "epoch": 0.12788655156162976,
      "grad_norm": 3.9557533264160156,
      "learning_rate": 9.983866974830794e-05,
      "loss": 2.6116329193115235,
      "memory(GiB)": 47.63,
      "step": 2985,
      "token_acc": 0.44807121661721067,
      "train_speed(iter/s)": 1.462362
    },
    {
      "epoch": 0.1281007668908787,
      "grad_norm": 4.995320796966553,
      "learning_rate": 9.983812911785238e-05,
      "loss": 2.72194881439209,
      "memory(GiB)": 47.63,
      "step": 2990,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.462744
    },
    {
      "epoch": 0.12831498222012766,
      "grad_norm": 4.00553035736084,
      "learning_rate": 9.983758758453238e-05,
      "loss": 2.504826545715332,
      "memory(GiB)": 47.63,
      "step": 2995,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.462946
    },
    {
      "epoch": 0.12852919754937664,
      "grad_norm": 4.180451393127441,
      "learning_rate": 9.983704514835771e-05,
      "loss": 2.6988336563110353,
      "memory(GiB)": 47.63,
      "step": 3000,
      "token_acc": 0.47019867549668876,
      "train_speed(iter/s)": 1.463163
    },
    {
      "epoch": 0.12852919754937664,
      "eval_loss": 2.293785333633423,
      "eval_runtime": 14.3443,
      "eval_samples_per_second": 6.971,
      "eval_steps_per_second": 6.971,
      "eval_token_acc": 0.453015427769986,
      "step": 3000
    },
    {
      "epoch": 0.1287434128786256,
      "grad_norm": 4.106462001800537,
      "learning_rate": 9.983650180933822e-05,
      "loss": 2.8215705871582033,
      "memory(GiB)": 47.63,
      "step": 3005,
      "token_acc": 0.44668008048289737,
      "train_speed(iter/s)": 1.452652
    },
    {
      "epoch": 0.12895762820787454,
      "grad_norm": 7.2187347412109375,
      "learning_rate": 9.983595756748376e-05,
      "loss": 2.584635543823242,
      "memory(GiB)": 47.63,
      "step": 3010,
      "token_acc": 0.4431818181818182,
      "train_speed(iter/s)": 1.452488
    },
    {
      "epoch": 0.12917184353712352,
      "grad_norm": 4.628119945526123,
      "learning_rate": 9.983541242280417e-05,
      "loss": 2.9025840759277344,
      "memory(GiB)": 47.63,
      "step": 3015,
      "token_acc": 0.42024539877300615,
      "train_speed(iter/s)": 1.452558
    },
    {
      "epoch": 0.12938605886637247,
      "grad_norm": 4.669239044189453,
      "learning_rate": 9.983486637530934e-05,
      "loss": 2.6678781509399414,
      "memory(GiB)": 47.63,
      "step": 3020,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.453222
    },
    {
      "epoch": 0.12960027419562145,
      "grad_norm": 3.1429362297058105,
      "learning_rate": 9.983431942500915e-05,
      "loss": 2.8684564590454102,
      "memory(GiB)": 47.63,
      "step": 3025,
      "token_acc": 0.41389728096676737,
      "train_speed(iter/s)": 1.453618
    },
    {
      "epoch": 0.1298144895248704,
      "grad_norm": 2.8368780612945557,
      "learning_rate": 9.983377157191352e-05,
      "loss": 2.491801643371582,
      "memory(GiB)": 47.63,
      "step": 3030,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.453491
    },
    {
      "epoch": 0.13002870485411935,
      "grad_norm": 2.881819486618042,
      "learning_rate": 9.983322281603235e-05,
      "loss": 2.719405746459961,
      "memory(GiB)": 47.63,
      "step": 3035,
      "token_acc": 0.4368231046931408,
      "train_speed(iter/s)": 1.453369
    },
    {
      "epoch": 0.13024292018336833,
      "grad_norm": 4.35675048828125,
      "learning_rate": 9.983267315737563e-05,
      "loss": 2.4980485916137694,
      "memory(GiB)": 47.63,
      "step": 3040,
      "token_acc": 0.45041322314049587,
      "train_speed(iter/s)": 1.454165
    },
    {
      "epoch": 0.13045713551261728,
      "grad_norm": 3.4245173931121826,
      "learning_rate": 9.983212259595328e-05,
      "loss": 2.6589332580566407,
      "memory(GiB)": 47.63,
      "step": 3045,
      "token_acc": 0.4355400696864111,
      "train_speed(iter/s)": 1.454215
    },
    {
      "epoch": 0.13067135084186624,
      "grad_norm": 4.292496204376221,
      "learning_rate": 9.983157113177528e-05,
      "loss": 2.5680091857910154,
      "memory(GiB)": 47.63,
      "step": 3050,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.454931
    },
    {
      "epoch": 0.13088556617111521,
      "grad_norm": 3.379364490509033,
      "learning_rate": 9.983101876485162e-05,
      "loss": 3.025213623046875,
      "memory(GiB)": 47.63,
      "step": 3055,
      "token_acc": 0.43217665615141954,
      "train_speed(iter/s)": 1.455643
    },
    {
      "epoch": 0.13109978150036417,
      "grad_norm": 5.147782325744629,
      "learning_rate": 9.983046549519232e-05,
      "loss": 2.453817367553711,
      "memory(GiB)": 47.63,
      "step": 3060,
      "token_acc": 0.45528455284552843,
      "train_speed(iter/s)": 1.455458
    },
    {
      "epoch": 0.13131399682961312,
      "grad_norm": 4.744100570678711,
      "learning_rate": 9.982991132280739e-05,
      "loss": 2.5531290054321287,
      "memory(GiB)": 47.63,
      "step": 3065,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.455646
    },
    {
      "epoch": 0.1315282121588621,
      "grad_norm": 4.192126750946045,
      "learning_rate": 9.982935624770687e-05,
      "loss": 2.2457679748535155,
      "memory(GiB)": 47.63,
      "step": 3070,
      "token_acc": 0.515527950310559,
      "train_speed(iter/s)": 1.455399
    },
    {
      "epoch": 0.13174242748811105,
      "grad_norm": 3.6630325317382812,
      "learning_rate": 9.98288002699008e-05,
      "loss": 2.8035654067993163,
      "memory(GiB)": 47.63,
      "step": 3075,
      "token_acc": 0.43727598566308246,
      "train_speed(iter/s)": 1.455998
    },
    {
      "epoch": 0.13195664281736,
      "grad_norm": 4.754130840301514,
      "learning_rate": 9.982824338939929e-05,
      "loss": 2.475682830810547,
      "memory(GiB)": 47.63,
      "step": 3080,
      "token_acc": 0.49606299212598426,
      "train_speed(iter/s)": 1.455822
    },
    {
      "epoch": 0.13217085814660898,
      "grad_norm": 4.164949893951416,
      "learning_rate": 9.98276856062124e-05,
      "loss": 2.553522300720215,
      "memory(GiB)": 47.63,
      "step": 3085,
      "token_acc": 0.4779116465863454,
      "train_speed(iter/s)": 1.455773
    },
    {
      "epoch": 0.13238507347585793,
      "grad_norm": 5.448624134063721,
      "learning_rate": 9.982712692035025e-05,
      "loss": 2.7034210205078124,
      "memory(GiB)": 47.63,
      "step": 3090,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.455792
    },
    {
      "epoch": 0.13259928880510688,
      "grad_norm": 3.43050217628479,
      "learning_rate": 9.982656733182296e-05,
      "loss": 2.0949424743652343,
      "memory(GiB)": 47.63,
      "step": 3095,
      "token_acc": 0.5330396475770925,
      "train_speed(iter/s)": 1.4562
    },
    {
      "epoch": 0.13281350413435586,
      "grad_norm": 4.507286071777344,
      "learning_rate": 9.982600684064065e-05,
      "loss": 2.70364933013916,
      "memory(GiB)": 47.63,
      "step": 3100,
      "token_acc": 0.4395973154362416,
      "train_speed(iter/s)": 1.456434
    },
    {
      "epoch": 0.1330277194636048,
      "grad_norm": 4.425159931182861,
      "learning_rate": 9.98254454468135e-05,
      "loss": 2.7447975158691404,
      "memory(GiB)": 47.63,
      "step": 3105,
      "token_acc": 0.391304347826087,
      "train_speed(iter/s)": 1.456779
    },
    {
      "epoch": 0.1332419347928538,
      "grad_norm": 3.55261492729187,
      "learning_rate": 9.982488315035166e-05,
      "loss": 2.6549182891845704,
      "memory(GiB)": 47.63,
      "step": 3110,
      "token_acc": 0.4664804469273743,
      "train_speed(iter/s)": 1.456559
    },
    {
      "epoch": 0.13345615012210274,
      "grad_norm": 4.19412899017334,
      "learning_rate": 9.982431995126531e-05,
      "loss": 2.496395492553711,
      "memory(GiB)": 47.63,
      "step": 3115,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.456263
    },
    {
      "epoch": 0.1336703654513517,
      "grad_norm": 3.5189948081970215,
      "learning_rate": 9.982375584956467e-05,
      "loss": 2.6310930252075195,
      "memory(GiB)": 47.63,
      "step": 3120,
      "token_acc": 0.42574257425742573,
      "train_speed(iter/s)": 1.456234
    },
    {
      "epoch": 0.13388458078060067,
      "grad_norm": 3.1445939540863037,
      "learning_rate": 9.982319084525995e-05,
      "loss": 2.954359436035156,
      "memory(GiB)": 47.63,
      "step": 3125,
      "token_acc": 0.42033898305084744,
      "train_speed(iter/s)": 1.456507
    },
    {
      "epoch": 0.13409879610984962,
      "grad_norm": 4.120058536529541,
      "learning_rate": 9.98226249383614e-05,
      "loss": 2.300580596923828,
      "memory(GiB)": 47.63,
      "step": 3130,
      "token_acc": 0.4618320610687023,
      "train_speed(iter/s)": 1.456747
    },
    {
      "epoch": 0.13431301143909857,
      "grad_norm": 3.6941018104553223,
      "learning_rate": 9.982205812887925e-05,
      "loss": 2.590833854675293,
      "memory(GiB)": 47.63,
      "step": 3135,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.457097
    },
    {
      "epoch": 0.13452722676834755,
      "grad_norm": 4.080898284912109,
      "learning_rate": 9.98214904168238e-05,
      "loss": 2.440216064453125,
      "memory(GiB)": 47.63,
      "step": 3140,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.45684
    },
    {
      "epoch": 0.1347414420975965,
      "grad_norm": 3.7610936164855957,
      "learning_rate": 9.98209218022053e-05,
      "loss": 2.539932632446289,
      "memory(GiB)": 47.63,
      "step": 3145,
      "token_acc": 0.46567164179104475,
      "train_speed(iter/s)": 1.456996
    },
    {
      "epoch": 0.13495565742684545,
      "grad_norm": 3.659843683242798,
      "learning_rate": 9.982035228503407e-05,
      "loss": 2.4131053924560546,
      "memory(GiB)": 47.63,
      "step": 3150,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.45732
    },
    {
      "epoch": 0.13516987275609443,
      "grad_norm": 4.357789993286133,
      "learning_rate": 9.981978186532041e-05,
      "loss": 2.670502853393555,
      "memory(GiB)": 47.63,
      "step": 3155,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.457363
    },
    {
      "epoch": 0.13538408808534338,
      "grad_norm": 3.6408872604370117,
      "learning_rate": 9.981921054307469e-05,
      "loss": 2.79071044921875,
      "memory(GiB)": 47.63,
      "step": 3160,
      "token_acc": 0.44886363636363635,
      "train_speed(iter/s)": 1.457951
    },
    {
      "epoch": 0.13559830341459234,
      "grad_norm": 3.7442562580108643,
      "learning_rate": 9.981863831830723e-05,
      "loss": 2.481209564208984,
      "memory(GiB)": 47.63,
      "step": 3165,
      "token_acc": 0.4968553459119497,
      "train_speed(iter/s)": 1.458041
    },
    {
      "epoch": 0.13581251874384132,
      "grad_norm": 3.842127799987793,
      "learning_rate": 9.98180651910284e-05,
      "loss": 2.656930923461914,
      "memory(GiB)": 47.63,
      "step": 3170,
      "token_acc": 0.45209580838323354,
      "train_speed(iter/s)": 1.45861
    },
    {
      "epoch": 0.13602673407309027,
      "grad_norm": 5.372028827667236,
      "learning_rate": 9.981749116124859e-05,
      "loss": 2.739255905151367,
      "memory(GiB)": 47.63,
      "step": 3175,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.458852
    },
    {
      "epoch": 0.13624094940233922,
      "grad_norm": 4.918059349060059,
      "learning_rate": 9.981691622897818e-05,
      "loss": 2.712474060058594,
      "memory(GiB)": 47.63,
      "step": 3180,
      "token_acc": 0.4389438943894389,
      "train_speed(iter/s)": 1.458461
    },
    {
      "epoch": 0.1364551647315882,
      "grad_norm": 3.887791156768799,
      "learning_rate": 9.981634039422761e-05,
      "loss": 2.9390422821044924,
      "memory(GiB)": 47.63,
      "step": 3185,
      "token_acc": 0.3894080996884735,
      "train_speed(iter/s)": 1.458451
    },
    {
      "epoch": 0.13666938006083715,
      "grad_norm": 3.610445976257324,
      "learning_rate": 9.98157636570073e-05,
      "loss": 2.300612449645996,
      "memory(GiB)": 47.63,
      "step": 3190,
      "token_acc": 0.5015479876160991,
      "train_speed(iter/s)": 1.458752
    },
    {
      "epoch": 0.13688359539008613,
      "grad_norm": 3.383845567703247,
      "learning_rate": 9.981518601732771e-05,
      "loss": 2.4171300888061524,
      "memory(GiB)": 47.63,
      "step": 3195,
      "token_acc": 0.5226480836236934,
      "train_speed(iter/s)": 1.458971
    },
    {
      "epoch": 0.13709781071933508,
      "grad_norm": 3.146437644958496,
      "learning_rate": 9.981460747519928e-05,
      "loss": 2.724361801147461,
      "memory(GiB)": 47.63,
      "step": 3200,
      "token_acc": 0.4158730158730159,
      "train_speed(iter/s)": 1.459388
    },
    {
      "epoch": 0.13731202604858403,
      "grad_norm": 3.7056562900543213,
      "learning_rate": 9.98140280306325e-05,
      "loss": 2.829375076293945,
      "memory(GiB)": 47.63,
      "step": 3205,
      "token_acc": 0.4489795918367347,
      "train_speed(iter/s)": 1.459722
    },
    {
      "epoch": 0.137526241377833,
      "grad_norm": 4.848848342895508,
      "learning_rate": 9.981344768363791e-05,
      "loss": 2.6239721298217775,
      "memory(GiB)": 47.63,
      "step": 3210,
      "token_acc": 0.43859649122807015,
      "train_speed(iter/s)": 1.458474
    },
    {
      "epoch": 0.13774045670708196,
      "grad_norm": 3.3288214206695557,
      "learning_rate": 9.981286643422596e-05,
      "loss": 2.5134204864501952,
      "memory(GiB)": 47.63,
      "step": 3215,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.459086
    },
    {
      "epoch": 0.1379546720363309,
      "grad_norm": 4.821554183959961,
      "learning_rate": 9.981228428240721e-05,
      "loss": 2.6251523971557615,
      "memory(GiB)": 47.63,
      "step": 3220,
      "token_acc": 0.4580152671755725,
      "train_speed(iter/s)": 1.45836
    },
    {
      "epoch": 0.1381688873655799,
      "grad_norm": 3.739990472793579,
      "learning_rate": 9.98117012281922e-05,
      "loss": 2.6130451202392577,
      "memory(GiB)": 47.63,
      "step": 3225,
      "token_acc": 0.49174917491749176,
      "train_speed(iter/s)": 1.458788
    },
    {
      "epoch": 0.13838310269482884,
      "grad_norm": 3.886059284210205,
      "learning_rate": 9.98111172715915e-05,
      "loss": 2.6881637573242188,
      "memory(GiB)": 47.63,
      "step": 3230,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.459028
    },
    {
      "epoch": 0.1385973180240778,
      "grad_norm": 3.547057867050171,
      "learning_rate": 9.981053241261567e-05,
      "loss": 2.7989946365356446,
      "memory(GiB)": 47.63,
      "step": 3235,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.458744
    },
    {
      "epoch": 0.13881153335332677,
      "grad_norm": 3.5991673469543457,
      "learning_rate": 9.980994665127535e-05,
      "loss": 2.447433853149414,
      "memory(GiB)": 47.63,
      "step": 3240,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.45873
    },
    {
      "epoch": 0.13902574868257572,
      "grad_norm": 4.359513282775879,
      "learning_rate": 9.980935998758109e-05,
      "loss": 2.5521196365356444,
      "memory(GiB)": 47.63,
      "step": 3245,
      "token_acc": 0.46946564885496184,
      "train_speed(iter/s)": 1.458775
    },
    {
      "epoch": 0.13923996401182467,
      "grad_norm": 3.155566692352295,
      "learning_rate": 9.980877242154356e-05,
      "loss": 2.5134252548217773,
      "memory(GiB)": 47.63,
      "step": 3250,
      "token_acc": 0.48328267477203646,
      "train_speed(iter/s)": 1.458647
    },
    {
      "epoch": 0.13945417934107365,
      "grad_norm": 3.671736717224121,
      "learning_rate": 9.980818395317339e-05,
      "loss": 2.5342523574829103,
      "memory(GiB)": 47.63,
      "step": 3255,
      "token_acc": 0.46886446886446886,
      "train_speed(iter/s)": 1.458844
    },
    {
      "epoch": 0.1396683946703226,
      "grad_norm": 2.9696121215820312,
      "learning_rate": 9.980759458248125e-05,
      "loss": 2.466375732421875,
      "memory(GiB)": 47.63,
      "step": 3260,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.458919
    },
    {
      "epoch": 0.13988260999957156,
      "grad_norm": 4.028738975524902,
      "learning_rate": 9.98070043094778e-05,
      "loss": 2.8555065155029298,
      "memory(GiB)": 47.63,
      "step": 3265,
      "token_acc": 0.45426829268292684,
      "train_speed(iter/s)": 1.458649
    },
    {
      "epoch": 0.14009682532882053,
      "grad_norm": 3.731355667114258,
      "learning_rate": 9.980641313417376e-05,
      "loss": 2.763540267944336,
      "memory(GiB)": 47.63,
      "step": 3270,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.458694
    },
    {
      "epoch": 0.14031104065806949,
      "grad_norm": 4.28919792175293,
      "learning_rate": 9.98058210565798e-05,
      "loss": 2.7386585235595704,
      "memory(GiB)": 47.63,
      "step": 3275,
      "token_acc": 0.46204620462046203,
      "train_speed(iter/s)": 1.45912
    },
    {
      "epoch": 0.14052525598731846,
      "grad_norm": 4.4661970138549805,
      "learning_rate": 9.980522807670669e-05,
      "loss": 2.326940155029297,
      "memory(GiB)": 47.63,
      "step": 3280,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.459468
    },
    {
      "epoch": 0.14073947131656742,
      "grad_norm": 4.619133949279785,
      "learning_rate": 9.980463419456516e-05,
      "loss": 2.806599998474121,
      "memory(GiB)": 47.63,
      "step": 3285,
      "token_acc": 0.4358974358974359,
      "train_speed(iter/s)": 1.46012
    },
    {
      "epoch": 0.14095368664581637,
      "grad_norm": 4.316283702850342,
      "learning_rate": 9.980403941016594e-05,
      "loss": 2.7667058944702148,
      "memory(GiB)": 47.63,
      "step": 3290,
      "token_acc": 0.4435146443514644,
      "train_speed(iter/s)": 1.46037
    },
    {
      "epoch": 0.14116790197506535,
      "grad_norm": 3.754873752593994,
      "learning_rate": 9.980344372351983e-05,
      "loss": 2.634560775756836,
      "memory(GiB)": 47.63,
      "step": 3295,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.460697
    },
    {
      "epoch": 0.1413821173043143,
      "grad_norm": 4.2558913230896,
      "learning_rate": 9.980284713463763e-05,
      "loss": 2.62609806060791,
      "memory(GiB)": 47.63,
      "step": 3300,
      "token_acc": 0.47876447876447875,
      "train_speed(iter/s)": 1.461006
    },
    {
      "epoch": 0.14159633263356325,
      "grad_norm": 4.759919166564941,
      "learning_rate": 9.980224964353014e-05,
      "loss": 2.6200397491455076,
      "memory(GiB)": 47.63,
      "step": 3305,
      "token_acc": 0.4217687074829932,
      "train_speed(iter/s)": 1.4612
    },
    {
      "epoch": 0.14181054796281223,
      "grad_norm": 3.752591848373413,
      "learning_rate": 9.980165125020816e-05,
      "loss": 2.603768539428711,
      "memory(GiB)": 47.63,
      "step": 3310,
      "token_acc": 0.4491803278688525,
      "train_speed(iter/s)": 1.461194
    },
    {
      "epoch": 0.14202476329206118,
      "grad_norm": 3.5088789463043213,
      "learning_rate": 9.980105195468256e-05,
      "loss": 2.9116655349731446,
      "memory(GiB)": 47.63,
      "step": 3315,
      "token_acc": 0.41785714285714287,
      "train_speed(iter/s)": 1.460879
    },
    {
      "epoch": 0.14223897862131013,
      "grad_norm": 5.019904613494873,
      "learning_rate": 9.980045175696418e-05,
      "loss": 2.826046371459961,
      "memory(GiB)": 47.63,
      "step": 3320,
      "token_acc": 0.4197183098591549,
      "train_speed(iter/s)": 1.460983
    },
    {
      "epoch": 0.1424531939505591,
      "grad_norm": 4.364889621734619,
      "learning_rate": 9.97998506570639e-05,
      "loss": 2.498710823059082,
      "memory(GiB)": 47.63,
      "step": 3325,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.461307
    },
    {
      "epoch": 0.14266740927980806,
      "grad_norm": 2.8879406452178955,
      "learning_rate": 9.979924865499262e-05,
      "loss": 2.739917182922363,
      "memory(GiB)": 47.63,
      "step": 3330,
      "token_acc": 0.44668587896253603,
      "train_speed(iter/s)": 1.461107
    },
    {
      "epoch": 0.142881624609057,
      "grad_norm": 3.873210906982422,
      "learning_rate": 9.979864575076124e-05,
      "loss": 2.6274206161499025,
      "memory(GiB)": 47.63,
      "step": 3335,
      "token_acc": 0.4588235294117647,
      "train_speed(iter/s)": 1.461477
    },
    {
      "epoch": 0.143095839938306,
      "grad_norm": 3.0949182510375977,
      "learning_rate": 9.979804194438065e-05,
      "loss": 2.3082038879394533,
      "memory(GiB)": 47.63,
      "step": 3340,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.461001
    },
    {
      "epoch": 0.14331005526755494,
      "grad_norm": 4.63090181350708,
      "learning_rate": 9.979743723586184e-05,
      "loss": 2.840785598754883,
      "memory(GiB)": 47.63,
      "step": 3345,
      "token_acc": 0.422360248447205,
      "train_speed(iter/s)": 1.460772
    },
    {
      "epoch": 0.1435242705968039,
      "grad_norm": 3.684236526489258,
      "learning_rate": 9.979683162521573e-05,
      "loss": 2.8608634948730467,
      "memory(GiB)": 47.63,
      "step": 3350,
      "token_acc": 0.4337748344370861,
      "train_speed(iter/s)": 1.461542
    },
    {
      "epoch": 0.14373848592605287,
      "grad_norm": 4.538915634155273,
      "learning_rate": 9.97962251124533e-05,
      "loss": 2.7592752456665037,
      "memory(GiB)": 47.63,
      "step": 3355,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.461603
    },
    {
      "epoch": 0.14395270125530182,
      "grad_norm": 4.658306121826172,
      "learning_rate": 9.979561769758555e-05,
      "loss": 2.6472780227661135,
      "memory(GiB)": 47.63,
      "step": 3360,
      "token_acc": 0.4359861591695502,
      "train_speed(iter/s)": 1.461189
    },
    {
      "epoch": 0.1441669165845508,
      "grad_norm": 3.5684494972229004,
      "learning_rate": 9.979500938062345e-05,
      "loss": 2.387424087524414,
      "memory(GiB)": 47.63,
      "step": 3365,
      "token_acc": 0.48134328358208955,
      "train_speed(iter/s)": 1.461779
    },
    {
      "epoch": 0.14438113191379975,
      "grad_norm": 6.197642803192139,
      "learning_rate": 9.979440016157805e-05,
      "loss": 2.476662445068359,
      "memory(GiB)": 47.63,
      "step": 3370,
      "token_acc": 0.43309859154929575,
      "train_speed(iter/s)": 1.461999
    },
    {
      "epoch": 0.1445953472430487,
      "grad_norm": 3.708441972732544,
      "learning_rate": 9.97937900404604e-05,
      "loss": 2.873057174682617,
      "memory(GiB)": 47.63,
      "step": 3375,
      "token_acc": 0.43354430379746833,
      "train_speed(iter/s)": 1.462099
    },
    {
      "epoch": 0.14480956257229768,
      "grad_norm": 4.045034408569336,
      "learning_rate": 9.979317901728153e-05,
      "loss": 2.8710140228271483,
      "memory(GiB)": 47.63,
      "step": 3380,
      "token_acc": 0.45936395759717313,
      "train_speed(iter/s)": 1.462166
    },
    {
      "epoch": 0.14502377790154664,
      "grad_norm": 3.011522054672241,
      "learning_rate": 9.979256709205251e-05,
      "loss": 2.744778633117676,
      "memory(GiB)": 47.63,
      "step": 3385,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.461826
    },
    {
      "epoch": 0.1452379932307956,
      "grad_norm": 3.169530153274536,
      "learning_rate": 9.979195426478443e-05,
      "loss": 2.4550779342651365,
      "memory(GiB)": 47.63,
      "step": 3390,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.462158
    },
    {
      "epoch": 0.14545220856004457,
      "grad_norm": 4.8921380043029785,
      "learning_rate": 9.979134053548838e-05,
      "loss": 2.5971981048583985,
      "memory(GiB)": 47.63,
      "step": 3395,
      "token_acc": 0.3935018050541516,
      "train_speed(iter/s)": 1.461687
    },
    {
      "epoch": 0.14566642388929352,
      "grad_norm": 3.600477695465088,
      "learning_rate": 9.979072590417549e-05,
      "loss": 2.3302986145019533,
      "memory(GiB)": 47.63,
      "step": 3400,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.462275
    },
    {
      "epoch": 0.14588063921854247,
      "grad_norm": 6.550259590148926,
      "learning_rate": 9.97901103708569e-05,
      "loss": 2.7483451843261717,
      "memory(GiB)": 47.63,
      "step": 3405,
      "token_acc": 0.48120300751879697,
      "train_speed(iter/s)": 1.462555
    },
    {
      "epoch": 0.14609485454779145,
      "grad_norm": 4.765453815460205,
      "learning_rate": 9.978949393554374e-05,
      "loss": 2.8481876373291017,
      "memory(GiB)": 47.63,
      "step": 3410,
      "token_acc": 0.4467455621301775,
      "train_speed(iter/s)": 1.462951
    },
    {
      "epoch": 0.1463090698770404,
      "grad_norm": 4.505581378936768,
      "learning_rate": 9.978887659824721e-05,
      "loss": 2.6652400970458983,
      "memory(GiB)": 47.63,
      "step": 3415,
      "token_acc": 0.48398576512455516,
      "train_speed(iter/s)": 1.462657
    },
    {
      "epoch": 0.14652328520628935,
      "grad_norm": 3.9403669834136963,
      "learning_rate": 9.978825835897844e-05,
      "loss": 2.4215864181518554,
      "memory(GiB)": 47.63,
      "step": 3420,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.462608
    },
    {
      "epoch": 0.14673750053553833,
      "grad_norm": 4.667635440826416,
      "learning_rate": 9.978763921774869e-05,
      "loss": 2.425082206726074,
      "memory(GiB)": 47.63,
      "step": 3425,
      "token_acc": 0.4536082474226804,
      "train_speed(iter/s)": 1.462776
    },
    {
      "epoch": 0.14695171586478728,
      "grad_norm": 3.6880412101745605,
      "learning_rate": 9.978701917456916e-05,
      "loss": 2.4595867156982423,
      "memory(GiB)": 47.63,
      "step": 3430,
      "token_acc": 0.49258160237388726,
      "train_speed(iter/s)": 1.462755
    },
    {
      "epoch": 0.14716593119403626,
      "grad_norm": 3.2978837490081787,
      "learning_rate": 9.978639822945107e-05,
      "loss": 2.511236000061035,
      "memory(GiB)": 47.63,
      "step": 3435,
      "token_acc": 0.4511784511784512,
      "train_speed(iter/s)": 1.46316
    },
    {
      "epoch": 0.1473801465232852,
      "grad_norm": 3.917314291000366,
      "learning_rate": 9.978577638240567e-05,
      "loss": 2.7325916290283203,
      "memory(GiB)": 47.63,
      "step": 3440,
      "token_acc": 0.4581818181818182,
      "train_speed(iter/s)": 1.463603
    },
    {
      "epoch": 0.14759436185253416,
      "grad_norm": 3.8789377212524414,
      "learning_rate": 9.978515363344422e-05,
      "loss": 2.5562042236328124,
      "memory(GiB)": 47.63,
      "step": 3445,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.464349
    },
    {
      "epoch": 0.14780857718178314,
      "grad_norm": 7.125819206237793,
      "learning_rate": 9.978452998257801e-05,
      "loss": 2.5441293716430664,
      "memory(GiB)": 47.63,
      "step": 3450,
      "token_acc": 0.44871794871794873,
      "train_speed(iter/s)": 1.464385
    },
    {
      "epoch": 0.1480227925110321,
      "grad_norm": 4.757804870605469,
      "learning_rate": 9.978390542981835e-05,
      "loss": 2.660295104980469,
      "memory(GiB)": 47.63,
      "step": 3455,
      "token_acc": 0.43555555555555553,
      "train_speed(iter/s)": 1.464718
    },
    {
      "epoch": 0.14823700784028104,
      "grad_norm": 4.2779107093811035,
      "learning_rate": 9.978327997517652e-05,
      "loss": 2.7758384704589845,
      "memory(GiB)": 47.63,
      "step": 3460,
      "token_acc": 0.43795620437956206,
      "train_speed(iter/s)": 1.465205
    },
    {
      "epoch": 0.14845122316953002,
      "grad_norm": 3.928638219833374,
      "learning_rate": 9.978265361866389e-05,
      "loss": 2.592887115478516,
      "memory(GiB)": 47.63,
      "step": 3465,
      "token_acc": 0.4823943661971831,
      "train_speed(iter/s)": 1.465435
    },
    {
      "epoch": 0.14866543849877897,
      "grad_norm": 3.746091842651367,
      "learning_rate": 9.978202636029179e-05,
      "loss": 2.6098140716552733,
      "memory(GiB)": 47.63,
      "step": 3470,
      "token_acc": 0.4689265536723164,
      "train_speed(iter/s)": 1.465081
    },
    {
      "epoch": 0.14887965382802792,
      "grad_norm": 4.820042610168457,
      "learning_rate": 9.978139820007158e-05,
      "loss": 2.5279256820678713,
      "memory(GiB)": 47.63,
      "step": 3475,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.464572
    },
    {
      "epoch": 0.1490938691572769,
      "grad_norm": 4.189342975616455,
      "learning_rate": 9.978076913801464e-05,
      "loss": 2.3815740585327148,
      "memory(GiB)": 47.63,
      "step": 3480,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.464616
    },
    {
      "epoch": 0.14930808448652585,
      "grad_norm": 6.875144958496094,
      "learning_rate": 9.978013917413237e-05,
      "loss": 2.633605194091797,
      "memory(GiB)": 47.63,
      "step": 3485,
      "token_acc": 0.4875,
      "train_speed(iter/s)": 1.465166
    },
    {
      "epoch": 0.1495222998157748,
      "grad_norm": 2.6285159587860107,
      "learning_rate": 9.977950830843619e-05,
      "loss": 2.563618850708008,
      "memory(GiB)": 47.63,
      "step": 3490,
      "token_acc": 0.4487534626038781,
      "train_speed(iter/s)": 1.465612
    },
    {
      "epoch": 0.14973651514502379,
      "grad_norm": 7.238272190093994,
      "learning_rate": 9.977887654093751e-05,
      "loss": 2.6613895416259767,
      "memory(GiB)": 47.63,
      "step": 3495,
      "token_acc": 0.48398576512455516,
      "train_speed(iter/s)": 1.465451
    },
    {
      "epoch": 0.14995073047427274,
      "grad_norm": 3.0726985931396484,
      "learning_rate": 9.977824387164778e-05,
      "loss": 2.343222999572754,
      "memory(GiB)": 47.63,
      "step": 3500,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.465652
    },
    {
      "epoch": 0.14995073047427274,
      "eval_loss": 2.3537485599517822,
      "eval_runtime": 14.0822,
      "eval_samples_per_second": 7.101,
      "eval_steps_per_second": 7.101,
      "eval_token_acc": 0.4715447154471545,
      "step": 3500
    },
    {
      "epoch": 0.1501649458035217,
      "grad_norm": 3.4169580936431885,
      "learning_rate": 9.977761030057847e-05,
      "loss": 2.290047836303711,
      "memory(GiB)": 47.63,
      "step": 3505,
      "token_acc": 0.49447236180904525,
      "train_speed(iter/s)": 1.456895
    },
    {
      "epoch": 0.15037916113277067,
      "grad_norm": 3.6757216453552246,
      "learning_rate": 9.977697582774106e-05,
      "loss": 2.4777524948120115,
      "memory(GiB)": 47.63,
      "step": 3510,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.456862
    },
    {
      "epoch": 0.15059337646201962,
      "grad_norm": 4.6084513664245605,
      "learning_rate": 9.977634045314703e-05,
      "loss": 2.6394153594970704,
      "memory(GiB)": 47.63,
      "step": 3515,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.456717
    },
    {
      "epoch": 0.1508075917912686,
      "grad_norm": 3.7747673988342285,
      "learning_rate": 9.977570417680791e-05,
      "loss": 2.7754934310913084,
      "memory(GiB)": 47.63,
      "step": 3520,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.45684
    },
    {
      "epoch": 0.15102180712051755,
      "grad_norm": 3.9443018436431885,
      "learning_rate": 9.977506699873521e-05,
      "loss": 2.4865036010742188,
      "memory(GiB)": 47.63,
      "step": 3525,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.457102
    },
    {
      "epoch": 0.1512360224497665,
      "grad_norm": 3.8653881549835205,
      "learning_rate": 9.977442891894048e-05,
      "loss": 2.498517608642578,
      "memory(GiB)": 47.63,
      "step": 3530,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.457026
    },
    {
      "epoch": 0.15145023777901548,
      "grad_norm": 3.7123584747314453,
      "learning_rate": 9.977378993743528e-05,
      "loss": 2.8637248992919924,
      "memory(GiB)": 47.63,
      "step": 3535,
      "token_acc": 0.4596774193548387,
      "train_speed(iter/s)": 1.457719
    },
    {
      "epoch": 0.15166445310826443,
      "grad_norm": 4.383836269378662,
      "learning_rate": 9.977315005423117e-05,
      "loss": 2.6163511276245117,
      "memory(GiB)": 47.63,
      "step": 3540,
      "token_acc": 0.4537037037037037,
      "train_speed(iter/s)": 1.457578
    },
    {
      "epoch": 0.15187866843751338,
      "grad_norm": 2.944066047668457,
      "learning_rate": 9.977250926933977e-05,
      "loss": 2.6205913543701174,
      "memory(GiB)": 47.63,
      "step": 3545,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.457971
    },
    {
      "epoch": 0.15209288376676236,
      "grad_norm": 3.992919445037842,
      "learning_rate": 9.977186758277268e-05,
      "loss": 2.6933469772338867,
      "memory(GiB)": 47.63,
      "step": 3550,
      "token_acc": 0.41818181818181815,
      "train_speed(iter/s)": 1.457775
    },
    {
      "epoch": 0.1523070990960113,
      "grad_norm": 7.702213764190674,
      "learning_rate": 9.97712249945415e-05,
      "loss": 2.7108774185180664,
      "memory(GiB)": 47.63,
      "step": 3555,
      "token_acc": 0.4452054794520548,
      "train_speed(iter/s)": 1.457589
    },
    {
      "epoch": 0.15252131442526026,
      "grad_norm": 3.4477903842926025,
      "learning_rate": 9.97705815046579e-05,
      "loss": 2.6474851608276366,
      "memory(GiB)": 47.63,
      "step": 3560,
      "token_acc": 0.45357142857142857,
      "train_speed(iter/s)": 1.457768
    },
    {
      "epoch": 0.15273552975450924,
      "grad_norm": 3.7272629737854004,
      "learning_rate": 9.976993711313354e-05,
      "loss": 2.744158935546875,
      "memory(GiB)": 47.63,
      "step": 3565,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.458
    },
    {
      "epoch": 0.1529497450837582,
      "grad_norm": 3.5521841049194336,
      "learning_rate": 9.976929181998007e-05,
      "loss": 2.6562219619750977,
      "memory(GiB)": 47.63,
      "step": 3570,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.458203
    },
    {
      "epoch": 0.15316396041300714,
      "grad_norm": 4.787902355194092,
      "learning_rate": 9.976864562520918e-05,
      "loss": 2.75551815032959,
      "memory(GiB)": 47.63,
      "step": 3575,
      "token_acc": 0.4262948207171315,
      "train_speed(iter/s)": 1.458356
    },
    {
      "epoch": 0.15337817574225612,
      "grad_norm": 2.84358286857605,
      "learning_rate": 9.976799852883261e-05,
      "loss": 2.6625064849853515,
      "memory(GiB)": 47.63,
      "step": 3580,
      "token_acc": 0.4413793103448276,
      "train_speed(iter/s)": 1.458068
    },
    {
      "epoch": 0.15359239107150507,
      "grad_norm": 4.125617027282715,
      "learning_rate": 9.976735053086204e-05,
      "loss": 2.428995895385742,
      "memory(GiB)": 47.63,
      "step": 3585,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.457472
    },
    {
      "epoch": 0.15380660640075403,
      "grad_norm": 3.760828733444214,
      "learning_rate": 9.976670163130924e-05,
      "loss": 2.9189868927001954,
      "memory(GiB)": 47.63,
      "step": 3590,
      "token_acc": 0.4415204678362573,
      "train_speed(iter/s)": 1.45773
    },
    {
      "epoch": 0.154020821730003,
      "grad_norm": 3.6647214889526367,
      "learning_rate": 9.976605183018594e-05,
      "loss": 2.695289421081543,
      "memory(GiB)": 47.63,
      "step": 3595,
      "token_acc": 0.43573667711598746,
      "train_speed(iter/s)": 1.458066
    },
    {
      "epoch": 0.15423503705925196,
      "grad_norm": 3.5985748767852783,
      "learning_rate": 9.976540112750394e-05,
      "loss": 2.5852956771850586,
      "memory(GiB)": 47.63,
      "step": 3600,
      "token_acc": 0.4532871972318339,
      "train_speed(iter/s)": 1.457399
    },
    {
      "epoch": 0.15444925238850093,
      "grad_norm": 3.0911688804626465,
      "learning_rate": 9.976474952327502e-05,
      "loss": 2.802461051940918,
      "memory(GiB)": 47.63,
      "step": 3605,
      "token_acc": 0.4758364312267658,
      "train_speed(iter/s)": 1.457697
    },
    {
      "epoch": 0.1546634677177499,
      "grad_norm": 4.63688325881958,
      "learning_rate": 9.976409701751097e-05,
      "loss": 2.487857627868652,
      "memory(GiB)": 47.63,
      "step": 3610,
      "token_acc": 0.48398576512455516,
      "train_speed(iter/s)": 1.457331
    },
    {
      "epoch": 0.15487768304699884,
      "grad_norm": 3.1934165954589844,
      "learning_rate": 9.976344361022363e-05,
      "loss": 2.3247928619384766,
      "memory(GiB)": 47.63,
      "step": 3615,
      "token_acc": 0.5115384615384615,
      "train_speed(iter/s)": 1.457743
    },
    {
      "epoch": 0.15509189837624782,
      "grad_norm": 3.4557573795318604,
      "learning_rate": 9.976278930142482e-05,
      "loss": 2.5618444442749024,
      "memory(GiB)": 47.63,
      "step": 3620,
      "token_acc": 0.49224806201550386,
      "train_speed(iter/s)": 1.458065
    },
    {
      "epoch": 0.15530611370549677,
      "grad_norm": 4.077261447906494,
      "learning_rate": 9.97621340911264e-05,
      "loss": 2.6343793869018555,
      "memory(GiB)": 47.63,
      "step": 3625,
      "token_acc": 0.43944636678200694,
      "train_speed(iter/s)": 1.458295
    },
    {
      "epoch": 0.15552032903474572,
      "grad_norm": 3.412285566329956,
      "learning_rate": 9.976147797934024e-05,
      "loss": 2.5630825042724608,
      "memory(GiB)": 47.63,
      "step": 3630,
      "token_acc": 0.44375,
      "train_speed(iter/s)": 1.457827
    },
    {
      "epoch": 0.1557345443639947,
      "grad_norm": 3.526050329208374,
      "learning_rate": 9.976082096607823e-05,
      "loss": 2.741041564941406,
      "memory(GiB)": 47.63,
      "step": 3635,
      "token_acc": 0.4104938271604938,
      "train_speed(iter/s)": 1.457739
    },
    {
      "epoch": 0.15594875969324365,
      "grad_norm": 5.315544605255127,
      "learning_rate": 9.976016305135228e-05,
      "loss": 2.279109573364258,
      "memory(GiB)": 47.63,
      "step": 3640,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.457882
    },
    {
      "epoch": 0.1561629750224926,
      "grad_norm": 3.2124979496002197,
      "learning_rate": 9.975950423517426e-05,
      "loss": 2.6411577224731446,
      "memory(GiB)": 47.63,
      "step": 3645,
      "token_acc": 0.45182724252491696,
      "train_speed(iter/s)": 1.4577
    },
    {
      "epoch": 0.15637719035174158,
      "grad_norm": 7.270047187805176,
      "learning_rate": 9.975884451755618e-05,
      "loss": 2.6146183013916016,
      "memory(GiB)": 47.63,
      "step": 3650,
      "token_acc": 0.44983818770226536,
      "train_speed(iter/s)": 1.457684
    },
    {
      "epoch": 0.15659140568099053,
      "grad_norm": 3.6394429206848145,
      "learning_rate": 9.975818389850993e-05,
      "loss": 2.5720624923706055,
      "memory(GiB)": 47.63,
      "step": 3655,
      "token_acc": 0.4732824427480916,
      "train_speed(iter/s)": 1.457808
    },
    {
      "epoch": 0.15680562101023948,
      "grad_norm": 3.7785966396331787,
      "learning_rate": 9.975752237804752e-05,
      "loss": 2.522658920288086,
      "memory(GiB)": 47.63,
      "step": 3660,
      "token_acc": 0.43006993006993005,
      "train_speed(iter/s)": 1.457458
    },
    {
      "epoch": 0.15701983633948846,
      "grad_norm": 3.6492886543273926,
      "learning_rate": 9.975685995618092e-05,
      "loss": 2.747486686706543,
      "memory(GiB)": 47.63,
      "step": 3665,
      "token_acc": 0.4624505928853755,
      "train_speed(iter/s)": 1.457678
    },
    {
      "epoch": 0.1572340516687374,
      "grad_norm": 4.997968673706055,
      "learning_rate": 9.97561966329221e-05,
      "loss": 2.707684135437012,
      "memory(GiB)": 47.63,
      "step": 3670,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.457882
    },
    {
      "epoch": 0.15744826699798636,
      "grad_norm": 2.9105184078216553,
      "learning_rate": 9.975553240828312e-05,
      "loss": 2.459530258178711,
      "memory(GiB)": 47.63,
      "step": 3675,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.458198
    },
    {
      "epoch": 0.15766248232723534,
      "grad_norm": 3.2676429748535156,
      "learning_rate": 9.975486728227599e-05,
      "loss": 2.582370567321777,
      "memory(GiB)": 47.63,
      "step": 3680,
      "token_acc": 0.4420485175202156,
      "train_speed(iter/s)": 1.458089
    },
    {
      "epoch": 0.1578766976564843,
      "grad_norm": 4.302821159362793,
      "learning_rate": 9.975420125491277e-05,
      "loss": 2.7656349182128905,
      "memory(GiB)": 47.63,
      "step": 3685,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.45836
    },
    {
      "epoch": 0.15809091298573327,
      "grad_norm": 3.3045613765716553,
      "learning_rate": 9.975353432620552e-05,
      "loss": 2.489249038696289,
      "memory(GiB)": 47.63,
      "step": 3690,
      "token_acc": 0.4940239043824701,
      "train_speed(iter/s)": 1.458707
    },
    {
      "epoch": 0.15830512831498222,
      "grad_norm": 5.259227275848389,
      "learning_rate": 9.975286649616633e-05,
      "loss": 2.4811010360717773,
      "memory(GiB)": 47.63,
      "step": 3695,
      "token_acc": 0.48739495798319327,
      "train_speed(iter/s)": 1.458281
    },
    {
      "epoch": 0.15851934364423118,
      "grad_norm": 5.344473361968994,
      "learning_rate": 9.975219776480727e-05,
      "loss": 2.438692092895508,
      "memory(GiB)": 47.63,
      "step": 3700,
      "token_acc": 0.5073529411764706,
      "train_speed(iter/s)": 1.458013
    },
    {
      "epoch": 0.15873355897348015,
      "grad_norm": 4.790504455566406,
      "learning_rate": 9.97515281321405e-05,
      "loss": 2.609920120239258,
      "memory(GiB)": 47.63,
      "step": 3705,
      "token_acc": 0.4855072463768116,
      "train_speed(iter/s)": 1.458119
    },
    {
      "epoch": 0.1589477743027291,
      "grad_norm": 3.29830002784729,
      "learning_rate": 9.975085759817811e-05,
      "loss": 2.884556770324707,
      "memory(GiB)": 47.63,
      "step": 3710,
      "token_acc": 0.4303030303030303,
      "train_speed(iter/s)": 1.457928
    },
    {
      "epoch": 0.15916198963197806,
      "grad_norm": 6.27782678604126,
      "learning_rate": 9.975018616293228e-05,
      "loss": 2.6151010513305666,
      "memory(GiB)": 47.63,
      "step": 3715,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.458229
    },
    {
      "epoch": 0.15937620496122704,
      "grad_norm": 4.553070545196533,
      "learning_rate": 9.974951382641516e-05,
      "loss": 2.7687604904174803,
      "memory(GiB)": 47.63,
      "step": 3720,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.458198
    },
    {
      "epoch": 0.159590420290476,
      "grad_norm": 5.753072261810303,
      "learning_rate": 9.97488405886389e-05,
      "loss": 2.497258758544922,
      "memory(GiB)": 47.63,
      "step": 3725,
      "token_acc": 0.47232472324723246,
      "train_speed(iter/s)": 1.45841
    },
    {
      "epoch": 0.15980463561972494,
      "grad_norm": 5.164744853973389,
      "learning_rate": 9.974816644961575e-05,
      "loss": 2.6047916412353516,
      "memory(GiB)": 47.63,
      "step": 3730,
      "token_acc": 0.42391304347826086,
      "train_speed(iter/s)": 1.459016
    },
    {
      "epoch": 0.16001885094897392,
      "grad_norm": 4.997974395751953,
      "learning_rate": 9.974749140935789e-05,
      "loss": 2.688969612121582,
      "memory(GiB)": 47.63,
      "step": 3735,
      "token_acc": 0.4460431654676259,
      "train_speed(iter/s)": 1.459304
    },
    {
      "epoch": 0.16023306627822287,
      "grad_norm": 3.5237510204315186,
      "learning_rate": 9.974681546787755e-05,
      "loss": 2.5835519790649415,
      "memory(GiB)": 47.63,
      "step": 3740,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.459492
    },
    {
      "epoch": 0.16044728160747182,
      "grad_norm": 4.978376388549805,
      "learning_rate": 9.974613862518699e-05,
      "loss": 2.5920886993408203,
      "memory(GiB)": 47.63,
      "step": 3745,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.459628
    },
    {
      "epoch": 0.1606614969367208,
      "grad_norm": 3.2934536933898926,
      "learning_rate": 9.974546088129846e-05,
      "loss": 2.704619598388672,
      "memory(GiB)": 47.63,
      "step": 3750,
      "token_acc": 0.4688427299703264,
      "train_speed(iter/s)": 1.459236
    },
    {
      "epoch": 0.16087571226596975,
      "grad_norm": 3.007472515106201,
      "learning_rate": 9.974478223622424e-05,
      "loss": 2.632056999206543,
      "memory(GiB)": 47.63,
      "step": 3755,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.459154
    },
    {
      "epoch": 0.1610899275952187,
      "grad_norm": 3.6066036224365234,
      "learning_rate": 9.974410268997662e-05,
      "loss": 2.520039367675781,
      "memory(GiB)": 47.63,
      "step": 3760,
      "token_acc": 0.4674329501915709,
      "train_speed(iter/s)": 1.459432
    },
    {
      "epoch": 0.16130414292446768,
      "grad_norm": 4.061606407165527,
      "learning_rate": 9.974342224256795e-05,
      "loss": 2.6116653442382813,
      "memory(GiB)": 47.63,
      "step": 3765,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.459787
    },
    {
      "epoch": 0.16151835825371663,
      "grad_norm": 3.367438554763794,
      "learning_rate": 9.974274089401048e-05,
      "loss": 2.678365707397461,
      "memory(GiB)": 47.63,
      "step": 3770,
      "token_acc": 0.44664031620553357,
      "train_speed(iter/s)": 1.460093
    },
    {
      "epoch": 0.1617325735829656,
      "grad_norm": 7.6087727546691895,
      "learning_rate": 9.974205864431661e-05,
      "loss": 2.593564033508301,
      "memory(GiB)": 47.63,
      "step": 3775,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.460482
    },
    {
      "epoch": 0.16194678891221456,
      "grad_norm": 4.246761798858643,
      "learning_rate": 9.974137549349869e-05,
      "loss": 2.468999671936035,
      "memory(GiB)": 47.63,
      "step": 3780,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.460402
    },
    {
      "epoch": 0.1621610042414635,
      "grad_norm": 3.301891803741455,
      "learning_rate": 9.97406914415691e-05,
      "loss": 2.7166805267333984,
      "memory(GiB)": 47.63,
      "step": 3785,
      "token_acc": 0.45180722891566266,
      "train_speed(iter/s)": 1.460596
    },
    {
      "epoch": 0.1623752195707125,
      "grad_norm": 4.316013813018799,
      "learning_rate": 9.97400064885402e-05,
      "loss": 2.4938470840454103,
      "memory(GiB)": 47.63,
      "step": 3790,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.460808
    },
    {
      "epoch": 0.16258943489996144,
      "grad_norm": 6.042535781860352,
      "learning_rate": 9.973932063442444e-05,
      "loss": 2.47210750579834,
      "memory(GiB)": 47.63,
      "step": 3795,
      "token_acc": 0.5083333333333333,
      "train_speed(iter/s)": 1.461412
    },
    {
      "epoch": 0.1628036502292104,
      "grad_norm": 3.7296314239501953,
      "learning_rate": 9.973863387923423e-05,
      "loss": 2.7515296936035156,
      "memory(GiB)": 47.63,
      "step": 3800,
      "token_acc": 0.4540059347181009,
      "train_speed(iter/s)": 1.461468
    },
    {
      "epoch": 0.16301786555845937,
      "grad_norm": 3.433135747909546,
      "learning_rate": 9.9737946222982e-05,
      "loss": 2.6556316375732423,
      "memory(GiB)": 47.63,
      "step": 3805,
      "token_acc": 0.4712328767123288,
      "train_speed(iter/s)": 1.461323
    },
    {
      "epoch": 0.16323208088770833,
      "grad_norm": 3.6524879932403564,
      "learning_rate": 9.973725766568023e-05,
      "loss": 2.7036582946777346,
      "memory(GiB)": 47.63,
      "step": 3810,
      "token_acc": 0.4463087248322148,
      "train_speed(iter/s)": 1.461134
    },
    {
      "epoch": 0.16344629621695728,
      "grad_norm": 3.3620219230651855,
      "learning_rate": 9.973656820734136e-05,
      "loss": 2.4934627532958986,
      "memory(GiB)": 47.63,
      "step": 3815,
      "token_acc": 0.5117056856187291,
      "train_speed(iter/s)": 1.461772
    },
    {
      "epoch": 0.16366051154620626,
      "grad_norm": 3.825281858444214,
      "learning_rate": 9.973587784797791e-05,
      "loss": 2.373024749755859,
      "memory(GiB)": 47.63,
      "step": 3820,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.462295
    },
    {
      "epoch": 0.1638747268754552,
      "grad_norm": 3.9945309162139893,
      "learning_rate": 9.973518658760236e-05,
      "loss": 2.5321113586425783,
      "memory(GiB)": 47.63,
      "step": 3825,
      "token_acc": 0.4844961240310077,
      "train_speed(iter/s)": 1.462047
    },
    {
      "epoch": 0.16408894220470416,
      "grad_norm": 3.375340223312378,
      "learning_rate": 9.973449442622729e-05,
      "loss": 3.001796913146973,
      "memory(GiB)": 47.63,
      "step": 3830,
      "token_acc": 0.40641711229946526,
      "train_speed(iter/s)": 1.461967
    },
    {
      "epoch": 0.16430315753395314,
      "grad_norm": 4.272892475128174,
      "learning_rate": 9.973380136386516e-05,
      "loss": 2.889964294433594,
      "memory(GiB)": 47.63,
      "step": 3835,
      "token_acc": 0.44964028776978415,
      "train_speed(iter/s)": 1.462502
    },
    {
      "epoch": 0.1645173728632021,
      "grad_norm": 4.154975414276123,
      "learning_rate": 9.973310740052859e-05,
      "loss": 2.9697288513183593,
      "memory(GiB)": 47.63,
      "step": 3840,
      "token_acc": 0.43617021276595747,
      "train_speed(iter/s)": 1.462468
    },
    {
      "epoch": 0.16473158819245104,
      "grad_norm": 3.1028828620910645,
      "learning_rate": 9.973241253623011e-05,
      "loss": 2.0401206970214845,
      "memory(GiB)": 47.63,
      "step": 3845,
      "token_acc": 0.5498154981549815,
      "train_speed(iter/s)": 1.462328
    },
    {
      "epoch": 0.16494580352170002,
      "grad_norm": 8.306822776794434,
      "learning_rate": 9.973171677098233e-05,
      "loss": 2.6747810363769533,
      "memory(GiB)": 47.63,
      "step": 3850,
      "token_acc": 0.44485294117647056,
      "train_speed(iter/s)": 1.462838
    },
    {
      "epoch": 0.16516001885094897,
      "grad_norm": 3.0384585857391357,
      "learning_rate": 9.973102010479784e-05,
      "loss": 2.571843147277832,
      "memory(GiB)": 47.63,
      "step": 3855,
      "token_acc": 0.4664179104477612,
      "train_speed(iter/s)": 1.4628
    },
    {
      "epoch": 0.16537423418019795,
      "grad_norm": 4.891858100891113,
      "learning_rate": 9.973032253768927e-05,
      "loss": 2.5152931213378906,
      "memory(GiB)": 47.63,
      "step": 3860,
      "token_acc": 0.44528301886792454,
      "train_speed(iter/s)": 1.463491
    },
    {
      "epoch": 0.1655884495094469,
      "grad_norm": 5.790170192718506,
      "learning_rate": 9.972962406966928e-05,
      "loss": 3.0690141677856446,
      "memory(GiB)": 47.63,
      "step": 3865,
      "token_acc": 0.38481675392670156,
      "train_speed(iter/s)": 1.463805
    },
    {
      "epoch": 0.16580266483869585,
      "grad_norm": 4.68450403213501,
      "learning_rate": 9.972892470075048e-05,
      "loss": 2.5472696304321287,
      "memory(GiB)": 47.63,
      "step": 3870,
      "token_acc": 0.45864661654135336,
      "train_speed(iter/s)": 1.464278
    },
    {
      "epoch": 0.16601688016794483,
      "grad_norm": 9.12940788269043,
      "learning_rate": 9.972822443094557e-05,
      "loss": 2.8336633682250976,
      "memory(GiB)": 47.63,
      "step": 3875,
      "token_acc": 0.4109090909090909,
      "train_speed(iter/s)": 1.464145
    },
    {
      "epoch": 0.16623109549719378,
      "grad_norm": 3.497802972793579,
      "learning_rate": 9.972752326026722e-05,
      "loss": 2.9579750061035157,
      "memory(GiB)": 47.63,
      "step": 3880,
      "token_acc": 0.47115384615384615,
      "train_speed(iter/s)": 1.463643
    },
    {
      "epoch": 0.16644531082644273,
      "grad_norm": 4.052250862121582,
      "learning_rate": 9.972682118872813e-05,
      "loss": 2.5573352813720702,
      "memory(GiB)": 47.63,
      "step": 3885,
      "token_acc": 0.4524714828897338,
      "train_speed(iter/s)": 1.463967
    },
    {
      "epoch": 0.1666595261556917,
      "grad_norm": 3.2251203060150146,
      "learning_rate": 9.972611821634104e-05,
      "loss": 2.8575668334960938,
      "memory(GiB)": 47.63,
      "step": 3890,
      "token_acc": 0.44807121661721067,
      "train_speed(iter/s)": 1.463968
    },
    {
      "epoch": 0.16687374148494066,
      "grad_norm": 3.248079538345337,
      "learning_rate": 9.972541434311866e-05,
      "loss": 2.5865352630615233,
      "memory(GiB)": 47.63,
      "step": 3895,
      "token_acc": 0.4626334519572954,
      "train_speed(iter/s)": 1.46376
    },
    {
      "epoch": 0.16708795681418961,
      "grad_norm": 4.117458343505859,
      "learning_rate": 9.972470956907375e-05,
      "loss": 2.7083208084106447,
      "memory(GiB)": 47.63,
      "step": 3900,
      "token_acc": 0.43597560975609756,
      "train_speed(iter/s)": 1.464029
    },
    {
      "epoch": 0.1673021721434386,
      "grad_norm": 2.8918769359588623,
      "learning_rate": 9.97240038942191e-05,
      "loss": 2.768160820007324,
      "memory(GiB)": 47.63,
      "step": 3905,
      "token_acc": 0.4117647058823529,
      "train_speed(iter/s)": 1.463813
    },
    {
      "epoch": 0.16751638747268754,
      "grad_norm": 3.165897846221924,
      "learning_rate": 9.972329731856745e-05,
      "loss": 2.3745092391967773,
      "memory(GiB)": 47.63,
      "step": 3910,
      "token_acc": 0.5363636363636364,
      "train_speed(iter/s)": 1.463548
    },
    {
      "epoch": 0.1677306028019365,
      "grad_norm": 5.167481899261475,
      "learning_rate": 9.972258984213164e-05,
      "loss": 2.7123252868652346,
      "memory(GiB)": 47.63,
      "step": 3915,
      "token_acc": 0.4430379746835443,
      "train_speed(iter/s)": 1.463868
    },
    {
      "epoch": 0.16794481813118547,
      "grad_norm": 7.225919723510742,
      "learning_rate": 9.972188146492446e-05,
      "loss": 2.6176177978515627,
      "memory(GiB)": 47.63,
      "step": 3920,
      "token_acc": 0.4486301369863014,
      "train_speed(iter/s)": 1.463912
    },
    {
      "epoch": 0.16815903346043443,
      "grad_norm": 4.501922607421875,
      "learning_rate": 9.972117218695877e-05,
      "loss": 2.619853973388672,
      "memory(GiB)": 47.63,
      "step": 3925,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.46408
    },
    {
      "epoch": 0.16837324878968338,
      "grad_norm": 3.6003787517547607,
      "learning_rate": 9.97204620082474e-05,
      "loss": 2.591394233703613,
      "memory(GiB)": 47.63,
      "step": 3930,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.464035
    },
    {
      "epoch": 0.16858746411893236,
      "grad_norm": 3.0413100719451904,
      "learning_rate": 9.971975092880321e-05,
      "loss": 2.5557174682617188,
      "memory(GiB)": 47.63,
      "step": 3935,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.46356
    },
    {
      "epoch": 0.1688016794481813,
      "grad_norm": 3.5041685104370117,
      "learning_rate": 9.97190389486391e-05,
      "loss": 2.6853059768676757,
      "memory(GiB)": 47.63,
      "step": 3940,
      "token_acc": 0.4619883040935672,
      "train_speed(iter/s)": 1.463606
    },
    {
      "epoch": 0.1690158947774303,
      "grad_norm": 3.9365909099578857,
      "learning_rate": 9.971832606776797e-05,
      "loss": 2.800808334350586,
      "memory(GiB)": 47.63,
      "step": 3945,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.463303
    },
    {
      "epoch": 0.16923011010667924,
      "grad_norm": 3.8099727630615234,
      "learning_rate": 9.97176122862027e-05,
      "loss": 2.8381914138793944,
      "memory(GiB)": 47.63,
      "step": 3950,
      "token_acc": 0.4297082228116711,
      "train_speed(iter/s)": 1.463091
    },
    {
      "epoch": 0.1694443254359282,
      "grad_norm": 3.864774465560913,
      "learning_rate": 9.971689760395626e-05,
      "loss": 2.587199592590332,
      "memory(GiB)": 47.63,
      "step": 3955,
      "token_acc": 0.44518272425249167,
      "train_speed(iter/s)": 1.462596
    },
    {
      "epoch": 0.16965854076517717,
      "grad_norm": 3.4290759563446045,
      "learning_rate": 9.971618202104157e-05,
      "loss": 2.573192024230957,
      "memory(GiB)": 47.63,
      "step": 3960,
      "token_acc": 0.4574780058651026,
      "train_speed(iter/s)": 1.462598
    },
    {
      "epoch": 0.16987275609442612,
      "grad_norm": 3.7943031787872314,
      "learning_rate": 9.971546553747163e-05,
      "loss": 2.7710811614990236,
      "memory(GiB)": 47.63,
      "step": 3965,
      "token_acc": 0.46387832699619774,
      "train_speed(iter/s)": 1.462764
    },
    {
      "epoch": 0.17008697142367507,
      "grad_norm": 3.480952501296997,
      "learning_rate": 9.971474815325936e-05,
      "loss": 2.6828598022460937,
      "memory(GiB)": 47.63,
      "step": 3970,
      "token_acc": 0.4290657439446367,
      "train_speed(iter/s)": 1.462971
    },
    {
      "epoch": 0.17030118675292405,
      "grad_norm": 4.311180591583252,
      "learning_rate": 9.97140298684178e-05,
      "loss": 2.555287170410156,
      "memory(GiB)": 47.63,
      "step": 3975,
      "token_acc": 0.5282258064516129,
      "train_speed(iter/s)": 1.462515
    },
    {
      "epoch": 0.170515402082173,
      "grad_norm": 4.069502830505371,
      "learning_rate": 9.971331068295998e-05,
      "loss": 2.439258575439453,
      "memory(GiB)": 47.63,
      "step": 3980,
      "token_acc": 0.49508196721311476,
      "train_speed(iter/s)": 1.462453
    },
    {
      "epoch": 0.17072961741142195,
      "grad_norm": 3.7039411067962646,
      "learning_rate": 9.971259059689887e-05,
      "loss": 3.065530014038086,
      "memory(GiB)": 47.63,
      "step": 3985,
      "token_acc": 0.4068241469816273,
      "train_speed(iter/s)": 1.462715
    },
    {
      "epoch": 0.17094383274067093,
      "grad_norm": 4.358893394470215,
      "learning_rate": 9.971186961024756e-05,
      "loss": 2.7284961700439454,
      "memory(GiB)": 47.63,
      "step": 3990,
      "token_acc": 0.4368231046931408,
      "train_speed(iter/s)": 1.462657
    },
    {
      "epoch": 0.17115804806991988,
      "grad_norm": 3.1369922161102295,
      "learning_rate": 9.97111477230191e-05,
      "loss": 2.567704200744629,
      "memory(GiB)": 47.63,
      "step": 3995,
      "token_acc": 0.45652173913043476,
      "train_speed(iter/s)": 1.463042
    },
    {
      "epoch": 0.17137226339916883,
      "grad_norm": 4.521172523498535,
      "learning_rate": 9.971042493522655e-05,
      "loss": 2.6029748916625977,
      "memory(GiB)": 47.63,
      "step": 4000,
      "token_acc": 0.466403162055336,
      "train_speed(iter/s)": 1.463397
    },
    {
      "epoch": 0.17137226339916883,
      "eval_loss": 2.2883613109588623,
      "eval_runtime": 14.1768,
      "eval_samples_per_second": 7.054,
      "eval_steps_per_second": 7.054,
      "eval_token_acc": 0.4913294797687861,
      "step": 4000
    },
    {
      "epoch": 0.1715864787284178,
      "grad_norm": 9.697166442871094,
      "learning_rate": 9.970970124688305e-05,
      "loss": 2.6056758880615236,
      "memory(GiB)": 47.63,
      "step": 4005,
      "token_acc": 0.4711155378486056,
      "train_speed(iter/s)": 1.455794
    },
    {
      "epoch": 0.17180069405766676,
      "grad_norm": 4.103949069976807,
      "learning_rate": 9.970897665800167e-05,
      "loss": 2.529444122314453,
      "memory(GiB)": 47.63,
      "step": 4010,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.456129
    },
    {
      "epoch": 0.17201490938691572,
      "grad_norm": 4.441743850708008,
      "learning_rate": 9.970825116859554e-05,
      "loss": 3.038761329650879,
      "memory(GiB)": 47.63,
      "step": 4015,
      "token_acc": 0.44375,
      "train_speed(iter/s)": 1.456316
    },
    {
      "epoch": 0.1722291247161647,
      "grad_norm": 3.7059218883514404,
      "learning_rate": 9.97075247786778e-05,
      "loss": 2.5436962127685545,
      "memory(GiB)": 47.63,
      "step": 4020,
      "token_acc": 0.5100401606425703,
      "train_speed(iter/s)": 1.456573
    },
    {
      "epoch": 0.17244334004541365,
      "grad_norm": 5.761212348937988,
      "learning_rate": 9.970679748826164e-05,
      "loss": 3.0638465881347656,
      "memory(GiB)": 47.63,
      "step": 4025,
      "token_acc": 0.4084084084084084,
      "train_speed(iter/s)": 1.45652
    },
    {
      "epoch": 0.17265755537466262,
      "grad_norm": 4.596331596374512,
      "learning_rate": 9.97060692973602e-05,
      "loss": 2.9444766998291017,
      "memory(GiB)": 47.63,
      "step": 4030,
      "token_acc": 0.42214532871972316,
      "train_speed(iter/s)": 1.456569
    },
    {
      "epoch": 0.17287177070391158,
      "grad_norm": 5.4121856689453125,
      "learning_rate": 9.97053402059867e-05,
      "loss": 2.8367212295532225,
      "memory(GiB)": 47.63,
      "step": 4035,
      "token_acc": 0.4276729559748428,
      "train_speed(iter/s)": 1.456212
    },
    {
      "epoch": 0.17308598603316053,
      "grad_norm": 5.506450653076172,
      "learning_rate": 9.970461021415433e-05,
      "loss": 2.3924999237060547,
      "memory(GiB)": 47.63,
      "step": 4040,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.455904
    },
    {
      "epoch": 0.1733002013624095,
      "grad_norm": 4.011473655700684,
      "learning_rate": 9.97038793218763e-05,
      "loss": 2.673995590209961,
      "memory(GiB)": 47.63,
      "step": 4045,
      "token_acc": 0.45255474452554745,
      "train_speed(iter/s)": 1.456114
    },
    {
      "epoch": 0.17351441669165846,
      "grad_norm": 4.279806613922119,
      "learning_rate": 9.97031475291659e-05,
      "loss": 2.7830909729003905,
      "memory(GiB)": 47.63,
      "step": 4050,
      "token_acc": 0.462406015037594,
      "train_speed(iter/s)": 1.456281
    },
    {
      "epoch": 0.1737286320209074,
      "grad_norm": 5.206503391265869,
      "learning_rate": 9.970241483603635e-05,
      "loss": 2.5369184494018553,
      "memory(GiB)": 47.63,
      "step": 4055,
      "token_acc": 0.45849802371541504,
      "train_speed(iter/s)": 1.456209
    },
    {
      "epoch": 0.1739428473501564,
      "grad_norm": 3.9950671195983887,
      "learning_rate": 9.970168124250093e-05,
      "loss": 2.4039979934692384,
      "memory(GiB)": 47.63,
      "step": 4060,
      "token_acc": 0.49609375,
      "train_speed(iter/s)": 1.456348
    },
    {
      "epoch": 0.17415706267940534,
      "grad_norm": 6.458606243133545,
      "learning_rate": 9.970094674857291e-05,
      "loss": 2.686125373840332,
      "memory(GiB)": 47.63,
      "step": 4065,
      "token_acc": 0.43609022556390975,
      "train_speed(iter/s)": 1.456607
    },
    {
      "epoch": 0.1743712780086543,
      "grad_norm": 7.220560550689697,
      "learning_rate": 9.970021135426564e-05,
      "loss": 3.1351200103759767,
      "memory(GiB)": 47.63,
      "step": 4070,
      "token_acc": 0.43508771929824563,
      "train_speed(iter/s)": 1.456987
    },
    {
      "epoch": 0.17458549333790327,
      "grad_norm": 3.4459733963012695,
      "learning_rate": 9.969947505959241e-05,
      "loss": 2.503296661376953,
      "memory(GiB)": 47.63,
      "step": 4075,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.456934
    },
    {
      "epoch": 0.17479970866715222,
      "grad_norm": 9.989091873168945,
      "learning_rate": 9.969873786456656e-05,
      "loss": 2.734407424926758,
      "memory(GiB)": 47.63,
      "step": 4080,
      "token_acc": 0.4431137724550898,
      "train_speed(iter/s)": 1.456992
    },
    {
      "epoch": 0.17501392399640117,
      "grad_norm": 3.8455822467803955,
      "learning_rate": 9.969799976920147e-05,
      "loss": 2.4487541198730467,
      "memory(GiB)": 47.63,
      "step": 4085,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.456482
    },
    {
      "epoch": 0.17522813932565015,
      "grad_norm": 4.026459693908691,
      "learning_rate": 9.969726077351047e-05,
      "loss": 2.525579833984375,
      "memory(GiB)": 47.63,
      "step": 4090,
      "token_acc": 0.4620253164556962,
      "train_speed(iter/s)": 1.456669
    },
    {
      "epoch": 0.1754423546548991,
      "grad_norm": 3.5084986686706543,
      "learning_rate": 9.969652087750698e-05,
      "loss": 2.494241142272949,
      "memory(GiB)": 47.63,
      "step": 4095,
      "token_acc": 0.4641638225255973,
      "train_speed(iter/s)": 1.456741
    },
    {
      "epoch": 0.17565656998414805,
      "grad_norm": 4.460583686828613,
      "learning_rate": 9.969578008120441e-05,
      "loss": 2.6435054779052733,
      "memory(GiB)": 47.63,
      "step": 4100,
      "token_acc": 0.4395973154362416,
      "train_speed(iter/s)": 1.456208
    },
    {
      "epoch": 0.17587078531339703,
      "grad_norm": 2.9879324436187744,
      "learning_rate": 9.969503838461615e-05,
      "loss": 2.456716537475586,
      "memory(GiB)": 47.63,
      "step": 4105,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.456558
    },
    {
      "epoch": 0.17608500064264598,
      "grad_norm": 2.53955078125,
      "learning_rate": 9.969429578775567e-05,
      "loss": 2.7428861618041993,
      "memory(GiB)": 47.63,
      "step": 4110,
      "token_acc": 0.42450142450142453,
      "train_speed(iter/s)": 1.457029
    },
    {
      "epoch": 0.17629921597189496,
      "grad_norm": 8.147093772888184,
      "learning_rate": 9.969355229063638e-05,
      "loss": 2.627157211303711,
      "memory(GiB)": 47.63,
      "step": 4115,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.456923
    },
    {
      "epoch": 0.1765134313011439,
      "grad_norm": 4.598494529724121,
      "learning_rate": 9.969280789327179e-05,
      "loss": 2.4965232849121093,
      "memory(GiB)": 47.63,
      "step": 4120,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.457068
    },
    {
      "epoch": 0.17672764663039287,
      "grad_norm": 4.967527389526367,
      "learning_rate": 9.969206259567537e-05,
      "loss": 2.524370574951172,
      "memory(GiB)": 47.63,
      "step": 4125,
      "token_acc": 0.46905537459283386,
      "train_speed(iter/s)": 1.457255
    },
    {
      "epoch": 0.17694186195964184,
      "grad_norm": 4.3180928230285645,
      "learning_rate": 9.969131639786061e-05,
      "loss": 2.4822004318237303,
      "memory(GiB)": 47.63,
      "step": 4130,
      "token_acc": 0.4910394265232975,
      "train_speed(iter/s)": 1.456867
    },
    {
      "epoch": 0.1771560772888908,
      "grad_norm": 3.8205959796905518,
      "learning_rate": 9.969056929984105e-05,
      "loss": 2.5836681365966796,
      "memory(GiB)": 47.63,
      "step": 4135,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.456748
    },
    {
      "epoch": 0.17737029261813975,
      "grad_norm": 3.3553004264831543,
      "learning_rate": 9.968982130163021e-05,
      "loss": 2.7959674835205077,
      "memory(GiB)": 47.63,
      "step": 4140,
      "token_acc": 0.4578313253012048,
      "train_speed(iter/s)": 1.456909
    },
    {
      "epoch": 0.17758450794738873,
      "grad_norm": 3.382815361022949,
      "learning_rate": 9.968907240324165e-05,
      "loss": 2.6812389373779295,
      "memory(GiB)": 47.63,
      "step": 4145,
      "token_acc": 0.4835820895522388,
      "train_speed(iter/s)": 1.457057
    },
    {
      "epoch": 0.17779872327663768,
      "grad_norm": 3.5095791816711426,
      "learning_rate": 9.968832260468892e-05,
      "loss": 2.73370304107666,
      "memory(GiB)": 47.63,
      "step": 4150,
      "token_acc": 0.4253246753246753,
      "train_speed(iter/s)": 1.457179
    },
    {
      "epoch": 0.17801293860588663,
      "grad_norm": 6.752483367919922,
      "learning_rate": 9.96875719059856e-05,
      "loss": 2.489898109436035,
      "memory(GiB)": 47.63,
      "step": 4155,
      "token_acc": 0.47580645161290325,
      "train_speed(iter/s)": 1.457324
    },
    {
      "epoch": 0.1782271539351356,
      "grad_norm": 3.5551066398620605,
      "learning_rate": 9.968682030714534e-05,
      "loss": 2.5549184799194338,
      "memory(GiB)": 47.63,
      "step": 4160,
      "token_acc": 0.46200607902735563,
      "train_speed(iter/s)": 1.457832
    },
    {
      "epoch": 0.17844136926438456,
      "grad_norm": 3.5505824089050293,
      "learning_rate": 9.96860678081817e-05,
      "loss": 2.686321258544922,
      "memory(GiB)": 47.63,
      "step": 4165,
      "token_acc": 0.4382716049382716,
      "train_speed(iter/s)": 1.458368
    },
    {
      "epoch": 0.1786555845936335,
      "grad_norm": 4.782723426818848,
      "learning_rate": 9.968531440910835e-05,
      "loss": 2.4822921752929688,
      "memory(GiB)": 47.63,
      "step": 4170,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.45876
    },
    {
      "epoch": 0.1788697999228825,
      "grad_norm": 3.174055337905884,
      "learning_rate": 9.96845601099389e-05,
      "loss": 2.507801818847656,
      "memory(GiB)": 47.63,
      "step": 4175,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.458457
    },
    {
      "epoch": 0.17908401525213144,
      "grad_norm": 4.528962135314941,
      "learning_rate": 9.968380491068705e-05,
      "loss": 2.70302734375,
      "memory(GiB)": 47.63,
      "step": 4180,
      "token_acc": 0.426056338028169,
      "train_speed(iter/s)": 1.458921
    },
    {
      "epoch": 0.1792982305813804,
      "grad_norm": 4.106651306152344,
      "learning_rate": 9.968304881136645e-05,
      "loss": 2.530369758605957,
      "memory(GiB)": 47.63,
      "step": 4185,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.459121
    },
    {
      "epoch": 0.17951244591062937,
      "grad_norm": 4.651950836181641,
      "learning_rate": 9.968229181199083e-05,
      "loss": 2.392143440246582,
      "memory(GiB)": 47.63,
      "step": 4190,
      "token_acc": 0.4497991967871486,
      "train_speed(iter/s)": 1.459139
    },
    {
      "epoch": 0.17972666123987832,
      "grad_norm": 2.965430498123169,
      "learning_rate": 9.968153391257388e-05,
      "loss": 2.497489166259766,
      "memory(GiB)": 47.63,
      "step": 4195,
      "token_acc": 0.4421052631578947,
      "train_speed(iter/s)": 1.459019
    },
    {
      "epoch": 0.1799408765691273,
      "grad_norm": 3.2989437580108643,
      "learning_rate": 9.968077511312934e-05,
      "loss": 2.5011953353881835,
      "memory(GiB)": 47.63,
      "step": 4200,
      "token_acc": 0.4797507788161994,
      "train_speed(iter/s)": 1.459391
    },
    {
      "epoch": 0.18015509189837625,
      "grad_norm": 3.1855950355529785,
      "learning_rate": 9.968001541367096e-05,
      "loss": 2.597939682006836,
      "memory(GiB)": 47.63,
      "step": 4205,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.45943
    },
    {
      "epoch": 0.1803693072276252,
      "grad_norm": 6.3596086502075195,
      "learning_rate": 9.967925481421249e-05,
      "loss": 2.8404872894287108,
      "memory(GiB)": 47.63,
      "step": 4210,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.459484
    },
    {
      "epoch": 0.18058352255687418,
      "grad_norm": 3.202695608139038,
      "learning_rate": 9.967849331476771e-05,
      "loss": 2.640762519836426,
      "memory(GiB)": 47.63,
      "step": 4215,
      "token_acc": 0.4360655737704918,
      "train_speed(iter/s)": 1.459632
    },
    {
      "epoch": 0.18079773788612313,
      "grad_norm": 3.3546950817108154,
      "learning_rate": 9.967773091535045e-05,
      "loss": 2.3101552963256835,
      "memory(GiB)": 47.63,
      "step": 4220,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.459579
    },
    {
      "epoch": 0.18101195321537208,
      "grad_norm": 3.6231908798217773,
      "learning_rate": 9.967696761597446e-05,
      "loss": 2.50335578918457,
      "memory(GiB)": 47.63,
      "step": 4225,
      "token_acc": 0.46360153256704983,
      "train_speed(iter/s)": 1.460172
    },
    {
      "epoch": 0.18122616854462106,
      "grad_norm": 4.873368740081787,
      "learning_rate": 9.967620341665361e-05,
      "loss": 2.5752490997314452,
      "memory(GiB)": 47.63,
      "step": 4230,
      "token_acc": 0.45317220543806647,
      "train_speed(iter/s)": 1.460317
    },
    {
      "epoch": 0.18144038387387001,
      "grad_norm": 4.936944484710693,
      "learning_rate": 9.967543831740174e-05,
      "loss": 2.3862064361572264,
      "memory(GiB)": 47.63,
      "step": 4235,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.460182
    },
    {
      "epoch": 0.18165459920311897,
      "grad_norm": 3.8961172103881836,
      "learning_rate": 9.967467231823271e-05,
      "loss": 2.3928709030151367,
      "memory(GiB)": 47.63,
      "step": 4240,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.460028
    },
    {
      "epoch": 0.18186881453236795,
      "grad_norm": 3.8946142196655273,
      "learning_rate": 9.967390541916037e-05,
      "loss": 2.912725639343262,
      "memory(GiB)": 47.63,
      "step": 4245,
      "token_acc": 0.4098939929328622,
      "train_speed(iter/s)": 1.46028
    },
    {
      "epoch": 0.1820830298616169,
      "grad_norm": 3.5400397777557373,
      "learning_rate": 9.967313762019866e-05,
      "loss": 2.9137847900390623,
      "memory(GiB)": 47.63,
      "step": 4250,
      "token_acc": 0.4354243542435424,
      "train_speed(iter/s)": 1.460411
    },
    {
      "epoch": 0.18229724519086585,
      "grad_norm": 2.751776695251465,
      "learning_rate": 9.967236892136145e-05,
      "loss": 2.5477777481079102,
      "memory(GiB)": 47.63,
      "step": 4255,
      "token_acc": 0.4567901234567901,
      "train_speed(iter/s)": 1.460698
    },
    {
      "epoch": 0.18251146052011483,
      "grad_norm": 3.9068148136138916,
      "learning_rate": 9.967159932266271e-05,
      "loss": 2.6640533447265624,
      "memory(GiB)": 47.63,
      "step": 4260,
      "token_acc": 0.4575645756457565,
      "train_speed(iter/s)": 1.460887
    },
    {
      "epoch": 0.18272567584936378,
      "grad_norm": 4.075963020324707,
      "learning_rate": 9.967082882411631e-05,
      "loss": 2.9097997665405275,
      "memory(GiB)": 47.63,
      "step": 4265,
      "token_acc": 0.40397350993377484,
      "train_speed(iter/s)": 1.46056
    },
    {
      "epoch": 0.18293989117861273,
      "grad_norm": 3.5479114055633545,
      "learning_rate": 9.967005742573627e-05,
      "loss": 2.8735851287841796,
      "memory(GiB)": 47.63,
      "step": 4270,
      "token_acc": 0.4472049689440994,
      "train_speed(iter/s)": 1.460882
    },
    {
      "epoch": 0.1831541065078617,
      "grad_norm": 3.406883478164673,
      "learning_rate": 9.966928512753656e-05,
      "loss": 2.4399322509765624,
      "memory(GiB)": 47.63,
      "step": 4275,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.461111
    },
    {
      "epoch": 0.18336832183711066,
      "grad_norm": 5.102316379547119,
      "learning_rate": 9.966851192953114e-05,
      "loss": 2.696395683288574,
      "memory(GiB)": 47.63,
      "step": 4280,
      "token_acc": 0.4774774774774775,
      "train_speed(iter/s)": 1.460907
    },
    {
      "epoch": 0.18358253716635964,
      "grad_norm": 4.178129196166992,
      "learning_rate": 9.966773783173403e-05,
      "loss": 2.9731023788452147,
      "memory(GiB)": 47.63,
      "step": 4285,
      "token_acc": 0.44609665427509293,
      "train_speed(iter/s)": 1.460806
    },
    {
      "epoch": 0.1837967524956086,
      "grad_norm": 4.420988082885742,
      "learning_rate": 9.966696283415926e-05,
      "loss": 2.349736785888672,
      "memory(GiB)": 47.63,
      "step": 4290,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.460971
    },
    {
      "epoch": 0.18401096782485754,
      "grad_norm": 3.1192681789398193,
      "learning_rate": 9.966618693682089e-05,
      "loss": 2.9653976440429686,
      "memory(GiB)": 47.63,
      "step": 4295,
      "token_acc": 0.4080267558528428,
      "train_speed(iter/s)": 1.461057
    },
    {
      "epoch": 0.18422518315410652,
      "grad_norm": 4.423130035400391,
      "learning_rate": 9.966541013973291e-05,
      "loss": 2.4411151885986326,
      "memory(GiB)": 47.63,
      "step": 4300,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.461167
    },
    {
      "epoch": 0.18443939848335547,
      "grad_norm": 3.7723641395568848,
      "learning_rate": 9.966463244290947e-05,
      "loss": 2.408475875854492,
      "memory(GiB)": 47.63,
      "step": 4305,
      "token_acc": 0.4875,
      "train_speed(iter/s)": 1.461318
    },
    {
      "epoch": 0.18465361381260442,
      "grad_norm": 4.908487796783447,
      "learning_rate": 9.966385384636461e-05,
      "loss": 2.6260196685791017,
      "memory(GiB)": 47.63,
      "step": 4310,
      "token_acc": 0.41603053435114506,
      "train_speed(iter/s)": 1.461368
    },
    {
      "epoch": 0.1848678291418534,
      "grad_norm": 4.305746555328369,
      "learning_rate": 9.966307435011245e-05,
      "loss": 2.60677547454834,
      "memory(GiB)": 47.63,
      "step": 4315,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.460971
    },
    {
      "epoch": 0.18508204447110235,
      "grad_norm": 4.457401752471924,
      "learning_rate": 9.96622939541671e-05,
      "loss": 2.923782730102539,
      "memory(GiB)": 47.63,
      "step": 4320,
      "token_acc": 0.45723684210526316,
      "train_speed(iter/s)": 1.461007
    },
    {
      "epoch": 0.1852962598003513,
      "grad_norm": 4.0793914794921875,
      "learning_rate": 9.966151265854273e-05,
      "loss": 2.727579116821289,
      "memory(GiB)": 47.63,
      "step": 4325,
      "token_acc": 0.4541832669322709,
      "train_speed(iter/s)": 1.461072
    },
    {
      "epoch": 0.18551047512960028,
      "grad_norm": 5.2035017013549805,
      "learning_rate": 9.966073046325346e-05,
      "loss": 2.4274356842041014,
      "memory(GiB)": 47.63,
      "step": 4330,
      "token_acc": 0.49361702127659574,
      "train_speed(iter/s)": 1.460589
    },
    {
      "epoch": 0.18572469045884923,
      "grad_norm": 4.7049880027771,
      "learning_rate": 9.965994736831348e-05,
      "loss": 2.6569192886352537,
      "memory(GiB)": 47.63,
      "step": 4335,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.461008
    },
    {
      "epoch": 0.18593890578809819,
      "grad_norm": 3.325430154800415,
      "learning_rate": 9.965916337373696e-05,
      "loss": 2.6684452056884767,
      "memory(GiB)": 47.63,
      "step": 4340,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.461264
    },
    {
      "epoch": 0.18615312111734716,
      "grad_norm": 2.775076389312744,
      "learning_rate": 9.965837847953812e-05,
      "loss": 2.372978401184082,
      "memory(GiB)": 47.63,
      "step": 4345,
      "token_acc": 0.48881789137380194,
      "train_speed(iter/s)": 1.461303
    },
    {
      "epoch": 0.18636733644659612,
      "grad_norm": 5.524483680725098,
      "learning_rate": 9.965759268573117e-05,
      "loss": 2.436813545227051,
      "memory(GiB)": 47.63,
      "step": 4350,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.461202
    },
    {
      "epoch": 0.18658155177584507,
      "grad_norm": 3.795135974884033,
      "learning_rate": 9.965680599233034e-05,
      "loss": 2.4541547775268553,
      "memory(GiB)": 47.63,
      "step": 4355,
      "token_acc": 0.5038759689922481,
      "train_speed(iter/s)": 1.461225
    },
    {
      "epoch": 0.18679576710509405,
      "grad_norm": 2.6173551082611084,
      "learning_rate": 9.965601839934988e-05,
      "loss": 2.671848487854004,
      "memory(GiB)": 47.63,
      "step": 4360,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.461331
    },
    {
      "epoch": 0.187009982434343,
      "grad_norm": 4.096261024475098,
      "learning_rate": 9.965522990680406e-05,
      "loss": 2.9269435882568358,
      "memory(GiB)": 47.63,
      "step": 4365,
      "token_acc": 0.4387755102040816,
      "train_speed(iter/s)": 1.461532
    },
    {
      "epoch": 0.18722419776359198,
      "grad_norm": 4.488049507141113,
      "learning_rate": 9.965444051470721e-05,
      "loss": 3.037659454345703,
      "memory(GiB)": 47.63,
      "step": 4370,
      "token_acc": 0.3881578947368421,
      "train_speed(iter/s)": 1.462003
    },
    {
      "epoch": 0.18743841309284093,
      "grad_norm": 2.9177701473236084,
      "learning_rate": 9.965365022307356e-05,
      "loss": 2.3297451019287108,
      "memory(GiB)": 47.63,
      "step": 4375,
      "token_acc": 0.45936395759717313,
      "train_speed(iter/s)": 1.46206
    },
    {
      "epoch": 0.18765262842208988,
      "grad_norm": 3.5849967002868652,
      "learning_rate": 9.965285903191746e-05,
      "loss": 2.6896245956420897,
      "memory(GiB)": 47.63,
      "step": 4380,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.461985
    },
    {
      "epoch": 0.18786684375133886,
      "grad_norm": 4.3653154373168945,
      "learning_rate": 9.965206694125324e-05,
      "loss": 2.6202011108398438,
      "memory(GiB)": 47.63,
      "step": 4385,
      "token_acc": 0.465625,
      "train_speed(iter/s)": 1.461821
    },
    {
      "epoch": 0.1880810590805878,
      "grad_norm": 3.7579469680786133,
      "learning_rate": 9.965127395109525e-05,
      "loss": 2.800714874267578,
      "memory(GiB)": 47.63,
      "step": 4390,
      "token_acc": 0.43853820598006643,
      "train_speed(iter/s)": 1.461655
    },
    {
      "epoch": 0.18829527440983676,
      "grad_norm": 4.8983259201049805,
      "learning_rate": 9.965048006145785e-05,
      "loss": 2.5570486068725584,
      "memory(GiB)": 47.63,
      "step": 4395,
      "token_acc": 0.4621212121212121,
      "train_speed(iter/s)": 1.462053
    },
    {
      "epoch": 0.18850948973908574,
      "grad_norm": 5.242818355560303,
      "learning_rate": 9.964968527235544e-05,
      "loss": 2.620072364807129,
      "memory(GiB)": 47.63,
      "step": 4400,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.462177
    },
    {
      "epoch": 0.1887237050683347,
      "grad_norm": 3.5841519832611084,
      "learning_rate": 9.96488895838024e-05,
      "loss": 2.581031608581543,
      "memory(GiB)": 47.63,
      "step": 4405,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.461995
    },
    {
      "epoch": 0.18893792039758364,
      "grad_norm": 3.4121358394622803,
      "learning_rate": 9.964809299581315e-05,
      "loss": 2.467431831359863,
      "memory(GiB)": 47.63,
      "step": 4410,
      "token_acc": 0.46494464944649444,
      "train_speed(iter/s)": 1.462257
    },
    {
      "epoch": 0.18915213572683262,
      "grad_norm": 5.127429008483887,
      "learning_rate": 9.964729550840212e-05,
      "loss": 2.5965818405151366,
      "memory(GiB)": 47.63,
      "step": 4415,
      "token_acc": 0.43302180685358255,
      "train_speed(iter/s)": 1.462006
    },
    {
      "epoch": 0.18936635105608157,
      "grad_norm": 3.961388111114502,
      "learning_rate": 9.964649712158377e-05,
      "loss": 2.677463150024414,
      "memory(GiB)": 47.63,
      "step": 4420,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.462467
    },
    {
      "epoch": 0.18958056638533052,
      "grad_norm": 3.791003942489624,
      "learning_rate": 9.964569783537255e-05,
      "loss": 2.9339937210083007,
      "memory(GiB)": 47.63,
      "step": 4425,
      "token_acc": 0.4678111587982833,
      "train_speed(iter/s)": 1.462977
    },
    {
      "epoch": 0.1897947817145795,
      "grad_norm": 3.5306408405303955,
      "learning_rate": 9.964489764978293e-05,
      "loss": 2.5060840606689454,
      "memory(GiB)": 47.63,
      "step": 4430,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.463159
    },
    {
      "epoch": 0.19000899704382845,
      "grad_norm": 3.824587821960449,
      "learning_rate": 9.964409656482943e-05,
      "loss": 2.289750862121582,
      "memory(GiB)": 47.63,
      "step": 4435,
      "token_acc": 0.49624060150375937,
      "train_speed(iter/s)": 1.463251
    },
    {
      "epoch": 0.1902232123730774,
      "grad_norm": 3.808020830154419,
      "learning_rate": 9.964329458052655e-05,
      "loss": 2.6161310195922853,
      "memory(GiB)": 47.63,
      "step": 4440,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.46351
    },
    {
      "epoch": 0.19043742770232638,
      "grad_norm": 3.0389132499694824,
      "learning_rate": 9.964249169688882e-05,
      "loss": 2.7402992248535156,
      "memory(GiB)": 47.63,
      "step": 4445,
      "token_acc": 0.43934426229508194,
      "train_speed(iter/s)": 1.463621
    },
    {
      "epoch": 0.19065164303157534,
      "grad_norm": 3.4146127700805664,
      "learning_rate": 9.964168791393077e-05,
      "loss": 2.8018726348876952,
      "memory(GiB)": 47.63,
      "step": 4450,
      "token_acc": 0.43272727272727274,
      "train_speed(iter/s)": 1.464128
    },
    {
      "epoch": 0.19086585836082431,
      "grad_norm": 3.163907527923584,
      "learning_rate": 9.964088323166698e-05,
      "loss": 3.381665802001953,
      "memory(GiB)": 47.63,
      "step": 4455,
      "token_acc": 0.415625,
      "train_speed(iter/s)": 1.463784
    },
    {
      "epoch": 0.19108007369007327,
      "grad_norm": 3.469853162765503,
      "learning_rate": 9.964007765011204e-05,
      "loss": 2.964613914489746,
      "memory(GiB)": 47.63,
      "step": 4460,
      "token_acc": 0.3939393939393939,
      "train_speed(iter/s)": 1.464038
    },
    {
      "epoch": 0.19129428901932222,
      "grad_norm": 4.974525451660156,
      "learning_rate": 9.963927116928051e-05,
      "loss": 2.4353404998779298,
      "memory(GiB)": 47.63,
      "step": 4465,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.464027
    },
    {
      "epoch": 0.1915085043485712,
      "grad_norm": 4.314286231994629,
      "learning_rate": 9.963846378918702e-05,
      "loss": 2.8339258193969727,
      "memory(GiB)": 47.63,
      "step": 4470,
      "token_acc": 0.4349315068493151,
      "train_speed(iter/s)": 1.463985
    },
    {
      "epoch": 0.19172271967782015,
      "grad_norm": 4.252258777618408,
      "learning_rate": 9.96376555098462e-05,
      "loss": 2.8118757247924804,
      "memory(GiB)": 47.63,
      "step": 4475,
      "token_acc": 0.4327956989247312,
      "train_speed(iter/s)": 1.464139
    },
    {
      "epoch": 0.1919369350070691,
      "grad_norm": 4.8369011878967285,
      "learning_rate": 9.963684633127269e-05,
      "loss": 2.559932327270508,
      "memory(GiB)": 47.63,
      "step": 4480,
      "token_acc": 0.4536741214057508,
      "train_speed(iter/s)": 1.46394
    },
    {
      "epoch": 0.19215115033631808,
      "grad_norm": 3.3614726066589355,
      "learning_rate": 9.963603625348114e-05,
      "loss": 2.5462081909179686,
      "memory(GiB)": 47.63,
      "step": 4485,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.463864
    },
    {
      "epoch": 0.19236536566556703,
      "grad_norm": 4.094184398651123,
      "learning_rate": 9.963522527648623e-05,
      "loss": 2.351723289489746,
      "memory(GiB)": 47.63,
      "step": 4490,
      "token_acc": 0.4961832061068702,
      "train_speed(iter/s)": 1.463849
    },
    {
      "epoch": 0.19257958099481598,
      "grad_norm": 3.6614444255828857,
      "learning_rate": 9.963441340030267e-05,
      "loss": 2.748393249511719,
      "memory(GiB)": 47.63,
      "step": 4495,
      "token_acc": 0.43213296398891965,
      "train_speed(iter/s)": 1.463856
    },
    {
      "epoch": 0.19279379632406496,
      "grad_norm": 3.257753372192383,
      "learning_rate": 9.963360062494512e-05,
      "loss": 2.688405990600586,
      "memory(GiB)": 47.63,
      "step": 4500,
      "token_acc": 0.4554794520547945,
      "train_speed(iter/s)": 1.464127
    },
    {
      "epoch": 0.19279379632406496,
      "eval_loss": 2.1965346336364746,
      "eval_runtime": 14.2949,
      "eval_samples_per_second": 6.995,
      "eval_steps_per_second": 6.995,
      "eval_token_acc": 0.4822888283378747,
      "step": 4500
    },
    {
      "epoch": 0.1930080116533139,
      "grad_norm": 3.3588240146636963,
      "learning_rate": 9.963278695042835e-05,
      "loss": 2.779301071166992,
      "memory(GiB)": 47.63,
      "step": 4505,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.457289
    },
    {
      "epoch": 0.19322222698256286,
      "grad_norm": 4.67567777633667,
      "learning_rate": 9.963197237676709e-05,
      "loss": 2.8437389373779296,
      "memory(GiB)": 47.63,
      "step": 4510,
      "token_acc": 0.4560669456066946,
      "train_speed(iter/s)": 1.457184
    },
    {
      "epoch": 0.19343644231181184,
      "grad_norm": 3.164679527282715,
      "learning_rate": 9.963115690397608e-05,
      "loss": 2.7683971405029295,
      "memory(GiB)": 47.63,
      "step": 4515,
      "token_acc": 0.4384858044164038,
      "train_speed(iter/s)": 1.457478
    },
    {
      "epoch": 0.1936506576410608,
      "grad_norm": 5.027650833129883,
      "learning_rate": 9.96303405320701e-05,
      "loss": 2.2967552185058593,
      "memory(GiB)": 47.63,
      "step": 4520,
      "token_acc": 0.525691699604743,
      "train_speed(iter/s)": 1.457404
    },
    {
      "epoch": 0.19386487297030974,
      "grad_norm": 5.658149242401123,
      "learning_rate": 9.962952326106396e-05,
      "loss": 2.7964462280273437,
      "memory(GiB)": 47.63,
      "step": 4525,
      "token_acc": 0.4463087248322148,
      "train_speed(iter/s)": 1.457647
    },
    {
      "epoch": 0.19407908829955872,
      "grad_norm": 3.484639883041382,
      "learning_rate": 9.962870509097245e-05,
      "loss": 2.441773223876953,
      "memory(GiB)": 47.63,
      "step": 4530,
      "token_acc": 0.4708333333333333,
      "train_speed(iter/s)": 1.457409
    },
    {
      "epoch": 0.19429330362880767,
      "grad_norm": 3.731177568435669,
      "learning_rate": 9.962788602181039e-05,
      "loss": 2.435281181335449,
      "memory(GiB)": 47.63,
      "step": 4535,
      "token_acc": 0.43197278911564624,
      "train_speed(iter/s)": 1.457833
    },
    {
      "epoch": 0.19450751895805665,
      "grad_norm": 2.832918167114258,
      "learning_rate": 9.962706605359262e-05,
      "loss": 2.6169305801391602,
      "memory(GiB)": 47.63,
      "step": 4540,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.457925
    },
    {
      "epoch": 0.1947217342873056,
      "grad_norm": 3.472694158554077,
      "learning_rate": 9.962624518633399e-05,
      "loss": 2.596920967102051,
      "memory(GiB)": 47.63,
      "step": 4545,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.45815
    },
    {
      "epoch": 0.19493594961655455,
      "grad_norm": 3.1194911003112793,
      "learning_rate": 9.962542342004937e-05,
      "loss": 2.5030025482177733,
      "memory(GiB)": 47.63,
      "step": 4550,
      "token_acc": 0.5093167701863354,
      "train_speed(iter/s)": 1.458317
    },
    {
      "epoch": 0.19515016494580353,
      "grad_norm": 3.147874116897583,
      "learning_rate": 9.962460075475366e-05,
      "loss": 2.5221424102783203,
      "memory(GiB)": 47.63,
      "step": 4555,
      "token_acc": 0.43410852713178294,
      "train_speed(iter/s)": 1.458664
    },
    {
      "epoch": 0.19536438027505248,
      "grad_norm": 3.951706886291504,
      "learning_rate": 9.962377719046176e-05,
      "loss": 2.5376556396484373,
      "memory(GiB)": 47.63,
      "step": 4560,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.458976
    },
    {
      "epoch": 0.19557859560430144,
      "grad_norm": 4.471684455871582,
      "learning_rate": 9.962295272718858e-05,
      "loss": 2.3781890869140625,
      "memory(GiB)": 47.63,
      "step": 4565,
      "token_acc": 0.5638766519823789,
      "train_speed(iter/s)": 1.458898
    },
    {
      "epoch": 0.19579281093355042,
      "grad_norm": 3.68644118309021,
      "learning_rate": 9.962212736494905e-05,
      "loss": 2.5323280334472655,
      "memory(GiB)": 47.63,
      "step": 4570,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.458497
    },
    {
      "epoch": 0.19600702626279937,
      "grad_norm": 4.825692176818848,
      "learning_rate": 9.962130110375815e-05,
      "loss": 2.6095434188842774,
      "memory(GiB)": 47.63,
      "step": 4575,
      "token_acc": 0.4528985507246377,
      "train_speed(iter/s)": 1.458933
    },
    {
      "epoch": 0.19622124159204832,
      "grad_norm": 3.770578384399414,
      "learning_rate": 9.962047394363083e-05,
      "loss": 2.5597244262695313,
      "memory(GiB)": 47.63,
      "step": 4580,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.458866
    },
    {
      "epoch": 0.1964354569212973,
      "grad_norm": 3.474961996078491,
      "learning_rate": 9.961964588458207e-05,
      "loss": 2.66544189453125,
      "memory(GiB)": 47.63,
      "step": 4585,
      "token_acc": 0.516,
      "train_speed(iter/s)": 1.459118
    },
    {
      "epoch": 0.19664967225054625,
      "grad_norm": 3.6821129322052,
      "learning_rate": 9.961881692662688e-05,
      "loss": 2.726749610900879,
      "memory(GiB)": 47.63,
      "step": 4590,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.459355
    },
    {
      "epoch": 0.1968638875797952,
      "grad_norm": 5.148923397064209,
      "learning_rate": 9.961798706978027e-05,
      "loss": 2.8032838821411135,
      "memory(GiB)": 47.63,
      "step": 4595,
      "token_acc": 0.4405144694533762,
      "train_speed(iter/s)": 1.459654
    },
    {
      "epoch": 0.19707810290904418,
      "grad_norm": 4.147294044494629,
      "learning_rate": 9.96171563140573e-05,
      "loss": 2.798273277282715,
      "memory(GiB)": 47.63,
      "step": 4600,
      "token_acc": 0.4262295081967213,
      "train_speed(iter/s)": 1.459743
    },
    {
      "epoch": 0.19729231823829313,
      "grad_norm": 4.429314613342285,
      "learning_rate": 9.961632465947297e-05,
      "loss": 2.8597497940063477,
      "memory(GiB)": 47.63,
      "step": 4605,
      "token_acc": 0.4368932038834951,
      "train_speed(iter/s)": 1.459741
    },
    {
      "epoch": 0.19750653356754208,
      "grad_norm": 3.4142091274261475,
      "learning_rate": 9.96154921060424e-05,
      "loss": 2.630633544921875,
      "memory(GiB)": 47.63,
      "step": 4610,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.459614
    },
    {
      "epoch": 0.19772074889679106,
      "grad_norm": 6.562750339508057,
      "learning_rate": 9.961465865378063e-05,
      "loss": 2.3304882049560547,
      "memory(GiB)": 47.63,
      "step": 4615,
      "token_acc": 0.49034749034749037,
      "train_speed(iter/s)": 1.459435
    },
    {
      "epoch": 0.19793496422604,
      "grad_norm": 5.5141282081604,
      "learning_rate": 9.961382430270278e-05,
      "loss": 2.7084377288818358,
      "memory(GiB)": 47.63,
      "step": 4620,
      "token_acc": 0.4161676646706587,
      "train_speed(iter/s)": 1.459577
    },
    {
      "epoch": 0.198149179555289,
      "grad_norm": 4.459146022796631,
      "learning_rate": 9.961298905282397e-05,
      "loss": 2.5132150650024414,
      "memory(GiB)": 47.63,
      "step": 4625,
      "token_acc": 0.4175084175084175,
      "train_speed(iter/s)": 1.459566
    },
    {
      "epoch": 0.19836339488453794,
      "grad_norm": 4.930593967437744,
      "learning_rate": 9.961215290415931e-05,
      "loss": 2.896724891662598,
      "memory(GiB)": 47.63,
      "step": 4630,
      "token_acc": 0.42585551330798477,
      "train_speed(iter/s)": 1.459714
    },
    {
      "epoch": 0.1985776102137869,
      "grad_norm": 5.480079650878906,
      "learning_rate": 9.961131585672396e-05,
      "loss": 2.681516647338867,
      "memory(GiB)": 47.63,
      "step": 4635,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.460097
    },
    {
      "epoch": 0.19879182554303587,
      "grad_norm": 5.128887176513672,
      "learning_rate": 9.961047791053309e-05,
      "loss": 2.6390451431274413,
      "memory(GiB)": 47.63,
      "step": 4640,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.460115
    },
    {
      "epoch": 0.19900604087228482,
      "grad_norm": 3.6200530529022217,
      "learning_rate": 9.960963906560188e-05,
      "loss": 2.7532272338867188,
      "memory(GiB)": 47.63,
      "step": 4645,
      "token_acc": 0.4426751592356688,
      "train_speed(iter/s)": 1.459993
    },
    {
      "epoch": 0.19922025620153377,
      "grad_norm": 3.444218873977661,
      "learning_rate": 9.96087993219455e-05,
      "loss": 2.4867788314819337,
      "memory(GiB)": 47.63,
      "step": 4650,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.459665
    },
    {
      "epoch": 0.19943447153078275,
      "grad_norm": 4.417496681213379,
      "learning_rate": 9.960795867957921e-05,
      "loss": 2.3958377838134766,
      "memory(GiB)": 47.63,
      "step": 4655,
      "token_acc": 0.5261194029850746,
      "train_speed(iter/s)": 1.459639
    },
    {
      "epoch": 0.1996486868600317,
      "grad_norm": 3.7539174556732178,
      "learning_rate": 9.96071171385182e-05,
      "loss": 2.6219974517822267,
      "memory(GiB)": 47.63,
      "step": 4660,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.459999
    },
    {
      "epoch": 0.19986290218928066,
      "grad_norm": 3.0214028358459473,
      "learning_rate": 9.960627469877773e-05,
      "loss": 2.370345115661621,
      "memory(GiB)": 47.63,
      "step": 4665,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.460206
    },
    {
      "epoch": 0.20007711751852963,
      "grad_norm": 6.439914226531982,
      "learning_rate": 9.960543136037306e-05,
      "loss": 2.6158199310302734,
      "memory(GiB)": 47.63,
      "step": 4670,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.4604
    },
    {
      "epoch": 0.20029133284777859,
      "grad_norm": 4.3249993324279785,
      "learning_rate": 9.960458712331946e-05,
      "loss": 2.337392807006836,
      "memory(GiB)": 47.63,
      "step": 4675,
      "token_acc": 0.5520833333333334,
      "train_speed(iter/s)": 1.460867
    },
    {
      "epoch": 0.20050554817702754,
      "grad_norm": 3.847118377685547,
      "learning_rate": 9.960374198763225e-05,
      "loss": 2.4481176376342773,
      "memory(GiB)": 47.63,
      "step": 4680,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.460848
    },
    {
      "epoch": 0.20071976350627652,
      "grad_norm": 3.2964940071105957,
      "learning_rate": 9.96028959533267e-05,
      "loss": 2.810785102844238,
      "memory(GiB)": 47.63,
      "step": 4685,
      "token_acc": 0.431266846361186,
      "train_speed(iter/s)": 1.461136
    },
    {
      "epoch": 0.20093397883552547,
      "grad_norm": 4.322447299957275,
      "learning_rate": 9.960204902041816e-05,
      "loss": 2.8656097412109376,
      "memory(GiB)": 47.63,
      "step": 4690,
      "token_acc": 0.467680608365019,
      "train_speed(iter/s)": 1.461308
    },
    {
      "epoch": 0.20114819416477442,
      "grad_norm": 5.639817714691162,
      "learning_rate": 9.960120118892199e-05,
      "loss": 2.6024860382080077,
      "memory(GiB)": 47.63,
      "step": 4695,
      "token_acc": 0.44106463878326996,
      "train_speed(iter/s)": 1.460901
    },
    {
      "epoch": 0.2013624094940234,
      "grad_norm": 3.873734951019287,
      "learning_rate": 9.960035245885352e-05,
      "loss": 2.8162139892578124,
      "memory(GiB)": 47.63,
      "step": 4700,
      "token_acc": 0.4626038781163435,
      "train_speed(iter/s)": 1.460887
    },
    {
      "epoch": 0.20157662482327235,
      "grad_norm": 4.02492618560791,
      "learning_rate": 9.959950283022813e-05,
      "loss": 2.5812231063842774,
      "memory(GiB)": 47.63,
      "step": 4705,
      "token_acc": 0.4837758112094395,
      "train_speed(iter/s)": 1.46106
    },
    {
      "epoch": 0.20179084015252133,
      "grad_norm": 4.980051517486572,
      "learning_rate": 9.959865230306122e-05,
      "loss": 2.585065460205078,
      "memory(GiB)": 47.63,
      "step": 4710,
      "token_acc": 0.4830188679245283,
      "train_speed(iter/s)": 1.460656
    },
    {
      "epoch": 0.20200505548177028,
      "grad_norm": 3.0168967247009277,
      "learning_rate": 9.959780087736821e-05,
      "loss": 2.552627372741699,
      "memory(GiB)": 47.63,
      "step": 4715,
      "token_acc": 0.46283783783783783,
      "train_speed(iter/s)": 1.460619
    },
    {
      "epoch": 0.20221927081101923,
      "grad_norm": 7.62388277053833,
      "learning_rate": 9.959694855316451e-05,
      "loss": 2.8916736602783204,
      "memory(GiB)": 47.63,
      "step": 4720,
      "token_acc": 0.4303030303030303,
      "train_speed(iter/s)": 1.46052
    },
    {
      "epoch": 0.2024334861402682,
      "grad_norm": 4.516824722290039,
      "learning_rate": 9.959609533046554e-05,
      "loss": 2.6287261962890627,
      "memory(GiB)": 47.63,
      "step": 4725,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.460771
    },
    {
      "epoch": 0.20264770146951716,
      "grad_norm": 4.003058910369873,
      "learning_rate": 9.959524120928678e-05,
      "loss": 2.6882965087890627,
      "memory(GiB)": 47.63,
      "step": 4730,
      "token_acc": 0.4155124653739612,
      "train_speed(iter/s)": 1.46094
    },
    {
      "epoch": 0.2028619167987661,
      "grad_norm": 4.212945938110352,
      "learning_rate": 9.959438618964372e-05,
      "loss": 2.74023551940918,
      "memory(GiB)": 47.63,
      "step": 4735,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.460756
    },
    {
      "epoch": 0.2030761321280151,
      "grad_norm": 4.152052879333496,
      "learning_rate": 9.959353027155183e-05,
      "loss": 2.5586921691894533,
      "memory(GiB)": 47.63,
      "step": 4740,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.461207
    },
    {
      "epoch": 0.20329034745726404,
      "grad_norm": 4.194866180419922,
      "learning_rate": 9.959267345502661e-05,
      "loss": 2.62713623046875,
      "memory(GiB)": 47.63,
      "step": 4745,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.461366
    },
    {
      "epoch": 0.203504562786513,
      "grad_norm": 4.050347805023193,
      "learning_rate": 9.959181574008358e-05,
      "loss": 2.79455509185791,
      "memory(GiB)": 47.63,
      "step": 4750,
      "token_acc": 0.42424242424242425,
      "train_speed(iter/s)": 1.461483
    },
    {
      "epoch": 0.20371877811576197,
      "grad_norm": 5.109241962432861,
      "learning_rate": 9.95909571267383e-05,
      "loss": 2.6086854934692383,
      "memory(GiB)": 47.63,
      "step": 4755,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.460906
    },
    {
      "epoch": 0.20393299344501092,
      "grad_norm": 4.411006927490234,
      "learning_rate": 9.95900976150063e-05,
      "loss": 2.4530099868774413,
      "memory(GiB)": 47.63,
      "step": 4760,
      "token_acc": 0.458955223880597,
      "train_speed(iter/s)": 1.460873
    },
    {
      "epoch": 0.20414720877425988,
      "grad_norm": 4.208810329437256,
      "learning_rate": 9.958923720490317e-05,
      "loss": 2.8047142028808594,
      "memory(GiB)": 47.63,
      "step": 4765,
      "token_acc": 0.43050847457627117,
      "train_speed(iter/s)": 1.461025
    },
    {
      "epoch": 0.20436142410350885,
      "grad_norm": 5.848051071166992,
      "learning_rate": 9.958837589644449e-05,
      "loss": 2.3469343185424805,
      "memory(GiB)": 47.63,
      "step": 4770,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.460712
    },
    {
      "epoch": 0.2045756394327578,
      "grad_norm": 3.1968307495117188,
      "learning_rate": 9.958751368964585e-05,
      "loss": 2.4620996475219727,
      "memory(GiB)": 47.63,
      "step": 4775,
      "token_acc": 0.5038759689922481,
      "train_speed(iter/s)": 1.460871
    },
    {
      "epoch": 0.20478985476200676,
      "grad_norm": 4.222325325012207,
      "learning_rate": 9.95866505845229e-05,
      "loss": 2.704343795776367,
      "memory(GiB)": 47.63,
      "step": 4780,
      "token_acc": 0.4359861591695502,
      "train_speed(iter/s)": 1.461128
    },
    {
      "epoch": 0.20500407009125574,
      "grad_norm": 3.7278714179992676,
      "learning_rate": 9.958578658109125e-05,
      "loss": 2.838206100463867,
      "memory(GiB)": 47.63,
      "step": 4785,
      "token_acc": 0.45323741007194246,
      "train_speed(iter/s)": 1.461544
    },
    {
      "epoch": 0.2052182854205047,
      "grad_norm": 3.182396650314331,
      "learning_rate": 9.958492167936658e-05,
      "loss": 2.4889122009277345,
      "memory(GiB)": 47.63,
      "step": 4790,
      "token_acc": 0.4872611464968153,
      "train_speed(iter/s)": 1.461955
    },
    {
      "epoch": 0.20543250074975367,
      "grad_norm": 3.691033363342285,
      "learning_rate": 9.958405587936452e-05,
      "loss": 2.598044013977051,
      "memory(GiB)": 47.63,
      "step": 4795,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.461961
    },
    {
      "epoch": 0.20564671607900262,
      "grad_norm": 4.164496898651123,
      "learning_rate": 9.958318918110077e-05,
      "loss": 2.354471206665039,
      "memory(GiB)": 47.63,
      "step": 4800,
      "token_acc": 0.5313531353135313,
      "train_speed(iter/s)": 1.461688
    },
    {
      "epoch": 0.20586093140825157,
      "grad_norm": 5.045453071594238,
      "learning_rate": 9.958232158459104e-05,
      "loss": 2.482447052001953,
      "memory(GiB)": 47.63,
      "step": 4805,
      "token_acc": 0.47639484978540775,
      "train_speed(iter/s)": 1.461909
    },
    {
      "epoch": 0.20607514673750055,
      "grad_norm": 3.663247585296631,
      "learning_rate": 9.958145308985105e-05,
      "loss": 2.739253616333008,
      "memory(GiB)": 47.63,
      "step": 4810,
      "token_acc": 0.4456140350877193,
      "train_speed(iter/s)": 1.461809
    },
    {
      "epoch": 0.2062893620667495,
      "grad_norm": 3.318002462387085,
      "learning_rate": 9.958058369689651e-05,
      "loss": 2.5637796401977537,
      "memory(GiB)": 47.63,
      "step": 4815,
      "token_acc": 0.491869918699187,
      "train_speed(iter/s)": 1.462036
    },
    {
      "epoch": 0.20650357739599845,
      "grad_norm": 3.3069710731506348,
      "learning_rate": 9.95797134057432e-05,
      "loss": 2.7852516174316406,
      "memory(GiB)": 47.63,
      "step": 4820,
      "token_acc": 0.4592833876221498,
      "train_speed(iter/s)": 1.462307
    },
    {
      "epoch": 0.20671779272524743,
      "grad_norm": 2.944810628890991,
      "learning_rate": 9.957884221640687e-05,
      "loss": 2.959232711791992,
      "memory(GiB)": 47.63,
      "step": 4825,
      "token_acc": 0.4103448275862069,
      "train_speed(iter/s)": 1.462467
    },
    {
      "epoch": 0.20693200805449638,
      "grad_norm": 5.038721084594727,
      "learning_rate": 9.95779701289033e-05,
      "loss": 2.544814682006836,
      "memory(GiB)": 47.63,
      "step": 4830,
      "token_acc": 0.49606299212598426,
      "train_speed(iter/s)": 1.462689
    },
    {
      "epoch": 0.20714622338374533,
      "grad_norm": 4.850725173950195,
      "learning_rate": 9.95770971432483e-05,
      "loss": 2.634608268737793,
      "memory(GiB)": 47.63,
      "step": 4835,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.462719
    },
    {
      "epoch": 0.2073604387129943,
      "grad_norm": 4.1277174949646,
      "learning_rate": 9.957622325945768e-05,
      "loss": 2.6970439910888673,
      "memory(GiB)": 47.63,
      "step": 4840,
      "token_acc": 0.4519230769230769,
      "train_speed(iter/s)": 1.46275
    },
    {
      "epoch": 0.20757465404224326,
      "grad_norm": 3.763849973678589,
      "learning_rate": 9.957534847754726e-05,
      "loss": 2.8271854400634764,
      "memory(GiB)": 47.63,
      "step": 4845,
      "token_acc": 0.44805194805194803,
      "train_speed(iter/s)": 1.462949
    },
    {
      "epoch": 0.2077888693714922,
      "grad_norm": 4.403916835784912,
      "learning_rate": 9.95744727975329e-05,
      "loss": 2.81970100402832,
      "memory(GiB)": 47.63,
      "step": 4850,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.463301
    },
    {
      "epoch": 0.2080030847007412,
      "grad_norm": 6.187358379364014,
      "learning_rate": 9.957359621943045e-05,
      "loss": 2.487234878540039,
      "memory(GiB)": 47.63,
      "step": 4855,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.463579
    },
    {
      "epoch": 0.20821730002999014,
      "grad_norm": 3.3152196407318115,
      "learning_rate": 9.957271874325582e-05,
      "loss": 2.536726379394531,
      "memory(GiB)": 47.63,
      "step": 4860,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.464034
    },
    {
      "epoch": 0.2084315153592391,
      "grad_norm": 3.83278489112854,
      "learning_rate": 9.957184036902488e-05,
      "loss": 2.633724594116211,
      "memory(GiB)": 47.63,
      "step": 4865,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.464009
    },
    {
      "epoch": 0.20864573068848807,
      "grad_norm": 5.216366291046143,
      "learning_rate": 9.957096109675354e-05,
      "loss": 2.8043491363525392,
      "memory(GiB)": 47.63,
      "step": 4870,
      "token_acc": 0.46706586826347307,
      "train_speed(iter/s)": 1.463764
    },
    {
      "epoch": 0.20885994601773702,
      "grad_norm": 3.232802629470825,
      "learning_rate": 9.957008092645777e-05,
      "loss": 2.591070365905762,
      "memory(GiB)": 47.63,
      "step": 4875,
      "token_acc": 0.4454828660436137,
      "train_speed(iter/s)": 1.463808
    },
    {
      "epoch": 0.209074161346986,
      "grad_norm": 20.843944549560547,
      "learning_rate": 9.956919985815345e-05,
      "loss": 2.6538322448730467,
      "memory(GiB)": 47.63,
      "step": 4880,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.463636
    },
    {
      "epoch": 0.20928837667623496,
      "grad_norm": 3.754133701324463,
      "learning_rate": 9.956831789185659e-05,
      "loss": 2.4146175384521484,
      "memory(GiB)": 47.63,
      "step": 4885,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.463367
    },
    {
      "epoch": 0.2095025920054839,
      "grad_norm": 7.558262825012207,
      "learning_rate": 9.956743502758315e-05,
      "loss": 2.7335124969482423,
      "memory(GiB)": 47.63,
      "step": 4890,
      "token_acc": 0.4652777777777778,
      "train_speed(iter/s)": 1.463483
    },
    {
      "epoch": 0.20971680733473289,
      "grad_norm": 2.960104465484619,
      "learning_rate": 9.956655126534911e-05,
      "loss": 2.3875946044921874,
      "memory(GiB)": 47.63,
      "step": 4895,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.463796
    },
    {
      "epoch": 0.20993102266398184,
      "grad_norm": 5.942025661468506,
      "learning_rate": 9.956566660517053e-05,
      "loss": 2.312450408935547,
      "memory(GiB)": 47.63,
      "step": 4900,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.463928
    },
    {
      "epoch": 0.2101452379932308,
      "grad_norm": 5.1008687019348145,
      "learning_rate": 9.956478104706337e-05,
      "loss": 2.6202777862548827,
      "memory(GiB)": 47.63,
      "step": 4905,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.464154
    },
    {
      "epoch": 0.21035945332247977,
      "grad_norm": 4.184366226196289,
      "learning_rate": 9.956389459104372e-05,
      "loss": 2.444474220275879,
      "memory(GiB)": 47.63,
      "step": 4910,
      "token_acc": 0.43272727272727274,
      "train_speed(iter/s)": 1.464534
    },
    {
      "epoch": 0.21057366865172872,
      "grad_norm": 3.3858139514923096,
      "learning_rate": 9.956300723712764e-05,
      "loss": 2.7998870849609374,
      "memory(GiB)": 47.63,
      "step": 4915,
      "token_acc": 0.41040462427745666,
      "train_speed(iter/s)": 1.464655
    },
    {
      "epoch": 0.21078788398097767,
      "grad_norm": 3.690751552581787,
      "learning_rate": 9.956211898533117e-05,
      "loss": 2.331569290161133,
      "memory(GiB)": 47.63,
      "step": 4920,
      "token_acc": 0.5143769968051118,
      "train_speed(iter/s)": 1.465006
    },
    {
      "epoch": 0.21100209931022665,
      "grad_norm": 6.2631354331970215,
      "learning_rate": 9.956122983567042e-05,
      "loss": 2.435498046875,
      "memory(GiB)": 47.63,
      "step": 4925,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.464621
    },
    {
      "epoch": 0.2112163146394756,
      "grad_norm": 3.335134744644165,
      "learning_rate": 9.95603397881615e-05,
      "loss": 2.3452857971191405,
      "memory(GiB)": 47.63,
      "step": 4930,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.464787
    },
    {
      "epoch": 0.21143052996872455,
      "grad_norm": 3.005373477935791,
      "learning_rate": 9.955944884282055e-05,
      "loss": 2.612366485595703,
      "memory(GiB)": 47.63,
      "step": 4935,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.465028
    },
    {
      "epoch": 0.21164474529797353,
      "grad_norm": 3.1305315494537354,
      "learning_rate": 9.955855699966368e-05,
      "loss": 2.6456478118896483,
      "memory(GiB)": 47.63,
      "step": 4940,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.465002
    },
    {
      "epoch": 0.21185896062722248,
      "grad_norm": 3.9591519832611084,
      "learning_rate": 9.955766425870706e-05,
      "loss": 3.0028560638427733,
      "memory(GiB)": 47.63,
      "step": 4945,
      "token_acc": 0.4119496855345912,
      "train_speed(iter/s)": 1.464684
    },
    {
      "epoch": 0.21207317595647143,
      "grad_norm": 5.0495758056640625,
      "learning_rate": 9.955677061996687e-05,
      "loss": 2.730988311767578,
      "memory(GiB)": 47.63,
      "step": 4950,
      "token_acc": 0.43636363636363634,
      "train_speed(iter/s)": 1.464715
    },
    {
      "epoch": 0.2122873912857204,
      "grad_norm": 3.475385904312134,
      "learning_rate": 9.955587608345928e-05,
      "loss": 2.548473930358887,
      "memory(GiB)": 47.63,
      "step": 4955,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.464675
    },
    {
      "epoch": 0.21250160661496936,
      "grad_norm": 3.494689702987671,
      "learning_rate": 9.955498064920052e-05,
      "loss": 2.870347595214844,
      "memory(GiB)": 47.63,
      "step": 4960,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.464637
    },
    {
      "epoch": 0.21271582194421834,
      "grad_norm": 3.259819507598877,
      "learning_rate": 9.955408431720681e-05,
      "loss": 2.629207992553711,
      "memory(GiB)": 47.63,
      "step": 4965,
      "token_acc": 0.43934426229508194,
      "train_speed(iter/s)": 1.464807
    },
    {
      "epoch": 0.2129300372734673,
      "grad_norm": 4.76029634475708,
      "learning_rate": 9.955318708749435e-05,
      "loss": 2.643509864807129,
      "memory(GiB)": 47.63,
      "step": 4970,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.464783
    },
    {
      "epoch": 0.21314425260271624,
      "grad_norm": 4.266506671905518,
      "learning_rate": 9.955228896007944e-05,
      "loss": 2.6341203689575194,
      "memory(GiB)": 47.63,
      "step": 4975,
      "token_acc": 0.44313725490196076,
      "train_speed(iter/s)": 1.464767
    },
    {
      "epoch": 0.21335846793196522,
      "grad_norm": 4.462551593780518,
      "learning_rate": 9.955138993497832e-05,
      "loss": 2.8413829803466797,
      "memory(GiB)": 47.63,
      "step": 4980,
      "token_acc": 0.42905405405405406,
      "train_speed(iter/s)": 1.464523
    },
    {
      "epoch": 0.21357268326121417,
      "grad_norm": 4.464143753051758,
      "learning_rate": 9.95504900122073e-05,
      "loss": 2.2535449981689455,
      "memory(GiB)": 47.63,
      "step": 4985,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.464764
    },
    {
      "epoch": 0.21378689859046313,
      "grad_norm": 4.146317481994629,
      "learning_rate": 9.954958919178265e-05,
      "loss": 2.2576440811157226,
      "memory(GiB)": 47.63,
      "step": 4990,
      "token_acc": 0.5102880658436214,
      "train_speed(iter/s)": 1.464386
    },
    {
      "epoch": 0.2140011139197121,
      "grad_norm": 3.9403140544891357,
      "learning_rate": 9.954868747372074e-05,
      "loss": 2.723891830444336,
      "memory(GiB)": 47.63,
      "step": 4995,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.464076
    },
    {
      "epoch": 0.21421532924896106,
      "grad_norm": 3.9458513259887695,
      "learning_rate": 9.954778485803787e-05,
      "loss": 2.4841426849365233,
      "memory(GiB)": 47.63,
      "step": 5000,
      "token_acc": 0.45038167938931295,
      "train_speed(iter/s)": 1.463911
    },
    {
      "epoch": 0.21421532924896106,
      "eval_loss": 2.3903603553771973,
      "eval_runtime": 13.3836,
      "eval_samples_per_second": 7.472,
      "eval_steps_per_second": 7.472,
      "eval_token_acc": 0.4508816120906801,
      "step": 5000
    },
    {
      "epoch": 0.21442954457821,
      "grad_norm": 4.423367023468018,
      "learning_rate": 9.954688134475038e-05,
      "loss": 2.759851837158203,
      "memory(GiB)": 47.63,
      "step": 5005,
      "token_acc": 0.4473924977127173,
      "train_speed(iter/s)": 1.457879
    },
    {
      "epoch": 0.214643759907459,
      "grad_norm": 3.4486584663391113,
      "learning_rate": 9.954597693387467e-05,
      "loss": 2.809971046447754,
      "memory(GiB)": 47.63,
      "step": 5010,
      "token_acc": 0.40853658536585363,
      "train_speed(iter/s)": 1.457742
    },
    {
      "epoch": 0.21485797523670794,
      "grad_norm": 3.5653340816497803,
      "learning_rate": 9.95450716254271e-05,
      "loss": 2.5606967926025392,
      "memory(GiB)": 47.63,
      "step": 5015,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.45784
    },
    {
      "epoch": 0.2150721905659569,
      "grad_norm": 6.181088924407959,
      "learning_rate": 9.954416541942408e-05,
      "loss": 3.0325870513916016,
      "memory(GiB)": 47.63,
      "step": 5020,
      "token_acc": 0.4220532319391635,
      "train_speed(iter/s)": 1.458003
    },
    {
      "epoch": 0.21528640589520587,
      "grad_norm": 5.29486083984375,
      "learning_rate": 9.954325831588204e-05,
      "loss": 2.6247535705566407,
      "memory(GiB)": 47.63,
      "step": 5025,
      "token_acc": 0.45674740484429066,
      "train_speed(iter/s)": 1.458211
    },
    {
      "epoch": 0.21550062122445482,
      "grad_norm": 4.460639953613281,
      "learning_rate": 9.954235031481739e-05,
      "loss": 2.4387359619140625,
      "memory(GiB)": 47.63,
      "step": 5030,
      "token_acc": 0.4769874476987448,
      "train_speed(iter/s)": 1.45829
    },
    {
      "epoch": 0.21571483655370377,
      "grad_norm": 3.4331841468811035,
      "learning_rate": 9.95414414162466e-05,
      "loss": 2.5991966247558596,
      "memory(GiB)": 47.63,
      "step": 5035,
      "token_acc": 0.4754601226993865,
      "train_speed(iter/s)": 1.458297
    },
    {
      "epoch": 0.21592905188295275,
      "grad_norm": 6.638716697692871,
      "learning_rate": 9.95405316201861e-05,
      "loss": 2.6689929962158203,
      "memory(GiB)": 47.63,
      "step": 5040,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.458547
    },
    {
      "epoch": 0.2161432672122017,
      "grad_norm": 4.650533676147461,
      "learning_rate": 9.953962092665243e-05,
      "loss": 2.536483383178711,
      "memory(GiB)": 47.63,
      "step": 5045,
      "token_acc": 0.45149253731343286,
      "train_speed(iter/s)": 1.45859
    },
    {
      "epoch": 0.21635748254145068,
      "grad_norm": 4.859783172607422,
      "learning_rate": 9.953870933566203e-05,
      "loss": 2.8836605072021486,
      "memory(GiB)": 47.63,
      "step": 5050,
      "token_acc": 0.4358974358974359,
      "train_speed(iter/s)": 1.458408
    },
    {
      "epoch": 0.21657169787069963,
      "grad_norm": 3.7098331451416016,
      "learning_rate": 9.953779684723146e-05,
      "loss": 2.5904306411743163,
      "memory(GiB)": 47.63,
      "step": 5055,
      "token_acc": 0.444794952681388,
      "train_speed(iter/s)": 1.458505
    },
    {
      "epoch": 0.21678591319994858,
      "grad_norm": 3.6314024925231934,
      "learning_rate": 9.953688346137722e-05,
      "loss": 2.781103515625,
      "memory(GiB)": 47.63,
      "step": 5060,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.458329
    },
    {
      "epoch": 0.21700012852919756,
      "grad_norm": 6.57258939743042,
      "learning_rate": 9.953596917811586e-05,
      "loss": 2.7235851287841797,
      "memory(GiB)": 47.63,
      "step": 5065,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.458626
    },
    {
      "epoch": 0.2172143438584465,
      "grad_norm": 5.607734680175781,
      "learning_rate": 9.953505399746395e-05,
      "loss": 2.539692687988281,
      "memory(GiB)": 47.63,
      "step": 5070,
      "token_acc": 0.4384858044164038,
      "train_speed(iter/s)": 1.458775
    },
    {
      "epoch": 0.21742855918769546,
      "grad_norm": 3.690596103668213,
      "learning_rate": 9.953413791943808e-05,
      "loss": 2.4824382781982424,
      "memory(GiB)": 47.63,
      "step": 5075,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.45831
    },
    {
      "epoch": 0.21764277451694444,
      "grad_norm": 4.219900131225586,
      "learning_rate": 9.953322094405482e-05,
      "loss": 2.744942855834961,
      "memory(GiB)": 47.63,
      "step": 5080,
      "token_acc": 0.4326241134751773,
      "train_speed(iter/s)": 1.458451
    },
    {
      "epoch": 0.2178569898461934,
      "grad_norm": 4.080595970153809,
      "learning_rate": 9.953230307133082e-05,
      "loss": 2.258548355102539,
      "memory(GiB)": 47.63,
      "step": 5085,
      "token_acc": 0.5038759689922481,
      "train_speed(iter/s)": 1.458626
    },
    {
      "epoch": 0.21807120517544235,
      "grad_norm": 5.973474025726318,
      "learning_rate": 9.953138430128266e-05,
      "loss": 2.804652786254883,
      "memory(GiB)": 47.63,
      "step": 5090,
      "token_acc": 0.41124260355029585,
      "train_speed(iter/s)": 1.458938
    },
    {
      "epoch": 0.21828542050469132,
      "grad_norm": 3.681203603744507,
      "learning_rate": 9.953046463392703e-05,
      "loss": 2.58138370513916,
      "memory(GiB)": 47.63,
      "step": 5095,
      "token_acc": 0.42990654205607476,
      "train_speed(iter/s)": 1.458989
    },
    {
      "epoch": 0.21849963583394028,
      "grad_norm": 3.3350465297698975,
      "learning_rate": 9.952954406928056e-05,
      "loss": 2.6960800170898436,
      "memory(GiB)": 47.63,
      "step": 5100,
      "token_acc": 0.4306784660766962,
      "train_speed(iter/s)": 1.459144
    },
    {
      "epoch": 0.21871385116318923,
      "grad_norm": 3.524101495742798,
      "learning_rate": 9.952862260735993e-05,
      "loss": 2.7578704833984373,
      "memory(GiB)": 47.63,
      "step": 5105,
      "token_acc": 0.4548611111111111,
      "train_speed(iter/s)": 1.459363
    },
    {
      "epoch": 0.2189280664924382,
      "grad_norm": 4.6973876953125,
      "learning_rate": 9.952770024818185e-05,
      "loss": 2.4339035034179686,
      "memory(GiB)": 47.63,
      "step": 5110,
      "token_acc": 0.5060728744939271,
      "train_speed(iter/s)": 1.459395
    },
    {
      "epoch": 0.21914228182168716,
      "grad_norm": 4.874196529388428,
      "learning_rate": 9.952677699176301e-05,
      "loss": 2.8538881301879884,
      "memory(GiB)": 47.63,
      "step": 5115,
      "token_acc": 0.4121212121212121,
      "train_speed(iter/s)": 1.459289
    },
    {
      "epoch": 0.2193564971509361,
      "grad_norm": 4.868494510650635,
      "learning_rate": 9.952585283812015e-05,
      "loss": 2.8529348373413086,
      "memory(GiB)": 47.63,
      "step": 5120,
      "token_acc": 0.43253968253968256,
      "train_speed(iter/s)": 1.459651
    },
    {
      "epoch": 0.2195707124801851,
      "grad_norm": 3.379431962966919,
      "learning_rate": 9.952492778727e-05,
      "loss": 2.5719249725341795,
      "memory(GiB)": 47.63,
      "step": 5125,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.459488
    },
    {
      "epoch": 0.21978492780943404,
      "grad_norm": 5.336353302001953,
      "learning_rate": 9.952400183922933e-05,
      "loss": 2.547113037109375,
      "memory(GiB)": 47.63,
      "step": 5130,
      "token_acc": 0.44357976653696496,
      "train_speed(iter/s)": 1.459376
    },
    {
      "epoch": 0.21999914313868302,
      "grad_norm": 3.798307418823242,
      "learning_rate": 9.952307499401492e-05,
      "loss": 2.483036422729492,
      "memory(GiB)": 47.63,
      "step": 5135,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.459064
    },
    {
      "epoch": 0.22021335846793197,
      "grad_norm": 4.275089740753174,
      "learning_rate": 9.952214725164355e-05,
      "loss": 2.9046915054321287,
      "memory(GiB)": 47.63,
      "step": 5140,
      "token_acc": 0.43359375,
      "train_speed(iter/s)": 1.459014
    },
    {
      "epoch": 0.22042757379718092,
      "grad_norm": 3.687201499938965,
      "learning_rate": 9.952121861213202e-05,
      "loss": 2.355246734619141,
      "memory(GiB)": 47.63,
      "step": 5145,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.45946
    },
    {
      "epoch": 0.2206417891264299,
      "grad_norm": 4.156665325164795,
      "learning_rate": 9.952028907549715e-05,
      "loss": 2.884083557128906,
      "memory(GiB)": 47.63,
      "step": 5150,
      "token_acc": 0.4307228915662651,
      "train_speed(iter/s)": 1.459692
    },
    {
      "epoch": 0.22085600445567885,
      "grad_norm": 6.38535213470459,
      "learning_rate": 9.951935864175581e-05,
      "loss": 2.8706649780273437,
      "memory(GiB)": 47.63,
      "step": 5155,
      "token_acc": 0.4163934426229508,
      "train_speed(iter/s)": 1.459908
    },
    {
      "epoch": 0.2210702197849278,
      "grad_norm": 5.345571994781494,
      "learning_rate": 9.951842731092482e-05,
      "loss": 2.9905134201049806,
      "memory(GiB)": 47.63,
      "step": 5160,
      "token_acc": 0.4596774193548387,
      "train_speed(iter/s)": 1.459867
    },
    {
      "epoch": 0.22128443511417678,
      "grad_norm": 3.282136917114258,
      "learning_rate": 9.951749508302106e-05,
      "loss": 2.6577373504638673,
      "memory(GiB)": 47.63,
      "step": 5165,
      "token_acc": 0.45121951219512196,
      "train_speed(iter/s)": 1.459848
    },
    {
      "epoch": 0.22149865044342573,
      "grad_norm": 3.052436590194702,
      "learning_rate": 9.951656195806145e-05,
      "loss": 2.5006778717041014,
      "memory(GiB)": 47.63,
      "step": 5170,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.459664
    },
    {
      "epoch": 0.22171286577267468,
      "grad_norm": 4.098628997802734,
      "learning_rate": 9.951562793606286e-05,
      "loss": 2.9629280090332033,
      "memory(GiB)": 47.63,
      "step": 5175,
      "token_acc": 0.4275618374558304,
      "train_speed(iter/s)": 1.460055
    },
    {
      "epoch": 0.22192708110192366,
      "grad_norm": 4.604274272918701,
      "learning_rate": 9.951469301704221e-05,
      "loss": 2.59681396484375,
      "memory(GiB)": 47.63,
      "step": 5180,
      "token_acc": 0.461864406779661,
      "train_speed(iter/s)": 1.460104
    },
    {
      "epoch": 0.2221412964311726,
      "grad_norm": 3.2579591274261475,
      "learning_rate": 9.951375720101645e-05,
      "loss": 2.9324100494384764,
      "memory(GiB)": 47.63,
      "step": 5185,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.460165
    },
    {
      "epoch": 0.22235551176042156,
      "grad_norm": 3.7421793937683105,
      "learning_rate": 9.951282048800255e-05,
      "loss": 2.741208076477051,
      "memory(GiB)": 47.63,
      "step": 5190,
      "token_acc": 0.44166666666666665,
      "train_speed(iter/s)": 1.460467
    },
    {
      "epoch": 0.22256972708967054,
      "grad_norm": 3.7896907329559326,
      "learning_rate": 9.951188287801744e-05,
      "loss": 2.325700569152832,
      "memory(GiB)": 47.63,
      "step": 5195,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.460124
    },
    {
      "epoch": 0.2227839424189195,
      "grad_norm": 4.166207790374756,
      "learning_rate": 9.951094437107812e-05,
      "loss": 2.6317943572998046,
      "memory(GiB)": 47.63,
      "step": 5200,
      "token_acc": 0.4773662551440329,
      "train_speed(iter/s)": 1.459908
    },
    {
      "epoch": 0.22299815774816845,
      "grad_norm": 3.2314565181732178,
      "learning_rate": 9.951000496720162e-05,
      "loss": 2.9743555068969725,
      "memory(GiB)": 47.63,
      "step": 5205,
      "token_acc": 0.41883116883116883,
      "train_speed(iter/s)": 1.460172
    },
    {
      "epoch": 0.22321237307741743,
      "grad_norm": 3.753066301345825,
      "learning_rate": 9.950906466640493e-05,
      "loss": 2.726948356628418,
      "memory(GiB)": 47.63,
      "step": 5210,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.460003
    },
    {
      "epoch": 0.22342658840666638,
      "grad_norm": 3.9496212005615234,
      "learning_rate": 9.950812346870508e-05,
      "loss": 2.732045555114746,
      "memory(GiB)": 47.63,
      "step": 5215,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.460337
    },
    {
      "epoch": 0.22364080373591536,
      "grad_norm": 5.344078063964844,
      "learning_rate": 9.950718137411913e-05,
      "loss": 2.6071565628051756,
      "memory(GiB)": 47.63,
      "step": 5220,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.460541
    },
    {
      "epoch": 0.2238550190651643,
      "grad_norm": 8.683192253112793,
      "learning_rate": 9.950623838266415e-05,
      "loss": 2.644216537475586,
      "memory(GiB)": 47.63,
      "step": 5225,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.46044
    },
    {
      "epoch": 0.22406923439441326,
      "grad_norm": 4.4640583992004395,
      "learning_rate": 9.950529449435722e-05,
      "loss": 2.8277259826660157,
      "memory(GiB)": 47.63,
      "step": 5230,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.460534
    },
    {
      "epoch": 0.22428344972366224,
      "grad_norm": 4.138047218322754,
      "learning_rate": 9.950434970921544e-05,
      "loss": 2.5987066268920898,
      "memory(GiB)": 47.63,
      "step": 5235,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.460354
    },
    {
      "epoch": 0.2244976650529112,
      "grad_norm": 2.9940555095672607,
      "learning_rate": 9.950340402725596e-05,
      "loss": 2.5609241485595704,
      "memory(GiB)": 47.63,
      "step": 5240,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.460535
    },
    {
      "epoch": 0.22471188038216014,
      "grad_norm": 3.2722036838531494,
      "learning_rate": 9.950245744849583e-05,
      "loss": 2.501548004150391,
      "memory(GiB)": 47.63,
      "step": 5245,
      "token_acc": 0.43727598566308246,
      "train_speed(iter/s)": 1.46079
    },
    {
      "epoch": 0.22492609571140912,
      "grad_norm": 5.219012260437012,
      "learning_rate": 9.950150997295226e-05,
      "loss": 2.4772377014160156,
      "memory(GiB)": 47.63,
      "step": 5250,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.460816
    },
    {
      "epoch": 0.22514031104065807,
      "grad_norm": 4.120798110961914,
      "learning_rate": 9.950056160064242e-05,
      "loss": 2.575048637390137,
      "memory(GiB)": 47.63,
      "step": 5255,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.460547
    },
    {
      "epoch": 0.22535452636990702,
      "grad_norm": 3.0218071937561035,
      "learning_rate": 9.949961233158346e-05,
      "loss": 2.4449541091918947,
      "memory(GiB)": 47.63,
      "step": 5260,
      "token_acc": 0.46226415094339623,
      "train_speed(iter/s)": 1.46073
    },
    {
      "epoch": 0.225568741699156,
      "grad_norm": 4.176855564117432,
      "learning_rate": 9.949866216579258e-05,
      "loss": 2.899913787841797,
      "memory(GiB)": 47.63,
      "step": 5265,
      "token_acc": 0.43986254295532645,
      "train_speed(iter/s)": 1.460585
    },
    {
      "epoch": 0.22578295702840495,
      "grad_norm": 4.538032054901123,
      "learning_rate": 9.949771110328701e-05,
      "loss": 2.5551383972167967,
      "memory(GiB)": 47.63,
      "step": 5270,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.460747
    },
    {
      "epoch": 0.2259971723576539,
      "grad_norm": 3.67073392868042,
      "learning_rate": 9.949675914408396e-05,
      "loss": 2.813393402099609,
      "memory(GiB)": 47.63,
      "step": 5275,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.461141
    },
    {
      "epoch": 0.22621138768690288,
      "grad_norm": 4.147843837738037,
      "learning_rate": 9.94958062882007e-05,
      "loss": 2.7792625427246094,
      "memory(GiB)": 47.63,
      "step": 5280,
      "token_acc": 0.46394984326018807,
      "train_speed(iter/s)": 1.461271
    },
    {
      "epoch": 0.22642560301615183,
      "grad_norm": 3.92873477935791,
      "learning_rate": 9.949485253565446e-05,
      "loss": 2.63089542388916,
      "memory(GiB)": 47.63,
      "step": 5285,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.461306
    },
    {
      "epoch": 0.22663981834540078,
      "grad_norm": 3.239232301712036,
      "learning_rate": 9.949389788646257e-05,
      "loss": 2.5243959426879883,
      "memory(GiB)": 47.63,
      "step": 5290,
      "token_acc": 0.4674922600619195,
      "train_speed(iter/s)": 1.461332
    },
    {
      "epoch": 0.22685403367464976,
      "grad_norm": 3.9025206565856934,
      "learning_rate": 9.949294234064226e-05,
      "loss": 2.4677541732788084,
      "memory(GiB)": 47.63,
      "step": 5295,
      "token_acc": 0.47278911564625853,
      "train_speed(iter/s)": 1.461653
    },
    {
      "epoch": 0.22706824900389871,
      "grad_norm": 3.9914939403533936,
      "learning_rate": 9.949198589821089e-05,
      "loss": 2.453841972351074,
      "memory(GiB)": 47.63,
      "step": 5300,
      "token_acc": 0.4981549815498155,
      "train_speed(iter/s)": 1.46185
    },
    {
      "epoch": 0.2272824643331477,
      "grad_norm": 4.803912162780762,
      "learning_rate": 9.949102855918575e-05,
      "loss": 2.6950902938842773,
      "memory(GiB)": 47.63,
      "step": 5305,
      "token_acc": 0.4481707317073171,
      "train_speed(iter/s)": 1.462137
    },
    {
      "epoch": 0.22749667966239664,
      "grad_norm": 3.7592363357543945,
      "learning_rate": 9.949007032358422e-05,
      "loss": 2.448274040222168,
      "memory(GiB)": 47.63,
      "step": 5310,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.462024
    },
    {
      "epoch": 0.2277108949916456,
      "grad_norm": 7.815985202789307,
      "learning_rate": 9.948911119142363e-05,
      "loss": 2.5520002365112306,
      "memory(GiB)": 47.63,
      "step": 5315,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.461873
    },
    {
      "epoch": 0.22792511032089458,
      "grad_norm": 5.387121200561523,
      "learning_rate": 9.948815116272137e-05,
      "loss": 2.404407501220703,
      "memory(GiB)": 47.63,
      "step": 5320,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.462019
    },
    {
      "epoch": 0.22813932565014353,
      "grad_norm": 3.177846670150757,
      "learning_rate": 9.948719023749482e-05,
      "loss": 2.678445816040039,
      "memory(GiB)": 47.63,
      "step": 5325,
      "token_acc": 0.44656488549618323,
      "train_speed(iter/s)": 1.46235
    },
    {
      "epoch": 0.22835354097939248,
      "grad_norm": 5.241051197052002,
      "learning_rate": 9.94862284157614e-05,
      "loss": 2.7848968505859375,
      "memory(GiB)": 47.63,
      "step": 5330,
      "token_acc": 0.44402985074626866,
      "train_speed(iter/s)": 1.462502
    },
    {
      "epoch": 0.22856775630864146,
      "grad_norm": 2.985356569290161,
      "learning_rate": 9.948526569753853e-05,
      "loss": 2.830743598937988,
      "memory(GiB)": 47.63,
      "step": 5335,
      "token_acc": 0.4746268656716418,
      "train_speed(iter/s)": 1.462809
    },
    {
      "epoch": 0.2287819716378904,
      "grad_norm": 3.4994444847106934,
      "learning_rate": 9.948430208284366e-05,
      "loss": 2.6473003387451173,
      "memory(GiB)": 47.63,
      "step": 5340,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.462925
    },
    {
      "epoch": 0.22899618696713936,
      "grad_norm": 2.861715078353882,
      "learning_rate": 9.948333757169424e-05,
      "loss": 2.493571662902832,
      "memory(GiB)": 47.63,
      "step": 5345,
      "token_acc": 0.5181159420289855,
      "train_speed(iter/s)": 1.462864
    },
    {
      "epoch": 0.22921040229638834,
      "grad_norm": 5.085661888122559,
      "learning_rate": 9.948237216410771e-05,
      "loss": 2.2965946197509766,
      "memory(GiB)": 47.63,
      "step": 5350,
      "token_acc": 0.47249190938511326,
      "train_speed(iter/s)": 1.463124
    },
    {
      "epoch": 0.2294246176256373,
      "grad_norm": 4.955268383026123,
      "learning_rate": 9.948140586010162e-05,
      "loss": 2.217868614196777,
      "memory(GiB)": 47.63,
      "step": 5355,
      "token_acc": 0.555956678700361,
      "train_speed(iter/s)": 1.463258
    },
    {
      "epoch": 0.22963883295488624,
      "grad_norm": 6.237241268157959,
      "learning_rate": 9.948043865969344e-05,
      "loss": 2.8866138458251953,
      "memory(GiB)": 47.63,
      "step": 5360,
      "token_acc": 0.44871794871794873,
      "train_speed(iter/s)": 1.463327
    },
    {
      "epoch": 0.22985304828413522,
      "grad_norm": 3.9926109313964844,
      "learning_rate": 9.94794705629007e-05,
      "loss": 2.698740005493164,
      "memory(GiB)": 47.63,
      "step": 5365,
      "token_acc": 0.4380952380952381,
      "train_speed(iter/s)": 1.462953
    },
    {
      "epoch": 0.23006726361338417,
      "grad_norm": 5.1145339012146,
      "learning_rate": 9.947850156974093e-05,
      "loss": 2.566942596435547,
      "memory(GiB)": 47.63,
      "step": 5370,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.462972
    },
    {
      "epoch": 0.23028147894263312,
      "grad_norm": 5.715196132659912,
      "learning_rate": 9.947753168023168e-05,
      "loss": 2.5916378021240236,
      "memory(GiB)": 47.63,
      "step": 5375,
      "token_acc": 0.4714828897338403,
      "train_speed(iter/s)": 1.46289
    },
    {
      "epoch": 0.2304956942718821,
      "grad_norm": 3.0483667850494385,
      "learning_rate": 9.947656089439055e-05,
      "loss": 2.7588014602661133,
      "memory(GiB)": 47.63,
      "step": 5380,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.462634
    },
    {
      "epoch": 0.23070990960113105,
      "grad_norm": 2.7828853130340576,
      "learning_rate": 9.947558921223509e-05,
      "loss": 2.7146541595458986,
      "memory(GiB)": 47.63,
      "step": 5385,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.462516
    },
    {
      "epoch": 0.23092412493038003,
      "grad_norm": 3.149014472961426,
      "learning_rate": 9.947461663378292e-05,
      "loss": 2.404269981384277,
      "memory(GiB)": 47.63,
      "step": 5390,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.462651
    },
    {
      "epoch": 0.23113834025962898,
      "grad_norm": 4.664316177368164,
      "learning_rate": 9.947364315905168e-05,
      "loss": 2.6491329193115236,
      "memory(GiB)": 47.63,
      "step": 5395,
      "token_acc": 0.4368932038834951,
      "train_speed(iter/s)": 1.462678
    },
    {
      "epoch": 0.23135255558887793,
      "grad_norm": 4.375911712646484,
      "learning_rate": 9.947266878805896e-05,
      "loss": 2.630513381958008,
      "memory(GiB)": 47.63,
      "step": 5400,
      "token_acc": 0.4548611111111111,
      "train_speed(iter/s)": 1.462808
    },
    {
      "epoch": 0.2315667709181269,
      "grad_norm": 4.754401683807373,
      "learning_rate": 9.947169352082245e-05,
      "loss": 2.536018943786621,
      "memory(GiB)": 47.63,
      "step": 5405,
      "token_acc": 0.4770992366412214,
      "train_speed(iter/s)": 1.462484
    },
    {
      "epoch": 0.23178098624737586,
      "grad_norm": 3.4064626693725586,
      "learning_rate": 9.94707173573598e-05,
      "loss": 2.026996040344238,
      "memory(GiB)": 47.63,
      "step": 5410,
      "token_acc": 0.5372549019607843,
      "train_speed(iter/s)": 1.462017
    },
    {
      "epoch": 0.23199520157662482,
      "grad_norm": 3.4410533905029297,
      "learning_rate": 9.946974029768869e-05,
      "loss": 2.7539953231811523,
      "memory(GiB)": 47.63,
      "step": 5415,
      "token_acc": 0.4701492537313433,
      "train_speed(iter/s)": 1.46208
    },
    {
      "epoch": 0.2322094169058738,
      "grad_norm": 4.9195404052734375,
      "learning_rate": 9.946876234182682e-05,
      "loss": 2.7599765777587892,
      "memory(GiB)": 47.63,
      "step": 5420,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.462127
    },
    {
      "epoch": 0.23242363223512275,
      "grad_norm": 5.885335922241211,
      "learning_rate": 9.946778348979193e-05,
      "loss": 2.507585144042969,
      "memory(GiB)": 47.63,
      "step": 5425,
      "token_acc": 0.5119047619047619,
      "train_speed(iter/s)": 1.462147
    },
    {
      "epoch": 0.2326378475643717,
      "grad_norm": 3.6174395084381104,
      "learning_rate": 9.946680374160174e-05,
      "loss": 2.5303054809570313,
      "memory(GiB)": 47.63,
      "step": 5430,
      "token_acc": 0.4805194805194805,
      "train_speed(iter/s)": 1.462138
    },
    {
      "epoch": 0.23285206289362068,
      "grad_norm": 4.536384105682373,
      "learning_rate": 9.9465823097274e-05,
      "loss": 2.6479154586791993,
      "memory(GiB)": 47.63,
      "step": 5435,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.462136
    },
    {
      "epoch": 0.23306627822286963,
      "grad_norm": 3.7358009815216064,
      "learning_rate": 9.946484155682646e-05,
      "loss": 2.7107013702392577,
      "memory(GiB)": 47.63,
      "step": 5440,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.4625
    },
    {
      "epoch": 0.23328049355211858,
      "grad_norm": 4.361748695373535,
      "learning_rate": 9.946385912027692e-05,
      "loss": 2.82617244720459,
      "memory(GiB)": 47.63,
      "step": 5445,
      "token_acc": 0.42142857142857143,
      "train_speed(iter/s)": 1.462705
    },
    {
      "epoch": 0.23349470888136756,
      "grad_norm": 4.033404350280762,
      "learning_rate": 9.946287578764318e-05,
      "loss": 2.5749502182006836,
      "memory(GiB)": 47.63,
      "step": 5450,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.462794
    },
    {
      "epoch": 0.2337089242106165,
      "grad_norm": 4.20624303817749,
      "learning_rate": 9.946189155894303e-05,
      "loss": 2.6544652938842774,
      "memory(GiB)": 47.63,
      "step": 5455,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.462835
    },
    {
      "epoch": 0.23392313953986546,
      "grad_norm": 3.0296459197998047,
      "learning_rate": 9.946090643419432e-05,
      "loss": 2.552129554748535,
      "memory(GiB)": 47.63,
      "step": 5460,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.462884
    },
    {
      "epoch": 0.23413735486911444,
      "grad_norm": 4.592703342437744,
      "learning_rate": 9.945992041341489e-05,
      "loss": 2.430109214782715,
      "memory(GiB)": 47.63,
      "step": 5465,
      "token_acc": 0.47477744807121663,
      "train_speed(iter/s)": 1.463021
    },
    {
      "epoch": 0.2343515701983634,
      "grad_norm": 3.492677688598633,
      "learning_rate": 9.945893349662261e-05,
      "loss": 2.6183616638183596,
      "memory(GiB)": 47.63,
      "step": 5470,
      "token_acc": 0.46546546546546547,
      "train_speed(iter/s)": 1.462729
    },
    {
      "epoch": 0.23456578552761237,
      "grad_norm": 4.0725626945495605,
      "learning_rate": 9.945794568383534e-05,
      "loss": 2.785839080810547,
      "memory(GiB)": 47.63,
      "step": 5475,
      "token_acc": 0.444,
      "train_speed(iter/s)": 1.463084
    },
    {
      "epoch": 0.23478000085686132,
      "grad_norm": 3.672146797180176,
      "learning_rate": 9.945695697507101e-05,
      "loss": 2.4932071685791017,
      "memory(GiB)": 47.63,
      "step": 5480,
      "token_acc": 0.42586750788643535,
      "train_speed(iter/s)": 1.463114
    },
    {
      "epoch": 0.23499421618611027,
      "grad_norm": 4.550312519073486,
      "learning_rate": 9.945596737034748e-05,
      "loss": 2.583989715576172,
      "memory(GiB)": 47.63,
      "step": 5485,
      "token_acc": 0.4311594202898551,
      "train_speed(iter/s)": 1.463369
    },
    {
      "epoch": 0.23520843151535925,
      "grad_norm": 4.472805500030518,
      "learning_rate": 9.945497686968273e-05,
      "loss": 2.3705770492553713,
      "memory(GiB)": 47.63,
      "step": 5490,
      "token_acc": 0.46987951807228917,
      "train_speed(iter/s)": 1.463439
    },
    {
      "epoch": 0.2354226468446082,
      "grad_norm": 3.246840000152588,
      "learning_rate": 9.945398547309467e-05,
      "loss": 2.5402286529541014,
      "memory(GiB)": 47.63,
      "step": 5495,
      "token_acc": 0.46518105849582175,
      "train_speed(iter/s)": 1.463436
    },
    {
      "epoch": 0.23563686217385715,
      "grad_norm": 3.9132003784179688,
      "learning_rate": 9.945299318060129e-05,
      "loss": 2.4419858932495115,
      "memory(GiB)": 47.63,
      "step": 5500,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.463269
    },
    {
      "epoch": 0.23563686217385715,
      "eval_loss": 2.3884971141815186,
      "eval_runtime": 13.5173,
      "eval_samples_per_second": 7.398,
      "eval_steps_per_second": 7.398,
      "eval_token_acc": 0.48157894736842105,
      "step": 5500
    },
    {
      "epoch": 0.23585107750310613,
      "grad_norm": 3.4197490215301514,
      "learning_rate": 9.945199999222053e-05,
      "loss": 2.9500453948974608,
      "memory(GiB)": 47.63,
      "step": 5505,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.457209
    },
    {
      "epoch": 0.23606529283235508,
      "grad_norm": 5.53607177734375,
      "learning_rate": 9.945100590797041e-05,
      "loss": 2.5514347076416017,
      "memory(GiB)": 47.63,
      "step": 5510,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.457341
    },
    {
      "epoch": 0.23627950816160403,
      "grad_norm": 3.660693883895874,
      "learning_rate": 9.945001092786893e-05,
      "loss": 2.706028366088867,
      "memory(GiB)": 47.63,
      "step": 5515,
      "token_acc": 0.4261168384879725,
      "train_speed(iter/s)": 1.457104
    },
    {
      "epoch": 0.23649372349085301,
      "grad_norm": 3.8297765254974365,
      "learning_rate": 9.944901505193411e-05,
      "loss": 2.6862775802612306,
      "memory(GiB)": 47.63,
      "step": 5520,
      "token_acc": 0.4496124031007752,
      "train_speed(iter/s)": 1.457171
    },
    {
      "epoch": 0.23670793882010197,
      "grad_norm": 3.6341445446014404,
      "learning_rate": 9.9448018280184e-05,
      "loss": 2.7686746597290037,
      "memory(GiB)": 47.63,
      "step": 5525,
      "token_acc": 0.43213296398891965,
      "train_speed(iter/s)": 1.457003
    },
    {
      "epoch": 0.23692215414935092,
      "grad_norm": 3.6948294639587402,
      "learning_rate": 9.944702061263664e-05,
      "loss": 2.6286340713500977,
      "memory(GiB)": 47.63,
      "step": 5530,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.456811
    },
    {
      "epoch": 0.2371363694785999,
      "grad_norm": 5.83001708984375,
      "learning_rate": 9.944602204931013e-05,
      "loss": 2.410238838195801,
      "memory(GiB)": 47.63,
      "step": 5535,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.456941
    },
    {
      "epoch": 0.23735058480784885,
      "grad_norm": 4.257678031921387,
      "learning_rate": 9.944502259022255e-05,
      "loss": 2.2674959182739256,
      "memory(GiB)": 47.63,
      "step": 5540,
      "token_acc": 0.5576923076923077,
      "train_speed(iter/s)": 1.457264
    },
    {
      "epoch": 0.2375648001370978,
      "grad_norm": 4.245832443237305,
      "learning_rate": 9.944402223539199e-05,
      "loss": 2.7157052993774413,
      "memory(GiB)": 47.63,
      "step": 5545,
      "token_acc": 0.4554794520547945,
      "train_speed(iter/s)": 1.457246
    },
    {
      "epoch": 0.23777901546634678,
      "grad_norm": 4.118874549865723,
      "learning_rate": 9.944302098483659e-05,
      "loss": 2.604669952392578,
      "memory(GiB)": 47.63,
      "step": 5550,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.457545
    },
    {
      "epoch": 0.23799323079559573,
      "grad_norm": 4.63669490814209,
      "learning_rate": 9.944201883857449e-05,
      "loss": 2.7955175399780274,
      "memory(GiB)": 47.63,
      "step": 5555,
      "token_acc": 0.41935483870967744,
      "train_speed(iter/s)": 1.457627
    },
    {
      "epoch": 0.2382074461248447,
      "grad_norm": 3.287989616394043,
      "learning_rate": 9.944101579662381e-05,
      "loss": 2.568241500854492,
      "memory(GiB)": 47.63,
      "step": 5560,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.457715
    },
    {
      "epoch": 0.23842166145409366,
      "grad_norm": 3.317380666732788,
      "learning_rate": 9.944001185900278e-05,
      "loss": 2.580239486694336,
      "memory(GiB)": 47.63,
      "step": 5565,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.457854
    },
    {
      "epoch": 0.2386358767833426,
      "grad_norm": 4.391005039215088,
      "learning_rate": 9.943900702572955e-05,
      "loss": 2.5028066635131836,
      "memory(GiB)": 47.63,
      "step": 5570,
      "token_acc": 0.4892703862660944,
      "train_speed(iter/s)": 1.457292
    },
    {
      "epoch": 0.2388500921125916,
      "grad_norm": 3.535794496536255,
      "learning_rate": 9.943800129682233e-05,
      "loss": 2.5869670867919923,
      "memory(GiB)": 47.63,
      "step": 5575,
      "token_acc": 0.4872611464968153,
      "train_speed(iter/s)": 1.457505
    },
    {
      "epoch": 0.23906430744184054,
      "grad_norm": 3.965402364730835,
      "learning_rate": 9.943699467229935e-05,
      "loss": 2.7497854232788086,
      "memory(GiB)": 47.63,
      "step": 5580,
      "token_acc": 0.4396551724137931,
      "train_speed(iter/s)": 1.457407
    },
    {
      "epoch": 0.2392785227710895,
      "grad_norm": 4.6321258544921875,
      "learning_rate": 9.94359871521788e-05,
      "loss": 2.588104248046875,
      "memory(GiB)": 47.63,
      "step": 5585,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.457048
    },
    {
      "epoch": 0.23949273810033847,
      "grad_norm": 3.646897792816162,
      "learning_rate": 9.9434978736479e-05,
      "loss": 2.4612993240356444,
      "memory(GiB)": 47.63,
      "step": 5590,
      "token_acc": 0.4794952681388013,
      "train_speed(iter/s)": 1.457051
    },
    {
      "epoch": 0.23970695342958742,
      "grad_norm": 3.5167148113250732,
      "learning_rate": 9.943396942521818e-05,
      "loss": 2.779859161376953,
      "memory(GiB)": 47.63,
      "step": 5595,
      "token_acc": 0.43137254901960786,
      "train_speed(iter/s)": 1.457356
    },
    {
      "epoch": 0.23992116875883637,
      "grad_norm": 5.377708911895752,
      "learning_rate": 9.943295921841462e-05,
      "loss": 2.449590492248535,
      "memory(GiB)": 47.63,
      "step": 5600,
      "token_acc": 0.5252918287937743,
      "train_speed(iter/s)": 1.457577
    },
    {
      "epoch": 0.24013538408808535,
      "grad_norm": 4.4027204513549805,
      "learning_rate": 9.943194811608665e-05,
      "loss": 2.644864082336426,
      "memory(GiB)": 47.63,
      "step": 5605,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.457633
    },
    {
      "epoch": 0.2403495994173343,
      "grad_norm": 3.8333849906921387,
      "learning_rate": 9.943093611825256e-05,
      "loss": 2.370524597167969,
      "memory(GiB)": 47.63,
      "step": 5610,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.457995
    },
    {
      "epoch": 0.24056381474658325,
      "grad_norm": 2.767317771911621,
      "learning_rate": 9.942992322493068e-05,
      "loss": 2.7053028106689454,
      "memory(GiB)": 47.63,
      "step": 5615,
      "token_acc": 0.46405228758169936,
      "train_speed(iter/s)": 1.458169
    },
    {
      "epoch": 0.24077803007583223,
      "grad_norm": 3.6258201599121094,
      "learning_rate": 9.942890943613939e-05,
      "loss": 2.4264352798461912,
      "memory(GiB)": 47.63,
      "step": 5620,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.458285
    },
    {
      "epoch": 0.24099224540508118,
      "grad_norm": 4.235227584838867,
      "learning_rate": 9.942789475189704e-05,
      "loss": 2.6738864898681642,
      "memory(GiB)": 47.63,
      "step": 5625,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.458511
    },
    {
      "epoch": 0.24120646073433014,
      "grad_norm": 4.216747760772705,
      "learning_rate": 9.9426879172222e-05,
      "loss": 2.611791801452637,
      "memory(GiB)": 47.63,
      "step": 5630,
      "token_acc": 0.45555555555555555,
      "train_speed(iter/s)": 1.458891
    },
    {
      "epoch": 0.24142067606357911,
      "grad_norm": 5.329629421234131,
      "learning_rate": 9.942586269713268e-05,
      "loss": 2.5263603210449217,
      "memory(GiB)": 47.63,
      "step": 5635,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.459087
    },
    {
      "epoch": 0.24163489139282807,
      "grad_norm": 4.288660526275635,
      "learning_rate": 9.942484532664748e-05,
      "loss": 2.8504562377929688,
      "memory(GiB)": 47.63,
      "step": 5640,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.459321
    },
    {
      "epoch": 0.24184910672207705,
      "grad_norm": 3.7665090560913086,
      "learning_rate": 9.942382706078486e-05,
      "loss": 3.1255748748779295,
      "memory(GiB)": 47.63,
      "step": 5645,
      "token_acc": 0.3892857142857143,
      "train_speed(iter/s)": 1.459575
    },
    {
      "epoch": 0.242063322051326,
      "grad_norm": 4.620436191558838,
      "learning_rate": 9.942280789956325e-05,
      "loss": 2.322142791748047,
      "memory(GiB)": 47.63,
      "step": 5650,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.459908
    },
    {
      "epoch": 0.24227753738057495,
      "grad_norm": 4.055863380432129,
      "learning_rate": 9.942178784300109e-05,
      "loss": 2.8052928924560545,
      "memory(GiB)": 47.63,
      "step": 5655,
      "token_acc": 0.45722713864306785,
      "train_speed(iter/s)": 1.460168
    },
    {
      "epoch": 0.24249175270982393,
      "grad_norm": 3.8129770755767822,
      "learning_rate": 9.94207668911169e-05,
      "loss": 2.8385086059570312,
      "memory(GiB)": 47.63,
      "step": 5660,
      "token_acc": 0.4163568773234201,
      "train_speed(iter/s)": 1.460169
    },
    {
      "epoch": 0.24270596803907288,
      "grad_norm": 3.9301958084106445,
      "learning_rate": 9.941974504392916e-05,
      "loss": 2.569260025024414,
      "memory(GiB)": 47.63,
      "step": 5665,
      "token_acc": 0.4421364985163205,
      "train_speed(iter/s)": 1.459913
    },
    {
      "epoch": 0.24292018336832183,
      "grad_norm": 4.454365253448486,
      "learning_rate": 9.941872230145637e-05,
      "loss": 2.8422536849975586,
      "memory(GiB)": 47.63,
      "step": 5670,
      "token_acc": 0.4421768707482993,
      "train_speed(iter/s)": 1.460053
    },
    {
      "epoch": 0.2431343986975708,
      "grad_norm": 3.1430835723876953,
      "learning_rate": 9.941769866371708e-05,
      "loss": 2.2880783081054688,
      "memory(GiB)": 47.63,
      "step": 5675,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.460004
    },
    {
      "epoch": 0.24334861402681976,
      "grad_norm": 4.095118045806885,
      "learning_rate": 9.94166741307298e-05,
      "loss": 2.471821975708008,
      "memory(GiB)": 47.63,
      "step": 5680,
      "token_acc": 0.4440789473684211,
      "train_speed(iter/s)": 1.460028
    },
    {
      "epoch": 0.2435628293560687,
      "grad_norm": 3.6256887912750244,
      "learning_rate": 9.941564870251312e-05,
      "loss": 2.397996520996094,
      "memory(GiB)": 47.63,
      "step": 5685,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.459812
    },
    {
      "epoch": 0.2437770446853177,
      "grad_norm": 3.829911708831787,
      "learning_rate": 9.941462237908561e-05,
      "loss": 2.7572067260742186,
      "memory(GiB)": 47.63,
      "step": 5690,
      "token_acc": 0.4427710843373494,
      "train_speed(iter/s)": 1.460168
    },
    {
      "epoch": 0.24399126001456664,
      "grad_norm": 5.137447834014893,
      "learning_rate": 9.941359516046586e-05,
      "loss": 2.923348236083984,
      "memory(GiB)": 47.63,
      "step": 5695,
      "token_acc": 0.437984496124031,
      "train_speed(iter/s)": 1.460379
    },
    {
      "epoch": 0.2442054753438156,
      "grad_norm": 4.275759220123291,
      "learning_rate": 9.941256704667249e-05,
      "loss": 2.330388069152832,
      "memory(GiB)": 47.63,
      "step": 5700,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.460243
    },
    {
      "epoch": 0.24441969067306457,
      "grad_norm": 5.067497253417969,
      "learning_rate": 9.941153803772412e-05,
      "loss": 2.49843807220459,
      "memory(GiB)": 47.63,
      "step": 5705,
      "token_acc": 0.5083612040133779,
      "train_speed(iter/s)": 1.460151
    },
    {
      "epoch": 0.24463390600231352,
      "grad_norm": 4.475621223449707,
      "learning_rate": 9.941050813363937e-05,
      "loss": 2.559776496887207,
      "memory(GiB)": 47.63,
      "step": 5710,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.460197
    },
    {
      "epoch": 0.24484812133156247,
      "grad_norm": 5.182840824127197,
      "learning_rate": 9.94094773344369e-05,
      "loss": 2.7183162689208986,
      "memory(GiB)": 47.63,
      "step": 5715,
      "token_acc": 0.4575757575757576,
      "train_speed(iter/s)": 1.460034
    },
    {
      "epoch": 0.24506233666081145,
      "grad_norm": 6.315879821777344,
      "learning_rate": 9.940844564013542e-05,
      "loss": 2.696755790710449,
      "memory(GiB)": 47.63,
      "step": 5720,
      "token_acc": 0.4452054794520548,
      "train_speed(iter/s)": 1.460214
    },
    {
      "epoch": 0.2452765519900604,
      "grad_norm": 6.686441421508789,
      "learning_rate": 9.94074130507536e-05,
      "loss": 2.8254520416259767,
      "memory(GiB)": 47.63,
      "step": 5725,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.460206
    },
    {
      "epoch": 0.24549076731930938,
      "grad_norm": 4.55892276763916,
      "learning_rate": 9.940637956631013e-05,
      "loss": 2.3481155395507813,
      "memory(GiB)": 47.63,
      "step": 5730,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.460336
    },
    {
      "epoch": 0.24570498264855833,
      "grad_norm": 4.8279218673706055,
      "learning_rate": 9.940534518682376e-05,
      "loss": 2.682110595703125,
      "memory(GiB)": 47.63,
      "step": 5735,
      "token_acc": 0.45936395759717313,
      "train_speed(iter/s)": 1.460624
    },
    {
      "epoch": 0.24591919797780729,
      "grad_norm": 4.679349422454834,
      "learning_rate": 9.940430991231322e-05,
      "loss": 3.0118648529052736,
      "memory(GiB)": 47.63,
      "step": 5740,
      "token_acc": 0.4158730158730159,
      "train_speed(iter/s)": 1.460892
    },
    {
      "epoch": 0.24613341330705626,
      "grad_norm": 3.5831639766693115,
      "learning_rate": 9.940327374279725e-05,
      "loss": 2.7695430755615233,
      "memory(GiB)": 47.63,
      "step": 5745,
      "token_acc": 0.4199288256227758,
      "train_speed(iter/s)": 1.460889
    },
    {
      "epoch": 0.24634762863630522,
      "grad_norm": 4.248871803283691,
      "learning_rate": 9.940223667829465e-05,
      "loss": 2.6276681900024412,
      "memory(GiB)": 47.63,
      "step": 5750,
      "token_acc": 0.422680412371134,
      "train_speed(iter/s)": 1.460391
    },
    {
      "epoch": 0.24656184396555417,
      "grad_norm": 4.029080390930176,
      "learning_rate": 9.940119871882416e-05,
      "loss": 2.8304515838623048,
      "memory(GiB)": 47.63,
      "step": 5755,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.460395
    },
    {
      "epoch": 0.24677605929480315,
      "grad_norm": 5.108576774597168,
      "learning_rate": 9.940015986440464e-05,
      "loss": 2.7089282989501955,
      "memory(GiB)": 47.63,
      "step": 5760,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.460593
    },
    {
      "epoch": 0.2469902746240521,
      "grad_norm": 5.069339275360107,
      "learning_rate": 9.939912011505487e-05,
      "loss": 2.565351104736328,
      "memory(GiB)": 47.63,
      "step": 5765,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.460477
    },
    {
      "epoch": 0.24720448995330105,
      "grad_norm": 4.0777130126953125,
      "learning_rate": 9.939807947079369e-05,
      "loss": 2.4939748764038088,
      "memory(GiB)": 47.63,
      "step": 5770,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.460238
    },
    {
      "epoch": 0.24741870528255003,
      "grad_norm": 4.485532760620117,
      "learning_rate": 9.939703793163998e-05,
      "loss": 2.81497859954834,
      "memory(GiB)": 47.63,
      "step": 5775,
      "token_acc": 0.4171974522292994,
      "train_speed(iter/s)": 1.460413
    },
    {
      "epoch": 0.24763292061179898,
      "grad_norm": 3.2145578861236572,
      "learning_rate": 9.939599549761259e-05,
      "loss": 2.407269096374512,
      "memory(GiB)": 47.63,
      "step": 5780,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.460456
    },
    {
      "epoch": 0.24784713594104793,
      "grad_norm": 4.005586624145508,
      "learning_rate": 9.939495216873038e-05,
      "loss": 2.7351686477661135,
      "memory(GiB)": 47.63,
      "step": 5785,
      "token_acc": 0.4340277777777778,
      "train_speed(iter/s)": 1.460402
    },
    {
      "epoch": 0.2480613512702969,
      "grad_norm": 5.348986625671387,
      "learning_rate": 9.93939079450123e-05,
      "loss": 2.5608905792236327,
      "memory(GiB)": 47.63,
      "step": 5790,
      "token_acc": 0.4608150470219436,
      "train_speed(iter/s)": 1.460249
    },
    {
      "epoch": 0.24827556659954586,
      "grad_norm": 4.527293682098389,
      "learning_rate": 9.939286282647723e-05,
      "loss": 2.7762317657470703,
      "memory(GiB)": 47.63,
      "step": 5795,
      "token_acc": 0.4568345323741007,
      "train_speed(iter/s)": 1.460275
    },
    {
      "epoch": 0.2484897819287948,
      "grad_norm": 3.6915411949157715,
      "learning_rate": 9.939181681314411e-05,
      "loss": 2.6612146377563475,
      "memory(GiB)": 47.63,
      "step": 5800,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.460155
    },
    {
      "epoch": 0.2487039972580438,
      "grad_norm": 3.42191481590271,
      "learning_rate": 9.939076990503191e-05,
      "loss": 2.2639301300048826,
      "memory(GiB)": 47.63,
      "step": 5805,
      "token_acc": 0.5261194029850746,
      "train_speed(iter/s)": 1.460182
    },
    {
      "epoch": 0.24891821258729274,
      "grad_norm": 3.4134628772735596,
      "learning_rate": 9.938972210215958e-05,
      "loss": 2.7226140975952147,
      "memory(GiB)": 47.63,
      "step": 5810,
      "token_acc": 0.42567567567567566,
      "train_speed(iter/s)": 1.460077
    },
    {
      "epoch": 0.24913242791654172,
      "grad_norm": 3.800591230392456,
      "learning_rate": 9.938867340454609e-05,
      "loss": 2.8251644134521485,
      "memory(GiB)": 47.63,
      "step": 5815,
      "token_acc": 0.45874587458745875,
      "train_speed(iter/s)": 1.460465
    },
    {
      "epoch": 0.24934664324579067,
      "grad_norm": 4.803477764129639,
      "learning_rate": 9.938762381221047e-05,
      "loss": 2.6071054458618166,
      "memory(GiB)": 47.63,
      "step": 5820,
      "token_acc": 0.43490304709141275,
      "train_speed(iter/s)": 1.460566
    },
    {
      "epoch": 0.24956085857503962,
      "grad_norm": 4.181412696838379,
      "learning_rate": 9.93865733251717e-05,
      "loss": 2.6922088623046876,
      "memory(GiB)": 47.63,
      "step": 5825,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.460679
    },
    {
      "epoch": 0.2497750739042886,
      "grad_norm": 3.804574966430664,
      "learning_rate": 9.938552194344883e-05,
      "loss": 2.6233333587646483,
      "memory(GiB)": 47.63,
      "step": 5830,
      "token_acc": 0.41818181818181815,
      "train_speed(iter/s)": 1.460525
    },
    {
      "epoch": 0.24998928923353755,
      "grad_norm": 4.469377517700195,
      "learning_rate": 9.93844696670609e-05,
      "loss": 2.653834915161133,
      "memory(GiB)": 47.63,
      "step": 5835,
      "token_acc": 0.43582089552238806,
      "train_speed(iter/s)": 1.460463
    },
    {
      "epoch": 0.2502035045627865,
      "grad_norm": 3.8620765209198,
      "learning_rate": 9.938341649602698e-05,
      "loss": 2.595180130004883,
      "memory(GiB)": 47.63,
      "step": 5840,
      "token_acc": 0.462406015037594,
      "train_speed(iter/s)": 1.460496
    },
    {
      "epoch": 0.2504177198920355,
      "grad_norm": 5.089691162109375,
      "learning_rate": 9.938236243036613e-05,
      "loss": 2.6849374771118164,
      "memory(GiB)": 47.63,
      "step": 5845,
      "token_acc": 0.43727598566308246,
      "train_speed(iter/s)": 1.46056
    },
    {
      "epoch": 0.2506319352212844,
      "grad_norm": 4.5347418785095215,
      "learning_rate": 9.938130747009748e-05,
      "loss": 2.4504329681396486,
      "memory(GiB)": 47.63,
      "step": 5850,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.460624
    },
    {
      "epoch": 0.2508461505505334,
      "grad_norm": 4.512951850891113,
      "learning_rate": 9.938025161524012e-05,
      "loss": 2.6327817916870115,
      "memory(GiB)": 47.63,
      "step": 5855,
      "token_acc": 0.45724907063197023,
      "train_speed(iter/s)": 1.460624
    },
    {
      "epoch": 0.25106036587978237,
      "grad_norm": 3.6371476650238037,
      "learning_rate": 9.937919486581317e-05,
      "loss": 2.5981365203857423,
      "memory(GiB)": 47.63,
      "step": 5860,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.460525
    },
    {
      "epoch": 0.25127458120903134,
      "grad_norm": 3.2102556228637695,
      "learning_rate": 9.937813722183579e-05,
      "loss": 2.607102966308594,
      "memory(GiB)": 47.63,
      "step": 5865,
      "token_acc": 0.4658385093167702,
      "train_speed(iter/s)": 1.460289
    },
    {
      "epoch": 0.25148879653828027,
      "grad_norm": 3.3604695796966553,
      "learning_rate": 9.937707868332713e-05,
      "loss": 2.401707077026367,
      "memory(GiB)": 47.63,
      "step": 5870,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.460493
    },
    {
      "epoch": 0.25170301186752925,
      "grad_norm": 6.565609931945801,
      "learning_rate": 9.937601925030638e-05,
      "loss": 2.6921478271484376,
      "memory(GiB)": 47.63,
      "step": 5875,
      "token_acc": 0.4295774647887324,
      "train_speed(iter/s)": 1.460509
    },
    {
      "epoch": 0.2519172271967782,
      "grad_norm": 5.022071838378906,
      "learning_rate": 9.937495892279272e-05,
      "loss": 2.6079330444335938,
      "memory(GiB)": 47.63,
      "step": 5880,
      "token_acc": 0.4798387096774194,
      "train_speed(iter/s)": 1.460815
    },
    {
      "epoch": 0.25213144252602715,
      "grad_norm": 3.526156425476074,
      "learning_rate": 9.937389770080535e-05,
      "loss": 2.1729354858398438,
      "memory(GiB)": 47.63,
      "step": 5885,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.460567
    },
    {
      "epoch": 0.25234565785527613,
      "grad_norm": 5.314532279968262,
      "learning_rate": 9.937283558436352e-05,
      "loss": 2.5845405578613283,
      "memory(GiB)": 47.63,
      "step": 5890,
      "token_acc": 0.4676258992805755,
      "train_speed(iter/s)": 1.460932
    },
    {
      "epoch": 0.2525598731845251,
      "grad_norm": 3.1484577655792236,
      "learning_rate": 9.937177257348645e-05,
      "loss": 2.2195556640625,
      "memory(GiB)": 47.63,
      "step": 5895,
      "token_acc": 0.517799352750809,
      "train_speed(iter/s)": 1.460927
    },
    {
      "epoch": 0.25277408851377403,
      "grad_norm": 3.156454563140869,
      "learning_rate": 9.93707086681934e-05,
      "loss": 2.505203437805176,
      "memory(GiB)": 47.63,
      "step": 5900,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.460815
    },
    {
      "epoch": 0.252988303843023,
      "grad_norm": 3.0690927505493164,
      "learning_rate": 9.936964386850366e-05,
      "loss": 2.4637630462646483,
      "memory(GiB)": 47.63,
      "step": 5905,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.460764
    },
    {
      "epoch": 0.253202519172272,
      "grad_norm": 3.5255067348480225,
      "learning_rate": 9.93685781744365e-05,
      "loss": 2.435877227783203,
      "memory(GiB)": 47.63,
      "step": 5910,
      "token_acc": 0.458041958041958,
      "train_speed(iter/s)": 1.460675
    },
    {
      "epoch": 0.2534167345015209,
      "grad_norm": 3.420804023742676,
      "learning_rate": 9.936751158601124e-05,
      "loss": 2.7171825408935546,
      "memory(GiB)": 47.63,
      "step": 5915,
      "token_acc": 0.4075342465753425,
      "train_speed(iter/s)": 1.460974
    },
    {
      "epoch": 0.2536309498307699,
      "grad_norm": 7.083895206451416,
      "learning_rate": 9.93664441032472e-05,
      "loss": 2.4212947845458985,
      "memory(GiB)": 47.63,
      "step": 5920,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.461293
    },
    {
      "epoch": 0.25384516516001887,
      "grad_norm": 4.429103374481201,
      "learning_rate": 9.936537572616372e-05,
      "loss": 2.6738073348999025,
      "memory(GiB)": 47.63,
      "step": 5925,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.46146
    },
    {
      "epoch": 0.2540593804892678,
      "grad_norm": 4.744236946105957,
      "learning_rate": 9.936430645478014e-05,
      "loss": 2.6790178298950194,
      "memory(GiB)": 47.63,
      "step": 5930,
      "token_acc": 0.4440677966101695,
      "train_speed(iter/s)": 1.461148
    },
    {
      "epoch": 0.2542735958185168,
      "grad_norm": 4.812023162841797,
      "learning_rate": 9.936323628911584e-05,
      "loss": 2.839132308959961,
      "memory(GiB)": 47.63,
      "step": 5935,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.46138
    },
    {
      "epoch": 0.25448781114776575,
      "grad_norm": 5.341465950012207,
      "learning_rate": 9.936216522919021e-05,
      "loss": 2.2750164031982423,
      "memory(GiB)": 47.63,
      "step": 5940,
      "token_acc": 0.5450980392156862,
      "train_speed(iter/s)": 1.461362
    },
    {
      "epoch": 0.2547020264770147,
      "grad_norm": 3.449023723602295,
      "learning_rate": 9.936109327502266e-05,
      "loss": 2.6081863403320313,
      "memory(GiB)": 47.63,
      "step": 5945,
      "token_acc": 0.48255813953488375,
      "train_speed(iter/s)": 1.461467
    },
    {
      "epoch": 0.25491624180626365,
      "grad_norm": 5.677049160003662,
      "learning_rate": 9.936002042663258e-05,
      "loss": 2.8544898986816407,
      "memory(GiB)": 47.63,
      "step": 5950,
      "token_acc": 0.41843971631205673,
      "train_speed(iter/s)": 1.461554
    },
    {
      "epoch": 0.25513045713551263,
      "grad_norm": 4.046048641204834,
      "learning_rate": 9.935894668403945e-05,
      "loss": 2.7638923645019533,
      "memory(GiB)": 47.63,
      "step": 5955,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.461758
    },
    {
      "epoch": 0.25534467246476156,
      "grad_norm": 5.44657564163208,
      "learning_rate": 9.935787204726268e-05,
      "loss": 2.677348518371582,
      "memory(GiB)": 47.63,
      "step": 5960,
      "token_acc": 0.4426751592356688,
      "train_speed(iter/s)": 1.461933
    },
    {
      "epoch": 0.25555888779401054,
      "grad_norm": 3.4075331687927246,
      "learning_rate": 9.935679651632177e-05,
      "loss": 2.6634239196777343,
      "memory(GiB)": 47.63,
      "step": 5965,
      "token_acc": 0.4423676012461059,
      "train_speed(iter/s)": 1.46191
    },
    {
      "epoch": 0.2557731031232595,
      "grad_norm": 4.674412250518799,
      "learning_rate": 9.935572009123619e-05,
      "loss": 2.6897483825683595,
      "memory(GiB)": 47.63,
      "step": 5970,
      "token_acc": 0.4180064308681672,
      "train_speed(iter/s)": 1.461873
    },
    {
      "epoch": 0.25598731845250844,
      "grad_norm": 3.6903035640716553,
      "learning_rate": 9.935464277202544e-05,
      "loss": 2.708321189880371,
      "memory(GiB)": 47.63,
      "step": 5975,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.46182
    },
    {
      "epoch": 0.2562015337817574,
      "grad_norm": 3.725214958190918,
      "learning_rate": 9.935356455870904e-05,
      "loss": 2.417351722717285,
      "memory(GiB)": 47.63,
      "step": 5980,
      "token_acc": 0.4789272030651341,
      "train_speed(iter/s)": 1.461616
    },
    {
      "epoch": 0.2564157491110064,
      "grad_norm": 3.872201919555664,
      "learning_rate": 9.93524854513065e-05,
      "loss": 2.609356689453125,
      "memory(GiB)": 47.63,
      "step": 5985,
      "token_acc": 0.4548736462093863,
      "train_speed(iter/s)": 1.461727
    },
    {
      "epoch": 0.2566299644402553,
      "grad_norm": 3.2694787979125977,
      "learning_rate": 9.93514054498374e-05,
      "loss": 2.5014877319335938,
      "memory(GiB)": 47.63,
      "step": 5990,
      "token_acc": 0.49615384615384617,
      "train_speed(iter/s)": 1.461838
    },
    {
      "epoch": 0.2568441797695043,
      "grad_norm": 5.387357234954834,
      "learning_rate": 9.93503245543213e-05,
      "loss": 2.867146110534668,
      "memory(GiB)": 47.63,
      "step": 5995,
      "token_acc": 0.41754385964912283,
      "train_speed(iter/s)": 1.461973
    },
    {
      "epoch": 0.2570583950987533,
      "grad_norm": 4.068348407745361,
      "learning_rate": 9.934924276477779e-05,
      "loss": 2.9688154220581056,
      "memory(GiB)": 47.63,
      "step": 6000,
      "token_acc": 0.43521594684385384,
      "train_speed(iter/s)": 1.461975
    },
    {
      "epoch": 0.2570583950987533,
      "eval_loss": 2.512045383453369,
      "eval_runtime": 14.2773,
      "eval_samples_per_second": 7.004,
      "eval_steps_per_second": 7.004,
      "eval_token_acc": 0.4553140096618358,
      "step": 6000
    },
    {
      "epoch": 0.2572726104280022,
      "grad_norm": 3.5032174587249756,
      "learning_rate": 9.934816008122643e-05,
      "loss": 2.8273218154907225,
      "memory(GiB)": 47.63,
      "step": 6005,
      "token_acc": 0.4603463992707384,
      "train_speed(iter/s)": 1.456857
    },
    {
      "epoch": 0.2574868257572512,
      "grad_norm": 4.137598991394043,
      "learning_rate": 9.934707650368686e-05,
      "loss": 2.599644660949707,
      "memory(GiB)": 47.63,
      "step": 6010,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.456824
    },
    {
      "epoch": 0.25770104108650016,
      "grad_norm": 4.335493087768555,
      "learning_rate": 9.934599203217874e-05,
      "loss": 2.692888069152832,
      "memory(GiB)": 47.63,
      "step": 6015,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.456937
    },
    {
      "epoch": 0.2579152564157491,
      "grad_norm": 4.111094951629639,
      "learning_rate": 9.934490666672164e-05,
      "loss": 2.4740135192871096,
      "memory(GiB)": 47.63,
      "step": 6020,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.457194
    },
    {
      "epoch": 0.25812947174499806,
      "grad_norm": 5.307859420776367,
      "learning_rate": 9.93438204073353e-05,
      "loss": 2.5906599044799803,
      "memory(GiB)": 47.63,
      "step": 6025,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.456628
    },
    {
      "epoch": 0.25834368707424704,
      "grad_norm": 4.8536272048950195,
      "learning_rate": 9.934273325403935e-05,
      "loss": 2.8745859146118162,
      "memory(GiB)": 47.63,
      "step": 6030,
      "token_acc": 0.44656488549618323,
      "train_speed(iter/s)": 1.456801
    },
    {
      "epoch": 0.258557902403496,
      "grad_norm": 4.217504501342773,
      "learning_rate": 9.934164520685349e-05,
      "loss": 2.4962957382202147,
      "memory(GiB)": 47.63,
      "step": 6035,
      "token_acc": 0.4942084942084942,
      "train_speed(iter/s)": 1.45659
    },
    {
      "epoch": 0.25877211773274494,
      "grad_norm": 3.7887179851531982,
      "learning_rate": 9.934055626579746e-05,
      "loss": 2.66213436126709,
      "memory(GiB)": 47.63,
      "step": 6040,
      "token_acc": 0.5060728744939271,
      "train_speed(iter/s)": 1.456629
    },
    {
      "epoch": 0.2589863330619939,
      "grad_norm": 3.306767225265503,
      "learning_rate": 9.933946643089096e-05,
      "loss": 2.7682987213134767,
      "memory(GiB)": 47.63,
      "step": 6045,
      "token_acc": 0.42165242165242167,
      "train_speed(iter/s)": 1.456714
    },
    {
      "epoch": 0.2592005483912429,
      "grad_norm": 4.921095371246338,
      "learning_rate": 9.933837570215374e-05,
      "loss": 2.4791046142578126,
      "memory(GiB)": 47.63,
      "step": 6050,
      "token_acc": 0.4808510638297872,
      "train_speed(iter/s)": 1.456622
    },
    {
      "epoch": 0.2594147637204918,
      "grad_norm": 3.72432279586792,
      "learning_rate": 9.933728407960556e-05,
      "loss": 2.349840545654297,
      "memory(GiB)": 47.63,
      "step": 6055,
      "token_acc": 0.5272727272727272,
      "train_speed(iter/s)": 1.456687
    },
    {
      "epoch": 0.2596289790497408,
      "grad_norm": 3.4007792472839355,
      "learning_rate": 9.933619156326621e-05,
      "loss": 2.6064830780029298,
      "memory(GiB)": 47.63,
      "step": 6060,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.456835
    },
    {
      "epoch": 0.2598431943789898,
      "grad_norm": 4.636086940765381,
      "learning_rate": 9.933509815315545e-05,
      "loss": 2.4625308990478514,
      "memory(GiB)": 47.63,
      "step": 6065,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.456911
    },
    {
      "epoch": 0.2600574097082387,
      "grad_norm": 3.8582499027252197,
      "learning_rate": 9.933400384929313e-05,
      "loss": 2.564784812927246,
      "memory(GiB)": 47.63,
      "step": 6070,
      "token_acc": 0.4779874213836478,
      "train_speed(iter/s)": 1.45693
    },
    {
      "epoch": 0.2602716250374877,
      "grad_norm": 2.999443531036377,
      "learning_rate": 9.933290865169903e-05,
      "loss": 2.6505287170410154,
      "memory(GiB)": 47.63,
      "step": 6075,
      "token_acc": 0.46578947368421053,
      "train_speed(iter/s)": 1.457018
    },
    {
      "epoch": 0.26048584036673667,
      "grad_norm": 3.236865758895874,
      "learning_rate": 9.933181256039301e-05,
      "loss": 2.6455543518066404,
      "memory(GiB)": 47.63,
      "step": 6080,
      "token_acc": 0.4503105590062112,
      "train_speed(iter/s)": 1.457262
    },
    {
      "epoch": 0.2607000556959856,
      "grad_norm": 2.9523332118988037,
      "learning_rate": 9.933071557539494e-05,
      "loss": 2.6446355819702148,
      "memory(GiB)": 47.63,
      "step": 6085,
      "token_acc": 0.4862068965517241,
      "train_speed(iter/s)": 1.457529
    },
    {
      "epoch": 0.26091427102523457,
      "grad_norm": 3.2571287155151367,
      "learning_rate": 9.932961769672469e-05,
      "loss": 2.8434967041015624,
      "memory(GiB)": 47.63,
      "step": 6090,
      "token_acc": 0.4384858044164038,
      "train_speed(iter/s)": 1.457772
    },
    {
      "epoch": 0.26112848635448355,
      "grad_norm": 4.042792797088623,
      "learning_rate": 9.932851892440211e-05,
      "loss": 2.4596343994140626,
      "memory(GiB)": 47.63,
      "step": 6095,
      "token_acc": 0.45774647887323944,
      "train_speed(iter/s)": 1.458079
    },
    {
      "epoch": 0.26134270168373247,
      "grad_norm": 5.742770671844482,
      "learning_rate": 9.932741925844717e-05,
      "loss": 2.4574043273925783,
      "memory(GiB)": 47.63,
      "step": 6100,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.458153
    },
    {
      "epoch": 0.26155691701298145,
      "grad_norm": 3.689504623413086,
      "learning_rate": 9.932631869887974e-05,
      "loss": 2.8216827392578123,
      "memory(GiB)": 47.63,
      "step": 6105,
      "token_acc": 0.4440789473684211,
      "train_speed(iter/s)": 1.458459
    },
    {
      "epoch": 0.26177113234223043,
      "grad_norm": 3.9181625843048096,
      "learning_rate": 9.932521724571977e-05,
      "loss": 2.758797454833984,
      "memory(GiB)": 47.63,
      "step": 6110,
      "token_acc": 0.40672782874617736,
      "train_speed(iter/s)": 1.458645
    },
    {
      "epoch": 0.26198534767147935,
      "grad_norm": 3.422567367553711,
      "learning_rate": 9.932411489898723e-05,
      "loss": 2.3415912628173827,
      "memory(GiB)": 47.63,
      "step": 6115,
      "token_acc": 0.5125448028673835,
      "train_speed(iter/s)": 1.458501
    },
    {
      "epoch": 0.26219956300072833,
      "grad_norm": 3.795133590698242,
      "learning_rate": 9.932301165870206e-05,
      "loss": 2.775946617126465,
      "memory(GiB)": 47.63,
      "step": 6120,
      "token_acc": 0.4395973154362416,
      "train_speed(iter/s)": 1.458252
    },
    {
      "epoch": 0.2624137783299773,
      "grad_norm": 3.046976327896118,
      "learning_rate": 9.932190752488428e-05,
      "loss": 2.5555233001708983,
      "memory(GiB)": 47.63,
      "step": 6125,
      "token_acc": 0.4398826979472141,
      "train_speed(iter/s)": 1.457662
    },
    {
      "epoch": 0.26262799365922623,
      "grad_norm": 3.9895241260528564,
      "learning_rate": 9.932080249755389e-05,
      "loss": 2.3629072189331053,
      "memory(GiB)": 47.63,
      "step": 6130,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.457824
    },
    {
      "epoch": 0.2628422089884752,
      "grad_norm": 5.900223731994629,
      "learning_rate": 9.931969657673088e-05,
      "loss": 2.648196220397949,
      "memory(GiB)": 47.63,
      "step": 6135,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.458188
    },
    {
      "epoch": 0.2630564243177242,
      "grad_norm": 4.17408561706543,
      "learning_rate": 9.931858976243531e-05,
      "loss": 2.629598617553711,
      "memory(GiB)": 47.63,
      "step": 6140,
      "token_acc": 0.4323308270676692,
      "train_speed(iter/s)": 1.458345
    },
    {
      "epoch": 0.2632706396469731,
      "grad_norm": 5.355582237243652,
      "learning_rate": 9.931748205468721e-05,
      "loss": 2.7148637771606445,
      "memory(GiB)": 47.63,
      "step": 6145,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.458381
    },
    {
      "epoch": 0.2634848549762221,
      "grad_norm": 3.701125144958496,
      "learning_rate": 9.931637345350667e-05,
      "loss": 2.8413522720336912,
      "memory(GiB)": 47.63,
      "step": 6150,
      "token_acc": 0.42696629213483145,
      "train_speed(iter/s)": 1.458714
    },
    {
      "epoch": 0.2636990703054711,
      "grad_norm": 5.480071544647217,
      "learning_rate": 9.931526395891375e-05,
      "loss": 2.6477405548095705,
      "memory(GiB)": 47.63,
      "step": 6155,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.458807
    },
    {
      "epoch": 0.26391328563472,
      "grad_norm": 4.121298789978027,
      "learning_rate": 9.931415357092858e-05,
      "loss": 2.474870300292969,
      "memory(GiB)": 47.63,
      "step": 6160,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.458935
    },
    {
      "epoch": 0.264127500963969,
      "grad_norm": 4.323131084442139,
      "learning_rate": 9.931304228957123e-05,
      "loss": 2.642374038696289,
      "memory(GiB)": 47.63,
      "step": 6165,
      "token_acc": 0.4134275618374558,
      "train_speed(iter/s)": 1.459283
    },
    {
      "epoch": 0.26434171629321795,
      "grad_norm": 3.5837652683258057,
      "learning_rate": 9.931193011486188e-05,
      "loss": 2.4898326873779295,
      "memory(GiB)": 47.63,
      "step": 6170,
      "token_acc": 0.4978540772532189,
      "train_speed(iter/s)": 1.459232
    },
    {
      "epoch": 0.2645559316224669,
      "grad_norm": 3.816152811050415,
      "learning_rate": 9.931081704682066e-05,
      "loss": 2.5765811920166017,
      "memory(GiB)": 47.63,
      "step": 6175,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.459484
    },
    {
      "epoch": 0.26477014695171586,
      "grad_norm": 4.142642974853516,
      "learning_rate": 9.930970308546772e-05,
      "loss": 2.852547836303711,
      "memory(GiB)": 47.63,
      "step": 6180,
      "token_acc": 0.4564459930313589,
      "train_speed(iter/s)": 1.459496
    },
    {
      "epoch": 0.26498436228096484,
      "grad_norm": 4.207241058349609,
      "learning_rate": 9.930858823082327e-05,
      "loss": 2.589390754699707,
      "memory(GiB)": 47.63,
      "step": 6185,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.459538
    },
    {
      "epoch": 0.26519857761021376,
      "grad_norm": 3.89192795753479,
      "learning_rate": 9.930747248290747e-05,
      "loss": 2.2295835494995115,
      "memory(GiB)": 47.63,
      "step": 6190,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.459644
    },
    {
      "epoch": 0.26541279293946274,
      "grad_norm": 5.49457311630249,
      "learning_rate": 9.930635584174056e-05,
      "loss": 2.608285903930664,
      "memory(GiB)": 47.63,
      "step": 6195,
      "token_acc": 0.44333333333333336,
      "train_speed(iter/s)": 1.459845
    },
    {
      "epoch": 0.2656270082687117,
      "grad_norm": 3.7768564224243164,
      "learning_rate": 9.930523830734276e-05,
      "loss": 2.72476863861084,
      "memory(GiB)": 47.63,
      "step": 6200,
      "token_acc": 0.4560810810810811,
      "train_speed(iter/s)": 1.459594
    },
    {
      "epoch": 0.2658412235979607,
      "grad_norm": 7.339259147644043,
      "learning_rate": 9.930411987973431e-05,
      "loss": 2.422555923461914,
      "memory(GiB)": 47.63,
      "step": 6205,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.459532
    },
    {
      "epoch": 0.2660554389272096,
      "grad_norm": 3.4869730472564697,
      "learning_rate": 9.930300055893549e-05,
      "loss": 2.5603126525878905,
      "memory(GiB)": 47.63,
      "step": 6210,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.459648
    },
    {
      "epoch": 0.2662696542564586,
      "grad_norm": 5.080729007720947,
      "learning_rate": 9.930188034496655e-05,
      "loss": 2.684031105041504,
      "memory(GiB)": 47.63,
      "step": 6215,
      "token_acc": 0.478125,
      "train_speed(iter/s)": 1.459563
    },
    {
      "epoch": 0.2664838695857076,
      "grad_norm": 6.193282604217529,
      "learning_rate": 9.93007592378478e-05,
      "loss": 2.7133209228515627,
      "memory(GiB)": 47.63,
      "step": 6220,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.459394
    },
    {
      "epoch": 0.2666980849149565,
      "grad_norm": 7.4428205490112305,
      "learning_rate": 9.929963723759956e-05,
      "loss": 2.7466331481933595,
      "memory(GiB)": 47.63,
      "step": 6225,
      "token_acc": 0.450354609929078,
      "train_speed(iter/s)": 1.459673
    },
    {
      "epoch": 0.2669123002442055,
      "grad_norm": 3.6143174171447754,
      "learning_rate": 9.929851434424216e-05,
      "loss": 2.7537635803222655,
      "memory(GiB)": 47.63,
      "step": 6230,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.459388
    },
    {
      "epoch": 0.26712651557345446,
      "grad_norm": 3.4207046031951904,
      "learning_rate": 9.92973905577959e-05,
      "loss": 2.6948078155517576,
      "memory(GiB)": 47.63,
      "step": 6235,
      "token_acc": 0.4440894568690096,
      "train_speed(iter/s)": 1.459492
    },
    {
      "epoch": 0.2673407309027034,
      "grad_norm": 3.170596122741699,
      "learning_rate": 9.929626587828118e-05,
      "loss": 2.574795150756836,
      "memory(GiB)": 47.63,
      "step": 6240,
      "token_acc": 0.4430379746835443,
      "train_speed(iter/s)": 1.459642
    },
    {
      "epoch": 0.26755494623195236,
      "grad_norm": 7.684406280517578,
      "learning_rate": 9.929514030571834e-05,
      "loss": 2.733842468261719,
      "memory(GiB)": 47.63,
      "step": 6245,
      "token_acc": 0.43209876543209874,
      "train_speed(iter/s)": 1.459705
    },
    {
      "epoch": 0.26776916156120134,
      "grad_norm": 4.399084568023682,
      "learning_rate": 9.92940138401278e-05,
      "loss": 2.87264347076416,
      "memory(GiB)": 47.63,
      "step": 6250,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.459662
    },
    {
      "epoch": 0.26798337689045026,
      "grad_norm": 3.7682693004608154,
      "learning_rate": 9.929288648152997e-05,
      "loss": 3.175109100341797,
      "memory(GiB)": 47.63,
      "step": 6255,
      "token_acc": 0.4161676646706587,
      "train_speed(iter/s)": 1.459561
    },
    {
      "epoch": 0.26819759221969924,
      "grad_norm": 4.389972686767578,
      "learning_rate": 9.929175822994526e-05,
      "loss": 2.734465408325195,
      "memory(GiB)": 47.63,
      "step": 6260,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.459654
    },
    {
      "epoch": 0.2684118075489482,
      "grad_norm": 3.4758071899414062,
      "learning_rate": 9.929062908539408e-05,
      "loss": 2.669170951843262,
      "memory(GiB)": 47.63,
      "step": 6265,
      "token_acc": 0.44405594405594406,
      "train_speed(iter/s)": 1.459501
    },
    {
      "epoch": 0.26862602287819715,
      "grad_norm": 3.4275498390197754,
      "learning_rate": 9.928949904789695e-05,
      "loss": 2.6379329681396486,
      "memory(GiB)": 47.63,
      "step": 6270,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.459464
    },
    {
      "epoch": 0.2688402382074461,
      "grad_norm": 4.899318218231201,
      "learning_rate": 9.928836811747429e-05,
      "loss": 2.7687206268310547,
      "memory(GiB)": 47.63,
      "step": 6275,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.459217
    },
    {
      "epoch": 0.2690544535366951,
      "grad_norm": 4.113958358764648,
      "learning_rate": 9.928723629414662e-05,
      "loss": 2.7000185012817384,
      "memory(GiB)": 47.63,
      "step": 6280,
      "token_acc": 0.5037037037037037,
      "train_speed(iter/s)": 1.459385
    },
    {
      "epoch": 0.269268668865944,
      "grad_norm": 4.145396709442139,
      "learning_rate": 9.928610357793441e-05,
      "loss": 2.6391624450683593,
      "memory(GiB)": 47.63,
      "step": 6285,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.459432
    },
    {
      "epoch": 0.269482884195193,
      "grad_norm": 5.5968918800354,
      "learning_rate": 9.928496996885821e-05,
      "loss": 2.4019641876220703,
      "memory(GiB)": 47.63,
      "step": 6290,
      "token_acc": 0.5060728744939271,
      "train_speed(iter/s)": 1.459583
    },
    {
      "epoch": 0.269697099524442,
      "grad_norm": 5.587214946746826,
      "learning_rate": 9.928383546693854e-05,
      "loss": 2.5242671966552734,
      "memory(GiB)": 47.63,
      "step": 6295,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.45965
    },
    {
      "epoch": 0.2699113148536909,
      "grad_norm": 4.530307292938232,
      "learning_rate": 9.928270007219598e-05,
      "loss": 2.593612289428711,
      "memory(GiB)": 47.63,
      "step": 6300,
      "token_acc": 0.4790996784565916,
      "train_speed(iter/s)": 1.459693
    },
    {
      "epoch": 0.2701255301829399,
      "grad_norm": 5.821206569671631,
      "learning_rate": 9.928156378465106e-05,
      "loss": 2.5492328643798827,
      "memory(GiB)": 47.63,
      "step": 6305,
      "token_acc": 0.47843137254901963,
      "train_speed(iter/s)": 1.460083
    },
    {
      "epoch": 0.27033974551218887,
      "grad_norm": 4.19746732711792,
      "learning_rate": 9.928042660432437e-05,
      "loss": 2.6815792083740235,
      "memory(GiB)": 47.63,
      "step": 6310,
      "token_acc": 0.42729970326409494,
      "train_speed(iter/s)": 1.460085
    },
    {
      "epoch": 0.2705539608414378,
      "grad_norm": 3.615483283996582,
      "learning_rate": 9.927928853123654e-05,
      "loss": 2.443739891052246,
      "memory(GiB)": 47.63,
      "step": 6315,
      "token_acc": 0.49842271293375395,
      "train_speed(iter/s)": 1.460239
    },
    {
      "epoch": 0.27076817617068677,
      "grad_norm": 5.410396099090576,
      "learning_rate": 9.927814956540818e-05,
      "loss": 2.5455259323120116,
      "memory(GiB)": 47.63,
      "step": 6320,
      "token_acc": 0.44981412639405205,
      "train_speed(iter/s)": 1.460304
    },
    {
      "epoch": 0.27098239149993575,
      "grad_norm": 5.281170845031738,
      "learning_rate": 9.927700970685989e-05,
      "loss": 2.694020652770996,
      "memory(GiB)": 47.63,
      "step": 6325,
      "token_acc": 0.46586345381526106,
      "train_speed(iter/s)": 1.460188
    },
    {
      "epoch": 0.27119660682918467,
      "grad_norm": 2.8500008583068848,
      "learning_rate": 9.927586895561237e-05,
      "loss": 2.289530372619629,
      "memory(GiB)": 47.63,
      "step": 6330,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.460152
    },
    {
      "epoch": 0.27141082215843365,
      "grad_norm": 4.507504463195801,
      "learning_rate": 9.927472731168623e-05,
      "loss": 2.4930225372314454,
      "memory(GiB)": 47.63,
      "step": 6335,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.460259
    },
    {
      "epoch": 0.27162503748768263,
      "grad_norm": 3.69659161567688,
      "learning_rate": 9.92735847751022e-05,
      "loss": 2.4775733947753906,
      "memory(GiB)": 47.63,
      "step": 6340,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.460101
    },
    {
      "epoch": 0.27183925281693155,
      "grad_norm": 12.177428245544434,
      "learning_rate": 9.927244134588095e-05,
      "loss": 2.730167007446289,
      "memory(GiB)": 47.63,
      "step": 6345,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.460416
    },
    {
      "epoch": 0.27205346814618053,
      "grad_norm": 4.370859622955322,
      "learning_rate": 9.927129702404321e-05,
      "loss": 2.5016876220703126,
      "memory(GiB)": 47.63,
      "step": 6350,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.460325
    },
    {
      "epoch": 0.2722676834754295,
      "grad_norm": 3.9999990463256836,
      "learning_rate": 9.92701518096097e-05,
      "loss": 2.726361083984375,
      "memory(GiB)": 47.63,
      "step": 6355,
      "token_acc": 0.4317343173431734,
      "train_speed(iter/s)": 1.460069
    },
    {
      "epoch": 0.27248189880467844,
      "grad_norm": 4.737161159515381,
      "learning_rate": 9.926900570260118e-05,
      "loss": 2.558404541015625,
      "memory(GiB)": 47.63,
      "step": 6360,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.460098
    },
    {
      "epoch": 0.2726961141339274,
      "grad_norm": 4.153711795806885,
      "learning_rate": 9.926785870303839e-05,
      "loss": 2.630443572998047,
      "memory(GiB)": 47.63,
      "step": 6365,
      "token_acc": 0.4682080924855491,
      "train_speed(iter/s)": 1.460066
    },
    {
      "epoch": 0.2729103294631764,
      "grad_norm": 3.8597517013549805,
      "learning_rate": 9.926671081094214e-05,
      "loss": 2.488731575012207,
      "memory(GiB)": 47.63,
      "step": 6370,
      "token_acc": 0.4549019607843137,
      "train_speed(iter/s)": 1.460238
    },
    {
      "epoch": 0.27312454479242537,
      "grad_norm": 4.428126335144043,
      "learning_rate": 9.92655620263332e-05,
      "loss": 2.739374351501465,
      "memory(GiB)": 47.63,
      "step": 6375,
      "token_acc": 0.45396825396825397,
      "train_speed(iter/s)": 1.460059
    },
    {
      "epoch": 0.2733387601216743,
      "grad_norm": 4.17251443862915,
      "learning_rate": 9.926441234923239e-05,
      "loss": 2.6665557861328124,
      "memory(GiB)": 47.63,
      "step": 6380,
      "token_acc": 0.44135802469135804,
      "train_speed(iter/s)": 1.459965
    },
    {
      "epoch": 0.2735529754509233,
      "grad_norm": 4.361823558807373,
      "learning_rate": 9.926326177966052e-05,
      "loss": 2.818448066711426,
      "memory(GiB)": 47.63,
      "step": 6385,
      "token_acc": 0.44333333333333336,
      "train_speed(iter/s)": 1.460136
    },
    {
      "epoch": 0.27376719078017225,
      "grad_norm": 3.3004982471466064,
      "learning_rate": 9.926211031763846e-05,
      "loss": 2.5723459243774416,
      "memory(GiB)": 47.63,
      "step": 6390,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.460242
    },
    {
      "epoch": 0.2739814061094212,
      "grad_norm": 4.8515543937683105,
      "learning_rate": 9.926095796318706e-05,
      "loss": 2.7295555114746093,
      "memory(GiB)": 47.63,
      "step": 6395,
      "token_acc": 0.45634920634920634,
      "train_speed(iter/s)": 1.460357
    },
    {
      "epoch": 0.27419562143867016,
      "grad_norm": 4.730801582336426,
      "learning_rate": 9.92598047163272e-05,
      "loss": 2.5801937103271486,
      "memory(GiB)": 47.63,
      "step": 6400,
      "token_acc": 0.4541832669322709,
      "train_speed(iter/s)": 1.460409
    },
    {
      "epoch": 0.27440983676791914,
      "grad_norm": 3.8852200508117676,
      "learning_rate": 9.925865057707977e-05,
      "loss": 3.041835403442383,
      "memory(GiB)": 47.63,
      "step": 6405,
      "token_acc": 0.4192546583850932,
      "train_speed(iter/s)": 1.460437
    },
    {
      "epoch": 0.27462405209716806,
      "grad_norm": 4.520515441894531,
      "learning_rate": 9.925749554546568e-05,
      "loss": 2.643325996398926,
      "memory(GiB)": 47.63,
      "step": 6410,
      "token_acc": 0.46440677966101696,
      "train_speed(iter/s)": 1.460586
    },
    {
      "epoch": 0.27483826742641704,
      "grad_norm": 3.491898775100708,
      "learning_rate": 9.925633962150584e-05,
      "loss": 2.39678955078125,
      "memory(GiB)": 47.63,
      "step": 6415,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.460348
    },
    {
      "epoch": 0.275052482755666,
      "grad_norm": 3.603445291519165,
      "learning_rate": 9.925518280522121e-05,
      "loss": 2.7103759765625,
      "memory(GiB)": 47.63,
      "step": 6420,
      "token_acc": 0.42946708463949845,
      "train_speed(iter/s)": 1.46015
    },
    {
      "epoch": 0.27526669808491494,
      "grad_norm": 3.179222345352173,
      "learning_rate": 9.925402509663273e-05,
      "loss": 2.5286163330078124,
      "memory(GiB)": 47.63,
      "step": 6425,
      "token_acc": 0.45652173913043476,
      "train_speed(iter/s)": 1.460259
    },
    {
      "epoch": 0.2754809134141639,
      "grad_norm": 3.502445936203003,
      "learning_rate": 9.92528664957614e-05,
      "loss": 2.733730697631836,
      "memory(GiB)": 47.63,
      "step": 6430,
      "token_acc": 0.447887323943662,
      "train_speed(iter/s)": 1.460065
    },
    {
      "epoch": 0.2756951287434129,
      "grad_norm": 4.481658935546875,
      "learning_rate": 9.925170700262817e-05,
      "loss": 2.6959110260009767,
      "memory(GiB)": 47.63,
      "step": 6435,
      "token_acc": 0.4489795918367347,
      "train_speed(iter/s)": 1.46009
    },
    {
      "epoch": 0.2759093440726618,
      "grad_norm": 3.666271448135376,
      "learning_rate": 9.925054661725406e-05,
      "loss": 2.81857967376709,
      "memory(GiB)": 47.63,
      "step": 6440,
      "token_acc": 0.4232081911262799,
      "train_speed(iter/s)": 1.46017
    },
    {
      "epoch": 0.2761235594019108,
      "grad_norm": 5.204419136047363,
      "learning_rate": 9.924938533966012e-05,
      "loss": 2.518722152709961,
      "memory(GiB)": 47.63,
      "step": 6445,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.460288
    },
    {
      "epoch": 0.2763377747311598,
      "grad_norm": 3.9146623611450195,
      "learning_rate": 9.924822316986735e-05,
      "loss": 2.5668308258056642,
      "memory(GiB)": 47.63,
      "step": 6450,
      "token_acc": 0.4738955823293173,
      "train_speed(iter/s)": 1.46037
    },
    {
      "epoch": 0.2765519900604087,
      "grad_norm": 4.472992897033691,
      "learning_rate": 9.924706010789683e-05,
      "loss": 2.757154846191406,
      "memory(GiB)": 47.63,
      "step": 6455,
      "token_acc": 0.39928057553956836,
      "train_speed(iter/s)": 1.460136
    },
    {
      "epoch": 0.2767662053896577,
      "grad_norm": 4.2314605712890625,
      "learning_rate": 9.924589615376962e-05,
      "loss": 2.5920293807983397,
      "memory(GiB)": 47.63,
      "step": 6460,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.460015
    },
    {
      "epoch": 0.27698042071890666,
      "grad_norm": 3.437063217163086,
      "learning_rate": 9.92447313075068e-05,
      "loss": 2.541089630126953,
      "memory(GiB)": 47.63,
      "step": 6465,
      "token_acc": 0.4691358024691358,
      "train_speed(iter/s)": 1.459735
    },
    {
      "epoch": 0.2771946360481556,
      "grad_norm": 5.263180255889893,
      "learning_rate": 9.924356556912946e-05,
      "loss": 2.538657379150391,
      "memory(GiB)": 47.63,
      "step": 6470,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.459798
    },
    {
      "epoch": 0.27740885137740456,
      "grad_norm": 4.6755852699279785,
      "learning_rate": 9.924239893865874e-05,
      "loss": 2.6377643585205077,
      "memory(GiB)": 47.63,
      "step": 6475,
      "token_acc": 0.48616600790513836,
      "train_speed(iter/s)": 1.46002
    },
    {
      "epoch": 0.27762306670665354,
      "grad_norm": 3.1691155433654785,
      "learning_rate": 9.924123141611578e-05,
      "loss": 2.4227458953857424,
      "memory(GiB)": 47.63,
      "step": 6480,
      "token_acc": 0.4901185770750988,
      "train_speed(iter/s)": 1.460192
    },
    {
      "epoch": 0.27783728203590247,
      "grad_norm": 5.500218391418457,
      "learning_rate": 9.924006300152173e-05,
      "loss": 2.5818939208984375,
      "memory(GiB)": 47.63,
      "step": 6485,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.460192
    },
    {
      "epoch": 0.27805149736515145,
      "grad_norm": 3.576978921890259,
      "learning_rate": 9.923889369489774e-05,
      "loss": 2.544826126098633,
      "memory(GiB)": 47.63,
      "step": 6490,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.460026
    },
    {
      "epoch": 0.2782657126944004,
      "grad_norm": 4.320828914642334,
      "learning_rate": 9.9237723496265e-05,
      "loss": 2.27396297454834,
      "memory(GiB)": 47.63,
      "step": 6495,
      "token_acc": 0.504950495049505,
      "train_speed(iter/s)": 1.460027
    },
    {
      "epoch": 0.27847992802364935,
      "grad_norm": 5.061696529388428,
      "learning_rate": 9.923655240564472e-05,
      "loss": 2.7767749786376954,
      "memory(GiB)": 47.63,
      "step": 6500,
      "token_acc": 0.43661971830985913,
      "train_speed(iter/s)": 1.460112
    },
    {
      "epoch": 0.27847992802364935,
      "eval_loss": 2.2300853729248047,
      "eval_runtime": 13.9267,
      "eval_samples_per_second": 7.18,
      "eval_steps_per_second": 7.18,
      "eval_token_acc": 0.47348951911220716,
      "step": 6500
    },
    {
      "epoch": 0.2786941433528983,
      "grad_norm": 3.8164281845092773,
      "learning_rate": 9.92353804230581e-05,
      "loss": 2.7503490447998047,
      "memory(GiB)": 47.63,
      "step": 6505,
      "token_acc": 0.4629294755877034,
      "train_speed(iter/s)": 1.455262
    },
    {
      "epoch": 0.2789083586821473,
      "grad_norm": 5.132174015045166,
      "learning_rate": 9.923420754852634e-05,
      "loss": 2.845453453063965,
      "memory(GiB)": 47.63,
      "step": 6510,
      "token_acc": 0.4187725631768953,
      "train_speed(iter/s)": 1.455479
    },
    {
      "epoch": 0.27912257401139623,
      "grad_norm": 4.182595252990723,
      "learning_rate": 9.923303378207077e-05,
      "loss": 2.765598678588867,
      "memory(GiB)": 47.63,
      "step": 6515,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.455863
    },
    {
      "epoch": 0.2793367893406452,
      "grad_norm": 5.0708394050598145,
      "learning_rate": 9.92318591237126e-05,
      "loss": 2.082514762878418,
      "memory(GiB)": 47.63,
      "step": 6520,
      "token_acc": 0.5726141078838174,
      "train_speed(iter/s)": 1.455809
    },
    {
      "epoch": 0.2795510046698942,
      "grad_norm": 4.086470127105713,
      "learning_rate": 9.923068357347312e-05,
      "loss": 2.5974002838134767,
      "memory(GiB)": 47.63,
      "step": 6525,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.456167
    },
    {
      "epoch": 0.2797652199991431,
      "grad_norm": 10.000062942504883,
      "learning_rate": 9.92295071313736e-05,
      "loss": 2.7181060791015623,
      "memory(GiB)": 47.63,
      "step": 6530,
      "token_acc": 0.4554794520547945,
      "train_speed(iter/s)": 1.456112
    },
    {
      "epoch": 0.2799794353283921,
      "grad_norm": 3.311910629272461,
      "learning_rate": 9.922832979743542e-05,
      "loss": 2.5410552978515626,
      "memory(GiB)": 47.63,
      "step": 6535,
      "token_acc": 0.44598337950138506,
      "train_speed(iter/s)": 1.456212
    },
    {
      "epoch": 0.28019365065764107,
      "grad_norm": 4.854501724243164,
      "learning_rate": 9.922715157167984e-05,
      "loss": 2.5051097869873047,
      "memory(GiB)": 47.63,
      "step": 6540,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.456141
    },
    {
      "epoch": 0.28040786598689005,
      "grad_norm": 4.3800129890441895,
      "learning_rate": 9.922597245412822e-05,
      "loss": 2.8104381561279297,
      "memory(GiB)": 47.63,
      "step": 6545,
      "token_acc": 0.4405144694533762,
      "train_speed(iter/s)": 1.456161
    },
    {
      "epoch": 0.28062208131613897,
      "grad_norm": 3.968067169189453,
      "learning_rate": 9.922479244480194e-05,
      "loss": 2.8367366790771484,
      "memory(GiB)": 47.63,
      "step": 6550,
      "token_acc": 0.44155844155844154,
      "train_speed(iter/s)": 1.456125
    },
    {
      "epoch": 0.28083629664538795,
      "grad_norm": 3.8599963188171387,
      "learning_rate": 9.922361154372237e-05,
      "loss": 2.4901859283447267,
      "memory(GiB)": 47.63,
      "step": 6555,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.456231
    },
    {
      "epoch": 0.28105051197463693,
      "grad_norm": 4.0197296142578125,
      "learning_rate": 9.922242975091092e-05,
      "loss": 2.4429126739501954,
      "memory(GiB)": 47.63,
      "step": 6560,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.456392
    },
    {
      "epoch": 0.28126472730388585,
      "grad_norm": 5.221249580383301,
      "learning_rate": 9.922124706638896e-05,
      "loss": 2.9108280181884765,
      "memory(GiB)": 47.63,
      "step": 6565,
      "token_acc": 0.43859649122807015,
      "train_speed(iter/s)": 1.456277
    },
    {
      "epoch": 0.28147894263313483,
      "grad_norm": 3.7353672981262207,
      "learning_rate": 9.922006349017793e-05,
      "loss": 2.744065284729004,
      "memory(GiB)": 47.63,
      "step": 6570,
      "token_acc": 0.44274809160305345,
      "train_speed(iter/s)": 1.456425
    },
    {
      "epoch": 0.2816931579623838,
      "grad_norm": 3.9291000366210938,
      "learning_rate": 9.92188790222993e-05,
      "loss": 2.686320114135742,
      "memory(GiB)": 47.63,
      "step": 6575,
      "token_acc": 0.47388059701492535,
      "train_speed(iter/s)": 1.456198
    },
    {
      "epoch": 0.28190737329163273,
      "grad_norm": 4.82513427734375,
      "learning_rate": 9.921769366277449e-05,
      "loss": 2.505121612548828,
      "memory(GiB)": 47.63,
      "step": 6580,
      "token_acc": 0.5038461538461538,
      "train_speed(iter/s)": 1.456238
    },
    {
      "epoch": 0.2821215886208817,
      "grad_norm": 4.253800868988037,
      "learning_rate": 9.9216507411625e-05,
      "loss": 2.924699401855469,
      "memory(GiB)": 47.63,
      "step": 6585,
      "token_acc": 0.4198250728862974,
      "train_speed(iter/s)": 1.456429
    },
    {
      "epoch": 0.2823358039501307,
      "grad_norm": 4.070432186126709,
      "learning_rate": 9.92153202688723e-05,
      "loss": 2.3470638275146483,
      "memory(GiB)": 47.63,
      "step": 6590,
      "token_acc": 0.5409252669039146,
      "train_speed(iter/s)": 1.456437
    },
    {
      "epoch": 0.2825500192793796,
      "grad_norm": 3.762282371520996,
      "learning_rate": 9.921413223453791e-05,
      "loss": 2.666696548461914,
      "memory(GiB)": 47.63,
      "step": 6595,
      "token_acc": 0.4218289085545723,
      "train_speed(iter/s)": 1.456519
    },
    {
      "epoch": 0.2827642346086286,
      "grad_norm": 4.166807174682617,
      "learning_rate": 9.921294330864334e-05,
      "loss": 2.8448354721069338,
      "memory(GiB)": 47.63,
      "step": 6600,
      "token_acc": 0.46686746987951805,
      "train_speed(iter/s)": 1.456385
    },
    {
      "epoch": 0.2829784499378776,
      "grad_norm": 5.437010765075684,
      "learning_rate": 9.921175349121015e-05,
      "loss": 2.6816150665283205,
      "memory(GiB)": 47.63,
      "step": 6605,
      "token_acc": 0.4416058394160584,
      "train_speed(iter/s)": 1.456789
    },
    {
      "epoch": 0.2831926652671265,
      "grad_norm": 3.6787517070770264,
      "learning_rate": 9.921056278225986e-05,
      "loss": 2.4739849090576174,
      "memory(GiB)": 47.63,
      "step": 6610,
      "token_acc": 0.5101351351351351,
      "train_speed(iter/s)": 1.456958
    },
    {
      "epoch": 0.2834068805963755,
      "grad_norm": 2.8377857208251953,
      "learning_rate": 9.920937118181408e-05,
      "loss": 2.6592750549316406,
      "memory(GiB)": 47.63,
      "step": 6615,
      "token_acc": 0.4169278996865204,
      "train_speed(iter/s)": 1.456925
    },
    {
      "epoch": 0.28362109592562446,
      "grad_norm": 3.365974187850952,
      "learning_rate": 9.920817868989439e-05,
      "loss": 2.3467428207397463,
      "memory(GiB)": 47.63,
      "step": 6620,
      "token_acc": 0.5303514376996805,
      "train_speed(iter/s)": 1.457005
    },
    {
      "epoch": 0.2838353112548734,
      "grad_norm": 4.327791213989258,
      "learning_rate": 9.920698530652235e-05,
      "loss": 2.6440120697021485,
      "memory(GiB)": 47.63,
      "step": 6625,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.456917
    },
    {
      "epoch": 0.28404952658412236,
      "grad_norm": 2.8808248043060303,
      "learning_rate": 9.920579103171963e-05,
      "loss": 2.6981765747070314,
      "memory(GiB)": 47.63,
      "step": 6630,
      "token_acc": 0.5016835016835017,
      "train_speed(iter/s)": 1.457114
    },
    {
      "epoch": 0.28426374191337134,
      "grad_norm": 4.228165626525879,
      "learning_rate": 9.920459586550785e-05,
      "loss": 2.6802579879760744,
      "memory(GiB)": 47.63,
      "step": 6635,
      "token_acc": 0.45660377358490567,
      "train_speed(iter/s)": 1.457054
    },
    {
      "epoch": 0.28447795724262026,
      "grad_norm": 4.075457572937012,
      "learning_rate": 9.920339980790864e-05,
      "loss": 2.4691463470458985,
      "memory(GiB)": 47.63,
      "step": 6640,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.45718
    },
    {
      "epoch": 0.28469217257186924,
      "grad_norm": 3.6696760654449463,
      "learning_rate": 9.92022028589437e-05,
      "loss": 2.7696184158325194,
      "memory(GiB)": 47.63,
      "step": 6645,
      "token_acc": 0.46886446886446886,
      "train_speed(iter/s)": 1.457392
    },
    {
      "epoch": 0.2849063879011182,
      "grad_norm": 3.7545242309570312,
      "learning_rate": 9.920100501863472e-05,
      "loss": 2.781435012817383,
      "memory(GiB)": 47.63,
      "step": 6650,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.457431
    },
    {
      "epoch": 0.28512060323036714,
      "grad_norm": 3.9622879028320312,
      "learning_rate": 9.919980628700335e-05,
      "loss": 2.5763864517211914,
      "memory(GiB)": 47.63,
      "step": 6655,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.457464
    },
    {
      "epoch": 0.2853348185596161,
      "grad_norm": 5.135830879211426,
      "learning_rate": 9.919860666407135e-05,
      "loss": 2.903422546386719,
      "memory(GiB)": 47.63,
      "step": 6660,
      "token_acc": 0.4277456647398844,
      "train_speed(iter/s)": 1.457552
    },
    {
      "epoch": 0.2855490338888651,
      "grad_norm": 4.834655284881592,
      "learning_rate": 9.919740614986043e-05,
      "loss": 2.7380495071411133,
      "memory(GiB)": 47.63,
      "step": 6665,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.457405
    },
    {
      "epoch": 0.285763249218114,
      "grad_norm": 3.739605188369751,
      "learning_rate": 9.919620474439236e-05,
      "loss": 2.609376907348633,
      "memory(GiB)": 47.63,
      "step": 6670,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.457537
    },
    {
      "epoch": 0.285977464547363,
      "grad_norm": 3.5914204120635986,
      "learning_rate": 9.919500244768886e-05,
      "loss": 2.6851119995117188,
      "memory(GiB)": 47.63,
      "step": 6675,
      "token_acc": 0.43728813559322033,
      "train_speed(iter/s)": 1.457429
    },
    {
      "epoch": 0.286191679876612,
      "grad_norm": 3.5740292072296143,
      "learning_rate": 9.919379925977177e-05,
      "loss": 2.534268379211426,
      "memory(GiB)": 47.63,
      "step": 6680,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.457522
    },
    {
      "epoch": 0.2864058952058609,
      "grad_norm": 3.6544411182403564,
      "learning_rate": 9.919259518066285e-05,
      "loss": 2.413597297668457,
      "memory(GiB)": 47.63,
      "step": 6685,
      "token_acc": 0.4701195219123506,
      "train_speed(iter/s)": 1.457378
    },
    {
      "epoch": 0.2866201105351099,
      "grad_norm": 4.151614665985107,
      "learning_rate": 9.919139021038392e-05,
      "loss": 2.855759620666504,
      "memory(GiB)": 47.63,
      "step": 6690,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.457442
    },
    {
      "epoch": 0.28683432586435886,
      "grad_norm": 5.356203556060791,
      "learning_rate": 9.919018434895681e-05,
      "loss": 2.4592670440673827,
      "memory(GiB)": 47.63,
      "step": 6695,
      "token_acc": 0.4879032258064516,
      "train_speed(iter/s)": 1.457558
    },
    {
      "epoch": 0.2870485411936078,
      "grad_norm": 4.413278579711914,
      "learning_rate": 9.918897759640338e-05,
      "loss": 2.4328805923461916,
      "memory(GiB)": 47.63,
      "step": 6700,
      "token_acc": 0.496,
      "train_speed(iter/s)": 1.457612
    },
    {
      "epoch": 0.28726275652285677,
      "grad_norm": 4.578810691833496,
      "learning_rate": 9.918776995274547e-05,
      "loss": 2.7367570877075194,
      "memory(GiB)": 47.63,
      "step": 6705,
      "token_acc": 0.43506493506493504,
      "train_speed(iter/s)": 1.4576
    },
    {
      "epoch": 0.28747697185210574,
      "grad_norm": 4.948911190032959,
      "learning_rate": 9.918656141800496e-05,
      "loss": 2.6622766494750976,
      "memory(GiB)": 47.63,
      "step": 6710,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.45722
    },
    {
      "epoch": 0.2876911871813547,
      "grad_norm": 4.788825035095215,
      "learning_rate": 9.918535199220376e-05,
      "loss": 2.6083139419555663,
      "memory(GiB)": 47.63,
      "step": 6715,
      "token_acc": 0.44660194174757284,
      "train_speed(iter/s)": 1.457461
    },
    {
      "epoch": 0.28790540251060365,
      "grad_norm": 3.6893582344055176,
      "learning_rate": 9.918414167536376e-05,
      "loss": 2.7440122604370116,
      "memory(GiB)": 47.63,
      "step": 6720,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.457459
    },
    {
      "epoch": 0.2881196178398526,
      "grad_norm": 5.994373798370361,
      "learning_rate": 9.918293046750689e-05,
      "loss": 2.5064340591430665,
      "memory(GiB)": 47.63,
      "step": 6725,
      "token_acc": 0.4601449275362319,
      "train_speed(iter/s)": 1.45765
    },
    {
      "epoch": 0.2883338331691016,
      "grad_norm": 4.485553741455078,
      "learning_rate": 9.918171836865511e-05,
      "loss": 2.9492183685302735,
      "memory(GiB)": 47.63,
      "step": 6730,
      "token_acc": 0.4421052631578947,
      "train_speed(iter/s)": 1.457606
    },
    {
      "epoch": 0.28854804849835053,
      "grad_norm": 4.04892635345459,
      "learning_rate": 9.918050537883037e-05,
      "loss": 2.4176584243774415,
      "memory(GiB)": 47.63,
      "step": 6735,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.457516
    },
    {
      "epoch": 0.2887622638275995,
      "grad_norm": 3.8965468406677246,
      "learning_rate": 9.917929149805462e-05,
      "loss": 2.4605600357055666,
      "memory(GiB)": 47.63,
      "step": 6740,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.457132
    },
    {
      "epoch": 0.2889764791568485,
      "grad_norm": 5.634366989135742,
      "learning_rate": 9.917807672634989e-05,
      "loss": 2.6476724624633787,
      "memory(GiB)": 47.63,
      "step": 6745,
      "token_acc": 0.416988416988417,
      "train_speed(iter/s)": 1.457053
    },
    {
      "epoch": 0.2891906944860974,
      "grad_norm": 3.4423749446868896,
      "learning_rate": 9.917686106373816e-05,
      "loss": 2.393769454956055,
      "memory(GiB)": 47.63,
      "step": 6750,
      "token_acc": 0.5198412698412699,
      "train_speed(iter/s)": 1.457149
    },
    {
      "epoch": 0.2894049098153464,
      "grad_norm": 3.979480028152466,
      "learning_rate": 9.917564451024147e-05,
      "loss": 2.6255361557006838,
      "memory(GiB)": 47.63,
      "step": 6755,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.457225
    },
    {
      "epoch": 0.28961912514459537,
      "grad_norm": 3.3376545906066895,
      "learning_rate": 9.917442706588183e-05,
      "loss": 2.601759910583496,
      "memory(GiB)": 47.63,
      "step": 6760,
      "token_acc": 0.46107784431137727,
      "train_speed(iter/s)": 1.457244
    },
    {
      "epoch": 0.2898333404738443,
      "grad_norm": 3.493684768676758,
      "learning_rate": 9.917320873068132e-05,
      "loss": 2.507145118713379,
      "memory(GiB)": 47.63,
      "step": 6765,
      "token_acc": 0.4876712328767123,
      "train_speed(iter/s)": 1.457043
    },
    {
      "epoch": 0.29004755580309327,
      "grad_norm": 4.453976631164551,
      "learning_rate": 9.9171989504662e-05,
      "loss": 2.54097843170166,
      "memory(GiB)": 47.63,
      "step": 6770,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.457142
    },
    {
      "epoch": 0.29026177113234225,
      "grad_norm": 4.38257360458374,
      "learning_rate": 9.917076938784597e-05,
      "loss": 2.573847198486328,
      "memory(GiB)": 47.63,
      "step": 6775,
      "token_acc": 0.4459016393442623,
      "train_speed(iter/s)": 1.457361
    },
    {
      "epoch": 0.2904759864615912,
      "grad_norm": 4.03444766998291,
      "learning_rate": 9.916954838025533e-05,
      "loss": 2.2041826248168945,
      "memory(GiB)": 47.63,
      "step": 6780,
      "token_acc": 0.5365853658536586,
      "train_speed(iter/s)": 1.457458
    },
    {
      "epoch": 0.29069020179084015,
      "grad_norm": 4.5989789962768555,
      "learning_rate": 9.91683264819122e-05,
      "loss": 2.7307586669921875,
      "memory(GiB)": 47.63,
      "step": 6785,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.457661
    },
    {
      "epoch": 0.29090441712008913,
      "grad_norm": 6.401525497436523,
      "learning_rate": 9.91671036928387e-05,
      "loss": 2.385505485534668,
      "memory(GiB)": 47.63,
      "step": 6790,
      "token_acc": 0.48398576512455516,
      "train_speed(iter/s)": 1.457658
    },
    {
      "epoch": 0.29111863244933806,
      "grad_norm": 4.399469375610352,
      "learning_rate": 9.9165880013057e-05,
      "loss": 2.6143795013427735,
      "memory(GiB)": 47.63,
      "step": 6795,
      "token_acc": 0.45977011494252873,
      "train_speed(iter/s)": 1.457322
    },
    {
      "epoch": 0.29133284777858703,
      "grad_norm": 3.8054795265197754,
      "learning_rate": 9.916465544258926e-05,
      "loss": 2.313238525390625,
      "memory(GiB)": 47.63,
      "step": 6800,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.457467
    },
    {
      "epoch": 0.291547063107836,
      "grad_norm": 4.2663984298706055,
      "learning_rate": 9.916342998145766e-05,
      "loss": 2.6732725143432616,
      "memory(GiB)": 47.63,
      "step": 6805,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.457199
    },
    {
      "epoch": 0.29176127843708494,
      "grad_norm": 7.208676815032959,
      "learning_rate": 9.916220362968443e-05,
      "loss": 2.809499740600586,
      "memory(GiB)": 47.63,
      "step": 6810,
      "token_acc": 0.4392523364485981,
      "train_speed(iter/s)": 1.457302
    },
    {
      "epoch": 0.2919754937663339,
      "grad_norm": 5.0455427169799805,
      "learning_rate": 9.916097638729174e-05,
      "loss": 2.493167686462402,
      "memory(GiB)": 47.63,
      "step": 6815,
      "token_acc": 0.49226006191950467,
      "train_speed(iter/s)": 1.457444
    },
    {
      "epoch": 0.2921897090955829,
      "grad_norm": 3.188335418701172,
      "learning_rate": 9.915974825430187e-05,
      "loss": 2.6794239044189454,
      "memory(GiB)": 47.63,
      "step": 6820,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.457734
    },
    {
      "epoch": 0.2924039244248318,
      "grad_norm": 4.809487819671631,
      "learning_rate": 9.915851923073702e-05,
      "loss": 2.5735183715820313,
      "memory(GiB)": 47.63,
      "step": 6825,
      "token_acc": 0.49166666666666664,
      "train_speed(iter/s)": 1.457673
    },
    {
      "epoch": 0.2926181397540808,
      "grad_norm": 3.612699031829834,
      "learning_rate": 9.915728931661949e-05,
      "loss": 2.7919570922851564,
      "memory(GiB)": 47.63,
      "step": 6830,
      "token_acc": 0.43728813559322033,
      "train_speed(iter/s)": 1.457787
    },
    {
      "epoch": 0.2928323550833298,
      "grad_norm": 3.884958267211914,
      "learning_rate": 9.915605851197156e-05,
      "loss": 2.87333984375,
      "memory(GiB)": 47.63,
      "step": 6835,
      "token_acc": 0.4367816091954023,
      "train_speed(iter/s)": 1.457947
    },
    {
      "epoch": 0.2930465704125787,
      "grad_norm": 4.934978485107422,
      "learning_rate": 9.91548268168155e-05,
      "loss": 2.6373437881469726,
      "memory(GiB)": 47.63,
      "step": 6840,
      "token_acc": 0.4657534246575342,
      "train_speed(iter/s)": 1.457901
    },
    {
      "epoch": 0.2932607857418277,
      "grad_norm": 5.34118127822876,
      "learning_rate": 9.915359423117366e-05,
      "loss": 2.6255226135253906,
      "memory(GiB)": 47.63,
      "step": 6845,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.457858
    },
    {
      "epoch": 0.29347500107107666,
      "grad_norm": 3.7396907806396484,
      "learning_rate": 9.915236075506833e-05,
      "loss": 2.570332336425781,
      "memory(GiB)": 47.63,
      "step": 6850,
      "token_acc": 0.4318181818181818,
      "train_speed(iter/s)": 1.458032
    },
    {
      "epoch": 0.2936892164003256,
      "grad_norm": 4.2154059410095215,
      "learning_rate": 9.915112638852188e-05,
      "loss": 2.5960338592529295,
      "memory(GiB)": 47.63,
      "step": 6855,
      "token_acc": 0.47843137254901963,
      "train_speed(iter/s)": 1.457974
    },
    {
      "epoch": 0.29390343172957456,
      "grad_norm": 5.163266658782959,
      "learning_rate": 9.914989113155668e-05,
      "loss": 2.8140689849853517,
      "memory(GiB)": 47.63,
      "step": 6860,
      "token_acc": 0.40594059405940597,
      "train_speed(iter/s)": 1.458094
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 5.603063106536865,
      "learning_rate": 9.91486549841951e-05,
      "loss": 2.5724794387817385,
      "memory(GiB)": 47.63,
      "step": 6865,
      "token_acc": 0.440625,
      "train_speed(iter/s)": 1.458086
    },
    {
      "epoch": 0.2943318623880725,
      "grad_norm": 4.696154594421387,
      "learning_rate": 9.914741794645952e-05,
      "loss": 2.582904815673828,
      "memory(GiB)": 47.63,
      "step": 6870,
      "token_acc": 0.47115384615384615,
      "train_speed(iter/s)": 1.458003
    },
    {
      "epoch": 0.29454607771732144,
      "grad_norm": 4.207950115203857,
      "learning_rate": 9.914618001837235e-05,
      "loss": 2.408650016784668,
      "memory(GiB)": 47.63,
      "step": 6875,
      "token_acc": 0.5211267605633803,
      "train_speed(iter/s)": 1.457964
    },
    {
      "epoch": 0.2947602930465704,
      "grad_norm": 4.780416965484619,
      "learning_rate": 9.914494119995604e-05,
      "loss": 2.9929107666015624,
      "memory(GiB)": 47.63,
      "step": 6880,
      "token_acc": 0.40202702702702703,
      "train_speed(iter/s)": 1.457783
    },
    {
      "epoch": 0.2949745083758194,
      "grad_norm": 4.01948881149292,
      "learning_rate": 9.914370149123302e-05,
      "loss": 2.7052196502685546,
      "memory(GiB)": 47.63,
      "step": 6885,
      "token_acc": 0.46254071661237783,
      "train_speed(iter/s)": 1.45741
    },
    {
      "epoch": 0.2951887237050683,
      "grad_norm": 3.457794666290283,
      "learning_rate": 9.914246089222575e-05,
      "loss": 2.8521696090698243,
      "memory(GiB)": 47.63,
      "step": 6890,
      "token_acc": 0.42036553524804177,
      "train_speed(iter/s)": 1.457387
    },
    {
      "epoch": 0.2954029390343173,
      "grad_norm": 6.059695243835449,
      "learning_rate": 9.914121940295669e-05,
      "loss": 2.523684501647949,
      "memory(GiB)": 47.63,
      "step": 6895,
      "token_acc": 0.48046875,
      "train_speed(iter/s)": 1.457385
    },
    {
      "epoch": 0.2956171543635663,
      "grad_norm": 5.192733287811279,
      "learning_rate": 9.913997702344834e-05,
      "loss": 2.4845111846923826,
      "memory(GiB)": 47.63,
      "step": 6900,
      "token_acc": 0.5111940298507462,
      "train_speed(iter/s)": 1.457472
    },
    {
      "epoch": 0.2958313696928152,
      "grad_norm": 3.800037145614624,
      "learning_rate": 9.913873375372321e-05,
      "loss": 2.552079772949219,
      "memory(GiB)": 47.63,
      "step": 6905,
      "token_acc": 0.4645390070921986,
      "train_speed(iter/s)": 1.457481
    },
    {
      "epoch": 0.2960455850220642,
      "grad_norm": 3.646867513656616,
      "learning_rate": 9.913748959380382e-05,
      "loss": 2.5172801971435548,
      "memory(GiB)": 47.63,
      "step": 6910,
      "token_acc": 0.477124183006536,
      "train_speed(iter/s)": 1.457499
    },
    {
      "epoch": 0.29625980035131316,
      "grad_norm": 3.363344669342041,
      "learning_rate": 9.913624454371273e-05,
      "loss": 2.3248594284057615,
      "memory(GiB)": 47.63,
      "step": 6915,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.457524
    },
    {
      "epoch": 0.2964740156805621,
      "grad_norm": 3.9670493602752686,
      "learning_rate": 9.913499860347246e-05,
      "loss": 2.7420141220092775,
      "memory(GiB)": 47.63,
      "step": 6920,
      "token_acc": 0.41304347826086957,
      "train_speed(iter/s)": 1.457685
    },
    {
      "epoch": 0.29668823100981107,
      "grad_norm": 5.507998466491699,
      "learning_rate": 9.91337517731056e-05,
      "loss": 2.9616031646728516,
      "memory(GiB)": 47.63,
      "step": 6925,
      "token_acc": 0.4177215189873418,
      "train_speed(iter/s)": 1.457645
    },
    {
      "epoch": 0.29690244633906004,
      "grad_norm": 3.6019091606140137,
      "learning_rate": 9.913250405263474e-05,
      "loss": 2.7343515396118163,
      "memory(GiB)": 47.63,
      "step": 6930,
      "token_acc": 0.4696969696969697,
      "train_speed(iter/s)": 1.45789
    },
    {
      "epoch": 0.29711666166830897,
      "grad_norm": 4.727652549743652,
      "learning_rate": 9.913125544208248e-05,
      "loss": 2.784088897705078,
      "memory(GiB)": 47.63,
      "step": 6935,
      "token_acc": 0.4229390681003584,
      "train_speed(iter/s)": 1.458068
    },
    {
      "epoch": 0.29733087699755795,
      "grad_norm": 5.329716682434082,
      "learning_rate": 9.913000594147144e-05,
      "loss": 2.8175352096557615,
      "memory(GiB)": 47.63,
      "step": 6940,
      "token_acc": 0.4262295081967213,
      "train_speed(iter/s)": 1.458009
    },
    {
      "epoch": 0.2975450923268069,
      "grad_norm": 3.970320224761963,
      "learning_rate": 9.912875555082425e-05,
      "loss": 2.7154836654663086,
      "memory(GiB)": 47.63,
      "step": 6945,
      "token_acc": 0.43380281690140843,
      "train_speed(iter/s)": 1.457928
    },
    {
      "epoch": 0.29775930765605585,
      "grad_norm": 3.9871678352355957,
      "learning_rate": 9.912750427016356e-05,
      "loss": 2.4553781509399415,
      "memory(GiB)": 47.63,
      "step": 6950,
      "token_acc": 0.45660377358490567,
      "train_speed(iter/s)": 1.457848
    },
    {
      "epoch": 0.29797352298530483,
      "grad_norm": 3.7019360065460205,
      "learning_rate": 9.912625209951206e-05,
      "loss": 2.7154155731201173,
      "memory(GiB)": 47.63,
      "step": 6955,
      "token_acc": 0.4331983805668016,
      "train_speed(iter/s)": 1.458222
    },
    {
      "epoch": 0.2981877383145538,
      "grad_norm": 3.8082334995269775,
      "learning_rate": 9.91249990388924e-05,
      "loss": 2.478436279296875,
      "memory(GiB)": 47.63,
      "step": 6960,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.45805
    },
    {
      "epoch": 0.29840195364380273,
      "grad_norm": 2.990219831466675,
      "learning_rate": 9.912374508832732e-05,
      "loss": 2.3979909896850584,
      "memory(GiB)": 47.63,
      "step": 6965,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.457916
    },
    {
      "epoch": 0.2986161689730517,
      "grad_norm": 4.876752853393555,
      "learning_rate": 9.912249024783951e-05,
      "loss": 2.6312931060791014,
      "memory(GiB)": 47.63,
      "step": 6970,
      "token_acc": 0.4504792332268371,
      "train_speed(iter/s)": 1.457852
    },
    {
      "epoch": 0.2988303843023007,
      "grad_norm": 3.5672826766967773,
      "learning_rate": 9.91212345174517e-05,
      "loss": 2.378168487548828,
      "memory(GiB)": 47.63,
      "step": 6975,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.458002
    },
    {
      "epoch": 0.2990445996315496,
      "grad_norm": 3.7002193927764893,
      "learning_rate": 9.911997789718666e-05,
      "loss": 2.4658145904541016,
      "memory(GiB)": 47.63,
      "step": 6980,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.457727
    },
    {
      "epoch": 0.2992588149607986,
      "grad_norm": 3.7269675731658936,
      "learning_rate": 9.911872038706713e-05,
      "loss": 2.325260353088379,
      "memory(GiB)": 47.63,
      "step": 6985,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.45787
    },
    {
      "epoch": 0.29947303029004757,
      "grad_norm": 3.892512559890747,
      "learning_rate": 9.911746198711591e-05,
      "loss": 2.2504718780517576,
      "memory(GiB)": 47.63,
      "step": 6990,
      "token_acc": 0.5288888888888889,
      "train_speed(iter/s)": 1.457971
    },
    {
      "epoch": 0.2996872456192965,
      "grad_norm": 5.19597053527832,
      "learning_rate": 9.911620269735578e-05,
      "loss": 2.3118865966796873,
      "memory(GiB)": 47.63,
      "step": 6995,
      "token_acc": 0.5269709543568465,
      "train_speed(iter/s)": 1.457796
    },
    {
      "epoch": 0.2999014609485455,
      "grad_norm": 3.593531370162964,
      "learning_rate": 9.911494251780957e-05,
      "loss": 2.300238037109375,
      "memory(GiB)": 47.63,
      "step": 7000,
      "token_acc": 0.47333333333333333,
      "train_speed(iter/s)": 1.457821
    },
    {
      "epoch": 0.2999014609485455,
      "eval_loss": 2.3745665550231934,
      "eval_runtime": 14.7829,
      "eval_samples_per_second": 6.765,
      "eval_steps_per_second": 6.765,
      "eval_token_acc": 0.4563445867287544,
      "step": 7000
    },
    {
      "epoch": 0.30011567627779445,
      "grad_norm": 4.208698749542236,
      "learning_rate": 9.911368144850011e-05,
      "loss": 2.863749694824219,
      "memory(GiB)": 47.63,
      "step": 7005,
      "token_acc": 0.4390444810543657,
      "train_speed(iter/s)": 1.453302
    },
    {
      "epoch": 0.3003298916070434,
      "grad_norm": 3.87911319732666,
      "learning_rate": 9.911241948945022e-05,
      "loss": 2.394370269775391,
      "memory(GiB)": 47.63,
      "step": 7010,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.453202
    },
    {
      "epoch": 0.30054410693629235,
      "grad_norm": 4.127148151397705,
      "learning_rate": 9.91111566406828e-05,
      "loss": 2.476901626586914,
      "memory(GiB)": 47.63,
      "step": 7015,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.453335
    },
    {
      "epoch": 0.30075832226554133,
      "grad_norm": 5.481370449066162,
      "learning_rate": 9.91098929022207e-05,
      "loss": 3.0413379669189453,
      "memory(GiB)": 47.63,
      "step": 7020,
      "token_acc": 0.39416058394160586,
      "train_speed(iter/s)": 1.453233
    },
    {
      "epoch": 0.30097253759479026,
      "grad_norm": 3.688483238220215,
      "learning_rate": 9.910862827408682e-05,
      "loss": 2.909566879272461,
      "memory(GiB)": 47.63,
      "step": 7025,
      "token_acc": 0.4053156146179402,
      "train_speed(iter/s)": 1.453437
    },
    {
      "epoch": 0.30118675292403924,
      "grad_norm": 5.38715934753418,
      "learning_rate": 9.910736275630408e-05,
      "loss": 2.6324440002441407,
      "memory(GiB)": 47.63,
      "step": 7030,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.453355
    },
    {
      "epoch": 0.3014009682532882,
      "grad_norm": 3.6263034343719482,
      "learning_rate": 9.910609634889538e-05,
      "loss": 2.7652103424072267,
      "memory(GiB)": 47.63,
      "step": 7035,
      "token_acc": 0.4485294117647059,
      "train_speed(iter/s)": 1.453448
    },
    {
      "epoch": 0.3016151835825372,
      "grad_norm": 3.28666615486145,
      "learning_rate": 9.91048290518837e-05,
      "loss": 2.8325773239135743,
      "memory(GiB)": 47.63,
      "step": 7040,
      "token_acc": 0.44565217391304346,
      "train_speed(iter/s)": 1.4538
    },
    {
      "epoch": 0.3018293989117861,
      "grad_norm": 3.835930109024048,
      "learning_rate": 9.910356086529196e-05,
      "loss": 2.8261871337890625,
      "memory(GiB)": 47.63,
      "step": 7045,
      "token_acc": 0.42641509433962266,
      "train_speed(iter/s)": 1.453892
    },
    {
      "epoch": 0.3020436142410351,
      "grad_norm": 4.7793145179748535,
      "learning_rate": 9.910229178914317e-05,
      "loss": 2.817231369018555,
      "memory(GiB)": 47.63,
      "step": 7050,
      "token_acc": 0.436241610738255,
      "train_speed(iter/s)": 1.453946
    },
    {
      "epoch": 0.3022578295702841,
      "grad_norm": 3.891536235809326,
      "learning_rate": 9.910102182346029e-05,
      "loss": 2.7835426330566406,
      "memory(GiB)": 47.63,
      "step": 7055,
      "token_acc": 0.4294478527607362,
      "train_speed(iter/s)": 1.454071
    },
    {
      "epoch": 0.302472044899533,
      "grad_norm": 3.9592056274414062,
      "learning_rate": 9.909975096826634e-05,
      "loss": 2.850838851928711,
      "memory(GiB)": 47.63,
      "step": 7060,
      "token_acc": 0.4098360655737705,
      "train_speed(iter/s)": 1.454033
    },
    {
      "epoch": 0.302686260228782,
      "grad_norm": 4.836696147918701,
      "learning_rate": 9.909847922358432e-05,
      "loss": 2.7942623138427733,
      "memory(GiB)": 47.63,
      "step": 7065,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.45406
    },
    {
      "epoch": 0.30290047555803096,
      "grad_norm": 3.4888906478881836,
      "learning_rate": 9.909720658943733e-05,
      "loss": 2.695857810974121,
      "memory(GiB)": 47.63,
      "step": 7070,
      "token_acc": 0.4300341296928328,
      "train_speed(iter/s)": 1.454272
    },
    {
      "epoch": 0.3031146908872799,
      "grad_norm": 3.878528118133545,
      "learning_rate": 9.909593306584837e-05,
      "loss": 2.5962528228759765,
      "memory(GiB)": 47.63,
      "step": 7075,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.454447
    },
    {
      "epoch": 0.30332890621652886,
      "grad_norm": 4.596851348876953,
      "learning_rate": 9.909465865284052e-05,
      "loss": 2.633403778076172,
      "memory(GiB)": 47.63,
      "step": 7080,
      "token_acc": 0.4676258992805755,
      "train_speed(iter/s)": 1.454547
    },
    {
      "epoch": 0.30354312154577784,
      "grad_norm": 5.361253261566162,
      "learning_rate": 9.909338335043688e-05,
      "loss": 2.7792930603027344,
      "memory(GiB)": 47.63,
      "step": 7085,
      "token_acc": 0.40942028985507245,
      "train_speed(iter/s)": 1.4548
    },
    {
      "epoch": 0.30375733687502676,
      "grad_norm": 5.789522647857666,
      "learning_rate": 9.909210715866055e-05,
      "loss": 2.5569536209106447,
      "memory(GiB)": 47.63,
      "step": 7090,
      "token_acc": 0.5064935064935064,
      "train_speed(iter/s)": 1.455016
    },
    {
      "epoch": 0.30397155220427574,
      "grad_norm": 4.068614482879639,
      "learning_rate": 9.909083007753464e-05,
      "loss": 2.817095947265625,
      "memory(GiB)": 47.63,
      "step": 7095,
      "token_acc": 0.4180602006688963,
      "train_speed(iter/s)": 1.455218
    },
    {
      "epoch": 0.3041857675335247,
      "grad_norm": 3.8201069831848145,
      "learning_rate": 9.90895521070823e-05,
      "loss": 2.5596477508544924,
      "memory(GiB)": 47.63,
      "step": 7100,
      "token_acc": 0.42073170731707316,
      "train_speed(iter/s)": 1.454855
    },
    {
      "epoch": 0.30439998286277364,
      "grad_norm": 4.3132781982421875,
      "learning_rate": 9.908827324732667e-05,
      "loss": 2.786307525634766,
      "memory(GiB)": 47.63,
      "step": 7105,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.454885
    },
    {
      "epoch": 0.3046141981920226,
      "grad_norm": 4.772188186645508,
      "learning_rate": 9.908699349829091e-05,
      "loss": 2.690424346923828,
      "memory(GiB)": 47.63,
      "step": 7110,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.45481
    },
    {
      "epoch": 0.3048284135212716,
      "grad_norm": 3.452068567276001,
      "learning_rate": 9.908571285999824e-05,
      "loss": 2.748585891723633,
      "memory(GiB)": 47.63,
      "step": 7115,
      "token_acc": 0.4362017804154303,
      "train_speed(iter/s)": 1.454919
    },
    {
      "epoch": 0.3050426288505205,
      "grad_norm": 4.686628818511963,
      "learning_rate": 9.908443133247182e-05,
      "loss": 3.0903095245361327,
      "memory(GiB)": 47.63,
      "step": 7120,
      "token_acc": 0.42901234567901236,
      "train_speed(iter/s)": 1.455089
    },
    {
      "epoch": 0.3052568441797695,
      "grad_norm": 5.241159915924072,
      "learning_rate": 9.908314891573489e-05,
      "loss": 2.8816459655761717,
      "memory(GiB)": 47.63,
      "step": 7125,
      "token_acc": 0.41333333333333333,
      "train_speed(iter/s)": 1.455082
    },
    {
      "epoch": 0.3054710595090185,
      "grad_norm": 3.407045602798462,
      "learning_rate": 9.908186560981066e-05,
      "loss": 2.6806121826171876,
      "memory(GiB)": 47.63,
      "step": 7130,
      "token_acc": 0.5210727969348659,
      "train_speed(iter/s)": 1.455067
    },
    {
      "epoch": 0.3056852748382674,
      "grad_norm": 4.179222106933594,
      "learning_rate": 9.908058141472239e-05,
      "loss": 2.7187618255615233,
      "memory(GiB)": 47.63,
      "step": 7135,
      "token_acc": 0.4358108108108108,
      "train_speed(iter/s)": 1.455106
    },
    {
      "epoch": 0.3058994901675164,
      "grad_norm": 3.6603033542633057,
      "learning_rate": 9.907929633049336e-05,
      "loss": 2.6038034439086912,
      "memory(GiB)": 47.63,
      "step": 7140,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.455309
    },
    {
      "epoch": 0.30611370549676536,
      "grad_norm": 4.7150983810424805,
      "learning_rate": 9.907801035714684e-05,
      "loss": 2.654140281677246,
      "memory(GiB)": 47.63,
      "step": 7145,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.455513
    },
    {
      "epoch": 0.3063279208260143,
      "grad_norm": 4.321269989013672,
      "learning_rate": 9.907672349470612e-05,
      "loss": 3.074951934814453,
      "memory(GiB)": 47.63,
      "step": 7150,
      "token_acc": 0.43389830508474575,
      "train_speed(iter/s)": 1.455334
    },
    {
      "epoch": 0.30654213615526327,
      "grad_norm": 4.386789321899414,
      "learning_rate": 9.90754357431945e-05,
      "loss": 2.465946006774902,
      "memory(GiB)": 47.63,
      "step": 7155,
      "token_acc": 0.47633136094674555,
      "train_speed(iter/s)": 1.455479
    },
    {
      "epoch": 0.30675635148451225,
      "grad_norm": 3.6322429180145264,
      "learning_rate": 9.907414710263534e-05,
      "loss": 2.7125003814697264,
      "memory(GiB)": 47.63,
      "step": 7160,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.455717
    },
    {
      "epoch": 0.30697056681376117,
      "grad_norm": 3.60015606880188,
      "learning_rate": 9.907285757305198e-05,
      "loss": 2.580677795410156,
      "memory(GiB)": 47.63,
      "step": 7165,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.456008
    },
    {
      "epoch": 0.30718478214301015,
      "grad_norm": 5.091457366943359,
      "learning_rate": 9.907156715446775e-05,
      "loss": 2.790735054016113,
      "memory(GiB)": 47.63,
      "step": 7170,
      "token_acc": 0.4506172839506173,
      "train_speed(iter/s)": 1.456158
    },
    {
      "epoch": 0.30739899747225913,
      "grad_norm": 4.9130988121032715,
      "learning_rate": 9.907027584690605e-05,
      "loss": 2.609231376647949,
      "memory(GiB)": 47.63,
      "step": 7175,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.456166
    },
    {
      "epoch": 0.30761321280150805,
      "grad_norm": 4.512624740600586,
      "learning_rate": 9.906898365039027e-05,
      "loss": 2.864467239379883,
      "memory(GiB)": 47.63,
      "step": 7180,
      "token_acc": 0.43416370106761565,
      "train_speed(iter/s)": 1.45623
    },
    {
      "epoch": 0.30782742813075703,
      "grad_norm": 7.18649959564209,
      "learning_rate": 9.906769056494384e-05,
      "loss": 2.4311191558837892,
      "memory(GiB)": 47.63,
      "step": 7185,
      "token_acc": 0.4881656804733728,
      "train_speed(iter/s)": 1.456284
    },
    {
      "epoch": 0.308041643460006,
      "grad_norm": 6.2223920822143555,
      "learning_rate": 9.906639659059015e-05,
      "loss": 2.4184572219848635,
      "memory(GiB)": 47.63,
      "step": 7190,
      "token_acc": 0.4966442953020134,
      "train_speed(iter/s)": 1.456319
    },
    {
      "epoch": 0.30825585878925493,
      "grad_norm": 3.70603609085083,
      "learning_rate": 9.906510172735266e-05,
      "loss": 2.337209701538086,
      "memory(GiB)": 47.63,
      "step": 7195,
      "token_acc": 0.47129909365558914,
      "train_speed(iter/s)": 1.456077
    },
    {
      "epoch": 0.3084700741185039,
      "grad_norm": 5.951071739196777,
      "learning_rate": 9.906380597525484e-05,
      "loss": 2.6221616744995115,
      "memory(GiB)": 47.63,
      "step": 7200,
      "token_acc": 0.41338582677165353,
      "train_speed(iter/s)": 1.456293
    },
    {
      "epoch": 0.3086842894477529,
      "grad_norm": 3.1243138313293457,
      "learning_rate": 9.906250933432013e-05,
      "loss": 2.4911632537841797,
      "memory(GiB)": 47.63,
      "step": 7205,
      "token_acc": 0.45151515151515154,
      "train_speed(iter/s)": 1.456391
    },
    {
      "epoch": 0.30889850477700187,
      "grad_norm": 3.2041783332824707,
      "learning_rate": 9.906121180457204e-05,
      "loss": 2.439503860473633,
      "memory(GiB)": 47.63,
      "step": 7210,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.456254
    },
    {
      "epoch": 0.3091127201062508,
      "grad_norm": 5.2725934982299805,
      "learning_rate": 9.905991338603409e-05,
      "loss": 2.8739700317382812,
      "memory(GiB)": 47.63,
      "step": 7215,
      "token_acc": 0.43014705882352944,
      "train_speed(iter/s)": 1.456324
    },
    {
      "epoch": 0.3093269354354998,
      "grad_norm": 4.381966590881348,
      "learning_rate": 9.905861407872977e-05,
      "loss": 2.599055290222168,
      "memory(GiB)": 47.63,
      "step": 7220,
      "token_acc": 0.484,
      "train_speed(iter/s)": 1.456322
    },
    {
      "epoch": 0.30954115076474875,
      "grad_norm": 3.6992130279541016,
      "learning_rate": 9.905731388268265e-05,
      "loss": 2.710344123840332,
      "memory(GiB)": 47.63,
      "step": 7225,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.456415
    },
    {
      "epoch": 0.3097553660939977,
      "grad_norm": 5.55756139755249,
      "learning_rate": 9.905601279791626e-05,
      "loss": 2.60537052154541,
      "memory(GiB)": 47.63,
      "step": 7230,
      "token_acc": 0.49828178694158076,
      "train_speed(iter/s)": 1.45652
    },
    {
      "epoch": 0.30996958142324665,
      "grad_norm": 3.7960941791534424,
      "learning_rate": 9.905471082445419e-05,
      "loss": 2.5660427093505858,
      "memory(GiB)": 47.63,
      "step": 7235,
      "token_acc": 0.4300341296928328,
      "train_speed(iter/s)": 1.456691
    },
    {
      "epoch": 0.31018379675249563,
      "grad_norm": 4.023140907287598,
      "learning_rate": 9.905340796232e-05,
      "loss": 2.454560089111328,
      "memory(GiB)": 47.63,
      "step": 7240,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.456605
    },
    {
      "epoch": 0.31039801208174456,
      "grad_norm": 4.8561787605285645,
      "learning_rate": 9.905210421153732e-05,
      "loss": 2.681989860534668,
      "memory(GiB)": 47.63,
      "step": 7245,
      "token_acc": 0.4401294498381877,
      "train_speed(iter/s)": 1.45664
    },
    {
      "epoch": 0.31061222741099354,
      "grad_norm": 4.080799579620361,
      "learning_rate": 9.905079957212975e-05,
      "loss": 2.655870819091797,
      "memory(GiB)": 47.63,
      "step": 7250,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.456485
    },
    {
      "epoch": 0.3108264427402425,
      "grad_norm": 3.901381731033325,
      "learning_rate": 9.904949404412094e-05,
      "loss": 2.425362205505371,
      "memory(GiB)": 47.63,
      "step": 7255,
      "token_acc": 0.43666666666666665,
      "train_speed(iter/s)": 1.456531
    },
    {
      "epoch": 0.31104065806949144,
      "grad_norm": 4.594584941864014,
      "learning_rate": 9.904818762753454e-05,
      "loss": 2.3024227142333986,
      "memory(GiB)": 47.63,
      "step": 7260,
      "token_acc": 0.4896265560165975,
      "train_speed(iter/s)": 1.456676
    },
    {
      "epoch": 0.3112548733987404,
      "grad_norm": 5.4759392738342285,
      "learning_rate": 9.904688032239419e-05,
      "loss": 2.772439193725586,
      "memory(GiB)": 47.63,
      "step": 7265,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.456646
    },
    {
      "epoch": 0.3114690887279894,
      "grad_norm": 3.5301244258880615,
      "learning_rate": 9.904557212872361e-05,
      "loss": 2.629535675048828,
      "memory(GiB)": 47.63,
      "step": 7270,
      "token_acc": 0.43119266055045874,
      "train_speed(iter/s)": 1.456787
    },
    {
      "epoch": 0.3116833040572383,
      "grad_norm": 4.165886402130127,
      "learning_rate": 9.904426304654648e-05,
      "loss": 2.7619659423828127,
      "memory(GiB)": 47.63,
      "step": 7275,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.456813
    },
    {
      "epoch": 0.3118975193864873,
      "grad_norm": 4.079471111297607,
      "learning_rate": 9.904295307588651e-05,
      "loss": 2.648187446594238,
      "memory(GiB)": 47.63,
      "step": 7280,
      "token_acc": 0.4778481012658228,
      "train_speed(iter/s)": 1.456821
    },
    {
      "epoch": 0.3121117347157363,
      "grad_norm": 5.125851154327393,
      "learning_rate": 9.904164221676745e-05,
      "loss": 2.295179748535156,
      "memory(GiB)": 47.63,
      "step": 7285,
      "token_acc": 0.5087719298245614,
      "train_speed(iter/s)": 1.45704
    },
    {
      "epoch": 0.3123259500449852,
      "grad_norm": 4.464629173278809,
      "learning_rate": 9.904033046921303e-05,
      "loss": 2.5714160919189455,
      "memory(GiB)": 47.63,
      "step": 7290,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.457077
    },
    {
      "epoch": 0.3125401653742342,
      "grad_norm": 4.502624034881592,
      "learning_rate": 9.903901783324702e-05,
      "loss": 2.7460556030273438,
      "memory(GiB)": 47.63,
      "step": 7295,
      "token_acc": 0.45307443365695793,
      "train_speed(iter/s)": 1.457023
    },
    {
      "epoch": 0.31275438070348316,
      "grad_norm": 4.947858810424805,
      "learning_rate": 9.90377043088932e-05,
      "loss": 2.956756591796875,
      "memory(GiB)": 47.63,
      "step": 7300,
      "token_acc": 0.43174603174603177,
      "train_speed(iter/s)": 1.457335
    },
    {
      "epoch": 0.3129685960327321,
      "grad_norm": 4.027510166168213,
      "learning_rate": 9.903638989617537e-05,
      "loss": 2.6416173934936524,
      "memory(GiB)": 47.63,
      "step": 7305,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.457109
    },
    {
      "epoch": 0.31318281136198106,
      "grad_norm": 6.303963661193848,
      "learning_rate": 9.903507459511733e-05,
      "loss": 3.154567909240723,
      "memory(GiB)": 47.63,
      "step": 7310,
      "token_acc": 0.40217391304347827,
      "train_speed(iter/s)": 1.457338
    },
    {
      "epoch": 0.31339702669123004,
      "grad_norm": 5.36435604095459,
      "learning_rate": 9.903375840574291e-05,
      "loss": 2.681963348388672,
      "memory(GiB)": 47.63,
      "step": 7315,
      "token_acc": 0.44485294117647056,
      "train_speed(iter/s)": 1.457225
    },
    {
      "epoch": 0.31361124202047896,
      "grad_norm": 3.716139554977417,
      "learning_rate": 9.903244132807597e-05,
      "loss": 2.3914871215820312,
      "memory(GiB)": 47.63,
      "step": 7320,
      "token_acc": 0.5105633802816901,
      "train_speed(iter/s)": 1.457414
    },
    {
      "epoch": 0.31382545734972794,
      "grad_norm": 3.5286977291107178,
      "learning_rate": 9.903112336214035e-05,
      "loss": 2.6671157836914063,
      "memory(GiB)": 47.63,
      "step": 7325,
      "token_acc": 0.4407894736842105,
      "train_speed(iter/s)": 1.457429
    },
    {
      "epoch": 0.3140396726789769,
      "grad_norm": 5.585231304168701,
      "learning_rate": 9.902980450795996e-05,
      "loss": 2.8353412628173826,
      "memory(GiB)": 47.63,
      "step": 7330,
      "token_acc": 0.41823056300268097,
      "train_speed(iter/s)": 1.457398
    },
    {
      "epoch": 0.31425388800822585,
      "grad_norm": 3.8338727951049805,
      "learning_rate": 9.902848476555864e-05,
      "loss": 2.7938262939453127,
      "memory(GiB)": 47.63,
      "step": 7335,
      "token_acc": 0.39184952978056425,
      "train_speed(iter/s)": 1.457507
    },
    {
      "epoch": 0.3144681033374748,
      "grad_norm": 5.961755752563477,
      "learning_rate": 9.902716413496034e-05,
      "loss": 2.2917255401611327,
      "memory(GiB)": 47.63,
      "step": 7340,
      "token_acc": 0.47413793103448276,
      "train_speed(iter/s)": 1.457397
    },
    {
      "epoch": 0.3146823186667238,
      "grad_norm": 3.7651877403259277,
      "learning_rate": 9.902584261618896e-05,
      "loss": 2.5533000946044924,
      "memory(GiB)": 47.63,
      "step": 7345,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.45717
    },
    {
      "epoch": 0.3148965339959727,
      "grad_norm": 6.507554531097412,
      "learning_rate": 9.902452020926845e-05,
      "loss": 2.6301080703735353,
      "memory(GiB)": 47.63,
      "step": 7350,
      "token_acc": 0.43621399176954734,
      "train_speed(iter/s)": 1.457171
    },
    {
      "epoch": 0.3151107493252217,
      "grad_norm": 4.027996063232422,
      "learning_rate": 9.902319691422277e-05,
      "loss": 2.4271074295043946,
      "memory(GiB)": 47.63,
      "step": 7355,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.457035
    },
    {
      "epoch": 0.3153249646544707,
      "grad_norm": 3.8188610076904297,
      "learning_rate": 9.902187273107591e-05,
      "loss": 2.706253242492676,
      "memory(GiB)": 47.63,
      "step": 7360,
      "token_acc": 0.45874587458745875,
      "train_speed(iter/s)": 1.457046
    },
    {
      "epoch": 0.3155391799837196,
      "grad_norm": 3.748914957046509,
      "learning_rate": 9.902054765985182e-05,
      "loss": 2.91416015625,
      "memory(GiB)": 47.63,
      "step": 7365,
      "token_acc": 0.42990654205607476,
      "train_speed(iter/s)": 1.457168
    },
    {
      "epoch": 0.3157533953129686,
      "grad_norm": 3.667433023452759,
      "learning_rate": 9.901922170057452e-05,
      "loss": 2.3270805358886717,
      "memory(GiB)": 47.63,
      "step": 7370,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.45699
    },
    {
      "epoch": 0.31596761064221757,
      "grad_norm": 4.356845378875732,
      "learning_rate": 9.901789485326804e-05,
      "loss": 2.5035377502441407,
      "memory(GiB)": 47.63,
      "step": 7375,
      "token_acc": 0.4744525547445255,
      "train_speed(iter/s)": 1.457163
    },
    {
      "epoch": 0.31618182597146655,
      "grad_norm": 3.480801820755005,
      "learning_rate": 9.901656711795641e-05,
      "loss": 2.443022537231445,
      "memory(GiB)": 47.63,
      "step": 7380,
      "token_acc": 0.5050847457627119,
      "train_speed(iter/s)": 1.456968
    },
    {
      "epoch": 0.31639604130071547,
      "grad_norm": 4.483109474182129,
      "learning_rate": 9.90152384946637e-05,
      "loss": 2.6643272399902345,
      "memory(GiB)": 47.63,
      "step": 7385,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.456782
    },
    {
      "epoch": 0.31661025662996445,
      "grad_norm": 4.516879081726074,
      "learning_rate": 9.901390898341397e-05,
      "loss": 2.4153942108154296,
      "memory(GiB)": 47.63,
      "step": 7390,
      "token_acc": 0.47540983606557374,
      "train_speed(iter/s)": 1.456984
    },
    {
      "epoch": 0.3168244719592134,
      "grad_norm": 4.858016014099121,
      "learning_rate": 9.901257858423127e-05,
      "loss": 2.756536865234375,
      "memory(GiB)": 47.63,
      "step": 7395,
      "token_acc": 0.4555160142348754,
      "train_speed(iter/s)": 1.45716
    },
    {
      "epoch": 0.31703868728846235,
      "grad_norm": 4.1707844734191895,
      "learning_rate": 9.901124729713975e-05,
      "loss": 2.4957103729248047,
      "memory(GiB)": 52.0,
      "step": 7400,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.456692
    },
    {
      "epoch": 0.31725290261771133,
      "grad_norm": 3.6951959133148193,
      "learning_rate": 9.900991512216351e-05,
      "loss": 2.726665496826172,
      "memory(GiB)": 52.0,
      "step": 7405,
      "token_acc": 0.46496815286624205,
      "train_speed(iter/s)": 1.456565
    },
    {
      "epoch": 0.3174671179469603,
      "grad_norm": 3.8989763259887695,
      "learning_rate": 9.900858205932668e-05,
      "loss": 2.5814361572265625,
      "memory(GiB)": 52.0,
      "step": 7410,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.456628
    },
    {
      "epoch": 0.31768133327620923,
      "grad_norm": 5.1748785972595215,
      "learning_rate": 9.900724810865341e-05,
      "loss": 2.6261388778686525,
      "memory(GiB)": 52.0,
      "step": 7415,
      "token_acc": 0.44108761329305135,
      "train_speed(iter/s)": 1.456834
    },
    {
      "epoch": 0.3178955486054582,
      "grad_norm": 4.179869174957275,
      "learning_rate": 9.900591327016786e-05,
      "loss": 2.5476551055908203,
      "memory(GiB)": 52.0,
      "step": 7420,
      "token_acc": 0.4609053497942387,
      "train_speed(iter/s)": 1.456736
    },
    {
      "epoch": 0.3181097639347072,
      "grad_norm": 7.137809753417969,
      "learning_rate": 9.900457754389422e-05,
      "loss": 2.7810552597045897,
      "memory(GiB)": 52.0,
      "step": 7425,
      "token_acc": 0.44694533762057875,
      "train_speed(iter/s)": 1.456672
    },
    {
      "epoch": 0.3183239792639561,
      "grad_norm": 4.1653361320495605,
      "learning_rate": 9.900324092985671e-05,
      "loss": 2.253307914733887,
      "memory(GiB)": 52.0,
      "step": 7430,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.456618
    },
    {
      "epoch": 0.3185381945932051,
      "grad_norm": 3.961106538772583,
      "learning_rate": 9.900190342807951e-05,
      "loss": 2.568301963806152,
      "memory(GiB)": 52.0,
      "step": 7435,
      "token_acc": 0.44483985765124556,
      "train_speed(iter/s)": 1.456629
    },
    {
      "epoch": 0.31875240992245407,
      "grad_norm": 5.501273155212402,
      "learning_rate": 9.900056503858685e-05,
      "loss": 2.644911003112793,
      "memory(GiB)": 52.0,
      "step": 7440,
      "token_acc": 0.4618181818181818,
      "train_speed(iter/s)": 1.456846
    },
    {
      "epoch": 0.318966625251703,
      "grad_norm": 4.654253005981445,
      "learning_rate": 9.8999225761403e-05,
      "loss": 2.213212585449219,
      "memory(GiB)": 52.0,
      "step": 7445,
      "token_acc": 0.524390243902439,
      "train_speed(iter/s)": 1.456834
    },
    {
      "epoch": 0.319180840580952,
      "grad_norm": 3.2251601219177246,
      "learning_rate": 9.899788559655221e-05,
      "loss": 2.4558940887451173,
      "memory(GiB)": 52.0,
      "step": 7450,
      "token_acc": 0.5125448028673835,
      "train_speed(iter/s)": 1.456843
    },
    {
      "epoch": 0.31939505591020095,
      "grad_norm": 3.600304126739502,
      "learning_rate": 9.899654454405876e-05,
      "loss": 2.6671876907348633,
      "memory(GiB)": 52.0,
      "step": 7455,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.456869
    },
    {
      "epoch": 0.3196092712394499,
      "grad_norm": 4.009782314300537,
      "learning_rate": 9.899520260394695e-05,
      "loss": 2.7939844131469727,
      "memory(GiB)": 52.0,
      "step": 7460,
      "token_acc": 0.4354243542435424,
      "train_speed(iter/s)": 1.457085
    },
    {
      "epoch": 0.31982348656869886,
      "grad_norm": 3.5540733337402344,
      "learning_rate": 9.899385977624107e-05,
      "loss": 2.6768720626831053,
      "memory(GiB)": 52.0,
      "step": 7465,
      "token_acc": 0.4515235457063712,
      "train_speed(iter/s)": 1.45714
    },
    {
      "epoch": 0.32003770189794783,
      "grad_norm": 5.435524940490723,
      "learning_rate": 9.899251606096546e-05,
      "loss": 2.6305347442626954,
      "memory(GiB)": 52.0,
      "step": 7470,
      "token_acc": 0.4275618374558304,
      "train_speed(iter/s)": 1.457141
    },
    {
      "epoch": 0.32025191722719676,
      "grad_norm": 4.0572590827941895,
      "learning_rate": 9.899117145814448e-05,
      "loss": 2.443489646911621,
      "memory(GiB)": 52.0,
      "step": 7475,
      "token_acc": 0.45387453874538747,
      "train_speed(iter/s)": 1.457394
    },
    {
      "epoch": 0.32046613255644574,
      "grad_norm": 4.868009567260742,
      "learning_rate": 9.898982596780244e-05,
      "loss": 2.3461074829101562,
      "memory(GiB)": 52.0,
      "step": 7480,
      "token_acc": 0.4765625,
      "train_speed(iter/s)": 1.457273
    },
    {
      "epoch": 0.3206803478856947,
      "grad_norm": 3.9423940181732178,
      "learning_rate": 9.898847958996377e-05,
      "loss": 2.801871109008789,
      "memory(GiB)": 52.0,
      "step": 7485,
      "token_acc": 0.43529411764705883,
      "train_speed(iter/s)": 1.457453
    },
    {
      "epoch": 0.32089456321494364,
      "grad_norm": 8.300311088562012,
      "learning_rate": 9.898713232465283e-05,
      "loss": 2.9120140075683594,
      "memory(GiB)": 52.0,
      "step": 7490,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.45761
    },
    {
      "epoch": 0.3211087785441926,
      "grad_norm": 4.529783248901367,
      "learning_rate": 9.898578417189403e-05,
      "loss": 2.4431697845458986,
      "memory(GiB)": 52.0,
      "step": 7495,
      "token_acc": 0.4966442953020134,
      "train_speed(iter/s)": 1.457671
    },
    {
      "epoch": 0.3213229938734416,
      "grad_norm": 4.425193786621094,
      "learning_rate": 9.89844351317118e-05,
      "loss": 2.6823225021362305,
      "memory(GiB)": 52.0,
      "step": 7500,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.45785
    },
    {
      "epoch": 0.3213229938734416,
      "eval_loss": 2.142605781555176,
      "eval_runtime": 14.2023,
      "eval_samples_per_second": 7.041,
      "eval_steps_per_second": 7.041,
      "eval_token_acc": 0.46029776674937967,
      "step": 7500
    },
    {
      "epoch": 0.3215372092026905,
      "grad_norm": 4.624557971954346,
      "learning_rate": 9.89830852041306e-05,
      "loss": 2.642671012878418,
      "memory(GiB)": 52.0,
      "step": 7505,
      "token_acc": 0.461044912923923,
      "train_speed(iter/s)": 1.45373
    },
    {
      "epoch": 0.3217514245319395,
      "grad_norm": 4.579354286193848,
      "learning_rate": 9.898173438917485e-05,
      "loss": 2.476900100708008,
      "memory(GiB)": 52.0,
      "step": 7510,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.453855
    },
    {
      "epoch": 0.3219656398611885,
      "grad_norm": 3.5670883655548096,
      "learning_rate": 9.8980382686869e-05,
      "loss": 2.5919353485107424,
      "memory(GiB)": 52.0,
      "step": 7515,
      "token_acc": 0.43389830508474575,
      "train_speed(iter/s)": 1.453897
    },
    {
      "epoch": 0.3221798551904374,
      "grad_norm": 6.061057090759277,
      "learning_rate": 9.89790300972376e-05,
      "loss": 3.1762290954589845,
      "memory(GiB)": 52.0,
      "step": 7520,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.453838
    },
    {
      "epoch": 0.3223940705196864,
      "grad_norm": 4.990741729736328,
      "learning_rate": 9.897767662030512e-05,
      "loss": 2.7945816040039064,
      "memory(GiB)": 52.0,
      "step": 7525,
      "token_acc": 0.46360153256704983,
      "train_speed(iter/s)": 1.453979
    },
    {
      "epoch": 0.32260828584893536,
      "grad_norm": 3.867506742477417,
      "learning_rate": 9.897632225609607e-05,
      "loss": 2.8618717193603516,
      "memory(GiB)": 52.0,
      "step": 7530,
      "token_acc": 0.4107142857142857,
      "train_speed(iter/s)": 1.454086
    },
    {
      "epoch": 0.3228225011781843,
      "grad_norm": 3.3391501903533936,
      "learning_rate": 9.897496700463502e-05,
      "loss": 2.7338409423828125,
      "memory(GiB)": 52.0,
      "step": 7535,
      "token_acc": 0.4716417910447761,
      "train_speed(iter/s)": 1.454119
    },
    {
      "epoch": 0.32303671650743326,
      "grad_norm": 3.1852691173553467,
      "learning_rate": 9.897361086594649e-05,
      "loss": 2.7723899841308595,
      "memory(GiB)": 52.0,
      "step": 7540,
      "token_acc": 0.43490304709141275,
      "train_speed(iter/s)": 1.454214
    },
    {
      "epoch": 0.32325093183668224,
      "grad_norm": 3.827683687210083,
      "learning_rate": 9.897225384005507e-05,
      "loss": 2.6990550994873046,
      "memory(GiB)": 52.0,
      "step": 7545,
      "token_acc": 0.4304635761589404,
      "train_speed(iter/s)": 1.45433
    },
    {
      "epoch": 0.3234651471659312,
      "grad_norm": 3.6892833709716797,
      "learning_rate": 9.897089592698532e-05,
      "loss": 2.631779670715332,
      "memory(GiB)": 52.0,
      "step": 7550,
      "token_acc": 0.4479495268138801,
      "train_speed(iter/s)": 1.454572
    },
    {
      "epoch": 0.32367936249518015,
      "grad_norm": 3.134920835494995,
      "learning_rate": 9.896953712676184e-05,
      "loss": 2.7226795196533202,
      "memory(GiB)": 52.0,
      "step": 7555,
      "token_acc": 0.4253246753246753,
      "train_speed(iter/s)": 1.454461
    },
    {
      "epoch": 0.3238935778244291,
      "grad_norm": 3.281334400177002,
      "learning_rate": 9.896817743940928e-05,
      "loss": 2.4763965606689453,
      "memory(GiB)": 52.0,
      "step": 7560,
      "token_acc": 0.49146757679180886,
      "train_speed(iter/s)": 1.454641
    },
    {
      "epoch": 0.3241077931536781,
      "grad_norm": 4.33760929107666,
      "learning_rate": 9.896681686495224e-05,
      "loss": 2.911746025085449,
      "memory(GiB)": 52.0,
      "step": 7565,
      "token_acc": 0.4524714828897338,
      "train_speed(iter/s)": 1.454782
    },
    {
      "epoch": 0.324322008482927,
      "grad_norm": 4.222476959228516,
      "learning_rate": 9.896545540341538e-05,
      "loss": 2.5468311309814453,
      "memory(GiB)": 52.0,
      "step": 7570,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.454743
    },
    {
      "epoch": 0.324536223812176,
      "grad_norm": 5.119168758392334,
      "learning_rate": 9.896409305482336e-05,
      "loss": 2.677763748168945,
      "memory(GiB)": 52.0,
      "step": 7575,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.454568
    },
    {
      "epoch": 0.324750439141425,
      "grad_norm": 4.543260097503662,
      "learning_rate": 9.896272981920087e-05,
      "loss": 2.8203689575195314,
      "memory(GiB)": 52.0,
      "step": 7580,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.454539
    },
    {
      "epoch": 0.3249646544706739,
      "grad_norm": 4.372576713562012,
      "learning_rate": 9.89613656965726e-05,
      "loss": 2.743465805053711,
      "memory(GiB)": 52.0,
      "step": 7585,
      "token_acc": 0.4351145038167939,
      "train_speed(iter/s)": 1.454685
    },
    {
      "epoch": 0.3251788697999229,
      "grad_norm": 3.731536388397217,
      "learning_rate": 9.896000068696325e-05,
      "loss": 2.5834712982177734,
      "memory(GiB)": 52.0,
      "step": 7590,
      "token_acc": 0.49557522123893805,
      "train_speed(iter/s)": 1.454757
    },
    {
      "epoch": 0.32539308512917187,
      "grad_norm": 4.425579071044922,
      "learning_rate": 9.895863479039756e-05,
      "loss": 2.3880266189575194,
      "memory(GiB)": 52.0,
      "step": 7595,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.454846
    },
    {
      "epoch": 0.3256073004584208,
      "grad_norm": 3.3718419075012207,
      "learning_rate": 9.895726800690028e-05,
      "loss": 2.6899112701416015,
      "memory(GiB)": 52.0,
      "step": 7600,
      "token_acc": 0.4980694980694981,
      "train_speed(iter/s)": 1.454966
    },
    {
      "epoch": 0.32582151578766977,
      "grad_norm": 4.074090003967285,
      "learning_rate": 9.895590033649616e-05,
      "loss": 2.691311836242676,
      "memory(GiB)": 52.0,
      "step": 7605,
      "token_acc": 0.45733788395904434,
      "train_speed(iter/s)": 1.455153
    },
    {
      "epoch": 0.32603573111691875,
      "grad_norm": 4.140510082244873,
      "learning_rate": 9.895453177920997e-05,
      "loss": 2.578537368774414,
      "memory(GiB)": 52.0,
      "step": 7610,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.45514
    },
    {
      "epoch": 0.32624994644616767,
      "grad_norm": 6.439145088195801,
      "learning_rate": 9.895316233506653e-05,
      "loss": 2.5949462890625,
      "memory(GiB)": 52.0,
      "step": 7615,
      "token_acc": 0.4503311258278146,
      "train_speed(iter/s)": 1.45499
    },
    {
      "epoch": 0.32646416177541665,
      "grad_norm": 4.326587200164795,
      "learning_rate": 9.895179200409063e-05,
      "loss": 2.413687324523926,
      "memory(GiB)": 52.0,
      "step": 7620,
      "token_acc": 0.476038338658147,
      "train_speed(iter/s)": 1.454962
    },
    {
      "epoch": 0.32667837710466563,
      "grad_norm": 5.436228275299072,
      "learning_rate": 9.895042078630709e-05,
      "loss": 2.5811790466308593,
      "memory(GiB)": 52.0,
      "step": 7625,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.454725
    },
    {
      "epoch": 0.32689259243391455,
      "grad_norm": 3.5004026889801025,
      "learning_rate": 9.894904868174076e-05,
      "loss": 2.6833271026611327,
      "memory(GiB)": 52.0,
      "step": 7630,
      "token_acc": 0.4954128440366973,
      "train_speed(iter/s)": 1.454869
    },
    {
      "epoch": 0.32710680776316353,
      "grad_norm": 5.4990034103393555,
      "learning_rate": 9.89476756904165e-05,
      "loss": 2.5884525299072267,
      "memory(GiB)": 52.0,
      "step": 7635,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.454954
    },
    {
      "epoch": 0.3273210230924125,
      "grad_norm": 4.441423416137695,
      "learning_rate": 9.894630181235917e-05,
      "loss": 2.7414947509765626,
      "memory(GiB)": 52.0,
      "step": 7640,
      "token_acc": 0.4527687296416938,
      "train_speed(iter/s)": 1.455111
    },
    {
      "epoch": 0.32753523842166143,
      "grad_norm": 3.819051504135132,
      "learning_rate": 9.894492704759369e-05,
      "loss": 2.5649023056030273,
      "memory(GiB)": 52.0,
      "step": 7645,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.455221
    },
    {
      "epoch": 0.3277494537509104,
      "grad_norm": 5.052639007568359,
      "learning_rate": 9.894355139614493e-05,
      "loss": 2.536255645751953,
      "memory(GiB)": 52.0,
      "step": 7650,
      "token_acc": 0.4763779527559055,
      "train_speed(iter/s)": 1.455362
    },
    {
      "epoch": 0.3279636690801594,
      "grad_norm": 4.879740238189697,
      "learning_rate": 9.89421748580378e-05,
      "loss": 2.621837043762207,
      "memory(GiB)": 52.0,
      "step": 7655,
      "token_acc": 0.46886446886446886,
      "train_speed(iter/s)": 1.455426
    },
    {
      "epoch": 0.3281778844094083,
      "grad_norm": 7.240545272827148,
      "learning_rate": 9.894079743329729e-05,
      "loss": 2.7611183166503905,
      "memory(GiB)": 52.0,
      "step": 7660,
      "token_acc": 0.4370629370629371,
      "train_speed(iter/s)": 1.455601
    },
    {
      "epoch": 0.3283920997386573,
      "grad_norm": 3.6262354850769043,
      "learning_rate": 9.893941912194831e-05,
      "loss": 2.543458938598633,
      "memory(GiB)": 52.0,
      "step": 7665,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.455476
    },
    {
      "epoch": 0.3286063150679063,
      "grad_norm": 2.893031120300293,
      "learning_rate": 9.893803992401586e-05,
      "loss": 2.453871154785156,
      "memory(GiB)": 52.0,
      "step": 7670,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.455521
    },
    {
      "epoch": 0.3288205303971552,
      "grad_norm": 3.959601879119873,
      "learning_rate": 9.893665983952489e-05,
      "loss": 2.8583328247070314,
      "memory(GiB)": 52.0,
      "step": 7675,
      "token_acc": 0.402555910543131,
      "train_speed(iter/s)": 1.455505
    },
    {
      "epoch": 0.3290347457264042,
      "grad_norm": 3.764578104019165,
      "learning_rate": 9.893527886850044e-05,
      "loss": 2.5479013442993166,
      "memory(GiB)": 52.0,
      "step": 7680,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.455537
    },
    {
      "epoch": 0.32924896105565316,
      "grad_norm": 6.591464042663574,
      "learning_rate": 9.89338970109675e-05,
      "loss": 2.2869140625,
      "memory(GiB)": 52.0,
      "step": 7685,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.455264
    },
    {
      "epoch": 0.3294631763849021,
      "grad_norm": 4.059225082397461,
      "learning_rate": 9.893251426695111e-05,
      "loss": 2.933719444274902,
      "memory(GiB)": 52.0,
      "step": 7690,
      "token_acc": 0.4041916167664671,
      "train_speed(iter/s)": 1.455251
    },
    {
      "epoch": 0.32967739171415106,
      "grad_norm": 4.148004055023193,
      "learning_rate": 9.893113063647632e-05,
      "loss": 2.7961151123046877,
      "memory(GiB)": 52.0,
      "step": 7695,
      "token_acc": 0.4793388429752066,
      "train_speed(iter/s)": 1.455352
    },
    {
      "epoch": 0.32989160704340004,
      "grad_norm": 4.07358980178833,
      "learning_rate": 9.89297461195682e-05,
      "loss": 2.361439323425293,
      "memory(GiB)": 58.3,
      "step": 7700,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.455065
    },
    {
      "epoch": 0.33010582237264896,
      "grad_norm": 3.9765727519989014,
      "learning_rate": 9.892836071625182e-05,
      "loss": 2.282497787475586,
      "memory(GiB)": 58.3,
      "step": 7705,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.454959
    },
    {
      "epoch": 0.33032003770189794,
      "grad_norm": 3.9762468338012695,
      "learning_rate": 9.89269744265523e-05,
      "loss": 2.883592414855957,
      "memory(GiB)": 58.3,
      "step": 7710,
      "token_acc": 0.42443729903536975,
      "train_speed(iter/s)": 1.455027
    },
    {
      "epoch": 0.3305342530311469,
      "grad_norm": 4.177420139312744,
      "learning_rate": 9.892558725049474e-05,
      "loss": 2.6837316513061524,
      "memory(GiB)": 58.3,
      "step": 7715,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.455137
    },
    {
      "epoch": 0.3307484683603959,
      "grad_norm": 3.9853692054748535,
      "learning_rate": 9.892419918810426e-05,
      "loss": 2.7274417877197266,
      "memory(GiB)": 58.3,
      "step": 7720,
      "token_acc": 0.43214285714285716,
      "train_speed(iter/s)": 1.455245
    },
    {
      "epoch": 0.3309626836896448,
      "grad_norm": 5.162968158721924,
      "learning_rate": 9.892281023940602e-05,
      "loss": 2.5281951904296873,
      "memory(GiB)": 58.3,
      "step": 7725,
      "token_acc": 0.44727272727272727,
      "train_speed(iter/s)": 1.455536
    },
    {
      "epoch": 0.3311768990188938,
      "grad_norm": 3.5378823280334473,
      "learning_rate": 9.892142040442518e-05,
      "loss": 2.6750280380249025,
      "memory(GiB)": 58.3,
      "step": 7730,
      "token_acc": 0.47337278106508873,
      "train_speed(iter/s)": 1.455346
    },
    {
      "epoch": 0.3313911143481428,
      "grad_norm": 5.67416524887085,
      "learning_rate": 9.892002968318692e-05,
      "loss": 2.2620065689086912,
      "memory(GiB)": 58.3,
      "step": 7735,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.455051
    },
    {
      "epoch": 0.3316053296773917,
      "grad_norm": 6.915927410125732,
      "learning_rate": 9.891863807571644e-05,
      "loss": 2.4981914520263673,
      "memory(GiB)": 58.3,
      "step": 7740,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.455187
    },
    {
      "epoch": 0.3318195450066407,
      "grad_norm": 3.4285967350006104,
      "learning_rate": 9.891724558203893e-05,
      "loss": 2.540232276916504,
      "memory(GiB)": 58.3,
      "step": 7745,
      "token_acc": 0.44654088050314467,
      "train_speed(iter/s)": 1.455221
    },
    {
      "epoch": 0.33203376033588966,
      "grad_norm": 3.635664224624634,
      "learning_rate": 9.891585220217964e-05,
      "loss": 2.550495147705078,
      "memory(GiB)": 58.3,
      "step": 7750,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.455151
    },
    {
      "epoch": 0.3322479756651386,
      "grad_norm": 5.062612056732178,
      "learning_rate": 9.891445793616378e-05,
      "loss": 2.6595279693603517,
      "memory(GiB)": 58.3,
      "step": 7755,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.454911
    },
    {
      "epoch": 0.33246219099438756,
      "grad_norm": 3.8729732036590576,
      "learning_rate": 9.891306278401665e-05,
      "loss": 2.349718475341797,
      "memory(GiB)": 58.3,
      "step": 7760,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.454786
    },
    {
      "epoch": 0.33267640632363654,
      "grad_norm": 4.594541549682617,
      "learning_rate": 9.891166674576349e-05,
      "loss": 2.9694677352905274,
      "memory(GiB)": 58.3,
      "step": 7765,
      "token_acc": 0.4006734006734007,
      "train_speed(iter/s)": 1.454485
    },
    {
      "epoch": 0.33289062165288547,
      "grad_norm": 3.407460927963257,
      "learning_rate": 9.891026982142962e-05,
      "loss": 2.890412139892578,
      "memory(GiB)": 58.3,
      "step": 7770,
      "token_acc": 0.3989071038251366,
      "train_speed(iter/s)": 1.454381
    },
    {
      "epoch": 0.33310483698213444,
      "grad_norm": 4.638185501098633,
      "learning_rate": 9.890887201104032e-05,
      "loss": 2.8380165100097656,
      "memory(GiB)": 58.3,
      "step": 7775,
      "token_acc": 0.45224719101123595,
      "train_speed(iter/s)": 1.454308
    },
    {
      "epoch": 0.3333190523113834,
      "grad_norm": 4.219641208648682,
      "learning_rate": 9.890747331462092e-05,
      "loss": 2.702608108520508,
      "memory(GiB)": 58.3,
      "step": 7780,
      "token_acc": 0.428125,
      "train_speed(iter/s)": 1.454617
    },
    {
      "epoch": 0.33353326764063235,
      "grad_norm": 3.846561908721924,
      "learning_rate": 9.890607373219676e-05,
      "loss": 2.5522266387939454,
      "memory(GiB)": 58.3,
      "step": 7785,
      "token_acc": 0.4819277108433735,
      "train_speed(iter/s)": 1.454803
    },
    {
      "epoch": 0.3337474829698813,
      "grad_norm": 4.175077438354492,
      "learning_rate": 9.89046732637932e-05,
      "loss": 2.759039878845215,
      "memory(GiB)": 58.3,
      "step": 7790,
      "token_acc": 0.4370629370629371,
      "train_speed(iter/s)": 1.45498
    },
    {
      "epoch": 0.3339616982991303,
      "grad_norm": 3.8363518714904785,
      "learning_rate": 9.890327190943561e-05,
      "loss": 2.6114875793457033,
      "memory(GiB)": 58.3,
      "step": 7795,
      "token_acc": 0.4537037037037037,
      "train_speed(iter/s)": 1.454938
    },
    {
      "epoch": 0.33417591362837923,
      "grad_norm": 3.5491459369659424,
      "learning_rate": 9.890186966914938e-05,
      "loss": 2.532672882080078,
      "memory(GiB)": 58.3,
      "step": 7800,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.455006
    },
    {
      "epoch": 0.3343901289576282,
      "grad_norm": 3.699162244796753,
      "learning_rate": 9.89004665429599e-05,
      "loss": 2.1899948120117188,
      "memory(GiB)": 58.3,
      "step": 7805,
      "token_acc": 0.5341880341880342,
      "train_speed(iter/s)": 1.454635
    },
    {
      "epoch": 0.3346043442868772,
      "grad_norm": 3.653193950653076,
      "learning_rate": 9.88990625308926e-05,
      "loss": 2.282488250732422,
      "memory(GiB)": 58.3,
      "step": 7810,
      "token_acc": 0.5290322580645161,
      "train_speed(iter/s)": 1.454536
    },
    {
      "epoch": 0.3348185596161261,
      "grad_norm": 4.398053169250488,
      "learning_rate": 9.889765763297291e-05,
      "loss": 2.5617956161499023,
      "memory(GiB)": 58.3,
      "step": 7815,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.454735
    },
    {
      "epoch": 0.3350327749453751,
      "grad_norm": 4.201615333557129,
      "learning_rate": 9.889625184922628e-05,
      "loss": 2.499456787109375,
      "memory(GiB)": 58.3,
      "step": 7820,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.454658
    },
    {
      "epoch": 0.33524699027462407,
      "grad_norm": 3.0528340339660645,
      "learning_rate": 9.889484517967818e-05,
      "loss": 2.5310670852661135,
      "memory(GiB)": 58.3,
      "step": 7825,
      "token_acc": 0.49117647058823527,
      "train_speed(iter/s)": 1.454819
    },
    {
      "epoch": 0.335461205603873,
      "grad_norm": 6.417255401611328,
      "learning_rate": 9.889343762435409e-05,
      "loss": 2.919891929626465,
      "memory(GiB)": 58.3,
      "step": 7830,
      "token_acc": 0.43564356435643564,
      "train_speed(iter/s)": 1.454859
    },
    {
      "epoch": 0.33567542093312197,
      "grad_norm": 4.355087757110596,
      "learning_rate": 9.88920291832795e-05,
      "loss": 2.7533243179321287,
      "memory(GiB)": 58.3,
      "step": 7835,
      "token_acc": 0.41033434650455924,
      "train_speed(iter/s)": 1.454752
    },
    {
      "epoch": 0.33588963626237095,
      "grad_norm": 3.2876315116882324,
      "learning_rate": 9.889061985647996e-05,
      "loss": 2.9273612976074217,
      "memory(GiB)": 58.3,
      "step": 7840,
      "token_acc": 0.40540540540540543,
      "train_speed(iter/s)": 1.454844
    },
    {
      "epoch": 0.3361038515916199,
      "grad_norm": 4.384795665740967,
      "learning_rate": 9.888920964398099e-05,
      "loss": 2.6128000259399413,
      "memory(GiB)": 58.3,
      "step": 7845,
      "token_acc": 0.45733788395904434,
      "train_speed(iter/s)": 1.455004
    },
    {
      "epoch": 0.33631806692086885,
      "grad_norm": 3.9293158054351807,
      "learning_rate": 9.88877985458081e-05,
      "loss": 2.5912336349487304,
      "memory(GiB)": 58.3,
      "step": 7850,
      "token_acc": 0.4753521126760563,
      "train_speed(iter/s)": 1.455202
    },
    {
      "epoch": 0.33653228225011783,
      "grad_norm": 3.4973764419555664,
      "learning_rate": 9.888638656198688e-05,
      "loss": 2.6211244583129885,
      "memory(GiB)": 58.3,
      "step": 7855,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.455112
    },
    {
      "epoch": 0.33674649757936675,
      "grad_norm": 3.9123497009277344,
      "learning_rate": 9.88849736925429e-05,
      "loss": 2.5589523315429688,
      "memory(GiB)": 58.3,
      "step": 7860,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.455165
    },
    {
      "epoch": 0.33696071290861573,
      "grad_norm": 4.121347904205322,
      "learning_rate": 9.888355993750178e-05,
      "loss": 2.5394596099853515,
      "memory(GiB)": 58.3,
      "step": 7865,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.455397
    },
    {
      "epoch": 0.3371749282378647,
      "grad_norm": 5.938421726226807,
      "learning_rate": 9.888214529688912e-05,
      "loss": 2.6497512817382813,
      "memory(GiB)": 58.3,
      "step": 7870,
      "token_acc": 0.4609053497942387,
      "train_speed(iter/s)": 1.455437
    },
    {
      "epoch": 0.33738914356711364,
      "grad_norm": 3.8449952602386475,
      "learning_rate": 9.888072977073053e-05,
      "loss": 2.4433937072753906,
      "memory(GiB)": 58.3,
      "step": 7875,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.455498
    },
    {
      "epoch": 0.3376033588963626,
      "grad_norm": 3.967132806777954,
      "learning_rate": 9.887931335905168e-05,
      "loss": 2.3723243713378905,
      "memory(GiB)": 58.3,
      "step": 7880,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.455701
    },
    {
      "epoch": 0.3378175742256116,
      "grad_norm": 3.0878348350524902,
      "learning_rate": 9.887789606187819e-05,
      "loss": 2.5916696548461915,
      "memory(GiB)": 58.3,
      "step": 7885,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.455677
    },
    {
      "epoch": 0.3380317895548606,
      "grad_norm": 5.347700119018555,
      "learning_rate": 9.887647787923578e-05,
      "loss": 2.7415653228759767,
      "memory(GiB)": 58.3,
      "step": 7890,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.455828
    },
    {
      "epoch": 0.3382460048841095,
      "grad_norm": 3.3523037433624268,
      "learning_rate": 9.887505881115013e-05,
      "loss": 2.3701921463012696,
      "memory(GiB)": 58.3,
      "step": 7895,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.456003
    },
    {
      "epoch": 0.3384602202133585,
      "grad_norm": 3.8140828609466553,
      "learning_rate": 9.887363885764693e-05,
      "loss": 2.7807552337646486,
      "memory(GiB)": 58.3,
      "step": 7900,
      "token_acc": 0.4296028880866426,
      "train_speed(iter/s)": 1.455988
    },
    {
      "epoch": 0.33867443554260745,
      "grad_norm": 5.0800700187683105,
      "learning_rate": 9.887221801875192e-05,
      "loss": 2.481638717651367,
      "memory(GiB)": 58.3,
      "step": 7905,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.455828
    },
    {
      "epoch": 0.3388886508718564,
      "grad_norm": 4.114391326904297,
      "learning_rate": 9.887079629449083e-05,
      "loss": 2.49346923828125,
      "memory(GiB)": 58.3,
      "step": 7910,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.455905
    },
    {
      "epoch": 0.33910286620110536,
      "grad_norm": 4.894964218139648,
      "learning_rate": 9.886937368488942e-05,
      "loss": 2.7256725311279295,
      "memory(GiB)": 58.3,
      "step": 7915,
      "token_acc": 0.4205607476635514,
      "train_speed(iter/s)": 1.455999
    },
    {
      "epoch": 0.33931708153035434,
      "grad_norm": 4.770758628845215,
      "learning_rate": 9.886795018997347e-05,
      "loss": 2.708760643005371,
      "memory(GiB)": 58.3,
      "step": 7920,
      "token_acc": 0.426056338028169,
      "train_speed(iter/s)": 1.456035
    },
    {
      "epoch": 0.33953129685960326,
      "grad_norm": 2.974980354309082,
      "learning_rate": 9.886652580976876e-05,
      "loss": 2.669449806213379,
      "memory(GiB)": 58.3,
      "step": 7925,
      "token_acc": 0.4575645756457565,
      "train_speed(iter/s)": 1.456164
    },
    {
      "epoch": 0.33974551218885224,
      "grad_norm": 3.535210132598877,
      "learning_rate": 9.886510054430108e-05,
      "loss": 2.564958953857422,
      "memory(GiB)": 58.3,
      "step": 7930,
      "token_acc": 0.4389438943894389,
      "train_speed(iter/s)": 1.456303
    },
    {
      "epoch": 0.3399597275181012,
      "grad_norm": 6.0368733406066895,
      "learning_rate": 9.886367439359627e-05,
      "loss": 2.9167285919189454,
      "memory(GiB)": 58.3,
      "step": 7935,
      "token_acc": 0.42585551330798477,
      "train_speed(iter/s)": 1.456396
    },
    {
      "epoch": 0.34017394284735014,
      "grad_norm": 3.674422025680542,
      "learning_rate": 9.886224735768017e-05,
      "loss": 2.611805725097656,
      "memory(GiB)": 58.3,
      "step": 7940,
      "token_acc": 0.4486301369863014,
      "train_speed(iter/s)": 1.456485
    },
    {
      "epoch": 0.3403881581765991,
      "grad_norm": 5.602137565612793,
      "learning_rate": 9.886081943657862e-05,
      "loss": 2.7226539611816407,
      "memory(GiB)": 58.3,
      "step": 7945,
      "token_acc": 0.4490566037735849,
      "train_speed(iter/s)": 1.456601
    },
    {
      "epoch": 0.3406023735058481,
      "grad_norm": 5.106326580047607,
      "learning_rate": 9.885939063031748e-05,
      "loss": 2.40622615814209,
      "memory(GiB)": 58.3,
      "step": 7950,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.456585
    },
    {
      "epoch": 0.340816588835097,
      "grad_norm": 4.052453994750977,
      "learning_rate": 9.885796093892266e-05,
      "loss": 2.5365081787109376,
      "memory(GiB)": 58.3,
      "step": 7955,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.456634
    },
    {
      "epoch": 0.341030804164346,
      "grad_norm": 5.224856853485107,
      "learning_rate": 9.885653036242004e-05,
      "loss": 2.4870349884033205,
      "memory(GiB)": 58.3,
      "step": 7960,
      "token_acc": 0.4939759036144578,
      "train_speed(iter/s)": 1.456684
    },
    {
      "epoch": 0.341245019493595,
      "grad_norm": 4.263911724090576,
      "learning_rate": 9.885509890083555e-05,
      "loss": 2.5133150100708006,
      "memory(GiB)": 58.3,
      "step": 7965,
      "token_acc": 0.5038461538461538,
      "train_speed(iter/s)": 1.456588
    },
    {
      "epoch": 0.3414592348228439,
      "grad_norm": 3.6719298362731934,
      "learning_rate": 9.88536665541951e-05,
      "loss": 2.6517005920410157,
      "memory(GiB)": 58.3,
      "step": 7970,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.456716
    },
    {
      "epoch": 0.3416734501520929,
      "grad_norm": 3.5599942207336426,
      "learning_rate": 9.885223332252464e-05,
      "loss": 2.5277809143066405,
      "memory(GiB)": 58.3,
      "step": 7975,
      "token_acc": 0.44089456869009586,
      "train_speed(iter/s)": 1.456793
    },
    {
      "epoch": 0.34188766548134186,
      "grad_norm": 3.620069742202759,
      "learning_rate": 9.885079920585017e-05,
      "loss": 2.6460525512695314,
      "memory(GiB)": 58.3,
      "step": 7980,
      "token_acc": 0.45686900958466453,
      "train_speed(iter/s)": 1.456853
    },
    {
      "epoch": 0.3421018808105908,
      "grad_norm": 3.9305548667907715,
      "learning_rate": 9.884936420419763e-05,
      "loss": 2.7537521362304687,
      "memory(GiB)": 58.3,
      "step": 7985,
      "token_acc": 0.44983818770226536,
      "train_speed(iter/s)": 1.456874
    },
    {
      "epoch": 0.34231609613983977,
      "grad_norm": 5.006189346313477,
      "learning_rate": 9.884792831759305e-05,
      "loss": 2.5016101837158202,
      "memory(GiB)": 58.3,
      "step": 7990,
      "token_acc": 0.45934959349593496,
      "train_speed(iter/s)": 1.45684
    },
    {
      "epoch": 0.34253031146908874,
      "grad_norm": 4.487048149108887,
      "learning_rate": 9.884649154606242e-05,
      "loss": 2.4803647994995117,
      "memory(GiB)": 58.3,
      "step": 7995,
      "token_acc": 0.4472843450479233,
      "train_speed(iter/s)": 1.456952
    },
    {
      "epoch": 0.34274452679833767,
      "grad_norm": 3.460939645767212,
      "learning_rate": 9.884505388963176e-05,
      "loss": 2.5205123901367186,
      "memory(GiB)": 58.3,
      "step": 8000,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.456872
    },
    {
      "epoch": 0.34274452679833767,
      "eval_loss": 2.3589062690734863,
      "eval_runtime": 14.3657,
      "eval_samples_per_second": 6.961,
      "eval_steps_per_second": 6.961,
      "eval_token_acc": 0.47733333333333333,
      "step": 8000
    },
    {
      "epoch": 0.34295874212758665,
      "grad_norm": 3.464186191558838,
      "learning_rate": 9.884361534832716e-05,
      "loss": 2.453921890258789,
      "memory(GiB)": 58.3,
      "step": 8005,
      "token_acc": 0.47714285714285715,
      "train_speed(iter/s)": 1.452755
    },
    {
      "epoch": 0.3431729574568356,
      "grad_norm": 4.4978413581848145,
      "learning_rate": 9.884217592217461e-05,
      "loss": 2.5027889251708983,
      "memory(GiB)": 58.3,
      "step": 8010,
      "token_acc": 0.43962848297213625,
      "train_speed(iter/s)": 1.452906
    },
    {
      "epoch": 0.34338717278608455,
      "grad_norm": 4.614760875701904,
      "learning_rate": 9.884073561120026e-05,
      "loss": 2.2955766677856446,
      "memory(GiB)": 58.3,
      "step": 8015,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.452973
    },
    {
      "epoch": 0.34360138811533353,
      "grad_norm": 3.8762171268463135,
      "learning_rate": 9.883929441543014e-05,
      "loss": 2.495701217651367,
      "memory(GiB)": 58.3,
      "step": 8020,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.452594
    },
    {
      "epoch": 0.3438156034445825,
      "grad_norm": 4.507558345794678,
      "learning_rate": 9.88378523348904e-05,
      "loss": 2.6824146270751954,
      "memory(GiB)": 58.3,
      "step": 8025,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.452281
    },
    {
      "epoch": 0.34402981877383143,
      "grad_norm": 3.8340647220611572,
      "learning_rate": 9.883640936960716e-05,
      "loss": 2.4789306640625,
      "memory(GiB)": 58.3,
      "step": 8030,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.452378
    },
    {
      "epoch": 0.3442440341030804,
      "grad_norm": 4.548669815063477,
      "learning_rate": 9.883496551960654e-05,
      "loss": 2.6601104736328125,
      "memory(GiB)": 58.3,
      "step": 8035,
      "token_acc": 0.45686900958466453,
      "train_speed(iter/s)": 1.45245
    },
    {
      "epoch": 0.3444582494323294,
      "grad_norm": 3.9123921394348145,
      "learning_rate": 9.88335207849147e-05,
      "loss": 2.433611297607422,
      "memory(GiB)": 58.3,
      "step": 8040,
      "token_acc": 0.49523809523809526,
      "train_speed(iter/s)": 1.45224
    },
    {
      "epoch": 0.3446724647615783,
      "grad_norm": 4.004251003265381,
      "learning_rate": 9.883207516555784e-05,
      "loss": 2.5678157806396484,
      "memory(GiB)": 58.3,
      "step": 8045,
      "token_acc": 0.43234323432343236,
      "train_speed(iter/s)": 1.452301
    },
    {
      "epoch": 0.3448866800908273,
      "grad_norm": 4.088587760925293,
      "learning_rate": 9.883062866156213e-05,
      "loss": 2.7233516693115236,
      "memory(GiB)": 58.3,
      "step": 8050,
      "token_acc": 0.4301369863013699,
      "train_speed(iter/s)": 1.452181
    },
    {
      "epoch": 0.34510089542007627,
      "grad_norm": 5.49698543548584,
      "learning_rate": 9.882918127295376e-05,
      "loss": 2.473859405517578,
      "memory(GiB)": 58.3,
      "step": 8055,
      "token_acc": 0.4405594405594406,
      "train_speed(iter/s)": 1.452229
    },
    {
      "epoch": 0.34531511074932525,
      "grad_norm": 4.037271499633789,
      "learning_rate": 9.882773299975897e-05,
      "loss": 3.2395488739013674,
      "memory(GiB)": 58.3,
      "step": 8060,
      "token_acc": 0.4034090909090909,
      "train_speed(iter/s)": 1.452279
    },
    {
      "epoch": 0.3455293260785742,
      "grad_norm": 3.700503349304199,
      "learning_rate": 9.8826283842004e-05,
      "loss": 2.470033073425293,
      "memory(GiB)": 58.3,
      "step": 8065,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.452196
    },
    {
      "epoch": 0.34574354140782315,
      "grad_norm": 3.711183547973633,
      "learning_rate": 9.882483379971509e-05,
      "loss": 2.552614212036133,
      "memory(GiB)": 58.3,
      "step": 8070,
      "token_acc": 0.48501362397820164,
      "train_speed(iter/s)": 1.452313
    },
    {
      "epoch": 0.34595775673707213,
      "grad_norm": 3.106621026992798,
      "learning_rate": 9.882338287291851e-05,
      "loss": 2.7338134765625,
      "memory(GiB)": 58.3,
      "step": 8075,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.452415
    },
    {
      "epoch": 0.34617197206632105,
      "grad_norm": 4.303894519805908,
      "learning_rate": 9.882193106164055e-05,
      "loss": 3.065963554382324,
      "memory(GiB)": 58.3,
      "step": 8080,
      "token_acc": 0.3867403314917127,
      "train_speed(iter/s)": 1.45252
    },
    {
      "epoch": 0.34638618739557003,
      "grad_norm": 4.250564098358154,
      "learning_rate": 9.882047836590752e-05,
      "loss": 2.4909812927246096,
      "memory(GiB)": 58.3,
      "step": 8085,
      "token_acc": 0.4895833333333333,
      "train_speed(iter/s)": 1.452694
    },
    {
      "epoch": 0.346600402724819,
      "grad_norm": 3.559344530105591,
      "learning_rate": 9.881902478574571e-05,
      "loss": 2.9835681915283203,
      "memory(GiB)": 58.3,
      "step": 8090,
      "token_acc": 0.40625,
      "train_speed(iter/s)": 1.452842
    },
    {
      "epoch": 0.34681461805406794,
      "grad_norm": 3.629955768585205,
      "learning_rate": 9.88175703211815e-05,
      "loss": 2.6201688766479494,
      "memory(GiB)": 58.3,
      "step": 8095,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.452799
    },
    {
      "epoch": 0.3470288333833169,
      "grad_norm": 3.5932557582855225,
      "learning_rate": 9.88161149722412e-05,
      "loss": 2.4452743530273438,
      "memory(GiB)": 58.3,
      "step": 8100,
      "token_acc": 0.47416413373860183,
      "train_speed(iter/s)": 1.452743
    },
    {
      "epoch": 0.3472430487125659,
      "grad_norm": 3.6034882068634033,
      "learning_rate": 9.881465873895116e-05,
      "loss": 3.03234977722168,
      "memory(GiB)": 58.3,
      "step": 8105,
      "token_acc": 0.4391891891891892,
      "train_speed(iter/s)": 1.452636
    },
    {
      "epoch": 0.3474572640418148,
      "grad_norm": 5.087571620941162,
      "learning_rate": 9.881320162133781e-05,
      "loss": 2.5947038650512697,
      "memory(GiB)": 58.3,
      "step": 8110,
      "token_acc": 0.4332129963898917,
      "train_speed(iter/s)": 1.452818
    },
    {
      "epoch": 0.3476714793710638,
      "grad_norm": 3.706878423690796,
      "learning_rate": 9.881174361942751e-05,
      "loss": 2.3738880157470703,
      "memory(GiB)": 58.3,
      "step": 8115,
      "token_acc": 0.5370370370370371,
      "train_speed(iter/s)": 1.452831
    },
    {
      "epoch": 0.3478856947003128,
      "grad_norm": 3.635059356689453,
      "learning_rate": 9.881028473324669e-05,
      "loss": 2.608393096923828,
      "memory(GiB)": 58.3,
      "step": 8120,
      "token_acc": 0.4006514657980456,
      "train_speed(iter/s)": 1.452928
    },
    {
      "epoch": 0.3480999100295617,
      "grad_norm": 3.160557985305786,
      "learning_rate": 9.880882496282176e-05,
      "loss": 2.796361541748047,
      "memory(GiB)": 58.3,
      "step": 8125,
      "token_acc": 0.445141065830721,
      "train_speed(iter/s)": 1.453067
    },
    {
      "epoch": 0.3483141253588107,
      "grad_norm": 4.458441257476807,
      "learning_rate": 9.88073643081792e-05,
      "loss": 2.282944679260254,
      "memory(GiB)": 58.3,
      "step": 8130,
      "token_acc": 0.5196850393700787,
      "train_speed(iter/s)": 1.453111
    },
    {
      "epoch": 0.34852834068805966,
      "grad_norm": 4.312187671661377,
      "learning_rate": 9.880590276934543e-05,
      "loss": 2.600994873046875,
      "memory(GiB)": 58.3,
      "step": 8135,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.453122
    },
    {
      "epoch": 0.3487425560173086,
      "grad_norm": 3.5220563411712646,
      "learning_rate": 9.880444034634698e-05,
      "loss": 2.680997848510742,
      "memory(GiB)": 58.3,
      "step": 8140,
      "token_acc": 0.43661971830985913,
      "train_speed(iter/s)": 1.453279
    },
    {
      "epoch": 0.34895677134655756,
      "grad_norm": 3.735930919647217,
      "learning_rate": 9.880297703921027e-05,
      "loss": 2.503192901611328,
      "memory(GiB)": 58.3,
      "step": 8145,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.453363
    },
    {
      "epoch": 0.34917098667580654,
      "grad_norm": 5.526791572570801,
      "learning_rate": 9.880151284796187e-05,
      "loss": 2.8281320571899413,
      "memory(GiB)": 58.3,
      "step": 8150,
      "token_acc": 0.40794223826714804,
      "train_speed(iter/s)": 1.453434
    },
    {
      "epoch": 0.34938520200505546,
      "grad_norm": 6.912009239196777,
      "learning_rate": 9.880004777262829e-05,
      "loss": 2.6142414093017576,
      "memory(GiB)": 58.3,
      "step": 8155,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.453425
    },
    {
      "epoch": 0.34959941733430444,
      "grad_norm": 4.234181880950928,
      "learning_rate": 9.879858181323607e-05,
      "loss": 2.4705215454101563,
      "memory(GiB)": 58.3,
      "step": 8160,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.453481
    },
    {
      "epoch": 0.3498136326635534,
      "grad_norm": 4.864664554595947,
      "learning_rate": 9.879711496981174e-05,
      "loss": 2.624741554260254,
      "memory(GiB)": 58.3,
      "step": 8165,
      "token_acc": 0.46715328467153283,
      "train_speed(iter/s)": 1.453658
    },
    {
      "epoch": 0.35002784799280234,
      "grad_norm": 6.615359306335449,
      "learning_rate": 9.879564724238193e-05,
      "loss": 2.94897403717041,
      "memory(GiB)": 58.3,
      "step": 8170,
      "token_acc": 0.3763440860215054,
      "train_speed(iter/s)": 1.45379
    },
    {
      "epoch": 0.3502420633220513,
      "grad_norm": 3.722355365753174,
      "learning_rate": 9.879417863097318e-05,
      "loss": 2.5865571975708006,
      "memory(GiB)": 58.3,
      "step": 8175,
      "token_acc": 0.4351851851851852,
      "train_speed(iter/s)": 1.453813
    },
    {
      "epoch": 0.3504562786513003,
      "grad_norm": 4.878786563873291,
      "learning_rate": 9.879270913561209e-05,
      "loss": 2.4978450775146483,
      "memory(GiB)": 58.3,
      "step": 8180,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.453702
    },
    {
      "epoch": 0.3506704939805492,
      "grad_norm": 4.959492206573486,
      "learning_rate": 9.879123875632534e-05,
      "loss": 2.77236385345459,
      "memory(GiB)": 58.3,
      "step": 8185,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.453822
    },
    {
      "epoch": 0.3508847093097982,
      "grad_norm": 4.195497512817383,
      "learning_rate": 9.878976749313951e-05,
      "loss": 2.7344221115112304,
      "memory(GiB)": 58.3,
      "step": 8190,
      "token_acc": 0.4658385093167702,
      "train_speed(iter/s)": 1.453561
    },
    {
      "epoch": 0.3510989246390472,
      "grad_norm": 4.9788947105407715,
      "learning_rate": 9.878829534608127e-05,
      "loss": 2.6990888595581053,
      "memory(GiB)": 58.3,
      "step": 8195,
      "token_acc": 0.4493927125506073,
      "train_speed(iter/s)": 1.45361
    },
    {
      "epoch": 0.3513131399682961,
      "grad_norm": 4.959716796875,
      "learning_rate": 9.878682231517731e-05,
      "loss": 2.3707590103149414,
      "memory(GiB)": 58.3,
      "step": 8200,
      "token_acc": 0.5167173252279635,
      "train_speed(iter/s)": 1.453289
    },
    {
      "epoch": 0.3515273552975451,
      "grad_norm": 3.226439952850342,
      "learning_rate": 9.878534840045428e-05,
      "loss": 2.610869789123535,
      "memory(GiB)": 58.3,
      "step": 8205,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.453178
    },
    {
      "epoch": 0.35174157062679406,
      "grad_norm": 3.548597574234009,
      "learning_rate": 9.878387360193891e-05,
      "loss": 2.5473773956298826,
      "memory(GiB)": 58.3,
      "step": 8210,
      "token_acc": 0.4512987012987013,
      "train_speed(iter/s)": 1.453252
    },
    {
      "epoch": 0.351955785956043,
      "grad_norm": 6.055249214172363,
      "learning_rate": 9.87823979196579e-05,
      "loss": 2.609727478027344,
      "memory(GiB)": 58.3,
      "step": 8215,
      "token_acc": 0.45041322314049587,
      "train_speed(iter/s)": 1.453461
    },
    {
      "epoch": 0.35217000128529197,
      "grad_norm": 3.633150100708008,
      "learning_rate": 9.8780921353638e-05,
      "loss": 2.622083282470703,
      "memory(GiB)": 58.3,
      "step": 8220,
      "token_acc": 0.4603658536585366,
      "train_speed(iter/s)": 1.453509
    },
    {
      "epoch": 0.35238421661454095,
      "grad_norm": 4.759337902069092,
      "learning_rate": 9.877944390390594e-05,
      "loss": 2.794211769104004,
      "memory(GiB)": 58.3,
      "step": 8225,
      "token_acc": 0.4139194139194139,
      "train_speed(iter/s)": 1.453525
    },
    {
      "epoch": 0.3525984319437899,
      "grad_norm": 4.9054131507873535,
      "learning_rate": 9.87779655704885e-05,
      "loss": 2.510045051574707,
      "memory(GiB)": 58.3,
      "step": 8230,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.453481
    },
    {
      "epoch": 0.35281264727303885,
      "grad_norm": 3.7385787963867188,
      "learning_rate": 9.877648635341245e-05,
      "loss": 2.4643476486206053,
      "memory(GiB)": 58.3,
      "step": 8235,
      "token_acc": 0.44727272727272727,
      "train_speed(iter/s)": 1.453471
    },
    {
      "epoch": 0.3530268626022878,
      "grad_norm": 4.54408073425293,
      "learning_rate": 9.877500625270459e-05,
      "loss": 2.8306304931640627,
      "memory(GiB)": 58.3,
      "step": 8240,
      "token_acc": 0.4409722222222222,
      "train_speed(iter/s)": 1.453642
    },
    {
      "epoch": 0.3532410779315368,
      "grad_norm": 6.045342922210693,
      "learning_rate": 9.877352526839174e-05,
      "loss": 2.7010244369506835,
      "memory(GiB)": 58.3,
      "step": 8245,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.45362
    },
    {
      "epoch": 0.35345529326078573,
      "grad_norm": 3.9131600856781006,
      "learning_rate": 9.877204340050075e-05,
      "loss": 2.348368835449219,
      "memory(GiB)": 58.3,
      "step": 8250,
      "token_acc": 0.5098814229249012,
      "train_speed(iter/s)": 1.453505
    },
    {
      "epoch": 0.3536695085900347,
      "grad_norm": 4.147241115570068,
      "learning_rate": 9.87705606490584e-05,
      "loss": 2.5031055450439452,
      "memory(GiB)": 58.3,
      "step": 8255,
      "token_acc": 0.4629080118694362,
      "train_speed(iter/s)": 1.453343
    },
    {
      "epoch": 0.3538837239192837,
      "grad_norm": 5.277980327606201,
      "learning_rate": 9.876907701409164e-05,
      "loss": 2.503815841674805,
      "memory(GiB)": 58.3,
      "step": 8260,
      "token_acc": 0.4548736462093863,
      "train_speed(iter/s)": 1.453521
    },
    {
      "epoch": 0.3540979392485326,
      "grad_norm": 4.859445095062256,
      "learning_rate": 9.876759249562727e-05,
      "loss": 2.5020065307617188,
      "memory(GiB)": 58.3,
      "step": 8265,
      "token_acc": 0.44483985765124556,
      "train_speed(iter/s)": 1.453608
    },
    {
      "epoch": 0.3543121545777816,
      "grad_norm": 4.326780319213867,
      "learning_rate": 9.876610709369221e-05,
      "loss": 2.516596794128418,
      "memory(GiB)": 58.3,
      "step": 8270,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.453578
    },
    {
      "epoch": 0.35452636990703057,
      "grad_norm": 10.966876983642578,
      "learning_rate": 9.876462080831338e-05,
      "loss": 3.038375663757324,
      "memory(GiB)": 58.3,
      "step": 8275,
      "token_acc": 0.43356643356643354,
      "train_speed(iter/s)": 1.453699
    },
    {
      "epoch": 0.3547405852362795,
      "grad_norm": 5.078854084014893,
      "learning_rate": 9.876313363951772e-05,
      "loss": 2.286676788330078,
      "memory(GiB)": 58.3,
      "step": 8280,
      "token_acc": 0.46846846846846846,
      "train_speed(iter/s)": 1.453735
    },
    {
      "epoch": 0.35495480056552847,
      "grad_norm": 3.905805826187134,
      "learning_rate": 9.876164558733213e-05,
      "loss": 2.6840614318847655,
      "memory(GiB)": 58.3,
      "step": 8285,
      "token_acc": 0.42295081967213116,
      "train_speed(iter/s)": 1.45379
    },
    {
      "epoch": 0.35516901589477745,
      "grad_norm": 4.04298734664917,
      "learning_rate": 9.87601566517836e-05,
      "loss": 2.526031494140625,
      "memory(GiB)": 58.3,
      "step": 8290,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.453525
    },
    {
      "epoch": 0.3553832312240264,
      "grad_norm": 3.051178455352783,
      "learning_rate": 9.875866683289907e-05,
      "loss": 2.277154731750488,
      "memory(GiB)": 58.3,
      "step": 8295,
      "token_acc": 0.5374592833876222,
      "train_speed(iter/s)": 1.453497
    },
    {
      "epoch": 0.35559744655327535,
      "grad_norm": 3.419931173324585,
      "learning_rate": 9.875717613070558e-05,
      "loss": 2.6096355438232424,
      "memory(GiB)": 58.3,
      "step": 8300,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.45345
    },
    {
      "epoch": 0.35581166188252433,
      "grad_norm": 3.775873899459839,
      "learning_rate": 9.875568454523008e-05,
      "loss": 2.7648088455200197,
      "memory(GiB)": 58.3,
      "step": 8305,
      "token_acc": 0.41914191419141916,
      "train_speed(iter/s)": 1.453653
    },
    {
      "epoch": 0.35602587721177326,
      "grad_norm": 4.657072067260742,
      "learning_rate": 9.875419207649963e-05,
      "loss": 2.469293975830078,
      "memory(GiB)": 58.3,
      "step": 8310,
      "token_acc": 0.4318181818181818,
      "train_speed(iter/s)": 1.453677
    },
    {
      "epoch": 0.35624009254102224,
      "grad_norm": 3.663996934890747,
      "learning_rate": 9.875269872454127e-05,
      "loss": 2.6888759613037108,
      "memory(GiB)": 58.3,
      "step": 8315,
      "token_acc": 0.4407894736842105,
      "train_speed(iter/s)": 1.453757
    },
    {
      "epoch": 0.3564543078702712,
      "grad_norm": 4.971199035644531,
      "learning_rate": 9.875120448938201e-05,
      "loss": 2.319040870666504,
      "memory(GiB)": 58.3,
      "step": 8320,
      "token_acc": 0.5590551181102362,
      "train_speed(iter/s)": 1.453754
    },
    {
      "epoch": 0.35666852319952014,
      "grad_norm": 3.797389030456543,
      "learning_rate": 9.874970937104897e-05,
      "loss": 2.7418405532836916,
      "memory(GiB)": 58.3,
      "step": 8325,
      "token_acc": 0.4452296819787986,
      "train_speed(iter/s)": 1.453754
    },
    {
      "epoch": 0.3568827385287691,
      "grad_norm": 3.2229490280151367,
      "learning_rate": 9.87482133695692e-05,
      "loss": 2.6428028106689454,
      "memory(GiB)": 58.3,
      "step": 8330,
      "token_acc": 0.4491803278688525,
      "train_speed(iter/s)": 1.453805
    },
    {
      "epoch": 0.3570969538580181,
      "grad_norm": 5.507114887237549,
      "learning_rate": 9.874671648496983e-05,
      "loss": 2.785486602783203,
      "memory(GiB)": 58.3,
      "step": 8335,
      "token_acc": 0.4304635761589404,
      "train_speed(iter/s)": 1.453797
    },
    {
      "epoch": 0.357311169187267,
      "grad_norm": 4.2238311767578125,
      "learning_rate": 9.874521871727795e-05,
      "loss": 2.508288764953613,
      "memory(GiB)": 58.3,
      "step": 8340,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.453767
    },
    {
      "epoch": 0.357525384516516,
      "grad_norm": 3.8674449920654297,
      "learning_rate": 9.87437200665207e-05,
      "loss": 2.3915096282958985,
      "memory(GiB)": 58.3,
      "step": 8345,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.453642
    },
    {
      "epoch": 0.357739599845765,
      "grad_norm": 4.345756530761719,
      "learning_rate": 9.874222053272526e-05,
      "loss": 2.6997234344482424,
      "memory(GiB)": 58.3,
      "step": 8350,
      "token_acc": 0.4472049689440994,
      "train_speed(iter/s)": 1.453505
    },
    {
      "epoch": 0.3579538151750139,
      "grad_norm": 4.850778579711914,
      "learning_rate": 9.874072011591875e-05,
      "loss": 2.4386642456054686,
      "memory(GiB)": 58.3,
      "step": 8355,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.453669
    },
    {
      "epoch": 0.3581680305042629,
      "grad_norm": 3.9055323600769043,
      "learning_rate": 9.873921881612839e-05,
      "loss": 2.7483781814575194,
      "memory(GiB)": 58.3,
      "step": 8360,
      "token_acc": 0.46099290780141844,
      "train_speed(iter/s)": 1.453667
    },
    {
      "epoch": 0.35838224583351186,
      "grad_norm": 3.9249587059020996,
      "learning_rate": 9.873771663338134e-05,
      "loss": 2.520079231262207,
      "memory(GiB)": 58.3,
      "step": 8365,
      "token_acc": 0.46706586826347307,
      "train_speed(iter/s)": 1.453655
    },
    {
      "epoch": 0.3585964611627608,
      "grad_norm": 4.477032661437988,
      "learning_rate": 9.873621356770485e-05,
      "loss": 2.4322654724121096,
      "memory(GiB)": 58.3,
      "step": 8370,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.453563
    },
    {
      "epoch": 0.35881067649200976,
      "grad_norm": 4.751509666442871,
      "learning_rate": 9.873470961912612e-05,
      "loss": 2.9733051300048827,
      "memory(GiB)": 58.3,
      "step": 8375,
      "token_acc": 0.4370860927152318,
      "train_speed(iter/s)": 1.45348
    },
    {
      "epoch": 0.35902489182125874,
      "grad_norm": 5.375596523284912,
      "learning_rate": 9.873320478767242e-05,
      "loss": 2.792676544189453,
      "memory(GiB)": 58.3,
      "step": 8380,
      "token_acc": 0.4474708171206226,
      "train_speed(iter/s)": 1.453546
    },
    {
      "epoch": 0.35923910715050766,
      "grad_norm": 4.573060035705566,
      "learning_rate": 9.873169907337098e-05,
      "loss": 2.275337791442871,
      "memory(GiB)": 58.3,
      "step": 8385,
      "token_acc": 0.5224489795918368,
      "train_speed(iter/s)": 1.453406
    },
    {
      "epoch": 0.35945332247975664,
      "grad_norm": 4.999676704406738,
      "learning_rate": 9.873019247624913e-05,
      "loss": 2.385422134399414,
      "memory(GiB)": 58.3,
      "step": 8390,
      "token_acc": 0.48134328358208955,
      "train_speed(iter/s)": 1.45333
    },
    {
      "epoch": 0.3596675378090056,
      "grad_norm": 3.7001821994781494,
      "learning_rate": 9.87286849963341e-05,
      "loss": 2.6264087677001955,
      "memory(GiB)": 58.3,
      "step": 8395,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.453251
    },
    {
      "epoch": 0.3598817531382546,
      "grad_norm": 2.916172504425049,
      "learning_rate": 9.872717663365325e-05,
      "loss": 2.6320735931396486,
      "memory(GiB)": 58.3,
      "step": 8400,
      "token_acc": 0.46866485013623976,
      "train_speed(iter/s)": 1.453154
    },
    {
      "epoch": 0.3600959684675035,
      "grad_norm": 4.477244853973389,
      "learning_rate": 9.872566738823388e-05,
      "loss": 2.71956729888916,
      "memory(GiB)": 58.3,
      "step": 8405,
      "token_acc": 0.4406779661016949,
      "train_speed(iter/s)": 1.453321
    },
    {
      "epoch": 0.3603101837967525,
      "grad_norm": 3.813852310180664,
      "learning_rate": 9.872415726010334e-05,
      "loss": 2.7722780227661135,
      "memory(GiB)": 58.3,
      "step": 8410,
      "token_acc": 0.45918367346938777,
      "train_speed(iter/s)": 1.453045
    },
    {
      "epoch": 0.3605243991260015,
      "grad_norm": 4.583105564117432,
      "learning_rate": 9.872264624928898e-05,
      "loss": 2.407649803161621,
      "memory(GiB)": 58.3,
      "step": 8415,
      "token_acc": 0.5055762081784386,
      "train_speed(iter/s)": 1.452983
    },
    {
      "epoch": 0.3607386144552504,
      "grad_norm": 3.8841018676757812,
      "learning_rate": 9.872113435581818e-05,
      "loss": 2.5894634246826174,
      "memory(GiB)": 58.3,
      "step": 8420,
      "token_acc": 0.4743202416918429,
      "train_speed(iter/s)": 1.45283
    },
    {
      "epoch": 0.3609528297844994,
      "grad_norm": 5.313314437866211,
      "learning_rate": 9.871962157971832e-05,
      "loss": 2.122224235534668,
      "memory(GiB)": 58.3,
      "step": 8425,
      "token_acc": 0.5283842794759825,
      "train_speed(iter/s)": 1.452825
    },
    {
      "epoch": 0.36116704511374836,
      "grad_norm": 4.664624214172363,
      "learning_rate": 9.871810792101681e-05,
      "loss": 2.8059823989868162,
      "memory(GiB)": 58.3,
      "step": 8430,
      "token_acc": 0.5093167701863354,
      "train_speed(iter/s)": 1.452979
    },
    {
      "epoch": 0.3613812604429973,
      "grad_norm": 4.487034320831299,
      "learning_rate": 9.871659337974109e-05,
      "loss": 2.6467830657958986,
      "memory(GiB)": 58.3,
      "step": 8435,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.452817
    },
    {
      "epoch": 0.36159547577224627,
      "grad_norm": 4.632307529449463,
      "learning_rate": 9.871507795591857e-05,
      "loss": 2.727056694030762,
      "memory(GiB)": 58.3,
      "step": 8440,
      "token_acc": 0.4491803278688525,
      "train_speed(iter/s)": 1.452824
    },
    {
      "epoch": 0.36180969110149525,
      "grad_norm": 4.677274227142334,
      "learning_rate": 9.87135616495767e-05,
      "loss": 2.635938835144043,
      "memory(GiB)": 58.3,
      "step": 8445,
      "token_acc": 0.4959349593495935,
      "train_speed(iter/s)": 1.452819
    },
    {
      "epoch": 0.36202390643074417,
      "grad_norm": 4.1115336418151855,
      "learning_rate": 9.871204446074298e-05,
      "loss": 2.6627834320068358,
      "memory(GiB)": 58.3,
      "step": 8450,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.452615
    },
    {
      "epoch": 0.36223812175999315,
      "grad_norm": 4.5900750160217285,
      "learning_rate": 9.871052638944489e-05,
      "loss": 2.5138330459594727,
      "memory(GiB)": 58.3,
      "step": 8455,
      "token_acc": 0.4858156028368794,
      "train_speed(iter/s)": 1.452559
    },
    {
      "epoch": 0.3624523370892421,
      "grad_norm": 6.558152675628662,
      "learning_rate": 9.87090074357099e-05,
      "loss": 2.6777801513671875,
      "memory(GiB)": 58.3,
      "step": 8460,
      "token_acc": 0.4790996784565916,
      "train_speed(iter/s)": 1.452618
    },
    {
      "epoch": 0.36266655241849105,
      "grad_norm": 3.3310494422912598,
      "learning_rate": 9.870748759956556e-05,
      "loss": 2.4743141174316405,
      "memory(GiB)": 58.3,
      "step": 8465,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.452814
    },
    {
      "epoch": 0.36288076774774003,
      "grad_norm": 4.387814044952393,
      "learning_rate": 9.87059668810394e-05,
      "loss": 2.566577911376953,
      "memory(GiB)": 58.3,
      "step": 8470,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.452952
    },
    {
      "epoch": 0.363094983076989,
      "grad_norm": 4.381251811981201,
      "learning_rate": 9.870444528015895e-05,
      "loss": 2.745221710205078,
      "memory(GiB)": 58.3,
      "step": 8475,
      "token_acc": 0.41389728096676737,
      "train_speed(iter/s)": 1.452875
    },
    {
      "epoch": 0.36330919840623793,
      "grad_norm": 4.326045513153076,
      "learning_rate": 9.870292279695177e-05,
      "loss": 2.5108896255493165,
      "memory(GiB)": 58.3,
      "step": 8480,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.453156
    },
    {
      "epoch": 0.3635234137354869,
      "grad_norm": 3.560746908187866,
      "learning_rate": 9.870139943144547e-05,
      "loss": 2.908544731140137,
      "memory(GiB)": 58.3,
      "step": 8485,
      "token_acc": 0.43272727272727274,
      "train_speed(iter/s)": 1.453043
    },
    {
      "epoch": 0.3637376290647359,
      "grad_norm": 4.861599445343018,
      "learning_rate": 9.869987518366763e-05,
      "loss": 2.437498092651367,
      "memory(GiB)": 58.3,
      "step": 8490,
      "token_acc": 0.5034722222222222,
      "train_speed(iter/s)": 1.453108
    },
    {
      "epoch": 0.3639518443939848,
      "grad_norm": 4.313116550445557,
      "learning_rate": 9.869835005364587e-05,
      "loss": 2.867035484313965,
      "memory(GiB)": 58.3,
      "step": 8495,
      "token_acc": 0.4520547945205479,
      "train_speed(iter/s)": 1.453244
    },
    {
      "epoch": 0.3641660597232338,
      "grad_norm": 3.0654003620147705,
      "learning_rate": 9.869682404140781e-05,
      "loss": 2.5062629699707033,
      "memory(GiB)": 58.3,
      "step": 8500,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.453331
    },
    {
      "epoch": 0.3641660597232338,
      "eval_loss": 2.2773027420043945,
      "eval_runtime": 14.6047,
      "eval_samples_per_second": 6.847,
      "eval_steps_per_second": 6.847,
      "eval_token_acc": 0.4375,
      "step": 8500
    },
    {
      "epoch": 0.36438027505248277,
      "grad_norm": 4.361446857452393,
      "learning_rate": 9.869529714698111e-05,
      "loss": 2.8385292053222657,
      "memory(GiB)": 58.3,
      "step": 8505,
      "token_acc": 0.43667296786389415,
      "train_speed(iter/s)": 1.449559
    },
    {
      "epoch": 0.3645944903817317,
      "grad_norm": 4.428905963897705,
      "learning_rate": 9.869376937039342e-05,
      "loss": 2.4612777709960936,
      "memory(GiB)": 58.3,
      "step": 8510,
      "token_acc": 0.49624060150375937,
      "train_speed(iter/s)": 1.449048
    },
    {
      "epoch": 0.3648087057109807,
      "grad_norm": 3.629077911376953,
      "learning_rate": 9.869224071167242e-05,
      "loss": 2.7202875137329103,
      "memory(GiB)": 58.3,
      "step": 8515,
      "token_acc": 0.4522058823529412,
      "train_speed(iter/s)": 1.449052
    },
    {
      "epoch": 0.36502292104022965,
      "grad_norm": 3.568740129470825,
      "learning_rate": 9.869071117084581e-05,
      "loss": 2.969765472412109,
      "memory(GiB)": 58.3,
      "step": 8520,
      "token_acc": 0.41420118343195267,
      "train_speed(iter/s)": 1.448825
    },
    {
      "epoch": 0.3652371363694786,
      "grad_norm": 4.509570598602295,
      "learning_rate": 9.868918074794126e-05,
      "loss": 2.496523857116699,
      "memory(GiB)": 58.3,
      "step": 8525,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.448846
    },
    {
      "epoch": 0.36545135169872756,
      "grad_norm": 4.692948341369629,
      "learning_rate": 9.868764944298656e-05,
      "loss": 2.6465007781982424,
      "memory(GiB)": 58.3,
      "step": 8530,
      "token_acc": 0.4471299093655589,
      "train_speed(iter/s)": 1.448729
    },
    {
      "epoch": 0.36566556702797653,
      "grad_norm": 4.059881210327148,
      "learning_rate": 9.86861172560094e-05,
      "loss": 2.5645797729492186,
      "memory(GiB)": 58.3,
      "step": 8535,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.448923
    },
    {
      "epoch": 0.36587978235722546,
      "grad_norm": 4.927090644836426,
      "learning_rate": 9.868458418703756e-05,
      "loss": 2.505494499206543,
      "memory(GiB)": 58.3,
      "step": 8540,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.449056
    },
    {
      "epoch": 0.36609399768647444,
      "grad_norm": 5.3619866371154785,
      "learning_rate": 9.868305023609881e-05,
      "loss": 2.314160919189453,
      "memory(GiB)": 58.3,
      "step": 8545,
      "token_acc": 0.515527950310559,
      "train_speed(iter/s)": 1.44932
    },
    {
      "epoch": 0.3663082130157234,
      "grad_norm": 4.235127925872803,
      "learning_rate": 9.868151540322094e-05,
      "loss": 2.611054611206055,
      "memory(GiB)": 58.3,
      "step": 8550,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.449418
    },
    {
      "epoch": 0.36652242834497234,
      "grad_norm": 4.088761806488037,
      "learning_rate": 9.867997968843175e-05,
      "loss": 2.767421340942383,
      "memory(GiB)": 58.3,
      "step": 8555,
      "token_acc": 0.4199288256227758,
      "train_speed(iter/s)": 1.449525
    },
    {
      "epoch": 0.3667366436742213,
      "grad_norm": 5.123048782348633,
      "learning_rate": 9.867844309175906e-05,
      "loss": 2.5393394470214843,
      "memory(GiB)": 58.3,
      "step": 8560,
      "token_acc": 0.47076023391812866,
      "train_speed(iter/s)": 1.449602
    },
    {
      "epoch": 0.3669508590034703,
      "grad_norm": 3.409019947052002,
      "learning_rate": 9.86769056132307e-05,
      "loss": 2.4834274291992187,
      "memory(GiB)": 58.3,
      "step": 8565,
      "token_acc": 0.4709480122324159,
      "train_speed(iter/s)": 1.449539
    },
    {
      "epoch": 0.3671650743327193,
      "grad_norm": 4.338001251220703,
      "learning_rate": 9.867536725287455e-05,
      "loss": 2.516732406616211,
      "memory(GiB)": 58.3,
      "step": 8570,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.449536
    },
    {
      "epoch": 0.3673792896619682,
      "grad_norm": 3.4665451049804688,
      "learning_rate": 9.867382801071844e-05,
      "loss": 2.4565670013427736,
      "memory(GiB)": 58.3,
      "step": 8575,
      "token_acc": 0.43666666666666665,
      "train_speed(iter/s)": 1.449611
    },
    {
      "epoch": 0.3675935049912172,
      "grad_norm": 3.400789260864258,
      "learning_rate": 9.86722878867903e-05,
      "loss": 2.612316703796387,
      "memory(GiB)": 58.3,
      "step": 8580,
      "token_acc": 0.4290909090909091,
      "train_speed(iter/s)": 1.449648
    },
    {
      "epoch": 0.36780772032046616,
      "grad_norm": 4.158246040344238,
      "learning_rate": 9.867074688111799e-05,
      "loss": 2.8523075103759767,
      "memory(GiB)": 58.3,
      "step": 8585,
      "token_acc": 0.424812030075188,
      "train_speed(iter/s)": 1.449566
    },
    {
      "epoch": 0.3680219356497151,
      "grad_norm": 3.6310372352600098,
      "learning_rate": 9.866920499372944e-05,
      "loss": 2.381735610961914,
      "memory(GiB)": 58.3,
      "step": 8590,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.4495
    },
    {
      "epoch": 0.36823615097896406,
      "grad_norm": 4.189392566680908,
      "learning_rate": 9.86676622246526e-05,
      "loss": 2.3994396209716795,
      "memory(GiB)": 58.3,
      "step": 8595,
      "token_acc": 0.5145228215767634,
      "train_speed(iter/s)": 1.449502
    },
    {
      "epoch": 0.36845036630821304,
      "grad_norm": 5.331646919250488,
      "learning_rate": 9.866611857391541e-05,
      "loss": 2.479269027709961,
      "memory(GiB)": 58.3,
      "step": 8600,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.44951
    },
    {
      "epoch": 0.36866458163746196,
      "grad_norm": 3.725822687149048,
      "learning_rate": 9.866457404154581e-05,
      "loss": 2.485365867614746,
      "memory(GiB)": 58.3,
      "step": 8605,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.449552
    },
    {
      "epoch": 0.36887879696671094,
      "grad_norm": 4.105701923370361,
      "learning_rate": 9.86630286275718e-05,
      "loss": 2.7235858917236326,
      "memory(GiB)": 58.3,
      "step": 8610,
      "token_acc": 0.4441176470588235,
      "train_speed(iter/s)": 1.44967
    },
    {
      "epoch": 0.3690930122959599,
      "grad_norm": 4.021725177764893,
      "learning_rate": 9.866148233202139e-05,
      "loss": 2.4119684219360353,
      "memory(GiB)": 58.3,
      "step": 8615,
      "token_acc": 0.4858156028368794,
      "train_speed(iter/s)": 1.449672
    },
    {
      "epoch": 0.36930722762520884,
      "grad_norm": 3.6348023414611816,
      "learning_rate": 9.865993515492258e-05,
      "loss": 2.7645082473754883,
      "memory(GiB)": 58.3,
      "step": 8620,
      "token_acc": 0.4540059347181009,
      "train_speed(iter/s)": 1.449739
    },
    {
      "epoch": 0.3695214429544578,
      "grad_norm": 3.8476059436798096,
      "learning_rate": 9.865838709630339e-05,
      "loss": 2.7304811477661133,
      "memory(GiB)": 58.3,
      "step": 8625,
      "token_acc": 0.48985507246376814,
      "train_speed(iter/s)": 1.449832
    },
    {
      "epoch": 0.3697356582837068,
      "grad_norm": 5.059749126434326,
      "learning_rate": 9.865683815619188e-05,
      "loss": 2.665720748901367,
      "memory(GiB)": 58.3,
      "step": 8630,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.449935
    },
    {
      "epoch": 0.3699498736129557,
      "grad_norm": 3.95491099357605,
      "learning_rate": 9.865528833461611e-05,
      "loss": 2.3191860198974608,
      "memory(GiB)": 58.3,
      "step": 8635,
      "token_acc": 0.5070422535211268,
      "train_speed(iter/s)": 1.449937
    },
    {
      "epoch": 0.3701640889422047,
      "grad_norm": 3.9109537601470947,
      "learning_rate": 9.865373763160413e-05,
      "loss": 2.7906370162963867,
      "memory(GiB)": 58.3,
      "step": 8640,
      "token_acc": 0.39792387543252594,
      "train_speed(iter/s)": 1.449898
    },
    {
      "epoch": 0.3703783042714537,
      "grad_norm": 4.044189453125,
      "learning_rate": 9.865218604718405e-05,
      "loss": 2.555398368835449,
      "memory(GiB)": 58.3,
      "step": 8645,
      "token_acc": 0.490625,
      "train_speed(iter/s)": 1.450051
    },
    {
      "epoch": 0.3705925196007026,
      "grad_norm": 4.719189167022705,
      "learning_rate": 9.8650633581384e-05,
      "loss": 2.5803035736083983,
      "memory(GiB)": 58.3,
      "step": 8650,
      "token_acc": 0.4627831715210356,
      "train_speed(iter/s)": 1.449982
    },
    {
      "epoch": 0.3708067349299516,
      "grad_norm": 4.118319034576416,
      "learning_rate": 9.864908023423207e-05,
      "loss": 2.4400157928466797,
      "memory(GiB)": 58.3,
      "step": 8655,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.4499
    },
    {
      "epoch": 0.37102095025920057,
      "grad_norm": 3.280740261077881,
      "learning_rate": 9.864752600575641e-05,
      "loss": 2.8693151473999023,
      "memory(GiB)": 58.3,
      "step": 8660,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.45001
    },
    {
      "epoch": 0.3712351655884495,
      "grad_norm": 4.42762565612793,
      "learning_rate": 9.864597089598519e-05,
      "loss": 2.679347038269043,
      "memory(GiB)": 58.3,
      "step": 8665,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.449935
    },
    {
      "epoch": 0.37144938091769847,
      "grad_norm": 4.9428887367248535,
      "learning_rate": 9.864441490494658e-05,
      "loss": 2.621337127685547,
      "memory(GiB)": 58.3,
      "step": 8670,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.449815
    },
    {
      "epoch": 0.37166359624694745,
      "grad_norm": 3.82576584815979,
      "learning_rate": 9.864285803266876e-05,
      "loss": 2.7555059432983398,
      "memory(GiB)": 58.3,
      "step": 8675,
      "token_acc": 0.43130990415335463,
      "train_speed(iter/s)": 1.449837
    },
    {
      "epoch": 0.37187781157619637,
      "grad_norm": 3.818668842315674,
      "learning_rate": 9.864130027917993e-05,
      "loss": 2.8221467971801757,
      "memory(GiB)": 58.3,
      "step": 8680,
      "token_acc": 0.42813455657492355,
      "train_speed(iter/s)": 1.44998
    },
    {
      "epoch": 0.37209202690544535,
      "grad_norm": 4.1920485496521,
      "learning_rate": 9.863974164450833e-05,
      "loss": 2.164962959289551,
      "memory(GiB)": 58.3,
      "step": 8685,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.449936
    },
    {
      "epoch": 0.37230624223469433,
      "grad_norm": 3.3402833938598633,
      "learning_rate": 9.863818212868217e-05,
      "loss": 2.559827423095703,
      "memory(GiB)": 58.3,
      "step": 8690,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.449817
    },
    {
      "epoch": 0.37252045756394325,
      "grad_norm": 3.664734125137329,
      "learning_rate": 9.863662173172971e-05,
      "loss": 2.6619312286376955,
      "memory(GiB)": 58.3,
      "step": 8695,
      "token_acc": 0.4270833333333333,
      "train_speed(iter/s)": 1.449801
    },
    {
      "epoch": 0.37273467289319223,
      "grad_norm": 4.694035530090332,
      "learning_rate": 9.863506045367923e-05,
      "loss": 2.540768051147461,
      "memory(GiB)": 58.3,
      "step": 8700,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.449959
    },
    {
      "epoch": 0.3729488882224412,
      "grad_norm": 6.23726749420166,
      "learning_rate": 9.863349829455899e-05,
      "loss": 2.611031341552734,
      "memory(GiB)": 58.3,
      "step": 8705,
      "token_acc": 0.48091603053435117,
      "train_speed(iter/s)": 1.449981
    },
    {
      "epoch": 0.37316310355169013,
      "grad_norm": 3.676201105117798,
      "learning_rate": 9.863193525439734e-05,
      "loss": 2.4545129776000976,
      "memory(GiB)": 58.3,
      "step": 8710,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.449955
    },
    {
      "epoch": 0.3733773188809391,
      "grad_norm": 4.304929256439209,
      "learning_rate": 9.863037133322252e-05,
      "loss": 2.3013126373291017,
      "memory(GiB)": 58.3,
      "step": 8715,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.449877
    },
    {
      "epoch": 0.3735915342101881,
      "grad_norm": 4.57396125793457,
      "learning_rate": 9.862880653106294e-05,
      "loss": 2.7888824462890627,
      "memory(GiB)": 58.3,
      "step": 8720,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.449935
    },
    {
      "epoch": 0.373805749539437,
      "grad_norm": 4.094754695892334,
      "learning_rate": 9.86272408479469e-05,
      "loss": 2.3246604919433596,
      "memory(GiB)": 58.3,
      "step": 8725,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.449748
    },
    {
      "epoch": 0.374019964868686,
      "grad_norm": 5.775775909423828,
      "learning_rate": 9.862567428390277e-05,
      "loss": 2.8006607055664063,
      "memory(GiB)": 58.3,
      "step": 8730,
      "token_acc": 0.45416666666666666,
      "train_speed(iter/s)": 1.449714
    },
    {
      "epoch": 0.374234180197935,
      "grad_norm": 4.251965045928955,
      "learning_rate": 9.862410683895895e-05,
      "loss": 2.8104974746704103,
      "memory(GiB)": 58.3,
      "step": 8735,
      "token_acc": 0.41530054644808745,
      "train_speed(iter/s)": 1.449563
    },
    {
      "epoch": 0.37444839552718395,
      "grad_norm": 4.093268871307373,
      "learning_rate": 9.86225385131438e-05,
      "loss": 2.719206428527832,
      "memory(GiB)": 58.3,
      "step": 8740,
      "token_acc": 0.4472049689440994,
      "train_speed(iter/s)": 1.449646
    },
    {
      "epoch": 0.3746626108564329,
      "grad_norm": 3.7247488498687744,
      "learning_rate": 9.862096930648577e-05,
      "loss": 2.518401336669922,
      "memory(GiB)": 58.3,
      "step": 8745,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.449695
    },
    {
      "epoch": 0.37487682618568186,
      "grad_norm": 6.050350189208984,
      "learning_rate": 9.861939921901326e-05,
      "loss": 2.544095993041992,
      "memory(GiB)": 58.3,
      "step": 8750,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.449597
    },
    {
      "epoch": 0.37509104151493083,
      "grad_norm": 5.940671920776367,
      "learning_rate": 9.861782825075475e-05,
      "loss": 2.575368309020996,
      "memory(GiB)": 58.3,
      "step": 8755,
      "token_acc": 0.4591439688715953,
      "train_speed(iter/s)": 1.449682
    },
    {
      "epoch": 0.37530525684417976,
      "grad_norm": 4.251453399658203,
      "learning_rate": 9.861625640173865e-05,
      "loss": 2.481136703491211,
      "memory(GiB)": 58.3,
      "step": 8760,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.449969
    },
    {
      "epoch": 0.37551947217342874,
      "grad_norm": 4.797479152679443,
      "learning_rate": 9.861468367199346e-05,
      "loss": 2.675431823730469,
      "memory(GiB)": 58.3,
      "step": 8765,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.449976
    },
    {
      "epoch": 0.3757336875026777,
      "grad_norm": 4.372289657592773,
      "learning_rate": 9.861311006154767e-05,
      "loss": 2.4585329055786134,
      "memory(GiB)": 58.3,
      "step": 8770,
      "token_acc": 0.5057915057915058,
      "train_speed(iter/s)": 1.449499
    },
    {
      "epoch": 0.37594790283192664,
      "grad_norm": 4.571930885314941,
      "learning_rate": 9.86115355704298e-05,
      "loss": 2.6418178558349608,
      "memory(GiB)": 58.3,
      "step": 8775,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.44961
    },
    {
      "epoch": 0.3761621181611756,
      "grad_norm": 3.6882073879241943,
      "learning_rate": 9.860996019866836e-05,
      "loss": 2.5285486221313476,
      "memory(GiB)": 58.3,
      "step": 8780,
      "token_acc": 0.45871559633027525,
      "train_speed(iter/s)": 1.449899
    },
    {
      "epoch": 0.3763763334904246,
      "grad_norm": 3.6706016063690186,
      "learning_rate": 9.860838394629188e-05,
      "loss": 2.7798057556152345,
      "memory(GiB)": 58.3,
      "step": 8785,
      "token_acc": 0.46794871794871795,
      "train_speed(iter/s)": 1.449993
    },
    {
      "epoch": 0.3765905488196735,
      "grad_norm": 3.9987995624542236,
      "learning_rate": 9.860680681332894e-05,
      "loss": 2.616022491455078,
      "memory(GiB)": 58.3,
      "step": 8790,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.450214
    },
    {
      "epoch": 0.3768047641489225,
      "grad_norm": 2.946453094482422,
      "learning_rate": 9.860522879980809e-05,
      "loss": 2.4633434295654295,
      "memory(GiB)": 58.3,
      "step": 8795,
      "token_acc": 0.5241635687732342,
      "train_speed(iter/s)": 1.450324
    },
    {
      "epoch": 0.3770189794781715,
      "grad_norm": 4.215868949890137,
      "learning_rate": 9.860364990575792e-05,
      "loss": 2.6088342666625977,
      "memory(GiB)": 58.3,
      "step": 8800,
      "token_acc": 0.463768115942029,
      "train_speed(iter/s)": 1.450562
    },
    {
      "epoch": 0.3772331948074204,
      "grad_norm": 3.7749364376068115,
      "learning_rate": 9.860207013120706e-05,
      "loss": 2.602193832397461,
      "memory(GiB)": 58.3,
      "step": 8805,
      "token_acc": 0.4676258992805755,
      "train_speed(iter/s)": 1.450677
    },
    {
      "epoch": 0.3774474101366694,
      "grad_norm": 6.906366348266602,
      "learning_rate": 9.860048947618408e-05,
      "loss": 2.79365234375,
      "memory(GiB)": 58.3,
      "step": 8810,
      "token_acc": 0.4196078431372549,
      "train_speed(iter/s)": 1.450832
    },
    {
      "epoch": 0.37766162546591836,
      "grad_norm": 5.40219783782959,
      "learning_rate": 9.859890794071767e-05,
      "loss": 2.6922300338745115,
      "memory(GiB)": 58.3,
      "step": 8815,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.450909
    },
    {
      "epoch": 0.3778758407951673,
      "grad_norm": 4.223740100860596,
      "learning_rate": 9.859732552483642e-05,
      "loss": 2.652150344848633,
      "memory(GiB)": 58.3,
      "step": 8820,
      "token_acc": 0.4277456647398844,
      "train_speed(iter/s)": 1.450953
    },
    {
      "epoch": 0.37809005612441626,
      "grad_norm": 4.2968597412109375,
      "learning_rate": 9.859574222856905e-05,
      "loss": 2.3183837890625,
      "memory(GiB)": 58.3,
      "step": 8825,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.450927
    },
    {
      "epoch": 0.37830427145366524,
      "grad_norm": 5.012269973754883,
      "learning_rate": 9.859415805194422e-05,
      "loss": 2.4743255615234374,
      "memory(GiB)": 58.3,
      "step": 8830,
      "token_acc": 0.48220064724919093,
      "train_speed(iter/s)": 1.450937
    },
    {
      "epoch": 0.37851848678291417,
      "grad_norm": 3.280315399169922,
      "learning_rate": 9.859257299499064e-05,
      "loss": 2.5424777984619142,
      "memory(GiB)": 58.3,
      "step": 8835,
      "token_acc": 0.4530612244897959,
      "train_speed(iter/s)": 1.451087
    },
    {
      "epoch": 0.37873270211216314,
      "grad_norm": 3.276996612548828,
      "learning_rate": 9.859098705773701e-05,
      "loss": 2.8809751510620116,
      "memory(GiB)": 58.3,
      "step": 8840,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.451148
    },
    {
      "epoch": 0.3789469174414121,
      "grad_norm": 2.588827133178711,
      "learning_rate": 9.858940024021205e-05,
      "loss": 2.124911880493164,
      "memory(GiB)": 58.3,
      "step": 8845,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.451228
    },
    {
      "epoch": 0.37916113277066105,
      "grad_norm": 4.800065040588379,
      "learning_rate": 9.858781254244455e-05,
      "loss": 2.6147735595703123,
      "memory(GiB)": 58.3,
      "step": 8850,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.451035
    },
    {
      "epoch": 0.37937534809991,
      "grad_norm": 4.340216636657715,
      "learning_rate": 9.858622396446325e-05,
      "loss": 2.6263282775878904,
      "memory(GiB)": 58.3,
      "step": 8855,
      "token_acc": 0.45051194539249145,
      "train_speed(iter/s)": 1.451073
    },
    {
      "epoch": 0.379589563429159,
      "grad_norm": 6.374390602111816,
      "learning_rate": 9.85846345062969e-05,
      "loss": 2.4751708984375,
      "memory(GiB)": 58.3,
      "step": 8860,
      "token_acc": 0.4789272030651341,
      "train_speed(iter/s)": 1.451247
    },
    {
      "epoch": 0.37980377875840793,
      "grad_norm": 6.287930488586426,
      "learning_rate": 9.858304416797433e-05,
      "loss": 2.4763072967529296,
      "memory(GiB)": 58.3,
      "step": 8865,
      "token_acc": 0.46586345381526106,
      "train_speed(iter/s)": 1.451324
    },
    {
      "epoch": 0.3800179940876569,
      "grad_norm": 4.323367118835449,
      "learning_rate": 9.858145294952435e-05,
      "loss": 2.5795166015625,
      "memory(GiB)": 58.3,
      "step": 8870,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.451165
    },
    {
      "epoch": 0.3802322094169059,
      "grad_norm": 4.53786563873291,
      "learning_rate": 9.857986085097577e-05,
      "loss": 2.6223476409912108,
      "memory(GiB)": 58.3,
      "step": 8875,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.451014
    },
    {
      "epoch": 0.3804464247461548,
      "grad_norm": 3.346238851547241,
      "learning_rate": 9.857826787235744e-05,
      "loss": 2.848287010192871,
      "memory(GiB)": 58.3,
      "step": 8880,
      "token_acc": 0.4641638225255973,
      "train_speed(iter/s)": 1.451134
    },
    {
      "epoch": 0.3806606400754038,
      "grad_norm": 3.6543309688568115,
      "learning_rate": 9.85766740136982e-05,
      "loss": 2.6900163650512696,
      "memory(GiB)": 58.3,
      "step": 8885,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.451118
    },
    {
      "epoch": 0.38087485540465277,
      "grad_norm": 3.6415889263153076,
      "learning_rate": 9.857507927502697e-05,
      "loss": 2.757152557373047,
      "memory(GiB)": 58.3,
      "step": 8890,
      "token_acc": 0.46646341463414637,
      "train_speed(iter/s)": 1.45105
    },
    {
      "epoch": 0.3810890707339017,
      "grad_norm": 3.651844024658203,
      "learning_rate": 9.85734836563726e-05,
      "loss": 2.892853355407715,
      "memory(GiB)": 58.3,
      "step": 8895,
      "token_acc": 0.4171597633136095,
      "train_speed(iter/s)": 1.450978
    },
    {
      "epoch": 0.38130328606315067,
      "grad_norm": 4.364047527313232,
      "learning_rate": 9.857188715776401e-05,
      "loss": 2.6258750915527345,
      "memory(GiB)": 58.3,
      "step": 8900,
      "token_acc": 0.44891640866873067,
      "train_speed(iter/s)": 1.450983
    },
    {
      "epoch": 0.38151750139239965,
      "grad_norm": 3.3479998111724854,
      "learning_rate": 9.857028977923011e-05,
      "loss": 2.9866472244262696,
      "memory(GiB)": 58.3,
      "step": 8905,
      "token_acc": 0.43213296398891965,
      "train_speed(iter/s)": 1.451027
    },
    {
      "epoch": 0.38173171672164863,
      "grad_norm": 3.5985403060913086,
      "learning_rate": 9.856869152079986e-05,
      "loss": 2.602186393737793,
      "memory(GiB)": 58.3,
      "step": 8910,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.451058
    },
    {
      "epoch": 0.38194593205089755,
      "grad_norm": 3.907210350036621,
      "learning_rate": 9.85670923825022e-05,
      "loss": 2.5993650436401365,
      "memory(GiB)": 58.3,
      "step": 8915,
      "token_acc": 0.41,
      "train_speed(iter/s)": 1.45122
    },
    {
      "epoch": 0.38216014738014653,
      "grad_norm": 5.691684246063232,
      "learning_rate": 9.856549236436609e-05,
      "loss": 2.7065284729003904,
      "memory(GiB)": 58.3,
      "step": 8920,
      "token_acc": 0.4392857142857143,
      "train_speed(iter/s)": 1.451179
    },
    {
      "epoch": 0.3823743627093955,
      "grad_norm": 4.579247951507568,
      "learning_rate": 9.856389146642054e-05,
      "loss": 2.5624696731567385,
      "memory(GiB)": 58.3,
      "step": 8925,
      "token_acc": 0.5089605734767025,
      "train_speed(iter/s)": 1.451182
    },
    {
      "epoch": 0.38258857803864443,
      "grad_norm": 3.177950143814087,
      "learning_rate": 9.856228968869454e-05,
      "loss": 2.6077198028564452,
      "memory(GiB)": 58.3,
      "step": 8930,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.451272
    },
    {
      "epoch": 0.3828027933678934,
      "grad_norm": 6.067451000213623,
      "learning_rate": 9.85606870312171e-05,
      "loss": 2.707725715637207,
      "memory(GiB)": 58.3,
      "step": 8935,
      "token_acc": 0.4560260586319218,
      "train_speed(iter/s)": 1.45128
    },
    {
      "epoch": 0.3830170086971424,
      "grad_norm": 4.170866966247559,
      "learning_rate": 9.855908349401727e-05,
      "loss": 2.529203987121582,
      "memory(GiB)": 58.3,
      "step": 8940,
      "token_acc": 0.49050632911392406,
      "train_speed(iter/s)": 1.451371
    },
    {
      "epoch": 0.3832312240263913,
      "grad_norm": 4.348936557769775,
      "learning_rate": 9.855747907712408e-05,
      "loss": 2.575935935974121,
      "memory(GiB)": 58.3,
      "step": 8945,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.451507
    },
    {
      "epoch": 0.3834454393556403,
      "grad_norm": 5.326323986053467,
      "learning_rate": 9.855587378056661e-05,
      "loss": 2.651362991333008,
      "memory(GiB)": 58.3,
      "step": 8950,
      "token_acc": 0.45791245791245794,
      "train_speed(iter/s)": 1.451672
    },
    {
      "epoch": 0.3836596546848893,
      "grad_norm": 4.141667366027832,
      "learning_rate": 9.855426760437394e-05,
      "loss": 2.64807014465332,
      "memory(GiB)": 58.3,
      "step": 8955,
      "token_acc": 0.4721189591078067,
      "train_speed(iter/s)": 1.451907
    },
    {
      "epoch": 0.3838738700141382,
      "grad_norm": 4.396313667297363,
      "learning_rate": 9.855266054857518e-05,
      "loss": 2.696456718444824,
      "memory(GiB)": 58.3,
      "step": 8960,
      "token_acc": 0.4318181818181818,
      "train_speed(iter/s)": 1.452041
    },
    {
      "epoch": 0.3840880853433872,
      "grad_norm": 4.367549896240234,
      "learning_rate": 9.855105261319939e-05,
      "loss": 2.4935434341430662,
      "memory(GiB)": 58.3,
      "step": 8965,
      "token_acc": 0.49173553719008267,
      "train_speed(iter/s)": 1.452163
    },
    {
      "epoch": 0.38430230067263615,
      "grad_norm": 3.9490792751312256,
      "learning_rate": 9.854944379827577e-05,
      "loss": 2.7790895462036134,
      "memory(GiB)": 58.3,
      "step": 8970,
      "token_acc": 0.4520123839009288,
      "train_speed(iter/s)": 1.452369
    },
    {
      "epoch": 0.3845165160018851,
      "grad_norm": 5.052127838134766,
      "learning_rate": 9.854783410383341e-05,
      "loss": 2.434416961669922,
      "memory(GiB)": 58.3,
      "step": 8975,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.452607
    },
    {
      "epoch": 0.38473073133113406,
      "grad_norm": 5.247725486755371,
      "learning_rate": 9.85462235299015e-05,
      "loss": 2.680352973937988,
      "memory(GiB)": 58.3,
      "step": 8980,
      "token_acc": 0.49166666666666664,
      "train_speed(iter/s)": 1.452584
    },
    {
      "epoch": 0.38494494666038304,
      "grad_norm": 3.5917298793792725,
      "learning_rate": 9.854461207650922e-05,
      "loss": 2.3029912948608398,
      "memory(GiB)": 58.3,
      "step": 8985,
      "token_acc": 0.5160142348754448,
      "train_speed(iter/s)": 1.452604
    },
    {
      "epoch": 0.38515916198963196,
      "grad_norm": 5.086394309997559,
      "learning_rate": 9.854299974368575e-05,
      "loss": 2.502714157104492,
      "memory(GiB)": 58.3,
      "step": 8990,
      "token_acc": 0.5120967741935484,
      "train_speed(iter/s)": 1.452735
    },
    {
      "epoch": 0.38537337731888094,
      "grad_norm": 2.9717671871185303,
      "learning_rate": 9.85413865314603e-05,
      "loss": 2.5790231704711912,
      "memory(GiB)": 58.3,
      "step": 8995,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.452849
    },
    {
      "epoch": 0.3855875926481299,
      "grad_norm": 6.031939506530762,
      "learning_rate": 9.85397724398621e-05,
      "loss": 2.724561309814453,
      "memory(GiB)": 58.3,
      "step": 9000,
      "token_acc": 0.46558704453441296,
      "train_speed(iter/s)": 1.452797
    },
    {
      "epoch": 0.3855875926481299,
      "eval_loss": 2.1903438568115234,
      "eval_runtime": 13.7791,
      "eval_samples_per_second": 7.257,
      "eval_steps_per_second": 7.257,
      "eval_token_acc": 0.47277936962750716,
      "step": 9000
    },
    {
      "epoch": 0.38580180797737884,
      "grad_norm": 5.6124958992004395,
      "learning_rate": 9.853815746892039e-05,
      "loss": 2.979076385498047,
      "memory(GiB)": 58.3,
      "step": 9005,
      "token_acc": 0.45857418111753373,
      "train_speed(iter/s)": 1.449633
    },
    {
      "epoch": 0.3860160233066278,
      "grad_norm": 6.156009674072266,
      "learning_rate": 9.853654161866442e-05,
      "loss": 2.6359397888183596,
      "memory(GiB)": 58.3,
      "step": 9010,
      "token_acc": 0.452,
      "train_speed(iter/s)": 1.449689
    },
    {
      "epoch": 0.3862302386358768,
      "grad_norm": 3.8203330039978027,
      "learning_rate": 9.853492488912347e-05,
      "loss": 2.7224937438964845,
      "memory(GiB)": 58.3,
      "step": 9015,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.449882
    },
    {
      "epoch": 0.3864444539651257,
      "grad_norm": 4.893547534942627,
      "learning_rate": 9.853330728032682e-05,
      "loss": 2.5215385437011717,
      "memory(GiB)": 58.3,
      "step": 9020,
      "token_acc": 0.468503937007874,
      "train_speed(iter/s)": 1.449883
    },
    {
      "epoch": 0.3866586692943747,
      "grad_norm": 3.1681206226348877,
      "learning_rate": 9.853168879230379e-05,
      "loss": 2.557341957092285,
      "memory(GiB)": 58.3,
      "step": 9025,
      "token_acc": 0.4984894259818731,
      "train_speed(iter/s)": 1.449515
    },
    {
      "epoch": 0.3868728846236237,
      "grad_norm": 4.308308124542236,
      "learning_rate": 9.853006942508369e-05,
      "loss": 2.6847625732421876,
      "memory(GiB)": 58.3,
      "step": 9030,
      "token_acc": 0.4461942257217848,
      "train_speed(iter/s)": 1.449602
    },
    {
      "epoch": 0.3870870999528726,
      "grad_norm": 3.6191303730010986,
      "learning_rate": 9.852844917869585e-05,
      "loss": 2.3625089645385744,
      "memory(GiB)": 58.3,
      "step": 9035,
      "token_acc": 0.48945147679324896,
      "train_speed(iter/s)": 1.449776
    },
    {
      "epoch": 0.3873013152821216,
      "grad_norm": 3.9500842094421387,
      "learning_rate": 9.852682805316964e-05,
      "loss": 2.590120315551758,
      "memory(GiB)": 58.3,
      "step": 9040,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.449744
    },
    {
      "epoch": 0.38751553061137056,
      "grad_norm": 3.949265480041504,
      "learning_rate": 9.852520604853442e-05,
      "loss": 2.4950660705566405,
      "memory(GiB)": 58.3,
      "step": 9045,
      "token_acc": 0.49074074074074076,
      "train_speed(iter/s)": 1.449912
    },
    {
      "epoch": 0.3877297459406195,
      "grad_norm": 3.363978624343872,
      "learning_rate": 9.852358316481955e-05,
      "loss": 2.665091323852539,
      "memory(GiB)": 58.3,
      "step": 9050,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.450148
    },
    {
      "epoch": 0.38794396126986846,
      "grad_norm": 3.606065273284912,
      "learning_rate": 9.852195940205448e-05,
      "loss": 2.7039356231689453,
      "memory(GiB)": 58.3,
      "step": 9055,
      "token_acc": 0.44366197183098594,
      "train_speed(iter/s)": 1.450109
    },
    {
      "epoch": 0.38815817659911744,
      "grad_norm": 4.576938152313232,
      "learning_rate": 9.852033476026859e-05,
      "loss": 2.7347015380859374,
      "memory(GiB)": 58.3,
      "step": 9060,
      "token_acc": 0.43728813559322033,
      "train_speed(iter/s)": 1.450198
    },
    {
      "epoch": 0.38837239192836637,
      "grad_norm": 4.903129577636719,
      "learning_rate": 9.851870923949131e-05,
      "loss": 2.5407920837402345,
      "memory(GiB)": 58.3,
      "step": 9065,
      "token_acc": 0.43137254901960786,
      "train_speed(iter/s)": 1.450246
    },
    {
      "epoch": 0.38858660725761535,
      "grad_norm": 3.3986198902130127,
      "learning_rate": 9.85170828397521e-05,
      "loss": 2.535195159912109,
      "memory(GiB)": 58.3,
      "step": 9070,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.450421
    },
    {
      "epoch": 0.3888008225868643,
      "grad_norm": 4.44728422164917,
      "learning_rate": 9.851545556108042e-05,
      "loss": 2.296782684326172,
      "memory(GiB)": 58.3,
      "step": 9075,
      "token_acc": 0.4528985507246377,
      "train_speed(iter/s)": 1.450491
    },
    {
      "epoch": 0.3890150379161133,
      "grad_norm": 4.065402984619141,
      "learning_rate": 9.851382740350576e-05,
      "loss": 2.7275468826293947,
      "memory(GiB)": 58.3,
      "step": 9080,
      "token_acc": 0.43283582089552236,
      "train_speed(iter/s)": 1.450614
    },
    {
      "epoch": 0.38922925324536223,
      "grad_norm": 4.332919120788574,
      "learning_rate": 9.851219836705761e-05,
      "loss": 2.8587520599365233,
      "memory(GiB)": 58.3,
      "step": 9085,
      "token_acc": 0.41935483870967744,
      "train_speed(iter/s)": 1.450797
    },
    {
      "epoch": 0.3894434685746112,
      "grad_norm": 4.267101287841797,
      "learning_rate": 9.851056845176547e-05,
      "loss": 2.5675411224365234,
      "memory(GiB)": 58.3,
      "step": 9090,
      "token_acc": 0.4541984732824427,
      "train_speed(iter/s)": 1.450742
    },
    {
      "epoch": 0.3896576839038602,
      "grad_norm": 4.282329082489014,
      "learning_rate": 9.850893765765887e-05,
      "loss": 2.5712690353393555,
      "memory(GiB)": 58.3,
      "step": 9095,
      "token_acc": 0.4134275618374558,
      "train_speed(iter/s)": 1.450987
    },
    {
      "epoch": 0.3898718992331091,
      "grad_norm": 4.885793685913086,
      "learning_rate": 9.850730598476737e-05,
      "loss": 2.836735725402832,
      "memory(GiB)": 58.3,
      "step": 9100,
      "token_acc": 0.45387453874538747,
      "train_speed(iter/s)": 1.451068
    },
    {
      "epoch": 0.3900861145623581,
      "grad_norm": 3.5165767669677734,
      "learning_rate": 9.850567343312051e-05,
      "loss": 2.3478668212890623,
      "memory(GiB)": 58.3,
      "step": 9105,
      "token_acc": 0.545774647887324,
      "train_speed(iter/s)": 1.45102
    },
    {
      "epoch": 0.39030032989160707,
      "grad_norm": 3.6581642627716064,
      "learning_rate": 9.850404000274789e-05,
      "loss": 2.976247024536133,
      "memory(GiB)": 58.3,
      "step": 9110,
      "token_acc": 0.3993174061433447,
      "train_speed(iter/s)": 1.451054
    },
    {
      "epoch": 0.390514545220856,
      "grad_norm": 3.52714204788208,
      "learning_rate": 9.850240569367908e-05,
      "loss": 2.715854263305664,
      "memory(GiB)": 58.3,
      "step": 9115,
      "token_acc": 0.43515850144092216,
      "train_speed(iter/s)": 1.451117
    },
    {
      "epoch": 0.39072876055010497,
      "grad_norm": 3.781163454055786,
      "learning_rate": 9.850077050594368e-05,
      "loss": 2.555784225463867,
      "memory(GiB)": 58.3,
      "step": 9120,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.451243
    },
    {
      "epoch": 0.39094297587935395,
      "grad_norm": 4.061525821685791,
      "learning_rate": 9.849913443957133e-05,
      "loss": 2.7376424789428713,
      "memory(GiB)": 58.3,
      "step": 9125,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.451207
    },
    {
      "epoch": 0.3911571912086029,
      "grad_norm": 4.145669460296631,
      "learning_rate": 9.849749749459167e-05,
      "loss": 2.382442092895508,
      "memory(GiB)": 58.3,
      "step": 9130,
      "token_acc": 0.5129032258064516,
      "train_speed(iter/s)": 1.451246
    },
    {
      "epoch": 0.39137140653785185,
      "grad_norm": 5.671361923217773,
      "learning_rate": 9.849585967103434e-05,
      "loss": 2.8731260299682617,
      "memory(GiB)": 58.3,
      "step": 9135,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.451174
    },
    {
      "epoch": 0.39158562186710083,
      "grad_norm": 3.527975559234619,
      "learning_rate": 9.849422096892902e-05,
      "loss": 2.7227451324462892,
      "memory(GiB)": 58.3,
      "step": 9140,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.451381
    },
    {
      "epoch": 0.39179983719634975,
      "grad_norm": 5.232457160949707,
      "learning_rate": 9.84925813883054e-05,
      "loss": 2.906509780883789,
      "memory(GiB)": 58.3,
      "step": 9145,
      "token_acc": 0.4144144144144144,
      "train_speed(iter/s)": 1.451339
    },
    {
      "epoch": 0.39201405252559873,
      "grad_norm": 4.1620635986328125,
      "learning_rate": 9.849094092919318e-05,
      "loss": 2.8824880599975584,
      "memory(GiB)": 58.3,
      "step": 9150,
      "token_acc": 0.4209039548022599,
      "train_speed(iter/s)": 1.451499
    },
    {
      "epoch": 0.3922282678548477,
      "grad_norm": 4.052217960357666,
      "learning_rate": 9.848929959162207e-05,
      "loss": 2.8716266632080076,
      "memory(GiB)": 58.3,
      "step": 9155,
      "token_acc": 0.439873417721519,
      "train_speed(iter/s)": 1.451558
    },
    {
      "epoch": 0.39244248318409664,
      "grad_norm": 3.931018590927124,
      "learning_rate": 9.848765737562183e-05,
      "loss": 2.4714435577392577,
      "memory(GiB)": 58.3,
      "step": 9160,
      "token_acc": 0.5063291139240507,
      "train_speed(iter/s)": 1.451644
    },
    {
      "epoch": 0.3926566985133456,
      "grad_norm": 3.8844563961029053,
      "learning_rate": 9.848601428122217e-05,
      "loss": 2.8242166519165037,
      "memory(GiB)": 58.3,
      "step": 9165,
      "token_acc": 0.43209876543209874,
      "train_speed(iter/s)": 1.451748
    },
    {
      "epoch": 0.3928709138425946,
      "grad_norm": 5.687803268432617,
      "learning_rate": 9.848437030845288e-05,
      "loss": 2.89019718170166,
      "memory(GiB)": 58.3,
      "step": 9170,
      "token_acc": 0.44609665427509293,
      "train_speed(iter/s)": 1.451899
    },
    {
      "epoch": 0.3930851291718435,
      "grad_norm": 5.416097164154053,
      "learning_rate": 9.848272545734374e-05,
      "loss": 2.4476722717285155,
      "memory(GiB)": 58.3,
      "step": 9175,
      "token_acc": 0.4901315789473684,
      "train_speed(iter/s)": 1.451799
    },
    {
      "epoch": 0.3932993445010925,
      "grad_norm": 3.545675754547119,
      "learning_rate": 9.848107972792455e-05,
      "loss": 2.467362976074219,
      "memory(GiB)": 58.3,
      "step": 9180,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.451979
    },
    {
      "epoch": 0.3935135598303415,
      "grad_norm": 5.395636081695557,
      "learning_rate": 9.847943312022511e-05,
      "loss": 2.738362121582031,
      "memory(GiB)": 58.3,
      "step": 9185,
      "token_acc": 0.4155405405405405,
      "train_speed(iter/s)": 1.452146
    },
    {
      "epoch": 0.3937277751595904,
      "grad_norm": 4.503758430480957,
      "learning_rate": 9.847778563427529e-05,
      "loss": 2.4091026306152346,
      "memory(GiB)": 58.3,
      "step": 9190,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.452289
    },
    {
      "epoch": 0.3939419904888394,
      "grad_norm": 4.088322639465332,
      "learning_rate": 9.84761372701049e-05,
      "loss": 2.3184776306152344,
      "memory(GiB)": 58.3,
      "step": 9195,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.452244
    },
    {
      "epoch": 0.39415620581808836,
      "grad_norm": 3.4466676712036133,
      "learning_rate": 9.847448802774379e-05,
      "loss": 2.609083557128906,
      "memory(GiB)": 58.3,
      "step": 9200,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.452226
    },
    {
      "epoch": 0.3943704211473373,
      "grad_norm": 3.4711740016937256,
      "learning_rate": 9.847283790722187e-05,
      "loss": 2.7564361572265623,
      "memory(GiB)": 58.3,
      "step": 9205,
      "token_acc": 0.4405144694533762,
      "train_speed(iter/s)": 1.452387
    },
    {
      "epoch": 0.39458463647658626,
      "grad_norm": 5.136669158935547,
      "learning_rate": 9.847118690856903e-05,
      "loss": 2.4305177688598634,
      "memory(GiB)": 58.3,
      "step": 9210,
      "token_acc": 0.46381578947368424,
      "train_speed(iter/s)": 1.452453
    },
    {
      "epoch": 0.39479885180583524,
      "grad_norm": 3.6338794231414795,
      "learning_rate": 9.846953503181515e-05,
      "loss": 2.3469114303588867,
      "memory(GiB)": 58.3,
      "step": 9215,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.452559
    },
    {
      "epoch": 0.39501306713508416,
      "grad_norm": 3.952202796936035,
      "learning_rate": 9.846788227699017e-05,
      "loss": 2.719685745239258,
      "memory(GiB)": 58.3,
      "step": 9220,
      "token_acc": 0.41590214067278286,
      "train_speed(iter/s)": 1.452582
    },
    {
      "epoch": 0.39522728246433314,
      "grad_norm": 6.918583869934082,
      "learning_rate": 9.846622864412406e-05,
      "loss": 2.6086856842041017,
      "memory(GiB)": 58.3,
      "step": 9225,
      "token_acc": 0.45878136200716846,
      "train_speed(iter/s)": 1.452516
    },
    {
      "epoch": 0.3954414977935821,
      "grad_norm": 4.2701334953308105,
      "learning_rate": 9.846457413324675e-05,
      "loss": 2.6882152557373047,
      "memory(GiB)": 58.3,
      "step": 9230,
      "token_acc": 0.43097643097643096,
      "train_speed(iter/s)": 1.45238
    },
    {
      "epoch": 0.39565571312283104,
      "grad_norm": 3.582998037338257,
      "learning_rate": 9.84629187443882e-05,
      "loss": 2.6443161010742187,
      "memory(GiB)": 58.3,
      "step": 9235,
      "token_acc": 0.438871473354232,
      "train_speed(iter/s)": 1.452554
    },
    {
      "epoch": 0.39586992845208,
      "grad_norm": 4.341336250305176,
      "learning_rate": 9.846126247757843e-05,
      "loss": 2.3565851211547852,
      "memory(GiB)": 58.3,
      "step": 9240,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.452753
    },
    {
      "epoch": 0.396084143781329,
      "grad_norm": 4.659677982330322,
      "learning_rate": 9.845960533284742e-05,
      "loss": 2.480521392822266,
      "memory(GiB)": 58.3,
      "step": 9245,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.452623
    },
    {
      "epoch": 0.396298359110578,
      "grad_norm": 3.5144901275634766,
      "learning_rate": 9.84579473102252e-05,
      "loss": 1.987625503540039,
      "memory(GiB)": 58.3,
      "step": 9250,
      "token_acc": 0.5800865800865801,
      "train_speed(iter/s)": 1.452228
    },
    {
      "epoch": 0.3965125744398269,
      "grad_norm": 4.90177059173584,
      "learning_rate": 9.845628840974182e-05,
      "loss": 2.4235244750976563,
      "memory(GiB)": 58.3,
      "step": 9255,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.452361
    },
    {
      "epoch": 0.3967267897690759,
      "grad_norm": 4.903687477111816,
      "learning_rate": 9.84546286314273e-05,
      "loss": 2.8135444641113283,
      "memory(GiB)": 58.3,
      "step": 9260,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.452342
    },
    {
      "epoch": 0.39694100509832486,
      "grad_norm": 4.126767635345459,
      "learning_rate": 9.845296797531173e-05,
      "loss": 2.4977348327636717,
      "memory(GiB)": 58.3,
      "step": 9265,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.452403
    },
    {
      "epoch": 0.3971552204275738,
      "grad_norm": 3.942574977874756,
      "learning_rate": 9.845130644142522e-05,
      "loss": 2.568085861206055,
      "memory(GiB)": 58.3,
      "step": 9270,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.45252
    },
    {
      "epoch": 0.39736943575682276,
      "grad_norm": 4.052740573883057,
      "learning_rate": 9.844964402979781e-05,
      "loss": 2.3654411315917967,
      "memory(GiB)": 58.3,
      "step": 9275,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.452639
    },
    {
      "epoch": 0.39758365108607174,
      "grad_norm": 4.126075267791748,
      "learning_rate": 9.844798074045967e-05,
      "loss": 2.4733795166015624,
      "memory(GiB)": 58.3,
      "step": 9280,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.452459
    },
    {
      "epoch": 0.39779786641532067,
      "grad_norm": 5.828798294067383,
      "learning_rate": 9.84463165734409e-05,
      "loss": 2.325762176513672,
      "memory(GiB)": 58.3,
      "step": 9285,
      "token_acc": 0.5198675496688742,
      "train_speed(iter/s)": 1.452431
    },
    {
      "epoch": 0.39801208174456965,
      "grad_norm": 3.830606698989868,
      "learning_rate": 9.844465152877167e-05,
      "loss": 2.5024595260620117,
      "memory(GiB)": 58.3,
      "step": 9290,
      "token_acc": 0.4645390070921986,
      "train_speed(iter/s)": 1.452523
    },
    {
      "epoch": 0.3982262970738186,
      "grad_norm": 4.568037986755371,
      "learning_rate": 9.844298560648213e-05,
      "loss": 2.521292877197266,
      "memory(GiB)": 58.3,
      "step": 9295,
      "token_acc": 0.45323741007194246,
      "train_speed(iter/s)": 1.452491
    },
    {
      "epoch": 0.39844051240306755,
      "grad_norm": 3.5581140518188477,
      "learning_rate": 9.844131880660246e-05,
      "loss": 2.6647274017333986,
      "memory(GiB)": 58.3,
      "step": 9300,
      "token_acc": 0.4752851711026616,
      "train_speed(iter/s)": 1.452494
    },
    {
      "epoch": 0.3986547277323165,
      "grad_norm": 4.720825672149658,
      "learning_rate": 9.843965112916285e-05,
      "loss": 2.689658355712891,
      "memory(GiB)": 58.3,
      "step": 9305,
      "token_acc": 0.422680412371134,
      "train_speed(iter/s)": 1.45263
    },
    {
      "epoch": 0.3988689430615655,
      "grad_norm": 3.6050236225128174,
      "learning_rate": 9.843798257419352e-05,
      "loss": 2.657181167602539,
      "memory(GiB)": 58.3,
      "step": 9310,
      "token_acc": 0.46394984326018807,
      "train_speed(iter/s)": 1.452611
    },
    {
      "epoch": 0.39908315839081443,
      "grad_norm": 4.149864196777344,
      "learning_rate": 9.843631314172471e-05,
      "loss": 2.538443946838379,
      "memory(GiB)": 58.3,
      "step": 9315,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.452678
    },
    {
      "epoch": 0.3992973737200634,
      "grad_norm": 5.8810858726501465,
      "learning_rate": 9.843464283178665e-05,
      "loss": 2.6902797698974608,
      "memory(GiB)": 58.3,
      "step": 9320,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.452744
    },
    {
      "epoch": 0.3995115890493124,
      "grad_norm": 5.2606306076049805,
      "learning_rate": 9.843297164440959e-05,
      "loss": 2.446046829223633,
      "memory(GiB)": 58.3,
      "step": 9325,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.452755
    },
    {
      "epoch": 0.3997258043785613,
      "grad_norm": 3.9002749919891357,
      "learning_rate": 9.843129957962381e-05,
      "loss": 2.461764907836914,
      "memory(GiB)": 58.3,
      "step": 9330,
      "token_acc": 0.4511494252873563,
      "train_speed(iter/s)": 1.452673
    },
    {
      "epoch": 0.3999400197078103,
      "grad_norm": 3.7831766605377197,
      "learning_rate": 9.842962663745963e-05,
      "loss": 2.641637420654297,
      "memory(GiB)": 58.3,
      "step": 9335,
      "token_acc": 0.44299674267100975,
      "train_speed(iter/s)": 1.452426
    },
    {
      "epoch": 0.40015423503705927,
      "grad_norm": 4.262258529663086,
      "learning_rate": 9.842795281794732e-05,
      "loss": 2.7654655456542967,
      "memory(GiB)": 58.3,
      "step": 9340,
      "token_acc": 0.42718446601941745,
      "train_speed(iter/s)": 1.452414
    },
    {
      "epoch": 0.4003684503663082,
      "grad_norm": 3.4499847888946533,
      "learning_rate": 9.84262781211172e-05,
      "loss": 2.6401214599609375,
      "memory(GiB)": 58.3,
      "step": 9345,
      "token_acc": 0.41216216216216217,
      "train_speed(iter/s)": 1.452275
    },
    {
      "epoch": 0.40058266569555717,
      "grad_norm": 4.302866458892822,
      "learning_rate": 9.842460254699963e-05,
      "loss": 2.5645931243896483,
      "memory(GiB)": 58.3,
      "step": 9350,
      "token_acc": 0.484,
      "train_speed(iter/s)": 1.45239
    },
    {
      "epoch": 0.40079688102480615,
      "grad_norm": 4.590943813323975,
      "learning_rate": 9.842292609562498e-05,
      "loss": 2.5758602142333986,
      "memory(GiB)": 58.3,
      "step": 9355,
      "token_acc": 0.4460431654676259,
      "train_speed(iter/s)": 1.452489
    },
    {
      "epoch": 0.4010110963540551,
      "grad_norm": 3.751528263092041,
      "learning_rate": 9.84212487670236e-05,
      "loss": 2.50805778503418,
      "memory(GiB)": 58.3,
      "step": 9360,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.452515
    },
    {
      "epoch": 0.40122531168330405,
      "grad_norm": 2.8437740802764893,
      "learning_rate": 9.841957056122584e-05,
      "loss": 2.6894599914550783,
      "memory(GiB)": 58.3,
      "step": 9365,
      "token_acc": 0.4631268436578171,
      "train_speed(iter/s)": 1.452476
    },
    {
      "epoch": 0.40143952701255303,
      "grad_norm": 4.310125827789307,
      "learning_rate": 9.841789147826217e-05,
      "loss": 2.7040168762207033,
      "memory(GiB)": 58.3,
      "step": 9370,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.452376
    },
    {
      "epoch": 0.40165374234180196,
      "grad_norm": 3.55946683883667,
      "learning_rate": 9.841621151816296e-05,
      "loss": 2.5776132583618163,
      "memory(GiB)": 58.3,
      "step": 9375,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.452496
    },
    {
      "epoch": 0.40186795767105093,
      "grad_norm": 3.3330557346343994,
      "learning_rate": 9.841453068095867e-05,
      "loss": 2.650056838989258,
      "memory(GiB)": 58.3,
      "step": 9380,
      "token_acc": 0.397887323943662,
      "train_speed(iter/s)": 1.452514
    },
    {
      "epoch": 0.4020821730002999,
      "grad_norm": 4.043461322784424,
      "learning_rate": 9.841284896667973e-05,
      "loss": 2.3511322021484373,
      "memory(GiB)": 58.3,
      "step": 9385,
      "token_acc": 0.524,
      "train_speed(iter/s)": 1.452401
    },
    {
      "epoch": 0.40229638832954884,
      "grad_norm": 3.820551633834839,
      "learning_rate": 9.841116637535662e-05,
      "loss": 2.6962852478027344,
      "memory(GiB)": 58.3,
      "step": 9390,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.452407
    },
    {
      "epoch": 0.4025106036587978,
      "grad_norm": 3.8388638496398926,
      "learning_rate": 9.840948290701982e-05,
      "loss": 2.6646074295043944,
      "memory(GiB)": 58.3,
      "step": 9395,
      "token_acc": 0.4697986577181208,
      "train_speed(iter/s)": 1.452536
    },
    {
      "epoch": 0.4027248189880468,
      "grad_norm": 4.511655330657959,
      "learning_rate": 9.840779856169982e-05,
      "loss": 2.613489532470703,
      "memory(GiB)": 58.3,
      "step": 9400,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.4525
    },
    {
      "epoch": 0.4029390343172957,
      "grad_norm": 5.018468379974365,
      "learning_rate": 9.840611333942715e-05,
      "loss": 2.9536693572998045,
      "memory(GiB)": 58.3,
      "step": 9405,
      "token_acc": 0.44014084507042256,
      "train_speed(iter/s)": 1.452561
    },
    {
      "epoch": 0.4031532496465447,
      "grad_norm": 3.4255735874176025,
      "learning_rate": 9.840442724023232e-05,
      "loss": 2.7968017578125,
      "memory(GiB)": 58.3,
      "step": 9410,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.452641
    },
    {
      "epoch": 0.4033674649757937,
      "grad_norm": 6.153944969177246,
      "learning_rate": 9.840274026414589e-05,
      "loss": 2.592268943786621,
      "memory(GiB)": 58.3,
      "step": 9415,
      "token_acc": 0.4810126582278481,
      "train_speed(iter/s)": 1.452797
    },
    {
      "epoch": 0.40358168030504266,
      "grad_norm": 3.660299301147461,
      "learning_rate": 9.840105241119841e-05,
      "loss": 2.437905502319336,
      "memory(GiB)": 58.3,
      "step": 9420,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.452994
    },
    {
      "epoch": 0.4037958956342916,
      "grad_norm": 4.591060161590576,
      "learning_rate": 9.839936368142046e-05,
      "loss": 2.7226917266845705,
      "memory(GiB)": 58.3,
      "step": 9425,
      "token_acc": 0.45045045045045046,
      "train_speed(iter/s)": 1.453125
    },
    {
      "epoch": 0.40401011096354056,
      "grad_norm": 3.767965078353882,
      "learning_rate": 9.839767407484264e-05,
      "loss": 2.5389484405517577,
      "memory(GiB)": 58.3,
      "step": 9430,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.453173
    },
    {
      "epoch": 0.40422432629278954,
      "grad_norm": 4.654022693634033,
      "learning_rate": 9.839598359149556e-05,
      "loss": 2.8059436798095705,
      "memory(GiB)": 58.3,
      "step": 9435,
      "token_acc": 0.4511784511784512,
      "train_speed(iter/s)": 1.45322
    },
    {
      "epoch": 0.40443854162203846,
      "grad_norm": 3.7642242908477783,
      "learning_rate": 9.839429223140981e-05,
      "loss": 2.648666000366211,
      "memory(GiB)": 58.3,
      "step": 9440,
      "token_acc": 0.44886363636363635,
      "train_speed(iter/s)": 1.453134
    },
    {
      "epoch": 0.40465275695128744,
      "grad_norm": 3.5528132915496826,
      "learning_rate": 9.839259999461609e-05,
      "loss": 2.6834552764892576,
      "memory(GiB)": 58.3,
      "step": 9445,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.452979
    },
    {
      "epoch": 0.4048669722805364,
      "grad_norm": 3.637704372406006,
      "learning_rate": 9.839090688114501e-05,
      "loss": 2.749816131591797,
      "memory(GiB)": 58.3,
      "step": 9450,
      "token_acc": 0.43174603174603177,
      "train_speed(iter/s)": 1.452977
    },
    {
      "epoch": 0.40508118760978534,
      "grad_norm": 4.624016761779785,
      "learning_rate": 9.838921289102726e-05,
      "loss": 2.6001607894897463,
      "memory(GiB)": 58.3,
      "step": 9455,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.452984
    },
    {
      "epoch": 0.4052954029390343,
      "grad_norm": 4.223477363586426,
      "learning_rate": 9.838751802429352e-05,
      "loss": 2.517749214172363,
      "memory(GiB)": 58.3,
      "step": 9460,
      "token_acc": 0.4423791821561338,
      "train_speed(iter/s)": 1.452943
    },
    {
      "epoch": 0.4055096182682833,
      "grad_norm": 3.8644094467163086,
      "learning_rate": 9.83858222809745e-05,
      "loss": 2.6502376556396485,
      "memory(GiB)": 58.3,
      "step": 9465,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.452595
    },
    {
      "epoch": 0.4057238335975322,
      "grad_norm": 3.79787015914917,
      "learning_rate": 9.838412566110094e-05,
      "loss": 2.7505889892578126,
      "memory(GiB)": 58.3,
      "step": 9470,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.452699
    },
    {
      "epoch": 0.4059380489267812,
      "grad_norm": 4.230555534362793,
      "learning_rate": 9.838242816470353e-05,
      "loss": 2.776617431640625,
      "memory(GiB)": 58.3,
      "step": 9475,
      "token_acc": 0.43197278911564624,
      "train_speed(iter/s)": 1.45274
    },
    {
      "epoch": 0.4061522642560302,
      "grad_norm": 3.367751121520996,
      "learning_rate": 9.838072979181306e-05,
      "loss": 2.65602970123291,
      "memory(GiB)": 58.3,
      "step": 9480,
      "token_acc": 0.4222222222222222,
      "train_speed(iter/s)": 1.452776
    },
    {
      "epoch": 0.4063664795852791,
      "grad_norm": 4.935830593109131,
      "learning_rate": 9.83790305424603e-05,
      "loss": 2.667824935913086,
      "memory(GiB)": 58.3,
      "step": 9485,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.452834
    },
    {
      "epoch": 0.4065806949145281,
      "grad_norm": 4.7056965827941895,
      "learning_rate": 9.837733041667598e-05,
      "loss": 2.516107749938965,
      "memory(GiB)": 58.3,
      "step": 9490,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.452742
    },
    {
      "epoch": 0.40679491024377706,
      "grad_norm": 5.4926252365112305,
      "learning_rate": 9.837562941449096e-05,
      "loss": 2.472407913208008,
      "memory(GiB)": 58.3,
      "step": 9495,
      "token_acc": 0.44649446494464945,
      "train_speed(iter/s)": 1.452924
    },
    {
      "epoch": 0.407009125573026,
      "grad_norm": 3.6378047466278076,
      "learning_rate": 9.837392753593604e-05,
      "loss": 2.4419391632080076,
      "memory(GiB)": 58.3,
      "step": 9500,
      "token_acc": 0.5328185328185329,
      "train_speed(iter/s)": 1.452911
    },
    {
      "epoch": 0.407009125573026,
      "eval_loss": 2.20744252204895,
      "eval_runtime": 13.7695,
      "eval_samples_per_second": 7.262,
      "eval_steps_per_second": 7.262,
      "eval_token_acc": 0.47275204359673023,
      "step": 9500
    },
    {
      "epoch": 0.40722334090227497,
      "grad_norm": 4.563931941986084,
      "learning_rate": 9.837222478104205e-05,
      "loss": 2.410663032531738,
      "memory(GiB)": 58.3,
      "step": 9505,
      "token_acc": 0.4746450304259635,
      "train_speed(iter/s)": 1.449619
    },
    {
      "epoch": 0.40743755623152395,
      "grad_norm": 4.4739580154418945,
      "learning_rate": 9.837052114983982e-05,
      "loss": 2.921822738647461,
      "memory(GiB)": 58.3,
      "step": 9510,
      "token_acc": 0.44571428571428573,
      "train_speed(iter/s)": 1.449609
    },
    {
      "epoch": 0.40765177156077287,
      "grad_norm": 3.862518787384033,
      "learning_rate": 9.836881664236021e-05,
      "loss": 2.6591524124145507,
      "memory(GiB)": 58.3,
      "step": 9515,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.449542
    },
    {
      "epoch": 0.40786598689002185,
      "grad_norm": 3.1519999504089355,
      "learning_rate": 9.836711125863413e-05,
      "loss": 2.584940528869629,
      "memory(GiB)": 58.3,
      "step": 9520,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.449509
    },
    {
      "epoch": 0.4080802022192708,
      "grad_norm": 6.747684955596924,
      "learning_rate": 9.836540499869244e-05,
      "loss": 2.619213676452637,
      "memory(GiB)": 58.3,
      "step": 9525,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.449433
    },
    {
      "epoch": 0.40829441754851975,
      "grad_norm": 3.8937957286834717,
      "learning_rate": 9.83636978625661e-05,
      "loss": 2.3868877410888674,
      "memory(GiB)": 58.3,
      "step": 9530,
      "token_acc": 0.4637223974763407,
      "train_speed(iter/s)": 1.449305
    },
    {
      "epoch": 0.40850863287776873,
      "grad_norm": 3.600677013397217,
      "learning_rate": 9.836198985028597e-05,
      "loss": 2.573666572570801,
      "memory(GiB)": 58.3,
      "step": 9535,
      "token_acc": 0.46688741721854304,
      "train_speed(iter/s)": 1.449465
    },
    {
      "epoch": 0.4087228482070177,
      "grad_norm": 4.229767322540283,
      "learning_rate": 9.836028096188306e-05,
      "loss": 2.581707000732422,
      "memory(GiB)": 58.3,
      "step": 9540,
      "token_acc": 0.462882096069869,
      "train_speed(iter/s)": 1.449476
    },
    {
      "epoch": 0.40893706353626663,
      "grad_norm": 6.386398792266846,
      "learning_rate": 9.835857119738827e-05,
      "loss": 2.889179229736328,
      "memory(GiB)": 58.3,
      "step": 9545,
      "token_acc": 0.41379310344827586,
      "train_speed(iter/s)": 1.449569
    },
    {
      "epoch": 0.4091512788655156,
      "grad_norm": 4.296133995056152,
      "learning_rate": 9.83568605568326e-05,
      "loss": 2.600038528442383,
      "memory(GiB)": 58.3,
      "step": 9550,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.449703
    },
    {
      "epoch": 0.4093654941947646,
      "grad_norm": 3.547651767730713,
      "learning_rate": 9.835514904024705e-05,
      "loss": 2.411071014404297,
      "memory(GiB)": 58.3,
      "step": 9555,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.449956
    },
    {
      "epoch": 0.4095797095240135,
      "grad_norm": 4.404221534729004,
      "learning_rate": 9.83534366476626e-05,
      "loss": 2.6571138381958006,
      "memory(GiB)": 58.3,
      "step": 9560,
      "token_acc": 0.4294478527607362,
      "train_speed(iter/s)": 1.449929
    },
    {
      "epoch": 0.4097939248532625,
      "grad_norm": 3.6686346530914307,
      "learning_rate": 9.83517233791103e-05,
      "loss": 2.6325796127319334,
      "memory(GiB)": 58.3,
      "step": 9565,
      "token_acc": 0.426056338028169,
      "train_speed(iter/s)": 1.449801
    },
    {
      "epoch": 0.41000814018251147,
      "grad_norm": 3.781003952026367,
      "learning_rate": 9.835000923462117e-05,
      "loss": 2.7300729751586914,
      "memory(GiB)": 58.3,
      "step": 9570,
      "token_acc": 0.4334470989761092,
      "train_speed(iter/s)": 1.449634
    },
    {
      "epoch": 0.4102223555117604,
      "grad_norm": 4.024383068084717,
      "learning_rate": 9.834829421422627e-05,
      "loss": 2.4482383728027344,
      "memory(GiB)": 58.3,
      "step": 9575,
      "token_acc": 0.494949494949495,
      "train_speed(iter/s)": 1.449674
    },
    {
      "epoch": 0.4104365708410094,
      "grad_norm": 3.7689874172210693,
      "learning_rate": 9.834657831795666e-05,
      "loss": 2.5864463806152345,
      "memory(GiB)": 58.3,
      "step": 9580,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.449801
    },
    {
      "epoch": 0.41065078617025835,
      "grad_norm": 4.039371490478516,
      "learning_rate": 9.834486154584342e-05,
      "loss": 2.506942367553711,
      "memory(GiB)": 58.3,
      "step": 9585,
      "token_acc": 0.5259515570934256,
      "train_speed(iter/s)": 1.44988
    },
    {
      "epoch": 0.41086500149950733,
      "grad_norm": 5.111746311187744,
      "learning_rate": 9.834314389791767e-05,
      "loss": 2.5287267684936525,
      "memory(GiB)": 58.3,
      "step": 9590,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.450055
    },
    {
      "epoch": 0.41107921682875626,
      "grad_norm": 3.6341724395751953,
      "learning_rate": 9.834142537421053e-05,
      "loss": 2.613294982910156,
      "memory(GiB)": 58.3,
      "step": 9595,
      "token_acc": 0.445141065830721,
      "train_speed(iter/s)": 1.450211
    },
    {
      "epoch": 0.41129343215800523,
      "grad_norm": 4.9661359786987305,
      "learning_rate": 9.833970597475311e-05,
      "loss": 2.364989471435547,
      "memory(GiB)": 58.3,
      "step": 9600,
      "token_acc": 0.4942084942084942,
      "train_speed(iter/s)": 1.450216
    },
    {
      "epoch": 0.4115076474872542,
      "grad_norm": 3.4044036865234375,
      "learning_rate": 9.833798569957657e-05,
      "loss": 2.380361557006836,
      "memory(GiB)": 58.3,
      "step": 9605,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.450301
    },
    {
      "epoch": 0.41172186281650314,
      "grad_norm": 2.8980321884155273,
      "learning_rate": 9.833626454871208e-05,
      "loss": 2.6926889419555664,
      "memory(GiB)": 58.3,
      "step": 9610,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.450566
    },
    {
      "epoch": 0.4119360781457521,
      "grad_norm": 3.441521644592285,
      "learning_rate": 9.833454252219082e-05,
      "loss": 2.644577217102051,
      "memory(GiB)": 58.3,
      "step": 9615,
      "token_acc": 0.49814126394052044,
      "train_speed(iter/s)": 1.450409
    },
    {
      "epoch": 0.4121502934750011,
      "grad_norm": 4.102413654327393,
      "learning_rate": 9.833281962004397e-05,
      "loss": 2.4470069885253904,
      "memory(GiB)": 58.3,
      "step": 9620,
      "token_acc": 0.5175097276264592,
      "train_speed(iter/s)": 1.450341
    },
    {
      "epoch": 0.41236450880425,
      "grad_norm": 4.925978183746338,
      "learning_rate": 9.833109584230275e-05,
      "loss": 2.8542598724365233,
      "memory(GiB)": 58.3,
      "step": 9625,
      "token_acc": 0.44715447154471544,
      "train_speed(iter/s)": 1.450291
    },
    {
      "epoch": 0.412578724133499,
      "grad_norm": 3.057582139968872,
      "learning_rate": 9.832937118899842e-05,
      "loss": 2.4771472930908205,
      "memory(GiB)": 58.3,
      "step": 9630,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.450556
    },
    {
      "epoch": 0.412792939462748,
      "grad_norm": 4.964419841766357,
      "learning_rate": 9.832764566016216e-05,
      "loss": 2.4388219833374025,
      "memory(GiB)": 58.3,
      "step": 9635,
      "token_acc": 0.5145985401459854,
      "train_speed(iter/s)": 1.450656
    },
    {
      "epoch": 0.4130071547919969,
      "grad_norm": 4.488737106323242,
      "learning_rate": 9.832591925582527e-05,
      "loss": 2.5482051849365233,
      "memory(GiB)": 58.3,
      "step": 9640,
      "token_acc": 0.4981549815498155,
      "train_speed(iter/s)": 1.450624
    },
    {
      "epoch": 0.4132213701212459,
      "grad_norm": 5.323754787445068,
      "learning_rate": 9.832419197601903e-05,
      "loss": 2.298923873901367,
      "memory(GiB)": 58.3,
      "step": 9645,
      "token_acc": 0.4779116465863454,
      "train_speed(iter/s)": 1.450678
    },
    {
      "epoch": 0.41343558545049486,
      "grad_norm": 4.610477924346924,
      "learning_rate": 9.832246382077471e-05,
      "loss": 2.7490882873535156,
      "memory(GiB)": 58.3,
      "step": 9650,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.450778
    },
    {
      "epoch": 0.4136498007797438,
      "grad_norm": 5.109628677368164,
      "learning_rate": 9.832073479012364e-05,
      "loss": 2.6343645095825194,
      "memory(GiB)": 58.3,
      "step": 9655,
      "token_acc": 0.45136186770428016,
      "train_speed(iter/s)": 1.450964
    },
    {
      "epoch": 0.41386401610899276,
      "grad_norm": 4.2545390129089355,
      "learning_rate": 9.83190048840971e-05,
      "loss": 2.8213424682617188,
      "memory(GiB)": 58.3,
      "step": 9660,
      "token_acc": 0.4553846153846154,
      "train_speed(iter/s)": 1.450974
    },
    {
      "epoch": 0.41407823143824174,
      "grad_norm": 3.911921501159668,
      "learning_rate": 9.83172741027265e-05,
      "loss": 2.620610237121582,
      "memory(GiB)": 58.3,
      "step": 9665,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.450997
    },
    {
      "epoch": 0.41429244676749066,
      "grad_norm": 3.68625545501709,
      "learning_rate": 9.831554244604313e-05,
      "loss": 2.453909492492676,
      "memory(GiB)": 58.3,
      "step": 9670,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.451004
    },
    {
      "epoch": 0.41450666209673964,
      "grad_norm": 4.247152328491211,
      "learning_rate": 9.831380991407841e-05,
      "loss": 2.911936569213867,
      "memory(GiB)": 58.3,
      "step": 9675,
      "token_acc": 0.4398496240601504,
      "train_speed(iter/s)": 1.45096
    },
    {
      "epoch": 0.4147208774259886,
      "grad_norm": 4.288780212402344,
      "learning_rate": 9.831207650686367e-05,
      "loss": 2.635361099243164,
      "memory(GiB)": 58.3,
      "step": 9680,
      "token_acc": 0.4391891891891892,
      "train_speed(iter/s)": 1.451075
    },
    {
      "epoch": 0.41493509275523754,
      "grad_norm": 3.842052459716797,
      "learning_rate": 9.831034222443037e-05,
      "loss": 2.533888244628906,
      "memory(GiB)": 58.3,
      "step": 9685,
      "token_acc": 0.524,
      "train_speed(iter/s)": 1.451013
    },
    {
      "epoch": 0.4151493080844865,
      "grad_norm": 3.9143149852752686,
      "learning_rate": 9.830860706680989e-05,
      "loss": 2.6112188339233398,
      "memory(GiB)": 58.3,
      "step": 9690,
      "token_acc": 0.49691358024691357,
      "train_speed(iter/s)": 1.451177
    },
    {
      "epoch": 0.4153635234137355,
      "grad_norm": 4.415402889251709,
      "learning_rate": 9.830687103403367e-05,
      "loss": 2.455311584472656,
      "memory(GiB)": 58.3,
      "step": 9695,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.451305
    },
    {
      "epoch": 0.4155777387429844,
      "grad_norm": 3.138246536254883,
      "learning_rate": 9.830513412613318e-05,
      "loss": 2.635904884338379,
      "memory(GiB)": 58.3,
      "step": 9700,
      "token_acc": 0.4420289855072464,
      "train_speed(iter/s)": 1.45111
    },
    {
      "epoch": 0.4157919540722334,
      "grad_norm": 5.393638610839844,
      "learning_rate": 9.830339634313985e-05,
      "loss": 2.5539560317993164,
      "memory(GiB)": 58.3,
      "step": 9705,
      "token_acc": 0.4887459807073955,
      "train_speed(iter/s)": 1.451068
    },
    {
      "epoch": 0.4160061694014824,
      "grad_norm": 3.9030025005340576,
      "learning_rate": 9.830165768508519e-05,
      "loss": 2.660310173034668,
      "memory(GiB)": 58.3,
      "step": 9710,
      "token_acc": 0.47413793103448276,
      "train_speed(iter/s)": 1.450961
    },
    {
      "epoch": 0.4162203847307313,
      "grad_norm": 6.049018859863281,
      "learning_rate": 9.82999181520007e-05,
      "loss": 2.5478237152099608,
      "memory(GiB)": 58.3,
      "step": 9715,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.451001
    },
    {
      "epoch": 0.4164346000599803,
      "grad_norm": 3.858489751815796,
      "learning_rate": 9.829817774391788e-05,
      "loss": 2.5838314056396485,
      "memory(GiB)": 58.3,
      "step": 9720,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.451008
    },
    {
      "epoch": 0.41664881538922927,
      "grad_norm": 4.7568440437316895,
      "learning_rate": 9.829643646086826e-05,
      "loss": 2.889063262939453,
      "memory(GiB)": 58.3,
      "step": 9725,
      "token_acc": 0.41924398625429554,
      "train_speed(iter/s)": 1.451026
    },
    {
      "epoch": 0.4168630307184782,
      "grad_norm": 4.264733791351318,
      "learning_rate": 9.829469430288338e-05,
      "loss": 2.687091064453125,
      "memory(GiB)": 58.3,
      "step": 9730,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.451126
    },
    {
      "epoch": 0.41707724604772717,
      "grad_norm": 3.2557778358459473,
      "learning_rate": 9.829295126999482e-05,
      "loss": 2.403067779541016,
      "memory(GiB)": 58.3,
      "step": 9735,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.451114
    },
    {
      "epoch": 0.41729146137697615,
      "grad_norm": 4.914039611816406,
      "learning_rate": 9.829120736223417e-05,
      "loss": 2.671504020690918,
      "memory(GiB)": 58.3,
      "step": 9740,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.451288
    },
    {
      "epoch": 0.41750567670622507,
      "grad_norm": 3.6416661739349365,
      "learning_rate": 9.828946257963296e-05,
      "loss": 2.2212257385253906,
      "memory(GiB)": 58.3,
      "step": 9745,
      "token_acc": 0.5387755102040817,
      "train_speed(iter/s)": 1.45145
    },
    {
      "epoch": 0.41771989203547405,
      "grad_norm": 8.937661170959473,
      "learning_rate": 9.828771692222284e-05,
      "loss": 2.4570201873779296,
      "memory(GiB)": 58.3,
      "step": 9750,
      "token_acc": 0.5113636363636364,
      "train_speed(iter/s)": 1.451537
    },
    {
      "epoch": 0.41793410736472303,
      "grad_norm": 3.8024401664733887,
      "learning_rate": 9.828597039003544e-05,
      "loss": 2.5888465881347655,
      "memory(GiB)": 58.3,
      "step": 9755,
      "token_acc": 0.5019607843137255,
      "train_speed(iter/s)": 1.451665
    },
    {
      "epoch": 0.418148322693972,
      "grad_norm": 4.451294422149658,
      "learning_rate": 9.82842229831024e-05,
      "loss": 2.936477851867676,
      "memory(GiB)": 58.3,
      "step": 9760,
      "token_acc": 0.4294478527607362,
      "train_speed(iter/s)": 1.451735
    },
    {
      "epoch": 0.41836253802322093,
      "grad_norm": 4.3740434646606445,
      "learning_rate": 9.828247470145535e-05,
      "loss": 2.576313781738281,
      "memory(GiB)": 58.3,
      "step": 9765,
      "token_acc": 0.4979757085020243,
      "train_speed(iter/s)": 1.451647
    },
    {
      "epoch": 0.4185767533524699,
      "grad_norm": 5.103072166442871,
      "learning_rate": 9.828072554512597e-05,
      "loss": 2.6400827407836913,
      "memory(GiB)": 58.3,
      "step": 9770,
      "token_acc": 0.4847560975609756,
      "train_speed(iter/s)": 1.451642
    },
    {
      "epoch": 0.4187909686817189,
      "grad_norm": 4.11093282699585,
      "learning_rate": 9.827897551414598e-05,
      "loss": 2.6991107940673826,
      "memory(GiB)": 58.3,
      "step": 9775,
      "token_acc": 0.45544554455445546,
      "train_speed(iter/s)": 1.451696
    },
    {
      "epoch": 0.4190051840109678,
      "grad_norm": 4.389251232147217,
      "learning_rate": 9.827722460854705e-05,
      "loss": 2.7348217010498046,
      "memory(GiB)": 58.3,
      "step": 9780,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.451829
    },
    {
      "epoch": 0.4192193993402168,
      "grad_norm": 4.366090774536133,
      "learning_rate": 9.82754728283609e-05,
      "loss": 2.5567569732666016,
      "memory(GiB)": 58.3,
      "step": 9785,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.451781
    },
    {
      "epoch": 0.41943361466946577,
      "grad_norm": 3.707515239715576,
      "learning_rate": 9.827372017361929e-05,
      "loss": 2.6124616622924806,
      "memory(GiB)": 58.3,
      "step": 9790,
      "token_acc": 0.44936708860759494,
      "train_speed(iter/s)": 1.451855
    },
    {
      "epoch": 0.4196478299987147,
      "grad_norm": 4.596794605255127,
      "learning_rate": 9.827196664435394e-05,
      "loss": 2.4758602142333985,
      "memory(GiB)": 58.3,
      "step": 9795,
      "token_acc": 0.5219123505976095,
      "train_speed(iter/s)": 1.451924
    },
    {
      "epoch": 0.4198620453279637,
      "grad_norm": 5.132248401641846,
      "learning_rate": 9.827021224059663e-05,
      "loss": 2.538256072998047,
      "memory(GiB)": 58.3,
      "step": 9800,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.452101
    },
    {
      "epoch": 0.42007626065721265,
      "grad_norm": 3.3905787467956543,
      "learning_rate": 9.826845696237917e-05,
      "loss": 2.232204818725586,
      "memory(GiB)": 58.3,
      "step": 9805,
      "token_acc": 0.47580645161290325,
      "train_speed(iter/s)": 1.451935
    },
    {
      "epoch": 0.4202904759864616,
      "grad_norm": 3.9138851165771484,
      "learning_rate": 9.826670080973331e-05,
      "loss": 2.61285400390625,
      "memory(GiB)": 58.3,
      "step": 9810,
      "token_acc": 0.44150943396226416,
      "train_speed(iter/s)": 1.452012
    },
    {
      "epoch": 0.42050469131571055,
      "grad_norm": 5.183884620666504,
      "learning_rate": 9.826494378269088e-05,
      "loss": 2.547484588623047,
      "memory(GiB)": 58.3,
      "step": 9815,
      "token_acc": 0.467680608365019,
      "train_speed(iter/s)": 1.452109
    },
    {
      "epoch": 0.42071890664495953,
      "grad_norm": 4.3976850509643555,
      "learning_rate": 9.826318588128373e-05,
      "loss": 2.4847038269042967,
      "memory(GiB)": 58.3,
      "step": 9820,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.452278
    },
    {
      "epoch": 0.42093312197420846,
      "grad_norm": 4.496557712554932,
      "learning_rate": 9.82614271055437e-05,
      "loss": 2.5820926666259765,
      "memory(GiB)": 58.3,
      "step": 9825,
      "token_acc": 0.46441947565543074,
      "train_speed(iter/s)": 1.452321
    },
    {
      "epoch": 0.42114733730345744,
      "grad_norm": 6.010549068450928,
      "learning_rate": 9.825966745550262e-05,
      "loss": 2.8377025604248045,
      "memory(GiB)": 58.3,
      "step": 9830,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.45234
    },
    {
      "epoch": 0.4213615526327064,
      "grad_norm": 3.693732261657715,
      "learning_rate": 9.825790693119241e-05,
      "loss": 2.646694564819336,
      "memory(GiB)": 58.3,
      "step": 9835,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.452264
    },
    {
      "epoch": 0.42157576796195534,
      "grad_norm": 5.219442844390869,
      "learning_rate": 9.825614553264495e-05,
      "loss": 3.0060569763183596,
      "memory(GiB)": 58.3,
      "step": 9840,
      "token_acc": 0.42657342657342656,
      "train_speed(iter/s)": 1.45242
    },
    {
      "epoch": 0.4217899832912043,
      "grad_norm": 3.7272820472717285,
      "learning_rate": 9.825438325989214e-05,
      "loss": 2.4056961059570314,
      "memory(GiB)": 58.3,
      "step": 9845,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.452337
    },
    {
      "epoch": 0.4220041986204533,
      "grad_norm": 4.125674724578857,
      "learning_rate": 9.825262011296591e-05,
      "loss": 2.541624069213867,
      "memory(GiB)": 58.3,
      "step": 9850,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.452179
    },
    {
      "epoch": 0.4222184139497022,
      "grad_norm": 3.719965696334839,
      "learning_rate": 9.825085609189821e-05,
      "loss": 2.3237314224243164,
      "memory(GiB)": 58.3,
      "step": 9855,
      "token_acc": 0.5020746887966805,
      "train_speed(iter/s)": 1.452181
    },
    {
      "epoch": 0.4224326292789512,
      "grad_norm": 5.1334028244018555,
      "learning_rate": 9.824909119672098e-05,
      "loss": 2.254361152648926,
      "memory(GiB)": 58.3,
      "step": 9860,
      "token_acc": 0.4981132075471698,
      "train_speed(iter/s)": 1.452126
    },
    {
      "epoch": 0.4226468446082002,
      "grad_norm": 5.238258361816406,
      "learning_rate": 9.82473254274662e-05,
      "loss": 2.55759334564209,
      "memory(GiB)": 58.3,
      "step": 9865,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.452199
    },
    {
      "epoch": 0.4228610599374491,
      "grad_norm": 5.095531940460205,
      "learning_rate": 9.824555878416586e-05,
      "loss": 2.494092559814453,
      "memory(GiB)": 58.3,
      "step": 9870,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.452066
    },
    {
      "epoch": 0.4230752752666981,
      "grad_norm": 4.849524974822998,
      "learning_rate": 9.824379126685197e-05,
      "loss": 2.3354536056518556,
      "memory(GiB)": 58.3,
      "step": 9875,
      "token_acc": 0.5038461538461538,
      "train_speed(iter/s)": 1.452032
    },
    {
      "epoch": 0.42328949059594706,
      "grad_norm": 3.707350730895996,
      "learning_rate": 9.824202287555655e-05,
      "loss": 2.878459167480469,
      "memory(GiB)": 58.3,
      "step": 9880,
      "token_acc": 0.41208791208791207,
      "train_speed(iter/s)": 1.452004
    },
    {
      "epoch": 0.423503705925196,
      "grad_norm": 4.141573905944824,
      "learning_rate": 9.824025361031162e-05,
      "loss": 2.334977722167969,
      "memory(GiB)": 58.3,
      "step": 9885,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.451923
    },
    {
      "epoch": 0.42371792125444496,
      "grad_norm": 4.0235209465026855,
      "learning_rate": 9.823848347114925e-05,
      "loss": 2.7896957397460938,
      "memory(GiB)": 58.3,
      "step": 9890,
      "token_acc": 0.4198717948717949,
      "train_speed(iter/s)": 1.451988
    },
    {
      "epoch": 0.42393213658369394,
      "grad_norm": 3.8881659507751465,
      "learning_rate": 9.82367124581015e-05,
      "loss": 2.3459598541259767,
      "memory(GiB)": 58.3,
      "step": 9895,
      "token_acc": 0.48412698412698413,
      "train_speed(iter/s)": 1.4522
    },
    {
      "epoch": 0.42414635191294286,
      "grad_norm": 4.352472305297852,
      "learning_rate": 9.823494057120046e-05,
      "loss": 2.568310356140137,
      "memory(GiB)": 58.3,
      "step": 9900,
      "token_acc": 0.43214285714285716,
      "train_speed(iter/s)": 1.45229
    },
    {
      "epoch": 0.42436056724219184,
      "grad_norm": 4.939019680023193,
      "learning_rate": 9.82331678104782e-05,
      "loss": 2.5175588607788084,
      "memory(GiB)": 58.3,
      "step": 9905,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.452451
    },
    {
      "epoch": 0.4245747825714408,
      "grad_norm": 6.635680198669434,
      "learning_rate": 9.823139417596688e-05,
      "loss": 2.3019681930541993,
      "memory(GiB)": 58.3,
      "step": 9910,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.452353
    },
    {
      "epoch": 0.42478899790068975,
      "grad_norm": 4.656494617462158,
      "learning_rate": 9.822961966769861e-05,
      "loss": 2.536408042907715,
      "memory(GiB)": 58.3,
      "step": 9915,
      "token_acc": 0.45692883895131087,
      "train_speed(iter/s)": 1.452396
    },
    {
      "epoch": 0.4250032132299387,
      "grad_norm": 5.094083309173584,
      "learning_rate": 9.822784428570552e-05,
      "loss": 2.776628112792969,
      "memory(GiB)": 58.3,
      "step": 9920,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.452285
    },
    {
      "epoch": 0.4252174285591877,
      "grad_norm": 3.064948081970215,
      "learning_rate": 9.822606803001981e-05,
      "loss": 2.715955924987793,
      "memory(GiB)": 58.3,
      "step": 9925,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.452241
    },
    {
      "epoch": 0.4254316438884367,
      "grad_norm": 3.244896650314331,
      "learning_rate": 9.822429090067363e-05,
      "loss": 2.6002098083496095,
      "memory(GiB)": 58.3,
      "step": 9930,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.452365
    },
    {
      "epoch": 0.4256458592176856,
      "grad_norm": 4.495660305023193,
      "learning_rate": 9.822251289769917e-05,
      "loss": 2.604629707336426,
      "memory(GiB)": 58.3,
      "step": 9935,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.452571
    },
    {
      "epoch": 0.4258600745469346,
      "grad_norm": 6.743414878845215,
      "learning_rate": 9.822073402112867e-05,
      "loss": 2.359467697143555,
      "memory(GiB)": 58.3,
      "step": 9940,
      "token_acc": 0.46568627450980393,
      "train_speed(iter/s)": 1.452555
    },
    {
      "epoch": 0.42607428987618357,
      "grad_norm": 3.630054235458374,
      "learning_rate": 9.821895427099434e-05,
      "loss": 2.600129318237305,
      "memory(GiB)": 58.3,
      "step": 9945,
      "token_acc": 0.4326647564469914,
      "train_speed(iter/s)": 1.452691
    },
    {
      "epoch": 0.4262885052054325,
      "grad_norm": 3.3595852851867676,
      "learning_rate": 9.821717364732841e-05,
      "loss": 2.707257843017578,
      "memory(GiB)": 58.3,
      "step": 9950,
      "token_acc": 0.43653250773993807,
      "train_speed(iter/s)": 1.452842
    },
    {
      "epoch": 0.42650272053468147,
      "grad_norm": 4.772958755493164,
      "learning_rate": 9.821539215016314e-05,
      "loss": 2.568878936767578,
      "memory(GiB)": 58.3,
      "step": 9955,
      "token_acc": 0.42765273311897106,
      "train_speed(iter/s)": 1.452812
    },
    {
      "epoch": 0.42671693586393045,
      "grad_norm": 4.185490131378174,
      "learning_rate": 9.821360977953083e-05,
      "loss": 2.379243087768555,
      "memory(GiB)": 58.3,
      "step": 9960,
      "token_acc": 0.5126353790613718,
      "train_speed(iter/s)": 1.452913
    },
    {
      "epoch": 0.42693115119317937,
      "grad_norm": 3.699676752090454,
      "learning_rate": 9.821182653546374e-05,
      "loss": 2.3603675842285154,
      "memory(GiB)": 58.3,
      "step": 9965,
      "token_acc": 0.4533333333333333,
      "train_speed(iter/s)": 1.452909
    },
    {
      "epoch": 0.42714536652242835,
      "grad_norm": 4.556610107421875,
      "learning_rate": 9.821004241799419e-05,
      "loss": 2.3917613983154298,
      "memory(GiB)": 58.3,
      "step": 9970,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.452939
    },
    {
      "epoch": 0.42735958185167733,
      "grad_norm": 3.3329644203186035,
      "learning_rate": 9.82082574271545e-05,
      "loss": 2.701329231262207,
      "memory(GiB)": 58.3,
      "step": 9975,
      "token_acc": 0.4268774703557312,
      "train_speed(iter/s)": 1.452964
    },
    {
      "epoch": 0.42757379718092625,
      "grad_norm": 4.63325309753418,
      "learning_rate": 9.8206471562977e-05,
      "loss": 2.4283470153808593,
      "memory(GiB)": 58.3,
      "step": 9980,
      "token_acc": 0.4788135593220339,
      "train_speed(iter/s)": 1.45299
    },
    {
      "epoch": 0.42778801251017523,
      "grad_norm": 5.624799728393555,
      "learning_rate": 9.820468482549403e-05,
      "loss": 2.7866586685180663,
      "memory(GiB)": 58.3,
      "step": 9985,
      "token_acc": 0.4316109422492401,
      "train_speed(iter/s)": 1.453091
    },
    {
      "epoch": 0.4280022278394242,
      "grad_norm": 10.387460708618164,
      "learning_rate": 9.8202897214738e-05,
      "loss": 2.660850715637207,
      "memory(GiB)": 58.3,
      "step": 9990,
      "token_acc": 0.4364406779661017,
      "train_speed(iter/s)": 1.45329
    },
    {
      "epoch": 0.42821644316867313,
      "grad_norm": 5.224329471588135,
      "learning_rate": 9.820110873074127e-05,
      "loss": 2.52262020111084,
      "memory(GiB)": 58.3,
      "step": 9995,
      "token_acc": 0.4845814977973568,
      "train_speed(iter/s)": 1.453339
    },
    {
      "epoch": 0.4284306584979221,
      "grad_norm": 4.111954212188721,
      "learning_rate": 9.819931937353622e-05,
      "loss": 2.7449642181396485,
      "memory(GiB)": 58.3,
      "step": 10000,
      "token_acc": 0.448559670781893,
      "train_speed(iter/s)": 1.453436
    },
    {
      "epoch": 0.4284306584979221,
      "eval_loss": 2.2310776710510254,
      "eval_runtime": 14.2185,
      "eval_samples_per_second": 7.033,
      "eval_steps_per_second": 7.033,
      "eval_token_acc": 0.46911764705882353,
      "step": 10000
    },
    {
      "epoch": 0.4286448738271711,
      "grad_norm": 3.7548162937164307,
      "learning_rate": 9.81975291431553e-05,
      "loss": 2.677891159057617,
      "memory(GiB)": 58.3,
      "step": 10005,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.450476
    },
    {
      "epoch": 0.42885908915642,
      "grad_norm": 4.450655937194824,
      "learning_rate": 9.819573803963092e-05,
      "loss": 2.459446144104004,
      "memory(GiB)": 58.3,
      "step": 10010,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.450607
    },
    {
      "epoch": 0.429073304485669,
      "grad_norm": 5.074817180633545,
      "learning_rate": 9.819394606299552e-05,
      "loss": 2.340096092224121,
      "memory(GiB)": 58.3,
      "step": 10015,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.450687
    },
    {
      "epoch": 0.429287519814918,
      "grad_norm": 3.4895269870758057,
      "learning_rate": 9.819215321328159e-05,
      "loss": 2.4784250259399414,
      "memory(GiB)": 58.3,
      "step": 10020,
      "token_acc": 0.45791245791245794,
      "train_speed(iter/s)": 1.450856
    },
    {
      "epoch": 0.4295017351441669,
      "grad_norm": 4.1343488693237305,
      "learning_rate": 9.81903594905216e-05,
      "loss": 2.873653030395508,
      "memory(GiB)": 58.3,
      "step": 10025,
      "token_acc": 0.4229390681003584,
      "train_speed(iter/s)": 1.450903
    },
    {
      "epoch": 0.4297159504734159,
      "grad_norm": 3.646547794342041,
      "learning_rate": 9.818856489474803e-05,
      "loss": 2.7377740859985353,
      "memory(GiB)": 58.3,
      "step": 10030,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.451047
    },
    {
      "epoch": 0.42993016580266485,
      "grad_norm": 4.119231700897217,
      "learning_rate": 9.818676942599343e-05,
      "loss": 2.670035552978516,
      "memory(GiB)": 58.3,
      "step": 10035,
      "token_acc": 0.43526170798898073,
      "train_speed(iter/s)": 1.450986
    },
    {
      "epoch": 0.4301443811319138,
      "grad_norm": 5.240797519683838,
      "learning_rate": 9.818497308429028e-05,
      "loss": 2.6698062896728514,
      "memory(GiB)": 58.3,
      "step": 10040,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.451002
    },
    {
      "epoch": 0.43035859646116276,
      "grad_norm": 5.5719895362854,
      "learning_rate": 9.818317586967114e-05,
      "loss": 2.6206993103027343,
      "memory(GiB)": 58.3,
      "step": 10045,
      "token_acc": 0.46394984326018807,
      "train_speed(iter/s)": 1.450987
    },
    {
      "epoch": 0.43057281179041174,
      "grad_norm": 4.11723518371582,
      "learning_rate": 9.818137778216857e-05,
      "loss": 2.7491432189941407,
      "memory(GiB)": 58.3,
      "step": 10050,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.451135
    },
    {
      "epoch": 0.43078702711966066,
      "grad_norm": 4.059296131134033,
      "learning_rate": 9.817957882181514e-05,
      "loss": 2.349420166015625,
      "memory(GiB)": 58.3,
      "step": 10055,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.451191
    },
    {
      "epoch": 0.43100124244890964,
      "grad_norm": 4.097110271453857,
      "learning_rate": 9.817777898864345e-05,
      "loss": 2.798237991333008,
      "memory(GiB)": 58.3,
      "step": 10060,
      "token_acc": 0.4364820846905538,
      "train_speed(iter/s)": 1.451312
    },
    {
      "epoch": 0.4312154577781586,
      "grad_norm": 5.689593315124512,
      "learning_rate": 9.81759782826861e-05,
      "loss": 2.465549659729004,
      "memory(GiB)": 58.3,
      "step": 10065,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.451293
    },
    {
      "epoch": 0.43142967310740754,
      "grad_norm": 4.548140525817871,
      "learning_rate": 9.817417670397571e-05,
      "loss": 2.3398454666137694,
      "memory(GiB)": 58.3,
      "step": 10070,
      "token_acc": 0.5106382978723404,
      "train_speed(iter/s)": 1.4512
    },
    {
      "epoch": 0.4316438884366565,
      "grad_norm": 6.6089863777160645,
      "learning_rate": 9.817237425254492e-05,
      "loss": 2.405929946899414,
      "memory(GiB)": 58.3,
      "step": 10075,
      "token_acc": 0.44654088050314467,
      "train_speed(iter/s)": 1.451241
    },
    {
      "epoch": 0.4318581037659055,
      "grad_norm": 5.215310573577881,
      "learning_rate": 9.817057092842639e-05,
      "loss": 2.610513114929199,
      "memory(GiB)": 58.3,
      "step": 10080,
      "token_acc": 0.44025157232704404,
      "train_speed(iter/s)": 1.451249
    },
    {
      "epoch": 0.4320723190951544,
      "grad_norm": 3.731516122817993,
      "learning_rate": 9.816876673165276e-05,
      "loss": 2.3214963912963866,
      "memory(GiB)": 58.3,
      "step": 10085,
      "token_acc": 0.4574898785425101,
      "train_speed(iter/s)": 1.451396
    },
    {
      "epoch": 0.4322865344244034,
      "grad_norm": 4.801403522491455,
      "learning_rate": 9.816696166225674e-05,
      "loss": 2.7218461990356446,
      "memory(GiB)": 58.3,
      "step": 10090,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.451422
    },
    {
      "epoch": 0.4325007497536524,
      "grad_norm": 4.4056878089904785,
      "learning_rate": 9.816515572027103e-05,
      "loss": 2.6250595092773437,
      "memory(GiB)": 58.3,
      "step": 10095,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.451457
    },
    {
      "epoch": 0.43271496508290136,
      "grad_norm": 5.308544158935547,
      "learning_rate": 9.816334890572834e-05,
      "loss": 2.5006891250610352,
      "memory(GiB)": 58.3,
      "step": 10100,
      "token_acc": 0.486013986013986,
      "train_speed(iter/s)": 1.451602
    },
    {
      "epoch": 0.4329291804121503,
      "grad_norm": 7.511530876159668,
      "learning_rate": 9.81615412186614e-05,
      "loss": 2.5856000900268556,
      "memory(GiB)": 58.3,
      "step": 10105,
      "token_acc": 0.46122448979591835,
      "train_speed(iter/s)": 1.451599
    },
    {
      "epoch": 0.43314339574139926,
      "grad_norm": 4.691821098327637,
      "learning_rate": 9.815973265910296e-05,
      "loss": 2.6388996124267576,
      "memory(GiB)": 58.3,
      "step": 10110,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.451403
    },
    {
      "epoch": 0.43335761107064824,
      "grad_norm": 4.410147190093994,
      "learning_rate": 9.815792322708579e-05,
      "loss": 2.643895721435547,
      "memory(GiB)": 58.3,
      "step": 10115,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.451398
    },
    {
      "epoch": 0.43357182639989716,
      "grad_norm": 4.606101989746094,
      "learning_rate": 9.815611292264267e-05,
      "loss": 2.6611881256103516,
      "memory(GiB)": 58.3,
      "step": 10120,
      "token_acc": 0.4620938628158845,
      "train_speed(iter/s)": 1.451477
    },
    {
      "epoch": 0.43378604172914614,
      "grad_norm": 4.348424911499023,
      "learning_rate": 9.815430174580638e-05,
      "loss": 2.3891780853271483,
      "memory(GiB)": 58.3,
      "step": 10125,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.451523
    },
    {
      "epoch": 0.4340002570583951,
      "grad_norm": 4.787311553955078,
      "learning_rate": 9.815248969660975e-05,
      "loss": 2.6099613189697264,
      "memory(GiB)": 58.3,
      "step": 10130,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.451583
    },
    {
      "epoch": 0.43421447238764405,
      "grad_norm": 4.865276336669922,
      "learning_rate": 9.81506767750856e-05,
      "loss": 2.622056770324707,
      "memory(GiB)": 58.3,
      "step": 10135,
      "token_acc": 0.47265625,
      "train_speed(iter/s)": 1.451635
    },
    {
      "epoch": 0.434428687716893,
      "grad_norm": 4.160851001739502,
      "learning_rate": 9.814886298126678e-05,
      "loss": 2.6428096771240233,
      "memory(GiB)": 58.3,
      "step": 10140,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.451525
    },
    {
      "epoch": 0.434642903046142,
      "grad_norm": 3.2754552364349365,
      "learning_rate": 9.814704831518612e-05,
      "loss": 2.5365121841430662,
      "memory(GiB)": 58.3,
      "step": 10145,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.451612
    },
    {
      "epoch": 0.4348571183753909,
      "grad_norm": 3.87518310546875,
      "learning_rate": 9.814523277687651e-05,
      "loss": 2.6110063552856446,
      "memory(GiB)": 58.3,
      "step": 10150,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.45182
    },
    {
      "epoch": 0.4350713337046399,
      "grad_norm": 4.890989303588867,
      "learning_rate": 9.814341636637085e-05,
      "loss": 3.039305305480957,
      "memory(GiB)": 58.3,
      "step": 10155,
      "token_acc": 0.460431654676259,
      "train_speed(iter/s)": 1.45179
    },
    {
      "epoch": 0.4352855490338889,
      "grad_norm": 4.779794692993164,
      "learning_rate": 9.814159908370206e-05,
      "loss": 2.5788080215454103,
      "memory(GiB)": 58.3,
      "step": 10160,
      "token_acc": 0.4601449275362319,
      "train_speed(iter/s)": 1.451743
    },
    {
      "epoch": 0.4354997643631378,
      "grad_norm": 4.143568992614746,
      "learning_rate": 9.813978092890302e-05,
      "loss": 2.623064422607422,
      "memory(GiB)": 58.3,
      "step": 10165,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.451692
    },
    {
      "epoch": 0.4357139796923868,
      "grad_norm": 4.308685302734375,
      "learning_rate": 9.813796190200671e-05,
      "loss": 2.2172544479370115,
      "memory(GiB)": 58.3,
      "step": 10170,
      "token_acc": 0.5130111524163569,
      "train_speed(iter/s)": 1.451685
    },
    {
      "epoch": 0.43592819502163577,
      "grad_norm": 5.4449543952941895,
      "learning_rate": 9.813614200304604e-05,
      "loss": 2.7361270904541017,
      "memory(GiB)": 58.3,
      "step": 10175,
      "token_acc": 0.4318181818181818,
      "train_speed(iter/s)": 1.451728
    },
    {
      "epoch": 0.4361424103508847,
      "grad_norm": 4.251591682434082,
      "learning_rate": 9.813432123205401e-05,
      "loss": 2.722416114807129,
      "memory(GiB)": 58.3,
      "step": 10180,
      "token_acc": 0.44011976047904194,
      "train_speed(iter/s)": 1.451803
    },
    {
      "epoch": 0.43635662568013367,
      "grad_norm": 5.440887928009033,
      "learning_rate": 9.813249958906362e-05,
      "loss": 2.454940414428711,
      "memory(GiB)": 58.3,
      "step": 10185,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.451828
    },
    {
      "epoch": 0.43657084100938265,
      "grad_norm": 4.593027114868164,
      "learning_rate": 9.813067707410781e-05,
      "loss": 2.643841361999512,
      "memory(GiB)": 58.3,
      "step": 10190,
      "token_acc": 0.4645390070921986,
      "train_speed(iter/s)": 1.451931
    },
    {
      "epoch": 0.43678505633863157,
      "grad_norm": 5.123544692993164,
      "learning_rate": 9.812885368721966e-05,
      "loss": 2.8350969314575196,
      "memory(GiB)": 58.3,
      "step": 10195,
      "token_acc": 0.4200626959247649,
      "train_speed(iter/s)": 1.451857
    },
    {
      "epoch": 0.43699927166788055,
      "grad_norm": 5.503055572509766,
      "learning_rate": 9.812702942843218e-05,
      "loss": 2.770107460021973,
      "memory(GiB)": 58.3,
      "step": 10200,
      "token_acc": 0.4659400544959128,
      "train_speed(iter/s)": 1.45171
    },
    {
      "epoch": 0.43721348699712953,
      "grad_norm": 4.224015235900879,
      "learning_rate": 9.812520429777839e-05,
      "loss": 2.5481491088867188,
      "memory(GiB)": 58.3,
      "step": 10205,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.451527
    },
    {
      "epoch": 0.43742770232637845,
      "grad_norm": 4.281762599945068,
      "learning_rate": 9.81233782952914e-05,
      "loss": 2.680375671386719,
      "memory(GiB)": 58.3,
      "step": 10210,
      "token_acc": 0.4630681818181818,
      "train_speed(iter/s)": 1.451532
    },
    {
      "epoch": 0.43764191765562743,
      "grad_norm": 4.127501964569092,
      "learning_rate": 9.812155142100425e-05,
      "loss": 2.637327194213867,
      "memory(GiB)": 58.3,
      "step": 10215,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.451541
    },
    {
      "epoch": 0.4378561329848764,
      "grad_norm": 5.3888325691223145,
      "learning_rate": 9.811972367495008e-05,
      "loss": 2.6785909652709963,
      "memory(GiB)": 58.3,
      "step": 10220,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.451575
    },
    {
      "epoch": 0.43807034831412534,
      "grad_norm": 3.7547709941864014,
      "learning_rate": 9.811789505716195e-05,
      "loss": 2.6827533721923826,
      "memory(GiB)": 58.3,
      "step": 10225,
      "token_acc": 0.4696969696969697,
      "train_speed(iter/s)": 1.451577
    },
    {
      "epoch": 0.4382845636433743,
      "grad_norm": 6.084743499755859,
      "learning_rate": 9.811606556767303e-05,
      "loss": 2.6391664505004884,
      "memory(GiB)": 58.3,
      "step": 10230,
      "token_acc": 0.4576923076923077,
      "train_speed(iter/s)": 1.451655
    },
    {
      "epoch": 0.4384987789726233,
      "grad_norm": 3.2450554370880127,
      "learning_rate": 9.811423520651644e-05,
      "loss": 2.8701803207397463,
      "memory(GiB)": 58.3,
      "step": 10235,
      "token_acc": 0.4391691394658754,
      "train_speed(iter/s)": 1.451551
    },
    {
      "epoch": 0.4387129943018722,
      "grad_norm": 3.278779983520508,
      "learning_rate": 9.811240397372535e-05,
      "loss": 2.6099533081054687,
      "memory(GiB)": 58.3,
      "step": 10240,
      "token_acc": 0.42295081967213116,
      "train_speed(iter/s)": 1.4517
    },
    {
      "epoch": 0.4389272096311212,
      "grad_norm": 4.576379299163818,
      "learning_rate": 9.811057186933293e-05,
      "loss": 2.3236488342285155,
      "memory(GiB)": 58.3,
      "step": 10245,
      "token_acc": 0.5287356321839081,
      "train_speed(iter/s)": 1.451766
    },
    {
      "epoch": 0.4391414249603702,
      "grad_norm": 3.7729880809783936,
      "learning_rate": 9.810873889337235e-05,
      "loss": 2.6522123336791994,
      "memory(GiB)": 58.3,
      "step": 10250,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.451955
    },
    {
      "epoch": 0.4393556402896191,
      "grad_norm": 4.309666633605957,
      "learning_rate": 9.810690504587685e-05,
      "loss": 2.790571594238281,
      "memory(GiB)": 58.3,
      "step": 10255,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.452002
    },
    {
      "epoch": 0.4395698556188681,
      "grad_norm": 4.178102016448975,
      "learning_rate": 9.810507032687964e-05,
      "loss": 2.5907960891723634,
      "memory(GiB)": 58.3,
      "step": 10260,
      "token_acc": 0.45666666666666667,
      "train_speed(iter/s)": 1.451945
    },
    {
      "epoch": 0.43978407094811706,
      "grad_norm": 3.921598196029663,
      "learning_rate": 9.810323473641395e-05,
      "loss": 2.7393951416015625,
      "memory(GiB)": 58.3,
      "step": 10265,
      "token_acc": 0.42524916943521596,
      "train_speed(iter/s)": 1.451981
    },
    {
      "epoch": 0.43999828627736604,
      "grad_norm": 3.2552382946014404,
      "learning_rate": 9.810139827451305e-05,
      "loss": 2.5158832550048826,
      "memory(GiB)": 58.3,
      "step": 10270,
      "token_acc": 0.46785714285714286,
      "train_speed(iter/s)": 1.45196
    },
    {
      "epoch": 0.44021250160661496,
      "grad_norm": 3.122969627380371,
      "learning_rate": 9.809956094121017e-05,
      "loss": 2.4390892028808593,
      "memory(GiB)": 58.3,
      "step": 10275,
      "token_acc": 0.4697986577181208,
      "train_speed(iter/s)": 1.452051
    },
    {
      "epoch": 0.44042671693586394,
      "grad_norm": 3.823323965072632,
      "learning_rate": 9.809772273653866e-05,
      "loss": 2.5466094970703126,
      "memory(GiB)": 58.3,
      "step": 10280,
      "token_acc": 0.45093457943925236,
      "train_speed(iter/s)": 1.452017
    },
    {
      "epoch": 0.4406409322651129,
      "grad_norm": 3.602243661880493,
      "learning_rate": 9.809588366053175e-05,
      "loss": 2.7130664825439452,
      "memory(GiB)": 58.3,
      "step": 10285,
      "token_acc": 0.4392523364485981,
      "train_speed(iter/s)": 1.452128
    },
    {
      "epoch": 0.44085514759436184,
      "grad_norm": 3.871955156326294,
      "learning_rate": 9.80940437132228e-05,
      "loss": 2.9540714263916015,
      "memory(GiB)": 58.3,
      "step": 10290,
      "token_acc": 0.427536231884058,
      "train_speed(iter/s)": 1.45217
    },
    {
      "epoch": 0.4410693629236108,
      "grad_norm": 3.894803285598755,
      "learning_rate": 9.809220289464513e-05,
      "loss": 2.9123489379882814,
      "memory(GiB)": 58.3,
      "step": 10295,
      "token_acc": 0.4298780487804878,
      "train_speed(iter/s)": 1.451978
    },
    {
      "epoch": 0.4412835782528598,
      "grad_norm": 3.7303433418273926,
      "learning_rate": 9.809036120483211e-05,
      "loss": 2.538544845581055,
      "memory(GiB)": 58.3,
      "step": 10300,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.452051
    },
    {
      "epoch": 0.4414977935821087,
      "grad_norm": 4.130242347717285,
      "learning_rate": 9.808851864381706e-05,
      "loss": 2.7464969635009764,
      "memory(GiB)": 58.3,
      "step": 10305,
      "token_acc": 0.4792746113989637,
      "train_speed(iter/s)": 1.452174
    },
    {
      "epoch": 0.4417120089113577,
      "grad_norm": 3.701951742172241,
      "learning_rate": 9.80866752116334e-05,
      "loss": 2.5881362915039063,
      "memory(GiB)": 58.3,
      "step": 10310,
      "token_acc": 0.44984802431610943,
      "train_speed(iter/s)": 1.452229
    },
    {
      "epoch": 0.4419262242406067,
      "grad_norm": 4.804937839508057,
      "learning_rate": 9.80848309083145e-05,
      "loss": 2.648440361022949,
      "memory(GiB)": 58.3,
      "step": 10315,
      "token_acc": 0.44206008583690987,
      "train_speed(iter/s)": 1.452319
    },
    {
      "epoch": 0.4421404395698556,
      "grad_norm": 4.358564853668213,
      "learning_rate": 9.808298573389379e-05,
      "loss": 2.5335319519042967,
      "memory(GiB)": 58.3,
      "step": 10320,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.45236
    },
    {
      "epoch": 0.4423546548991046,
      "grad_norm": 4.186276912689209,
      "learning_rate": 9.808113968840468e-05,
      "loss": 2.5856439590454103,
      "memory(GiB)": 58.3,
      "step": 10325,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.452504
    },
    {
      "epoch": 0.44256887022835356,
      "grad_norm": 4.695544242858887,
      "learning_rate": 9.807929277188061e-05,
      "loss": 2.5881237030029296,
      "memory(GiB)": 58.3,
      "step": 10330,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.452594
    },
    {
      "epoch": 0.4427830855576025,
      "grad_norm": 3.1374800205230713,
      "learning_rate": 9.807744498435507e-05,
      "loss": 2.614379119873047,
      "memory(GiB)": 58.3,
      "step": 10335,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.452636
    },
    {
      "epoch": 0.44299730088685146,
      "grad_norm": 4.29791259765625,
      "learning_rate": 9.80755963258615e-05,
      "loss": 2.776248741149902,
      "memory(GiB)": 58.3,
      "step": 10340,
      "token_acc": 0.44285714285714284,
      "train_speed(iter/s)": 1.452854
    },
    {
      "epoch": 0.44321151621610044,
      "grad_norm": 4.541945934295654,
      "learning_rate": 9.807374679643342e-05,
      "loss": 2.270546531677246,
      "memory(GiB)": 58.3,
      "step": 10345,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.452839
    },
    {
      "epoch": 0.44342573154534937,
      "grad_norm": 3.594449520111084,
      "learning_rate": 9.80718963961043e-05,
      "loss": 2.5613216400146483,
      "memory(GiB)": 58.3,
      "step": 10350,
      "token_acc": 0.4953560371517028,
      "train_speed(iter/s)": 1.452746
    },
    {
      "epoch": 0.44363994687459835,
      "grad_norm": 3.237761974334717,
      "learning_rate": 9.807004512490769e-05,
      "loss": 2.365348052978516,
      "memory(GiB)": 58.3,
      "step": 10355,
      "token_acc": 0.5299145299145299,
      "train_speed(iter/s)": 1.452781
    },
    {
      "epoch": 0.4438541622038473,
      "grad_norm": 4.93245267868042,
      "learning_rate": 9.806819298287713e-05,
      "loss": 2.696757698059082,
      "memory(GiB)": 58.3,
      "step": 10360,
      "token_acc": 0.4382716049382716,
      "train_speed(iter/s)": 1.452898
    },
    {
      "epoch": 0.44406837753309625,
      "grad_norm": 5.638664722442627,
      "learning_rate": 9.806633997004615e-05,
      "loss": 2.884890556335449,
      "memory(GiB)": 58.3,
      "step": 10365,
      "token_acc": 0.42124542124542125,
      "train_speed(iter/s)": 1.45285
    },
    {
      "epoch": 0.4442825928623452,
      "grad_norm": 3.3431310653686523,
      "learning_rate": 9.806448608644834e-05,
      "loss": 2.3671947479248048,
      "memory(GiB)": 58.3,
      "step": 10370,
      "token_acc": 0.504424778761062,
      "train_speed(iter/s)": 1.452998
    },
    {
      "epoch": 0.4444968081915942,
      "grad_norm": 3.615079879760742,
      "learning_rate": 9.806263133211728e-05,
      "loss": 2.407727813720703,
      "memory(GiB)": 58.3,
      "step": 10375,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.453045
    },
    {
      "epoch": 0.44471102352084313,
      "grad_norm": 6.307542324066162,
      "learning_rate": 9.806077570708654e-05,
      "loss": 2.398736763000488,
      "memory(GiB)": 58.3,
      "step": 10380,
      "token_acc": 0.5284280936454849,
      "train_speed(iter/s)": 1.453167
    },
    {
      "epoch": 0.4449252388500921,
      "grad_norm": 6.1022443771362305,
      "learning_rate": 9.805891921138979e-05,
      "loss": 2.2987117767333984,
      "memory(GiB)": 58.3,
      "step": 10385,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.453223
    },
    {
      "epoch": 0.4451394541793411,
      "grad_norm": 5.117356300354004,
      "learning_rate": 9.805706184506062e-05,
      "loss": 2.508358955383301,
      "memory(GiB)": 58.3,
      "step": 10390,
      "token_acc": 0.46835443037974683,
      "train_speed(iter/s)": 1.453035
    },
    {
      "epoch": 0.44535366950859,
      "grad_norm": 4.126290321350098,
      "learning_rate": 9.805520360813272e-05,
      "loss": 2.717280960083008,
      "memory(GiB)": 58.3,
      "step": 10395,
      "token_acc": 0.44072948328267475,
      "train_speed(iter/s)": 1.4529
    },
    {
      "epoch": 0.445567884837839,
      "grad_norm": 5.412095069885254,
      "learning_rate": 9.80533445006397e-05,
      "loss": 2.6317028045654296,
      "memory(GiB)": 58.3,
      "step": 10400,
      "token_acc": 0.44074074074074077,
      "train_speed(iter/s)": 1.453116
    },
    {
      "epoch": 0.44578210016708797,
      "grad_norm": 8.25463581085205,
      "learning_rate": 9.805148452261528e-05,
      "loss": 2.9181276321411134,
      "memory(GiB)": 58.3,
      "step": 10405,
      "token_acc": 0.43812709030100333,
      "train_speed(iter/s)": 1.453152
    },
    {
      "epoch": 0.4459963154963369,
      "grad_norm": 3.7849009037017822,
      "learning_rate": 9.804962367409313e-05,
      "loss": 2.539438247680664,
      "memory(GiB)": 58.3,
      "step": 10410,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.453323
    },
    {
      "epoch": 0.44621053082558587,
      "grad_norm": 3.961578130722046,
      "learning_rate": 9.804776195510699e-05,
      "loss": 2.5199079513549805,
      "memory(GiB)": 58.3,
      "step": 10415,
      "token_acc": 0.5508474576271186,
      "train_speed(iter/s)": 1.453375
    },
    {
      "epoch": 0.44642474615483485,
      "grad_norm": 4.937049388885498,
      "learning_rate": 9.804589936569055e-05,
      "loss": 2.718492889404297,
      "memory(GiB)": 58.3,
      "step": 10420,
      "token_acc": 0.4318181818181818,
      "train_speed(iter/s)": 1.453055
    },
    {
      "epoch": 0.4466389614840838,
      "grad_norm": 3.9586424827575684,
      "learning_rate": 9.804403590587758e-05,
      "loss": 2.673126792907715,
      "memory(GiB)": 58.3,
      "step": 10425,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.452989
    },
    {
      "epoch": 0.44685317681333275,
      "grad_norm": 3.0467731952667236,
      "learning_rate": 9.804217157570184e-05,
      "loss": 2.875569152832031,
      "memory(GiB)": 58.3,
      "step": 10430,
      "token_acc": 0.4103448275862069,
      "train_speed(iter/s)": 1.452952
    },
    {
      "epoch": 0.44706739214258173,
      "grad_norm": 2.768760919570923,
      "learning_rate": 9.804030637519708e-05,
      "loss": 2.668725776672363,
      "memory(GiB)": 58.3,
      "step": 10435,
      "token_acc": 0.46175637393767704,
      "train_speed(iter/s)": 1.453009
    },
    {
      "epoch": 0.4472816074718307,
      "grad_norm": 3.534268617630005,
      "learning_rate": 9.803844030439711e-05,
      "loss": 2.4038856506347654,
      "memory(GiB)": 58.3,
      "step": 10440,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.453162
    },
    {
      "epoch": 0.44749582280107963,
      "grad_norm": 4.700990676879883,
      "learning_rate": 9.803657336333574e-05,
      "loss": 2.8660232543945314,
      "memory(GiB)": 58.3,
      "step": 10445,
      "token_acc": 0.4234527687296417,
      "train_speed(iter/s)": 1.453075
    },
    {
      "epoch": 0.4477100381303286,
      "grad_norm": 5.253242015838623,
      "learning_rate": 9.803470555204676e-05,
      "loss": 2.3562389373779298,
      "memory(GiB)": 58.3,
      "step": 10450,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.452904
    },
    {
      "epoch": 0.4479242534595776,
      "grad_norm": 14.840235710144043,
      "learning_rate": 9.803283687056404e-05,
      "loss": 2.645217704772949,
      "memory(GiB)": 58.3,
      "step": 10455,
      "token_acc": 0.43097643097643096,
      "train_speed(iter/s)": 1.45275
    },
    {
      "epoch": 0.4481384687888265,
      "grad_norm": 4.14026403427124,
      "learning_rate": 9.803096731892142e-05,
      "loss": 2.549029541015625,
      "memory(GiB)": 58.3,
      "step": 10460,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.452964
    },
    {
      "epoch": 0.4483526841180755,
      "grad_norm": 5.461249351501465,
      "learning_rate": 9.802909689715278e-05,
      "loss": 2.739639091491699,
      "memory(GiB)": 58.3,
      "step": 10465,
      "token_acc": 0.4558011049723757,
      "train_speed(iter/s)": 1.45317
    },
    {
      "epoch": 0.4485668994473245,
      "grad_norm": 4.478499412536621,
      "learning_rate": 9.802722560529199e-05,
      "loss": 2.686005401611328,
      "memory(GiB)": 58.3,
      "step": 10470,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.453096
    },
    {
      "epoch": 0.4487811147765734,
      "grad_norm": 6.938562870025635,
      "learning_rate": 9.802535344337296e-05,
      "loss": 2.4839391708374023,
      "memory(GiB)": 58.3,
      "step": 10475,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.452976
    },
    {
      "epoch": 0.4489953301058224,
      "grad_norm": 4.087958812713623,
      "learning_rate": 9.80234804114296e-05,
      "loss": 2.4675451278686524,
      "memory(GiB)": 58.3,
      "step": 10480,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.453059
    },
    {
      "epoch": 0.44920954543507136,
      "grad_norm": 4.140214443206787,
      "learning_rate": 9.802160650949584e-05,
      "loss": 2.826065254211426,
      "memory(GiB)": 58.3,
      "step": 10485,
      "token_acc": 0.44025157232704404,
      "train_speed(iter/s)": 1.453005
    },
    {
      "epoch": 0.4494237607643203,
      "grad_norm": 4.372977256774902,
      "learning_rate": 9.801973173760562e-05,
      "loss": 2.6633235931396486,
      "memory(GiB)": 58.3,
      "step": 10490,
      "token_acc": 0.459546925566343,
      "train_speed(iter/s)": 1.45318
    },
    {
      "epoch": 0.44963797609356926,
      "grad_norm": 6.438418865203857,
      "learning_rate": 9.801785609579292e-05,
      "loss": 2.941879653930664,
      "memory(GiB)": 58.3,
      "step": 10495,
      "token_acc": 0.44200626959247646,
      "train_speed(iter/s)": 1.453301
    },
    {
      "epoch": 0.44985219142281824,
      "grad_norm": 3.240180730819702,
      "learning_rate": 9.801597958409172e-05,
      "loss": 2.5285844802856445,
      "memory(GiB)": 58.3,
      "step": 10500,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.453377
    },
    {
      "epoch": 0.44985219142281824,
      "eval_loss": 2.126368522644043,
      "eval_runtime": 14.1191,
      "eval_samples_per_second": 7.083,
      "eval_steps_per_second": 7.083,
      "eval_token_acc": 0.49923195084485406,
      "step": 10500
    },
    {
      "epoch": 0.45006640675206716,
      "grad_norm": 3.5414857864379883,
      "learning_rate": 9.8014102202536e-05,
      "loss": 2.6892297744750975,
      "memory(GiB)": 58.3,
      "step": 10505,
      "token_acc": 0.48895582329317266,
      "train_speed(iter/s)": 1.450312
    },
    {
      "epoch": 0.45028062208131614,
      "grad_norm": 4.189624309539795,
      "learning_rate": 9.801222395115976e-05,
      "loss": 2.5900028228759764,
      "memory(GiB)": 58.3,
      "step": 10510,
      "token_acc": 0.45936395759717313,
      "train_speed(iter/s)": 1.450381
    },
    {
      "epoch": 0.4504948374105651,
      "grad_norm": 6.751174449920654,
      "learning_rate": 9.801034482999707e-05,
      "loss": 2.3130468368530273,
      "memory(GiB)": 58.3,
      "step": 10515,
      "token_acc": 0.4858757062146893,
      "train_speed(iter/s)": 1.450449
    },
    {
      "epoch": 0.45070905273981404,
      "grad_norm": 3.647382974624634,
      "learning_rate": 9.800846483908195e-05,
      "loss": 2.590544891357422,
      "memory(GiB)": 58.3,
      "step": 10520,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.450405
    },
    {
      "epoch": 0.450923268069063,
      "grad_norm": 5.126917362213135,
      "learning_rate": 9.800658397844844e-05,
      "loss": 2.6092994689941404,
      "memory(GiB)": 58.3,
      "step": 10525,
      "token_acc": 0.4742268041237113,
      "train_speed(iter/s)": 1.450581
    },
    {
      "epoch": 0.451137483398312,
      "grad_norm": 4.543416500091553,
      "learning_rate": 9.800470224813064e-05,
      "loss": 2.889688491821289,
      "memory(GiB)": 58.3,
      "step": 10530,
      "token_acc": 0.44481605351170567,
      "train_speed(iter/s)": 1.450654
    },
    {
      "epoch": 0.4513516987275609,
      "grad_norm": 4.379642963409424,
      "learning_rate": 9.80028196481626e-05,
      "loss": 2.648379898071289,
      "memory(GiB)": 58.3,
      "step": 10535,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.450449
    },
    {
      "epoch": 0.4515659140568099,
      "grad_norm": 3.2970504760742188,
      "learning_rate": 9.800093617857846e-05,
      "loss": 2.5996782302856447,
      "memory(GiB)": 58.3,
      "step": 10540,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.450531
    },
    {
      "epoch": 0.4517801293860589,
      "grad_norm": 5.440197467803955,
      "learning_rate": 9.799905183941236e-05,
      "loss": 2.53688907623291,
      "memory(GiB)": 58.3,
      "step": 10545,
      "token_acc": 0.45774647887323944,
      "train_speed(iter/s)": 1.450623
    },
    {
      "epoch": 0.4519943447153078,
      "grad_norm": 4.509754657745361,
      "learning_rate": 9.799716663069838e-05,
      "loss": 2.78790397644043,
      "memory(GiB)": 58.3,
      "step": 10550,
      "token_acc": 0.46204620462046203,
      "train_speed(iter/s)": 1.450625
    },
    {
      "epoch": 0.4522085600445568,
      "grad_norm": 4.187043190002441,
      "learning_rate": 9.799528055247071e-05,
      "loss": 2.655309867858887,
      "memory(GiB)": 58.3,
      "step": 10555,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.450604
    },
    {
      "epoch": 0.45242277537380576,
      "grad_norm": 4.815025329589844,
      "learning_rate": 9.799339360476352e-05,
      "loss": 2.7851726531982424,
      "memory(GiB)": 58.3,
      "step": 10560,
      "token_acc": 0.44360902255639095,
      "train_speed(iter/s)": 1.450622
    },
    {
      "epoch": 0.4526369907030547,
      "grad_norm": 3.575446128845215,
      "learning_rate": 9.799150578761098e-05,
      "loss": 2.543497848510742,
      "memory(GiB)": 58.3,
      "step": 10565,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.450637
    },
    {
      "epoch": 0.45285120603230367,
      "grad_norm": 4.181159496307373,
      "learning_rate": 9.798961710104728e-05,
      "loss": 2.623805618286133,
      "memory(GiB)": 58.3,
      "step": 10570,
      "token_acc": 0.4753521126760563,
      "train_speed(iter/s)": 1.450629
    },
    {
      "epoch": 0.45306542136155264,
      "grad_norm": 3.958008050918579,
      "learning_rate": 9.798772754510666e-05,
      "loss": 2.475885200500488,
      "memory(GiB)": 58.3,
      "step": 10575,
      "token_acc": 0.46827794561933533,
      "train_speed(iter/s)": 1.450384
    },
    {
      "epoch": 0.45327963669080157,
      "grad_norm": 4.189507007598877,
      "learning_rate": 9.798583711982332e-05,
      "loss": 2.7183481216430665,
      "memory(GiB)": 58.3,
      "step": 10580,
      "token_acc": 0.46229508196721314,
      "train_speed(iter/s)": 1.450573
    },
    {
      "epoch": 0.45349385202005055,
      "grad_norm": 4.074063301086426,
      "learning_rate": 9.798394582523154e-05,
      "loss": 2.802981948852539,
      "memory(GiB)": 58.3,
      "step": 10585,
      "token_acc": 0.41534391534391535,
      "train_speed(iter/s)": 1.450591
    },
    {
      "epoch": 0.4537080673492995,
      "grad_norm": 5.128859043121338,
      "learning_rate": 9.798205366136558e-05,
      "loss": 2.7304380416870115,
      "memory(GiB)": 58.3,
      "step": 10590,
      "token_acc": 0.4405594405594406,
      "train_speed(iter/s)": 1.450703
    },
    {
      "epoch": 0.45392228267854845,
      "grad_norm": 3.6847312450408936,
      "learning_rate": 9.79801606282597e-05,
      "loss": 2.2724870681762694,
      "memory(GiB)": 58.3,
      "step": 10595,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.450779
    },
    {
      "epoch": 0.45413649800779743,
      "grad_norm": 4.458700656890869,
      "learning_rate": 9.797826672594819e-05,
      "loss": 2.497824478149414,
      "memory(GiB)": 58.3,
      "step": 10600,
      "token_acc": 0.48638132295719844,
      "train_speed(iter/s)": 1.450905
    },
    {
      "epoch": 0.4543507133370464,
      "grad_norm": 4.180672645568848,
      "learning_rate": 9.797637195446538e-05,
      "loss": 2.615709686279297,
      "memory(GiB)": 58.3,
      "step": 10605,
      "token_acc": 0.477124183006536,
      "train_speed(iter/s)": 1.450776
    },
    {
      "epoch": 0.4545649286662954,
      "grad_norm": 4.374013423919678,
      "learning_rate": 9.797447631384559e-05,
      "loss": 2.6567962646484373,
      "memory(GiB)": 58.3,
      "step": 10610,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.450747
    },
    {
      "epoch": 0.4547791439955443,
      "grad_norm": 4.041299343109131,
      "learning_rate": 9.797257980412316e-05,
      "loss": 2.621622657775879,
      "memory(GiB)": 58.3,
      "step": 10615,
      "token_acc": 0.4584837545126354,
      "train_speed(iter/s)": 1.450717
    },
    {
      "epoch": 0.4549933593247933,
      "grad_norm": 4.688261985778809,
      "learning_rate": 9.797068242533243e-05,
      "loss": 2.5129161834716798,
      "memory(GiB)": 58.3,
      "step": 10620,
      "token_acc": 0.46255506607929514,
      "train_speed(iter/s)": 1.450356
    },
    {
      "epoch": 0.45520757465404227,
      "grad_norm": 3.561720132827759,
      "learning_rate": 9.79687841775078e-05,
      "loss": 2.6358680725097656,
      "memory(GiB)": 58.3,
      "step": 10625,
      "token_acc": 0.4423791821561338,
      "train_speed(iter/s)": 1.450467
    },
    {
      "epoch": 0.4554217899832912,
      "grad_norm": 4.223676681518555,
      "learning_rate": 9.796688506068364e-05,
      "loss": 2.379704475402832,
      "memory(GiB)": 58.3,
      "step": 10630,
      "token_acc": 0.4536741214057508,
      "train_speed(iter/s)": 1.45058
    },
    {
      "epoch": 0.45563600531254017,
      "grad_norm": 5.094300746917725,
      "learning_rate": 9.796498507489436e-05,
      "loss": 2.7889373779296873,
      "memory(GiB)": 58.3,
      "step": 10635,
      "token_acc": 0.4532871972318339,
      "train_speed(iter/s)": 1.450619
    },
    {
      "epoch": 0.45585022064178915,
      "grad_norm": 4.634842395782471,
      "learning_rate": 9.79630842201744e-05,
      "loss": 2.478668785095215,
      "memory(GiB)": 58.3,
      "step": 10640,
      "token_acc": 0.46540880503144655,
      "train_speed(iter/s)": 1.450697
    },
    {
      "epoch": 0.4560644359710381,
      "grad_norm": 4.784429550170898,
      "learning_rate": 9.796118249655814e-05,
      "loss": 2.6598688125610352,
      "memory(GiB)": 58.3,
      "step": 10645,
      "token_acc": 0.486404833836858,
      "train_speed(iter/s)": 1.450567
    },
    {
      "epoch": 0.45627865130028705,
      "grad_norm": 3.1336419582366943,
      "learning_rate": 9.795927990408009e-05,
      "loss": 2.588586616516113,
      "memory(GiB)": 58.3,
      "step": 10650,
      "token_acc": 0.4773413897280967,
      "train_speed(iter/s)": 1.450478
    },
    {
      "epoch": 0.45649286662953603,
      "grad_norm": 4.033613204956055,
      "learning_rate": 9.79573764427747e-05,
      "loss": 2.7217081069946287,
      "memory(GiB)": 58.3,
      "step": 10655,
      "token_acc": 0.48638132295719844,
      "train_speed(iter/s)": 1.450615
    },
    {
      "epoch": 0.45670708195878496,
      "grad_norm": 5.214172840118408,
      "learning_rate": 9.795547211267643e-05,
      "loss": 2.4614585876464843,
      "memory(GiB)": 58.3,
      "step": 10660,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.450557
    },
    {
      "epoch": 0.45692129728803393,
      "grad_norm": 6.807183742523193,
      "learning_rate": 9.795356691381983e-05,
      "loss": 2.5423999786376954,
      "memory(GiB)": 58.3,
      "step": 10665,
      "token_acc": 0.4530612244897959,
      "train_speed(iter/s)": 1.450437
    },
    {
      "epoch": 0.4571355126172829,
      "grad_norm": 3.8348653316497803,
      "learning_rate": 9.795166084623934e-05,
      "loss": 2.6810192108154296,
      "memory(GiB)": 58.3,
      "step": 10670,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.450467
    },
    {
      "epoch": 0.45734972794653184,
      "grad_norm": 6.727038860321045,
      "learning_rate": 9.794975390996956e-05,
      "loss": 2.6809814453125,
      "memory(GiB)": 58.3,
      "step": 10675,
      "token_acc": 0.4462809917355372,
      "train_speed(iter/s)": 1.450499
    },
    {
      "epoch": 0.4575639432757808,
      "grad_norm": 4.346179008483887,
      "learning_rate": 9.7947846105045e-05,
      "loss": 2.444858932495117,
      "memory(GiB)": 58.3,
      "step": 10680,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.450585
    },
    {
      "epoch": 0.4577781586050298,
      "grad_norm": 6.229421615600586,
      "learning_rate": 9.794593743150022e-05,
      "loss": 2.836994171142578,
      "memory(GiB)": 58.3,
      "step": 10685,
      "token_acc": 0.43911439114391143,
      "train_speed(iter/s)": 1.450806
    },
    {
      "epoch": 0.4579923739342787,
      "grad_norm": 3.387035846710205,
      "learning_rate": 9.794402788936983e-05,
      "loss": 2.5257070541381834,
      "memory(GiB)": 58.3,
      "step": 10690,
      "token_acc": 0.48582995951417,
      "train_speed(iter/s)": 1.450953
    },
    {
      "epoch": 0.4582065892635277,
      "grad_norm": 3.9601640701293945,
      "learning_rate": 9.794211747868838e-05,
      "loss": 2.578817939758301,
      "memory(GiB)": 58.3,
      "step": 10695,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.450979
    },
    {
      "epoch": 0.4584208045927767,
      "grad_norm": 5.536830425262451,
      "learning_rate": 9.79402061994905e-05,
      "loss": 2.96250057220459,
      "memory(GiB)": 58.3,
      "step": 10700,
      "token_acc": 0.41132075471698115,
      "train_speed(iter/s)": 1.451014
    },
    {
      "epoch": 0.4586350199220256,
      "grad_norm": 4.493021488189697,
      "learning_rate": 9.793829405181081e-05,
      "loss": 2.4744644165039062,
      "memory(GiB)": 58.3,
      "step": 10705,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.451083
    },
    {
      "epoch": 0.4588492352512746,
      "grad_norm": 4.748714923858643,
      "learning_rate": 9.793638103568397e-05,
      "loss": 2.5041988372802733,
      "memory(GiB)": 58.3,
      "step": 10710,
      "token_acc": 0.46629213483146065,
      "train_speed(iter/s)": 1.451054
    },
    {
      "epoch": 0.45906345058052356,
      "grad_norm": 3.1664113998413086,
      "learning_rate": 9.79344671511446e-05,
      "loss": 2.6128122329711916,
      "memory(GiB)": 58.3,
      "step": 10715,
      "token_acc": 0.46006389776357826,
      "train_speed(iter/s)": 1.450868
    },
    {
      "epoch": 0.4592776659097725,
      "grad_norm": 5.172250747680664,
      "learning_rate": 9.79325523982274e-05,
      "loss": 2.7653314590454103,
      "memory(GiB)": 58.3,
      "step": 10720,
      "token_acc": 0.4651898734177215,
      "train_speed(iter/s)": 1.450748
    },
    {
      "epoch": 0.45949188123902146,
      "grad_norm": 4.8607001304626465,
      "learning_rate": 9.793063677696706e-05,
      "loss": 2.7596765518188477,
      "memory(GiB)": 58.3,
      "step": 10725,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.450681
    },
    {
      "epoch": 0.45970609656827044,
      "grad_norm": 4.228477954864502,
      "learning_rate": 9.792872028739826e-05,
      "loss": 2.715163803100586,
      "memory(GiB)": 58.3,
      "step": 10730,
      "token_acc": 0.43234323432343236,
      "train_speed(iter/s)": 1.450609
    },
    {
      "epoch": 0.45992031189751936,
      "grad_norm": 4.469358444213867,
      "learning_rate": 9.792680292955571e-05,
      "loss": 2.675069046020508,
      "memory(GiB)": 58.3,
      "step": 10735,
      "token_acc": 0.4405594405594406,
      "train_speed(iter/s)": 1.450748
    },
    {
      "epoch": 0.46013452722676834,
      "grad_norm": 3.619096517562866,
      "learning_rate": 9.792488470347421e-05,
      "loss": 2.6168188095092773,
      "memory(GiB)": 58.3,
      "step": 10740,
      "token_acc": 0.45084745762711864,
      "train_speed(iter/s)": 1.450713
    },
    {
      "epoch": 0.4603487425560173,
      "grad_norm": 3.301767587661743,
      "learning_rate": 9.792296560918844e-05,
      "loss": 2.9165679931640627,
      "memory(GiB)": 58.3,
      "step": 10745,
      "token_acc": 0.42084942084942084,
      "train_speed(iter/s)": 1.450789
    },
    {
      "epoch": 0.46056295788526624,
      "grad_norm": 4.956804275512695,
      "learning_rate": 9.792104564673319e-05,
      "loss": 2.60153751373291,
      "memory(GiB)": 58.3,
      "step": 10750,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.450813
    },
    {
      "epoch": 0.4607771732145152,
      "grad_norm": 3.9421226978302,
      "learning_rate": 9.791912481614324e-05,
      "loss": 2.554246520996094,
      "memory(GiB)": 58.3,
      "step": 10755,
      "token_acc": 0.4550898203592814,
      "train_speed(iter/s)": 1.450818
    },
    {
      "epoch": 0.4609913885437642,
      "grad_norm": 6.13618278503418,
      "learning_rate": 9.791720311745342e-05,
      "loss": 2.6295148849487306,
      "memory(GiB)": 58.3,
      "step": 10760,
      "token_acc": 0.5245901639344263,
      "train_speed(iter/s)": 1.450973
    },
    {
      "epoch": 0.4612056038730131,
      "grad_norm": 4.67579460144043,
      "learning_rate": 9.791528055069849e-05,
      "loss": 2.4764760971069335,
      "memory(GiB)": 58.3,
      "step": 10765,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.451022
    },
    {
      "epoch": 0.4614198192022621,
      "grad_norm": 5.211319446563721,
      "learning_rate": 9.791335711591332e-05,
      "loss": 2.703862190246582,
      "memory(GiB)": 58.3,
      "step": 10770,
      "token_acc": 0.42948717948717946,
      "train_speed(iter/s)": 1.45096
    },
    {
      "epoch": 0.4616340345315111,
      "grad_norm": 4.508678913116455,
      "learning_rate": 9.791143281313274e-05,
      "loss": 2.5066123962402345,
      "memory(GiB)": 58.3,
      "step": 10775,
      "token_acc": 0.4508670520231214,
      "train_speed(iter/s)": 1.450878
    },
    {
      "epoch": 0.46184824986076006,
      "grad_norm": 3.7897000312805176,
      "learning_rate": 9.79095076423916e-05,
      "loss": 2.6497310638427733,
      "memory(GiB)": 58.3,
      "step": 10780,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.450857
    },
    {
      "epoch": 0.462062465190009,
      "grad_norm": 3.7451624870300293,
      "learning_rate": 9.790758160372479e-05,
      "loss": 2.488443374633789,
      "memory(GiB)": 58.3,
      "step": 10785,
      "token_acc": 0.47232472324723246,
      "train_speed(iter/s)": 1.450973
    },
    {
      "epoch": 0.46227668051925797,
      "grad_norm": 5.758881568908691,
      "learning_rate": 9.79056546971672e-05,
      "loss": 2.585683059692383,
      "memory(GiB)": 58.3,
      "step": 10790,
      "token_acc": 0.44565217391304346,
      "train_speed(iter/s)": 1.450978
    },
    {
      "epoch": 0.46249089584850694,
      "grad_norm": 3.644552230834961,
      "learning_rate": 9.790372692275374e-05,
      "loss": 2.5159221649169923,
      "memory(GiB)": 58.3,
      "step": 10795,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.451049
    },
    {
      "epoch": 0.46270511117775587,
      "grad_norm": 4.923215389251709,
      "learning_rate": 9.790179828051931e-05,
      "loss": 2.3791568756103514,
      "memory(GiB)": 58.3,
      "step": 10800,
      "token_acc": 0.4703703703703704,
      "train_speed(iter/s)": 1.451007
    },
    {
      "epoch": 0.46291932650700485,
      "grad_norm": 3.903329849243164,
      "learning_rate": 9.789986877049888e-05,
      "loss": 2.612640380859375,
      "memory(GiB)": 58.3,
      "step": 10805,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.451101
    },
    {
      "epoch": 0.4631335418362538,
      "grad_norm": 4.178727149963379,
      "learning_rate": 9.78979383927274e-05,
      "loss": 2.5980998992919924,
      "memory(GiB)": 58.3,
      "step": 10810,
      "token_acc": 0.44193548387096776,
      "train_speed(iter/s)": 1.450968
    },
    {
      "epoch": 0.46334775716550275,
      "grad_norm": 3.5005838871002197,
      "learning_rate": 9.789600714723983e-05,
      "loss": 2.5595413208007813,
      "memory(GiB)": 58.3,
      "step": 10815,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.450794
    },
    {
      "epoch": 0.46356197249475173,
      "grad_norm": 3.065406560897827,
      "learning_rate": 9.789407503407115e-05,
      "loss": 2.8204328536987306,
      "memory(GiB)": 58.3,
      "step": 10820,
      "token_acc": 0.4253164556962025,
      "train_speed(iter/s)": 1.451003
    },
    {
      "epoch": 0.4637761878240007,
      "grad_norm": 5.285351276397705,
      "learning_rate": 9.789214205325638e-05,
      "loss": 2.522801971435547,
      "memory(GiB)": 58.3,
      "step": 10825,
      "token_acc": 0.46229508196721314,
      "train_speed(iter/s)": 1.450962
    },
    {
      "epoch": 0.46399040315324963,
      "grad_norm": 3.872533082962036,
      "learning_rate": 9.789020820483055e-05,
      "loss": 2.9009300231933595,
      "memory(GiB)": 58.3,
      "step": 10830,
      "token_acc": 0.375,
      "train_speed(iter/s)": 1.450856
    },
    {
      "epoch": 0.4642046184824986,
      "grad_norm": 3.1430745124816895,
      "learning_rate": 9.788827348882865e-05,
      "loss": 2.362409973144531,
      "memory(GiB)": 58.3,
      "step": 10835,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.451044
    },
    {
      "epoch": 0.4644188338117476,
      "grad_norm": 3.7982048988342285,
      "learning_rate": 9.788633790528576e-05,
      "loss": 2.4230241775512695,
      "memory(GiB)": 58.3,
      "step": 10840,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.451088
    },
    {
      "epoch": 0.4646330491409965,
      "grad_norm": 5.107214450836182,
      "learning_rate": 9.788440145423695e-05,
      "loss": 2.7976070404052735,
      "memory(GiB)": 58.3,
      "step": 10845,
      "token_acc": 0.4944649446494465,
      "train_speed(iter/s)": 1.45126
    },
    {
      "epoch": 0.4648472644702455,
      "grad_norm": 4.19752311706543,
      "learning_rate": 9.788246413571727e-05,
      "loss": 2.6039764404296877,
      "memory(GiB)": 58.3,
      "step": 10850,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.451428
    },
    {
      "epoch": 0.46506147979949447,
      "grad_norm": 9.007368087768555,
      "learning_rate": 9.788052594976184e-05,
      "loss": 2.68975715637207,
      "memory(GiB)": 58.3,
      "step": 10855,
      "token_acc": 0.4396887159533074,
      "train_speed(iter/s)": 1.451448
    },
    {
      "epoch": 0.4652756951287434,
      "grad_norm": 6.284673690795898,
      "learning_rate": 9.787858689640577e-05,
      "loss": 2.6068992614746094,
      "memory(GiB)": 58.3,
      "step": 10860,
      "token_acc": 0.44285714285714284,
      "train_speed(iter/s)": 1.451358
    },
    {
      "epoch": 0.4654899104579924,
      "grad_norm": 4.84188985824585,
      "learning_rate": 9.787664697568418e-05,
      "loss": 2.7556400299072266,
      "memory(GiB)": 58.3,
      "step": 10865,
      "token_acc": 0.4426229508196721,
      "train_speed(iter/s)": 1.451505
    },
    {
      "epoch": 0.46570412578724135,
      "grad_norm": 4.268581867218018,
      "learning_rate": 9.787470618763222e-05,
      "loss": 2.7689865112304686,
      "memory(GiB)": 58.3,
      "step": 10870,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.451717
    },
    {
      "epoch": 0.4659183411164903,
      "grad_norm": 3.183107614517212,
      "learning_rate": 9.787276453228504e-05,
      "loss": 2.609415817260742,
      "memory(GiB)": 58.3,
      "step": 10875,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.451829
    },
    {
      "epoch": 0.46613255644573925,
      "grad_norm": 8.195131301879883,
      "learning_rate": 9.787082200967784e-05,
      "loss": 2.664288330078125,
      "memory(GiB)": 58.3,
      "step": 10880,
      "token_acc": 0.4251497005988024,
      "train_speed(iter/s)": 1.452029
    },
    {
      "epoch": 0.46634677177498823,
      "grad_norm": 4.695074558258057,
      "learning_rate": 9.786887861984578e-05,
      "loss": 2.6576133728027345,
      "memory(GiB)": 58.3,
      "step": 10885,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.452024
    },
    {
      "epoch": 0.46656098710423716,
      "grad_norm": 4.121676445007324,
      "learning_rate": 9.786693436282408e-05,
      "loss": 2.577968978881836,
      "memory(GiB)": 58.3,
      "step": 10890,
      "token_acc": 0.4979919678714859,
      "train_speed(iter/s)": 1.452152
    },
    {
      "epoch": 0.46677520243348614,
      "grad_norm": 3.376992702484131,
      "learning_rate": 9.786498923864796e-05,
      "loss": 2.4011980056762696,
      "memory(GiB)": 58.3,
      "step": 10895,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.45232
    },
    {
      "epoch": 0.4669894177627351,
      "grad_norm": 3.7245006561279297,
      "learning_rate": 9.786304324735267e-05,
      "loss": 2.729457473754883,
      "memory(GiB)": 58.3,
      "step": 10900,
      "token_acc": 0.48623853211009177,
      "train_speed(iter/s)": 1.452395
    },
    {
      "epoch": 0.46720363309198404,
      "grad_norm": 3.90267276763916,
      "learning_rate": 9.786109638897344e-05,
      "loss": 2.938713264465332,
      "memory(GiB)": 58.3,
      "step": 10905,
      "token_acc": 0.44108761329305135,
      "train_speed(iter/s)": 1.452512
    },
    {
      "epoch": 0.467417848421233,
      "grad_norm": 4.368808269500732,
      "learning_rate": 9.785914866354556e-05,
      "loss": 3.125321960449219,
      "memory(GiB)": 58.3,
      "step": 10910,
      "token_acc": 0.3785900783289817,
      "train_speed(iter/s)": 1.452511
    },
    {
      "epoch": 0.467632063750482,
      "grad_norm": 5.022944927215576,
      "learning_rate": 9.78572000711043e-05,
      "loss": 2.748683738708496,
      "memory(GiB)": 58.3,
      "step": 10915,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.452626
    },
    {
      "epoch": 0.4678462790797309,
      "grad_norm": 3.6183857917785645,
      "learning_rate": 9.785525061168497e-05,
      "loss": 2.657187652587891,
      "memory(GiB)": 58.3,
      "step": 10920,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.452518
    },
    {
      "epoch": 0.4680604944089799,
      "grad_norm": 4.752320289611816,
      "learning_rate": 9.785330028532288e-05,
      "loss": 2.315539741516113,
      "memory(GiB)": 58.3,
      "step": 10925,
      "token_acc": 0.5018181818181818,
      "train_speed(iter/s)": 1.45255
    },
    {
      "epoch": 0.4682747097382289,
      "grad_norm": 4.117373943328857,
      "learning_rate": 9.785134909205337e-05,
      "loss": 2.528944396972656,
      "memory(GiB)": 58.3,
      "step": 10930,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.4526
    },
    {
      "epoch": 0.4684889250674778,
      "grad_norm": 5.804306507110596,
      "learning_rate": 9.784939703191179e-05,
      "loss": 2.4984342575073244,
      "memory(GiB)": 58.3,
      "step": 10935,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.452635
    },
    {
      "epoch": 0.4687031403967268,
      "grad_norm": 3.236720561981201,
      "learning_rate": 9.784744410493348e-05,
      "loss": 2.6287277221679686,
      "memory(GiB)": 58.3,
      "step": 10940,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.452843
    },
    {
      "epoch": 0.46891735572597576,
      "grad_norm": 5.333811283111572,
      "learning_rate": 9.784549031115384e-05,
      "loss": 2.7113250732421874,
      "memory(GiB)": 58.3,
      "step": 10945,
      "token_acc": 0.44964028776978415,
      "train_speed(iter/s)": 1.452899
    },
    {
      "epoch": 0.46913157105522474,
      "grad_norm": 4.313626766204834,
      "learning_rate": 9.784353565060826e-05,
      "loss": 2.402693176269531,
      "memory(GiB)": 58.3,
      "step": 10950,
      "token_acc": 0.49421965317919075,
      "train_speed(iter/s)": 1.452959
    },
    {
      "epoch": 0.46934578638447366,
      "grad_norm": 3.7378852367401123,
      "learning_rate": 9.784158012333216e-05,
      "loss": 2.800084686279297,
      "memory(GiB)": 58.3,
      "step": 10955,
      "token_acc": 0.4424778761061947,
      "train_speed(iter/s)": 1.452848
    },
    {
      "epoch": 0.46956000171372264,
      "grad_norm": 3.3986082077026367,
      "learning_rate": 9.783962372936095e-05,
      "loss": 2.5746301651000976,
      "memory(GiB)": 58.3,
      "step": 10960,
      "token_acc": 0.4608695652173913,
      "train_speed(iter/s)": 1.453001
    },
    {
      "epoch": 0.4697742170429716,
      "grad_norm": 5.421916484832764,
      "learning_rate": 9.783766646873008e-05,
      "loss": 2.3186120986938477,
      "memory(GiB)": 58.3,
      "step": 10965,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.452974
    },
    {
      "epoch": 0.46998843237222054,
      "grad_norm": 4.455541133880615,
      "learning_rate": 9.7835708341475e-05,
      "loss": 2.440431594848633,
      "memory(GiB)": 58.3,
      "step": 10970,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.453093
    },
    {
      "epoch": 0.4702026477014695,
      "grad_norm": 4.785274028778076,
      "learning_rate": 9.78337493476312e-05,
      "loss": 2.561728858947754,
      "memory(GiB)": 58.3,
      "step": 10975,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.453272
    },
    {
      "epoch": 0.4704168630307185,
      "grad_norm": 6.452266693115234,
      "learning_rate": 9.783178948723415e-05,
      "loss": 2.5045833587646484,
      "memory(GiB)": 58.3,
      "step": 10980,
      "token_acc": 0.43728813559322033,
      "train_speed(iter/s)": 1.453093
    },
    {
      "epoch": 0.4706310783599674,
      "grad_norm": 4.0317511558532715,
      "learning_rate": 9.782982876031938e-05,
      "loss": 2.4897130966186523,
      "memory(GiB)": 58.3,
      "step": 10985,
      "token_acc": 0.44727272727272727,
      "train_speed(iter/s)": 1.452998
    },
    {
      "epoch": 0.4708452936892164,
      "grad_norm": 9.115554809570312,
      "learning_rate": 9.782786716692239e-05,
      "loss": 2.3984521865844726,
      "memory(GiB)": 58.3,
      "step": 10990,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.452987
    },
    {
      "epoch": 0.4710595090184654,
      "grad_norm": 4.206314563751221,
      "learning_rate": 9.782590470707871e-05,
      "loss": 2.9820087432861326,
      "memory(GiB)": 58.3,
      "step": 10995,
      "token_acc": 0.4431137724550898,
      "train_speed(iter/s)": 1.45302
    },
    {
      "epoch": 0.4712737243477143,
      "grad_norm": 5.37673282623291,
      "learning_rate": 9.78239413808239e-05,
      "loss": 2.4091548919677734,
      "memory(GiB)": 58.3,
      "step": 11000,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.45301
    },
    {
      "epoch": 0.4712737243477143,
      "eval_loss": 2.286198616027832,
      "eval_runtime": 14.3672,
      "eval_samples_per_second": 6.96,
      "eval_steps_per_second": 6.96,
      "eval_token_acc": 0.49477806788511747,
      "step": 11000
    },
    {
      "epoch": 0.4714879396769633,
      "grad_norm": 4.41751766204834,
      "learning_rate": 9.782197718819352e-05,
      "loss": 2.7890464782714846,
      "memory(GiB)": 58.3,
      "step": 11005,
      "token_acc": 0.48725212464589235,
      "train_speed(iter/s)": 1.450004
    },
    {
      "epoch": 0.47170215500621226,
      "grad_norm": 5.613940238952637,
      "learning_rate": 9.782001212922319e-05,
      "loss": 2.665138053894043,
      "memory(GiB)": 58.3,
      "step": 11010,
      "token_acc": 0.4372623574144487,
      "train_speed(iter/s)": 1.450127
    },
    {
      "epoch": 0.4719163703354612,
      "grad_norm": 5.2741265296936035,
      "learning_rate": 9.781804620394847e-05,
      "loss": 2.850432586669922,
      "memory(GiB)": 58.3,
      "step": 11015,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.450154
    },
    {
      "epoch": 0.47213058566471017,
      "grad_norm": 4.864946365356445,
      "learning_rate": 9.781607941240498e-05,
      "loss": 2.886791229248047,
      "memory(GiB)": 58.3,
      "step": 11020,
      "token_acc": 0.41904761904761906,
      "train_speed(iter/s)": 1.450107
    },
    {
      "epoch": 0.47234480099395915,
      "grad_norm": 3.490731716156006,
      "learning_rate": 9.781411175462836e-05,
      "loss": 2.668185234069824,
      "memory(GiB)": 58.3,
      "step": 11025,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.450274
    },
    {
      "epoch": 0.47255901632320807,
      "grad_norm": 4.4292097091674805,
      "learning_rate": 9.781214323065426e-05,
      "loss": 2.3660659790039062,
      "memory(GiB)": 58.3,
      "step": 11030,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.450323
    },
    {
      "epoch": 0.47277323165245705,
      "grad_norm": 9.111448287963867,
      "learning_rate": 9.781017384051832e-05,
      "loss": 2.639826774597168,
      "memory(GiB)": 58.3,
      "step": 11035,
      "token_acc": 0.4574898785425101,
      "train_speed(iter/s)": 1.450164
    },
    {
      "epoch": 0.47298744698170603,
      "grad_norm": 4.101097106933594,
      "learning_rate": 9.780820358425625e-05,
      "loss": 2.7708057403564452,
      "memory(GiB)": 58.3,
      "step": 11040,
      "token_acc": 0.4119718309859155,
      "train_speed(iter/s)": 1.450106
    },
    {
      "epoch": 0.47320166231095495,
      "grad_norm": 4.074931621551514,
      "learning_rate": 9.780623246190371e-05,
      "loss": 2.6240386962890625,
      "memory(GiB)": 58.3,
      "step": 11045,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.450164
    },
    {
      "epoch": 0.47341587764020393,
      "grad_norm": 4.386037349700928,
      "learning_rate": 9.780426047349642e-05,
      "loss": 2.390812301635742,
      "memory(GiB)": 58.3,
      "step": 11050,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.450212
    },
    {
      "epoch": 0.4736300929694529,
      "grad_norm": 3.773947238922119,
      "learning_rate": 9.78022876190701e-05,
      "loss": 2.5393697738647463,
      "memory(GiB)": 58.3,
      "step": 11055,
      "token_acc": 0.44360902255639095,
      "train_speed(iter/s)": 1.450069
    },
    {
      "epoch": 0.47384430829870183,
      "grad_norm": 4.300468921661377,
      "learning_rate": 9.780031389866053e-05,
      "loss": 2.258570098876953,
      "memory(GiB)": 58.3,
      "step": 11060,
      "token_acc": 0.5281385281385281,
      "train_speed(iter/s)": 1.449887
    },
    {
      "epoch": 0.4740585236279508,
      "grad_norm": 4.8900628089904785,
      "learning_rate": 9.77983393123034e-05,
      "loss": 2.48463134765625,
      "memory(GiB)": 58.3,
      "step": 11065,
      "token_acc": 0.4559386973180077,
      "train_speed(iter/s)": 1.450048
    },
    {
      "epoch": 0.4742727389571998,
      "grad_norm": 5.431396007537842,
      "learning_rate": 9.779636386003453e-05,
      "loss": 2.3649187088012695,
      "memory(GiB)": 58.3,
      "step": 11070,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.45013
    },
    {
      "epoch": 0.4744869542864487,
      "grad_norm": 3.571594715118408,
      "learning_rate": 9.77943875418897e-05,
      "loss": 2.6671390533447266,
      "memory(GiB)": 58.3,
      "step": 11075,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.450237
    },
    {
      "epoch": 0.4747011696156977,
      "grad_norm": 3.7833197116851807,
      "learning_rate": 9.77924103579047e-05,
      "loss": 2.6256244659423826,
      "memory(GiB)": 58.3,
      "step": 11080,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.450348
    },
    {
      "epoch": 0.4749153849449467,
      "grad_norm": 3.9306674003601074,
      "learning_rate": 9.779043230811534e-05,
      "loss": 2.424663543701172,
      "memory(GiB)": 58.3,
      "step": 11085,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.450376
    },
    {
      "epoch": 0.4751296002741956,
      "grad_norm": 8.253090858459473,
      "learning_rate": 9.778845339255749e-05,
      "loss": 2.9180736541748047,
      "memory(GiB)": 58.3,
      "step": 11090,
      "token_acc": 0.4290909090909091,
      "train_speed(iter/s)": 1.45026
    },
    {
      "epoch": 0.4753438156034446,
      "grad_norm": 4.530609607696533,
      "learning_rate": 9.778647361126696e-05,
      "loss": 2.809162712097168,
      "memory(GiB)": 58.3,
      "step": 11095,
      "token_acc": 0.4057971014492754,
      "train_speed(iter/s)": 1.450212
    },
    {
      "epoch": 0.47555803093269355,
      "grad_norm": 3.136714458465576,
      "learning_rate": 9.778449296427962e-05,
      "loss": 2.52059268951416,
      "memory(GiB)": 58.3,
      "step": 11100,
      "token_acc": 0.46048109965635736,
      "train_speed(iter/s)": 1.450245
    },
    {
      "epoch": 0.4757722462619425,
      "grad_norm": 4.584779262542725,
      "learning_rate": 9.778251145163139e-05,
      "loss": 2.684937286376953,
      "memory(GiB)": 58.3,
      "step": 11105,
      "token_acc": 0.46886446886446886,
      "train_speed(iter/s)": 1.450349
    },
    {
      "epoch": 0.47598646159119146,
      "grad_norm": 4.542789459228516,
      "learning_rate": 9.778052907335814e-05,
      "loss": 2.3787593841552734,
      "memory(GiB)": 58.3,
      "step": 11110,
      "token_acc": 0.48955223880597015,
      "train_speed(iter/s)": 1.450429
    },
    {
      "epoch": 0.47620067692044044,
      "grad_norm": 5.453615188598633,
      "learning_rate": 9.777854582949578e-05,
      "loss": 2.5061763763427733,
      "memory(GiB)": 58.3,
      "step": 11115,
      "token_acc": 0.46646341463414637,
      "train_speed(iter/s)": 1.45045
    },
    {
      "epoch": 0.4764148922496894,
      "grad_norm": 3.53247332572937,
      "learning_rate": 9.777656172008023e-05,
      "loss": 2.703365707397461,
      "memory(GiB)": 58.3,
      "step": 11120,
      "token_acc": 0.4650537634408602,
      "train_speed(iter/s)": 1.450509
    },
    {
      "epoch": 0.47662910757893834,
      "grad_norm": 3.5506818294525146,
      "learning_rate": 9.777457674514748e-05,
      "loss": 2.737448310852051,
      "memory(GiB)": 58.3,
      "step": 11125,
      "token_acc": 0.46107784431137727,
      "train_speed(iter/s)": 1.450372
    },
    {
      "epoch": 0.4768433229081873,
      "grad_norm": 4.337513446807861,
      "learning_rate": 9.777259090473341e-05,
      "loss": 2.4103200912475584,
      "memory(GiB)": 58.3,
      "step": 11130,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.450535
    },
    {
      "epoch": 0.4770575382374363,
      "grad_norm": 5.818124771118164,
      "learning_rate": 9.777060419887407e-05,
      "loss": 2.7565093994140626,
      "memory(GiB)": 58.3,
      "step": 11135,
      "token_acc": 0.4339622641509434,
      "train_speed(iter/s)": 1.450591
    },
    {
      "epoch": 0.4772717535666852,
      "grad_norm": 3.3356211185455322,
      "learning_rate": 9.776861662760541e-05,
      "loss": 2.607416534423828,
      "memory(GiB)": 58.3,
      "step": 11140,
      "token_acc": 0.47232472324723246,
      "train_speed(iter/s)": 1.450469
    },
    {
      "epoch": 0.4774859688959342,
      "grad_norm": 4.544053077697754,
      "learning_rate": 9.776662819096347e-05,
      "loss": 2.7298851013183594,
      "memory(GiB)": 58.3,
      "step": 11145,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.450424
    },
    {
      "epoch": 0.4777001842251832,
      "grad_norm": 5.424997329711914,
      "learning_rate": 9.776463888898423e-05,
      "loss": 2.64554443359375,
      "memory(GiB)": 58.3,
      "step": 11150,
      "token_acc": 0.392,
      "train_speed(iter/s)": 1.450504
    },
    {
      "epoch": 0.4779143995544321,
      "grad_norm": 4.7738542556762695,
      "learning_rate": 9.776264872170376e-05,
      "loss": 2.76705379486084,
      "memory(GiB)": 58.3,
      "step": 11155,
      "token_acc": 0.4379310344827586,
      "train_speed(iter/s)": 1.450424
    },
    {
      "epoch": 0.4781286148836811,
      "grad_norm": 4.916574478149414,
      "learning_rate": 9.77606576891581e-05,
      "loss": 2.580837631225586,
      "memory(GiB)": 58.3,
      "step": 11160,
      "token_acc": 0.452,
      "train_speed(iter/s)": 1.450398
    },
    {
      "epoch": 0.47834283021293006,
      "grad_norm": 5.020221710205078,
      "learning_rate": 9.775866579138332e-05,
      "loss": 2.681121253967285,
      "memory(GiB)": 58.3,
      "step": 11165,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.450373
    },
    {
      "epoch": 0.478557045542179,
      "grad_norm": 4.253873825073242,
      "learning_rate": 9.775667302841551e-05,
      "loss": 2.583617401123047,
      "memory(GiB)": 58.3,
      "step": 11170,
      "token_acc": 0.45722713864306785,
      "train_speed(iter/s)": 1.45027
    },
    {
      "epoch": 0.47877126087142796,
      "grad_norm": 5.4516167640686035,
      "learning_rate": 9.775467940029077e-05,
      "loss": 2.4560674667358398,
      "memory(GiB)": 58.3,
      "step": 11175,
      "token_acc": 0.4793388429752066,
      "train_speed(iter/s)": 1.450318
    },
    {
      "epoch": 0.47898547620067694,
      "grad_norm": 4.386411190032959,
      "learning_rate": 9.775268490704522e-05,
      "loss": 2.5876808166503906,
      "memory(GiB)": 58.3,
      "step": 11180,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.450354
    },
    {
      "epoch": 0.47919969152992586,
      "grad_norm": 4.466586589813232,
      "learning_rate": 9.775068954871498e-05,
      "loss": 2.7462501525878906,
      "memory(GiB)": 58.3,
      "step": 11185,
      "token_acc": 0.46863468634686345,
      "train_speed(iter/s)": 1.450413
    },
    {
      "epoch": 0.47941390685917484,
      "grad_norm": 5.115545749664307,
      "learning_rate": 9.774869332533622e-05,
      "loss": 2.7964197158813477,
      "memory(GiB)": 58.3,
      "step": 11190,
      "token_acc": 0.45625,
      "train_speed(iter/s)": 1.450541
    },
    {
      "epoch": 0.4796281221884238,
      "grad_norm": 4.668806076049805,
      "learning_rate": 9.774669623694507e-05,
      "loss": 2.9263195037841796,
      "memory(GiB)": 58.3,
      "step": 11195,
      "token_acc": 0.4169381107491857,
      "train_speed(iter/s)": 1.450589
    },
    {
      "epoch": 0.47984233751767275,
      "grad_norm": 4.503677845001221,
      "learning_rate": 9.774469828357773e-05,
      "loss": 2.2816009521484375,
      "memory(GiB)": 58.3,
      "step": 11200,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.450739
    },
    {
      "epoch": 0.4800565528469217,
      "grad_norm": 4.348319053649902,
      "learning_rate": 9.77426994652704e-05,
      "loss": 2.1784135818481447,
      "memory(GiB)": 58.3,
      "step": 11205,
      "token_acc": 0.5164473684210527,
      "train_speed(iter/s)": 1.450811
    },
    {
      "epoch": 0.4802707681761707,
      "grad_norm": 5.390242576599121,
      "learning_rate": 9.774069978205928e-05,
      "loss": 2.409230422973633,
      "memory(GiB)": 58.3,
      "step": 11210,
      "token_acc": 0.5022026431718062,
      "train_speed(iter/s)": 1.450864
    },
    {
      "epoch": 0.4804849835054196,
      "grad_norm": 3.67338490486145,
      "learning_rate": 9.77386992339806e-05,
      "loss": 2.435714912414551,
      "memory(GiB)": 58.3,
      "step": 11215,
      "token_acc": 0.5285714285714286,
      "train_speed(iter/s)": 1.450881
    },
    {
      "epoch": 0.4806991988346686,
      "grad_norm": 4.455507278442383,
      "learning_rate": 9.77366978210706e-05,
      "loss": 2.4204294204711916,
      "memory(GiB)": 58.3,
      "step": 11220,
      "token_acc": 0.45819397993311034,
      "train_speed(iter/s)": 1.450766
    },
    {
      "epoch": 0.4809134141639176,
      "grad_norm": 3.2890052795410156,
      "learning_rate": 9.773469554336553e-05,
      "loss": 2.987748718261719,
      "memory(GiB)": 58.3,
      "step": 11225,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.451006
    },
    {
      "epoch": 0.4811276294931665,
      "grad_norm": 3.461409330368042,
      "learning_rate": 9.773269240090169e-05,
      "loss": 2.811074066162109,
      "memory(GiB)": 58.3,
      "step": 11230,
      "token_acc": 0.4746268656716418,
      "train_speed(iter/s)": 1.451177
    },
    {
      "epoch": 0.4813418448224155,
      "grad_norm": 6.8326311111450195,
      "learning_rate": 9.773068839371534e-05,
      "loss": 2.378412628173828,
      "memory(GiB)": 58.3,
      "step": 11235,
      "token_acc": 0.5037878787878788,
      "train_speed(iter/s)": 1.451157
    },
    {
      "epoch": 0.48155606015166447,
      "grad_norm": 4.739665985107422,
      "learning_rate": 9.772868352184279e-05,
      "loss": 2.4834747314453125,
      "memory(GiB)": 58.3,
      "step": 11240,
      "token_acc": 0.4959677419354839,
      "train_speed(iter/s)": 1.451184
    },
    {
      "epoch": 0.4817702754809134,
      "grad_norm": 4.117530345916748,
      "learning_rate": 9.772667778532036e-05,
      "loss": 2.51016845703125,
      "memory(GiB)": 58.3,
      "step": 11245,
      "token_acc": 0.4717514124293785,
      "train_speed(iter/s)": 1.451296
    },
    {
      "epoch": 0.48198449081016237,
      "grad_norm": 4.38101863861084,
      "learning_rate": 9.77246711841844e-05,
      "loss": 2.8449834823608398,
      "memory(GiB)": 58.3,
      "step": 11250,
      "token_acc": 0.42207792207792205,
      "train_speed(iter/s)": 1.45116
    },
    {
      "epoch": 0.48219870613941135,
      "grad_norm": 5.53923225402832,
      "learning_rate": 9.772266371847125e-05,
      "loss": 2.6350059509277344,
      "memory(GiB)": 58.3,
      "step": 11255,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.451289
    },
    {
      "epoch": 0.48241292146866027,
      "grad_norm": 3.890557050704956,
      "learning_rate": 9.772065538821728e-05,
      "loss": 2.357133483886719,
      "memory(GiB)": 58.3,
      "step": 11260,
      "token_acc": 0.516,
      "train_speed(iter/s)": 1.451357
    },
    {
      "epoch": 0.48262713679790925,
      "grad_norm": 3.507786750793457,
      "learning_rate": 9.771864619345888e-05,
      "loss": 2.5293365478515626,
      "memory(GiB)": 58.3,
      "step": 11265,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.451479
    },
    {
      "epoch": 0.48284135212715823,
      "grad_norm": 3.9261813163757324,
      "learning_rate": 9.771663613423243e-05,
      "loss": 2.661027526855469,
      "memory(GiB)": 58.3,
      "step": 11270,
      "token_acc": 0.422360248447205,
      "train_speed(iter/s)": 1.451399
    },
    {
      "epoch": 0.48305556745640715,
      "grad_norm": 4.198917388916016,
      "learning_rate": 9.771462521057436e-05,
      "loss": 2.894261932373047,
      "memory(GiB)": 58.3,
      "step": 11275,
      "token_acc": 0.43312101910828027,
      "train_speed(iter/s)": 1.451358
    },
    {
      "epoch": 0.48326978278565613,
      "grad_norm": 3.858447551727295,
      "learning_rate": 9.771261342252109e-05,
      "loss": 2.839884567260742,
      "memory(GiB)": 58.3,
      "step": 11280,
      "token_acc": 0.43309859154929575,
      "train_speed(iter/s)": 1.451374
    },
    {
      "epoch": 0.4834839981149051,
      "grad_norm": 4.041822910308838,
      "learning_rate": 9.771060077010907e-05,
      "loss": 2.575081443786621,
      "memory(GiB)": 58.3,
      "step": 11285,
      "token_acc": 0.4692982456140351,
      "train_speed(iter/s)": 1.451444
    },
    {
      "epoch": 0.4836982134441541,
      "grad_norm": 4.165565490722656,
      "learning_rate": 9.770858725337477e-05,
      "loss": 2.52838020324707,
      "memory(GiB)": 58.3,
      "step": 11290,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.451671
    },
    {
      "epoch": 0.483912428773403,
      "grad_norm": 4.5889105796813965,
      "learning_rate": 9.770657287235465e-05,
      "loss": 2.9204345703125,
      "memory(GiB)": 58.3,
      "step": 11295,
      "token_acc": 0.421875,
      "train_speed(iter/s)": 1.451666
    },
    {
      "epoch": 0.484126644102652,
      "grad_norm": 4.789783477783203,
      "learning_rate": 9.770455762708521e-05,
      "loss": 2.543626594543457,
      "memory(GiB)": 58.3,
      "step": 11300,
      "token_acc": 0.46229508196721314,
      "train_speed(iter/s)": 1.451668
    },
    {
      "epoch": 0.48434085943190097,
      "grad_norm": 4.947669982910156,
      "learning_rate": 9.770254151760297e-05,
      "loss": 2.5403892517089846,
      "memory(GiB)": 58.3,
      "step": 11305,
      "token_acc": 0.4505928853754941,
      "train_speed(iter/s)": 1.45176
    },
    {
      "epoch": 0.4845550747611499,
      "grad_norm": 5.263824939727783,
      "learning_rate": 9.770052454394443e-05,
      "loss": 2.775040054321289,
      "memory(GiB)": 58.3,
      "step": 11310,
      "token_acc": 0.43548387096774194,
      "train_speed(iter/s)": 1.451967
    },
    {
      "epoch": 0.4847692900903989,
      "grad_norm": 5.169901371002197,
      "learning_rate": 9.769850670614613e-05,
      "loss": 2.853218841552734,
      "memory(GiB)": 58.3,
      "step": 11315,
      "token_acc": 0.4235294117647059,
      "train_speed(iter/s)": 1.451824
    },
    {
      "epoch": 0.48498350541964785,
      "grad_norm": 3.248640298843384,
      "learning_rate": 9.769648800424465e-05,
      "loss": 2.5355600357055663,
      "memory(GiB)": 58.3,
      "step": 11320,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.451889
    },
    {
      "epoch": 0.4851977207488968,
      "grad_norm": 3.849860668182373,
      "learning_rate": 9.769446843827655e-05,
      "loss": 2.318301773071289,
      "memory(GiB)": 58.3,
      "step": 11325,
      "token_acc": 0.538135593220339,
      "train_speed(iter/s)": 1.451945
    },
    {
      "epoch": 0.48541193607814576,
      "grad_norm": 4.20799446105957,
      "learning_rate": 9.769244800827841e-05,
      "loss": 2.681135368347168,
      "memory(GiB)": 58.3,
      "step": 11330,
      "token_acc": 0.4567901234567901,
      "train_speed(iter/s)": 1.452038
    },
    {
      "epoch": 0.48562615140739473,
      "grad_norm": 5.085803031921387,
      "learning_rate": 9.769042671428682e-05,
      "loss": 2.6751615524291994,
      "memory(GiB)": 58.3,
      "step": 11335,
      "token_acc": 0.4470198675496689,
      "train_speed(iter/s)": 1.452087
    },
    {
      "epoch": 0.48584036673664366,
      "grad_norm": 3.483921766281128,
      "learning_rate": 9.768840455633842e-05,
      "loss": 2.9844387054443358,
      "memory(GiB)": 58.3,
      "step": 11340,
      "token_acc": 0.4158415841584158,
      "train_speed(iter/s)": 1.452105
    },
    {
      "epoch": 0.48605458206589264,
      "grad_norm": 4.813082218170166,
      "learning_rate": 9.768638153446986e-05,
      "loss": 2.9042110443115234,
      "memory(GiB)": 58.3,
      "step": 11345,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.452154
    },
    {
      "epoch": 0.4862687973951416,
      "grad_norm": 4.587554931640625,
      "learning_rate": 9.768435764871775e-05,
      "loss": 2.5322193145751952,
      "memory(GiB)": 58.3,
      "step": 11350,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.452256
    },
    {
      "epoch": 0.48648301272439054,
      "grad_norm": 43.5738410949707,
      "learning_rate": 9.768233289911877e-05,
      "loss": 2.3911769866943358,
      "memory(GiB)": 58.3,
      "step": 11355,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.452392
    },
    {
      "epoch": 0.4866972280536395,
      "grad_norm": 4.606513977050781,
      "learning_rate": 9.768030728570959e-05,
      "loss": 2.770819664001465,
      "memory(GiB)": 58.3,
      "step": 11360,
      "token_acc": 0.4409722222222222,
      "train_speed(iter/s)": 1.452449
    },
    {
      "epoch": 0.4869114433828885,
      "grad_norm": 5.725339889526367,
      "learning_rate": 9.767828080852693e-05,
      "loss": 2.4176609039306642,
      "memory(GiB)": 58.3,
      "step": 11365,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.45258
    },
    {
      "epoch": 0.4871256587121374,
      "grad_norm": 4.472254753112793,
      "learning_rate": 9.767625346760748e-05,
      "loss": 2.5790700912475586,
      "memory(GiB)": 58.3,
      "step": 11370,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.452634
    },
    {
      "epoch": 0.4873398740413864,
      "grad_norm": 4.8378167152404785,
      "learning_rate": 9.767422526298798e-05,
      "loss": 2.669243812561035,
      "memory(GiB)": 58.3,
      "step": 11375,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.45254
    },
    {
      "epoch": 0.4875540893706354,
      "grad_norm": 3.5412521362304688,
      "learning_rate": 9.767219619470516e-05,
      "loss": 2.5998735427856445,
      "memory(GiB)": 58.3,
      "step": 11380,
      "token_acc": 0.46439628482972134,
      "train_speed(iter/s)": 1.452617
    },
    {
      "epoch": 0.4877683046998843,
      "grad_norm": 4.090600490570068,
      "learning_rate": 9.767016626279577e-05,
      "loss": 2.7880121231079102,
      "memory(GiB)": 58.3,
      "step": 11385,
      "token_acc": 0.4384057971014493,
      "train_speed(iter/s)": 1.452739
    },
    {
      "epoch": 0.4879825200291333,
      "grad_norm": 4.44712495803833,
      "learning_rate": 9.766813546729663e-05,
      "loss": 2.4887435913085936,
      "memory(GiB)": 58.3,
      "step": 11390,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.452517
    },
    {
      "epoch": 0.48819673535838226,
      "grad_norm": 4.05634069442749,
      "learning_rate": 9.766610380824448e-05,
      "loss": 2.8442813873291017,
      "memory(GiB)": 58.3,
      "step": 11395,
      "token_acc": 0.414985590778098,
      "train_speed(iter/s)": 1.45259
    },
    {
      "epoch": 0.4884109506876312,
      "grad_norm": 4.251303672790527,
      "learning_rate": 9.766407128567617e-05,
      "loss": 2.378520202636719,
      "memory(GiB)": 58.3,
      "step": 11400,
      "token_acc": 0.451505016722408,
      "train_speed(iter/s)": 1.452563
    },
    {
      "epoch": 0.48862516601688016,
      "grad_norm": 4.743833541870117,
      "learning_rate": 9.766203789962846e-05,
      "loss": 2.995096778869629,
      "memory(GiB)": 58.3,
      "step": 11405,
      "token_acc": 0.38823529411764707,
      "train_speed(iter/s)": 1.452374
    },
    {
      "epoch": 0.48883938134612914,
      "grad_norm": 5.054455757141113,
      "learning_rate": 9.766000365013824e-05,
      "loss": 2.7642860412597656,
      "memory(GiB)": 58.3,
      "step": 11410,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.452323
    },
    {
      "epoch": 0.48905359667537807,
      "grad_norm": 4.550734996795654,
      "learning_rate": 9.765796853724235e-05,
      "loss": 2.798261260986328,
      "memory(GiB)": 58.3,
      "step": 11415,
      "token_acc": 0.40350877192982454,
      "train_speed(iter/s)": 1.452376
    },
    {
      "epoch": 0.48926781200462705,
      "grad_norm": 3.1900641918182373,
      "learning_rate": 9.765593256097764e-05,
      "loss": 2.6664859771728517,
      "memory(GiB)": 58.3,
      "step": 11420,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.452396
    },
    {
      "epoch": 0.489482027333876,
      "grad_norm": 4.950425624847412,
      "learning_rate": 9.765389572138103e-05,
      "loss": 2.4235610961914062,
      "memory(GiB)": 58.3,
      "step": 11425,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.452476
    },
    {
      "epoch": 0.48969624266312495,
      "grad_norm": 4.0858154296875,
      "learning_rate": 9.765185801848938e-05,
      "loss": 2.7007659912109374,
      "memory(GiB)": 58.3,
      "step": 11430,
      "token_acc": 0.4520123839009288,
      "train_speed(iter/s)": 1.45247
    },
    {
      "epoch": 0.4899104579923739,
      "grad_norm": 5.181708335876465,
      "learning_rate": 9.764981945233962e-05,
      "loss": 2.6881858825683596,
      "memory(GiB)": 58.3,
      "step": 11435,
      "token_acc": 0.4355400696864111,
      "train_speed(iter/s)": 1.452572
    },
    {
      "epoch": 0.4901246733216229,
      "grad_norm": 4.984737396240234,
      "learning_rate": 9.764778002296868e-05,
      "loss": 2.8407032012939455,
      "memory(GiB)": 58.3,
      "step": 11440,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.452719
    },
    {
      "epoch": 0.49033888865087183,
      "grad_norm": 4.215512752532959,
      "learning_rate": 9.764573973041352e-05,
      "loss": 2.6679666519165037,
      "memory(GiB)": 58.3,
      "step": 11445,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.452746
    },
    {
      "epoch": 0.4905531039801208,
      "grad_norm": 5.14586877822876,
      "learning_rate": 9.764369857471109e-05,
      "loss": 2.652199363708496,
      "memory(GiB)": 58.3,
      "step": 11450,
      "token_acc": 0.43260188087774293,
      "train_speed(iter/s)": 1.452642
    },
    {
      "epoch": 0.4907673193093698,
      "grad_norm": 4.5310845375061035,
      "learning_rate": 9.764165655589835e-05,
      "loss": 2.650895118713379,
      "memory(GiB)": 58.3,
      "step": 11455,
      "token_acc": 0.4753623188405797,
      "train_speed(iter/s)": 1.452825
    },
    {
      "epoch": 0.49098153463861877,
      "grad_norm": 5.773660182952881,
      "learning_rate": 9.763961367401231e-05,
      "loss": 2.690845489501953,
      "memory(GiB)": 58.3,
      "step": 11460,
      "token_acc": 0.4157706093189964,
      "train_speed(iter/s)": 1.452876
    },
    {
      "epoch": 0.4911957499678677,
      "grad_norm": 3.6825263500213623,
      "learning_rate": 9.763756992909e-05,
      "loss": 2.5997386932373048,
      "memory(GiB)": 58.3,
      "step": 11465,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.452998
    },
    {
      "epoch": 0.49140996529711667,
      "grad_norm": 5.825430870056152,
      "learning_rate": 9.763552532116841e-05,
      "loss": 2.6364845275878905,
      "memory(GiB)": 58.3,
      "step": 11470,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.453105
    },
    {
      "epoch": 0.49162418062636565,
      "grad_norm": 4.11679220199585,
      "learning_rate": 9.76334798502846e-05,
      "loss": 2.563291549682617,
      "memory(GiB)": 58.3,
      "step": 11475,
      "token_acc": 0.4262295081967213,
      "train_speed(iter/s)": 1.45313
    },
    {
      "epoch": 0.49183839595561457,
      "grad_norm": 4.311800003051758,
      "learning_rate": 9.763143351647561e-05,
      "loss": 2.289215850830078,
      "memory(GiB)": 58.3,
      "step": 11480,
      "token_acc": 0.46494464944649444,
      "train_speed(iter/s)": 1.453215
    },
    {
      "epoch": 0.49205261128486355,
      "grad_norm": 7.280170440673828,
      "learning_rate": 9.762938631977852e-05,
      "loss": 2.361956024169922,
      "memory(GiB)": 58.3,
      "step": 11485,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.453248
    },
    {
      "epoch": 0.49226682661411253,
      "grad_norm": 7.868833541870117,
      "learning_rate": 9.762733826023042e-05,
      "loss": 2.538089370727539,
      "memory(GiB)": 58.3,
      "step": 11490,
      "token_acc": 0.4628099173553719,
      "train_speed(iter/s)": 1.453321
    },
    {
      "epoch": 0.49248104194336145,
      "grad_norm": 4.240434169769287,
      "learning_rate": 9.76252893378684e-05,
      "loss": 2.8499610900878904,
      "memory(GiB)": 58.3,
      "step": 11495,
      "token_acc": 0.43309859154929575,
      "train_speed(iter/s)": 1.453397
    },
    {
      "epoch": 0.49269525727261043,
      "grad_norm": 4.263349533081055,
      "learning_rate": 9.76232395527296e-05,
      "loss": 2.6419279098510744,
      "memory(GiB)": 58.3,
      "step": 11500,
      "token_acc": 0.4895833333333333,
      "train_speed(iter/s)": 1.453492
    },
    {
      "epoch": 0.49269525727261043,
      "eval_loss": 2.122830867767334,
      "eval_runtime": 13.3411,
      "eval_samples_per_second": 7.496,
      "eval_steps_per_second": 7.496,
      "eval_token_acc": 0.47485207100591714,
      "step": 11500
    },
    {
      "epoch": 0.4929094726018594,
      "grad_norm": 3.7236857414245605,
      "learning_rate": 9.762118890485115e-05,
      "loss": 2.3778526306152346,
      "memory(GiB)": 58.3,
      "step": 11505,
      "token_acc": 0.4754601226993865,
      "train_speed(iter/s)": 1.45099
    },
    {
      "epoch": 0.49312368793110833,
      "grad_norm": 12.476144790649414,
      "learning_rate": 9.761913739427017e-05,
      "loss": 2.415938377380371,
      "memory(GiB)": 58.3,
      "step": 11510,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.451143
    },
    {
      "epoch": 0.4933379032603573,
      "grad_norm": 3.7864506244659424,
      "learning_rate": 9.761708502102384e-05,
      "loss": 2.613703727722168,
      "memory(GiB)": 58.3,
      "step": 11515,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.451055
    },
    {
      "epoch": 0.4935521185896063,
      "grad_norm": 7.46455192565918,
      "learning_rate": 9.761503178514938e-05,
      "loss": 2.7990486145019533,
      "memory(GiB)": 58.3,
      "step": 11520,
      "token_acc": 0.4426751592356688,
      "train_speed(iter/s)": 1.451047
    },
    {
      "epoch": 0.4937663339188552,
      "grad_norm": 4.192196846008301,
      "learning_rate": 9.761297768668393e-05,
      "loss": 2.8287534713745117,
      "memory(GiB)": 58.3,
      "step": 11525,
      "token_acc": 0.445141065830721,
      "train_speed(iter/s)": 1.450996
    },
    {
      "epoch": 0.4939805492481042,
      "grad_norm": 3.70595121383667,
      "learning_rate": 9.761092272566472e-05,
      "loss": 2.641410064697266,
      "memory(GiB)": 58.3,
      "step": 11530,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.450973
    },
    {
      "epoch": 0.4941947645773532,
      "grad_norm": 6.178360939025879,
      "learning_rate": 9.7608866902129e-05,
      "loss": 2.670266342163086,
      "memory(GiB)": 58.3,
      "step": 11535,
      "token_acc": 0.45075757575757575,
      "train_speed(iter/s)": 1.451131
    },
    {
      "epoch": 0.4944089799066021,
      "grad_norm": 3.9855008125305176,
      "learning_rate": 9.760681021611398e-05,
      "loss": 2.3463743209838865,
      "memory(GiB)": 58.3,
      "step": 11540,
      "token_acc": 0.5064516129032258,
      "train_speed(iter/s)": 1.45128
    },
    {
      "epoch": 0.4946231952358511,
      "grad_norm": 3.2524452209472656,
      "learning_rate": 9.760475266765694e-05,
      "loss": 2.5913551330566404,
      "memory(GiB)": 58.3,
      "step": 11545,
      "token_acc": 0.3978102189781022,
      "train_speed(iter/s)": 1.451253
    },
    {
      "epoch": 0.49483741056510006,
      "grad_norm": 6.220343589782715,
      "learning_rate": 9.760269425679514e-05,
      "loss": 2.8090896606445312,
      "memory(GiB)": 58.3,
      "step": 11550,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.451357
    },
    {
      "epoch": 0.495051625894349,
      "grad_norm": 4.260661602020264,
      "learning_rate": 9.760063498356589e-05,
      "loss": 2.58850154876709,
      "memory(GiB)": 58.3,
      "step": 11555,
      "token_acc": 0.48598130841121495,
      "train_speed(iter/s)": 1.451495
    },
    {
      "epoch": 0.49526584122359796,
      "grad_norm": 3.7763559818267822,
      "learning_rate": 9.759857484800647e-05,
      "loss": 2.461856460571289,
      "memory(GiB)": 58.3,
      "step": 11560,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.451426
    },
    {
      "epoch": 0.49548005655284694,
      "grad_norm": 4.351131439208984,
      "learning_rate": 9.759651385015423e-05,
      "loss": 2.551533317565918,
      "memory(GiB)": 58.3,
      "step": 11565,
      "token_acc": 0.46449704142011833,
      "train_speed(iter/s)": 1.451246
    },
    {
      "epoch": 0.49569427188209586,
      "grad_norm": 4.711762428283691,
      "learning_rate": 9.759445199004649e-05,
      "loss": 2.5739336013793945,
      "memory(GiB)": 58.3,
      "step": 11570,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.451283
    },
    {
      "epoch": 0.49590848721134484,
      "grad_norm": 4.900021553039551,
      "learning_rate": 9.75923892677206e-05,
      "loss": 2.4031457901000977,
      "memory(GiB)": 58.3,
      "step": 11575,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.451165
    },
    {
      "epoch": 0.4961227025405938,
      "grad_norm": 4.870184898376465,
      "learning_rate": 9.759032568321395e-05,
      "loss": 2.3678571701049806,
      "memory(GiB)": 58.3,
      "step": 11580,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.451148
    },
    {
      "epoch": 0.49633691786984274,
      "grad_norm": 4.825766086578369,
      "learning_rate": 9.758826123656388e-05,
      "loss": 2.607929801940918,
      "memory(GiB)": 58.3,
      "step": 11585,
      "token_acc": 0.45819397993311034,
      "train_speed(iter/s)": 1.451158
    },
    {
      "epoch": 0.4965511331990917,
      "grad_norm": 3.561554193496704,
      "learning_rate": 9.758619592780784e-05,
      "loss": 2.5645885467529297,
      "memory(GiB)": 58.3,
      "step": 11590,
      "token_acc": 0.49258160237388726,
      "train_speed(iter/s)": 1.451167
    },
    {
      "epoch": 0.4967653485283407,
      "grad_norm": 5.883943557739258,
      "learning_rate": 9.758412975698321e-05,
      "loss": 2.891803169250488,
      "memory(GiB)": 58.3,
      "step": 11595,
      "token_acc": 0.4241573033707865,
      "train_speed(iter/s)": 1.451143
    },
    {
      "epoch": 0.4969795638575896,
      "grad_norm": 4.955366611480713,
      "learning_rate": 9.758206272412743e-05,
      "loss": 2.5933320999145506,
      "memory(GiB)": 58.3,
      "step": 11600,
      "token_acc": 0.46062992125984253,
      "train_speed(iter/s)": 1.451096
    },
    {
      "epoch": 0.4971937791868386,
      "grad_norm": 3.6083242893218994,
      "learning_rate": 9.757999482927795e-05,
      "loss": 2.7880538940429687,
      "memory(GiB)": 58.3,
      "step": 11605,
      "token_acc": 0.3843843843843844,
      "train_speed(iter/s)": 1.450971
    },
    {
      "epoch": 0.4974079945160876,
      "grad_norm": 3.6765027046203613,
      "learning_rate": 9.757792607247224e-05,
      "loss": 2.4341617584228517,
      "memory(GiB)": 58.3,
      "step": 11610,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.450936
    },
    {
      "epoch": 0.4976222098453365,
      "grad_norm": 4.222357749938965,
      "learning_rate": 9.757585645374777e-05,
      "loss": 2.667819023132324,
      "memory(GiB)": 58.3,
      "step": 11615,
      "token_acc": 0.4618181818181818,
      "train_speed(iter/s)": 1.451034
    },
    {
      "epoch": 0.4978364251745855,
      "grad_norm": 4.218835353851318,
      "learning_rate": 9.757378597314201e-05,
      "loss": 2.8918861389160155,
      "memory(GiB)": 58.3,
      "step": 11620,
      "token_acc": 0.4200626959247649,
      "train_speed(iter/s)": 1.451137
    },
    {
      "epoch": 0.49805064050383446,
      "grad_norm": 4.356986999511719,
      "learning_rate": 9.75717146306925e-05,
      "loss": 2.3594308853149415,
      "memory(GiB)": 58.3,
      "step": 11625,
      "token_acc": 0.5103734439834025,
      "train_speed(iter/s)": 1.451228
    },
    {
      "epoch": 0.49826485583308344,
      "grad_norm": 5.578523635864258,
      "learning_rate": 9.756964242643674e-05,
      "loss": 2.4100622177124023,
      "memory(GiB)": 58.3,
      "step": 11630,
      "token_acc": 0.483271375464684,
      "train_speed(iter/s)": 1.451132
    },
    {
      "epoch": 0.49847907116233237,
      "grad_norm": 4.419356346130371,
      "learning_rate": 9.75675693604123e-05,
      "loss": 2.533791351318359,
      "memory(GiB)": 58.3,
      "step": 11635,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.45122
    },
    {
      "epoch": 0.49869328649158134,
      "grad_norm": 3.9564931392669678,
      "learning_rate": 9.756549543265671e-05,
      "loss": 2.726533126831055,
      "memory(GiB)": 58.3,
      "step": 11640,
      "token_acc": 0.4353312302839117,
      "train_speed(iter/s)": 1.451084
    },
    {
      "epoch": 0.4989075018208303,
      "grad_norm": 4.707624435424805,
      "learning_rate": 9.756342064320756e-05,
      "loss": 2.6874515533447267,
      "memory(GiB)": 58.3,
      "step": 11645,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.451192
    },
    {
      "epoch": 0.49912171715007925,
      "grad_norm": 3.3866236209869385,
      "learning_rate": 9.756134499210243e-05,
      "loss": 2.8155418395996095,
      "memory(GiB)": 58.3,
      "step": 11650,
      "token_acc": 0.42363112391930835,
      "train_speed(iter/s)": 1.451229
    },
    {
      "epoch": 0.4993359324793282,
      "grad_norm": 5.313538074493408,
      "learning_rate": 9.755926847937891e-05,
      "loss": 2.5449777603149415,
      "memory(GiB)": 58.3,
      "step": 11655,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.451112
    },
    {
      "epoch": 0.4995501478085772,
      "grad_norm": 3.765958547592163,
      "learning_rate": 9.755719110507462e-05,
      "loss": 2.7128576278686523,
      "memory(GiB)": 58.3,
      "step": 11660,
      "token_acc": 0.45307443365695793,
      "train_speed(iter/s)": 1.451031
    },
    {
      "epoch": 0.49976436313782613,
      "grad_norm": 4.323287010192871,
      "learning_rate": 9.75551128692272e-05,
      "loss": 2.391266632080078,
      "memory(GiB)": 58.3,
      "step": 11665,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.450972
    },
    {
      "epoch": 0.4999785784670751,
      "grad_norm": 3.935492753982544,
      "learning_rate": 9.755303377187433e-05,
      "loss": 2.717921829223633,
      "memory(GiB)": 58.3,
      "step": 11670,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.451131
    },
    {
      "epoch": 0.5001927937963241,
      "grad_norm": 5.324994087219238,
      "learning_rate": 9.755095381305362e-05,
      "loss": 2.6325710296630858,
      "memory(GiB)": 58.3,
      "step": 11675,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.451327
    },
    {
      "epoch": 0.500407009125573,
      "grad_norm": 5.840517997741699,
      "learning_rate": 9.754887299280277e-05,
      "loss": 2.7376882553100588,
      "memory(GiB)": 58.3,
      "step": 11680,
      "token_acc": 0.43097643097643096,
      "train_speed(iter/s)": 1.451365
    },
    {
      "epoch": 0.5006212244548219,
      "grad_norm": 4.677224159240723,
      "learning_rate": 9.754679131115949e-05,
      "loss": 2.4044061660766602,
      "memory(GiB)": 58.3,
      "step": 11685,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.451416
    },
    {
      "epoch": 0.500835439784071,
      "grad_norm": 4.544371128082275,
      "learning_rate": 9.754470876816148e-05,
      "loss": 2.551678276062012,
      "memory(GiB)": 58.3,
      "step": 11690,
      "token_acc": 0.44396551724137934,
      "train_speed(iter/s)": 1.451492
    },
    {
      "epoch": 0.5010496551133199,
      "grad_norm": 4.3035054206848145,
      "learning_rate": 9.754262536384649e-05,
      "loss": 2.9027379989624023,
      "memory(GiB)": 58.3,
      "step": 11695,
      "token_acc": 0.4576923076923077,
      "train_speed(iter/s)": 1.451529
    },
    {
      "epoch": 0.5012638704425688,
      "grad_norm": 4.433408737182617,
      "learning_rate": 9.754054109825223e-05,
      "loss": 2.643610191345215,
      "memory(GiB)": 58.3,
      "step": 11700,
      "token_acc": 0.4669260700389105,
      "train_speed(iter/s)": 1.451467
    },
    {
      "epoch": 0.5014780857718178,
      "grad_norm": 3.3892669677734375,
      "learning_rate": 9.753845597141647e-05,
      "loss": 2.4351217269897463,
      "memory(GiB)": 58.3,
      "step": 11705,
      "token_acc": 0.5020080321285141,
      "train_speed(iter/s)": 1.451427
    },
    {
      "epoch": 0.5016923011010668,
      "grad_norm": 4.115910053253174,
      "learning_rate": 9.753636998337698e-05,
      "loss": 2.770868682861328,
      "memory(GiB)": 58.3,
      "step": 11710,
      "token_acc": 0.44314868804664725,
      "train_speed(iter/s)": 1.451487
    },
    {
      "epoch": 0.5019065164303158,
      "grad_norm": 4.731926918029785,
      "learning_rate": 9.753428313417155e-05,
      "loss": 3.0838483810424804,
      "memory(GiB)": 58.3,
      "step": 11715,
      "token_acc": 0.4070175438596491,
      "train_speed(iter/s)": 1.45147
    },
    {
      "epoch": 0.5021207317595647,
      "grad_norm": 4.004453182220459,
      "learning_rate": 9.7532195423838e-05,
      "loss": 2.358524131774902,
      "memory(GiB)": 58.3,
      "step": 11720,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.451402
    },
    {
      "epoch": 0.5023349470888137,
      "grad_norm": 4.105185508728027,
      "learning_rate": 9.753010685241415e-05,
      "loss": 2.7120088577270507,
      "memory(GiB)": 58.3,
      "step": 11725,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.451454
    },
    {
      "epoch": 0.5025491624180627,
      "grad_norm": 4.423943519592285,
      "learning_rate": 9.752801741993781e-05,
      "loss": 2.531999206542969,
      "memory(GiB)": 58.3,
      "step": 11730,
      "token_acc": 0.48497854077253216,
      "train_speed(iter/s)": 1.451519
    },
    {
      "epoch": 0.5027633777473116,
      "grad_norm": 4.0634846687316895,
      "learning_rate": 9.752592712644686e-05,
      "loss": 2.777116394042969,
      "memory(GiB)": 58.3,
      "step": 11735,
      "token_acc": 0.42382271468144045,
      "train_speed(iter/s)": 1.451661
    },
    {
      "epoch": 0.5029775930765605,
      "grad_norm": 5.358497619628906,
      "learning_rate": 9.752383597197916e-05,
      "loss": 2.463951301574707,
      "memory(GiB)": 58.3,
      "step": 11740,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.451653
    },
    {
      "epoch": 0.5031918084058096,
      "grad_norm": 3.7360455989837646,
      "learning_rate": 9.752174395657259e-05,
      "loss": 2.750463104248047,
      "memory(GiB)": 58.3,
      "step": 11745,
      "token_acc": 0.4212218649517685,
      "train_speed(iter/s)": 1.451706
    },
    {
      "epoch": 0.5034060237350585,
      "grad_norm": 3.7836999893188477,
      "learning_rate": 9.751965108026505e-05,
      "loss": 2.4442745208740235,
      "memory(GiB)": 58.3,
      "step": 11750,
      "token_acc": 0.511400651465798,
      "train_speed(iter/s)": 1.451742
    },
    {
      "epoch": 0.5036202390643074,
      "grad_norm": 4.425050735473633,
      "learning_rate": 9.751755734309443e-05,
      "loss": 2.639969825744629,
      "memory(GiB)": 58.3,
      "step": 11755,
      "token_acc": 0.4353312302839117,
      "train_speed(iter/s)": 1.451744
    },
    {
      "epoch": 0.5038344543935565,
      "grad_norm": 4.007545471191406,
      "learning_rate": 9.751546274509871e-05,
      "loss": 2.988727569580078,
      "memory(GiB)": 58.3,
      "step": 11760,
      "token_acc": 0.45907473309608543,
      "train_speed(iter/s)": 1.451953
    },
    {
      "epoch": 0.5040486697228054,
      "grad_norm": 4.643023490905762,
      "learning_rate": 9.75133672863158e-05,
      "loss": 2.9101932525634764,
      "memory(GiB)": 58.3,
      "step": 11765,
      "token_acc": 0.43934426229508194,
      "train_speed(iter/s)": 1.452005
    },
    {
      "epoch": 0.5042628850520543,
      "grad_norm": 4.114360809326172,
      "learning_rate": 9.751127096678366e-05,
      "loss": 2.6409418106079103,
      "memory(GiB)": 58.3,
      "step": 11770,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.452116
    },
    {
      "epoch": 0.5044771003813033,
      "grad_norm": 3.6016287803649902,
      "learning_rate": 9.75091737865403e-05,
      "loss": 2.69439811706543,
      "memory(GiB)": 58.3,
      "step": 11775,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.452114
    },
    {
      "epoch": 0.5046913157105523,
      "grad_norm": 3.8433313369750977,
      "learning_rate": 9.750707574562368e-05,
      "loss": 2.542442512512207,
      "memory(GiB)": 58.3,
      "step": 11780,
      "token_acc": 0.5082508250825083,
      "train_speed(iter/s)": 1.452184
    },
    {
      "epoch": 0.5049055310398012,
      "grad_norm": 4.735212802886963,
      "learning_rate": 9.75049768440718e-05,
      "loss": 2.599843215942383,
      "memory(GiB)": 58.3,
      "step": 11785,
      "token_acc": 0.511326860841424,
      "train_speed(iter/s)": 1.452248
    },
    {
      "epoch": 0.5051197463690502,
      "grad_norm": 4.743094444274902,
      "learning_rate": 9.750287708192272e-05,
      "loss": 2.8545053482055662,
      "memory(GiB)": 58.3,
      "step": 11790,
      "token_acc": 0.3988439306358382,
      "train_speed(iter/s)": 1.452284
    },
    {
      "epoch": 0.5053339616982991,
      "grad_norm": 4.007525444030762,
      "learning_rate": 9.750077645921445e-05,
      "loss": 2.507660675048828,
      "memory(GiB)": 58.3,
      "step": 11795,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.452403
    },
    {
      "epoch": 0.5055481770275481,
      "grad_norm": 3.828134536743164,
      "learning_rate": 9.749867497598508e-05,
      "loss": 2.9148635864257812,
      "memory(GiB)": 58.3,
      "step": 11800,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.452611
    },
    {
      "epoch": 0.5057623923567971,
      "grad_norm": 4.234160900115967,
      "learning_rate": 9.749657263227263e-05,
      "loss": 2.424146270751953,
      "memory(GiB)": 58.3,
      "step": 11805,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.452753
    },
    {
      "epoch": 0.505976607686046,
      "grad_norm": 8.152174949645996,
      "learning_rate": 9.749446942811522e-05,
      "loss": 2.65775146484375,
      "memory(GiB)": 58.3,
      "step": 11810,
      "token_acc": 0.4632352941176471,
      "train_speed(iter/s)": 1.452743
    },
    {
      "epoch": 0.5061908230152949,
      "grad_norm": 4.062210559844971,
      "learning_rate": 9.749236536355094e-05,
      "loss": 2.568818473815918,
      "memory(GiB)": 58.3,
      "step": 11815,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.452694
    },
    {
      "epoch": 0.506405038344544,
      "grad_norm": 5.438949108123779,
      "learning_rate": 9.749026043861791e-05,
      "loss": 2.648557662963867,
      "memory(GiB)": 58.3,
      "step": 11820,
      "token_acc": 0.4491525423728814,
      "train_speed(iter/s)": 1.452805
    },
    {
      "epoch": 0.5066192536737929,
      "grad_norm": 5.653181076049805,
      "learning_rate": 9.748815465335426e-05,
      "loss": 2.7346431732177736,
      "memory(GiB)": 58.3,
      "step": 11825,
      "token_acc": 0.4573643410852713,
      "train_speed(iter/s)": 1.452829
    },
    {
      "epoch": 0.5068334690030418,
      "grad_norm": 4.348195552825928,
      "learning_rate": 9.748604800779814e-05,
      "loss": 2.5485939025878905,
      "memory(GiB)": 58.3,
      "step": 11830,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.452821
    },
    {
      "epoch": 0.5070476843322909,
      "grad_norm": 4.624939441680908,
      "learning_rate": 9.748394050198773e-05,
      "loss": 2.787402534484863,
      "memory(GiB)": 58.3,
      "step": 11835,
      "token_acc": 0.4076655052264808,
      "train_speed(iter/s)": 1.452931
    },
    {
      "epoch": 0.5072618996615398,
      "grad_norm": 4.1721882820129395,
      "learning_rate": 9.748183213596118e-05,
      "loss": 2.7364587783813477,
      "memory(GiB)": 58.3,
      "step": 11840,
      "token_acc": 0.45084745762711864,
      "train_speed(iter/s)": 1.452854
    },
    {
      "epoch": 0.5074761149907887,
      "grad_norm": 3.9521102905273438,
      "learning_rate": 9.747972290975671e-05,
      "loss": 2.575947952270508,
      "memory(GiB)": 58.3,
      "step": 11845,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.452885
    },
    {
      "epoch": 0.5076903303200377,
      "grad_norm": 3.7558672428131104,
      "learning_rate": 9.747761282341253e-05,
      "loss": 2.4672239303588865,
      "memory(GiB)": 58.3,
      "step": 11850,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.452844
    },
    {
      "epoch": 0.5079045456492867,
      "grad_norm": 4.29282283782959,
      "learning_rate": 9.747550187696684e-05,
      "loss": 2.7997127532958985,
      "memory(GiB)": 58.3,
      "step": 11855,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.452977
    },
    {
      "epoch": 0.5081187609785356,
      "grad_norm": 3.626417398452759,
      "learning_rate": 9.747339007045789e-05,
      "loss": 2.789495086669922,
      "memory(GiB)": 58.3,
      "step": 11860,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.453052
    },
    {
      "epoch": 0.5083329763077846,
      "grad_norm": 4.912052154541016,
      "learning_rate": 9.747127740392394e-05,
      "loss": 2.5460906982421876,
      "memory(GiB)": 58.3,
      "step": 11865,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.45313
    },
    {
      "epoch": 0.5085471916370335,
      "grad_norm": 4.189931392669678,
      "learning_rate": 9.746916387740329e-05,
      "loss": 2.536775588989258,
      "memory(GiB)": 58.3,
      "step": 11870,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.45317
    },
    {
      "epoch": 0.5087614069662825,
      "grad_norm": 3.7563111782073975,
      "learning_rate": 9.74670494909342e-05,
      "loss": 2.6769477844238283,
      "memory(GiB)": 58.3,
      "step": 11875,
      "token_acc": 0.45209580838323354,
      "train_speed(iter/s)": 1.453222
    },
    {
      "epoch": 0.5089756222955315,
      "grad_norm": 3.833993673324585,
      "learning_rate": 9.746493424455498e-05,
      "loss": 2.463917350769043,
      "memory(GiB)": 58.3,
      "step": 11880,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.453369
    },
    {
      "epoch": 0.5091898376247804,
      "grad_norm": 4.479302406311035,
      "learning_rate": 9.746281813830394e-05,
      "loss": 2.7719192504882812,
      "memory(GiB)": 58.3,
      "step": 11885,
      "token_acc": 0.43283582089552236,
      "train_speed(iter/s)": 1.453378
    },
    {
      "epoch": 0.5094040529540294,
      "grad_norm": 4.144737720489502,
      "learning_rate": 9.746070117221944e-05,
      "loss": 2.903744697570801,
      "memory(GiB)": 58.3,
      "step": 11890,
      "token_acc": 0.428125,
      "train_speed(iter/s)": 1.453447
    },
    {
      "epoch": 0.5096182682832784,
      "grad_norm": 3.634902000427246,
      "learning_rate": 9.745858334633982e-05,
      "loss": 2.30502872467041,
      "memory(GiB)": 58.3,
      "step": 11895,
      "token_acc": 0.5122950819672131,
      "train_speed(iter/s)": 1.453514
    },
    {
      "epoch": 0.5098324836125273,
      "grad_norm": 4.992306709289551,
      "learning_rate": 9.745646466070342e-05,
      "loss": 2.5915502548217773,
      "memory(GiB)": 58.3,
      "step": 11900,
      "token_acc": 0.44569288389513106,
      "train_speed(iter/s)": 1.453516
    },
    {
      "epoch": 0.5100466989417762,
      "grad_norm": 4.262874603271484,
      "learning_rate": 9.745434511534867e-05,
      "loss": 2.4882678985595703,
      "memory(GiB)": 58.3,
      "step": 11905,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.453402
    },
    {
      "epoch": 0.5102609142710253,
      "grad_norm": 3.4171600341796875,
      "learning_rate": 9.745222471031392e-05,
      "loss": 2.9526262283325195,
      "memory(GiB)": 58.3,
      "step": 11910,
      "token_acc": 0.4292803970223325,
      "train_speed(iter/s)": 1.453441
    },
    {
      "epoch": 0.5104751296002742,
      "grad_norm": 4.713479518890381,
      "learning_rate": 9.745010344563763e-05,
      "loss": 2.548380661010742,
      "memory(GiB)": 58.3,
      "step": 11915,
      "token_acc": 0.4652567975830816,
      "train_speed(iter/s)": 1.453429
    },
    {
      "epoch": 0.5106893449295231,
      "grad_norm": 3.457210063934326,
      "learning_rate": 9.744798132135819e-05,
      "loss": 2.721689224243164,
      "memory(GiB)": 58.3,
      "step": 11920,
      "token_acc": 0.46779661016949153,
      "train_speed(iter/s)": 1.453146
    },
    {
      "epoch": 0.5109035602587721,
      "grad_norm": 4.835439682006836,
      "learning_rate": 9.744585833751405e-05,
      "loss": 2.7133157730102537,
      "memory(GiB)": 58.3,
      "step": 11925,
      "token_acc": 0.43416370106761565,
      "train_speed(iter/s)": 1.453285
    },
    {
      "epoch": 0.5111177755880211,
      "grad_norm": 5.564133644104004,
      "learning_rate": 9.74437344941437e-05,
      "loss": 2.6488723754882812,
      "memory(GiB)": 58.3,
      "step": 11930,
      "token_acc": 0.4509090909090909,
      "train_speed(iter/s)": 1.45303
    },
    {
      "epoch": 0.51133199091727,
      "grad_norm": 4.037095069885254,
      "learning_rate": 9.744160979128559e-05,
      "loss": 2.5098573684692385,
      "memory(GiB)": 58.3,
      "step": 11935,
      "token_acc": 0.4766355140186916,
      "train_speed(iter/s)": 1.452984
    },
    {
      "epoch": 0.511546206246519,
      "grad_norm": 4.985520362854004,
      "learning_rate": 9.743948422897819e-05,
      "loss": 2.3646467208862303,
      "memory(GiB)": 58.3,
      "step": 11940,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.45315
    },
    {
      "epoch": 0.511760421575768,
      "grad_norm": 4.7432355880737305,
      "learning_rate": 9.743735780726005e-05,
      "loss": 2.8441465377807615,
      "memory(GiB)": 58.3,
      "step": 11945,
      "token_acc": 0.4,
      "train_speed(iter/s)": 1.453117
    },
    {
      "epoch": 0.5119746369050169,
      "grad_norm": 4.232179641723633,
      "learning_rate": 9.743523052616968e-05,
      "loss": 2.597243881225586,
      "memory(GiB)": 58.3,
      "step": 11950,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.453083
    },
    {
      "epoch": 0.5121888522342659,
      "grad_norm": 3.3830976486206055,
      "learning_rate": 9.743310238574561e-05,
      "loss": 2.7058849334716797,
      "memory(GiB)": 58.3,
      "step": 11955,
      "token_acc": 0.44528301886792454,
      "train_speed(iter/s)": 1.453081
    },
    {
      "epoch": 0.5124030675635148,
      "grad_norm": 4.269236087799072,
      "learning_rate": 9.743097338602638e-05,
      "loss": 2.4047061920166017,
      "memory(GiB)": 58.3,
      "step": 11960,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.453046
    },
    {
      "epoch": 0.5126172828927638,
      "grad_norm": 4.670446395874023,
      "learning_rate": 9.742884352705057e-05,
      "loss": 2.70489387512207,
      "memory(GiB)": 58.3,
      "step": 11965,
      "token_acc": 0.4307228915662651,
      "train_speed(iter/s)": 1.452803
    },
    {
      "epoch": 0.5128314982220128,
      "grad_norm": 5.872429847717285,
      "learning_rate": 9.742671280885678e-05,
      "loss": 2.67957763671875,
      "memory(GiB)": 58.3,
      "step": 11970,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.452902
    },
    {
      "epoch": 0.5130457135512617,
      "grad_norm": 4.57229471206665,
      "learning_rate": 9.74245812314836e-05,
      "loss": 2.635705757141113,
      "memory(GiB)": 58.3,
      "step": 11975,
      "token_acc": 0.41007194244604317,
      "train_speed(iter/s)": 1.452959
    },
    {
      "epoch": 0.5132599288805106,
      "grad_norm": 6.656675815582275,
      "learning_rate": 9.742244879496964e-05,
      "loss": 2.6367509841918944,
      "memory(GiB)": 58.3,
      "step": 11980,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.453006
    },
    {
      "epoch": 0.5134741442097597,
      "grad_norm": 7.522746562957764,
      "learning_rate": 9.742031549935353e-05,
      "loss": 2.6589969635009765,
      "memory(GiB)": 58.3,
      "step": 11985,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.453102
    },
    {
      "epoch": 0.5136883595390086,
      "grad_norm": 5.553147792816162,
      "learning_rate": 9.741818134467393e-05,
      "loss": 2.8405023574829102,
      "memory(GiB)": 58.3,
      "step": 11990,
      "token_acc": 0.4222972972972973,
      "train_speed(iter/s)": 1.453165
    },
    {
      "epoch": 0.5139025748682575,
      "grad_norm": 4.80103063583374,
      "learning_rate": 9.741604633096947e-05,
      "loss": 2.891823577880859,
      "memory(GiB)": 58.3,
      "step": 11995,
      "token_acc": 0.44280442804428044,
      "train_speed(iter/s)": 1.452883
    },
    {
      "epoch": 0.5141167901975066,
      "grad_norm": 6.593349933624268,
      "learning_rate": 9.741391045827888e-05,
      "loss": 2.5027820587158205,
      "memory(GiB)": 58.3,
      "step": 12000,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.452948
    },
    {
      "epoch": 0.5141167901975066,
      "eval_loss": 2.1061935424804688,
      "eval_runtime": 13.9199,
      "eval_samples_per_second": 7.184,
      "eval_steps_per_second": 7.184,
      "eval_token_acc": 0.4846715328467153,
      "step": 12000
    },
    {
      "epoch": 0.5143310055267555,
      "grad_norm": 7.821264266967773,
      "learning_rate": 9.741177372664081e-05,
      "loss": 2.4967153549194334,
      "memory(GiB)": 58.3,
      "step": 12005,
      "token_acc": 0.48237885462555063,
      "train_speed(iter/s)": 1.450436
    },
    {
      "epoch": 0.5145452208560044,
      "grad_norm": 4.144789695739746,
      "learning_rate": 9.7409636136094e-05,
      "loss": 2.572602081298828,
      "memory(GiB)": 58.3,
      "step": 12010,
      "token_acc": 0.4794952681388013,
      "train_speed(iter/s)": 1.450444
    },
    {
      "epoch": 0.5147594361852534,
      "grad_norm": 6.287311553955078,
      "learning_rate": 9.740749768667715e-05,
      "loss": 2.457497978210449,
      "memory(GiB)": 58.3,
      "step": 12015,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.450458
    },
    {
      "epoch": 0.5149736515145024,
      "grad_norm": 4.425909042358398,
      "learning_rate": 9.740535837842899e-05,
      "loss": 2.573872947692871,
      "memory(GiB)": 58.3,
      "step": 12020,
      "token_acc": 0.4307692307692308,
      "train_speed(iter/s)": 1.450501
    },
    {
      "epoch": 0.5151878668437513,
      "grad_norm": 4.586042881011963,
      "learning_rate": 9.74032182113883e-05,
      "loss": 2.8564168930053713,
      "memory(GiB)": 58.3,
      "step": 12025,
      "token_acc": 0.42366412213740456,
      "train_speed(iter/s)": 1.450662
    },
    {
      "epoch": 0.5154020821730003,
      "grad_norm": 5.136746406555176,
      "learning_rate": 9.740107718559385e-05,
      "loss": 2.612055206298828,
      "memory(GiB)": 58.3,
      "step": 12030,
      "token_acc": 0.4673202614379085,
      "train_speed(iter/s)": 1.450771
    },
    {
      "epoch": 0.5156162975022492,
      "grad_norm": 5.197052478790283,
      "learning_rate": 9.739893530108442e-05,
      "loss": 2.603532600402832,
      "memory(GiB)": 58.3,
      "step": 12035,
      "token_acc": 0.47257383966244726,
      "train_speed(iter/s)": 1.450714
    },
    {
      "epoch": 0.5158305128314982,
      "grad_norm": 5.981902599334717,
      "learning_rate": 9.739679255789881e-05,
      "loss": 2.574320602416992,
      "memory(GiB)": 58.3,
      "step": 12040,
      "token_acc": 0.4785992217898833,
      "train_speed(iter/s)": 1.450684
    },
    {
      "epoch": 0.5160447281607472,
      "grad_norm": 3.7969653606414795,
      "learning_rate": 9.739464895607584e-05,
      "loss": 2.3138635635375975,
      "memory(GiB)": 58.3,
      "step": 12045,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.450664
    },
    {
      "epoch": 0.5162589434899961,
      "grad_norm": 3.929399013519287,
      "learning_rate": 9.739250449565436e-05,
      "loss": 3.0134450912475588,
      "memory(GiB)": 58.3,
      "step": 12050,
      "token_acc": 0.42105263157894735,
      "train_speed(iter/s)": 1.45073
    },
    {
      "epoch": 0.5164731588192452,
      "grad_norm": 3.7105753421783447,
      "learning_rate": 9.739035917667319e-05,
      "loss": 2.5796510696411135,
      "memory(GiB)": 58.3,
      "step": 12055,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.45072
    },
    {
      "epoch": 0.5166873741484941,
      "grad_norm": 5.453879356384277,
      "learning_rate": 9.738821299917121e-05,
      "loss": 2.9756711959838866,
      "memory(GiB)": 58.3,
      "step": 12060,
      "token_acc": 0.4421768707482993,
      "train_speed(iter/s)": 1.450885
    },
    {
      "epoch": 0.516901589477743,
      "grad_norm": 4.616195201873779,
      "learning_rate": 9.738606596318728e-05,
      "loss": 2.7886936187744142,
      "memory(GiB)": 58.3,
      "step": 12065,
      "token_acc": 0.44982698961937717,
      "train_speed(iter/s)": 1.450952
    },
    {
      "epoch": 0.517115804806992,
      "grad_norm": 3.2179694175720215,
      "learning_rate": 9.738391806876034e-05,
      "loss": 2.543000602722168,
      "memory(GiB)": 58.3,
      "step": 12070,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.450944
    },
    {
      "epoch": 0.517330020136241,
      "grad_norm": 3.4223556518554688,
      "learning_rate": 9.738176931592926e-05,
      "loss": 2.7501804351806642,
      "memory(GiB)": 58.3,
      "step": 12075,
      "token_acc": 0.45689655172413796,
      "train_speed(iter/s)": 1.450925
    },
    {
      "epoch": 0.5175442354654899,
      "grad_norm": 2.8262991905212402,
      "learning_rate": 9.7379619704733e-05,
      "loss": 2.565086555480957,
      "memory(GiB)": 58.3,
      "step": 12080,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.450993
    },
    {
      "epoch": 0.5177584507947389,
      "grad_norm": 4.205055236816406,
      "learning_rate": 9.737746923521046e-05,
      "loss": 2.6331724166870116,
      "memory(GiB)": 58.3,
      "step": 12085,
      "token_acc": 0.45051194539249145,
      "train_speed(iter/s)": 1.451112
    },
    {
      "epoch": 0.5179726661239878,
      "grad_norm": 4.022721290588379,
      "learning_rate": 9.737531790740062e-05,
      "loss": 2.4493757247924806,
      "memory(GiB)": 58.3,
      "step": 12090,
      "token_acc": 0.5211267605633803,
      "train_speed(iter/s)": 1.451162
    },
    {
      "epoch": 0.5181868814532368,
      "grad_norm": 12.431967735290527,
      "learning_rate": 9.737316572134247e-05,
      "loss": 2.714704132080078,
      "memory(GiB)": 58.3,
      "step": 12095,
      "token_acc": 0.43730886850152906,
      "train_speed(iter/s)": 1.451171
    },
    {
      "epoch": 0.5184010967824858,
      "grad_norm": 4.202085971832275,
      "learning_rate": 9.737101267707498e-05,
      "loss": 2.7350027084350588,
      "memory(GiB)": 58.3,
      "step": 12100,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.451307
    },
    {
      "epoch": 0.5186153121117347,
      "grad_norm": 4.404984951019287,
      "learning_rate": 9.736885877463716e-05,
      "loss": 2.9046512603759767,
      "memory(GiB)": 58.3,
      "step": 12105,
      "token_acc": 0.46096654275092935,
      "train_speed(iter/s)": 1.451393
    },
    {
      "epoch": 0.5188295274409837,
      "grad_norm": 4.3891191482543945,
      "learning_rate": 9.736670401406802e-05,
      "loss": 2.510883903503418,
      "memory(GiB)": 58.3,
      "step": 12110,
      "token_acc": 0.5211726384364821,
      "train_speed(iter/s)": 1.451349
    },
    {
      "epoch": 0.5190437427702327,
      "grad_norm": 4.217000484466553,
      "learning_rate": 9.736454839540661e-05,
      "loss": 2.7631343841552733,
      "memory(GiB)": 58.3,
      "step": 12115,
      "token_acc": 0.42394822006472493,
      "train_speed(iter/s)": 1.451441
    },
    {
      "epoch": 0.5192579580994816,
      "grad_norm": 4.029565334320068,
      "learning_rate": 9.736239191869197e-05,
      "loss": 2.4568695068359374,
      "memory(GiB)": 58.3,
      "step": 12120,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.451457
    },
    {
      "epoch": 0.5194721734287305,
      "grad_norm": 3.3806543350219727,
      "learning_rate": 9.736023458396318e-05,
      "loss": 2.556052398681641,
      "memory(GiB)": 58.3,
      "step": 12125,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.451233
    },
    {
      "epoch": 0.5196863887579796,
      "grad_norm": 4.3811564445495605,
      "learning_rate": 9.735807639125928e-05,
      "loss": 2.3946300506591798,
      "memory(GiB)": 58.3,
      "step": 12130,
      "token_acc": 0.4430769230769231,
      "train_speed(iter/s)": 1.451196
    },
    {
      "epoch": 0.5199006040872285,
      "grad_norm": 4.903264999389648,
      "learning_rate": 9.735591734061945e-05,
      "loss": 2.472573089599609,
      "memory(GiB)": 58.3,
      "step": 12135,
      "token_acc": 0.4505928853754941,
      "train_speed(iter/s)": 1.451045
    },
    {
      "epoch": 0.5201148194164774,
      "grad_norm": 4.18032169342041,
      "learning_rate": 9.735375743208273e-05,
      "loss": 2.7119796752929686,
      "memory(GiB)": 58.3,
      "step": 12140,
      "token_acc": 0.49377593360995853,
      "train_speed(iter/s)": 1.451109
    },
    {
      "epoch": 0.5203290347457264,
      "grad_norm": 4.25031042098999,
      "learning_rate": 9.735159666568826e-05,
      "loss": 2.797419548034668,
      "memory(GiB)": 58.3,
      "step": 12145,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.451134
    },
    {
      "epoch": 0.5205432500749754,
      "grad_norm": 3.4540727138519287,
      "learning_rate": 9.734943504147521e-05,
      "loss": 2.6920007705688476,
      "memory(GiB)": 58.3,
      "step": 12150,
      "token_acc": 0.42033898305084744,
      "train_speed(iter/s)": 1.451134
    },
    {
      "epoch": 0.5207574654042243,
      "grad_norm": 3.4848079681396484,
      "learning_rate": 9.734727255948273e-05,
      "loss": 2.753078269958496,
      "memory(GiB)": 58.3,
      "step": 12155,
      "token_acc": 0.4416961130742049,
      "train_speed(iter/s)": 1.450984
    },
    {
      "epoch": 0.5209716807334733,
      "grad_norm": 3.585627555847168,
      "learning_rate": 9.734510921974998e-05,
      "loss": 2.6603424072265627,
      "memory(GiB)": 58.3,
      "step": 12160,
      "token_acc": 0.46774193548387094,
      "train_speed(iter/s)": 1.450981
    },
    {
      "epoch": 0.5211858960627223,
      "grad_norm": 4.278395175933838,
      "learning_rate": 9.734294502231618e-05,
      "loss": 2.345458984375,
      "memory(GiB)": 58.3,
      "step": 12165,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.45096
    },
    {
      "epoch": 0.5214001113919712,
      "grad_norm": 3.882774829864502,
      "learning_rate": 9.73407799672205e-05,
      "loss": 2.5837369918823243,
      "memory(GiB)": 58.3,
      "step": 12170,
      "token_acc": 0.4485049833887043,
      "train_speed(iter/s)": 1.450963
    },
    {
      "epoch": 0.5216143267212202,
      "grad_norm": 6.81789493560791,
      "learning_rate": 9.73386140545022e-05,
      "loss": 2.9007122039794924,
      "memory(GiB)": 58.3,
      "step": 12175,
      "token_acc": 0.3978102189781022,
      "train_speed(iter/s)": 1.450844
    },
    {
      "epoch": 0.5218285420504691,
      "grad_norm": 3.43770432472229,
      "learning_rate": 9.73364472842005e-05,
      "loss": 2.546364974975586,
      "memory(GiB)": 58.3,
      "step": 12180,
      "token_acc": 0.5113636363636364,
      "train_speed(iter/s)": 1.450865
    },
    {
      "epoch": 0.5220427573797181,
      "grad_norm": 4.896144866943359,
      "learning_rate": 9.733427965635465e-05,
      "loss": 2.6740402221679687,
      "memory(GiB)": 58.3,
      "step": 12185,
      "token_acc": 0.4507936507936508,
      "train_speed(iter/s)": 1.450979
    },
    {
      "epoch": 0.5222569727089671,
      "grad_norm": 5.150900363922119,
      "learning_rate": 9.73321111710039e-05,
      "loss": 2.641109085083008,
      "memory(GiB)": 58.3,
      "step": 12190,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.45109
    },
    {
      "epoch": 0.522471188038216,
      "grad_norm": 4.9929046630859375,
      "learning_rate": 9.732994182818757e-05,
      "loss": 2.557425880432129,
      "memory(GiB)": 58.3,
      "step": 12195,
      "token_acc": 0.4370860927152318,
      "train_speed(iter/s)": 1.451104
    },
    {
      "epoch": 0.5226854033674649,
      "grad_norm": 3.6203742027282715,
      "learning_rate": 9.732777162794496e-05,
      "loss": 2.441172790527344,
      "memory(GiB)": 58.3,
      "step": 12200,
      "token_acc": 0.43174603174603177,
      "train_speed(iter/s)": 1.451158
    },
    {
      "epoch": 0.522899618696714,
      "grad_norm": 4.222485065460205,
      "learning_rate": 9.732560057031534e-05,
      "loss": 2.535055923461914,
      "memory(GiB)": 58.3,
      "step": 12205,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.451187
    },
    {
      "epoch": 0.5231138340259629,
      "grad_norm": 5.140800952911377,
      "learning_rate": 9.732342865533809e-05,
      "loss": 2.519024658203125,
      "memory(GiB)": 58.3,
      "step": 12210,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.451206
    },
    {
      "epoch": 0.5233280493552118,
      "grad_norm": 7.724201202392578,
      "learning_rate": 9.732125588305252e-05,
      "loss": 2.722677993774414,
      "memory(GiB)": 58.3,
      "step": 12215,
      "token_acc": 0.429042904290429,
      "train_speed(iter/s)": 1.451296
    },
    {
      "epoch": 0.5235422646844609,
      "grad_norm": 4.023663520812988,
      "learning_rate": 9.731908225349803e-05,
      "loss": 2.4608808517456056,
      "memory(GiB)": 58.3,
      "step": 12220,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.451089
    },
    {
      "epoch": 0.5237564800137098,
      "grad_norm": 3.1600120067596436,
      "learning_rate": 9.731690776671396e-05,
      "loss": 2.505036544799805,
      "memory(GiB)": 58.3,
      "step": 12225,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.451101
    },
    {
      "epoch": 0.5239706953429587,
      "grad_norm": 6.346956729888916,
      "learning_rate": 9.731473242273974e-05,
      "loss": 2.5708852767944337,
      "memory(GiB)": 58.3,
      "step": 12230,
      "token_acc": 0.4618181818181818,
      "train_speed(iter/s)": 1.451158
    },
    {
      "epoch": 0.5241849106722077,
      "grad_norm": 3.5241777896881104,
      "learning_rate": 9.731255622161474e-05,
      "loss": 2.58681640625,
      "memory(GiB)": 58.3,
      "step": 12235,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.451249
    },
    {
      "epoch": 0.5243991260014567,
      "grad_norm": 4.737888336181641,
      "learning_rate": 9.731037916337839e-05,
      "loss": 2.253646659851074,
      "memory(GiB)": 58.3,
      "step": 12240,
      "token_acc": 0.5341365461847389,
      "train_speed(iter/s)": 1.451189
    },
    {
      "epoch": 0.5246133413307056,
      "grad_norm": 3.948641300201416,
      "learning_rate": 9.730820124807017e-05,
      "loss": 2.4099130630493164,
      "memory(GiB)": 58.3,
      "step": 12245,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.451249
    },
    {
      "epoch": 0.5248275566599546,
      "grad_norm": 4.104923725128174,
      "learning_rate": 9.730602247572948e-05,
      "loss": 2.530244827270508,
      "memory(GiB)": 58.3,
      "step": 12250,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.451226
    },
    {
      "epoch": 0.5250417719892035,
      "grad_norm": 4.411331653594971,
      "learning_rate": 9.730384284639584e-05,
      "loss": 2.8195140838623045,
      "memory(GiB)": 58.3,
      "step": 12255,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.451425
    },
    {
      "epoch": 0.5252559873184525,
      "grad_norm": 4.188760757446289,
      "learning_rate": 9.730166236010869e-05,
      "loss": 2.518187141418457,
      "memory(GiB)": 58.3,
      "step": 12260,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.451508
    },
    {
      "epoch": 0.5254702026477015,
      "grad_norm": 5.70860481262207,
      "learning_rate": 9.729948101690757e-05,
      "loss": 2.4894786834716798,
      "memory(GiB)": 58.3,
      "step": 12265,
      "token_acc": 0.5203252032520326,
      "train_speed(iter/s)": 1.451656
    },
    {
      "epoch": 0.5256844179769504,
      "grad_norm": 4.290028095245361,
      "learning_rate": 9.729729881683197e-05,
      "loss": 2.852960395812988,
      "memory(GiB)": 58.3,
      "step": 12270,
      "token_acc": 0.4372623574144487,
      "train_speed(iter/s)": 1.451747
    },
    {
      "epoch": 0.5258986333061993,
      "grad_norm": 5.143388271331787,
      "learning_rate": 9.729511575992145e-05,
      "loss": 2.5350303649902344,
      "memory(GiB)": 58.3,
      "step": 12275,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.451699
    },
    {
      "epoch": 0.5261128486354484,
      "grad_norm": 4.118575096130371,
      "learning_rate": 9.729293184621552e-05,
      "loss": 2.620782661437988,
      "memory(GiB)": 58.3,
      "step": 12280,
      "token_acc": 0.4716417910447761,
      "train_speed(iter/s)": 1.451748
    },
    {
      "epoch": 0.5263270639646973,
      "grad_norm": 3.6123764514923096,
      "learning_rate": 9.729074707575377e-05,
      "loss": 2.578450012207031,
      "memory(GiB)": 58.3,
      "step": 12285,
      "token_acc": 0.4676923076923077,
      "train_speed(iter/s)": 1.451596
    },
    {
      "epoch": 0.5265412792939462,
      "grad_norm": 5.424049377441406,
      "learning_rate": 9.728856144857579e-05,
      "loss": 2.8362285614013674,
      "memory(GiB)": 58.3,
      "step": 12290,
      "token_acc": 0.42574257425742573,
      "train_speed(iter/s)": 1.451664
    },
    {
      "epoch": 0.5267554946231953,
      "grad_norm": 4.256664276123047,
      "learning_rate": 9.728637496472115e-05,
      "loss": 2.4898361206054687,
      "memory(GiB)": 58.3,
      "step": 12295,
      "token_acc": 0.45255474452554745,
      "train_speed(iter/s)": 1.451589
    },
    {
      "epoch": 0.5269697099524442,
      "grad_norm": 6.542172431945801,
      "learning_rate": 9.728418762422947e-05,
      "loss": 2.6374752044677736,
      "memory(GiB)": 58.3,
      "step": 12300,
      "token_acc": 0.3870967741935484,
      "train_speed(iter/s)": 1.451678
    },
    {
      "epoch": 0.5271839252816931,
      "grad_norm": 3.748230218887329,
      "learning_rate": 9.728199942714036e-05,
      "loss": 2.5267093658447264,
      "memory(GiB)": 58.3,
      "step": 12305,
      "token_acc": 0.5050167224080268,
      "train_speed(iter/s)": 1.451577
    },
    {
      "epoch": 0.5273981406109421,
      "grad_norm": 3.6318066120147705,
      "learning_rate": 9.727981037349349e-05,
      "loss": 2.425985336303711,
      "memory(GiB)": 58.3,
      "step": 12310,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.451635
    },
    {
      "epoch": 0.5276123559401911,
      "grad_norm": 6.197627067565918,
      "learning_rate": 9.727762046332851e-05,
      "loss": 2.6292070388793944,
      "memory(GiB)": 58.3,
      "step": 12315,
      "token_acc": 0.4686192468619247,
      "train_speed(iter/s)": 1.451699
    },
    {
      "epoch": 0.52782657126944,
      "grad_norm": 4.748229026794434,
      "learning_rate": 9.727542969668507e-05,
      "loss": 2.525190734863281,
      "memory(GiB)": 58.3,
      "step": 12320,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.451718
    },
    {
      "epoch": 0.528040786598689,
      "grad_norm": 4.905789852142334,
      "learning_rate": 9.72732380736029e-05,
      "loss": 2.280447769165039,
      "memory(GiB)": 58.3,
      "step": 12325,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.451725
    },
    {
      "epoch": 0.528255001927938,
      "grad_norm": 5.149145603179932,
      "learning_rate": 9.727104559412163e-05,
      "loss": 2.813428497314453,
      "memory(GiB)": 58.3,
      "step": 12330,
      "token_acc": 0.4246575342465753,
      "train_speed(iter/s)": 1.45169
    },
    {
      "epoch": 0.5284692172571869,
      "grad_norm": 4.573469161987305,
      "learning_rate": 9.726885225828106e-05,
      "loss": 2.6495691299438477,
      "memory(GiB)": 58.3,
      "step": 12335,
      "token_acc": 0.44402985074626866,
      "train_speed(iter/s)": 1.451755
    },
    {
      "epoch": 0.5286834325864359,
      "grad_norm": 4.365417957305908,
      "learning_rate": 9.726665806612087e-05,
      "loss": 2.465804100036621,
      "memory(GiB)": 58.3,
      "step": 12340,
      "token_acc": 0.5125448028673835,
      "train_speed(iter/s)": 1.451658
    },
    {
      "epoch": 0.5288976479156848,
      "grad_norm": 3.845370292663574,
      "learning_rate": 9.726446301768085e-05,
      "loss": 2.7484079360961915,
      "memory(GiB)": 58.3,
      "step": 12345,
      "token_acc": 0.46273291925465837,
      "train_speed(iter/s)": 1.451639
    },
    {
      "epoch": 0.5291118632449338,
      "grad_norm": 3.590235471725464,
      "learning_rate": 9.726226711300074e-05,
      "loss": 2.6285877227783203,
      "memory(GiB)": 58.3,
      "step": 12350,
      "token_acc": 0.40584415584415584,
      "train_speed(iter/s)": 1.451685
    },
    {
      "epoch": 0.5293260785741828,
      "grad_norm": 4.054399013519287,
      "learning_rate": 9.726007035212033e-05,
      "loss": 2.486698532104492,
      "memory(GiB)": 58.3,
      "step": 12355,
      "token_acc": 0.44694533762057875,
      "train_speed(iter/s)": 1.451579
    },
    {
      "epoch": 0.5295402939034317,
      "grad_norm": 5.202297687530518,
      "learning_rate": 9.725787273507939e-05,
      "loss": 2.4875463485717773,
      "memory(GiB)": 58.3,
      "step": 12360,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.451589
    },
    {
      "epoch": 0.5297545092326806,
      "grad_norm": 4.829860210418701,
      "learning_rate": 9.725567426191777e-05,
      "loss": 2.2026813507080076,
      "memory(GiB)": 58.3,
      "step": 12365,
      "token_acc": 0.5458015267175572,
      "train_speed(iter/s)": 1.45161
    },
    {
      "epoch": 0.5299687245619297,
      "grad_norm": 6.145134449005127,
      "learning_rate": 9.725347493267529e-05,
      "loss": 2.4365734100341796,
      "memory(GiB)": 58.3,
      "step": 12370,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.451589
    },
    {
      "epoch": 0.5301829398911786,
      "grad_norm": 5.194461822509766,
      "learning_rate": 9.725127474739176e-05,
      "loss": 2.5591482162475585,
      "memory(GiB)": 58.3,
      "step": 12375,
      "token_acc": 0.41379310344827586,
      "train_speed(iter/s)": 1.451754
    },
    {
      "epoch": 0.5303971552204275,
      "grad_norm": 3.0579028129577637,
      "learning_rate": 9.724907370610707e-05,
      "loss": 2.559821891784668,
      "memory(GiB)": 58.3,
      "step": 12380,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.451796
    },
    {
      "epoch": 0.5306113705496766,
      "grad_norm": 4.266158580780029,
      "learning_rate": 9.724687180886109e-05,
      "loss": 2.6127849578857423,
      "memory(GiB)": 58.3,
      "step": 12385,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.451756
    },
    {
      "epoch": 0.5308255858789255,
      "grad_norm": 5.282445430755615,
      "learning_rate": 9.724466905569372e-05,
      "loss": 2.6352052688598633,
      "memory(GiB)": 58.3,
      "step": 12390,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.451722
    },
    {
      "epoch": 0.5310398012081745,
      "grad_norm": 4.208840370178223,
      "learning_rate": 9.724246544664484e-05,
      "loss": 2.904363250732422,
      "memory(GiB)": 58.3,
      "step": 12395,
      "token_acc": 0.41471571906354515,
      "train_speed(iter/s)": 1.451688
    },
    {
      "epoch": 0.5312540165374234,
      "grad_norm": 3.660950183868408,
      "learning_rate": 9.724026098175438e-05,
      "loss": 2.935175323486328,
      "memory(GiB)": 58.3,
      "step": 12400,
      "token_acc": 0.4363143631436314,
      "train_speed(iter/s)": 1.451635
    },
    {
      "epoch": 0.5314682318666724,
      "grad_norm": 5.2977519035339355,
      "learning_rate": 9.723805566106227e-05,
      "loss": 2.7877475738525392,
      "memory(GiB)": 58.3,
      "step": 12405,
      "token_acc": 0.44649446494464945,
      "train_speed(iter/s)": 1.451639
    },
    {
      "epoch": 0.5316824471959214,
      "grad_norm": 4.768824577331543,
      "learning_rate": 9.723584948460848e-05,
      "loss": 2.896093559265137,
      "memory(GiB)": 58.3,
      "step": 12410,
      "token_acc": 0.42424242424242425,
      "train_speed(iter/s)": 1.451483
    },
    {
      "epoch": 0.5318966625251703,
      "grad_norm": 5.282357215881348,
      "learning_rate": 9.723364245243296e-05,
      "loss": 2.7335073471069338,
      "memory(GiB)": 58.3,
      "step": 12415,
      "token_acc": 0.4423076923076923,
      "train_speed(iter/s)": 1.451515
    },
    {
      "epoch": 0.5321108778544192,
      "grad_norm": 6.384718894958496,
      "learning_rate": 9.723143456457571e-05,
      "loss": 2.750216484069824,
      "memory(GiB)": 58.3,
      "step": 12420,
      "token_acc": 0.4296028880866426,
      "train_speed(iter/s)": 1.451517
    },
    {
      "epoch": 0.5323250931836683,
      "grad_norm": 5.642190933227539,
      "learning_rate": 9.72292258210767e-05,
      "loss": 2.60189208984375,
      "memory(GiB)": 58.3,
      "step": 12425,
      "token_acc": 0.5043103448275862,
      "train_speed(iter/s)": 1.451657
    },
    {
      "epoch": 0.5325393085129172,
      "grad_norm": 5.021167755126953,
      "learning_rate": 9.722701622197596e-05,
      "loss": 2.67369384765625,
      "memory(GiB)": 58.3,
      "step": 12430,
      "token_acc": 0.4664429530201342,
      "train_speed(iter/s)": 1.451786
    },
    {
      "epoch": 0.5327535238421661,
      "grad_norm": 5.418324947357178,
      "learning_rate": 9.722480576731354e-05,
      "loss": 2.748818206787109,
      "memory(GiB)": 58.3,
      "step": 12435,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.45174
    },
    {
      "epoch": 0.5329677391714152,
      "grad_norm": 4.710017681121826,
      "learning_rate": 9.722259445712945e-05,
      "loss": 2.6717979431152346,
      "memory(GiB)": 58.3,
      "step": 12440,
      "token_acc": 0.45660377358490567,
      "train_speed(iter/s)": 1.451499
    },
    {
      "epoch": 0.5331819545006641,
      "grad_norm": 4.8615031242370605,
      "learning_rate": 9.722038229146375e-05,
      "loss": 2.6751766204833984,
      "memory(GiB)": 58.3,
      "step": 12445,
      "token_acc": 0.4383116883116883,
      "train_speed(iter/s)": 1.451323
    },
    {
      "epoch": 0.533396169829913,
      "grad_norm": 4.771420001983643,
      "learning_rate": 9.721816927035654e-05,
      "loss": 2.6626399993896483,
      "memory(GiB)": 58.3,
      "step": 12450,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.451212
    },
    {
      "epoch": 0.533610385159162,
      "grad_norm": 5.13985538482666,
      "learning_rate": 9.721595539384791e-05,
      "loss": 2.873168182373047,
      "memory(GiB)": 58.3,
      "step": 12455,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.45132
    },
    {
      "epoch": 0.533824600488411,
      "grad_norm": 3.6609761714935303,
      "learning_rate": 9.721374066197795e-05,
      "loss": 2.7462268829345704,
      "memory(GiB)": 58.3,
      "step": 12460,
      "token_acc": 0.4612903225806452,
      "train_speed(iter/s)": 1.451409
    },
    {
      "epoch": 0.5340388158176599,
      "grad_norm": 6.595832824707031,
      "learning_rate": 9.721152507478677e-05,
      "loss": 2.399384117126465,
      "memory(GiB)": 58.3,
      "step": 12465,
      "token_acc": 0.4855072463768116,
      "train_speed(iter/s)": 1.451458
    },
    {
      "epoch": 0.5342530311469089,
      "grad_norm": 3.449709415435791,
      "learning_rate": 9.720930863231454e-05,
      "loss": 2.1814918518066406,
      "memory(GiB)": 58.3,
      "step": 12470,
      "token_acc": 0.5165562913907285,
      "train_speed(iter/s)": 1.451342
    },
    {
      "epoch": 0.5344672464761578,
      "grad_norm": 4.599881172180176,
      "learning_rate": 9.72070913346014e-05,
      "loss": 2.8916391372680663,
      "memory(GiB)": 58.3,
      "step": 12475,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.451297
    },
    {
      "epoch": 0.5346814618054068,
      "grad_norm": 4.440439224243164,
      "learning_rate": 9.720487318168752e-05,
      "loss": 2.7131805419921875,
      "memory(GiB)": 58.3,
      "step": 12480,
      "token_acc": 0.4690909090909091,
      "train_speed(iter/s)": 1.451327
    },
    {
      "epoch": 0.5348956771346558,
      "grad_norm": 4.202776908874512,
      "learning_rate": 9.720265417361306e-05,
      "loss": 2.696521759033203,
      "memory(GiB)": 58.3,
      "step": 12485,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.451387
    },
    {
      "epoch": 0.5351098924639047,
      "grad_norm": 4.543173789978027,
      "learning_rate": 9.720043431041825e-05,
      "loss": 2.8303363800048826,
      "memory(GiB)": 58.3,
      "step": 12490,
      "token_acc": 0.42955326460481097,
      "train_speed(iter/s)": 1.451482
    },
    {
      "epoch": 0.5353241077931536,
      "grad_norm": 3.8472940921783447,
      "learning_rate": 9.719821359214331e-05,
      "loss": 2.638447952270508,
      "memory(GiB)": 58.3,
      "step": 12495,
      "token_acc": 0.4908424908424908,
      "train_speed(iter/s)": 1.451592
    },
    {
      "epoch": 0.5355383231224027,
      "grad_norm": 6.116992950439453,
      "learning_rate": 9.719599201882845e-05,
      "loss": 2.5670450210571287,
      "memory(GiB)": 58.3,
      "step": 12500,
      "token_acc": 0.4927007299270073,
      "train_speed(iter/s)": 1.45172
    },
    {
      "epoch": 0.5355383231224027,
      "eval_loss": 1.9528933763504028,
      "eval_runtime": 13.6132,
      "eval_samples_per_second": 7.346,
      "eval_steps_per_second": 7.346,
      "eval_token_acc": 0.5,
      "step": 12500
    },
    {
      "epoch": 0.5357525384516516,
      "grad_norm": 4.405342102050781,
      "learning_rate": 9.71937695905139e-05,
      "loss": 2.6192043304443358,
      "memory(GiB)": 58.3,
      "step": 12505,
      "token_acc": 0.4769392033542977,
      "train_speed(iter/s)": 1.449136
    },
    {
      "epoch": 0.5359667537809005,
      "grad_norm": 4.714395046234131,
      "learning_rate": 9.719154630723996e-05,
      "loss": 2.841592025756836,
      "memory(GiB)": 58.3,
      "step": 12510,
      "token_acc": 0.4608433734939759,
      "train_speed(iter/s)": 1.449103
    },
    {
      "epoch": 0.5361809691101496,
      "grad_norm": 4.565947532653809,
      "learning_rate": 9.718932216904689e-05,
      "loss": 2.6948455810546874,
      "memory(GiB)": 58.3,
      "step": 12515,
      "token_acc": 0.42248062015503873,
      "train_speed(iter/s)": 1.449295
    },
    {
      "epoch": 0.5363951844393985,
      "grad_norm": 5.4101996421813965,
      "learning_rate": 9.718709717597498e-05,
      "loss": 2.7623340606689455,
      "memory(GiB)": 58.3,
      "step": 12520,
      "token_acc": 0.42045454545454547,
      "train_speed(iter/s)": 1.449142
    },
    {
      "epoch": 0.5366093997686474,
      "grad_norm": 3.920884132385254,
      "learning_rate": 9.718487132806453e-05,
      "loss": 2.7843671798706056,
      "memory(GiB)": 58.3,
      "step": 12525,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.449283
    },
    {
      "epoch": 0.5368236150978964,
      "grad_norm": 3.6120080947875977,
      "learning_rate": 9.71826446253559e-05,
      "loss": 2.5558847427368163,
      "memory(GiB)": 58.3,
      "step": 12530,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.449465
    },
    {
      "epoch": 0.5370378304271454,
      "grad_norm": 4.816959857940674,
      "learning_rate": 9.718041706788937e-05,
      "loss": 2.7386369705200195,
      "memory(GiB)": 58.3,
      "step": 12535,
      "token_acc": 0.47701149425287354,
      "train_speed(iter/s)": 1.449543
    },
    {
      "epoch": 0.5372520457563943,
      "grad_norm": 3.709352731704712,
      "learning_rate": 9.717818865570534e-05,
      "loss": 2.7484809875488283,
      "memory(GiB)": 58.3,
      "step": 12540,
      "token_acc": 0.4481707317073171,
      "train_speed(iter/s)": 1.449612
    },
    {
      "epoch": 0.5374662610856433,
      "grad_norm": 3.888885736465454,
      "learning_rate": 9.717595938884416e-05,
      "loss": 2.809458541870117,
      "memory(GiB)": 58.3,
      "step": 12545,
      "token_acc": 0.43843843843843844,
      "train_speed(iter/s)": 1.449496
    },
    {
      "epoch": 0.5376804764148922,
      "grad_norm": 4.670475482940674,
      "learning_rate": 9.717372926734623e-05,
      "loss": 3.0961544036865236,
      "memory(GiB)": 58.3,
      "step": 12550,
      "token_acc": 0.37942122186495175,
      "train_speed(iter/s)": 1.449608
    },
    {
      "epoch": 0.5378946917441412,
      "grad_norm": 4.811488628387451,
      "learning_rate": 9.717149829125194e-05,
      "loss": 2.8523914337158205,
      "memory(GiB)": 58.3,
      "step": 12555,
      "token_acc": 0.44802867383512546,
      "train_speed(iter/s)": 1.449638
    },
    {
      "epoch": 0.5381089070733902,
      "grad_norm": 5.6197428703308105,
      "learning_rate": 9.71692664606017e-05,
      "loss": 2.5427404403686524,
      "memory(GiB)": 58.3,
      "step": 12560,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.449733
    },
    {
      "epoch": 0.5383231224026391,
      "grad_norm": 5.28272819519043,
      "learning_rate": 9.716703377543595e-05,
      "loss": 2.851909637451172,
      "memory(GiB)": 58.3,
      "step": 12565,
      "token_acc": 0.4454828660436137,
      "train_speed(iter/s)": 1.449798
    },
    {
      "epoch": 0.538537337731888,
      "grad_norm": 4.3687591552734375,
      "learning_rate": 9.716480023579516e-05,
      "loss": 2.531698226928711,
      "memory(GiB)": 58.3,
      "step": 12570,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.449789
    },
    {
      "epoch": 0.5387515530611371,
      "grad_norm": 3.5607926845550537,
      "learning_rate": 9.716256584171974e-05,
      "loss": 2.599342918395996,
      "memory(GiB)": 58.3,
      "step": 12575,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.449691
    },
    {
      "epoch": 0.538965768390386,
      "grad_norm": 4.8093085289001465,
      "learning_rate": 9.71603305932502e-05,
      "loss": 2.6707332611083983,
      "memory(GiB)": 58.3,
      "step": 12580,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.449786
    },
    {
      "epoch": 0.5391799837196349,
      "grad_norm": 3.1623497009277344,
      "learning_rate": 9.715809449042706e-05,
      "loss": 2.695049285888672,
      "memory(GiB)": 58.3,
      "step": 12585,
      "token_acc": 0.44744744744744747,
      "train_speed(iter/s)": 1.449648
    },
    {
      "epoch": 0.539394199048884,
      "grad_norm": 4.399327278137207,
      "learning_rate": 9.715585753329077e-05,
      "loss": 2.579286575317383,
      "memory(GiB)": 58.3,
      "step": 12590,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.449724
    },
    {
      "epoch": 0.5396084143781329,
      "grad_norm": 5.561838150024414,
      "learning_rate": 9.715361972188191e-05,
      "loss": 2.3928447723388673,
      "memory(GiB)": 58.3,
      "step": 12595,
      "token_acc": 0.48828125,
      "train_speed(iter/s)": 1.44968
    },
    {
      "epoch": 0.5398226297073818,
      "grad_norm": 4.708993434906006,
      "learning_rate": 9.715138105624097e-05,
      "loss": 2.908864212036133,
      "memory(GiB)": 58.3,
      "step": 12600,
      "token_acc": 0.4104938271604938,
      "train_speed(iter/s)": 1.449764
    },
    {
      "epoch": 0.5400368450366309,
      "grad_norm": 6.453427314758301,
      "learning_rate": 9.714914153640854e-05,
      "loss": 2.405851364135742,
      "memory(GiB)": 58.3,
      "step": 12605,
      "token_acc": 0.4942084942084942,
      "train_speed(iter/s)": 1.449807
    },
    {
      "epoch": 0.5402510603658798,
      "grad_norm": 7.729811191558838,
      "learning_rate": 9.714690116242518e-05,
      "loss": 2.809157943725586,
      "memory(GiB)": 58.3,
      "step": 12610,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.449851
    },
    {
      "epoch": 0.5404652756951287,
      "grad_norm": 3.0981762409210205,
      "learning_rate": 9.714465993433148e-05,
      "loss": 3.0204233169555663,
      "memory(GiB)": 58.3,
      "step": 12615,
      "token_acc": 0.44542772861356933,
      "train_speed(iter/s)": 1.449977
    },
    {
      "epoch": 0.5406794910243777,
      "grad_norm": 4.510212421417236,
      "learning_rate": 9.714241785216804e-05,
      "loss": 2.4773700714111326,
      "memory(GiB)": 58.3,
      "step": 12620,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.450007
    },
    {
      "epoch": 0.5408937063536267,
      "grad_norm": 3.8259317874908447,
      "learning_rate": 9.714017491597548e-05,
      "loss": 2.7129066467285154,
      "memory(GiB)": 58.3,
      "step": 12625,
      "token_acc": 0.4621212121212121,
      "train_speed(iter/s)": 1.450023
    },
    {
      "epoch": 0.5411079216828756,
      "grad_norm": 4.733958721160889,
      "learning_rate": 9.713793112579443e-05,
      "loss": 2.3526430130004883,
      "memory(GiB)": 58.3,
      "step": 12630,
      "token_acc": 0.5100401606425703,
      "train_speed(iter/s)": 1.45012
    },
    {
      "epoch": 0.5413221370121246,
      "grad_norm": 4.693044185638428,
      "learning_rate": 9.713568648166555e-05,
      "loss": 2.585987091064453,
      "memory(GiB)": 58.3,
      "step": 12635,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.450247
    },
    {
      "epoch": 0.5415363523413735,
      "grad_norm": 5.047013759613037,
      "learning_rate": 9.713344098362948e-05,
      "loss": 3.068511962890625,
      "memory(GiB)": 58.3,
      "step": 12640,
      "token_acc": 0.40977443609022557,
      "train_speed(iter/s)": 1.450129
    },
    {
      "epoch": 0.5417505676706225,
      "grad_norm": 3.8072783946990967,
      "learning_rate": 9.71311946317269e-05,
      "loss": 2.135857582092285,
      "memory(GiB)": 58.3,
      "step": 12645,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.450077
    },
    {
      "epoch": 0.5419647829998715,
      "grad_norm": 3.554363489151001,
      "learning_rate": 9.712894742599854e-05,
      "loss": 2.5715707778930663,
      "memory(GiB)": 58.3,
      "step": 12650,
      "token_acc": 0.46254071661237783,
      "train_speed(iter/s)": 1.450074
    },
    {
      "epoch": 0.5421789983291204,
      "grad_norm": 3.843611240386963,
      "learning_rate": 9.712669936648507e-05,
      "loss": 2.515215301513672,
      "memory(GiB)": 58.3,
      "step": 12655,
      "token_acc": 0.4559386973180077,
      "train_speed(iter/s)": 1.45004
    },
    {
      "epoch": 0.5423932136583693,
      "grad_norm": 5.072500228881836,
      "learning_rate": 9.712445045322725e-05,
      "loss": 2.6251152038574217,
      "memory(GiB)": 58.3,
      "step": 12660,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.450226
    },
    {
      "epoch": 0.5426074289876184,
      "grad_norm": 5.165890693664551,
      "learning_rate": 9.712220068626578e-05,
      "loss": 2.5669994354248047,
      "memory(GiB)": 58.3,
      "step": 12665,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.450245
    },
    {
      "epoch": 0.5428216443168673,
      "grad_norm": 7.052614688873291,
      "learning_rate": 9.711995006564145e-05,
      "loss": 2.443429946899414,
      "memory(GiB)": 58.3,
      "step": 12670,
      "token_acc": 0.4979919678714859,
      "train_speed(iter/s)": 1.450332
    },
    {
      "epoch": 0.5430358596461162,
      "grad_norm": 4.854347229003906,
      "learning_rate": 9.711769859139503e-05,
      "loss": 2.286446952819824,
      "memory(GiB)": 58.3,
      "step": 12675,
      "token_acc": 0.5258620689655172,
      "train_speed(iter/s)": 1.450368
    },
    {
      "epoch": 0.5432500749753653,
      "grad_norm": 4.551116466522217,
      "learning_rate": 9.71154462635673e-05,
      "loss": 2.7156274795532225,
      "memory(GiB)": 58.3,
      "step": 12680,
      "token_acc": 0.4751552795031056,
      "train_speed(iter/s)": 1.450338
    },
    {
      "epoch": 0.5434642903046142,
      "grad_norm": 4.808742046356201,
      "learning_rate": 9.711319308219906e-05,
      "loss": 2.6372392654418944,
      "memory(GiB)": 58.3,
      "step": 12685,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.450455
    },
    {
      "epoch": 0.5436785056338631,
      "grad_norm": 3.753126859664917,
      "learning_rate": 9.711093904733113e-05,
      "loss": 2.5791189193725588,
      "memory(GiB)": 58.3,
      "step": 12690,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.450517
    },
    {
      "epoch": 0.5438927209631121,
      "grad_norm": 3.754671573638916,
      "learning_rate": 9.710868415900433e-05,
      "loss": 2.4302879333496095,
      "memory(GiB)": 58.3,
      "step": 12695,
      "token_acc": 0.45955882352941174,
      "train_speed(iter/s)": 1.450481
    },
    {
      "epoch": 0.5441069362923611,
      "grad_norm": 4.710944175720215,
      "learning_rate": 9.710642841725954e-05,
      "loss": 2.624073791503906,
      "memory(GiB)": 58.3,
      "step": 12700,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.450594
    },
    {
      "epoch": 0.54432115162161,
      "grad_norm": 6.307151794433594,
      "learning_rate": 9.710417182213762e-05,
      "loss": 2.7054271697998047,
      "memory(GiB)": 58.3,
      "step": 12705,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.450679
    },
    {
      "epoch": 0.544535366950859,
      "grad_norm": 7.1200361251831055,
      "learning_rate": 9.710191437367942e-05,
      "loss": 2.6211063385009767,
      "memory(GiB)": 58.3,
      "step": 12710,
      "token_acc": 0.4323308270676692,
      "train_speed(iter/s)": 1.450773
    },
    {
      "epoch": 0.544749582280108,
      "grad_norm": 4.841012001037598,
      "learning_rate": 9.709965607192589e-05,
      "loss": 2.5262451171875,
      "memory(GiB)": 58.3,
      "step": 12715,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.450709
    },
    {
      "epoch": 0.5449637976093569,
      "grad_norm": 3.5466911792755127,
      "learning_rate": 9.709739691691787e-05,
      "loss": 2.6461902618408204,
      "memory(GiB)": 58.3,
      "step": 12720,
      "token_acc": 0.4659400544959128,
      "train_speed(iter/s)": 1.450779
    },
    {
      "epoch": 0.5451780129386059,
      "grad_norm": 4.010507106781006,
      "learning_rate": 9.709513690869634e-05,
      "loss": 2.596065139770508,
      "memory(GiB)": 58.3,
      "step": 12725,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.450676
    },
    {
      "epoch": 0.5453922282678548,
      "grad_norm": 3.871396780014038,
      "learning_rate": 9.709287604730222e-05,
      "loss": 2.5589183807373046,
      "memory(GiB)": 58.3,
      "step": 12730,
      "token_acc": 0.4380165289256198,
      "train_speed(iter/s)": 1.450752
    },
    {
      "epoch": 0.5456064435971039,
      "grad_norm": 4.161299705505371,
      "learning_rate": 9.709061433277647e-05,
      "loss": 2.575978088378906,
      "memory(GiB)": 58.3,
      "step": 12735,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.450644
    },
    {
      "epoch": 0.5458206589263528,
      "grad_norm": 5.430095672607422,
      "learning_rate": 9.708835176516008e-05,
      "loss": 2.4091758728027344,
      "memory(GiB)": 58.3,
      "step": 12740,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.450437
    },
    {
      "epoch": 0.5460348742556017,
      "grad_norm": 5.0184245109558105,
      "learning_rate": 9.708608834449402e-05,
      "loss": 2.7799776077270506,
      "memory(GiB)": 58.3,
      "step": 12745,
      "token_acc": 0.4637096774193548,
      "train_speed(iter/s)": 1.450624
    },
    {
      "epoch": 0.5462490895848507,
      "grad_norm": 7.726380825042725,
      "learning_rate": 9.708382407081929e-05,
      "loss": 2.339492988586426,
      "memory(GiB)": 58.3,
      "step": 12750,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.450706
    },
    {
      "epoch": 0.5464633049140997,
      "grad_norm": 5.161066055297852,
      "learning_rate": 9.708155894417693e-05,
      "loss": 2.641633224487305,
      "memory(GiB)": 58.3,
      "step": 12755,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.450709
    },
    {
      "epoch": 0.5466775202433486,
      "grad_norm": 3.77683424949646,
      "learning_rate": 9.707929296460796e-05,
      "loss": 2.5671504974365233,
      "memory(GiB)": 58.3,
      "step": 12760,
      "token_acc": 0.47953216374269003,
      "train_speed(iter/s)": 1.450634
    },
    {
      "epoch": 0.5468917355725976,
      "grad_norm": 5.168262958526611,
      "learning_rate": 9.707702613215344e-05,
      "loss": 2.7779285430908205,
      "memory(GiB)": 58.3,
      "step": 12765,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.450815
    },
    {
      "epoch": 0.5471059509018465,
      "grad_norm": 4.115829944610596,
      "learning_rate": 9.707475844685443e-05,
      "loss": 2.7260311126708983,
      "memory(GiB)": 58.3,
      "step": 12770,
      "token_acc": 0.46981627296587924,
      "train_speed(iter/s)": 1.450769
    },
    {
      "epoch": 0.5473201662310955,
      "grad_norm": 4.663558006286621,
      "learning_rate": 9.7072489908752e-05,
      "loss": 2.793963623046875,
      "memory(GiB)": 58.3,
      "step": 12775,
      "token_acc": 0.4245614035087719,
      "train_speed(iter/s)": 1.450838
    },
    {
      "epoch": 0.5475343815603445,
      "grad_norm": 3.709885835647583,
      "learning_rate": 9.707022051788725e-05,
      "loss": 2.8072214126586914,
      "memory(GiB)": 58.3,
      "step": 12780,
      "token_acc": 0.4431818181818182,
      "train_speed(iter/s)": 1.450803
    },
    {
      "epoch": 0.5477485968895934,
      "grad_norm": 4.756540298461914,
      "learning_rate": 9.706795027430133e-05,
      "loss": 2.7352317810058593,
      "memory(GiB)": 58.3,
      "step": 12785,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.450834
    },
    {
      "epoch": 0.5479628122188424,
      "grad_norm": 4.949060916900635,
      "learning_rate": 9.706567917803532e-05,
      "loss": 2.598540115356445,
      "memory(GiB)": 58.3,
      "step": 12790,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.450694
    },
    {
      "epoch": 0.5481770275480914,
      "grad_norm": 4.608774185180664,
      "learning_rate": 9.706340722913037e-05,
      "loss": 2.7334430694580076,
      "memory(GiB)": 58.3,
      "step": 12795,
      "token_acc": 0.48923076923076925,
      "train_speed(iter/s)": 1.450803
    },
    {
      "epoch": 0.5483912428773403,
      "grad_norm": 3.5583767890930176,
      "learning_rate": 9.706113442762768e-05,
      "loss": 2.512015533447266,
      "memory(GiB)": 58.3,
      "step": 12800,
      "token_acc": 0.5236363636363637,
      "train_speed(iter/s)": 1.450884
    },
    {
      "epoch": 0.5486054582065892,
      "grad_norm": 5.349429130554199,
      "learning_rate": 9.705886077356837e-05,
      "loss": 2.61452693939209,
      "memory(GiB)": 58.3,
      "step": 12805,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.451024
    },
    {
      "epoch": 0.5488196735358383,
      "grad_norm": 3.242171287536621,
      "learning_rate": 9.705658626699365e-05,
      "loss": 2.4409656524658203,
      "memory(GiB)": 58.3,
      "step": 12810,
      "token_acc": 0.4884393063583815,
      "train_speed(iter/s)": 1.450948
    },
    {
      "epoch": 0.5490338888650872,
      "grad_norm": 3.8514227867126465,
      "learning_rate": 9.705431090794472e-05,
      "loss": 2.807894706726074,
      "memory(GiB)": 58.3,
      "step": 12815,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.451088
    },
    {
      "epoch": 0.5492481041943361,
      "grad_norm": 4.336859703063965,
      "learning_rate": 9.705203469646282e-05,
      "loss": 2.4014379501342775,
      "memory(GiB)": 58.3,
      "step": 12820,
      "token_acc": 0.5188284518828452,
      "train_speed(iter/s)": 1.451135
    },
    {
      "epoch": 0.5494623195235852,
      "grad_norm": 5.311072826385498,
      "learning_rate": 9.704975763258919e-05,
      "loss": 2.6091053009033205,
      "memory(GiB)": 58.3,
      "step": 12825,
      "token_acc": 0.45188284518828453,
      "train_speed(iter/s)": 1.451138
    },
    {
      "epoch": 0.5496765348528341,
      "grad_norm": 5.685634613037109,
      "learning_rate": 9.704747971636504e-05,
      "loss": 2.8710981369018556,
      "memory(GiB)": 58.3,
      "step": 12830,
      "token_acc": 0.45318352059925093,
      "train_speed(iter/s)": 1.451154
    },
    {
      "epoch": 0.549890750182083,
      "grad_norm": 4.341067790985107,
      "learning_rate": 9.704520094783167e-05,
      "loss": 2.6993837356567383,
      "memory(GiB)": 58.3,
      "step": 12835,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.451161
    },
    {
      "epoch": 0.550104965511332,
      "grad_norm": 4.665333271026611,
      "learning_rate": 9.704292132703034e-05,
      "loss": 2.406839370727539,
      "memory(GiB)": 58.3,
      "step": 12840,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.45117
    },
    {
      "epoch": 0.550319180840581,
      "grad_norm": 4.147494316101074,
      "learning_rate": 9.704064085400237e-05,
      "loss": 2.5593090057373047,
      "memory(GiB)": 58.3,
      "step": 12845,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.451185
    },
    {
      "epoch": 0.5505333961698299,
      "grad_norm": 8.908293724060059,
      "learning_rate": 9.703835952878908e-05,
      "loss": 2.672134208679199,
      "memory(GiB)": 58.3,
      "step": 12850,
      "token_acc": 0.45110410094637227,
      "train_speed(iter/s)": 1.451197
    },
    {
      "epoch": 0.5507476114990789,
      "grad_norm": 5.46074104309082,
      "learning_rate": 9.703607735143176e-05,
      "loss": 2.5653865814208983,
      "memory(GiB)": 58.3,
      "step": 12855,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.451356
    },
    {
      "epoch": 0.5509618268283278,
      "grad_norm": 5.539285659790039,
      "learning_rate": 9.703379432197178e-05,
      "loss": 2.5961772918701174,
      "memory(GiB)": 58.3,
      "step": 12860,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.45141
    },
    {
      "epoch": 0.5511760421575768,
      "grad_norm": 3.4348926544189453,
      "learning_rate": 9.70315104404505e-05,
      "loss": 2.52852783203125,
      "memory(GiB)": 58.3,
      "step": 12865,
      "token_acc": 0.453125,
      "train_speed(iter/s)": 1.451285
    },
    {
      "epoch": 0.5513902574868258,
      "grad_norm": 4.793821811676025,
      "learning_rate": 9.70292257069093e-05,
      "loss": 2.3236888885498046,
      "memory(GiB)": 58.3,
      "step": 12870,
      "token_acc": 0.5065359477124183,
      "train_speed(iter/s)": 1.451377
    },
    {
      "epoch": 0.5516044728160747,
      "grad_norm": 4.095990180969238,
      "learning_rate": 9.702694012138953e-05,
      "loss": 2.354779815673828,
      "memory(GiB)": 58.3,
      "step": 12875,
      "token_acc": 0.43450479233226835,
      "train_speed(iter/s)": 1.451371
    },
    {
      "epoch": 0.5518186881453236,
      "grad_norm": 3.842207193374634,
      "learning_rate": 9.702465368393264e-05,
      "loss": 2.506814956665039,
      "memory(GiB)": 58.3,
      "step": 12880,
      "token_acc": 0.48757763975155277,
      "train_speed(iter/s)": 1.451259
    },
    {
      "epoch": 0.5520329034745727,
      "grad_norm": 8.617725372314453,
      "learning_rate": 9.702236639458003e-05,
      "loss": 2.399854850769043,
      "memory(GiB)": 58.3,
      "step": 12885,
      "token_acc": 0.434640522875817,
      "train_speed(iter/s)": 1.45107
    },
    {
      "epoch": 0.5522471188038216,
      "grad_norm": 4.077536582946777,
      "learning_rate": 9.702007825337316e-05,
      "loss": 2.788869094848633,
      "memory(GiB)": 58.3,
      "step": 12890,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.45096
    },
    {
      "epoch": 0.5524613341330705,
      "grad_norm": 4.308690547943115,
      "learning_rate": 9.701778926035344e-05,
      "loss": 2.6484283447265624,
      "memory(GiB)": 58.3,
      "step": 12895,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.45103
    },
    {
      "epoch": 0.5526755494623196,
      "grad_norm": 4.967340469360352,
      "learning_rate": 9.701549941556238e-05,
      "loss": 2.627827453613281,
      "memory(GiB)": 58.3,
      "step": 12900,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.450966
    },
    {
      "epoch": 0.5528897647915685,
      "grad_norm": 3.6724588871002197,
      "learning_rate": 9.701320871904143e-05,
      "loss": 2.8314285278320312,
      "memory(GiB)": 58.3,
      "step": 12905,
      "token_acc": 0.444141689373297,
      "train_speed(iter/s)": 1.451098
    },
    {
      "epoch": 0.5531039801208174,
      "grad_norm": 4.921066761016846,
      "learning_rate": 9.70109171708321e-05,
      "loss": 2.488862228393555,
      "memory(GiB)": 58.3,
      "step": 12910,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.451267
    },
    {
      "epoch": 0.5533181954500664,
      "grad_norm": 3.338116407394409,
      "learning_rate": 9.700862477097592e-05,
      "loss": 2.708690643310547,
      "memory(GiB)": 58.3,
      "step": 12915,
      "token_acc": 0.44931506849315067,
      "train_speed(iter/s)": 1.451401
    },
    {
      "epoch": 0.5535324107793154,
      "grad_norm": 5.2589874267578125,
      "learning_rate": 9.700633151951441e-05,
      "loss": 2.536569595336914,
      "memory(GiB)": 58.3,
      "step": 12920,
      "token_acc": 0.5054151624548736,
      "train_speed(iter/s)": 1.451434
    },
    {
      "epoch": 0.5537466261085643,
      "grad_norm": 4.2836785316467285,
      "learning_rate": 9.700403741648909e-05,
      "loss": 2.6208091735839845,
      "memory(GiB)": 58.3,
      "step": 12925,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.451421
    },
    {
      "epoch": 0.5539608414378133,
      "grad_norm": 4.430039882659912,
      "learning_rate": 9.700174246194155e-05,
      "loss": 2.7207714080810548,
      "memory(GiB)": 58.3,
      "step": 12930,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.451375
    },
    {
      "epoch": 0.5541750567670622,
      "grad_norm": 4.568837642669678,
      "learning_rate": 9.699944665591336e-05,
      "loss": 2.876577377319336,
      "memory(GiB)": 58.3,
      "step": 12935,
      "token_acc": 0.4377358490566038,
      "train_speed(iter/s)": 1.45135
    },
    {
      "epoch": 0.5543892720963112,
      "grad_norm": 4.635406494140625,
      "learning_rate": 9.699714999844608e-05,
      "loss": 2.5632776260375976,
      "memory(GiB)": 58.3,
      "step": 12940,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.451482
    },
    {
      "epoch": 0.5546034874255602,
      "grad_norm": 3.9506866931915283,
      "learning_rate": 9.699485248958137e-05,
      "loss": 3.1370044708251954,
      "memory(GiB)": 58.3,
      "step": 12945,
      "token_acc": 0.40524781341107874,
      "train_speed(iter/s)": 1.451616
    },
    {
      "epoch": 0.5548177027548091,
      "grad_norm": 4.8641133308410645,
      "learning_rate": 9.69925541293608e-05,
      "loss": 2.539725494384766,
      "memory(GiB)": 58.3,
      "step": 12950,
      "token_acc": 0.512,
      "train_speed(iter/s)": 1.451677
    },
    {
      "epoch": 0.555031918084058,
      "grad_norm": 4.505123138427734,
      "learning_rate": 9.699025491782606e-05,
      "loss": 2.677878570556641,
      "memory(GiB)": 58.3,
      "step": 12955,
      "token_acc": 0.43795620437956206,
      "train_speed(iter/s)": 1.451786
    },
    {
      "epoch": 0.5552461334133071,
      "grad_norm": 9.601716995239258,
      "learning_rate": 9.698795485501873e-05,
      "loss": 2.6862897872924805,
      "memory(GiB)": 58.3,
      "step": 12960,
      "token_acc": 0.43700787401574803,
      "train_speed(iter/s)": 1.451697
    },
    {
      "epoch": 0.555460348742556,
      "grad_norm": 4.412141799926758,
      "learning_rate": 9.698565394098054e-05,
      "loss": 2.8015703201293944,
      "memory(GiB)": 58.3,
      "step": 12965,
      "token_acc": 0.41637010676156583,
      "train_speed(iter/s)": 1.451778
    },
    {
      "epoch": 0.5556745640718049,
      "grad_norm": 4.217938423156738,
      "learning_rate": 9.698335217575316e-05,
      "loss": 2.674403762817383,
      "memory(GiB)": 58.3,
      "step": 12970,
      "token_acc": 0.44696969696969696,
      "train_speed(iter/s)": 1.451799
    },
    {
      "epoch": 0.555888779401054,
      "grad_norm": 4.214227676391602,
      "learning_rate": 9.698104955937827e-05,
      "loss": 2.4468994140625,
      "memory(GiB)": 58.3,
      "step": 12975,
      "token_acc": 0.4435146443514644,
      "train_speed(iter/s)": 1.451741
    },
    {
      "epoch": 0.5561029947303029,
      "grad_norm": 4.063323497772217,
      "learning_rate": 9.697874609189759e-05,
      "loss": 2.7058795928955077,
      "memory(GiB)": 58.3,
      "step": 12980,
      "token_acc": 0.4377104377104377,
      "train_speed(iter/s)": 1.451781
    },
    {
      "epoch": 0.5563172100595518,
      "grad_norm": 5.039035797119141,
      "learning_rate": 9.697644177335287e-05,
      "loss": 2.4908977508544923,
      "memory(GiB)": 58.3,
      "step": 12985,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.45187
    },
    {
      "epoch": 0.5565314253888008,
      "grad_norm": 4.5800604820251465,
      "learning_rate": 9.697413660378584e-05,
      "loss": 2.4564802169799806,
      "memory(GiB)": 58.3,
      "step": 12990,
      "token_acc": 0.4573643410852713,
      "train_speed(iter/s)": 1.451969
    },
    {
      "epoch": 0.5567456407180498,
      "grad_norm": 5.657577991485596,
      "learning_rate": 9.697183058323826e-05,
      "loss": 2.9000558853149414,
      "memory(GiB)": 58.3,
      "step": 12995,
      "token_acc": 0.43617021276595747,
      "train_speed(iter/s)": 1.451962
    },
    {
      "epoch": 0.5569598560472987,
      "grad_norm": 4.3039655685424805,
      "learning_rate": 9.69695237117519e-05,
      "loss": 2.6337804794311523,
      "memory(GiB)": 58.3,
      "step": 13000,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.451979
    },
    {
      "epoch": 0.5569598560472987,
      "eval_loss": 2.31534481048584,
      "eval_runtime": 15.1459,
      "eval_samples_per_second": 6.602,
      "eval_steps_per_second": 6.602,
      "eval_token_acc": 0.45041322314049587,
      "step": 13000
    },
    {
      "epoch": 0.5571740713765477,
      "grad_norm": 4.509006977081299,
      "learning_rate": 9.696721598936856e-05,
      "loss": 2.721030426025391,
      "memory(GiB)": 58.3,
      "step": 13005,
      "token_acc": 0.4441233140655106,
      "train_speed(iter/s)": 1.449492
    },
    {
      "epoch": 0.5573882867057967,
      "grad_norm": 6.866149425506592,
      "learning_rate": 9.696490741613002e-05,
      "loss": 2.7085390090942383,
      "memory(GiB)": 58.3,
      "step": 13010,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.449568
    },
    {
      "epoch": 0.5576025020350456,
      "grad_norm": 3.926135778427124,
      "learning_rate": 9.696259799207816e-05,
      "loss": 2.2735612869262694,
      "memory(GiB)": 58.3,
      "step": 13015,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.449593
    },
    {
      "epoch": 0.5578167173642946,
      "grad_norm": 4.030943870544434,
      "learning_rate": 9.696028771725476e-05,
      "loss": 2.5962993621826174,
      "memory(GiB)": 58.3,
      "step": 13020,
      "token_acc": 0.4501510574018127,
      "train_speed(iter/s)": 1.449596
    },
    {
      "epoch": 0.5580309326935435,
      "grad_norm": 4.363567352294922,
      "learning_rate": 9.695797659170171e-05,
      "loss": 2.469046974182129,
      "memory(GiB)": 58.3,
      "step": 13025,
      "token_acc": 0.45429362880886426,
      "train_speed(iter/s)": 1.449647
    },
    {
      "epoch": 0.5582451480227925,
      "grad_norm": 3.859976291656494,
      "learning_rate": 9.695566461546086e-05,
      "loss": 2.4266117095947264,
      "memory(GiB)": 58.3,
      "step": 13030,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.449571
    },
    {
      "epoch": 0.5584593633520415,
      "grad_norm": 3.6077330112457275,
      "learning_rate": 9.695335178857409e-05,
      "loss": 2.3070953369140623,
      "memory(GiB)": 58.3,
      "step": 13035,
      "token_acc": 0.5175097276264592,
      "train_speed(iter/s)": 1.449596
    },
    {
      "epoch": 0.5586735786812904,
      "grad_norm": 3.4801628589630127,
      "learning_rate": 9.695103811108333e-05,
      "loss": 2.861098861694336,
      "memory(GiB)": 58.3,
      "step": 13040,
      "token_acc": 0.41754385964912283,
      "train_speed(iter/s)": 1.449626
    },
    {
      "epoch": 0.5588877940105393,
      "grad_norm": 4.068077564239502,
      "learning_rate": 9.694872358303044e-05,
      "loss": 2.5247350692749024,
      "memory(GiB)": 58.3,
      "step": 13045,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.449642
    },
    {
      "epoch": 0.5591020093397884,
      "grad_norm": 7.038202285766602,
      "learning_rate": 9.69464082044574e-05,
      "loss": 2.351951789855957,
      "memory(GiB)": 58.3,
      "step": 13050,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.449544
    },
    {
      "epoch": 0.5593162246690373,
      "grad_norm": 2.840205669403076,
      "learning_rate": 9.694409197540613e-05,
      "loss": 2.796143341064453,
      "memory(GiB)": 58.3,
      "step": 13055,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.449556
    },
    {
      "epoch": 0.5595304399982862,
      "grad_norm": 4.22896671295166,
      "learning_rate": 9.69417748959186e-05,
      "loss": 2.82263298034668,
      "memory(GiB)": 58.3,
      "step": 13060,
      "token_acc": 0.42258064516129035,
      "train_speed(iter/s)": 1.449657
    },
    {
      "epoch": 0.5597446553275353,
      "grad_norm": 4.6137003898620605,
      "learning_rate": 9.693945696603678e-05,
      "loss": 2.875554084777832,
      "memory(GiB)": 58.3,
      "step": 13065,
      "token_acc": 0.4317343173431734,
      "train_speed(iter/s)": 1.449754
    },
    {
      "epoch": 0.5599588706567842,
      "grad_norm": 4.3805694580078125,
      "learning_rate": 9.693713818580265e-05,
      "loss": 2.7376270294189453,
      "memory(GiB)": 58.3,
      "step": 13070,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.449686
    },
    {
      "epoch": 0.5601730859860332,
      "grad_norm": 3.852978467941284,
      "learning_rate": 9.693481855525826e-05,
      "loss": 2.638378143310547,
      "memory(GiB)": 58.3,
      "step": 13075,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.449615
    },
    {
      "epoch": 0.5603873013152821,
      "grad_norm": 4.143217086791992,
      "learning_rate": 9.693249807444559e-05,
      "loss": 2.5436470031738283,
      "memory(GiB)": 58.3,
      "step": 13080,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.449628
    },
    {
      "epoch": 0.5606015166445311,
      "grad_norm": 4.595394134521484,
      "learning_rate": 9.693017674340669e-05,
      "loss": 2.449018096923828,
      "memory(GiB)": 58.3,
      "step": 13085,
      "token_acc": 0.4981684981684982,
      "train_speed(iter/s)": 1.449735
    },
    {
      "epoch": 0.5608157319737801,
      "grad_norm": 5.33073616027832,
      "learning_rate": 9.69278545621836e-05,
      "loss": 2.654311752319336,
      "memory(GiB)": 58.3,
      "step": 13090,
      "token_acc": 0.45149253731343286,
      "train_speed(iter/s)": 1.449899
    },
    {
      "epoch": 0.561029947303029,
      "grad_norm": 5.2792558670043945,
      "learning_rate": 9.692553153081842e-05,
      "loss": 2.7055185317993162,
      "memory(GiB)": 58.3,
      "step": 13095,
      "token_acc": 0.46613545816733065,
      "train_speed(iter/s)": 1.449902
    },
    {
      "epoch": 0.5612441626322779,
      "grad_norm": 6.160458087921143,
      "learning_rate": 9.692320764935322e-05,
      "loss": 2.9085187911987305,
      "memory(GiB)": 58.3,
      "step": 13100,
      "token_acc": 0.376425855513308,
      "train_speed(iter/s)": 1.44995
    },
    {
      "epoch": 0.561458377961527,
      "grad_norm": 4.582027435302734,
      "learning_rate": 9.692088291783009e-05,
      "loss": 2.349048614501953,
      "memory(GiB)": 58.3,
      "step": 13105,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.450039
    },
    {
      "epoch": 0.5616725932907759,
      "grad_norm": 4.76365327835083,
      "learning_rate": 9.691855733629115e-05,
      "loss": 2.264979934692383,
      "memory(GiB)": 58.3,
      "step": 13110,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.450036
    },
    {
      "epoch": 0.5618868086200248,
      "grad_norm": 6.014882564544678,
      "learning_rate": 9.691623090477852e-05,
      "loss": 2.1806407928466798,
      "memory(GiB)": 58.3,
      "step": 13115,
      "token_acc": 0.5083333333333333,
      "train_speed(iter/s)": 1.450114
    },
    {
      "epoch": 0.5621010239492739,
      "grad_norm": 4.662960529327393,
      "learning_rate": 9.691390362333437e-05,
      "loss": 2.566688346862793,
      "memory(GiB)": 58.3,
      "step": 13120,
      "token_acc": 0.47289156626506024,
      "train_speed(iter/s)": 1.450212
    },
    {
      "epoch": 0.5623152392785228,
      "grad_norm": 4.14929723739624,
      "learning_rate": 9.691157549200085e-05,
      "loss": 3.0323604583740233,
      "memory(GiB)": 58.3,
      "step": 13125,
      "token_acc": 0.4,
      "train_speed(iter/s)": 1.450343
    },
    {
      "epoch": 0.5625294546077717,
      "grad_norm": 3.6472291946411133,
      "learning_rate": 9.690924651082014e-05,
      "loss": 2.609294891357422,
      "memory(GiB)": 58.3,
      "step": 13130,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.450296
    },
    {
      "epoch": 0.5627436699370207,
      "grad_norm": 6.209442615509033,
      "learning_rate": 9.69069166798344e-05,
      "loss": 2.5425687789916993,
      "memory(GiB)": 58.3,
      "step": 13135,
      "token_acc": 0.4628099173553719,
      "train_speed(iter/s)": 1.450398
    },
    {
      "epoch": 0.5629578852662697,
      "grad_norm": 4.023408889770508,
      "learning_rate": 9.690458599908588e-05,
      "loss": 2.6907154083251954,
      "memory(GiB)": 58.3,
      "step": 13140,
      "token_acc": 0.40784313725490196,
      "train_speed(iter/s)": 1.450402
    },
    {
      "epoch": 0.5631721005955186,
      "grad_norm": 3.9270732402801514,
      "learning_rate": 9.690225446861678e-05,
      "loss": 2.7321781158447265,
      "memory(GiB)": 58.3,
      "step": 13145,
      "token_acc": 0.45323741007194246,
      "train_speed(iter/s)": 1.450516
    },
    {
      "epoch": 0.5633863159247676,
      "grad_norm": 5.218646049499512,
      "learning_rate": 9.689992208846934e-05,
      "loss": 2.55145206451416,
      "memory(GiB)": 58.3,
      "step": 13150,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.450647
    },
    {
      "epoch": 0.5636005312540165,
      "grad_norm": 4.007495880126953,
      "learning_rate": 9.689758885868582e-05,
      "loss": 2.489964485168457,
      "memory(GiB)": 58.3,
      "step": 13155,
      "token_acc": 0.47794117647058826,
      "train_speed(iter/s)": 1.450663
    },
    {
      "epoch": 0.5638147465832655,
      "grad_norm": 3.621326208114624,
      "learning_rate": 9.689525477930848e-05,
      "loss": 2.556059646606445,
      "memory(GiB)": 58.3,
      "step": 13160,
      "token_acc": 0.4597014925373134,
      "train_speed(iter/s)": 1.450761
    },
    {
      "epoch": 0.5640289619125145,
      "grad_norm": 4.4627838134765625,
      "learning_rate": 9.68929198503796e-05,
      "loss": 2.153653144836426,
      "memory(GiB)": 58.3,
      "step": 13165,
      "token_acc": 0.5205992509363296,
      "train_speed(iter/s)": 1.45079
    },
    {
      "epoch": 0.5642431772417634,
      "grad_norm": 4.90310525894165,
      "learning_rate": 9.689058407194152e-05,
      "loss": 2.8451507568359373,
      "memory(GiB)": 58.3,
      "step": 13170,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.450798
    },
    {
      "epoch": 0.5644573925710124,
      "grad_norm": 4.354957580566406,
      "learning_rate": 9.688824744403648e-05,
      "loss": 3.0005460739135743,
      "memory(GiB)": 58.3,
      "step": 13175,
      "token_acc": 0.39501779359430605,
      "train_speed(iter/s)": 1.450831
    },
    {
      "epoch": 0.5646716079002614,
      "grad_norm": 3.2764077186584473,
      "learning_rate": 9.688590996670688e-05,
      "loss": 2.5787830352783203,
      "memory(GiB)": 58.3,
      "step": 13180,
      "token_acc": 0.44765342960288806,
      "train_speed(iter/s)": 1.45096
    },
    {
      "epoch": 0.5648858232295103,
      "grad_norm": 4.579800128936768,
      "learning_rate": 9.688357163999504e-05,
      "loss": 2.542692184448242,
      "memory(GiB)": 58.3,
      "step": 13185,
      "token_acc": 0.4756554307116105,
      "train_speed(iter/s)": 1.451041
    },
    {
      "epoch": 0.5651000385587592,
      "grad_norm": 4.687442779541016,
      "learning_rate": 9.68812324639433e-05,
      "loss": 2.341318893432617,
      "memory(GiB)": 58.3,
      "step": 13190,
      "token_acc": 0.47808764940239046,
      "train_speed(iter/s)": 1.451113
    },
    {
      "epoch": 0.5653142538880083,
      "grad_norm": 4.083542823791504,
      "learning_rate": 9.687889243859407e-05,
      "loss": 2.4852067947387697,
      "memory(GiB)": 58.3,
      "step": 13195,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.451116
    },
    {
      "epoch": 0.5655284692172572,
      "grad_norm": 5.367619037628174,
      "learning_rate": 9.687655156398972e-05,
      "loss": 2.3516740798950195,
      "memory(GiB)": 58.3,
      "step": 13200,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.451069
    },
    {
      "epoch": 0.5657426845465061,
      "grad_norm": 5.573348522186279,
      "learning_rate": 9.687420984017268e-05,
      "loss": 2.6972415924072264,
      "memory(GiB)": 58.3,
      "step": 13205,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.451185
    },
    {
      "epoch": 0.5659568998757551,
      "grad_norm": 4.3836774826049805,
      "learning_rate": 9.687186726718533e-05,
      "loss": 2.289139747619629,
      "memory(GiB)": 58.3,
      "step": 13210,
      "token_acc": 0.5296610169491526,
      "train_speed(iter/s)": 1.451183
    },
    {
      "epoch": 0.5661711152050041,
      "grad_norm": 4.545618057250977,
      "learning_rate": 9.686952384507017e-05,
      "loss": 3.006224822998047,
      "memory(GiB)": 58.3,
      "step": 13215,
      "token_acc": 0.43666666666666665,
      "train_speed(iter/s)": 1.451122
    },
    {
      "epoch": 0.566385330534253,
      "grad_norm": 6.145777225494385,
      "learning_rate": 9.686717957386959e-05,
      "loss": 2.8068992614746096,
      "memory(GiB)": 58.3,
      "step": 13220,
      "token_acc": 0.45394736842105265,
      "train_speed(iter/s)": 1.451137
    },
    {
      "epoch": 0.566599545863502,
      "grad_norm": 4.864924430847168,
      "learning_rate": 9.686483445362611e-05,
      "loss": 2.664941596984863,
      "memory(GiB)": 58.3,
      "step": 13225,
      "token_acc": 0.4318181818181818,
      "train_speed(iter/s)": 1.45113
    },
    {
      "epoch": 0.566813761192751,
      "grad_norm": 4.378489971160889,
      "learning_rate": 9.686248848438217e-05,
      "loss": 2.6865907669067384,
      "memory(GiB)": 58.3,
      "step": 13230,
      "token_acc": 0.4778481012658228,
      "train_speed(iter/s)": 1.451299
    },
    {
      "epoch": 0.5670279765219999,
      "grad_norm": 3.1974430084228516,
      "learning_rate": 9.68601416661803e-05,
      "loss": 2.5594970703125,
      "memory(GiB)": 58.3,
      "step": 13235,
      "token_acc": 0.4489795918367347,
      "train_speed(iter/s)": 1.451306
    },
    {
      "epoch": 0.5672421918512489,
      "grad_norm": 3.9626917839050293,
      "learning_rate": 9.6857793999063e-05,
      "loss": 2.546051025390625,
      "memory(GiB)": 58.3,
      "step": 13240,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.451345
    },
    {
      "epoch": 0.5674564071804978,
      "grad_norm": 4.272452354431152,
      "learning_rate": 9.685544548307281e-05,
      "loss": 2.6064426422119142,
      "memory(GiB)": 58.3,
      "step": 13245,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.451342
    },
    {
      "epoch": 0.5676706225097468,
      "grad_norm": 8.32752513885498,
      "learning_rate": 9.685309611825226e-05,
      "loss": 2.732959175109863,
      "memory(GiB)": 58.3,
      "step": 13250,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.451461
    },
    {
      "epoch": 0.5678848378389958,
      "grad_norm": 4.122401714324951,
      "learning_rate": 9.685074590464394e-05,
      "loss": 2.6859151840209963,
      "memory(GiB)": 58.3,
      "step": 13255,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.451412
    },
    {
      "epoch": 0.5680990531682447,
      "grad_norm": 3.7030677795410156,
      "learning_rate": 9.68483948422904e-05,
      "loss": 2.4719486236572266,
      "memory(GiB)": 58.3,
      "step": 13260,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.451446
    },
    {
      "epoch": 0.5683132684974936,
      "grad_norm": 4.952679634094238,
      "learning_rate": 9.684604293123425e-05,
      "loss": 2.449727249145508,
      "memory(GiB)": 58.3,
      "step": 13265,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.451449
    },
    {
      "epoch": 0.5685274838267427,
      "grad_norm": 4.573519229888916,
      "learning_rate": 9.684369017151806e-05,
      "loss": 2.691756820678711,
      "memory(GiB)": 58.3,
      "step": 13270,
      "token_acc": 0.45387453874538747,
      "train_speed(iter/s)": 1.451499
    },
    {
      "epoch": 0.5687416991559916,
      "grad_norm": 3.881869077682495,
      "learning_rate": 9.68413365631845e-05,
      "loss": 2.715593719482422,
      "memory(GiB)": 58.3,
      "step": 13275,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.451538
    },
    {
      "epoch": 0.5689559144852405,
      "grad_norm": 3.503706216812134,
      "learning_rate": 9.68389821062762e-05,
      "loss": 2.540977668762207,
      "memory(GiB)": 58.3,
      "step": 13280,
      "token_acc": 0.45878136200716846,
      "train_speed(iter/s)": 1.451594
    },
    {
      "epoch": 0.5691701298144896,
      "grad_norm": 5.725363254547119,
      "learning_rate": 9.683662680083578e-05,
      "loss": 2.4994224548339843,
      "memory(GiB)": 58.3,
      "step": 13285,
      "token_acc": 0.49603174603174605,
      "train_speed(iter/s)": 1.451763
    },
    {
      "epoch": 0.5693843451437385,
      "grad_norm": 4.425648212432861,
      "learning_rate": 9.683427064690593e-05,
      "loss": 2.2694299697875975,
      "memory(GiB)": 58.3,
      "step": 13290,
      "token_acc": 0.5179856115107914,
      "train_speed(iter/s)": 1.451885
    },
    {
      "epoch": 0.5695985604729874,
      "grad_norm": 3.9148318767547607,
      "learning_rate": 9.683191364452934e-05,
      "loss": 2.630241584777832,
      "memory(GiB)": 58.3,
      "step": 13295,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.452041
    },
    {
      "epoch": 0.5698127758022364,
      "grad_norm": 4.354046821594238,
      "learning_rate": 9.68295557937487e-05,
      "loss": 2.615975570678711,
      "memory(GiB)": 58.3,
      "step": 13300,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.452075
    },
    {
      "epoch": 0.5700269911314854,
      "grad_norm": 3.723904609680176,
      "learning_rate": 9.682719709460672e-05,
      "loss": 2.825900650024414,
      "memory(GiB)": 58.3,
      "step": 13305,
      "token_acc": 0.4507042253521127,
      "train_speed(iter/s)": 1.452179
    },
    {
      "epoch": 0.5702412064607343,
      "grad_norm": 3.485006809234619,
      "learning_rate": 9.682483754714615e-05,
      "loss": 2.4839103698730467,
      "memory(GiB)": 58.3,
      "step": 13310,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.452157
    },
    {
      "epoch": 0.5704554217899833,
      "grad_norm": 5.618500232696533,
      "learning_rate": 9.682247715140974e-05,
      "loss": 2.64285945892334,
      "memory(GiB)": 58.3,
      "step": 13315,
      "token_acc": 0.49224806201550386,
      "train_speed(iter/s)": 1.452165
    },
    {
      "epoch": 0.5706696371192322,
      "grad_norm": 4.187808036804199,
      "learning_rate": 9.68201159074402e-05,
      "loss": 2.7868972778320313,
      "memory(GiB)": 58.3,
      "step": 13320,
      "token_acc": 0.4222873900293255,
      "train_speed(iter/s)": 1.452189
    },
    {
      "epoch": 0.5708838524484812,
      "grad_norm": 5.1764397621154785,
      "learning_rate": 9.681775381528034e-05,
      "loss": 2.9550350189208983,
      "memory(GiB)": 58.3,
      "step": 13325,
      "token_acc": 0.4054878048780488,
      "train_speed(iter/s)": 1.452265
    },
    {
      "epoch": 0.5710980677777302,
      "grad_norm": 4.35557746887207,
      "learning_rate": 9.681539087497296e-05,
      "loss": 2.9351924896240233,
      "memory(GiB)": 58.3,
      "step": 13330,
      "token_acc": 0.3776978417266187,
      "train_speed(iter/s)": 1.45238
    },
    {
      "epoch": 0.5713122831069791,
      "grad_norm": 3.86228346824646,
      "learning_rate": 9.681302708656086e-05,
      "loss": 2.2315444946289062,
      "memory(GiB)": 58.3,
      "step": 13335,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.452455
    },
    {
      "epoch": 0.571526498436228,
      "grad_norm": 5.441580772399902,
      "learning_rate": 9.681066245008687e-05,
      "loss": 2.4947357177734375,
      "memory(GiB)": 58.3,
      "step": 13340,
      "token_acc": 0.4745098039215686,
      "train_speed(iter/s)": 1.452475
    },
    {
      "epoch": 0.5717407137654771,
      "grad_norm": 4.5027618408203125,
      "learning_rate": 9.68082969655938e-05,
      "loss": 2.5578094482421876,
      "memory(GiB)": 58.3,
      "step": 13345,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.452481
    },
    {
      "epoch": 0.571954929094726,
      "grad_norm": 5.213467597961426,
      "learning_rate": 9.680593063312454e-05,
      "loss": 2.4859502792358397,
      "memory(GiB)": 58.3,
      "step": 13350,
      "token_acc": 0.43853820598006643,
      "train_speed(iter/s)": 1.452356
    },
    {
      "epoch": 0.5721691444239749,
      "grad_norm": 4.5291242599487305,
      "learning_rate": 9.680356345272193e-05,
      "loss": 2.4200429916381836,
      "memory(GiB)": 58.3,
      "step": 13355,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.452537
    },
    {
      "epoch": 0.572383359753224,
      "grad_norm": 5.182436943054199,
      "learning_rate": 9.680119542442885e-05,
      "loss": 2.8287117004394533,
      "memory(GiB)": 58.3,
      "step": 13360,
      "token_acc": 0.39666666666666667,
      "train_speed(iter/s)": 1.452691
    },
    {
      "epoch": 0.5725975750824729,
      "grad_norm": 4.105990886688232,
      "learning_rate": 9.679882654828822e-05,
      "loss": 2.871139144897461,
      "memory(GiB)": 58.3,
      "step": 13365,
      "token_acc": 0.3916349809885932,
      "train_speed(iter/s)": 1.452636
    },
    {
      "epoch": 0.5728117904117218,
      "grad_norm": 4.6735711097717285,
      "learning_rate": 9.679645682434295e-05,
      "loss": 2.658009910583496,
      "memory(GiB)": 58.3,
      "step": 13370,
      "token_acc": 0.4355400696864111,
      "train_speed(iter/s)": 1.452619
    },
    {
      "epoch": 0.5730260057409708,
      "grad_norm": 4.655219078063965,
      "learning_rate": 9.679408625263596e-05,
      "loss": 3.003887176513672,
      "memory(GiB)": 58.3,
      "step": 13375,
      "token_acc": 0.40058479532163743,
      "train_speed(iter/s)": 1.452713
    },
    {
      "epoch": 0.5732402210702198,
      "grad_norm": 3.323198080062866,
      "learning_rate": 9.679171483321022e-05,
      "loss": 2.4898963928222657,
      "memory(GiB)": 58.3,
      "step": 13380,
      "token_acc": 0.4980544747081712,
      "train_speed(iter/s)": 1.45276
    },
    {
      "epoch": 0.5734544363994687,
      "grad_norm": 9.692115783691406,
      "learning_rate": 9.678934256610864e-05,
      "loss": 2.6398977279663085,
      "memory(GiB)": 58.3,
      "step": 13385,
      "token_acc": 0.45625,
      "train_speed(iter/s)": 1.452801
    },
    {
      "epoch": 0.5736686517287177,
      "grad_norm": 6.924264907836914,
      "learning_rate": 9.678696945137424e-05,
      "loss": 2.2850440979003905,
      "memory(GiB)": 58.3,
      "step": 13390,
      "token_acc": 0.4789915966386555,
      "train_speed(iter/s)": 1.452856
    },
    {
      "epoch": 0.5738828670579667,
      "grad_norm": 3.8131535053253174,
      "learning_rate": 9.678459548905001e-05,
      "loss": 2.346193313598633,
      "memory(GiB)": 58.3,
      "step": 13395,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.452807
    },
    {
      "epoch": 0.5740970823872156,
      "grad_norm": 4.590564250946045,
      "learning_rate": 9.678222067917893e-05,
      "loss": 2.387767219543457,
      "memory(GiB)": 58.3,
      "step": 13400,
      "token_acc": 0.4979253112033195,
      "train_speed(iter/s)": 1.452786
    },
    {
      "epoch": 0.5743112977164646,
      "grad_norm": 4.188224792480469,
      "learning_rate": 9.677984502180405e-05,
      "loss": 2.72305965423584,
      "memory(GiB)": 58.3,
      "step": 13405,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.452682
    },
    {
      "epoch": 0.5745255130457135,
      "grad_norm": 3.7970058917999268,
      "learning_rate": 9.677746851696838e-05,
      "loss": 2.6192707061767577,
      "memory(GiB)": 58.3,
      "step": 13410,
      "token_acc": 0.45871559633027525,
      "train_speed(iter/s)": 1.452724
    },
    {
      "epoch": 0.5747397283749626,
      "grad_norm": 3.8336691856384277,
      "learning_rate": 9.6775091164715e-05,
      "loss": 2.6549158096313477,
      "memory(GiB)": 58.3,
      "step": 13415,
      "token_acc": 0.47289156626506024,
      "train_speed(iter/s)": 1.452819
    },
    {
      "epoch": 0.5749539437042115,
      "grad_norm": 4.714938163757324,
      "learning_rate": 9.677271296508697e-05,
      "loss": 2.6755447387695312,
      "memory(GiB)": 58.3,
      "step": 13420,
      "token_acc": 0.4357142857142857,
      "train_speed(iter/s)": 1.45289
    },
    {
      "epoch": 0.5751681590334604,
      "grad_norm": 5.060439109802246,
      "learning_rate": 9.677033391812736e-05,
      "loss": 2.711924362182617,
      "memory(GiB)": 58.3,
      "step": 13425,
      "token_acc": 0.4730290456431535,
      "train_speed(iter/s)": 1.452918
    },
    {
      "epoch": 0.5753823743627094,
      "grad_norm": 5.794318675994873,
      "learning_rate": 9.676795402387927e-05,
      "loss": 2.855929374694824,
      "memory(GiB)": 58.3,
      "step": 13430,
      "token_acc": 0.46381578947368424,
      "train_speed(iter/s)": 1.452941
    },
    {
      "epoch": 0.5755965896919584,
      "grad_norm": 6.0406951904296875,
      "learning_rate": 9.676557328238582e-05,
      "loss": 2.692300796508789,
      "memory(GiB)": 58.3,
      "step": 13435,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.453039
    },
    {
      "epoch": 0.5758108050212073,
      "grad_norm": 4.528384685516357,
      "learning_rate": 9.676319169369016e-05,
      "loss": 2.774568557739258,
      "memory(GiB)": 58.3,
      "step": 13440,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.45302
    },
    {
      "epoch": 0.5760250203504563,
      "grad_norm": 4.644454002380371,
      "learning_rate": 9.67608092578354e-05,
      "loss": 2.7741506576538084,
      "memory(GiB)": 58.3,
      "step": 13445,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.453008
    },
    {
      "epoch": 0.5762392356797053,
      "grad_norm": 4.520377159118652,
      "learning_rate": 9.675842597486471e-05,
      "loss": 2.6387094497680663,
      "memory(GiB)": 58.3,
      "step": 13450,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.45299
    },
    {
      "epoch": 0.5764534510089542,
      "grad_norm": 3.4945321083068848,
      "learning_rate": 9.675604184482129e-05,
      "loss": 2.656844711303711,
      "memory(GiB)": 58.3,
      "step": 13455,
      "token_acc": 0.4525316455696203,
      "train_speed(iter/s)": 1.453088
    },
    {
      "epoch": 0.5766676663382032,
      "grad_norm": 3.8861958980560303,
      "learning_rate": 9.67536568677483e-05,
      "loss": 2.69125919342041,
      "memory(GiB)": 58.3,
      "step": 13460,
      "token_acc": 0.4608150470219436,
      "train_speed(iter/s)": 1.453107
    },
    {
      "epoch": 0.5768818816674521,
      "grad_norm": 4.431004047393799,
      "learning_rate": 9.675127104368895e-05,
      "loss": 2.7479740142822267,
      "memory(GiB)": 58.3,
      "step": 13465,
      "token_acc": 0.46303501945525294,
      "train_speed(iter/s)": 1.453098
    },
    {
      "epoch": 0.5770960969967011,
      "grad_norm": 3.8639848232269287,
      "learning_rate": 9.674888437268648e-05,
      "loss": 3.028243637084961,
      "memory(GiB)": 58.3,
      "step": 13470,
      "token_acc": 0.4275092936802974,
      "train_speed(iter/s)": 1.453157
    },
    {
      "epoch": 0.5773103123259501,
      "grad_norm": 4.122154712677002,
      "learning_rate": 9.674649685478412e-05,
      "loss": 2.6113021850585936,
      "memory(GiB)": 58.3,
      "step": 13475,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.45307
    },
    {
      "epoch": 0.577524527655199,
      "grad_norm": 4.971024036407471,
      "learning_rate": 9.67441084900251e-05,
      "loss": 2.4940296173095704,
      "memory(GiB)": 58.3,
      "step": 13480,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.453074
    },
    {
      "epoch": 0.5777387429844479,
      "grad_norm": 4.5956130027771,
      "learning_rate": 9.674171927845272e-05,
      "loss": 2.6294580459594727,
      "memory(GiB)": 58.3,
      "step": 13485,
      "token_acc": 0.4591549295774648,
      "train_speed(iter/s)": 1.453102
    },
    {
      "epoch": 0.577952958313697,
      "grad_norm": 5.023074626922607,
      "learning_rate": 9.673932922011024e-05,
      "loss": 2.645197868347168,
      "memory(GiB)": 58.3,
      "step": 13490,
      "token_acc": 0.4437869822485207,
      "train_speed(iter/s)": 1.453032
    },
    {
      "epoch": 0.5781671736429459,
      "grad_norm": 4.95644474029541,
      "learning_rate": 9.673693831504096e-05,
      "loss": 2.6930280685424806,
      "memory(GiB)": 58.3,
      "step": 13495,
      "token_acc": 0.4349442379182156,
      "train_speed(iter/s)": 1.453145
    },
    {
      "epoch": 0.5783813889721948,
      "grad_norm": 6.048409461975098,
      "learning_rate": 9.673454656328823e-05,
      "loss": 2.411910057067871,
      "memory(GiB)": 58.3,
      "step": 13500,
      "token_acc": 0.5265017667844523,
      "train_speed(iter/s)": 1.453176
    },
    {
      "epoch": 0.5783813889721948,
      "eval_loss": 2.1747255325317383,
      "eval_runtime": 14.1219,
      "eval_samples_per_second": 7.081,
      "eval_steps_per_second": 7.081,
      "eval_token_acc": 0.4956268221574344,
      "step": 13500
    },
    {
      "epoch": 0.5785956043014439,
      "grad_norm": 5.690944671630859,
      "learning_rate": 9.673215396489531e-05,
      "loss": 2.3216552734375,
      "memory(GiB)": 58.3,
      "step": 13505,
      "token_acc": 0.5005452562704471,
      "train_speed(iter/s)": 1.450877
    },
    {
      "epoch": 0.5788098196306928,
      "grad_norm": 6.579805850982666,
      "learning_rate": 9.67297605199056e-05,
      "loss": 2.703226089477539,
      "memory(GiB)": 58.3,
      "step": 13510,
      "token_acc": 0.4708333333333333,
      "train_speed(iter/s)": 1.450922
    },
    {
      "epoch": 0.5790240349599417,
      "grad_norm": 4.9528656005859375,
      "learning_rate": 9.672736622836245e-05,
      "loss": 2.45620174407959,
      "memory(GiB)": 58.3,
      "step": 13515,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.450966
    },
    {
      "epoch": 0.5792382502891907,
      "grad_norm": 3.16727614402771,
      "learning_rate": 9.672497109030922e-05,
      "loss": 2.566771125793457,
      "memory(GiB)": 58.3,
      "step": 13520,
      "token_acc": 0.4623955431754875,
      "train_speed(iter/s)": 1.451063
    },
    {
      "epoch": 0.5794524656184397,
      "grad_norm": 3.9224820137023926,
      "learning_rate": 9.67225751057893e-05,
      "loss": 2.4791086196899412,
      "memory(GiB)": 58.3,
      "step": 13525,
      "token_acc": 0.4576923076923077,
      "train_speed(iter/s)": 1.451165
    },
    {
      "epoch": 0.5796666809476886,
      "grad_norm": 4.189754962921143,
      "learning_rate": 9.672017827484611e-05,
      "loss": 2.561331367492676,
      "memory(GiB)": 58.3,
      "step": 13530,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.451234
    },
    {
      "epoch": 0.5798808962769376,
      "grad_norm": 4.538440704345703,
      "learning_rate": 9.671778059752305e-05,
      "loss": 2.524432373046875,
      "memory(GiB)": 58.3,
      "step": 13535,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.451346
    },
    {
      "epoch": 0.5800951116061865,
      "grad_norm": 5.3694376945495605,
      "learning_rate": 9.67153820738636e-05,
      "loss": 2.5391387939453125,
      "memory(GiB)": 58.3,
      "step": 13540,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.451429
    },
    {
      "epoch": 0.5803093269354355,
      "grad_norm": 5.829346179962158,
      "learning_rate": 9.671298270391114e-05,
      "loss": 2.6271434783935548,
      "memory(GiB)": 58.3,
      "step": 13545,
      "token_acc": 0.45151515151515154,
      "train_speed(iter/s)": 1.451451
    },
    {
      "epoch": 0.5805235422646845,
      "grad_norm": 4.982487678527832,
      "learning_rate": 9.671058248770922e-05,
      "loss": 2.541544532775879,
      "memory(GiB)": 58.3,
      "step": 13550,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.451254
    },
    {
      "epoch": 0.5807377575939334,
      "grad_norm": 4.437562942504883,
      "learning_rate": 9.670818142530125e-05,
      "loss": 2.4570968627929686,
      "memory(GiB)": 58.3,
      "step": 13555,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.450998
    },
    {
      "epoch": 0.5809519729231823,
      "grad_norm": 5.505538463592529,
      "learning_rate": 9.670577951673076e-05,
      "loss": 2.573525238037109,
      "memory(GiB)": 58.3,
      "step": 13560,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.45109
    },
    {
      "epoch": 0.5811661882524314,
      "grad_norm": 5.254263401031494,
      "learning_rate": 9.670337676204127e-05,
      "loss": 2.2848575592041014,
      "memory(GiB)": 58.3,
      "step": 13565,
      "token_acc": 0.5342465753424658,
      "train_speed(iter/s)": 1.451095
    },
    {
      "epoch": 0.5813804035816803,
      "grad_norm": 4.772819519042969,
      "learning_rate": 9.670097316127631e-05,
      "loss": 2.5541152954101562,
      "memory(GiB)": 58.3,
      "step": 13570,
      "token_acc": 0.4662756598240469,
      "train_speed(iter/s)": 1.451135
    },
    {
      "epoch": 0.5815946189109292,
      "grad_norm": 3.9432716369628906,
      "learning_rate": 9.66985687144794e-05,
      "loss": 2.6513044357299806,
      "memory(GiB)": 58.3,
      "step": 13575,
      "token_acc": 0.45555555555555555,
      "train_speed(iter/s)": 1.451187
    },
    {
      "epoch": 0.5818088342401783,
      "grad_norm": 4.397954940795898,
      "learning_rate": 9.66961634216941e-05,
      "loss": 2.5172119140625,
      "memory(GiB)": 58.3,
      "step": 13580,
      "token_acc": 0.46794871794871795,
      "train_speed(iter/s)": 1.451316
    },
    {
      "epoch": 0.5820230495694272,
      "grad_norm": 4.439431190490723,
      "learning_rate": 9.669375728296402e-05,
      "loss": 2.5462577819824217,
      "memory(GiB)": 58.3,
      "step": 13585,
      "token_acc": 0.4646153846153846,
      "train_speed(iter/s)": 1.451258
    },
    {
      "epoch": 0.5822372648986761,
      "grad_norm": 5.510266304016113,
      "learning_rate": 9.66913502983327e-05,
      "loss": 2.8166122436523438,
      "memory(GiB)": 58.3,
      "step": 13590,
      "token_acc": 0.436046511627907,
      "train_speed(iter/s)": 1.451351
    },
    {
      "epoch": 0.5824514802279251,
      "grad_norm": 4.007482051849365,
      "learning_rate": 9.668894246784378e-05,
      "loss": 2.554096984863281,
      "memory(GiB)": 58.3,
      "step": 13595,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.451545
    },
    {
      "epoch": 0.5826656955571741,
      "grad_norm": 5.670605182647705,
      "learning_rate": 9.668653379154086e-05,
      "loss": 2.6899135589599608,
      "memory(GiB)": 58.3,
      "step": 13600,
      "token_acc": 0.4155844155844156,
      "train_speed(iter/s)": 1.45173
    },
    {
      "epoch": 0.582879910886423,
      "grad_norm": 6.042681694030762,
      "learning_rate": 9.66841242694676e-05,
      "loss": 2.5125286102294924,
      "memory(GiB)": 58.3,
      "step": 13605,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.451881
    },
    {
      "epoch": 0.583094126215672,
      "grad_norm": 5.8735480308532715,
      "learning_rate": 9.668171390166763e-05,
      "loss": 2.506675148010254,
      "memory(GiB)": 58.3,
      "step": 13610,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.451908
    },
    {
      "epoch": 0.583308341544921,
      "grad_norm": 4.454692840576172,
      "learning_rate": 9.667930268818462e-05,
      "loss": 2.5537179946899413,
      "memory(GiB)": 58.3,
      "step": 13615,
      "token_acc": 0.4377224199288256,
      "train_speed(iter/s)": 1.451952
    },
    {
      "epoch": 0.5835225568741699,
      "grad_norm": 3.755244493484497,
      "learning_rate": 9.667689062906226e-05,
      "loss": 2.945166015625,
      "memory(GiB)": 58.3,
      "step": 13620,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.452108
    },
    {
      "epoch": 0.5837367722034189,
      "grad_norm": 3.8499667644500732,
      "learning_rate": 9.667447772434423e-05,
      "loss": 2.7186246871948243,
      "memory(GiB)": 58.3,
      "step": 13625,
      "token_acc": 0.4720670391061452,
      "train_speed(iter/s)": 1.452228
    },
    {
      "epoch": 0.5839509875326678,
      "grad_norm": 4.239790439605713,
      "learning_rate": 9.667206397407426e-05,
      "loss": 2.6613414764404295,
      "memory(GiB)": 58.3,
      "step": 13630,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.452201
    },
    {
      "epoch": 0.5841652028619168,
      "grad_norm": 4.446804046630859,
      "learning_rate": 9.666964937829606e-05,
      "loss": 2.292164611816406,
      "memory(GiB)": 58.3,
      "step": 13635,
      "token_acc": 0.5393258426966292,
      "train_speed(iter/s)": 1.452264
    },
    {
      "epoch": 0.5843794181911658,
      "grad_norm": 4.99111270904541,
      "learning_rate": 9.666723393705339e-05,
      "loss": 2.611367034912109,
      "memory(GiB)": 58.3,
      "step": 13640,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.452283
    },
    {
      "epoch": 0.5845936335204147,
      "grad_norm": 5.620711803436279,
      "learning_rate": 9.666481765038999e-05,
      "loss": 2.590936851501465,
      "memory(GiB)": 58.3,
      "step": 13645,
      "token_acc": 0.5054151624548736,
      "train_speed(iter/s)": 1.452348
    },
    {
      "epoch": 0.5848078488496636,
      "grad_norm": 3.6960132122039795,
      "learning_rate": 9.666240051834965e-05,
      "loss": 2.449190139770508,
      "memory(GiB)": 58.3,
      "step": 13650,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.452411
    },
    {
      "epoch": 0.5850220641789127,
      "grad_norm": 4.801359176635742,
      "learning_rate": 9.665998254097616e-05,
      "loss": 2.8072816848754885,
      "memory(GiB)": 58.3,
      "step": 13655,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.452259
    },
    {
      "epoch": 0.5852362795081616,
      "grad_norm": 6.511500358581543,
      "learning_rate": 9.66575637183133e-05,
      "loss": 2.620912551879883,
      "memory(GiB)": 58.3,
      "step": 13660,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.452318
    },
    {
      "epoch": 0.5854504948374105,
      "grad_norm": 4.2787652015686035,
      "learning_rate": 9.665514405040491e-05,
      "loss": 2.427029037475586,
      "memory(GiB)": 58.3,
      "step": 13665,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.4523
    },
    {
      "epoch": 0.5856647101666596,
      "grad_norm": 5.601728916168213,
      "learning_rate": 9.665272353729482e-05,
      "loss": 2.292038917541504,
      "memory(GiB)": 58.3,
      "step": 13670,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.452293
    },
    {
      "epoch": 0.5858789254959085,
      "grad_norm": 5.865860462188721,
      "learning_rate": 9.665030217902688e-05,
      "loss": 2.4172096252441406,
      "memory(GiB)": 58.3,
      "step": 13675,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.452235
    },
    {
      "epoch": 0.5860931408251574,
      "grad_norm": 4.575709342956543,
      "learning_rate": 9.664787997564496e-05,
      "loss": 2.5697406768798827,
      "memory(GiB)": 58.3,
      "step": 13680,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.452081
    },
    {
      "epoch": 0.5863073561544064,
      "grad_norm": 3.8804333209991455,
      "learning_rate": 9.664545692719293e-05,
      "loss": 2.788030815124512,
      "memory(GiB)": 58.3,
      "step": 13685,
      "token_acc": 0.4523076923076923,
      "train_speed(iter/s)": 1.452111
    },
    {
      "epoch": 0.5865215714836554,
      "grad_norm": 4.499039649963379,
      "learning_rate": 9.66430330337147e-05,
      "loss": 2.9364303588867187,
      "memory(GiB)": 58.3,
      "step": 13690,
      "token_acc": 0.4273255813953488,
      "train_speed(iter/s)": 1.452193
    },
    {
      "epoch": 0.5867357868129043,
      "grad_norm": 3.6702146530151367,
      "learning_rate": 9.664060829525416e-05,
      "loss": 2.805882453918457,
      "memory(GiB)": 58.3,
      "step": 13695,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.452258
    },
    {
      "epoch": 0.5869500021421533,
      "grad_norm": 2.9204540252685547,
      "learning_rate": 9.663818271185525e-05,
      "loss": 2.2844181060791016,
      "memory(GiB)": 58.3,
      "step": 13700,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.452122
    },
    {
      "epoch": 0.5871642174714022,
      "grad_norm": 4.366288661956787,
      "learning_rate": 9.663575628356191e-05,
      "loss": 2.7558219909667967,
      "memory(GiB)": 58.3,
      "step": 13705,
      "token_acc": 0.41479099678456594,
      "train_speed(iter/s)": 1.452145
    },
    {
      "epoch": 0.5873784328006512,
      "grad_norm": 3.9321718215942383,
      "learning_rate": 9.663332901041809e-05,
      "loss": 2.560669708251953,
      "memory(GiB)": 58.3,
      "step": 13710,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.452137
    },
    {
      "epoch": 0.5875926481299002,
      "grad_norm": 4.446691513061523,
      "learning_rate": 9.663090089246778e-05,
      "loss": 2.5037235260009765,
      "memory(GiB)": 58.3,
      "step": 13715,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.452131
    },
    {
      "epoch": 0.5878068634591491,
      "grad_norm": 5.75213098526001,
      "learning_rate": 9.662847192975496e-05,
      "loss": 2.622756004333496,
      "memory(GiB)": 58.3,
      "step": 13720,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.452148
    },
    {
      "epoch": 0.588021078788398,
      "grad_norm": 3.5477755069732666,
      "learning_rate": 9.662604212232362e-05,
      "loss": 2.5508455276489257,
      "memory(GiB)": 58.3,
      "step": 13725,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.452162
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 4.282444953918457,
      "learning_rate": 9.662361147021779e-05,
      "loss": 2.6218746185302733,
      "memory(GiB)": 58.3,
      "step": 13730,
      "token_acc": 0.45396825396825397,
      "train_speed(iter/s)": 1.452029
    },
    {
      "epoch": 0.588449509446896,
      "grad_norm": 6.025940895080566,
      "learning_rate": 9.662117997348151e-05,
      "loss": 2.7067113876342774,
      "memory(GiB)": 58.3,
      "step": 13735,
      "token_acc": 0.4275092936802974,
      "train_speed(iter/s)": 1.452049
    },
    {
      "epoch": 0.588663724776145,
      "grad_norm": 5.109095573425293,
      "learning_rate": 9.661874763215881e-05,
      "loss": 2.4650619506835936,
      "memory(GiB)": 58.3,
      "step": 13740,
      "token_acc": 0.49836065573770494,
      "train_speed(iter/s)": 1.452106
    },
    {
      "epoch": 0.588877940105394,
      "grad_norm": 3.9077296257019043,
      "learning_rate": 9.661631444629378e-05,
      "loss": 2.58374080657959,
      "memory(GiB)": 58.3,
      "step": 13745,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.452199
    },
    {
      "epoch": 0.5890921554346429,
      "grad_norm": 4.706995964050293,
      "learning_rate": 9.661388041593047e-05,
      "loss": 2.468706512451172,
      "memory(GiB)": 58.3,
      "step": 13750,
      "token_acc": 0.4770992366412214,
      "train_speed(iter/s)": 1.452121
    },
    {
      "epoch": 0.5893063707638919,
      "grad_norm": 6.8934197425842285,
      "learning_rate": 9.661144554111298e-05,
      "loss": 2.5233896255493162,
      "memory(GiB)": 58.3,
      "step": 13755,
      "token_acc": 0.4315068493150685,
      "train_speed(iter/s)": 1.452287
    },
    {
      "epoch": 0.5895205860931408,
      "grad_norm": 5.084076404571533,
      "learning_rate": 9.660900982188545e-05,
      "loss": 2.863209533691406,
      "memory(GiB)": 58.3,
      "step": 13760,
      "token_acc": 0.41721854304635764,
      "train_speed(iter/s)": 1.452329
    },
    {
      "epoch": 0.5897348014223898,
      "grad_norm": 4.910953998565674,
      "learning_rate": 9.660657325829198e-05,
      "loss": 2.768596649169922,
      "memory(GiB)": 58.3,
      "step": 13765,
      "token_acc": 0.4542124542124542,
      "train_speed(iter/s)": 1.452159
    },
    {
      "epoch": 0.5899490167516388,
      "grad_norm": 4.866161823272705,
      "learning_rate": 9.660413585037671e-05,
      "loss": 2.6919881820678713,
      "memory(GiB)": 58.3,
      "step": 13770,
      "token_acc": 0.44285714285714284,
      "train_speed(iter/s)": 1.452161
    },
    {
      "epoch": 0.5901632320808877,
      "grad_norm": 4.18895959854126,
      "learning_rate": 9.66016975981838e-05,
      "loss": 2.540228843688965,
      "memory(GiB)": 58.3,
      "step": 13775,
      "token_acc": 0.46099290780141844,
      "train_speed(iter/s)": 1.452056
    },
    {
      "epoch": 0.5903774474101366,
      "grad_norm": 3.4329934120178223,
      "learning_rate": 9.659925850175742e-05,
      "loss": 2.3678192138671874,
      "memory(GiB)": 58.3,
      "step": 13780,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.452087
    },
    {
      "epoch": 0.5905916627393857,
      "grad_norm": 5.119359016418457,
      "learning_rate": 9.659681856114177e-05,
      "loss": 2.637171173095703,
      "memory(GiB)": 58.3,
      "step": 13785,
      "token_acc": 0.4624505928853755,
      "train_speed(iter/s)": 1.452083
    },
    {
      "epoch": 0.5908058780686346,
      "grad_norm": 5.700256824493408,
      "learning_rate": 9.659437777638102e-05,
      "loss": 2.883258819580078,
      "memory(GiB)": 58.3,
      "step": 13790,
      "token_acc": 0.42962962962962964,
      "train_speed(iter/s)": 1.452093
    },
    {
      "epoch": 0.5910200933978835,
      "grad_norm": 5.514988899230957,
      "learning_rate": 9.659193614751942e-05,
      "loss": 2.347756195068359,
      "memory(GiB)": 58.3,
      "step": 13795,
      "token_acc": 0.528957528957529,
      "train_speed(iter/s)": 1.45209
    },
    {
      "epoch": 0.5912343087271326,
      "grad_norm": 5.882185935974121,
      "learning_rate": 9.658949367460119e-05,
      "loss": 2.6247514724731444,
      "memory(GiB)": 58.3,
      "step": 13800,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.452121
    },
    {
      "epoch": 0.5914485240563815,
      "grad_norm": 8.835413932800293,
      "learning_rate": 9.658705035767059e-05,
      "loss": 2.618433380126953,
      "memory(GiB)": 58.3,
      "step": 13805,
      "token_acc": 0.47651006711409394,
      "train_speed(iter/s)": 1.452129
    },
    {
      "epoch": 0.5916627393856304,
      "grad_norm": 4.470920085906982,
      "learning_rate": 9.658460619677185e-05,
      "loss": 2.285435676574707,
      "memory(GiB)": 58.3,
      "step": 13810,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.45213
    },
    {
      "epoch": 0.5918769547148794,
      "grad_norm": 4.125425815582275,
      "learning_rate": 9.658216119194929e-05,
      "loss": 2.206900978088379,
      "memory(GiB)": 58.3,
      "step": 13815,
      "token_acc": 0.5318352059925093,
      "train_speed(iter/s)": 1.452085
    },
    {
      "epoch": 0.5920911700441284,
      "grad_norm": 3.692875862121582,
      "learning_rate": 9.657971534324717e-05,
      "loss": 2.577462005615234,
      "memory(GiB)": 58.3,
      "step": 13820,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.452077
    },
    {
      "epoch": 0.5923053853733773,
      "grad_norm": 3.980250835418701,
      "learning_rate": 9.657726865070981e-05,
      "loss": 2.7860286712646483,
      "memory(GiB)": 58.3,
      "step": 13825,
      "token_acc": 0.40540540540540543,
      "train_speed(iter/s)": 1.452021
    },
    {
      "epoch": 0.5925196007026263,
      "grad_norm": 4.529167175292969,
      "learning_rate": 9.657482111438154e-05,
      "loss": 2.538942337036133,
      "memory(GiB)": 58.3,
      "step": 13830,
      "token_acc": 0.4866920152091255,
      "train_speed(iter/s)": 1.452063
    },
    {
      "epoch": 0.5927338160318752,
      "grad_norm": 3.8707027435302734,
      "learning_rate": 9.657237273430669e-05,
      "loss": 2.565081024169922,
      "memory(GiB)": 58.3,
      "step": 13835,
      "token_acc": 0.4653846153846154,
      "train_speed(iter/s)": 1.452147
    },
    {
      "epoch": 0.5929480313611242,
      "grad_norm": 4.658082962036133,
      "learning_rate": 9.656992351052961e-05,
      "loss": 2.672258758544922,
      "memory(GiB)": 58.3,
      "step": 13840,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.452335
    },
    {
      "epoch": 0.5931622466903732,
      "grad_norm": 4.3584465980529785,
      "learning_rate": 9.656747344309469e-05,
      "loss": 2.639421081542969,
      "memory(GiB)": 58.3,
      "step": 13845,
      "token_acc": 0.42105263157894735,
      "train_speed(iter/s)": 1.45237
    },
    {
      "epoch": 0.5933764620196221,
      "grad_norm": 3.647871255874634,
      "learning_rate": 9.656502253204632e-05,
      "loss": 2.8180770874023438,
      "memory(GiB)": 58.3,
      "step": 13850,
      "token_acc": 0.44155844155844154,
      "train_speed(iter/s)": 1.45213
    },
    {
      "epoch": 0.593590677348871,
      "grad_norm": 4.493406772613525,
      "learning_rate": 9.656257077742886e-05,
      "loss": 2.958848571777344,
      "memory(GiB)": 58.3,
      "step": 13855,
      "token_acc": 0.40304182509505704,
      "train_speed(iter/s)": 1.452213
    },
    {
      "epoch": 0.5938048926781201,
      "grad_norm": 4.315821170806885,
      "learning_rate": 9.656011817928676e-05,
      "loss": 2.5926551818847656,
      "memory(GiB)": 58.3,
      "step": 13860,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.452184
    },
    {
      "epoch": 0.594019108007369,
      "grad_norm": 3.80910325050354,
      "learning_rate": 9.655766473766444e-05,
      "loss": 2.6649192810058593,
      "memory(GiB)": 58.3,
      "step": 13865,
      "token_acc": 0.490625,
      "train_speed(iter/s)": 1.452193
    },
    {
      "epoch": 0.5942333233366179,
      "grad_norm": 4.4840288162231445,
      "learning_rate": 9.655521045260636e-05,
      "loss": 2.464204216003418,
      "memory(GiB)": 58.3,
      "step": 13870,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.452183
    },
    {
      "epoch": 0.594447538665867,
      "grad_norm": 4.357649803161621,
      "learning_rate": 9.655275532415696e-05,
      "loss": 2.6859384536743165,
      "memory(GiB)": 58.3,
      "step": 13875,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.452042
    },
    {
      "epoch": 0.5946617539951159,
      "grad_norm": 3.79764986038208,
      "learning_rate": 9.655029935236073e-05,
      "loss": 2.5860641479492186,
      "memory(GiB)": 58.3,
      "step": 13880,
      "token_acc": 0.4727272727272727,
      "train_speed(iter/s)": 1.452174
    },
    {
      "epoch": 0.5948759693243648,
      "grad_norm": 4.252235412597656,
      "learning_rate": 9.654784253726216e-05,
      "loss": 2.846329689025879,
      "memory(GiB)": 58.3,
      "step": 13885,
      "token_acc": 0.4377358490566038,
      "train_speed(iter/s)": 1.452104
    },
    {
      "epoch": 0.5950901846536139,
      "grad_norm": 4.272499084472656,
      "learning_rate": 9.654538487890577e-05,
      "loss": 2.596055603027344,
      "memory(GiB)": 58.3,
      "step": 13890,
      "token_acc": 0.5150501672240803,
      "train_speed(iter/s)": 1.452166
    },
    {
      "epoch": 0.5953043999828628,
      "grad_norm": 4.841752052307129,
      "learning_rate": 9.654292637733605e-05,
      "loss": 2.573365592956543,
      "memory(GiB)": 58.3,
      "step": 13895,
      "token_acc": 0.44074074074074077,
      "train_speed(iter/s)": 1.452163
    },
    {
      "epoch": 0.5955186153121117,
      "grad_norm": 3.2819955348968506,
      "learning_rate": 9.654046703259755e-05,
      "loss": 2.2940391540527343,
      "memory(GiB)": 58.3,
      "step": 13900,
      "token_acc": 0.49044585987261147,
      "train_speed(iter/s)": 1.452121
    },
    {
      "epoch": 0.5957328306413607,
      "grad_norm": 4.382943153381348,
      "learning_rate": 9.653800684473485e-05,
      "loss": 2.7047683715820314,
      "memory(GiB)": 58.3,
      "step": 13905,
      "token_acc": 0.461038961038961,
      "train_speed(iter/s)": 1.452076
    },
    {
      "epoch": 0.5959470459706097,
      "grad_norm": 3.7000436782836914,
      "learning_rate": 9.653554581379249e-05,
      "loss": 2.6412469863891603,
      "memory(GiB)": 58.3,
      "step": 13910,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.452066
    },
    {
      "epoch": 0.5961612612998586,
      "grad_norm": 3.6861205101013184,
      "learning_rate": 9.653308393981508e-05,
      "loss": 2.781342697143555,
      "memory(GiB)": 58.3,
      "step": 13915,
      "token_acc": 0.4152823920265781,
      "train_speed(iter/s)": 1.45207
    },
    {
      "epoch": 0.5963754766291076,
      "grad_norm": 4.117077827453613,
      "learning_rate": 9.65306212228472e-05,
      "loss": 2.6871110916137697,
      "memory(GiB)": 58.3,
      "step": 13920,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.452144
    },
    {
      "epoch": 0.5965896919583565,
      "grad_norm": 4.764352798461914,
      "learning_rate": 9.652815766293344e-05,
      "loss": 2.4248533248901367,
      "memory(GiB)": 58.3,
      "step": 13925,
      "token_acc": 0.4983277591973244,
      "train_speed(iter/s)": 1.452105
    },
    {
      "epoch": 0.5968039072876055,
      "grad_norm": 4.078861236572266,
      "learning_rate": 9.652569326011849e-05,
      "loss": 2.328511619567871,
      "memory(GiB)": 58.3,
      "step": 13930,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.452074
    },
    {
      "epoch": 0.5970181226168545,
      "grad_norm": 3.8641281127929688,
      "learning_rate": 9.652322801444695e-05,
      "loss": 2.7165510177612306,
      "memory(GiB)": 58.3,
      "step": 13935,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.452143
    },
    {
      "epoch": 0.5972323379461034,
      "grad_norm": 5.55017614364624,
      "learning_rate": 9.652076192596349e-05,
      "loss": 2.7470321655273438,
      "memory(GiB)": 58.3,
      "step": 13940,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.452266
    },
    {
      "epoch": 0.5974465532753523,
      "grad_norm": 3.43029522895813,
      "learning_rate": 9.65182949947128e-05,
      "loss": 2.4457880020141602,
      "memory(GiB)": 58.3,
      "step": 13945,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.452376
    },
    {
      "epoch": 0.5976607686046014,
      "grad_norm": 3.934824228286743,
      "learning_rate": 9.651582722073953e-05,
      "loss": 2.572235870361328,
      "memory(GiB)": 58.3,
      "step": 13950,
      "token_acc": 0.4645390070921986,
      "train_speed(iter/s)": 1.452451
    },
    {
      "epoch": 0.5978749839338503,
      "grad_norm": 4.444162845611572,
      "learning_rate": 9.651335860408843e-05,
      "loss": 2.563711929321289,
      "memory(GiB)": 58.3,
      "step": 13955,
      "token_acc": 0.5056179775280899,
      "train_speed(iter/s)": 1.452566
    },
    {
      "epoch": 0.5980891992630992,
      "grad_norm": 5.351120948791504,
      "learning_rate": 9.651088914480421e-05,
      "loss": 2.1993560791015625,
      "memory(GiB)": 58.3,
      "step": 13960,
      "token_acc": 0.5021834061135371,
      "train_speed(iter/s)": 1.452669
    },
    {
      "epoch": 0.5983034145923483,
      "grad_norm": 5.675091743469238,
      "learning_rate": 9.650841884293159e-05,
      "loss": 2.4855014801025392,
      "memory(GiB)": 58.3,
      "step": 13965,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.452715
    },
    {
      "epoch": 0.5985176299215972,
      "grad_norm": 6.399936676025391,
      "learning_rate": 9.650594769851535e-05,
      "loss": 2.689995002746582,
      "memory(GiB)": 58.3,
      "step": 13970,
      "token_acc": 0.4892703862660944,
      "train_speed(iter/s)": 1.452744
    },
    {
      "epoch": 0.5987318452508461,
      "grad_norm": 4.217350959777832,
      "learning_rate": 9.650347571160023e-05,
      "loss": 2.794980049133301,
      "memory(GiB)": 58.3,
      "step": 13975,
      "token_acc": 0.423728813559322,
      "train_speed(iter/s)": 1.452855
    },
    {
      "epoch": 0.5989460605800951,
      "grad_norm": 3.8354604244232178,
      "learning_rate": 9.650100288223102e-05,
      "loss": 2.725836181640625,
      "memory(GiB)": 58.3,
      "step": 13980,
      "token_acc": 0.42902208201892744,
      "train_speed(iter/s)": 1.453026
    },
    {
      "epoch": 0.5991602759093441,
      "grad_norm": 4.5916900634765625,
      "learning_rate": 9.649852921045254e-05,
      "loss": 2.405137634277344,
      "memory(GiB)": 58.3,
      "step": 13985,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.453117
    },
    {
      "epoch": 0.599374491238593,
      "grad_norm": 3.5865020751953125,
      "learning_rate": 9.649605469630958e-05,
      "loss": 2.7852060317993166,
      "memory(GiB)": 58.3,
      "step": 13990,
      "token_acc": 0.4230769230769231,
      "train_speed(iter/s)": 1.453108
    },
    {
      "epoch": 0.599588706567842,
      "grad_norm": 4.899942874908447,
      "learning_rate": 9.649357933984697e-05,
      "loss": 2.580856704711914,
      "memory(GiB)": 58.3,
      "step": 13995,
      "token_acc": 0.4206008583690987,
      "train_speed(iter/s)": 1.453176
    },
    {
      "epoch": 0.599802921897091,
      "grad_norm": 4.871976375579834,
      "learning_rate": 9.649110314110954e-05,
      "loss": 2.4768436431884764,
      "memory(GiB)": 58.3,
      "step": 14000,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.453175
    },
    {
      "epoch": 0.599802921897091,
      "eval_loss": 2.266469717025757,
      "eval_runtime": 14.8183,
      "eval_samples_per_second": 6.748,
      "eval_steps_per_second": 6.748,
      "eval_token_acc": 0.45633456334563344,
      "step": 14000
    },
    {
      "epoch": 0.6000171372263399,
      "grad_norm": 4.842276096343994,
      "learning_rate": 9.648862610014218e-05,
      "loss": 2.5272029876708983,
      "memory(GiB)": 58.3,
      "step": 14005,
      "token_acc": 0.45356176735798015,
      "train_speed(iter/s)": 1.450832
    },
    {
      "epoch": 0.6002313525555889,
      "grad_norm": 5.21980619430542,
      "learning_rate": 9.648614821698975e-05,
      "loss": 2.73854923248291,
      "memory(GiB)": 58.3,
      "step": 14010,
      "token_acc": 0.4532871972318339,
      "train_speed(iter/s)": 1.450948
    },
    {
      "epoch": 0.6004455678848378,
      "grad_norm": 5.694094181060791,
      "learning_rate": 9.648366949169712e-05,
      "loss": 2.8148332595825196,
      "memory(GiB)": 58.3,
      "step": 14015,
      "token_acc": 0.43125,
      "train_speed(iter/s)": 1.451003
    },
    {
      "epoch": 0.6006597832140868,
      "grad_norm": 4.2976484298706055,
      "learning_rate": 9.648118992430923e-05,
      "loss": 2.3018362045288088,
      "memory(GiB)": 58.3,
      "step": 14020,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.450999
    },
    {
      "epoch": 0.6008739985433358,
      "grad_norm": 6.890940189361572,
      "learning_rate": 9.647870951487098e-05,
      "loss": 2.737139892578125,
      "memory(GiB)": 58.3,
      "step": 14025,
      "token_acc": 0.4107142857142857,
      "train_speed(iter/s)": 1.450921
    },
    {
      "epoch": 0.6010882138725847,
      "grad_norm": 7.609458923339844,
      "learning_rate": 9.64762282634273e-05,
      "loss": 2.702621269226074,
      "memory(GiB)": 58.3,
      "step": 14030,
      "token_acc": 0.46332046332046334,
      "train_speed(iter/s)": 1.45093
    },
    {
      "epoch": 0.6013024292018336,
      "grad_norm": 4.618037223815918,
      "learning_rate": 9.647374617002314e-05,
      "loss": 2.8386764526367188,
      "memory(GiB)": 58.3,
      "step": 14035,
      "token_acc": 0.47854785478547857,
      "train_speed(iter/s)": 1.450964
    },
    {
      "epoch": 0.6015166445310827,
      "grad_norm": 5.122060298919678,
      "learning_rate": 9.647126323470349e-05,
      "loss": 2.4943979263305662,
      "memory(GiB)": 58.3,
      "step": 14040,
      "token_acc": 0.47580645161290325,
      "train_speed(iter/s)": 1.450872
    },
    {
      "epoch": 0.6017308598603316,
      "grad_norm": 4.990699291229248,
      "learning_rate": 9.646877945751332e-05,
      "loss": 2.692562484741211,
      "memory(GiB)": 58.3,
      "step": 14045,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.451002
    },
    {
      "epoch": 0.6019450751895805,
      "grad_norm": 4.44877290725708,
      "learning_rate": 9.646629483849758e-05,
      "loss": 2.585972213745117,
      "memory(GiB)": 58.3,
      "step": 14050,
      "token_acc": 0.4463276836158192,
      "train_speed(iter/s)": 1.450945
    },
    {
      "epoch": 0.6021592905188295,
      "grad_norm": 4.9418792724609375,
      "learning_rate": 9.646380937770135e-05,
      "loss": 2.6613460540771485,
      "memory(GiB)": 58.3,
      "step": 14055,
      "token_acc": 0.4407894736842105,
      "train_speed(iter/s)": 1.451073
    },
    {
      "epoch": 0.6023735058480785,
      "grad_norm": 5.462378978729248,
      "learning_rate": 9.646132307516961e-05,
      "loss": 2.4605152130126955,
      "memory(GiB)": 58.3,
      "step": 14060,
      "token_acc": 0.5171102661596958,
      "train_speed(iter/s)": 1.451083
    },
    {
      "epoch": 0.6025877211773274,
      "grad_norm": 4.256039142608643,
      "learning_rate": 9.645883593094745e-05,
      "loss": 2.7011783599853514,
      "memory(GiB)": 58.3,
      "step": 14065,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.451087
    },
    {
      "epoch": 0.6028019365065764,
      "grad_norm": 3.353818655014038,
      "learning_rate": 9.645634794507987e-05,
      "loss": 2.4486928939819337,
      "memory(GiB)": 58.3,
      "step": 14070,
      "token_acc": 0.4735202492211838,
      "train_speed(iter/s)": 1.451147
    },
    {
      "epoch": 0.6030161518358254,
      "grad_norm": 4.841371059417725,
      "learning_rate": 9.645385911761196e-05,
      "loss": 2.2656063079833983,
      "memory(GiB)": 58.3,
      "step": 14075,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.450907
    },
    {
      "epoch": 0.6032303671650744,
      "grad_norm": 5.12148904800415,
      "learning_rate": 9.645136944858883e-05,
      "loss": 2.6110525131225586,
      "memory(GiB)": 58.3,
      "step": 14080,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.450824
    },
    {
      "epoch": 0.6034445824943233,
      "grad_norm": 4.80199670791626,
      "learning_rate": 9.644887893805557e-05,
      "loss": 2.7595199584960937,
      "memory(GiB)": 58.3,
      "step": 14085,
      "token_acc": 0.4554794520547945,
      "train_speed(iter/s)": 1.450818
    },
    {
      "epoch": 0.6036587978235722,
      "grad_norm": 4.588850975036621,
      "learning_rate": 9.644638758605729e-05,
      "loss": 2.67706413269043,
      "memory(GiB)": 58.3,
      "step": 14090,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.450912
    },
    {
      "epoch": 0.6038730131528213,
      "grad_norm": 6.626927852630615,
      "learning_rate": 9.644389539263911e-05,
      "loss": 2.828997993469238,
      "memory(GiB)": 58.3,
      "step": 14095,
      "token_acc": 0.39622641509433965,
      "train_speed(iter/s)": 1.450938
    },
    {
      "epoch": 0.6040872284820702,
      "grad_norm": 4.618725776672363,
      "learning_rate": 9.644140235784623e-05,
      "loss": 2.8019035339355467,
      "memory(GiB)": 58.3,
      "step": 14100,
      "token_acc": 0.4652777777777778,
      "train_speed(iter/s)": 1.45098
    },
    {
      "epoch": 0.6043014438113191,
      "grad_norm": 3.681743860244751,
      "learning_rate": 9.643890848172377e-05,
      "loss": 2.538693428039551,
      "memory(GiB)": 58.3,
      "step": 14105,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.450892
    },
    {
      "epoch": 0.6045156591405682,
      "grad_norm": 4.0561842918396,
      "learning_rate": 9.643641376431691e-05,
      "loss": 2.4754875183105467,
      "memory(GiB)": 58.3,
      "step": 14110,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.450895
    },
    {
      "epoch": 0.6047298744698171,
      "grad_norm": 5.372085094451904,
      "learning_rate": 9.643391820567084e-05,
      "loss": 2.580911064147949,
      "memory(GiB)": 58.3,
      "step": 14115,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.450821
    },
    {
      "epoch": 0.604944089799066,
      "grad_norm": 10.50987720489502,
      "learning_rate": 9.643142180583081e-05,
      "loss": 2.5099327087402346,
      "memory(GiB)": 58.3,
      "step": 14120,
      "token_acc": 0.4489051094890511,
      "train_speed(iter/s)": 1.450813
    },
    {
      "epoch": 0.605158305128315,
      "grad_norm": 5.173738956451416,
      "learning_rate": 9.642892456484201e-05,
      "loss": 2.69002742767334,
      "memory(GiB)": 58.3,
      "step": 14125,
      "token_acc": 0.4037854889589905,
      "train_speed(iter/s)": 1.450797
    },
    {
      "epoch": 0.605372520457564,
      "grad_norm": 4.199275970458984,
      "learning_rate": 9.642642648274968e-05,
      "loss": 2.5305654525756838,
      "memory(GiB)": 58.3,
      "step": 14130,
      "token_acc": 0.4961832061068702,
      "train_speed(iter/s)": 1.450635
    },
    {
      "epoch": 0.6055867357868129,
      "grad_norm": 3.784954786300659,
      "learning_rate": 9.64239275595991e-05,
      "loss": 2.4811016082763673,
      "memory(GiB)": 58.3,
      "step": 14135,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.450555
    },
    {
      "epoch": 0.6058009511160619,
      "grad_norm": 6.651519298553467,
      "learning_rate": 9.642142779543551e-05,
      "loss": 2.6361852645874024,
      "memory(GiB)": 58.3,
      "step": 14140,
      "token_acc": 0.4892966360856269,
      "train_speed(iter/s)": 1.450674
    },
    {
      "epoch": 0.6060151664453108,
      "grad_norm": 3.5732951164245605,
      "learning_rate": 9.641892719030421e-05,
      "loss": 2.4061546325683594,
      "memory(GiB)": 58.3,
      "step": 14145,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.450729
    },
    {
      "epoch": 0.6062293817745598,
      "grad_norm": 3.4998037815093994,
      "learning_rate": 9.64164257442505e-05,
      "loss": 2.5439785003662108,
      "memory(GiB)": 58.3,
      "step": 14150,
      "token_acc": 0.44404332129963897,
      "train_speed(iter/s)": 1.450569
    },
    {
      "epoch": 0.6064435971038088,
      "grad_norm": 4.5941290855407715,
      "learning_rate": 9.64139234573197e-05,
      "loss": 2.0855281829833983,
      "memory(GiB)": 58.3,
      "step": 14155,
      "token_acc": 0.5274261603375527,
      "train_speed(iter/s)": 1.450455
    },
    {
      "epoch": 0.6066578124330577,
      "grad_norm": 5.384954452514648,
      "learning_rate": 9.641142032955714e-05,
      "loss": 2.896497344970703,
      "memory(GiB)": 58.3,
      "step": 14160,
      "token_acc": 0.4253731343283582,
      "train_speed(iter/s)": 1.450449
    },
    {
      "epoch": 0.6068720277623066,
      "grad_norm": 4.107150554656982,
      "learning_rate": 9.640891636100815e-05,
      "loss": 2.663096809387207,
      "memory(GiB)": 58.3,
      "step": 14165,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.450381
    },
    {
      "epoch": 0.6070862430915557,
      "grad_norm": 4.029058456420898,
      "learning_rate": 9.640641155171812e-05,
      "loss": 2.3572019577026366,
      "memory(GiB)": 58.3,
      "step": 14170,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.450346
    },
    {
      "epoch": 0.6073004584208046,
      "grad_norm": 4.141958236694336,
      "learning_rate": 9.64039059017324e-05,
      "loss": 2.4927330017089844,
      "memory(GiB)": 58.3,
      "step": 14175,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.450355
    },
    {
      "epoch": 0.6075146737500535,
      "grad_norm": 4.207716941833496,
      "learning_rate": 9.64013994110964e-05,
      "loss": 2.625646209716797,
      "memory(GiB)": 58.3,
      "step": 14180,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.450491
    },
    {
      "epoch": 0.6077288890793026,
      "grad_norm": 4.127267360687256,
      "learning_rate": 9.639889207985552e-05,
      "loss": 2.9878454208374023,
      "memory(GiB)": 58.3,
      "step": 14185,
      "token_acc": 0.44648318042813456,
      "train_speed(iter/s)": 1.450512
    },
    {
      "epoch": 0.6079431044085515,
      "grad_norm": 4.650112628936768,
      "learning_rate": 9.639638390805518e-05,
      "loss": 2.593561363220215,
      "memory(GiB)": 58.3,
      "step": 14190,
      "token_acc": 0.4911504424778761,
      "train_speed(iter/s)": 1.450517
    },
    {
      "epoch": 0.6081573197378004,
      "grad_norm": 4.303674697875977,
      "learning_rate": 9.639387489574083e-05,
      "loss": 2.535626220703125,
      "memory(GiB)": 58.3,
      "step": 14195,
      "token_acc": 0.46757679180887374,
      "train_speed(iter/s)": 1.450536
    },
    {
      "epoch": 0.6083715350670494,
      "grad_norm": 4.284163951873779,
      "learning_rate": 9.639136504295792e-05,
      "loss": 2.706220245361328,
      "memory(GiB)": 58.3,
      "step": 14200,
      "token_acc": 0.43670886075949367,
      "train_speed(iter/s)": 1.450551
    },
    {
      "epoch": 0.6085857503962984,
      "grad_norm": 4.55698823928833,
      "learning_rate": 9.638885434975189e-05,
      "loss": 2.5487178802490233,
      "memory(GiB)": 58.3,
      "step": 14205,
      "token_acc": 0.46405228758169936,
      "train_speed(iter/s)": 1.450581
    },
    {
      "epoch": 0.6087999657255473,
      "grad_norm": 4.081040382385254,
      "learning_rate": 9.638634281616827e-05,
      "loss": 2.6076353073120115,
      "memory(GiB)": 58.3,
      "step": 14210,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.450535
    },
    {
      "epoch": 0.6090141810547963,
      "grad_norm": 3.8725972175598145,
      "learning_rate": 9.638383044225252e-05,
      "loss": 2.4013492584228517,
      "memory(GiB)": 58.3,
      "step": 14215,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.450483
    },
    {
      "epoch": 0.6092283963840452,
      "grad_norm": 5.643509387969971,
      "learning_rate": 9.638131722805018e-05,
      "loss": 2.8583192825317383,
      "memory(GiB)": 58.3,
      "step": 14220,
      "token_acc": 0.41964285714285715,
      "train_speed(iter/s)": 1.450534
    },
    {
      "epoch": 0.6094426117132942,
      "grad_norm": 3.9989659786224365,
      "learning_rate": 9.637880317360677e-05,
      "loss": 2.5155548095703124,
      "memory(GiB)": 58.3,
      "step": 14225,
      "token_acc": 0.4356060606060606,
      "train_speed(iter/s)": 1.450504
    },
    {
      "epoch": 0.6096568270425432,
      "grad_norm": 4.759162902832031,
      "learning_rate": 9.637628827896784e-05,
      "loss": 2.6664148330688477,
      "memory(GiB)": 58.3,
      "step": 14230,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.450415
    },
    {
      "epoch": 0.6098710423717921,
      "grad_norm": 4.269876003265381,
      "learning_rate": 9.637377254417892e-05,
      "loss": 2.6349323272705076,
      "memory(GiB)": 58.3,
      "step": 14235,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.4504
    },
    {
      "epoch": 0.610085257701041,
      "grad_norm": 5.584471225738525,
      "learning_rate": 9.637125596928562e-05,
      "loss": 2.5454620361328124,
      "memory(GiB)": 58.3,
      "step": 14240,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.4503
    },
    {
      "epoch": 0.6102994730302901,
      "grad_norm": 3.627065658569336,
      "learning_rate": 9.636873855433353e-05,
      "loss": 2.417140769958496,
      "memory(GiB)": 58.3,
      "step": 14245,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.450325
    },
    {
      "epoch": 0.610513688359539,
      "grad_norm": 3.7807254791259766,
      "learning_rate": 9.636622029936823e-05,
      "loss": 2.7936275482177733,
      "memory(GiB)": 58.3,
      "step": 14250,
      "token_acc": 0.40350877192982454,
      "train_speed(iter/s)": 1.45039
    },
    {
      "epoch": 0.6107279036887879,
      "grad_norm": 5.293404579162598,
      "learning_rate": 9.636370120443536e-05,
      "loss": 2.5364202499389648,
      "memory(GiB)": 58.3,
      "step": 14255,
      "token_acc": 0.4746268656716418,
      "train_speed(iter/s)": 1.450284
    },
    {
      "epoch": 0.610942119018037,
      "grad_norm": 4.4679131507873535,
      "learning_rate": 9.636118126958056e-05,
      "loss": 2.5206331253051757,
      "memory(GiB)": 58.3,
      "step": 14260,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.450263
    },
    {
      "epoch": 0.6111563343472859,
      "grad_norm": 3.8023853302001953,
      "learning_rate": 9.635866049484945e-05,
      "loss": 2.8486602783203123,
      "memory(GiB)": 58.3,
      "step": 14265,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.450305
    },
    {
      "epoch": 0.6113705496765348,
      "grad_norm": 4.473516941070557,
      "learning_rate": 9.635613888028773e-05,
      "loss": 2.6485260009765623,
      "memory(GiB)": 58.3,
      "step": 14270,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.450278
    },
    {
      "epoch": 0.6115847650057838,
      "grad_norm": 3.4229116439819336,
      "learning_rate": 9.635361642594107e-05,
      "loss": 2.694051742553711,
      "memory(GiB)": 58.3,
      "step": 14275,
      "token_acc": 0.42900302114803623,
      "train_speed(iter/s)": 1.450324
    },
    {
      "epoch": 0.6117989803350328,
      "grad_norm": 5.261970043182373,
      "learning_rate": 9.635109313185515e-05,
      "loss": 2.8413217544555662,
      "memory(GiB)": 58.3,
      "step": 14280,
      "token_acc": 0.4208860759493671,
      "train_speed(iter/s)": 1.450424
    },
    {
      "epoch": 0.6120131956642817,
      "grad_norm": 3.982478380203247,
      "learning_rate": 9.634856899807571e-05,
      "loss": 2.7169370651245117,
      "memory(GiB)": 58.3,
      "step": 14285,
      "token_acc": 0.4269005847953216,
      "train_speed(iter/s)": 1.450419
    },
    {
      "epoch": 0.6122274109935307,
      "grad_norm": 5.789243221282959,
      "learning_rate": 9.634604402464846e-05,
      "loss": 2.6715490341186525,
      "memory(GiB)": 58.3,
      "step": 14290,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.450434
    },
    {
      "epoch": 0.6124416263227797,
      "grad_norm": 5.56578254699707,
      "learning_rate": 9.634351821161916e-05,
      "loss": 2.5312240600585936,
      "memory(GiB)": 58.3,
      "step": 14295,
      "token_acc": 0.46557377049180326,
      "train_speed(iter/s)": 1.450463
    },
    {
      "epoch": 0.6126558416520286,
      "grad_norm": 4.066060543060303,
      "learning_rate": 9.634099155903353e-05,
      "loss": 2.700741004943848,
      "memory(GiB)": 58.3,
      "step": 14300,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.450592
    },
    {
      "epoch": 0.6128700569812776,
      "grad_norm": 3.4789788722991943,
      "learning_rate": 9.633846406693738e-05,
      "loss": 2.366909980773926,
      "memory(GiB)": 58.3,
      "step": 14305,
      "token_acc": 0.4793650793650794,
      "train_speed(iter/s)": 1.450625
    },
    {
      "epoch": 0.6130842723105265,
      "grad_norm": 5.285825252532959,
      "learning_rate": 9.633593573537649e-05,
      "loss": 2.700864791870117,
      "memory(GiB)": 58.3,
      "step": 14310,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.450696
    },
    {
      "epoch": 0.6132984876397755,
      "grad_norm": 4.436102867126465,
      "learning_rate": 9.633340656439664e-05,
      "loss": 2.7850835800170897,
      "memory(GiB)": 58.3,
      "step": 14315,
      "token_acc": 0.4540059347181009,
      "train_speed(iter/s)": 1.450669
    },
    {
      "epoch": 0.6135127029690245,
      "grad_norm": 4.019272804260254,
      "learning_rate": 9.633087655404369e-05,
      "loss": 2.4706047058105467,
      "memory(GiB)": 58.3,
      "step": 14320,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.450561
    },
    {
      "epoch": 0.6137269182982734,
      "grad_norm": 4.5024824142456055,
      "learning_rate": 9.632834570436343e-05,
      "loss": 2.671927642822266,
      "memory(GiB)": 58.3,
      "step": 14325,
      "token_acc": 0.4380952380952381,
      "train_speed(iter/s)": 1.450663
    },
    {
      "epoch": 0.6139411336275223,
      "grad_norm": 3.2568390369415283,
      "learning_rate": 9.632581401540174e-05,
      "loss": 2.699192428588867,
      "memory(GiB)": 58.3,
      "step": 14330,
      "token_acc": 0.4506172839506173,
      "train_speed(iter/s)": 1.450665
    },
    {
      "epoch": 0.6141553489567714,
      "grad_norm": 7.013794898986816,
      "learning_rate": 9.632328148720447e-05,
      "loss": 2.422679901123047,
      "memory(GiB)": 58.3,
      "step": 14335,
      "token_acc": 0.5152838427947598,
      "train_speed(iter/s)": 1.450637
    },
    {
      "epoch": 0.6143695642860203,
      "grad_norm": 5.078404903411865,
      "learning_rate": 9.632074811981749e-05,
      "loss": 2.7318899154663088,
      "memory(GiB)": 58.3,
      "step": 14340,
      "token_acc": 0.4405144694533762,
      "train_speed(iter/s)": 1.450648
    },
    {
      "epoch": 0.6145837796152692,
      "grad_norm": 3.7661755084991455,
      "learning_rate": 9.631821391328671e-05,
      "loss": 3.045706367492676,
      "memory(GiB)": 58.3,
      "step": 14345,
      "token_acc": 0.40236686390532544,
      "train_speed(iter/s)": 1.450602
    },
    {
      "epoch": 0.6147979949445183,
      "grad_norm": 3.5946297645568848,
      "learning_rate": 9.631567886765804e-05,
      "loss": 2.706712341308594,
      "memory(GiB)": 58.3,
      "step": 14350,
      "token_acc": 0.48089171974522293,
      "train_speed(iter/s)": 1.450656
    },
    {
      "epoch": 0.6150122102737672,
      "grad_norm": 4.757500171661377,
      "learning_rate": 9.63131429829774e-05,
      "loss": 2.5729793548583983,
      "memory(GiB)": 58.3,
      "step": 14355,
      "token_acc": 0.4576271186440678,
      "train_speed(iter/s)": 1.45078
    },
    {
      "epoch": 0.6152264256030161,
      "grad_norm": 3.7429420948028564,
      "learning_rate": 9.631060625929073e-05,
      "loss": 2.5921089172363283,
      "memory(GiB)": 58.3,
      "step": 14360,
      "token_acc": 0.4492753623188406,
      "train_speed(iter/s)": 1.450923
    },
    {
      "epoch": 0.6154406409322651,
      "grad_norm": 4.863556385040283,
      "learning_rate": 9.630806869664397e-05,
      "loss": 2.179573822021484,
      "memory(GiB)": 58.3,
      "step": 14365,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.45097
    },
    {
      "epoch": 0.6156548562615141,
      "grad_norm": 4.245652198791504,
      "learning_rate": 9.630553029508312e-05,
      "loss": 2.7798912048339846,
      "memory(GiB)": 58.3,
      "step": 14370,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.451016
    },
    {
      "epoch": 0.615869071590763,
      "grad_norm": 6.306260585784912,
      "learning_rate": 9.630299105465414e-05,
      "loss": 2.9054943084716798,
      "memory(GiB)": 58.3,
      "step": 14375,
      "token_acc": 0.4653846153846154,
      "train_speed(iter/s)": 1.450915
    },
    {
      "epoch": 0.616083286920012,
      "grad_norm": 4.311077117919922,
      "learning_rate": 9.630045097540305e-05,
      "loss": 2.494624137878418,
      "memory(GiB)": 58.3,
      "step": 14380,
      "token_acc": 0.4725274725274725,
      "train_speed(iter/s)": 1.450996
    },
    {
      "epoch": 0.6162975022492609,
      "grad_norm": 5.55035400390625,
      "learning_rate": 9.629791005737587e-05,
      "loss": 2.644223213195801,
      "memory(GiB)": 58.3,
      "step": 14385,
      "token_acc": 0.4532871972318339,
      "train_speed(iter/s)": 1.451078
    },
    {
      "epoch": 0.6165117175785099,
      "grad_norm": 5.805172443389893,
      "learning_rate": 9.62953683006186e-05,
      "loss": 2.422890281677246,
      "memory(GiB)": 58.3,
      "step": 14390,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.451204
    },
    {
      "epoch": 0.6167259329077589,
      "grad_norm": 4.4684953689575195,
      "learning_rate": 9.62928257051773e-05,
      "loss": 2.603474426269531,
      "memory(GiB)": 58.3,
      "step": 14395,
      "token_acc": 0.45307443365695793,
      "train_speed(iter/s)": 1.451256
    },
    {
      "epoch": 0.6169401482370078,
      "grad_norm": 4.715904712677002,
      "learning_rate": 9.629028227109806e-05,
      "loss": 2.8599857330322265,
      "memory(GiB)": 58.3,
      "step": 14400,
      "token_acc": 0.436950146627566,
      "train_speed(iter/s)": 1.451093
    },
    {
      "epoch": 0.6171543635662567,
      "grad_norm": 4.722514629364014,
      "learning_rate": 9.628773799842692e-05,
      "loss": 2.5614501953125,
      "memory(GiB)": 58.3,
      "step": 14405,
      "token_acc": 0.48606811145510836,
      "train_speed(iter/s)": 1.451136
    },
    {
      "epoch": 0.6173685788955058,
      "grad_norm": 4.117427349090576,
      "learning_rate": 9.628519288720998e-05,
      "loss": 2.8555744171142576,
      "memory(GiB)": 58.3,
      "step": 14410,
      "token_acc": 0.4268292682926829,
      "train_speed(iter/s)": 1.451169
    },
    {
      "epoch": 0.6175827942247547,
      "grad_norm": 10.364761352539062,
      "learning_rate": 9.628264693749336e-05,
      "loss": 2.869075393676758,
      "memory(GiB)": 58.3,
      "step": 14415,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.451201
    },
    {
      "epoch": 0.6177970095540037,
      "grad_norm": 4.2784552574157715,
      "learning_rate": 9.628010014932315e-05,
      "loss": 2.4816051483154298,
      "memory(GiB)": 58.3,
      "step": 14420,
      "token_acc": 0.4858757062146893,
      "train_speed(iter/s)": 1.451221
    },
    {
      "epoch": 0.6180112248832527,
      "grad_norm": 3.3749730587005615,
      "learning_rate": 9.627755252274555e-05,
      "loss": 2.7301563262939452,
      "memory(GiB)": 58.3,
      "step": 14425,
      "token_acc": 0.4351851851851852,
      "train_speed(iter/s)": 1.451293
    },
    {
      "epoch": 0.6182254402125016,
      "grad_norm": 3.5255680084228516,
      "learning_rate": 9.627500405780664e-05,
      "loss": 2.773752975463867,
      "memory(GiB)": 58.3,
      "step": 14430,
      "token_acc": 0.4602272727272727,
      "train_speed(iter/s)": 1.451264
    },
    {
      "epoch": 0.6184396555417506,
      "grad_norm": 4.291807174682617,
      "learning_rate": 9.627245475455265e-05,
      "loss": 2.408559036254883,
      "memory(GiB)": 58.3,
      "step": 14435,
      "token_acc": 0.45484949832775917,
      "train_speed(iter/s)": 1.45117
    },
    {
      "epoch": 0.6186538708709995,
      "grad_norm": 4.134547710418701,
      "learning_rate": 9.626990461302972e-05,
      "loss": 2.3926017761230467,
      "memory(GiB)": 58.3,
      "step": 14440,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.451235
    },
    {
      "epoch": 0.6188680862002485,
      "grad_norm": 5.719235420227051,
      "learning_rate": 9.626735363328405e-05,
      "loss": 2.5206960678100585,
      "memory(GiB)": 58.3,
      "step": 14445,
      "token_acc": 0.44363636363636366,
      "train_speed(iter/s)": 1.45133
    },
    {
      "epoch": 0.6190823015294975,
      "grad_norm": 4.217067718505859,
      "learning_rate": 9.626480181536188e-05,
      "loss": 2.5698015213012697,
      "memory(GiB)": 58.3,
      "step": 14450,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.451226
    },
    {
      "epoch": 0.6192965168587464,
      "grad_norm": 5.145505428314209,
      "learning_rate": 9.626224915930943e-05,
      "loss": 2.5638763427734377,
      "memory(GiB)": 58.3,
      "step": 14455,
      "token_acc": 0.45051194539249145,
      "train_speed(iter/s)": 1.451161
    },
    {
      "epoch": 0.6195107321879953,
      "grad_norm": 4.336952209472656,
      "learning_rate": 9.625969566517294e-05,
      "loss": 2.7218080520629884,
      "memory(GiB)": 58.3,
      "step": 14460,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.451182
    },
    {
      "epoch": 0.6197249475172444,
      "grad_norm": 3.6626038551330566,
      "learning_rate": 9.625714133299866e-05,
      "loss": 2.8155496597290037,
      "memory(GiB)": 58.3,
      "step": 14465,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.451176
    },
    {
      "epoch": 0.6199391628464933,
      "grad_norm": 6.035238265991211,
      "learning_rate": 9.625458616283289e-05,
      "loss": 2.8277408599853517,
      "memory(GiB)": 58.3,
      "step": 14470,
      "token_acc": 0.4423676012461059,
      "train_speed(iter/s)": 1.451142
    },
    {
      "epoch": 0.6201533781757422,
      "grad_norm": 3.5719058513641357,
      "learning_rate": 9.625203015472189e-05,
      "loss": 2.819818115234375,
      "memory(GiB)": 58.3,
      "step": 14475,
      "token_acc": 0.42388059701492536,
      "train_speed(iter/s)": 1.45106
    },
    {
      "epoch": 0.6203675935049913,
      "grad_norm": 6.20145320892334,
      "learning_rate": 9.624947330871197e-05,
      "loss": 2.3607290267944334,
      "memory(GiB)": 58.3,
      "step": 14480,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.450937
    },
    {
      "epoch": 0.6205818088342402,
      "grad_norm": 4.41830587387085,
      "learning_rate": 9.624691562484945e-05,
      "loss": 2.8208541870117188,
      "memory(GiB)": 58.3,
      "step": 14485,
      "token_acc": 0.4205607476635514,
      "train_speed(iter/s)": 1.451021
    },
    {
      "epoch": 0.6207960241634891,
      "grad_norm": 4.027515888214111,
      "learning_rate": 9.624435710318067e-05,
      "loss": 2.599051666259766,
      "memory(GiB)": 58.3,
      "step": 14490,
      "token_acc": 0.44107744107744107,
      "train_speed(iter/s)": 1.450996
    },
    {
      "epoch": 0.6210102394927381,
      "grad_norm": 4.197312831878662,
      "learning_rate": 9.6241797743752e-05,
      "loss": 2.4655889511108398,
      "memory(GiB)": 58.3,
      "step": 14495,
      "token_acc": 0.4927007299270073,
      "train_speed(iter/s)": 1.451173
    },
    {
      "epoch": 0.6212244548219871,
      "grad_norm": 4.853488922119141,
      "learning_rate": 9.623923754660977e-05,
      "loss": 2.570265197753906,
      "memory(GiB)": 58.3,
      "step": 14500,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.451
    },
    {
      "epoch": 0.6212244548219871,
      "eval_loss": 2.4207069873809814,
      "eval_runtime": 14.1062,
      "eval_samples_per_second": 7.089,
      "eval_steps_per_second": 7.089,
      "eval_token_acc": 0.4683053040103493,
      "step": 14500
    },
    {
      "epoch": 0.621438670151236,
      "grad_norm": 5.254773139953613,
      "learning_rate": 9.623667651180039e-05,
      "loss": 2.7100574493408205,
      "memory(GiB)": 58.3,
      "step": 14505,
      "token_acc": 0.46378174976481656,
      "train_speed(iter/s)": 1.448937
    },
    {
      "epoch": 0.621652885480485,
      "grad_norm": 5.069192409515381,
      "learning_rate": 9.623411463937022e-05,
      "loss": 2.627546501159668,
      "memory(GiB)": 58.3,
      "step": 14510,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.448913
    },
    {
      "epoch": 0.621867100809734,
      "grad_norm": 4.279447078704834,
      "learning_rate": 9.62315519293657e-05,
      "loss": 2.289988327026367,
      "memory(GiB)": 58.3,
      "step": 14515,
      "token_acc": 0.49173553719008267,
      "train_speed(iter/s)": 1.448999
    },
    {
      "epoch": 0.6220813161389829,
      "grad_norm": 4.658717632293701,
      "learning_rate": 9.622898838183325e-05,
      "loss": 2.7342790603637694,
      "memory(GiB)": 58.3,
      "step": 14520,
      "token_acc": 0.45569620253164556,
      "train_speed(iter/s)": 1.449107
    },
    {
      "epoch": 0.6222955314682319,
      "grad_norm": 4.0352606773376465,
      "learning_rate": 9.622642399681931e-05,
      "loss": 2.62386531829834,
      "memory(GiB)": 58.3,
      "step": 14525,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.449165
    },
    {
      "epoch": 0.6225097467974808,
      "grad_norm": 4.412250995635986,
      "learning_rate": 9.622385877437034e-05,
      "loss": 2.4408428192138674,
      "memory(GiB)": 58.3,
      "step": 14530,
      "token_acc": 0.4591194968553459,
      "train_speed(iter/s)": 1.449163
    },
    {
      "epoch": 0.6227239621267298,
      "grad_norm": 4.083387851715088,
      "learning_rate": 9.622129271453281e-05,
      "loss": 2.6369386672973634,
      "memory(GiB)": 58.3,
      "step": 14535,
      "token_acc": 0.4539877300613497,
      "train_speed(iter/s)": 1.449091
    },
    {
      "epoch": 0.6229381774559788,
      "grad_norm": 5.386481285095215,
      "learning_rate": 9.62187258173532e-05,
      "loss": 3.10937614440918,
      "memory(GiB)": 58.3,
      "step": 14540,
      "token_acc": 0.39880952380952384,
      "train_speed(iter/s)": 1.449229
    },
    {
      "epoch": 0.6231523927852277,
      "grad_norm": 3.7416281700134277,
      "learning_rate": 9.621615808287801e-05,
      "loss": 2.6056060791015625,
      "memory(GiB)": 58.3,
      "step": 14545,
      "token_acc": 0.46616541353383456,
      "train_speed(iter/s)": 1.4492
    },
    {
      "epoch": 0.6233666081144766,
      "grad_norm": 5.48376989364624,
      "learning_rate": 9.621358951115377e-05,
      "loss": 2.794524574279785,
      "memory(GiB)": 58.3,
      "step": 14550,
      "token_acc": 0.44368600682593856,
      "train_speed(iter/s)": 1.449036
    },
    {
      "epoch": 0.6235808234437257,
      "grad_norm": 3.7273807525634766,
      "learning_rate": 9.6211020102227e-05,
      "loss": 2.8398351669311523,
      "memory(GiB)": 58.3,
      "step": 14555,
      "token_acc": 0.44368600682593856,
      "train_speed(iter/s)": 1.449018
    },
    {
      "epoch": 0.6237950387729746,
      "grad_norm": 5.19226598739624,
      "learning_rate": 9.620844985614425e-05,
      "loss": 2.773996353149414,
      "memory(GiB)": 58.3,
      "step": 14560,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.449044
    },
    {
      "epoch": 0.6240092541022235,
      "grad_norm": 4.169505596160889,
      "learning_rate": 9.620587877295209e-05,
      "loss": 2.681260108947754,
      "memory(GiB)": 58.3,
      "step": 14565,
      "token_acc": 0.45907473309608543,
      "train_speed(iter/s)": 1.448974
    },
    {
      "epoch": 0.6242234694314726,
      "grad_norm": 3.792816400527954,
      "learning_rate": 9.620330685269707e-05,
      "loss": 2.4349451065063477,
      "memory(GiB)": 58.3,
      "step": 14570,
      "token_acc": 0.453125,
      "train_speed(iter/s)": 1.448966
    },
    {
      "epoch": 0.6244376847607215,
      "grad_norm": 4.641355514526367,
      "learning_rate": 9.620073409542583e-05,
      "loss": 2.322515678405762,
      "memory(GiB)": 58.3,
      "step": 14575,
      "token_acc": 0.5234657039711191,
      "train_speed(iter/s)": 1.449039
    },
    {
      "epoch": 0.6246519000899704,
      "grad_norm": 5.689401149749756,
      "learning_rate": 9.619816050118494e-05,
      "loss": 2.7147125244140624,
      "memory(GiB)": 58.3,
      "step": 14580,
      "token_acc": 0.43548387096774194,
      "train_speed(iter/s)": 1.449087
    },
    {
      "epoch": 0.6248661154192194,
      "grad_norm": 5.070311069488525,
      "learning_rate": 9.619558607002103e-05,
      "loss": 2.847411346435547,
      "memory(GiB)": 58.3,
      "step": 14585,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.449143
    },
    {
      "epoch": 0.6250803307484684,
      "grad_norm": 4.0744524002075195,
      "learning_rate": 9.619301080198075e-05,
      "loss": 2.4808425903320312,
      "memory(GiB)": 58.3,
      "step": 14590,
      "token_acc": 0.5223367697594502,
      "train_speed(iter/s)": 1.449218
    },
    {
      "epoch": 0.6252945460777173,
      "grad_norm": 9.106696128845215,
      "learning_rate": 9.619043469711074e-05,
      "loss": 2.7112491607666014,
      "memory(GiB)": 58.3,
      "step": 14595,
      "token_acc": 0.4819672131147541,
      "train_speed(iter/s)": 1.449326
    },
    {
      "epoch": 0.6255087614069663,
      "grad_norm": 6.736730575561523,
      "learning_rate": 9.618785775545769e-05,
      "loss": 2.572785186767578,
      "memory(GiB)": 58.3,
      "step": 14600,
      "token_acc": 0.43829787234042555,
      "train_speed(iter/s)": 1.449366
    },
    {
      "epoch": 0.6257229767362152,
      "grad_norm": 4.710590362548828,
      "learning_rate": 9.618527997706825e-05,
      "loss": 2.7171817779541017,
      "memory(GiB)": 58.3,
      "step": 14605,
      "token_acc": 0.4584837545126354,
      "train_speed(iter/s)": 1.449359
    },
    {
      "epoch": 0.6259371920654642,
      "grad_norm": 8.038471221923828,
      "learning_rate": 9.618270136198915e-05,
      "loss": 2.5910783767700196,
      "memory(GiB)": 58.3,
      "step": 14610,
      "token_acc": 0.45864661654135336,
      "train_speed(iter/s)": 1.449265
    },
    {
      "epoch": 0.6261514073947132,
      "grad_norm": 3.7803711891174316,
      "learning_rate": 9.61801219102671e-05,
      "loss": 2.5998628616333006,
      "memory(GiB)": 58.3,
      "step": 14615,
      "token_acc": 0.425531914893617,
      "train_speed(iter/s)": 1.449428
    },
    {
      "epoch": 0.6263656227239621,
      "grad_norm": 4.469743728637695,
      "learning_rate": 9.617754162194881e-05,
      "loss": 2.6015525817871095,
      "memory(GiB)": 58.3,
      "step": 14620,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.449515
    },
    {
      "epoch": 0.626579838053211,
      "grad_norm": 5.250475883483887,
      "learning_rate": 9.617496049708103e-05,
      "loss": 2.6085208892822265,
      "memory(GiB)": 58.3,
      "step": 14625,
      "token_acc": 0.46785714285714286,
      "train_speed(iter/s)": 1.449516
    },
    {
      "epoch": 0.6267940533824601,
      "grad_norm": 3.4902091026306152,
      "learning_rate": 9.617237853571054e-05,
      "loss": 2.5155656814575194,
      "memory(GiB)": 58.3,
      "step": 14630,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.449539
    },
    {
      "epoch": 0.627008268711709,
      "grad_norm": 6.242879390716553,
      "learning_rate": 9.61697957378841e-05,
      "loss": 2.553343963623047,
      "memory(GiB)": 58.3,
      "step": 14635,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.449504
    },
    {
      "epoch": 0.6272224840409579,
      "grad_norm": 6.107500076293945,
      "learning_rate": 9.616721210364849e-05,
      "loss": 2.7408821105957033,
      "memory(GiB)": 58.3,
      "step": 14640,
      "token_acc": 0.4318936877076412,
      "train_speed(iter/s)": 1.449604
    },
    {
      "epoch": 0.627436699370207,
      "grad_norm": 5.43466854095459,
      "learning_rate": 9.616462763305052e-05,
      "loss": 2.3727228164672853,
      "memory(GiB)": 58.3,
      "step": 14645,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.449502
    },
    {
      "epoch": 0.6276509146994559,
      "grad_norm": 4.648040771484375,
      "learning_rate": 9.616204232613701e-05,
      "loss": 2.435367965698242,
      "memory(GiB)": 58.3,
      "step": 14650,
      "token_acc": 0.5127388535031847,
      "train_speed(iter/s)": 1.449526
    },
    {
      "epoch": 0.6278651300287048,
      "grad_norm": 4.420108318328857,
      "learning_rate": 9.615945618295483e-05,
      "loss": 2.3956180572509767,
      "memory(GiB)": 58.3,
      "step": 14655,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.449593
    },
    {
      "epoch": 0.6280793453579538,
      "grad_norm": 4.477234840393066,
      "learning_rate": 9.615686920355077e-05,
      "loss": 2.6692153930664064,
      "memory(GiB)": 58.3,
      "step": 14660,
      "token_acc": 0.42729970326409494,
      "train_speed(iter/s)": 1.449737
    },
    {
      "epoch": 0.6282935606872028,
      "grad_norm": 3.919339179992676,
      "learning_rate": 9.615428138797174e-05,
      "loss": 2.6346946716308595,
      "memory(GiB)": 58.3,
      "step": 14665,
      "token_acc": 0.46062992125984253,
      "train_speed(iter/s)": 1.449803
    },
    {
      "epoch": 0.6285077760164517,
      "grad_norm": 4.786696910858154,
      "learning_rate": 9.61516927362646e-05,
      "loss": 2.4836502075195312,
      "memory(GiB)": 58.3,
      "step": 14670,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.449779
    },
    {
      "epoch": 0.6287219913457007,
      "grad_norm": 4.565245151519775,
      "learning_rate": 9.614910324847626e-05,
      "loss": 2.596850776672363,
      "memory(GiB)": 58.3,
      "step": 14675,
      "token_acc": 0.4490566037735849,
      "train_speed(iter/s)": 1.449794
    },
    {
      "epoch": 0.6289362066749496,
      "grad_norm": 3.0799903869628906,
      "learning_rate": 9.614651292465361e-05,
      "loss": 2.465973472595215,
      "memory(GiB)": 58.3,
      "step": 14680,
      "token_acc": 0.5085227272727273,
      "train_speed(iter/s)": 1.449859
    },
    {
      "epoch": 0.6291504220041986,
      "grad_norm": 4.583732604980469,
      "learning_rate": 9.614392176484359e-05,
      "loss": 2.8886692047119142,
      "memory(GiB)": 58.3,
      "step": 14685,
      "token_acc": 0.410958904109589,
      "train_speed(iter/s)": 1.449801
    },
    {
      "epoch": 0.6293646373334476,
      "grad_norm": 3.2853856086730957,
      "learning_rate": 9.614132976909316e-05,
      "loss": 2.834272575378418,
      "memory(GiB)": 58.3,
      "step": 14690,
      "token_acc": 0.42105263157894735,
      "train_speed(iter/s)": 1.449768
    },
    {
      "epoch": 0.6295788526626965,
      "grad_norm": 4.822491645812988,
      "learning_rate": 9.613873693744924e-05,
      "loss": 2.610274314880371,
      "memory(GiB)": 58.3,
      "step": 14695,
      "token_acc": 0.41762452107279696,
      "train_speed(iter/s)": 1.449903
    },
    {
      "epoch": 0.6297930679919455,
      "grad_norm": 4.612882137298584,
      "learning_rate": 9.613614326995882e-05,
      "loss": 2.860098457336426,
      "memory(GiB)": 58.3,
      "step": 14700,
      "token_acc": 0.4253731343283582,
      "train_speed(iter/s)": 1.44996
    },
    {
      "epoch": 0.6300072833211945,
      "grad_norm": 4.541575908660889,
      "learning_rate": 9.613354876666889e-05,
      "loss": 2.702968406677246,
      "memory(GiB)": 58.3,
      "step": 14705,
      "token_acc": 0.44904458598726116,
      "train_speed(iter/s)": 1.450032
    },
    {
      "epoch": 0.6302214986504434,
      "grad_norm": 4.412137031555176,
      "learning_rate": 9.613095342762646e-05,
      "loss": 2.6839962005615234,
      "memory(GiB)": 58.3,
      "step": 14710,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.450089
    },
    {
      "epoch": 0.6304357139796923,
      "grad_norm": 4.365449905395508,
      "learning_rate": 9.612835725287852e-05,
      "loss": 2.8457691192626955,
      "memory(GiB)": 58.3,
      "step": 14715,
      "token_acc": 0.4389438943894389,
      "train_speed(iter/s)": 1.450139
    },
    {
      "epoch": 0.6306499293089414,
      "grad_norm": 4.461371898651123,
      "learning_rate": 9.61257602424721e-05,
      "loss": 2.569447326660156,
      "memory(GiB)": 58.3,
      "step": 14720,
      "token_acc": 0.4368932038834951,
      "train_speed(iter/s)": 1.450082
    },
    {
      "epoch": 0.6308641446381903,
      "grad_norm": 4.058807373046875,
      "learning_rate": 9.612316239645429e-05,
      "loss": 2.6915924072265627,
      "memory(GiB)": 58.3,
      "step": 14725,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.450199
    },
    {
      "epoch": 0.6310783599674392,
      "grad_norm": 3.8894944190979004,
      "learning_rate": 9.612056371487213e-05,
      "loss": 2.5796859741210936,
      "memory(GiB)": 58.3,
      "step": 14730,
      "token_acc": 0.463855421686747,
      "train_speed(iter/s)": 1.450191
    },
    {
      "epoch": 0.6312925752966883,
      "grad_norm": 3.9801249504089355,
      "learning_rate": 9.611796419777266e-05,
      "loss": 2.786220741271973,
      "memory(GiB)": 58.3,
      "step": 14735,
      "token_acc": 0.445578231292517,
      "train_speed(iter/s)": 1.450221
    },
    {
      "epoch": 0.6315067906259372,
      "grad_norm": 3.8651113510131836,
      "learning_rate": 9.611536384520303e-05,
      "loss": 2.436446952819824,
      "memory(GiB)": 58.3,
      "step": 14740,
      "token_acc": 0.4887459807073955,
      "train_speed(iter/s)": 1.450162
    },
    {
      "epoch": 0.6317210059551861,
      "grad_norm": 4.436505317687988,
      "learning_rate": 9.611276265721032e-05,
      "loss": 2.6532703399658204,
      "memory(GiB)": 58.3,
      "step": 14745,
      "token_acc": 0.4601449275362319,
      "train_speed(iter/s)": 1.449949
    },
    {
      "epoch": 0.6319352212844351,
      "grad_norm": 4.747355937957764,
      "learning_rate": 9.611016063384168e-05,
      "loss": 2.555825424194336,
      "memory(GiB)": 58.3,
      "step": 14750,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.449947
    },
    {
      "epoch": 0.6321494366136841,
      "grad_norm": 4.22128963470459,
      "learning_rate": 9.61075577751442e-05,
      "loss": 2.3227397918701174,
      "memory(GiB)": 58.3,
      "step": 14755,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.449919
    },
    {
      "epoch": 0.6323636519429331,
      "grad_norm": 5.194764137268066,
      "learning_rate": 9.610495408116507e-05,
      "loss": 2.728322410583496,
      "memory(GiB)": 58.3,
      "step": 14760,
      "token_acc": 0.42292490118577075,
      "train_speed(iter/s)": 1.449878
    },
    {
      "epoch": 0.632577867272182,
      "grad_norm": 4.795577049255371,
      "learning_rate": 9.610234955195143e-05,
      "loss": 2.3574304580688477,
      "memory(GiB)": 58.3,
      "step": 14765,
      "token_acc": 0.4978723404255319,
      "train_speed(iter/s)": 1.44982
    },
    {
      "epoch": 0.6327920826014309,
      "grad_norm": 5.83280086517334,
      "learning_rate": 9.60997441875505e-05,
      "loss": 2.8128021240234373,
      "memory(GiB)": 58.3,
      "step": 14770,
      "token_acc": 0.4305555555555556,
      "train_speed(iter/s)": 1.449905
    },
    {
      "epoch": 0.63300629793068,
      "grad_norm": 4.12320613861084,
      "learning_rate": 9.609713798800943e-05,
      "loss": 2.672073745727539,
      "memory(GiB)": 58.3,
      "step": 14775,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.44985
    },
    {
      "epoch": 0.6332205132599289,
      "grad_norm": 5.067982196807861,
      "learning_rate": 9.609453095337548e-05,
      "loss": 2.612883758544922,
      "memory(GiB)": 58.3,
      "step": 14780,
      "token_acc": 0.46794871794871795,
      "train_speed(iter/s)": 1.449889
    },
    {
      "epoch": 0.6334347285891778,
      "grad_norm": 4.375309944152832,
      "learning_rate": 9.609192308369588e-05,
      "loss": 2.7582046508789064,
      "memory(GiB)": 58.3,
      "step": 14785,
      "token_acc": 0.46417445482866043,
      "train_speed(iter/s)": 1.449929
    },
    {
      "epoch": 0.6336489439184269,
      "grad_norm": 5.104846954345703,
      "learning_rate": 9.608931437901782e-05,
      "loss": 2.6408689498901365,
      "memory(GiB)": 58.3,
      "step": 14790,
      "token_acc": 0.43564356435643564,
      "train_speed(iter/s)": 1.449887
    },
    {
      "epoch": 0.6338631592476758,
      "grad_norm": 4.74508810043335,
      "learning_rate": 9.608670483938862e-05,
      "loss": 2.599860191345215,
      "memory(GiB)": 58.3,
      "step": 14795,
      "token_acc": 0.4259818731117825,
      "train_speed(iter/s)": 1.449771
    },
    {
      "epoch": 0.6340773745769247,
      "grad_norm": 3.539123773574829,
      "learning_rate": 9.608409446485553e-05,
      "loss": 2.5515140533447265,
      "memory(GiB)": 58.3,
      "step": 14800,
      "token_acc": 0.44573643410852715,
      "train_speed(iter/s)": 1.44983
    },
    {
      "epoch": 0.6342915899061737,
      "grad_norm": 3.3713126182556152,
      "learning_rate": 9.608148325546583e-05,
      "loss": 2.586057662963867,
      "memory(GiB)": 58.3,
      "step": 14805,
      "token_acc": 0.4426751592356688,
      "train_speed(iter/s)": 1.44987
    },
    {
      "epoch": 0.6345058052354227,
      "grad_norm": 4.70167875289917,
      "learning_rate": 9.607887121126684e-05,
      "loss": 2.3151166915893553,
      "memory(GiB)": 58.3,
      "step": 14810,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.449931
    },
    {
      "epoch": 0.6347200205646716,
      "grad_norm": 3.8407154083251953,
      "learning_rate": 9.607625833230586e-05,
      "loss": 2.755192756652832,
      "memory(GiB)": 58.3,
      "step": 14815,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.449956
    },
    {
      "epoch": 0.6349342358939206,
      "grad_norm": 5.744650840759277,
      "learning_rate": 9.607364461863024e-05,
      "loss": 2.551161766052246,
      "memory(GiB)": 58.3,
      "step": 14820,
      "token_acc": 0.5113636363636364,
      "train_speed(iter/s)": 1.449978
    },
    {
      "epoch": 0.6351484512231695,
      "grad_norm": 4.699014663696289,
      "learning_rate": 9.607103007028733e-05,
      "loss": 2.8635089874267576,
      "memory(GiB)": 58.3,
      "step": 14825,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.450109
    },
    {
      "epoch": 0.6353626665524185,
      "grad_norm": 5.925031661987305,
      "learning_rate": 9.606841468732451e-05,
      "loss": 2.715308952331543,
      "memory(GiB)": 58.3,
      "step": 14830,
      "token_acc": 0.4635036496350365,
      "train_speed(iter/s)": 1.450097
    },
    {
      "epoch": 0.6355768818816675,
      "grad_norm": 5.102819919586182,
      "learning_rate": 9.606579846978912e-05,
      "loss": 2.9080530166625977,
      "memory(GiB)": 58.3,
      "step": 14835,
      "token_acc": 0.446875,
      "train_speed(iter/s)": 1.450097
    },
    {
      "epoch": 0.6357910972109164,
      "grad_norm": 3.8110156059265137,
      "learning_rate": 9.606318141772858e-05,
      "loss": 2.618484687805176,
      "memory(GiB)": 58.3,
      "step": 14840,
      "token_acc": 0.43354430379746833,
      "train_speed(iter/s)": 1.45006
    },
    {
      "epoch": 0.6360053125401653,
      "grad_norm": 5.4171366691589355,
      "learning_rate": 9.606056353119031e-05,
      "loss": 2.465500259399414,
      "memory(GiB)": 58.3,
      "step": 14845,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.44985
    },
    {
      "epoch": 0.6362195278694144,
      "grad_norm": 4.653716087341309,
      "learning_rate": 9.60579448102217e-05,
      "loss": 2.785238265991211,
      "memory(GiB)": 58.3,
      "step": 14850,
      "token_acc": 0.4456140350877193,
      "train_speed(iter/s)": 1.449817
    },
    {
      "epoch": 0.6364337431986633,
      "grad_norm": 5.076840877532959,
      "learning_rate": 9.605532525487024e-05,
      "loss": 2.428042411804199,
      "memory(GiB)": 58.3,
      "step": 14855,
      "token_acc": 0.5260869565217391,
      "train_speed(iter/s)": 1.44994
    },
    {
      "epoch": 0.6366479585279122,
      "grad_norm": 4.1579670906066895,
      "learning_rate": 9.605270486518335e-05,
      "loss": 2.4552005767822265,
      "memory(GiB)": 58.3,
      "step": 14860,
      "token_acc": 0.5064102564102564,
      "train_speed(iter/s)": 1.449914
    },
    {
      "epoch": 0.6368621738571613,
      "grad_norm": 4.259786128997803,
      "learning_rate": 9.60500836412085e-05,
      "loss": 2.512075424194336,
      "memory(GiB)": 58.3,
      "step": 14865,
      "token_acc": 0.4492753623188406,
      "train_speed(iter/s)": 1.449956
    },
    {
      "epoch": 0.6370763891864102,
      "grad_norm": 3.9307875633239746,
      "learning_rate": 9.604746158299319e-05,
      "loss": 2.515934181213379,
      "memory(GiB)": 58.3,
      "step": 14870,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.449822
    },
    {
      "epoch": 0.6372906045156591,
      "grad_norm": 4.962164402008057,
      "learning_rate": 9.604483869058492e-05,
      "loss": 2.844198226928711,
      "memory(GiB)": 58.3,
      "step": 14875,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.449927
    },
    {
      "epoch": 0.6375048198449081,
      "grad_norm": 5.517521381378174,
      "learning_rate": 9.604221496403121e-05,
      "loss": 2.4753753662109377,
      "memory(GiB)": 58.3,
      "step": 14880,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.449976
    },
    {
      "epoch": 0.6377190351741571,
      "grad_norm": 4.912667751312256,
      "learning_rate": 9.603959040337958e-05,
      "loss": 2.599565887451172,
      "memory(GiB)": 58.3,
      "step": 14885,
      "token_acc": 0.4143835616438356,
      "train_speed(iter/s)": 1.449923
    },
    {
      "epoch": 0.637933250503406,
      "grad_norm": 4.038173198699951,
      "learning_rate": 9.603696500867758e-05,
      "loss": 2.6866586685180662,
      "memory(GiB)": 58.3,
      "step": 14890,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.449931
    },
    {
      "epoch": 0.638147465832655,
      "grad_norm": 6.335878372192383,
      "learning_rate": 9.603433877997278e-05,
      "loss": 2.87451286315918,
      "memory(GiB)": 58.3,
      "step": 14895,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.45004
    },
    {
      "epoch": 0.638361681161904,
      "grad_norm": 6.428534507751465,
      "learning_rate": 9.603171171731273e-05,
      "loss": 2.958345413208008,
      "memory(GiB)": 58.3,
      "step": 14900,
      "token_acc": 0.46122448979591835,
      "train_speed(iter/s)": 1.450129
    },
    {
      "epoch": 0.6385758964911529,
      "grad_norm": 4.494062900543213,
      "learning_rate": 9.602908382074508e-05,
      "loss": 2.7423254013061524,
      "memory(GiB)": 58.3,
      "step": 14905,
      "token_acc": 0.45427728613569324,
      "train_speed(iter/s)": 1.450149
    },
    {
      "epoch": 0.6387901118204019,
      "grad_norm": 4.378876209259033,
      "learning_rate": 9.602645509031735e-05,
      "loss": 2.493967819213867,
      "memory(GiB)": 58.3,
      "step": 14910,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.450043
    },
    {
      "epoch": 0.6390043271496508,
      "grad_norm": 4.599621772766113,
      "learning_rate": 9.602382552607724e-05,
      "loss": 2.6918758392333983,
      "memory(GiB)": 58.3,
      "step": 14915,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.450114
    },
    {
      "epoch": 0.6392185424788998,
      "grad_norm": 4.914430141448975,
      "learning_rate": 9.602119512807234e-05,
      "loss": 2.548309326171875,
      "memory(GiB)": 58.3,
      "step": 14920,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.450072
    },
    {
      "epoch": 0.6394327578081488,
      "grad_norm": 5.141462326049805,
      "learning_rate": 9.601856389635034e-05,
      "loss": 2.8038703918457033,
      "memory(GiB)": 58.3,
      "step": 14925,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.450165
    },
    {
      "epoch": 0.6396469731373977,
      "grad_norm": 5.20515251159668,
      "learning_rate": 9.601593183095887e-05,
      "loss": 2.6030431747436524,
      "memory(GiB)": 58.3,
      "step": 14930,
      "token_acc": 0.44366197183098594,
      "train_speed(iter/s)": 1.450181
    },
    {
      "epoch": 0.6398611884666466,
      "grad_norm": 5.146335601806641,
      "learning_rate": 9.601329893194563e-05,
      "loss": 2.6439170837402344,
      "memory(GiB)": 58.3,
      "step": 14935,
      "token_acc": 0.4630225080385852,
      "train_speed(iter/s)": 1.450191
    },
    {
      "epoch": 0.6400754037958957,
      "grad_norm": 4.347159385681152,
      "learning_rate": 9.601066519935832e-05,
      "loss": 2.5498104095458984,
      "memory(GiB)": 58.3,
      "step": 14940,
      "token_acc": 0.47115384615384615,
      "train_speed(iter/s)": 1.450194
    },
    {
      "epoch": 0.6402896191251446,
      "grad_norm": 6.115302562713623,
      "learning_rate": 9.600803063324465e-05,
      "loss": 2.3640268325805662,
      "memory(GiB)": 58.3,
      "step": 14945,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.450126
    },
    {
      "epoch": 0.6405038344543935,
      "grad_norm": 5.782168865203857,
      "learning_rate": 9.600539523365234e-05,
      "loss": 2.961846923828125,
      "memory(GiB)": 58.3,
      "step": 14950,
      "token_acc": 0.4331395348837209,
      "train_speed(iter/s)": 1.450112
    },
    {
      "epoch": 0.6407180497836426,
      "grad_norm": 3.8622121810913086,
      "learning_rate": 9.600275900062915e-05,
      "loss": 2.6871658325195313,
      "memory(GiB)": 58.3,
      "step": 14955,
      "token_acc": 0.4438202247191011,
      "train_speed(iter/s)": 1.450199
    },
    {
      "epoch": 0.6409322651128915,
      "grad_norm": 3.9965529441833496,
      "learning_rate": 9.600012193422281e-05,
      "loss": 2.574698066711426,
      "memory(GiB)": 58.3,
      "step": 14960,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.450078
    },
    {
      "epoch": 0.6411464804421404,
      "grad_norm": 3.745694637298584,
      "learning_rate": 9.599748403448113e-05,
      "loss": 2.608061981201172,
      "memory(GiB)": 58.3,
      "step": 14965,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.4499
    },
    {
      "epoch": 0.6413606957713894,
      "grad_norm": 4.3401408195495605,
      "learning_rate": 9.599484530145186e-05,
      "loss": 2.2880584716796877,
      "memory(GiB)": 58.3,
      "step": 14970,
      "token_acc": 0.4837662337662338,
      "train_speed(iter/s)": 1.449876
    },
    {
      "epoch": 0.6415749111006384,
      "grad_norm": 4.041738986968994,
      "learning_rate": 9.599220573518282e-05,
      "loss": 2.6305240631103515,
      "memory(GiB)": 58.3,
      "step": 14975,
      "token_acc": 0.45084745762711864,
      "train_speed(iter/s)": 1.449915
    },
    {
      "epoch": 0.6417891264298873,
      "grad_norm": 5.006702423095703,
      "learning_rate": 9.598956533572183e-05,
      "loss": 2.7863771438598635,
      "memory(GiB)": 58.3,
      "step": 14980,
      "token_acc": 0.4483985765124555,
      "train_speed(iter/s)": 1.449805
    },
    {
      "epoch": 0.6420033417591363,
      "grad_norm": 5.555137634277344,
      "learning_rate": 9.598692410311674e-05,
      "loss": 2.6082275390625,
      "memory(GiB)": 58.3,
      "step": 14985,
      "token_acc": 0.4746376811594203,
      "train_speed(iter/s)": 1.449877
    },
    {
      "epoch": 0.6422175570883852,
      "grad_norm": 4.539990425109863,
      "learning_rate": 9.598428203741536e-05,
      "loss": 2.6776247024536133,
      "memory(GiB)": 58.3,
      "step": 14990,
      "token_acc": 0.44525547445255476,
      "train_speed(iter/s)": 1.449895
    },
    {
      "epoch": 0.6424317724176342,
      "grad_norm": 4.6724853515625,
      "learning_rate": 9.598163913866558e-05,
      "loss": 2.6252384185791016,
      "memory(GiB)": 58.3,
      "step": 14995,
      "token_acc": 0.49466192170818507,
      "train_speed(iter/s)": 1.450055
    },
    {
      "epoch": 0.6426459877468832,
      "grad_norm": 5.745842933654785,
      "learning_rate": 9.597899540691527e-05,
      "loss": 2.830762481689453,
      "memory(GiB)": 58.3,
      "step": 15000,
      "token_acc": 0.46254071661237783,
      "train_speed(iter/s)": 1.450112
    },
    {
      "epoch": 0.6426459877468832,
      "eval_loss": 2.21223521232605,
      "eval_runtime": 13.5047,
      "eval_samples_per_second": 7.405,
      "eval_steps_per_second": 7.405,
      "eval_token_acc": 0.48322147651006714,
      "step": 15000
    },
    {
      "epoch": 0.6428602030761321,
      "grad_norm": 5.049619674682617,
      "learning_rate": 9.597635084221234e-05,
      "loss": 2.7212646484375,
      "memory(GiB)": 58.3,
      "step": 15005,
      "token_acc": 0.47092469018112487,
      "train_speed(iter/s)": 1.448171
    },
    {
      "epoch": 0.643074418405381,
      "grad_norm": 4.307399749755859,
      "learning_rate": 9.597370544460467e-05,
      "loss": 2.6538330078125,
      "memory(GiB)": 58.3,
      "step": 15010,
      "token_acc": 0.4384057971014493,
      "train_speed(iter/s)": 1.448084
    },
    {
      "epoch": 0.6432886337346301,
      "grad_norm": 4.235497951507568,
      "learning_rate": 9.59710592141402e-05,
      "loss": 2.7035709381103517,
      "memory(GiB)": 58.3,
      "step": 15015,
      "token_acc": 0.45151515151515154,
      "train_speed(iter/s)": 1.448238
    },
    {
      "epoch": 0.643502849063879,
      "grad_norm": 6.353463649749756,
      "learning_rate": 9.596841215086689e-05,
      "loss": 2.3623008728027344,
      "memory(GiB)": 58.3,
      "step": 15020,
      "token_acc": 0.5127272727272727,
      "train_speed(iter/s)": 1.448341
    },
    {
      "epoch": 0.6437170643931279,
      "grad_norm": 8.712040901184082,
      "learning_rate": 9.596576425483264e-05,
      "loss": 2.5878549575805665,
      "memory(GiB)": 58.3,
      "step": 15025,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.448338
    },
    {
      "epoch": 0.643931279722377,
      "grad_norm": 4.22176456451416,
      "learning_rate": 9.596311552608547e-05,
      "loss": 2.953642654418945,
      "memory(GiB)": 58.3,
      "step": 15030,
      "token_acc": 0.44242424242424244,
      "train_speed(iter/s)": 1.448318
    },
    {
      "epoch": 0.6441454950516259,
      "grad_norm": 4.134119987487793,
      "learning_rate": 9.596046596467334e-05,
      "loss": 2.391531562805176,
      "memory(GiB)": 58.3,
      "step": 15035,
      "token_acc": 0.5073529411764706,
      "train_speed(iter/s)": 1.448516
    },
    {
      "epoch": 0.6443597103808748,
      "grad_norm": 3.641388177871704,
      "learning_rate": 9.595781557064427e-05,
      "loss": 2.477495574951172,
      "memory(GiB)": 58.3,
      "step": 15040,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.4485
    },
    {
      "epoch": 0.6445739257101238,
      "grad_norm": 7.13884162902832,
      "learning_rate": 9.595516434404624e-05,
      "loss": 2.774261474609375,
      "memory(GiB)": 58.3,
      "step": 15045,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.448479
    },
    {
      "epoch": 0.6447881410393728,
      "grad_norm": 3.5477068424224854,
      "learning_rate": 9.59525122849273e-05,
      "loss": 2.663995552062988,
      "memory(GiB)": 58.3,
      "step": 15050,
      "token_acc": 0.4753623188405797,
      "train_speed(iter/s)": 1.448425
    },
    {
      "epoch": 0.6450023563686217,
      "grad_norm": 4.730018615722656,
      "learning_rate": 9.594985939333549e-05,
      "loss": 2.543264389038086,
      "memory(GiB)": 58.3,
      "step": 15055,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.448499
    },
    {
      "epoch": 0.6452165716978707,
      "grad_norm": 4.0632195472717285,
      "learning_rate": 9.594720566931889e-05,
      "loss": 2.618559455871582,
      "memory(GiB)": 58.3,
      "step": 15060,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.448317
    },
    {
      "epoch": 0.6454307870271196,
      "grad_norm": 3.4372284412384033,
      "learning_rate": 9.594455111292555e-05,
      "loss": 2.6636754989624025,
      "memory(GiB)": 58.3,
      "step": 15065,
      "token_acc": 0.4296028880866426,
      "train_speed(iter/s)": 1.448291
    },
    {
      "epoch": 0.6456450023563686,
      "grad_norm": 4.137977123260498,
      "learning_rate": 9.594189572420356e-05,
      "loss": 2.300486183166504,
      "memory(GiB)": 58.3,
      "step": 15070,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.448254
    },
    {
      "epoch": 0.6458592176856176,
      "grad_norm": 4.151328086853027,
      "learning_rate": 9.593923950320104e-05,
      "loss": 2.7640867233276367,
      "memory(GiB)": 58.3,
      "step": 15075,
      "token_acc": 0.4405594405594406,
      "train_speed(iter/s)": 1.448279
    },
    {
      "epoch": 0.6460734330148665,
      "grad_norm": 4.091307163238525,
      "learning_rate": 9.593658244996609e-05,
      "loss": 2.4460733413696287,
      "memory(GiB)": 58.3,
      "step": 15080,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.448334
    },
    {
      "epoch": 0.6462876483441155,
      "grad_norm": 4.356914043426514,
      "learning_rate": 9.593392456454686e-05,
      "loss": 2.548112487792969,
      "memory(GiB)": 58.3,
      "step": 15085,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.44833
    },
    {
      "epoch": 0.6465018636733645,
      "grad_norm": 4.232750415802002,
      "learning_rate": 9.59312658469915e-05,
      "loss": 2.892285919189453,
      "memory(GiB)": 58.3,
      "step": 15090,
      "token_acc": 0.4280821917808219,
      "train_speed(iter/s)": 1.448328
    },
    {
      "epoch": 0.6467160790026134,
      "grad_norm": 4.230464458465576,
      "learning_rate": 9.592860629734819e-05,
      "loss": 2.692633628845215,
      "memory(GiB)": 58.3,
      "step": 15095,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.448443
    },
    {
      "epoch": 0.6469302943318624,
      "grad_norm": 5.196896553039551,
      "learning_rate": 9.592594591566508e-05,
      "loss": 2.826826477050781,
      "memory(GiB)": 58.3,
      "step": 15100,
      "token_acc": 0.4296028880866426,
      "train_speed(iter/s)": 1.448516
    },
    {
      "epoch": 0.6471445096611114,
      "grad_norm": 9.97440242767334,
      "learning_rate": 9.592328470199037e-05,
      "loss": 2.8210947036743166,
      "memory(GiB)": 58.3,
      "step": 15105,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.448551
    },
    {
      "epoch": 0.6473587249903603,
      "grad_norm": 4.933980941772461,
      "learning_rate": 9.592062265637227e-05,
      "loss": 2.51159725189209,
      "memory(GiB)": 58.3,
      "step": 15110,
      "token_acc": 0.46788990825688076,
      "train_speed(iter/s)": 1.448651
    },
    {
      "epoch": 0.6475729403196093,
      "grad_norm": 5.6142497062683105,
      "learning_rate": 9.591795977885903e-05,
      "loss": 2.496227264404297,
      "memory(GiB)": 58.3,
      "step": 15115,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.448722
    },
    {
      "epoch": 0.6477871556488582,
      "grad_norm": 6.159660339355469,
      "learning_rate": 9.591529606949887e-05,
      "loss": 3.0869821548461913,
      "memory(GiB)": 58.3,
      "step": 15120,
      "token_acc": 0.39492753623188404,
      "train_speed(iter/s)": 1.448821
    },
    {
      "epoch": 0.6480013709781072,
      "grad_norm": 6.724240779876709,
      "learning_rate": 9.591263152834005e-05,
      "loss": 2.614556312561035,
      "memory(GiB)": 58.3,
      "step": 15125,
      "token_acc": 0.452755905511811,
      "train_speed(iter/s)": 1.448823
    },
    {
      "epoch": 0.6482155863073562,
      "grad_norm": 4.72502326965332,
      "learning_rate": 9.590996615543084e-05,
      "loss": 2.723069953918457,
      "memory(GiB)": 58.3,
      "step": 15130,
      "token_acc": 0.44333333333333336,
      "train_speed(iter/s)": 1.448852
    },
    {
      "epoch": 0.6484298016366051,
      "grad_norm": 4.011580467224121,
      "learning_rate": 9.590729995081953e-05,
      "loss": 2.35583438873291,
      "memory(GiB)": 58.3,
      "step": 15135,
      "token_acc": 0.5017064846416383,
      "train_speed(iter/s)": 1.4489
    },
    {
      "epoch": 0.648644016965854,
      "grad_norm": 3.4947268962860107,
      "learning_rate": 9.590463291455442e-05,
      "loss": 2.531879425048828,
      "memory(GiB)": 58.3,
      "step": 15140,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.448927
    },
    {
      "epoch": 0.6488582322951031,
      "grad_norm": 4.407508373260498,
      "learning_rate": 9.59019650466838e-05,
      "loss": 2.477404022216797,
      "memory(GiB)": 58.3,
      "step": 15145,
      "token_acc": 0.5127272727272727,
      "train_speed(iter/s)": 1.449022
    },
    {
      "epoch": 0.649072447624352,
      "grad_norm": 3.6686882972717285,
      "learning_rate": 9.589929634725605e-05,
      "loss": 2.609175109863281,
      "memory(GiB)": 58.3,
      "step": 15150,
      "token_acc": 0.4743202416918429,
      "train_speed(iter/s)": 1.449006
    },
    {
      "epoch": 0.6492866629536009,
      "grad_norm": 5.289790630340576,
      "learning_rate": 9.589662681631948e-05,
      "loss": 2.7299274444580077,
      "memory(GiB)": 58.3,
      "step": 15155,
      "token_acc": 0.42402826855123676,
      "train_speed(iter/s)": 1.448868
    },
    {
      "epoch": 0.64950087828285,
      "grad_norm": 6.507251262664795,
      "learning_rate": 9.589395645392245e-05,
      "loss": 2.722701072692871,
      "memory(GiB)": 58.3,
      "step": 15160,
      "token_acc": 0.4263565891472868,
      "train_speed(iter/s)": 1.448838
    },
    {
      "epoch": 0.6497150936120989,
      "grad_norm": 4.34958028793335,
      "learning_rate": 9.589128526011336e-05,
      "loss": 2.349719429016113,
      "memory(GiB)": 58.3,
      "step": 15165,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.448831
    },
    {
      "epoch": 0.6499293089413478,
      "grad_norm": 3.771009922027588,
      "learning_rate": 9.588861323494058e-05,
      "loss": 2.47757511138916,
      "memory(GiB)": 58.3,
      "step": 15170,
      "token_acc": 0.45058139534883723,
      "train_speed(iter/s)": 1.448901
    },
    {
      "epoch": 0.6501435242705969,
      "grad_norm": 3.5361275672912598,
      "learning_rate": 9.588594037845254e-05,
      "loss": 2.5539297103881835,
      "memory(GiB)": 58.3,
      "step": 15175,
      "token_acc": 0.47878787878787876,
      "train_speed(iter/s)": 1.448977
    },
    {
      "epoch": 0.6503577395998458,
      "grad_norm": 4.314279556274414,
      "learning_rate": 9.588326669069763e-05,
      "loss": 2.497235107421875,
      "memory(GiB)": 58.3,
      "step": 15180,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.449097
    },
    {
      "epoch": 0.6505719549290947,
      "grad_norm": 9.188122749328613,
      "learning_rate": 9.588059217172432e-05,
      "loss": 2.6258724212646483,
      "memory(GiB)": 58.3,
      "step": 15185,
      "token_acc": 0.45625,
      "train_speed(iter/s)": 1.449022
    },
    {
      "epoch": 0.6507861702583437,
      "grad_norm": 4.531054973602295,
      "learning_rate": 9.587791682158102e-05,
      "loss": 2.5726291656494142,
      "memory(GiB)": 58.3,
      "step": 15190,
      "token_acc": 0.4820846905537459,
      "train_speed(iter/s)": 1.449143
    },
    {
      "epoch": 0.6510003855875927,
      "grad_norm": 4.9413604736328125,
      "learning_rate": 9.587524064031624e-05,
      "loss": 2.427939796447754,
      "memory(GiB)": 58.3,
      "step": 15195,
      "token_acc": 0.5181159420289855,
      "train_speed(iter/s)": 1.449209
    },
    {
      "epoch": 0.6512146009168416,
      "grad_norm": 5.501973628997803,
      "learning_rate": 9.587256362797842e-05,
      "loss": 2.771798515319824,
      "memory(GiB)": 58.3,
      "step": 15200,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.449229
    },
    {
      "epoch": 0.6514288162460906,
      "grad_norm": 3.9437625408172607,
      "learning_rate": 9.586988578461609e-05,
      "loss": 2.7026594161987303,
      "memory(GiB)": 58.3,
      "step": 15205,
      "token_acc": 0.4271523178807947,
      "train_speed(iter/s)": 1.449269
    },
    {
      "epoch": 0.6516430315753395,
      "grad_norm": 4.041426658630371,
      "learning_rate": 9.586720711027775e-05,
      "loss": 2.468884086608887,
      "memory(GiB)": 58.3,
      "step": 15210,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.449257
    },
    {
      "epoch": 0.6518572469045885,
      "grad_norm": 3.4535014629364014,
      "learning_rate": 9.586452760501193e-05,
      "loss": 2.4665542602539063,
      "memory(GiB)": 58.3,
      "step": 15215,
      "token_acc": 0.4745222929936306,
      "train_speed(iter/s)": 1.449229
    },
    {
      "epoch": 0.6520714622338375,
      "grad_norm": 5.335357666015625,
      "learning_rate": 9.586184726886715e-05,
      "loss": 2.476365661621094,
      "memory(GiB)": 58.3,
      "step": 15220,
      "token_acc": 0.5182186234817814,
      "train_speed(iter/s)": 1.449209
    },
    {
      "epoch": 0.6522856775630864,
      "grad_norm": 3.933119535446167,
      "learning_rate": 9.5859166101892e-05,
      "loss": 2.5218833923339843,
      "memory(GiB)": 58.3,
      "step": 15225,
      "token_acc": 0.4186046511627907,
      "train_speed(iter/s)": 1.449328
    },
    {
      "epoch": 0.6524998928923353,
      "grad_norm": 5.322347164154053,
      "learning_rate": 9.585648410413503e-05,
      "loss": 2.808527183532715,
      "memory(GiB)": 58.3,
      "step": 15230,
      "token_acc": 0.4392857142857143,
      "train_speed(iter/s)": 1.449424
    },
    {
      "epoch": 0.6527141082215844,
      "grad_norm": 4.473291873931885,
      "learning_rate": 9.585380127564484e-05,
      "loss": 2.8380367279052736,
      "memory(GiB)": 58.3,
      "step": 15235,
      "token_acc": 0.45396825396825397,
      "train_speed(iter/s)": 1.449552
    },
    {
      "epoch": 0.6529283235508333,
      "grad_norm": 5.80584192276001,
      "learning_rate": 9.585111761647002e-05,
      "loss": 2.588529586791992,
      "memory(GiB)": 58.3,
      "step": 15240,
      "token_acc": 0.45864661654135336,
      "train_speed(iter/s)": 1.449591
    },
    {
      "epoch": 0.6531425388800822,
      "grad_norm": 3.6384706497192383,
      "learning_rate": 9.58484331266592e-05,
      "loss": 2.661268615722656,
      "memory(GiB)": 58.3,
      "step": 15245,
      "token_acc": 0.4440894568690096,
      "train_speed(iter/s)": 1.449575
    },
    {
      "epoch": 0.6533567542093313,
      "grad_norm": 3.9359335899353027,
      "learning_rate": 9.5845747806261e-05,
      "loss": 2.6027481079101564,
      "memory(GiB)": 58.3,
      "step": 15250,
      "token_acc": 0.4472843450479233,
      "train_speed(iter/s)": 1.449665
    },
    {
      "epoch": 0.6535709695385802,
      "grad_norm": 3.7551069259643555,
      "learning_rate": 9.584306165532406e-05,
      "loss": 2.597681427001953,
      "memory(GiB)": 58.3,
      "step": 15255,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.449719
    },
    {
      "epoch": 0.6537851848678291,
      "grad_norm": 4.248190879821777,
      "learning_rate": 9.584037467389708e-05,
      "loss": 2.6357345581054688,
      "memory(GiB)": 58.3,
      "step": 15260,
      "token_acc": 0.42142857142857143,
      "train_speed(iter/s)": 1.449625
    },
    {
      "epoch": 0.6539994001970781,
      "grad_norm": 4.554111480712891,
      "learning_rate": 9.583768686202869e-05,
      "loss": 2.529732894897461,
      "memory(GiB)": 58.3,
      "step": 15265,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.449628
    },
    {
      "epoch": 0.6542136155263271,
      "grad_norm": 7.0177130699157715,
      "learning_rate": 9.583499821976762e-05,
      "loss": 2.3018617630004883,
      "memory(GiB)": 58.3,
      "step": 15270,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.44969
    },
    {
      "epoch": 0.654427830855576,
      "grad_norm": 3.57904314994812,
      "learning_rate": 9.583230874716253e-05,
      "loss": 2.4643199920654295,
      "memory(GiB)": 58.3,
      "step": 15275,
      "token_acc": 0.49337748344370863,
      "train_speed(iter/s)": 1.449541
    },
    {
      "epoch": 0.654642046184825,
      "grad_norm": 4.028859615325928,
      "learning_rate": 9.582961844426221e-05,
      "loss": 2.6705575942993165,
      "memory(GiB)": 58.3,
      "step": 15280,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.449535
    },
    {
      "epoch": 0.654856261514074,
      "grad_norm": 4.198547840118408,
      "learning_rate": 9.582692731111535e-05,
      "loss": 2.5124027252197267,
      "memory(GiB)": 58.3,
      "step": 15285,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.449683
    },
    {
      "epoch": 0.6550704768433229,
      "grad_norm": 5.1375532150268555,
      "learning_rate": 9.58242353477707e-05,
      "loss": 2.7559852600097656,
      "memory(GiB)": 58.3,
      "step": 15290,
      "token_acc": 0.42990654205607476,
      "train_speed(iter/s)": 1.449703
    },
    {
      "epoch": 0.6552846921725719,
      "grad_norm": 4.5739850997924805,
      "learning_rate": 9.582154255427705e-05,
      "loss": 2.802250289916992,
      "memory(GiB)": 58.3,
      "step": 15295,
      "token_acc": 0.47540983606557374,
      "train_speed(iter/s)": 1.449707
    },
    {
      "epoch": 0.6554989075018208,
      "grad_norm": 3.3840103149414062,
      "learning_rate": 9.581884893068315e-05,
      "loss": 2.700469398498535,
      "memory(GiB)": 58.3,
      "step": 15300,
      "token_acc": 0.4858757062146893,
      "train_speed(iter/s)": 1.449663
    },
    {
      "epoch": 0.6557131228310698,
      "grad_norm": 3.8742876052856445,
      "learning_rate": 9.581615447703784e-05,
      "loss": 2.60408878326416,
      "memory(GiB)": 58.3,
      "step": 15305,
      "token_acc": 0.45185185185185184,
      "train_speed(iter/s)": 1.449727
    },
    {
      "epoch": 0.6559273381603188,
      "grad_norm": 4.047839641571045,
      "learning_rate": 9.581345919338994e-05,
      "loss": 2.711720275878906,
      "memory(GiB)": 58.3,
      "step": 15310,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.449896
    },
    {
      "epoch": 0.6561415534895677,
      "grad_norm": 6.770996570587158,
      "learning_rate": 9.58107630797882e-05,
      "loss": 2.7397390365600587,
      "memory(GiB)": 58.3,
      "step": 15315,
      "token_acc": 0.45126353790613716,
      "train_speed(iter/s)": 1.449886
    },
    {
      "epoch": 0.6563557688188166,
      "grad_norm": 3.7116146087646484,
      "learning_rate": 9.580806613628155e-05,
      "loss": 2.1243192672729494,
      "memory(GiB)": 58.3,
      "step": 15320,
      "token_acc": 0.5271317829457365,
      "train_speed(iter/s)": 1.449861
    },
    {
      "epoch": 0.6565699841480657,
      "grad_norm": 4.117305278778076,
      "learning_rate": 9.580536836291878e-05,
      "loss": 2.8824741363525392,
      "memory(GiB)": 58.3,
      "step": 15325,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.449924
    },
    {
      "epoch": 0.6567841994773146,
      "grad_norm": 4.3368144035339355,
      "learning_rate": 9.58026697597488e-05,
      "loss": 2.6865615844726562,
      "memory(GiB)": 58.3,
      "step": 15330,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.449948
    },
    {
      "epoch": 0.6569984148065635,
      "grad_norm": 4.002928256988525,
      "learning_rate": 9.579997032682052e-05,
      "loss": 2.4928972244262697,
      "memory(GiB)": 58.3,
      "step": 15335,
      "token_acc": 0.459375,
      "train_speed(iter/s)": 1.450055
    },
    {
      "epoch": 0.6572126301358125,
      "grad_norm": 5.437620162963867,
      "learning_rate": 9.579727006418279e-05,
      "loss": 2.749365043640137,
      "memory(GiB)": 58.3,
      "step": 15340,
      "token_acc": 0.45723684210526316,
      "train_speed(iter/s)": 1.450044
    },
    {
      "epoch": 0.6574268454650615,
      "grad_norm": 4.67940616607666,
      "learning_rate": 9.579456897188455e-05,
      "loss": 2.677341651916504,
      "memory(GiB)": 58.3,
      "step": 15345,
      "token_acc": 0.43462897526501765,
      "train_speed(iter/s)": 1.450066
    },
    {
      "epoch": 0.6576410607943104,
      "grad_norm": 5.337660312652588,
      "learning_rate": 9.579186704997474e-05,
      "loss": 2.707118797302246,
      "memory(GiB)": 58.3,
      "step": 15350,
      "token_acc": 0.4539249146757679,
      "train_speed(iter/s)": 1.450126
    },
    {
      "epoch": 0.6578552761235594,
      "grad_norm": 4.947908878326416,
      "learning_rate": 9.57891642985023e-05,
      "loss": 2.550575065612793,
      "memory(GiB)": 58.3,
      "step": 15355,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.450199
    },
    {
      "epoch": 0.6580694914528084,
      "grad_norm": 3.9470489025115967,
      "learning_rate": 9.57864607175162e-05,
      "loss": 2.4807998657226564,
      "memory(GiB)": 58.3,
      "step": 15360,
      "token_acc": 0.4610169491525424,
      "train_speed(iter/s)": 1.45009
    },
    {
      "epoch": 0.6582837067820573,
      "grad_norm": 3.980710506439209,
      "learning_rate": 9.57837563070654e-05,
      "loss": 2.661483383178711,
      "memory(GiB)": 58.3,
      "step": 15365,
      "token_acc": 0.4486301369863014,
      "train_speed(iter/s)": 1.450136
    },
    {
      "epoch": 0.6584979221113063,
      "grad_norm": 5.150298595428467,
      "learning_rate": 9.578105106719893e-05,
      "loss": 2.567930221557617,
      "memory(GiB)": 58.3,
      "step": 15370,
      "token_acc": 0.4731182795698925,
      "train_speed(iter/s)": 1.450088
    },
    {
      "epoch": 0.6587121374405552,
      "grad_norm": 4.133373737335205,
      "learning_rate": 9.577834499796575e-05,
      "loss": 2.7865345001220705,
      "memory(GiB)": 58.3,
      "step": 15375,
      "token_acc": 0.48264984227129337,
      "train_speed(iter/s)": 1.450218
    },
    {
      "epoch": 0.6589263527698042,
      "grad_norm": 3.8452861309051514,
      "learning_rate": 9.577563809941492e-05,
      "loss": 2.725305938720703,
      "memory(GiB)": 58.3,
      "step": 15380,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.450212
    },
    {
      "epoch": 0.6591405680990532,
      "grad_norm": 3.7102482318878174,
      "learning_rate": 9.577293037159544e-05,
      "loss": 2.448910140991211,
      "memory(GiB)": 58.3,
      "step": 15385,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.450242
    },
    {
      "epoch": 0.6593547834283021,
      "grad_norm": 5.404450416564941,
      "learning_rate": 9.577022181455641e-05,
      "loss": 2.9752422332763673,
      "memory(GiB)": 58.3,
      "step": 15390,
      "token_acc": 0.4380952380952381,
      "train_speed(iter/s)": 1.450391
    },
    {
      "epoch": 0.659568998757551,
      "grad_norm": 7.460755348205566,
      "learning_rate": 9.57675124283469e-05,
      "loss": 3.09713134765625,
      "memory(GiB)": 58.3,
      "step": 15395,
      "token_acc": 0.42136498516320475,
      "train_speed(iter/s)": 1.450439
    },
    {
      "epoch": 0.6597832140868001,
      "grad_norm": 4.8870673179626465,
      "learning_rate": 9.576480221301593e-05,
      "loss": 2.526996040344238,
      "memory(GiB)": 58.3,
      "step": 15400,
      "token_acc": 0.4819277108433735,
      "train_speed(iter/s)": 1.450396
    },
    {
      "epoch": 0.659997429416049,
      "grad_norm": 5.4269938468933105,
      "learning_rate": 9.576209116861265e-05,
      "loss": 2.477591705322266,
      "memory(GiB)": 58.3,
      "step": 15405,
      "token_acc": 0.45229681978798586,
      "train_speed(iter/s)": 1.450455
    },
    {
      "epoch": 0.6602116447452979,
      "grad_norm": 4.405874252319336,
      "learning_rate": 9.575937929518616e-05,
      "loss": 2.419519233703613,
      "memory(GiB)": 58.3,
      "step": 15410,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.450492
    },
    {
      "epoch": 0.660425860074547,
      "grad_norm": 3.6095001697540283,
      "learning_rate": 9.575666659278559e-05,
      "loss": 2.7280147552490233,
      "memory(GiB)": 58.3,
      "step": 15415,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.450433
    },
    {
      "epoch": 0.6606400754037959,
      "grad_norm": 4.082696437835693,
      "learning_rate": 9.575395306146008e-05,
      "loss": 2.7314144134521485,
      "memory(GiB)": 58.3,
      "step": 15420,
      "token_acc": 0.4676923076923077,
      "train_speed(iter/s)": 1.450416
    },
    {
      "epoch": 0.6608542907330448,
      "grad_norm": 3.2907700538635254,
      "learning_rate": 9.575123870125879e-05,
      "loss": 2.544133758544922,
      "memory(GiB)": 58.3,
      "step": 15425,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.450361
    },
    {
      "epoch": 0.6610685060622938,
      "grad_norm": 4.113558769226074,
      "learning_rate": 9.574852351223089e-05,
      "loss": 2.8627801895141602,
      "memory(GiB)": 58.3,
      "step": 15430,
      "token_acc": 0.42934782608695654,
      "train_speed(iter/s)": 1.450241
    },
    {
      "epoch": 0.6612827213915428,
      "grad_norm": 5.359152317047119,
      "learning_rate": 9.574580749442557e-05,
      "loss": 2.5982944488525392,
      "memory(GiB)": 58.3,
      "step": 15435,
      "token_acc": 0.49794238683127573,
      "train_speed(iter/s)": 1.450292
    },
    {
      "epoch": 0.6614969367207918,
      "grad_norm": 3.745368242263794,
      "learning_rate": 9.574309064789205e-05,
      "loss": 2.4612525939941405,
      "memory(GiB)": 58.3,
      "step": 15440,
      "token_acc": 0.4808510638297872,
      "train_speed(iter/s)": 1.450211
    },
    {
      "epoch": 0.6617111520500407,
      "grad_norm": 5.729100227355957,
      "learning_rate": 9.57403729726795e-05,
      "loss": 2.8477020263671875,
      "memory(GiB)": 58.3,
      "step": 15445,
      "token_acc": 0.4117647058823529,
      "train_speed(iter/s)": 1.450283
    },
    {
      "epoch": 0.6619253673792896,
      "grad_norm": 3.986607074737549,
      "learning_rate": 9.573765446883722e-05,
      "loss": 2.642181396484375,
      "memory(GiB)": 58.3,
      "step": 15450,
      "token_acc": 0.4520123839009288,
      "train_speed(iter/s)": 1.450259
    },
    {
      "epoch": 0.6621395827085387,
      "grad_norm": 4.759144306182861,
      "learning_rate": 9.57349351364144e-05,
      "loss": 2.9675331115722656,
      "memory(GiB)": 58.3,
      "step": 15455,
      "token_acc": 0.43508771929824563,
      "train_speed(iter/s)": 1.45029
    },
    {
      "epoch": 0.6623537980377876,
      "grad_norm": 3.890176296234131,
      "learning_rate": 9.573221497546035e-05,
      "loss": 2.729171562194824,
      "memory(GiB)": 58.3,
      "step": 15460,
      "token_acc": 0.42948717948717946,
      "train_speed(iter/s)": 1.450391
    },
    {
      "epoch": 0.6625680133670365,
      "grad_norm": 3.833308458328247,
      "learning_rate": 9.57294939860243e-05,
      "loss": 2.7343345642089845,
      "memory(GiB)": 58.3,
      "step": 15465,
      "token_acc": 0.4409937888198758,
      "train_speed(iter/s)": 1.450206
    },
    {
      "epoch": 0.6627822286962856,
      "grad_norm": 5.782052993774414,
      "learning_rate": 9.572677216815559e-05,
      "loss": 2.8174747467041015,
      "memory(GiB)": 58.3,
      "step": 15470,
      "token_acc": 0.4420289855072464,
      "train_speed(iter/s)": 1.450226
    },
    {
      "epoch": 0.6629964440255345,
      "grad_norm": 4.530306339263916,
      "learning_rate": 9.572404952190349e-05,
      "loss": 2.4785350799560546,
      "memory(GiB)": 58.3,
      "step": 15475,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.450288
    },
    {
      "epoch": 0.6632106593547834,
      "grad_norm": 3.7064712047576904,
      "learning_rate": 9.572132604731735e-05,
      "loss": 2.698088836669922,
      "memory(GiB)": 58.3,
      "step": 15480,
      "token_acc": 0.42933333333333334,
      "train_speed(iter/s)": 1.450391
    },
    {
      "epoch": 0.6634248746840324,
      "grad_norm": 3.4631004333496094,
      "learning_rate": 9.571860174444649e-05,
      "loss": 2.378741455078125,
      "memory(GiB)": 58.3,
      "step": 15485,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.450334
    },
    {
      "epoch": 0.6636390900132814,
      "grad_norm": 4.903103351593018,
      "learning_rate": 9.571587661334028e-05,
      "loss": 2.87160701751709,
      "memory(GiB)": 58.3,
      "step": 15490,
      "token_acc": 0.42485549132947975,
      "train_speed(iter/s)": 1.450324
    },
    {
      "epoch": 0.6638533053425303,
      "grad_norm": 5.625885486602783,
      "learning_rate": 9.571315065404808e-05,
      "loss": 2.830636405944824,
      "memory(GiB)": 58.3,
      "step": 15495,
      "token_acc": 0.43302180685358255,
      "train_speed(iter/s)": 1.450328
    },
    {
      "epoch": 0.6640675206717793,
      "grad_norm": 3.764037847518921,
      "learning_rate": 9.571042386661928e-05,
      "loss": 2.3046358108520506,
      "memory(GiB)": 58.3,
      "step": 15500,
      "token_acc": 0.5148809523809523,
      "train_speed(iter/s)": 1.450257
    },
    {
      "epoch": 0.6640675206717793,
      "eval_loss": 2.1536643505096436,
      "eval_runtime": 14.8387,
      "eval_samples_per_second": 6.739,
      "eval_steps_per_second": 6.739,
      "eval_token_acc": 0.48297604035308955,
      "step": 15500
    },
    {
      "epoch": 0.6642817360010282,
      "grad_norm": 4.8557024002075195,
      "learning_rate": 9.570769625110325e-05,
      "loss": 2.5669265747070313,
      "memory(GiB)": 58.3,
      "step": 15505,
      "token_acc": 0.47685185185185186,
      "train_speed(iter/s)": 1.44813
    },
    {
      "epoch": 0.6644959513302772,
      "grad_norm": 4.15380859375,
      "learning_rate": 9.570496780754945e-05,
      "loss": 2.394583892822266,
      "memory(GiB)": 58.3,
      "step": 15510,
      "token_acc": 0.5425101214574899,
      "train_speed(iter/s)": 1.448189
    },
    {
      "epoch": 0.6647101666595262,
      "grad_norm": 3.6459875106811523,
      "learning_rate": 9.570223853600727e-05,
      "loss": 2.4049057006835937,
      "memory(GiB)": 58.3,
      "step": 15515,
      "token_acc": 0.46407185628742514,
      "train_speed(iter/s)": 1.448203
    },
    {
      "epoch": 0.6649243819887751,
      "grad_norm": 4.712183952331543,
      "learning_rate": 9.569950843652618e-05,
      "loss": 2.4937490463256835,
      "memory(GiB)": 58.3,
      "step": 15520,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.448305
    },
    {
      "epoch": 0.665138597318024,
      "grad_norm": 4.0576252937316895,
      "learning_rate": 9.56967775091556e-05,
      "loss": 2.5970298767089846,
      "memory(GiB)": 58.3,
      "step": 15525,
      "token_acc": 0.45481927710843373,
      "train_speed(iter/s)": 1.448361
    },
    {
      "epoch": 0.6653528126472731,
      "grad_norm": 7.180209159851074,
      "learning_rate": 9.569404575394505e-05,
      "loss": 2.9798162460327147,
      "memory(GiB)": 58.3,
      "step": 15530,
      "token_acc": 0.3971631205673759,
      "train_speed(iter/s)": 1.448372
    },
    {
      "epoch": 0.665567027976522,
      "grad_norm": 6.177602767944336,
      "learning_rate": 9.569131317094399e-05,
      "loss": 2.54278450012207,
      "memory(GiB)": 58.3,
      "step": 15535,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.448288
    },
    {
      "epoch": 0.6657812433057709,
      "grad_norm": 4.771605968475342,
      "learning_rate": 9.568857976020193e-05,
      "loss": 2.2341651916503906,
      "memory(GiB)": 58.3,
      "step": 15540,
      "token_acc": 0.5192307692307693,
      "train_speed(iter/s)": 1.448205
    },
    {
      "epoch": 0.66599545863502,
      "grad_norm": 6.403241157531738,
      "learning_rate": 9.568584552176838e-05,
      "loss": 2.461362838745117,
      "memory(GiB)": 58.3,
      "step": 15545,
      "token_acc": 0.458041958041958,
      "train_speed(iter/s)": 1.448194
    },
    {
      "epoch": 0.6662096739642689,
      "grad_norm": 6.478058338165283,
      "learning_rate": 9.568311045569289e-05,
      "loss": 2.625710296630859,
      "memory(GiB)": 58.3,
      "step": 15550,
      "token_acc": 0.4513888888888889,
      "train_speed(iter/s)": 1.448289
    },
    {
      "epoch": 0.6664238892935178,
      "grad_norm": 4.833446979522705,
      "learning_rate": 9.568037456202501e-05,
      "loss": 2.674976348876953,
      "memory(GiB)": 58.3,
      "step": 15555,
      "token_acc": 0.44404332129963897,
      "train_speed(iter/s)": 1.448339
    },
    {
      "epoch": 0.6666381046227668,
      "grad_norm": 5.176446437835693,
      "learning_rate": 9.567763784081428e-05,
      "loss": 2.465069389343262,
      "memory(GiB)": 58.3,
      "step": 15560,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.44838
    },
    {
      "epoch": 0.6668523199520158,
      "grad_norm": 4.159816741943359,
      "learning_rate": 9.567490029211029e-05,
      "loss": 2.8300382614135744,
      "memory(GiB)": 58.3,
      "step": 15565,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.448447
    },
    {
      "epoch": 0.6670665352812647,
      "grad_norm": 5.9345173835754395,
      "learning_rate": 9.567216191596264e-05,
      "loss": 2.7121652603149413,
      "memory(GiB)": 58.3,
      "step": 15570,
      "token_acc": 0.4022140221402214,
      "train_speed(iter/s)": 1.448456
    },
    {
      "epoch": 0.6672807506105137,
      "grad_norm": 5.538405895233154,
      "learning_rate": 9.566942271242093e-05,
      "loss": 2.6163633346557615,
      "memory(GiB)": 58.3,
      "step": 15575,
      "token_acc": 0.46987951807228917,
      "train_speed(iter/s)": 1.448494
    },
    {
      "epoch": 0.6674949659397627,
      "grad_norm": 4.507320404052734,
      "learning_rate": 9.566668268153479e-05,
      "loss": 2.7107967376708983,
      "memory(GiB)": 58.3,
      "step": 15580,
      "token_acc": 0.4542483660130719,
      "train_speed(iter/s)": 1.44853
    },
    {
      "epoch": 0.6677091812690116,
      "grad_norm": 4.423649311065674,
      "learning_rate": 9.566394182335385e-05,
      "loss": 2.3837303161621093,
      "memory(GiB)": 58.3,
      "step": 15585,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.448467
    },
    {
      "epoch": 0.6679233965982606,
      "grad_norm": 4.200186252593994,
      "learning_rate": 9.566120013792776e-05,
      "loss": 2.636524963378906,
      "memory(GiB)": 58.3,
      "step": 15590,
      "token_acc": 0.42948717948717946,
      "train_speed(iter/s)": 1.448527
    },
    {
      "epoch": 0.6681376119275095,
      "grad_norm": 4.650906562805176,
      "learning_rate": 9.56584576253062e-05,
      "loss": 2.7633777618408204,
      "memory(GiB)": 58.3,
      "step": 15595,
      "token_acc": 0.44866920152091255,
      "train_speed(iter/s)": 1.448562
    },
    {
      "epoch": 0.6683518272567585,
      "grad_norm": 4.903339385986328,
      "learning_rate": 9.565571428553887e-05,
      "loss": 2.8440830230712892,
      "memory(GiB)": 58.3,
      "step": 15600,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.448635
    },
    {
      "epoch": 0.6685660425860075,
      "grad_norm": 5.680325031280518,
      "learning_rate": 9.565297011867543e-05,
      "loss": 2.523952293395996,
      "memory(GiB)": 58.3,
      "step": 15605,
      "token_acc": 0.4872881355932203,
      "train_speed(iter/s)": 1.448532
    },
    {
      "epoch": 0.6687802579152564,
      "grad_norm": 3.57682728767395,
      "learning_rate": 9.565022512476561e-05,
      "loss": 2.3742599487304688,
      "memory(GiB)": 58.3,
      "step": 15610,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.448619
    },
    {
      "epoch": 0.6689944732445053,
      "grad_norm": 3.680056571960449,
      "learning_rate": 9.564747930385915e-05,
      "loss": 2.515506553649902,
      "memory(GiB)": 58.3,
      "step": 15615,
      "token_acc": 0.4365781710914454,
      "train_speed(iter/s)": 1.448648
    },
    {
      "epoch": 0.6692086885737544,
      "grad_norm": 4.250692367553711,
      "learning_rate": 9.564473265600576e-05,
      "loss": 2.7476613998413084,
      "memory(GiB)": 58.3,
      "step": 15620,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.448617
    },
    {
      "epoch": 0.6694229039030033,
      "grad_norm": 3.3645665645599365,
      "learning_rate": 9.564198518125523e-05,
      "loss": 2.7650035858154296,
      "memory(GiB)": 58.3,
      "step": 15625,
      "token_acc": 0.44200626959247646,
      "train_speed(iter/s)": 1.448681
    },
    {
      "epoch": 0.6696371192322522,
      "grad_norm": 4.581256866455078,
      "learning_rate": 9.563923687965733e-05,
      "loss": 2.4920093536376955,
      "memory(GiB)": 58.3,
      "step": 15630,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.448558
    },
    {
      "epoch": 0.6698513345615013,
      "grad_norm": 4.66019344329834,
      "learning_rate": 9.563648775126184e-05,
      "loss": 2.6510133743286133,
      "memory(GiB)": 58.3,
      "step": 15635,
      "token_acc": 0.41605839416058393,
      "train_speed(iter/s)": 1.448541
    },
    {
      "epoch": 0.6700655498907502,
      "grad_norm": 3.4649465084075928,
      "learning_rate": 9.563373779611855e-05,
      "loss": 3.2039012908935547,
      "memory(GiB)": 58.3,
      "step": 15640,
      "token_acc": 0.4186666666666667,
      "train_speed(iter/s)": 1.448552
    },
    {
      "epoch": 0.6702797652199991,
      "grad_norm": 3.6231043338775635,
      "learning_rate": 9.563098701427731e-05,
      "loss": 2.628450393676758,
      "memory(GiB)": 58.3,
      "step": 15645,
      "token_acc": 0.42805755395683454,
      "train_speed(iter/s)": 1.448482
    },
    {
      "epoch": 0.6704939805492481,
      "grad_norm": 3.9865915775299072,
      "learning_rate": 9.562823540578791e-05,
      "loss": 2.263797569274902,
      "memory(GiB)": 58.3,
      "step": 15650,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.448468
    },
    {
      "epoch": 0.6707081958784971,
      "grad_norm": 4.097041130065918,
      "learning_rate": 9.562548297070025e-05,
      "loss": 2.6353050231933595,
      "memory(GiB)": 58.3,
      "step": 15655,
      "token_acc": 0.4420289855072464,
      "train_speed(iter/s)": 1.448541
    },
    {
      "epoch": 0.670922411207746,
      "grad_norm": 4.2176947593688965,
      "learning_rate": 9.562272970906416e-05,
      "loss": 2.4473430633544924,
      "memory(GiB)": 58.3,
      "step": 15660,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.448567
    },
    {
      "epoch": 0.671136626536995,
      "grad_norm": 4.233860015869141,
      "learning_rate": 9.561997562092951e-05,
      "loss": 2.614794158935547,
      "memory(GiB)": 58.3,
      "step": 15665,
      "token_acc": 0.5070422535211268,
      "train_speed(iter/s)": 1.448615
    },
    {
      "epoch": 0.6713508418662439,
      "grad_norm": 5.902677059173584,
      "learning_rate": 9.561722070634623e-05,
      "loss": 2.4403099060058593,
      "memory(GiB)": 58.3,
      "step": 15670,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.448708
    },
    {
      "epoch": 0.6715650571954929,
      "grad_norm": 4.197824001312256,
      "learning_rate": 9.561446496536418e-05,
      "loss": 2.6506038665771485,
      "memory(GiB)": 58.3,
      "step": 15675,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.448782
    },
    {
      "epoch": 0.6717792725247419,
      "grad_norm": 4.982149600982666,
      "learning_rate": 9.561170839803331e-05,
      "loss": 2.8639751434326173,
      "memory(GiB)": 58.3,
      "step": 15680,
      "token_acc": 0.4394904458598726,
      "train_speed(iter/s)": 1.448724
    },
    {
      "epoch": 0.6719934878539908,
      "grad_norm": 3.7808735370635986,
      "learning_rate": 9.560895100440357e-05,
      "loss": 2.6068578720092774,
      "memory(GiB)": 58.3,
      "step": 15685,
      "token_acc": 0.47384615384615386,
      "train_speed(iter/s)": 1.448636
    },
    {
      "epoch": 0.6722077031832397,
      "grad_norm": 5.031792163848877,
      "learning_rate": 9.56061927845249e-05,
      "loss": 2.752046585083008,
      "memory(GiB)": 58.3,
      "step": 15690,
      "token_acc": 0.4415954415954416,
      "train_speed(iter/s)": 1.448581
    },
    {
      "epoch": 0.6724219185124888,
      "grad_norm": 4.1393866539001465,
      "learning_rate": 9.560343373844724e-05,
      "loss": 2.550790214538574,
      "memory(GiB)": 58.3,
      "step": 15695,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.448735
    },
    {
      "epoch": 0.6726361338417377,
      "grad_norm": 3.781517505645752,
      "learning_rate": 9.560067386622063e-05,
      "loss": 2.2200366973876955,
      "memory(GiB)": 58.3,
      "step": 15700,
      "token_acc": 0.4690909090909091,
      "train_speed(iter/s)": 1.448676
    },
    {
      "epoch": 0.6728503491709866,
      "grad_norm": 3.903384208679199,
      "learning_rate": 9.559791316789502e-05,
      "loss": 2.4927507400512696,
      "memory(GiB)": 58.3,
      "step": 15705,
      "token_acc": 0.5054945054945055,
      "train_speed(iter/s)": 1.448766
    },
    {
      "epoch": 0.6730645645002357,
      "grad_norm": 4.564765453338623,
      "learning_rate": 9.559515164352044e-05,
      "loss": 2.3333017349243166,
      "memory(GiB)": 58.3,
      "step": 15710,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.448836
    },
    {
      "epoch": 0.6732787798294846,
      "grad_norm": 4.970495223999023,
      "learning_rate": 9.559238929314692e-05,
      "loss": 2.400589942932129,
      "memory(GiB)": 58.3,
      "step": 15715,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.448981
    },
    {
      "epoch": 0.6734929951587335,
      "grad_norm": 6.2129316329956055,
      "learning_rate": 9.558962611682447e-05,
      "loss": 2.517072296142578,
      "memory(GiB)": 58.3,
      "step": 15720,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.448921
    },
    {
      "epoch": 0.6737072104879825,
      "grad_norm": 4.229959011077881,
      "learning_rate": 9.558686211460322e-05,
      "loss": 2.586163330078125,
      "memory(GiB)": 58.3,
      "step": 15725,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.449017
    },
    {
      "epoch": 0.6739214258172315,
      "grad_norm": 5.169749736785889,
      "learning_rate": 9.558409728653317e-05,
      "loss": 2.6079883575439453,
      "memory(GiB)": 58.3,
      "step": 15730,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.449006
    },
    {
      "epoch": 0.6741356411464804,
      "grad_norm": 4.491950988769531,
      "learning_rate": 9.558133163266444e-05,
      "loss": 2.63720817565918,
      "memory(GiB)": 58.3,
      "step": 15735,
      "token_acc": 0.42700729927007297,
      "train_speed(iter/s)": 1.449099
    },
    {
      "epoch": 0.6743498564757294,
      "grad_norm": 3.3280744552612305,
      "learning_rate": 9.557856515304713e-05,
      "loss": 2.5957775115966797,
      "memory(GiB)": 58.3,
      "step": 15740,
      "token_acc": 0.5095785440613027,
      "train_speed(iter/s)": 1.449122
    },
    {
      "epoch": 0.6745640718049783,
      "grad_norm": 6.450634479522705,
      "learning_rate": 9.557579784773137e-05,
      "loss": 2.9286909103393555,
      "memory(GiB)": 58.3,
      "step": 15745,
      "token_acc": 0.415929203539823,
      "train_speed(iter/s)": 1.449168
    },
    {
      "epoch": 0.6747782871342273,
      "grad_norm": 4.917341232299805,
      "learning_rate": 9.557302971676727e-05,
      "loss": 2.7538082122802736,
      "memory(GiB)": 58.3,
      "step": 15750,
      "token_acc": 0.44816053511705684,
      "train_speed(iter/s)": 1.44924
    },
    {
      "epoch": 0.6749925024634763,
      "grad_norm": 4.088737487792969,
      "learning_rate": 9.557026076020498e-05,
      "loss": 2.5165273666381838,
      "memory(GiB)": 58.3,
      "step": 15755,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.449304
    },
    {
      "epoch": 0.6752067177927252,
      "grad_norm": 5.825265407562256,
      "learning_rate": 9.556749097809468e-05,
      "loss": 2.7471025466918944,
      "memory(GiB)": 58.3,
      "step": 15760,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.449354
    },
    {
      "epoch": 0.6754209331219742,
      "grad_norm": 3.5806992053985596,
      "learning_rate": 9.556472037048651e-05,
      "loss": 2.5167312622070312,
      "memory(GiB)": 58.3,
      "step": 15765,
      "token_acc": 0.5060728744939271,
      "train_speed(iter/s)": 1.449396
    },
    {
      "epoch": 0.6756351484512232,
      "grad_norm": 5.661139965057373,
      "learning_rate": 9.556194893743071e-05,
      "loss": 2.584786224365234,
      "memory(GiB)": 58.3,
      "step": 15770,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.4495
    },
    {
      "epoch": 0.6758493637804721,
      "grad_norm": 3.8455541133880615,
      "learning_rate": 9.555917667897746e-05,
      "loss": 2.7654178619384764,
      "memory(GiB)": 58.3,
      "step": 15775,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.449355
    },
    {
      "epoch": 0.6760635791097211,
      "grad_norm": 3.6448280811309814,
      "learning_rate": 9.555640359517699e-05,
      "loss": 2.662990379333496,
      "memory(GiB)": 58.3,
      "step": 15780,
      "token_acc": 0.43217665615141954,
      "train_speed(iter/s)": 1.449363
    },
    {
      "epoch": 0.6762777944389701,
      "grad_norm": 5.703397274017334,
      "learning_rate": 9.555362968607952e-05,
      "loss": 2.5035566329956054,
      "memory(GiB)": 58.3,
      "step": 15785,
      "token_acc": 0.45864661654135336,
      "train_speed(iter/s)": 1.44935
    },
    {
      "epoch": 0.676492009768219,
      "grad_norm": 5.291288375854492,
      "learning_rate": 9.555085495173532e-05,
      "loss": 2.7179203033447266,
      "memory(GiB)": 58.3,
      "step": 15790,
      "token_acc": 0.44744744744744747,
      "train_speed(iter/s)": 1.449386
    },
    {
      "epoch": 0.676706225097468,
      "grad_norm": 3.1444194316864014,
      "learning_rate": 9.554807939219467e-05,
      "loss": 2.593650245666504,
      "memory(GiB)": 58.3,
      "step": 15795,
      "token_acc": 0.453551912568306,
      "train_speed(iter/s)": 1.449422
    },
    {
      "epoch": 0.676920440426717,
      "grad_norm": 4.930181503295898,
      "learning_rate": 9.554530300750782e-05,
      "loss": 2.3686145782470702,
      "memory(GiB)": 58.3,
      "step": 15800,
      "token_acc": 0.5096153846153846,
      "train_speed(iter/s)": 1.449459
    },
    {
      "epoch": 0.6771346557559659,
      "grad_norm": 3.371978282928467,
      "learning_rate": 9.554252579772509e-05,
      "loss": 2.7324657440185547,
      "memory(GiB)": 58.3,
      "step": 15805,
      "token_acc": 0.4409937888198758,
      "train_speed(iter/s)": 1.449497
    },
    {
      "epoch": 0.6773488710852149,
      "grad_norm": 4.526597499847412,
      "learning_rate": 9.553974776289678e-05,
      "loss": 2.4020816802978517,
      "memory(GiB)": 58.3,
      "step": 15810,
      "token_acc": 0.5228070175438596,
      "train_speed(iter/s)": 1.449483
    },
    {
      "epoch": 0.6775630864144638,
      "grad_norm": 5.530508518218994,
      "learning_rate": 9.55369689030732e-05,
      "loss": 2.7830686569213867,
      "memory(GiB)": 58.3,
      "step": 15815,
      "token_acc": 0.4423791821561338,
      "train_speed(iter/s)": 1.449503
    },
    {
      "epoch": 0.6777773017437128,
      "grad_norm": 5.447394847869873,
      "learning_rate": 9.553418921830473e-05,
      "loss": 2.8545257568359377,
      "memory(GiB)": 58.3,
      "step": 15820,
      "token_acc": 0.4367816091954023,
      "train_speed(iter/s)": 1.449539
    },
    {
      "epoch": 0.6779915170729618,
      "grad_norm": 4.551540374755859,
      "learning_rate": 9.55314087086417e-05,
      "loss": 2.6005834579467773,
      "memory(GiB)": 58.3,
      "step": 15825,
      "token_acc": 0.44722222222222224,
      "train_speed(iter/s)": 1.449545
    },
    {
      "epoch": 0.6782057324022107,
      "grad_norm": 4.18668794631958,
      "learning_rate": 9.552862737413449e-05,
      "loss": 2.5934478759765627,
      "memory(GiB)": 58.3,
      "step": 15830,
      "token_acc": 0.45934959349593496,
      "train_speed(iter/s)": 1.44949
    },
    {
      "epoch": 0.6784199477314596,
      "grad_norm": 4.853087425231934,
      "learning_rate": 9.55258452148335e-05,
      "loss": 2.9316173553466798,
      "memory(GiB)": 58.3,
      "step": 15835,
      "token_acc": 0.3720136518771331,
      "train_speed(iter/s)": 1.449685
    },
    {
      "epoch": 0.6786341630607087,
      "grad_norm": 4.226537227630615,
      "learning_rate": 9.552306223078911e-05,
      "loss": 2.6189754486083983,
      "memory(GiB)": 58.3,
      "step": 15840,
      "token_acc": 0.44814814814814813,
      "train_speed(iter/s)": 1.449792
    },
    {
      "epoch": 0.6788483783899576,
      "grad_norm": 4.9391984939575195,
      "learning_rate": 9.552027842205174e-05,
      "loss": 2.5882129669189453,
      "memory(GiB)": 58.3,
      "step": 15845,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.449899
    },
    {
      "epoch": 0.6790625937192065,
      "grad_norm": 4.2914886474609375,
      "learning_rate": 9.55174937886718e-05,
      "loss": 2.534084510803223,
      "memory(GiB)": 58.3,
      "step": 15850,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.449918
    },
    {
      "epoch": 0.6792768090484556,
      "grad_norm": 5.369961738586426,
      "learning_rate": 9.551470833069979e-05,
      "loss": 2.339730644226074,
      "memory(GiB)": 58.3,
      "step": 15855,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.44991
    },
    {
      "epoch": 0.6794910243777045,
      "grad_norm": 4.803164005279541,
      "learning_rate": 9.551192204818615e-05,
      "loss": 2.491539192199707,
      "memory(GiB)": 58.3,
      "step": 15860,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.449898
    },
    {
      "epoch": 0.6797052397069534,
      "grad_norm": 3.6270666122436523,
      "learning_rate": 9.550913494118133e-05,
      "loss": 2.540834426879883,
      "memory(GiB)": 58.3,
      "step": 15865,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.449983
    },
    {
      "epoch": 0.6799194550362024,
      "grad_norm": 4.375837802886963,
      "learning_rate": 9.550634700973585e-05,
      "loss": 2.5530811309814454,
      "memory(GiB)": 58.3,
      "step": 15870,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.450053
    },
    {
      "epoch": 0.6801336703654514,
      "grad_norm": 4.825113773345947,
      "learning_rate": 9.55035582539002e-05,
      "loss": 2.3854511260986326,
      "memory(GiB)": 58.3,
      "step": 15875,
      "token_acc": 0.495114006514658,
      "train_speed(iter/s)": 1.450144
    },
    {
      "epoch": 0.6803478856947003,
      "grad_norm": 4.216777801513672,
      "learning_rate": 9.550076867372491e-05,
      "loss": 2.5397287368774415,
      "memory(GiB)": 58.3,
      "step": 15880,
      "token_acc": 0.5020408163265306,
      "train_speed(iter/s)": 1.450225
    },
    {
      "epoch": 0.6805621010239493,
      "grad_norm": 3.958674907684326,
      "learning_rate": 9.549797826926052e-05,
      "loss": 3.0595335006713866,
      "memory(GiB)": 58.3,
      "step": 15885,
      "token_acc": 0.4208754208754209,
      "train_speed(iter/s)": 1.450291
    },
    {
      "epoch": 0.6807763163531982,
      "grad_norm": 4.0088629722595215,
      "learning_rate": 9.549518704055755e-05,
      "loss": 2.646309661865234,
      "memory(GiB)": 58.3,
      "step": 15890,
      "token_acc": 0.4847560975609756,
      "train_speed(iter/s)": 1.450348
    },
    {
      "epoch": 0.6809905316824472,
      "grad_norm": 3.2504427433013916,
      "learning_rate": 9.549239498766661e-05,
      "loss": 2.3795700073242188,
      "memory(GiB)": 58.3,
      "step": 15895,
      "token_acc": 0.5296167247386759,
      "train_speed(iter/s)": 1.450378
    },
    {
      "epoch": 0.6812047470116962,
      "grad_norm": 6.107132911682129,
      "learning_rate": 9.548960211063824e-05,
      "loss": 2.814947319030762,
      "memory(GiB)": 58.3,
      "step": 15900,
      "token_acc": 0.4377358490566038,
      "train_speed(iter/s)": 1.450523
    },
    {
      "epoch": 0.6814189623409451,
      "grad_norm": 4.817509174346924,
      "learning_rate": 9.548680840952308e-05,
      "loss": 2.4671628952026365,
      "memory(GiB)": 58.3,
      "step": 15905,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.450489
    },
    {
      "epoch": 0.681633177670194,
      "grad_norm": 3.37170672416687,
      "learning_rate": 9.548401388437169e-05,
      "loss": 2.4301380157470702,
      "memory(GiB)": 58.3,
      "step": 15910,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.450458
    },
    {
      "epoch": 0.6818473929994431,
      "grad_norm": 3.7406206130981445,
      "learning_rate": 9.548121853523475e-05,
      "loss": 2.5562255859375,
      "memory(GiB)": 58.3,
      "step": 15915,
      "token_acc": 0.4421364985163205,
      "train_speed(iter/s)": 1.450442
    },
    {
      "epoch": 0.682061608328692,
      "grad_norm": 5.093752861022949,
      "learning_rate": 9.547842236216285e-05,
      "loss": 2.5152847290039064,
      "memory(GiB)": 58.3,
      "step": 15920,
      "token_acc": 0.44074074074074077,
      "train_speed(iter/s)": 1.450432
    },
    {
      "epoch": 0.6822758236579409,
      "grad_norm": 7.331698894500732,
      "learning_rate": 9.547562536520667e-05,
      "loss": 2.437386894226074,
      "memory(GiB)": 58.3,
      "step": 15925,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.450507
    },
    {
      "epoch": 0.68249003898719,
      "grad_norm": 3.6611979007720947,
      "learning_rate": 9.547282754441687e-05,
      "loss": 2.376340866088867,
      "memory(GiB)": 58.3,
      "step": 15930,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.45038
    },
    {
      "epoch": 0.6827042543164389,
      "grad_norm": 4.315795421600342,
      "learning_rate": 9.547002889984415e-05,
      "loss": 2.6874521255493162,
      "memory(GiB)": 58.3,
      "step": 15935,
      "token_acc": 0.45808383233532934,
      "train_speed(iter/s)": 1.450423
    },
    {
      "epoch": 0.6829184696456878,
      "grad_norm": 3.830885648727417,
      "learning_rate": 9.546722943153921e-05,
      "loss": 2.7009918212890627,
      "memory(GiB)": 58.3,
      "step": 15940,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.450553
    },
    {
      "epoch": 0.6831326849749368,
      "grad_norm": 5.225132942199707,
      "learning_rate": 9.546442913955276e-05,
      "loss": 2.9436458587646483,
      "memory(GiB)": 58.3,
      "step": 15945,
      "token_acc": 0.42679127725856697,
      "train_speed(iter/s)": 1.450532
    },
    {
      "epoch": 0.6833469003041858,
      "grad_norm": 4.207620143890381,
      "learning_rate": 9.54616280239355e-05,
      "loss": 2.712986373901367,
      "memory(GiB)": 58.3,
      "step": 15950,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.450611
    },
    {
      "epoch": 0.6835611156334347,
      "grad_norm": 4.295711040496826,
      "learning_rate": 9.545882608473823e-05,
      "loss": 2.7001279830932616,
      "memory(GiB)": 58.3,
      "step": 15955,
      "token_acc": 0.41420118343195267,
      "train_speed(iter/s)": 1.450667
    },
    {
      "epoch": 0.6837753309626837,
      "grad_norm": 3.617671489715576,
      "learning_rate": 9.545602332201167e-05,
      "loss": 2.2829994201660155,
      "memory(GiB)": 58.3,
      "step": 15960,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.450622
    },
    {
      "epoch": 0.6839895462919326,
      "grad_norm": 6.1937456130981445,
      "learning_rate": 9.545321973580661e-05,
      "loss": 2.655887985229492,
      "memory(GiB)": 58.3,
      "step": 15965,
      "token_acc": 0.4813664596273292,
      "train_speed(iter/s)": 1.450634
    },
    {
      "epoch": 0.6842037616211816,
      "grad_norm": 4.067612648010254,
      "learning_rate": 9.545041532617382e-05,
      "loss": 2.786912536621094,
      "memory(GiB)": 58.3,
      "step": 15970,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.450701
    },
    {
      "epoch": 0.6844179769504306,
      "grad_norm": 4.4033989906311035,
      "learning_rate": 9.544761009316414e-05,
      "loss": 2.200278472900391,
      "memory(GiB)": 58.3,
      "step": 15975,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.450778
    },
    {
      "epoch": 0.6846321922796795,
      "grad_norm": 3.363515853881836,
      "learning_rate": 9.544480403682836e-05,
      "loss": 2.694540596008301,
      "memory(GiB)": 58.3,
      "step": 15980,
      "token_acc": 0.45016077170418006,
      "train_speed(iter/s)": 1.450812
    },
    {
      "epoch": 0.6848464076089285,
      "grad_norm": 4.277008533477783,
      "learning_rate": 9.544199715721734e-05,
      "loss": 2.2847898483276365,
      "memory(GiB)": 58.3,
      "step": 15985,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.450816
    },
    {
      "epoch": 0.6850606229381775,
      "grad_norm": 5.444164276123047,
      "learning_rate": 9.543918945438189e-05,
      "loss": 2.752191162109375,
      "memory(GiB)": 58.3,
      "step": 15990,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.450821
    },
    {
      "epoch": 0.6852748382674264,
      "grad_norm": 4.117075443267822,
      "learning_rate": 9.543638092837291e-05,
      "loss": 2.4991092681884766,
      "memory(GiB)": 58.3,
      "step": 15995,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.450865
    },
    {
      "epoch": 0.6854890535966753,
      "grad_norm": 5.4325151443481445,
      "learning_rate": 9.543357157924126e-05,
      "loss": 2.451528549194336,
      "memory(GiB)": 58.3,
      "step": 16000,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.450834
    },
    {
      "epoch": 0.6854890535966753,
      "eval_loss": 2.322693109512329,
      "eval_runtime": 13.6174,
      "eval_samples_per_second": 7.344,
      "eval_steps_per_second": 7.344,
      "eval_token_acc": 0.4449877750611247,
      "step": 16000
    },
    {
      "epoch": 0.6857032689259244,
      "grad_norm": 7.669858455657959,
      "learning_rate": 9.543076140703786e-05,
      "loss": 2.4565792083740234,
      "memory(GiB)": 58.3,
      "step": 16005,
      "token_acc": 0.45146520146520147,
      "train_speed(iter/s)": 1.448885
    },
    {
      "epoch": 0.6859174842551733,
      "grad_norm": 4.37327241897583,
      "learning_rate": 9.542795041181359e-05,
      "loss": 2.625646209716797,
      "memory(GiB)": 58.3,
      "step": 16010,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.449003
    },
    {
      "epoch": 0.6861316995844222,
      "grad_norm": 3.5756986141204834,
      "learning_rate": 9.542513859361937e-05,
      "loss": 2.615800476074219,
      "memory(GiB)": 58.3,
      "step": 16015,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.449111
    },
    {
      "epoch": 0.6863459149136713,
      "grad_norm": 3.9372990131378174,
      "learning_rate": 9.542232595250615e-05,
      "loss": 2.419285011291504,
      "memory(GiB)": 58.3,
      "step": 16020,
      "token_acc": 0.5186440677966102,
      "train_speed(iter/s)": 1.449101
    },
    {
      "epoch": 0.6865601302429202,
      "grad_norm": 3.793991804122925,
      "learning_rate": 9.54195124885249e-05,
      "loss": 2.5227998733520507,
      "memory(GiB)": 58.3,
      "step": 16025,
      "token_acc": 0.46715328467153283,
      "train_speed(iter/s)": 1.449144
    },
    {
      "epoch": 0.6867743455721691,
      "grad_norm": 5.286299705505371,
      "learning_rate": 9.541669820172656e-05,
      "loss": 2.799925422668457,
      "memory(GiB)": 58.3,
      "step": 16030,
      "token_acc": 0.4266666666666667,
      "train_speed(iter/s)": 1.449197
    },
    {
      "epoch": 0.6869885609014181,
      "grad_norm": 3.6555585861206055,
      "learning_rate": 9.541388309216214e-05,
      "loss": 2.8805471420288087,
      "memory(GiB)": 58.3,
      "step": 16035,
      "token_acc": 0.4252199413489736,
      "train_speed(iter/s)": 1.449221
    },
    {
      "epoch": 0.6872027762306671,
      "grad_norm": 5.6552252769470215,
      "learning_rate": 9.541106715988263e-05,
      "loss": 2.4908044815063475,
      "memory(GiB)": 58.3,
      "step": 16040,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.449226
    },
    {
      "epoch": 0.687416991559916,
      "grad_norm": 4.04453182220459,
      "learning_rate": 9.540825040493903e-05,
      "loss": 2.737270736694336,
      "memory(GiB)": 58.3,
      "step": 16045,
      "token_acc": 0.4623955431754875,
      "train_speed(iter/s)": 1.449303
    },
    {
      "epoch": 0.687631206889165,
      "grad_norm": 6.701990604400635,
      "learning_rate": 9.540543282738239e-05,
      "loss": 2.8500442504882812,
      "memory(GiB)": 58.3,
      "step": 16050,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.44938
    },
    {
      "epoch": 0.6878454222184139,
      "grad_norm": 3.8837637901306152,
      "learning_rate": 9.540261442726373e-05,
      "loss": 2.3897096633911135,
      "memory(GiB)": 58.3,
      "step": 16055,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.449392
    },
    {
      "epoch": 0.6880596375476629,
      "grad_norm": 4.747652530670166,
      "learning_rate": 9.539979520463413e-05,
      "loss": 2.5115901947021486,
      "memory(GiB)": 58.3,
      "step": 16060,
      "token_acc": 0.48046875,
      "train_speed(iter/s)": 1.449424
    },
    {
      "epoch": 0.6882738528769119,
      "grad_norm": 3.818439245223999,
      "learning_rate": 9.539697515954465e-05,
      "loss": 2.573995590209961,
      "memory(GiB)": 58.3,
      "step": 16065,
      "token_acc": 0.5274725274725275,
      "train_speed(iter/s)": 1.449487
    },
    {
      "epoch": 0.6884880682061608,
      "grad_norm": 5.686701774597168,
      "learning_rate": 9.539415429204636e-05,
      "loss": 2.7241022109985353,
      "memory(GiB)": 58.3,
      "step": 16070,
      "token_acc": 0.43986254295532645,
      "train_speed(iter/s)": 1.449489
    },
    {
      "epoch": 0.6887022835354097,
      "grad_norm": 4.41595983505249,
      "learning_rate": 9.53913326021904e-05,
      "loss": 2.47973518371582,
      "memory(GiB)": 58.3,
      "step": 16075,
      "token_acc": 0.49377593360995853,
      "train_speed(iter/s)": 1.449536
    },
    {
      "epoch": 0.6889164988646588,
      "grad_norm": 4.110328197479248,
      "learning_rate": 9.538851009002785e-05,
      "loss": 2.469658851623535,
      "memory(GiB)": 58.3,
      "step": 16080,
      "token_acc": 0.49693251533742333,
      "train_speed(iter/s)": 1.44951
    },
    {
      "epoch": 0.6891307141939077,
      "grad_norm": 3.547121047973633,
      "learning_rate": 9.538568675560988e-05,
      "loss": 2.4900867462158205,
      "memory(GiB)": 58.3,
      "step": 16085,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.449508
    },
    {
      "epoch": 0.6893449295231566,
      "grad_norm": 4.928086757659912,
      "learning_rate": 9.538286259898762e-05,
      "loss": 2.4570621490478515,
      "memory(GiB)": 58.3,
      "step": 16090,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.449573
    },
    {
      "epoch": 0.6895591448524057,
      "grad_norm": 4.7055583000183105,
      "learning_rate": 9.538003762021221e-05,
      "loss": 2.6836000442504884,
      "memory(GiB)": 58.3,
      "step": 16095,
      "token_acc": 0.4236111111111111,
      "train_speed(iter/s)": 1.449735
    },
    {
      "epoch": 0.6897733601816546,
      "grad_norm": 4.162377834320068,
      "learning_rate": 9.537721181933488e-05,
      "loss": 2.3971416473388674,
      "memory(GiB)": 58.3,
      "step": 16100,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.449735
    },
    {
      "epoch": 0.6899875755109035,
      "grad_norm": 5.180971145629883,
      "learning_rate": 9.537438519640675e-05,
      "loss": 2.8276094436645507,
      "memory(GiB)": 58.3,
      "step": 16105,
      "token_acc": 0.4219269102990033,
      "train_speed(iter/s)": 1.449773
    },
    {
      "epoch": 0.6902017908401525,
      "grad_norm": 4.417033672332764,
      "learning_rate": 9.53715577514791e-05,
      "loss": 2.6529945373535155,
      "memory(GiB)": 58.3,
      "step": 16110,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.449903
    },
    {
      "epoch": 0.6904160061694015,
      "grad_norm": 4.234143257141113,
      "learning_rate": 9.536872948460312e-05,
      "loss": 3.0863554000854494,
      "memory(GiB)": 58.3,
      "step": 16115,
      "token_acc": 0.3931034482758621,
      "train_speed(iter/s)": 1.449982
    },
    {
      "epoch": 0.6906302214986505,
      "grad_norm": 5.599666595458984,
      "learning_rate": 9.536590039583004e-05,
      "loss": 3.0715717315673827,
      "memory(GiB)": 58.3,
      "step": 16120,
      "token_acc": 0.428169014084507,
      "train_speed(iter/s)": 1.449985
    },
    {
      "epoch": 0.6908444368278994,
      "grad_norm": 4.459110736846924,
      "learning_rate": 9.536307048521112e-05,
      "loss": 2.604759407043457,
      "memory(GiB)": 58.3,
      "step": 16125,
      "token_acc": 0.44569288389513106,
      "train_speed(iter/s)": 1.449951
    },
    {
      "epoch": 0.6910586521571483,
      "grad_norm": 3.9584357738494873,
      "learning_rate": 9.53602397527976e-05,
      "loss": 2.5693227767944338,
      "memory(GiB)": 58.3,
      "step": 16130,
      "token_acc": 0.4906832298136646,
      "train_speed(iter/s)": 1.449953
    },
    {
      "epoch": 0.6912728674863974,
      "grad_norm": 4.391383647918701,
      "learning_rate": 9.535740819864081e-05,
      "loss": 2.773303985595703,
      "memory(GiB)": 58.3,
      "step": 16135,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.449884
    },
    {
      "epoch": 0.6914870828156463,
      "grad_norm": 5.701388359069824,
      "learning_rate": 9.535457582279203e-05,
      "loss": 2.2828033447265623,
      "memory(GiB)": 58.3,
      "step": 16140,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.449785
    },
    {
      "epoch": 0.6917012981448952,
      "grad_norm": 5.432440280914307,
      "learning_rate": 9.535174262530254e-05,
      "loss": 2.720505714416504,
      "memory(GiB)": 58.3,
      "step": 16145,
      "token_acc": 0.4357142857142857,
      "train_speed(iter/s)": 1.449811
    },
    {
      "epoch": 0.6919155134741443,
      "grad_norm": 3.811840534210205,
      "learning_rate": 9.53489086062237e-05,
      "loss": 2.417336654663086,
      "memory(GiB)": 58.3,
      "step": 16150,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.449639
    },
    {
      "epoch": 0.6921297288033932,
      "grad_norm": 4.712042331695557,
      "learning_rate": 9.534607376560684e-05,
      "loss": 2.4939800262451173,
      "memory(GiB)": 58.3,
      "step": 16155,
      "token_acc": 0.4440789473684211,
      "train_speed(iter/s)": 1.449653
    },
    {
      "epoch": 0.6923439441326421,
      "grad_norm": 4.263582706451416,
      "learning_rate": 9.534323810350332e-05,
      "loss": 2.709640884399414,
      "memory(GiB)": 58.3,
      "step": 16160,
      "token_acc": 0.43370165745856354,
      "train_speed(iter/s)": 1.449606
    },
    {
      "epoch": 0.6925581594618911,
      "grad_norm": 3.9528398513793945,
      "learning_rate": 9.534040161996449e-05,
      "loss": 2.777052307128906,
      "memory(GiB)": 58.3,
      "step": 16165,
      "token_acc": 0.44753086419753085,
      "train_speed(iter/s)": 1.449681
    },
    {
      "epoch": 0.6927723747911401,
      "grad_norm": 5.489184379577637,
      "learning_rate": 9.533756431504177e-05,
      "loss": 2.6040952682495115,
      "memory(GiB)": 58.3,
      "step": 16170,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.449658
    },
    {
      "epoch": 0.692986590120389,
      "grad_norm": 4.068277359008789,
      "learning_rate": 9.533472618878653e-05,
      "loss": 2.476105499267578,
      "memory(GiB)": 58.3,
      "step": 16175,
      "token_acc": 0.47854785478547857,
      "train_speed(iter/s)": 1.449665
    },
    {
      "epoch": 0.693200805449638,
      "grad_norm": 4.737843036651611,
      "learning_rate": 9.53318872412502e-05,
      "loss": 2.7528892517089845,
      "memory(GiB)": 58.31,
      "step": 16180,
      "token_acc": 0.43389830508474575,
      "train_speed(iter/s)": 1.44956
    },
    {
      "epoch": 0.693415020778887,
      "grad_norm": 4.122523784637451,
      "learning_rate": 9.532904747248422e-05,
      "loss": 2.7778581619262694,
      "memory(GiB)": 58.31,
      "step": 16185,
      "token_acc": 0.44363636363636366,
      "train_speed(iter/s)": 1.449671
    },
    {
      "epoch": 0.6936292361081359,
      "grad_norm": 3.963362216949463,
      "learning_rate": 9.532620688254e-05,
      "loss": 2.5131149291992188,
      "memory(GiB)": 58.31,
      "step": 16190,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.44966
    },
    {
      "epoch": 0.6938434514373849,
      "grad_norm": 4.278669357299805,
      "learning_rate": 9.532336547146904e-05,
      "loss": 2.8307126998901366,
      "memory(GiB)": 58.31,
      "step": 16195,
      "token_acc": 0.4340175953079179,
      "train_speed(iter/s)": 1.449623
    },
    {
      "epoch": 0.6940576667666338,
      "grad_norm": 3.4925103187561035,
      "learning_rate": 9.532052323932279e-05,
      "loss": 2.4903324127197264,
      "memory(GiB)": 58.31,
      "step": 16200,
      "token_acc": 0.48024316109422494,
      "train_speed(iter/s)": 1.449628
    },
    {
      "epoch": 0.6942718820958828,
      "grad_norm": 3.750378370285034,
      "learning_rate": 9.531768018615276e-05,
      "loss": 2.578135108947754,
      "memory(GiB)": 58.31,
      "step": 16205,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.449617
    },
    {
      "epoch": 0.6944860974251318,
      "grad_norm": 3.7137584686279297,
      "learning_rate": 9.531483631201044e-05,
      "loss": 2.89431209564209,
      "memory(GiB)": 58.31,
      "step": 16210,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.449604
    },
    {
      "epoch": 0.6947003127543807,
      "grad_norm": 4.538028240203857,
      "learning_rate": 9.531199161694732e-05,
      "loss": 2.784868049621582,
      "memory(GiB)": 58.31,
      "step": 16215,
      "token_acc": 0.42813455657492355,
      "train_speed(iter/s)": 1.449597
    },
    {
      "epoch": 0.6949145280836296,
      "grad_norm": 3.5266575813293457,
      "learning_rate": 9.5309146101015e-05,
      "loss": 2.85870418548584,
      "memory(GiB)": 58.31,
      "step": 16220,
      "token_acc": 0.45321637426900585,
      "train_speed(iter/s)": 1.449693
    },
    {
      "epoch": 0.6951287434128787,
      "grad_norm": 4.1946845054626465,
      "learning_rate": 9.530629976426499e-05,
      "loss": 2.6545372009277344,
      "memory(GiB)": 58.31,
      "step": 16225,
      "token_acc": 0.43302180685358255,
      "train_speed(iter/s)": 1.449739
    },
    {
      "epoch": 0.6953429587421276,
      "grad_norm": 4.715703964233398,
      "learning_rate": 9.530345260674885e-05,
      "loss": 2.595602798461914,
      "memory(GiB)": 58.31,
      "step": 16230,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.449611
    },
    {
      "epoch": 0.6955571740713765,
      "grad_norm": 5.864632606506348,
      "learning_rate": 9.530060462851818e-05,
      "loss": 2.577613639831543,
      "memory(GiB)": 58.31,
      "step": 16235,
      "token_acc": 0.4309210526315789,
      "train_speed(iter/s)": 1.449668
    },
    {
      "epoch": 0.6957713894006256,
      "grad_norm": 4.584555625915527,
      "learning_rate": 9.529775582962455e-05,
      "loss": 2.4517467498779295,
      "memory(GiB)": 58.31,
      "step": 16240,
      "token_acc": 0.4746376811594203,
      "train_speed(iter/s)": 1.449669
    },
    {
      "epoch": 0.6959856047298745,
      "grad_norm": 3.677097797393799,
      "learning_rate": 9.529490621011958e-05,
      "loss": 2.4341060638427736,
      "memory(GiB)": 58.31,
      "step": 16245,
      "token_acc": 0.5103857566765578,
      "train_speed(iter/s)": 1.449694
    },
    {
      "epoch": 0.6961998200591234,
      "grad_norm": 4.16178560256958,
      "learning_rate": 9.529205577005491e-05,
      "loss": 2.491594696044922,
      "memory(GiB)": 58.31,
      "step": 16250,
      "token_acc": 0.48493975903614456,
      "train_speed(iter/s)": 1.449732
    },
    {
      "epoch": 0.6964140353883724,
      "grad_norm": 3.9551873207092285,
      "learning_rate": 9.528920450948215e-05,
      "loss": 2.1626350402832033,
      "memory(GiB)": 58.31,
      "step": 16255,
      "token_acc": 0.5194805194805194,
      "train_speed(iter/s)": 1.449701
    },
    {
      "epoch": 0.6966282507176214,
      "grad_norm": 3.9807679653167725,
      "learning_rate": 9.528635242845297e-05,
      "loss": 2.726948547363281,
      "memory(GiB)": 58.31,
      "step": 16260,
      "token_acc": 0.45660377358490567,
      "train_speed(iter/s)": 1.449736
    },
    {
      "epoch": 0.6968424660468703,
      "grad_norm": 2.9111335277557373,
      "learning_rate": 9.528349952701902e-05,
      "loss": 2.7875213623046875,
      "memory(GiB)": 58.31,
      "step": 16265,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.449802
    },
    {
      "epoch": 0.6970566813761193,
      "grad_norm": 3.767491102218628,
      "learning_rate": 9.528064580523201e-05,
      "loss": 2.675881195068359,
      "memory(GiB)": 58.31,
      "step": 16270,
      "token_acc": 0.4174454828660436,
      "train_speed(iter/s)": 1.449799
    },
    {
      "epoch": 0.6972708967053682,
      "grad_norm": 6.079091548919678,
      "learning_rate": 9.527779126314362e-05,
      "loss": 2.9438419342041016,
      "memory(GiB)": 58.31,
      "step": 16275,
      "token_acc": 0.4125,
      "train_speed(iter/s)": 1.449884
    },
    {
      "epoch": 0.6974851120346172,
      "grad_norm": 4.782175064086914,
      "learning_rate": 9.527493590080557e-05,
      "loss": 2.6365474700927733,
      "memory(GiB)": 58.31,
      "step": 16280,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.449817
    },
    {
      "epoch": 0.6976993273638662,
      "grad_norm": 4.4325432777404785,
      "learning_rate": 9.527207971826959e-05,
      "loss": 2.650682830810547,
      "memory(GiB)": 58.31,
      "step": 16285,
      "token_acc": 0.4522058823529412,
      "train_speed(iter/s)": 1.449787
    },
    {
      "epoch": 0.6979135426931151,
      "grad_norm": 4.810183048248291,
      "learning_rate": 9.52692227155874e-05,
      "loss": 2.6694765090942383,
      "memory(GiB)": 58.31,
      "step": 16290,
      "token_acc": 0.43653250773993807,
      "train_speed(iter/s)": 1.449844
    },
    {
      "epoch": 0.698127758022364,
      "grad_norm": 4.229180812835693,
      "learning_rate": 9.526636489281078e-05,
      "loss": 2.7364078521728517,
      "memory(GiB)": 58.31,
      "step": 16295,
      "token_acc": 0.3884057971014493,
      "train_speed(iter/s)": 1.449878
    },
    {
      "epoch": 0.6983419733516131,
      "grad_norm": 4.053791522979736,
      "learning_rate": 9.526350624999153e-05,
      "loss": 2.5006582260131838,
      "memory(GiB)": 58.31,
      "step": 16300,
      "token_acc": 0.527027027027027,
      "train_speed(iter/s)": 1.449816
    },
    {
      "epoch": 0.698556188680862,
      "grad_norm": 6.355086326599121,
      "learning_rate": 9.526064678718137e-05,
      "loss": 2.558262252807617,
      "memory(GiB)": 58.31,
      "step": 16305,
      "token_acc": 0.47580645161290325,
      "train_speed(iter/s)": 1.449899
    },
    {
      "epoch": 0.6987704040101109,
      "grad_norm": 4.354630947113037,
      "learning_rate": 9.525778650443214e-05,
      "loss": 2.4823657989501955,
      "memory(GiB)": 58.31,
      "step": 16310,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.449947
    },
    {
      "epoch": 0.69898461933936,
      "grad_norm": 5.398983478546143,
      "learning_rate": 9.525492540179563e-05,
      "loss": 2.803253936767578,
      "memory(GiB)": 58.31,
      "step": 16315,
      "token_acc": 0.4020979020979021,
      "train_speed(iter/s)": 1.449923
    },
    {
      "epoch": 0.6991988346686089,
      "grad_norm": 4.110025405883789,
      "learning_rate": 9.525206347932373e-05,
      "loss": 2.8333145141601563,
      "memory(GiB)": 58.31,
      "step": 16320,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.449874
    },
    {
      "epoch": 0.6994130499978578,
      "grad_norm": 3.702155828475952,
      "learning_rate": 9.524920073706824e-05,
      "loss": 2.674330711364746,
      "memory(GiB)": 58.31,
      "step": 16325,
      "token_acc": 0.44171779141104295,
      "train_speed(iter/s)": 1.449955
    },
    {
      "epoch": 0.6996272653271068,
      "grad_norm": 5.382178783416748,
      "learning_rate": 9.524633717508103e-05,
      "loss": 2.4870960235595705,
      "memory(GiB)": 58.31,
      "step": 16330,
      "token_acc": 0.48328267477203646,
      "train_speed(iter/s)": 1.449953
    },
    {
      "epoch": 0.6998414806563558,
      "grad_norm": 3.2930479049682617,
      "learning_rate": 9.524347279341397e-05,
      "loss": 2.3145076751708986,
      "memory(GiB)": 58.31,
      "step": 16335,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.449905
    },
    {
      "epoch": 0.7000556959856047,
      "grad_norm": 3.5885164737701416,
      "learning_rate": 9.524060759211897e-05,
      "loss": 2.5117565155029298,
      "memory(GiB)": 58.31,
      "step": 16340,
      "token_acc": 0.48739495798319327,
      "train_speed(iter/s)": 1.449904
    },
    {
      "epoch": 0.7002699113148537,
      "grad_norm": 5.129047393798828,
      "learning_rate": 9.523774157124791e-05,
      "loss": 2.7829776763916017,
      "memory(GiB)": 58.31,
      "step": 16345,
      "token_acc": 0.4142857142857143,
      "train_speed(iter/s)": 1.449922
    },
    {
      "epoch": 0.7004841266441026,
      "grad_norm": 3.988901376724243,
      "learning_rate": 9.523487473085274e-05,
      "loss": 2.69129638671875,
      "memory(GiB)": 58.31,
      "step": 16350,
      "token_acc": 0.4462025316455696,
      "train_speed(iter/s)": 1.450021
    },
    {
      "epoch": 0.7006983419733516,
      "grad_norm": 4.978018760681152,
      "learning_rate": 9.523200707098537e-05,
      "loss": 2.768937683105469,
      "memory(GiB)": 58.31,
      "step": 16355,
      "token_acc": 0.4961832061068702,
      "train_speed(iter/s)": 1.449972
    },
    {
      "epoch": 0.7009125573026006,
      "grad_norm": 3.4455044269561768,
      "learning_rate": 9.522913859169778e-05,
      "loss": 2.683019828796387,
      "memory(GiB)": 58.31,
      "step": 16360,
      "token_acc": 0.45535714285714285,
      "train_speed(iter/s)": 1.449999
    },
    {
      "epoch": 0.7011267726318495,
      "grad_norm": 5.0793776512146,
      "learning_rate": 9.52262692930419e-05,
      "loss": 2.762808418273926,
      "memory(GiB)": 58.31,
      "step": 16365,
      "token_acc": 0.46441947565543074,
      "train_speed(iter/s)": 1.450002
    },
    {
      "epoch": 0.7013409879610984,
      "grad_norm": 5.498508930206299,
      "learning_rate": 9.522339917506973e-05,
      "loss": 2.7171464920043946,
      "memory(GiB)": 58.31,
      "step": 16370,
      "token_acc": 0.4653846153846154,
      "train_speed(iter/s)": 1.450021
    },
    {
      "epoch": 0.7015552032903475,
      "grad_norm": 5.40051794052124,
      "learning_rate": 9.522052823783325e-05,
      "loss": 2.6595659255981445,
      "memory(GiB)": 58.31,
      "step": 16375,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.450067
    },
    {
      "epoch": 0.7017694186195964,
      "grad_norm": 4.447750568389893,
      "learning_rate": 9.521765648138449e-05,
      "loss": 2.330704689025879,
      "memory(GiB)": 58.31,
      "step": 16380,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.450206
    },
    {
      "epoch": 0.7019836339488453,
      "grad_norm": 3.448335886001587,
      "learning_rate": 9.521478390577546e-05,
      "loss": 2.478415298461914,
      "memory(GiB)": 58.31,
      "step": 16385,
      "token_acc": 0.4696969696969697,
      "train_speed(iter/s)": 1.450255
    },
    {
      "epoch": 0.7021978492780944,
      "grad_norm": 4.745573043823242,
      "learning_rate": 9.521191051105823e-05,
      "loss": 2.713380241394043,
      "memory(GiB)": 58.31,
      "step": 16390,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.450227
    },
    {
      "epoch": 0.7024120646073433,
      "grad_norm": 5.548971652984619,
      "learning_rate": 9.520903629728479e-05,
      "loss": 2.829987907409668,
      "memory(GiB)": 58.31,
      "step": 16395,
      "token_acc": 0.42028985507246375,
      "train_speed(iter/s)": 1.450252
    },
    {
      "epoch": 0.7026262799365922,
      "grad_norm": 6.130307674407959,
      "learning_rate": 9.520616126450726e-05,
      "loss": 2.919237518310547,
      "memory(GiB)": 58.31,
      "step": 16400,
      "token_acc": 0.42379182156133827,
      "train_speed(iter/s)": 1.450224
    },
    {
      "epoch": 0.7028404952658412,
      "grad_norm": 3.8262250423431396,
      "learning_rate": 9.520328541277772e-05,
      "loss": 2.577581214904785,
      "memory(GiB)": 58.31,
      "step": 16405,
      "token_acc": 0.46037735849056605,
      "train_speed(iter/s)": 1.450253
    },
    {
      "epoch": 0.7030547105950902,
      "grad_norm": 4.228984832763672,
      "learning_rate": 9.520040874214827e-05,
      "loss": 2.5994098663330076,
      "memory(GiB)": 66.02,
      "step": 16410,
      "token_acc": 0.4126984126984127,
      "train_speed(iter/s)": 1.450028
    },
    {
      "epoch": 0.7032689259243391,
      "grad_norm": 5.115041732788086,
      "learning_rate": 9.519753125267101e-05,
      "loss": 2.663760185241699,
      "memory(GiB)": 66.02,
      "step": 16415,
      "token_acc": 0.45084745762711864,
      "train_speed(iter/s)": 1.449944
    },
    {
      "epoch": 0.7034831412535881,
      "grad_norm": 4.100739479064941,
      "learning_rate": 9.519465294439805e-05,
      "loss": 2.630366325378418,
      "memory(GiB)": 66.02,
      "step": 16420,
      "token_acc": 0.4627831715210356,
      "train_speed(iter/s)": 1.449906
    },
    {
      "epoch": 0.703697356582837,
      "grad_norm": 4.6938300132751465,
      "learning_rate": 9.519177381738158e-05,
      "loss": 2.6072622299194337,
      "memory(GiB)": 66.02,
      "step": 16425,
      "token_acc": 0.48770491803278687,
      "train_speed(iter/s)": 1.449985
    },
    {
      "epoch": 0.703911571912086,
      "grad_norm": 3.35369873046875,
      "learning_rate": 9.518889387167373e-05,
      "loss": 2.489060974121094,
      "memory(GiB)": 66.02,
      "step": 16430,
      "token_acc": 0.48398576512455516,
      "train_speed(iter/s)": 1.45005
    },
    {
      "epoch": 0.704125787241335,
      "grad_norm": 4.35598087310791,
      "learning_rate": 9.518601310732667e-05,
      "loss": 2.4032833099365236,
      "memory(GiB)": 66.02,
      "step": 16435,
      "token_acc": 0.49096385542168675,
      "train_speed(iter/s)": 1.450024
    },
    {
      "epoch": 0.7043400025705839,
      "grad_norm": 4.064105987548828,
      "learning_rate": 9.518313152439259e-05,
      "loss": 2.6915843963623045,
      "memory(GiB)": 66.02,
      "step": 16440,
      "token_acc": 0.43597560975609756,
      "train_speed(iter/s)": 1.450105
    },
    {
      "epoch": 0.7045542178998329,
      "grad_norm": 4.789132595062256,
      "learning_rate": 9.51802491229237e-05,
      "loss": 2.5562313079833983,
      "memory(GiB)": 66.02,
      "step": 16445,
      "token_acc": 0.4228187919463087,
      "train_speed(iter/s)": 1.450243
    },
    {
      "epoch": 0.7047684332290819,
      "grad_norm": 3.658273935317993,
      "learning_rate": 9.517736590297223e-05,
      "loss": 2.533077430725098,
      "memory(GiB)": 66.02,
      "step": 16450,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.45024
    },
    {
      "epoch": 0.7049826485583308,
      "grad_norm": 3.736837148666382,
      "learning_rate": 9.517448186459038e-05,
      "loss": 2.8293155670166015,
      "memory(GiB)": 66.02,
      "step": 16455,
      "token_acc": 0.40502793296089384,
      "train_speed(iter/s)": 1.450317
    },
    {
      "epoch": 0.7051968638875799,
      "grad_norm": 5.096014976501465,
      "learning_rate": 9.517159700783042e-05,
      "loss": 2.6939470291137697,
      "memory(GiB)": 66.02,
      "step": 16460,
      "token_acc": 0.4471830985915493,
      "train_speed(iter/s)": 1.450365
    },
    {
      "epoch": 0.7054110792168288,
      "grad_norm": 3.996666431427002,
      "learning_rate": 9.516871133274461e-05,
      "loss": 2.4760942459106445,
      "memory(GiB)": 66.02,
      "step": 16465,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.450455
    },
    {
      "epoch": 0.7056252945460777,
      "grad_norm": 4.290310382843018,
      "learning_rate": 9.51658248393852e-05,
      "loss": 2.5249183654785154,
      "memory(GiB)": 66.02,
      "step": 16470,
      "token_acc": 0.44483985765124556,
      "train_speed(iter/s)": 1.450324
    },
    {
      "epoch": 0.7058395098753267,
      "grad_norm": 3.6787147521972656,
      "learning_rate": 9.516293752780454e-05,
      "loss": 2.4754077911376955,
      "memory(GiB)": 66.02,
      "step": 16475,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.450288
    },
    {
      "epoch": 0.7060537252045757,
      "grad_norm": 6.543824195861816,
      "learning_rate": 9.516004939805488e-05,
      "loss": 2.9514930725097654,
      "memory(GiB)": 66.02,
      "step": 16480,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.450242
    },
    {
      "epoch": 0.7062679405338246,
      "grad_norm": 4.192415237426758,
      "learning_rate": 9.515716045018856e-05,
      "loss": 2.3985437393188476,
      "memory(GiB)": 66.02,
      "step": 16485,
      "token_acc": 0.4765957446808511,
      "train_speed(iter/s)": 1.450307
    },
    {
      "epoch": 0.7064821558630736,
      "grad_norm": 3.979339122772217,
      "learning_rate": 9.515427068425793e-05,
      "loss": 2.457245635986328,
      "memory(GiB)": 66.02,
      "step": 16490,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.450247
    },
    {
      "epoch": 0.7066963711923225,
      "grad_norm": 3.7735750675201416,
      "learning_rate": 9.515138010031532e-05,
      "loss": 2.5792287826538085,
      "memory(GiB)": 66.02,
      "step": 16495,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.450258
    },
    {
      "epoch": 0.7069105865215715,
      "grad_norm": 4.328885555267334,
      "learning_rate": 9.51484886984131e-05,
      "loss": 2.369606590270996,
      "memory(GiB)": 66.02,
      "step": 16500,
      "token_acc": 0.47388059701492535,
      "train_speed(iter/s)": 1.450326
    },
    {
      "epoch": 0.7069105865215715,
      "eval_loss": 2.395688533782959,
      "eval_runtime": 14.091,
      "eval_samples_per_second": 7.097,
      "eval_steps_per_second": 7.097,
      "eval_token_acc": 0.46732429099876693,
      "step": 16500
    },
    {
      "epoch": 0.7071248018508205,
      "grad_norm": 4.743238925933838,
      "learning_rate": 9.514559647860366e-05,
      "loss": 2.4409147262573243,
      "memory(GiB)": 66.02,
      "step": 16505,
      "token_acc": 0.47569113441372735,
      "train_speed(iter/s)": 1.448363
    },
    {
      "epoch": 0.7073390171800694,
      "grad_norm": 3.831484079360962,
      "learning_rate": 9.514270344093939e-05,
      "loss": 2.3554454803466798,
      "memory(GiB)": 66.02,
      "step": 16510,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.448306
    },
    {
      "epoch": 0.7075532325093183,
      "grad_norm": 3.9123921394348145,
      "learning_rate": 9.513980958547269e-05,
      "loss": 2.552790069580078,
      "memory(GiB)": 66.02,
      "step": 16515,
      "token_acc": 0.4637223974763407,
      "train_speed(iter/s)": 1.448271
    },
    {
      "epoch": 0.7077674478385674,
      "grad_norm": 4.251744747161865,
      "learning_rate": 9.5136914912256e-05,
      "loss": 2.697835922241211,
      "memory(GiB)": 66.02,
      "step": 16520,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.448233
    },
    {
      "epoch": 0.7079816631678163,
      "grad_norm": 6.784512042999268,
      "learning_rate": 9.513401942134177e-05,
      "loss": 2.770741844177246,
      "memory(GiB)": 66.02,
      "step": 16525,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.448175
    },
    {
      "epoch": 0.7081958784970652,
      "grad_norm": 6.408402442932129,
      "learning_rate": 9.513112311278243e-05,
      "loss": 2.4230630874633787,
      "memory(GiB)": 66.02,
      "step": 16530,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.44824
    },
    {
      "epoch": 0.7084100938263143,
      "grad_norm": 4.877574443817139,
      "learning_rate": 9.512822598663045e-05,
      "loss": 2.4652992248535157,
      "memory(GiB)": 66.02,
      "step": 16535,
      "token_acc": 0.45514950166112955,
      "train_speed(iter/s)": 1.448293
    },
    {
      "epoch": 0.7086243091555632,
      "grad_norm": 7.459070682525635,
      "learning_rate": 9.512532804293832e-05,
      "loss": 2.2614675521850587,
      "memory(GiB)": 66.02,
      "step": 16540,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.448332
    },
    {
      "epoch": 0.7088385244848121,
      "grad_norm": 4.735755920410156,
      "learning_rate": 9.512242928175857e-05,
      "loss": 2.4417387008666993,
      "memory(GiB)": 66.02,
      "step": 16545,
      "token_acc": 0.44981412639405205,
      "train_speed(iter/s)": 1.448261
    },
    {
      "epoch": 0.7090527398140611,
      "grad_norm": 4.738497734069824,
      "learning_rate": 9.511952970314365e-05,
      "loss": 2.6786081314086916,
      "memory(GiB)": 66.02,
      "step": 16550,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.448292
    },
    {
      "epoch": 0.7092669551433101,
      "grad_norm": 3.969362735748291,
      "learning_rate": 9.511662930714614e-05,
      "loss": 2.6648998260498047,
      "memory(GiB)": 66.02,
      "step": 16555,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.448269
    },
    {
      "epoch": 0.709481170472559,
      "grad_norm": 3.7801785469055176,
      "learning_rate": 9.511372809381856e-05,
      "loss": 2.4989370346069335,
      "memory(GiB)": 66.02,
      "step": 16560,
      "token_acc": 0.4901315789473684,
      "train_speed(iter/s)": 1.448182
    },
    {
      "epoch": 0.709695385801808,
      "grad_norm": 5.0340423583984375,
      "learning_rate": 9.511082606321348e-05,
      "loss": 2.795084571838379,
      "memory(GiB)": 66.02,
      "step": 16565,
      "token_acc": 0.46325878594249204,
      "train_speed(iter/s)": 1.448032
    },
    {
      "epoch": 0.7099096011310569,
      "grad_norm": 5.707828521728516,
      "learning_rate": 9.510792321538348e-05,
      "loss": 2.8757003784179687,
      "memory(GiB)": 66.02,
      "step": 16570,
      "token_acc": 0.411371237458194,
      "train_speed(iter/s)": 1.44814
    },
    {
      "epoch": 0.7101238164603059,
      "grad_norm": 4.380470275878906,
      "learning_rate": 9.510501955038112e-05,
      "loss": 2.385051727294922,
      "memory(GiB)": 66.02,
      "step": 16575,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.448197
    },
    {
      "epoch": 0.7103380317895549,
      "grad_norm": 4.149919509887695,
      "learning_rate": 9.510211506825904e-05,
      "loss": 2.710375213623047,
      "memory(GiB)": 66.02,
      "step": 16580,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.448253
    },
    {
      "epoch": 0.7105522471188038,
      "grad_norm": 3.3036952018737793,
      "learning_rate": 9.509920976906981e-05,
      "loss": 2.7132644653320312,
      "memory(GiB)": 66.02,
      "step": 16585,
      "token_acc": 0.43440233236151604,
      "train_speed(iter/s)": 1.448354
    },
    {
      "epoch": 0.7107664624480527,
      "grad_norm": 3.1686246395111084,
      "learning_rate": 9.50963036528661e-05,
      "loss": 2.793349266052246,
      "memory(GiB)": 66.02,
      "step": 16590,
      "token_acc": 0.3955431754874652,
      "train_speed(iter/s)": 1.448285
    },
    {
      "epoch": 0.7109806777773018,
      "grad_norm": 3.1137187480926514,
      "learning_rate": 9.509339671970054e-05,
      "loss": 2.4462833404541016,
      "memory(GiB)": 66.02,
      "step": 16595,
      "token_acc": 0.5120481927710844,
      "train_speed(iter/s)": 1.448261
    },
    {
      "epoch": 0.7111948931065507,
      "grad_norm": 4.400082111358643,
      "learning_rate": 9.50904889696258e-05,
      "loss": 2.8768001556396485,
      "memory(GiB)": 66.02,
      "step": 16600,
      "token_acc": 0.4296577946768061,
      "train_speed(iter/s)": 1.448281
    },
    {
      "epoch": 0.7114091084357996,
      "grad_norm": 3.2652695178985596,
      "learning_rate": 9.508758040269457e-05,
      "loss": 2.4992000579833986,
      "memory(GiB)": 66.02,
      "step": 16605,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.44834
    },
    {
      "epoch": 0.7116233237650487,
      "grad_norm": 5.064366340637207,
      "learning_rate": 9.50846710189595e-05,
      "loss": 2.4636676788330076,
      "memory(GiB)": 66.02,
      "step": 16610,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.448332
    },
    {
      "epoch": 0.7118375390942976,
      "grad_norm": 3.61647891998291,
      "learning_rate": 9.508176081847333e-05,
      "loss": 2.437930679321289,
      "memory(GiB)": 66.02,
      "step": 16615,
      "token_acc": 0.47305389221556887,
      "train_speed(iter/s)": 1.448347
    },
    {
      "epoch": 0.7120517544235465,
      "grad_norm": 4.5554280281066895,
      "learning_rate": 9.507884980128879e-05,
      "loss": 2.6646337509155273,
      "memory(GiB)": 66.02,
      "step": 16620,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.448246
    },
    {
      "epoch": 0.7122659697527955,
      "grad_norm": 4.849153518676758,
      "learning_rate": 9.507593796745858e-05,
      "loss": 2.6109994888305663,
      "memory(GiB)": 66.02,
      "step": 16625,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.448316
    },
    {
      "epoch": 0.7124801850820445,
      "grad_norm": 4.082527160644531,
      "learning_rate": 9.507302531703549e-05,
      "loss": 2.654677391052246,
      "memory(GiB)": 66.02,
      "step": 16630,
      "token_acc": 0.4781021897810219,
      "train_speed(iter/s)": 1.448201
    },
    {
      "epoch": 0.7126944004112934,
      "grad_norm": 6.488663196563721,
      "learning_rate": 9.507011185007224e-05,
      "loss": 2.562953567504883,
      "memory(GiB)": 66.02,
      "step": 16635,
      "token_acc": 0.458041958041958,
      "train_speed(iter/s)": 1.448269
    },
    {
      "epoch": 0.7129086157405424,
      "grad_norm": 3.6953341960906982,
      "learning_rate": 9.506719756662163e-05,
      "loss": 2.6324296951293946,
      "memory(GiB)": 66.02,
      "step": 16640,
      "token_acc": 0.509090909090909,
      "train_speed(iter/s)": 1.448295
    },
    {
      "epoch": 0.7131228310697914,
      "grad_norm": 4.222318172454834,
      "learning_rate": 9.50642824667365e-05,
      "loss": 2.5764081954956053,
      "memory(GiB)": 66.02,
      "step": 16645,
      "token_acc": 0.41389728096676737,
      "train_speed(iter/s)": 1.448312
    },
    {
      "epoch": 0.7133370463990403,
      "grad_norm": 5.272609233856201,
      "learning_rate": 9.506136655046957e-05,
      "loss": 2.3582765579223635,
      "memory(GiB)": 66.02,
      "step": 16650,
      "token_acc": 0.5316455696202531,
      "train_speed(iter/s)": 1.448299
    },
    {
      "epoch": 0.7135512617282893,
      "grad_norm": 5.22477912902832,
      "learning_rate": 9.505844981787374e-05,
      "loss": 2.6418216705322264,
      "memory(GiB)": 66.02,
      "step": 16655,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.448346
    },
    {
      "epoch": 0.7137654770575382,
      "grad_norm": 3.1178903579711914,
      "learning_rate": 9.505553226900181e-05,
      "loss": 2.271025848388672,
      "memory(GiB)": 66.02,
      "step": 16660,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.448402
    },
    {
      "epoch": 0.7139796923867872,
      "grad_norm": 4.743771076202393,
      "learning_rate": 9.505261390390667e-05,
      "loss": 2.644114685058594,
      "memory(GiB)": 66.02,
      "step": 16665,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.448367
    },
    {
      "epoch": 0.7141939077160362,
      "grad_norm": 3.894268274307251,
      "learning_rate": 9.504969472264114e-05,
      "loss": 2.396790313720703,
      "memory(GiB)": 66.02,
      "step": 16670,
      "token_acc": 0.48923076923076925,
      "train_speed(iter/s)": 1.448485
    },
    {
      "epoch": 0.7144081230452851,
      "grad_norm": 3.8684892654418945,
      "learning_rate": 9.504677472525816e-05,
      "loss": 2.4858880996704102,
      "memory(GiB)": 66.02,
      "step": 16675,
      "token_acc": 0.4696969696969697,
      "train_speed(iter/s)": 1.448527
    },
    {
      "epoch": 0.714622338374534,
      "grad_norm": 4.849381446838379,
      "learning_rate": 9.504385391181058e-05,
      "loss": 2.515666198730469,
      "memory(GiB)": 66.02,
      "step": 16680,
      "token_acc": 0.48221343873517786,
      "train_speed(iter/s)": 1.4485
    },
    {
      "epoch": 0.7148365537037831,
      "grad_norm": 4.635295391082764,
      "learning_rate": 9.504093228235132e-05,
      "loss": 2.6861574172973635,
      "memory(GiB)": 66.02,
      "step": 16685,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.44858
    },
    {
      "epoch": 0.715050769033032,
      "grad_norm": 4.157852649688721,
      "learning_rate": 9.503800983693334e-05,
      "loss": 2.3195526123046877,
      "memory(GiB)": 66.02,
      "step": 16690,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.448612
    },
    {
      "epoch": 0.7152649843622809,
      "grad_norm": 4.365450382232666,
      "learning_rate": 9.503508657560956e-05,
      "loss": 2.799073600769043,
      "memory(GiB)": 66.02,
      "step": 16695,
      "token_acc": 0.4124629080118694,
      "train_speed(iter/s)": 1.44873
    },
    {
      "epoch": 0.71547919969153,
      "grad_norm": 4.813438415527344,
      "learning_rate": 9.503216249843294e-05,
      "loss": 2.6884090423583986,
      "memory(GiB)": 66.02,
      "step": 16700,
      "token_acc": 0.4265927977839335,
      "train_speed(iter/s)": 1.448762
    },
    {
      "epoch": 0.7156934150207789,
      "grad_norm": 4.648043155670166,
      "learning_rate": 9.502923760545644e-05,
      "loss": 2.622831344604492,
      "memory(GiB)": 66.02,
      "step": 16705,
      "token_acc": 0.4862068965517241,
      "train_speed(iter/s)": 1.448763
    },
    {
      "epoch": 0.7159076303500278,
      "grad_norm": 4.139097690582275,
      "learning_rate": 9.502631189673307e-05,
      "loss": 2.3960453033447267,
      "memory(GiB)": 66.02,
      "step": 16710,
      "token_acc": 0.5043859649122807,
      "train_speed(iter/s)": 1.448877
    },
    {
      "epoch": 0.7161218456792768,
      "grad_norm": 3.825942277908325,
      "learning_rate": 9.502338537231583e-05,
      "loss": 2.6495628356933594,
      "memory(GiB)": 66.02,
      "step": 16715,
      "token_acc": 0.44868035190615835,
      "train_speed(iter/s)": 1.448897
    },
    {
      "epoch": 0.7163360610085258,
      "grad_norm": 4.570528030395508,
      "learning_rate": 9.502045803225772e-05,
      "loss": 2.729033660888672,
      "memory(GiB)": 66.02,
      "step": 16720,
      "token_acc": 0.4196078431372549,
      "train_speed(iter/s)": 1.448988
    },
    {
      "epoch": 0.7165502763377747,
      "grad_norm": 4.131954669952393,
      "learning_rate": 9.501752987661177e-05,
      "loss": 2.60513973236084,
      "memory(GiB)": 66.02,
      "step": 16725,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.449049
    },
    {
      "epoch": 0.7167644916670237,
      "grad_norm": 3.556879758834839,
      "learning_rate": 9.501460090543105e-05,
      "loss": 2.2864448547363283,
      "memory(GiB)": 66.02,
      "step": 16730,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.44903
    },
    {
      "epoch": 0.7169787069962726,
      "grad_norm": 6.622162818908691,
      "learning_rate": 9.50116711187686e-05,
      "loss": 2.583823394775391,
      "memory(GiB)": 66.02,
      "step": 16735,
      "token_acc": 0.432,
      "train_speed(iter/s)": 1.449011
    },
    {
      "epoch": 0.7171929223255216,
      "grad_norm": 4.226747989654541,
      "learning_rate": 9.500874051667751e-05,
      "loss": 2.53782958984375,
      "memory(GiB)": 66.02,
      "step": 16740,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.449067
    },
    {
      "epoch": 0.7174071376547706,
      "grad_norm": 3.340885877609253,
      "learning_rate": 9.500580909921086e-05,
      "loss": 2.438665580749512,
      "memory(GiB)": 66.02,
      "step": 16745,
      "token_acc": 0.44594594594594594,
      "train_speed(iter/s)": 1.449148
    },
    {
      "epoch": 0.7176213529840195,
      "grad_norm": 4.5980424880981445,
      "learning_rate": 9.500287686642177e-05,
      "loss": 2.5203458786010744,
      "memory(GiB)": 66.02,
      "step": 16750,
      "token_acc": 0.45517241379310347,
      "train_speed(iter/s)": 1.449138
    },
    {
      "epoch": 0.7178355683132684,
      "grad_norm": 3.902045726776123,
      "learning_rate": 9.499994381836334e-05,
      "loss": 2.7687450408935548,
      "memory(GiB)": 66.02,
      "step": 16755,
      "token_acc": 0.48746518105849584,
      "train_speed(iter/s)": 1.449159
    },
    {
      "epoch": 0.7180497836425175,
      "grad_norm": 3.6287145614624023,
      "learning_rate": 9.499700995508871e-05,
      "loss": 2.5825531005859377,
      "memory(GiB)": 66.02,
      "step": 16760,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.449072
    },
    {
      "epoch": 0.7182639989717664,
      "grad_norm": 4.56061315536499,
      "learning_rate": 9.499407527665103e-05,
      "loss": 2.476934242248535,
      "memory(GiB)": 66.02,
      "step": 16765,
      "token_acc": 0.4881889763779528,
      "train_speed(iter/s)": 1.448946
    },
    {
      "epoch": 0.7184782143010153,
      "grad_norm": 3.9306862354278564,
      "learning_rate": 9.499113978310348e-05,
      "loss": 2.6438152313232424,
      "memory(GiB)": 66.02,
      "step": 16770,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.448964
    },
    {
      "epoch": 0.7186924296302644,
      "grad_norm": 4.159835338592529,
      "learning_rate": 9.498820347449923e-05,
      "loss": 2.686585807800293,
      "memory(GiB)": 66.02,
      "step": 16775,
      "token_acc": 0.44528301886792454,
      "train_speed(iter/s)": 1.448902
    },
    {
      "epoch": 0.7189066449595133,
      "grad_norm": 4.010687351226807,
      "learning_rate": 9.498526635089147e-05,
      "loss": 2.5310346603393556,
      "memory(GiB)": 66.02,
      "step": 16780,
      "token_acc": 0.4402730375426621,
      "train_speed(iter/s)": 1.44893
    },
    {
      "epoch": 0.7191208602887622,
      "grad_norm": 5.286407470703125,
      "learning_rate": 9.498232841233341e-05,
      "loss": 2.6868690490722655,
      "memory(GiB)": 66.02,
      "step": 16785,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.449007
    },
    {
      "epoch": 0.7193350756180112,
      "grad_norm": 5.136654376983643,
      "learning_rate": 9.497938965887827e-05,
      "loss": 2.529734230041504,
      "memory(GiB)": 66.02,
      "step": 16790,
      "token_acc": 0.4669421487603306,
      "train_speed(iter/s)": 1.448982
    },
    {
      "epoch": 0.7195492909472602,
      "grad_norm": 3.7295262813568115,
      "learning_rate": 9.497645009057929e-05,
      "loss": 2.6471458435058595,
      "memory(GiB)": 66.02,
      "step": 16795,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.449068
    },
    {
      "epoch": 0.7197635062765092,
      "grad_norm": 3.391190767288208,
      "learning_rate": 9.497350970748973e-05,
      "loss": 2.6329538345336916,
      "memory(GiB)": 66.02,
      "step": 16800,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.449178
    },
    {
      "epoch": 0.7199777216057581,
      "grad_norm": 4.856434345245361,
      "learning_rate": 9.497056850966286e-05,
      "loss": 2.485714149475098,
      "memory(GiB)": 66.02,
      "step": 16805,
      "token_acc": 0.4978723404255319,
      "train_speed(iter/s)": 1.449291
    },
    {
      "epoch": 0.720191936935007,
      "grad_norm": 4.344027519226074,
      "learning_rate": 9.496762649715194e-05,
      "loss": 2.601712226867676,
      "memory(GiB)": 66.02,
      "step": 16810,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.449369
    },
    {
      "epoch": 0.7204061522642561,
      "grad_norm": 5.066960334777832,
      "learning_rate": 9.496468367001027e-05,
      "loss": 2.5768354415893553,
      "memory(GiB)": 66.02,
      "step": 16815,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.449416
    },
    {
      "epoch": 0.720620367593505,
      "grad_norm": 5.329759120941162,
      "learning_rate": 9.496174002829121e-05,
      "loss": 2.587724304199219,
      "memory(GiB)": 66.02,
      "step": 16820,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.449337
    },
    {
      "epoch": 0.7208345829227539,
      "grad_norm": 4.92232084274292,
      "learning_rate": 9.495879557204803e-05,
      "loss": 2.499320220947266,
      "memory(GiB)": 66.02,
      "step": 16825,
      "token_acc": 0.45849802371541504,
      "train_speed(iter/s)": 1.449149
    },
    {
      "epoch": 0.721048798252003,
      "grad_norm": 4.489537239074707,
      "learning_rate": 9.49558503013341e-05,
      "loss": 2.6504798889160157,
      "memory(GiB)": 66.02,
      "step": 16830,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.449243
    },
    {
      "epoch": 0.7212630135812519,
      "grad_norm": 4.168896675109863,
      "learning_rate": 9.495290421620278e-05,
      "loss": 2.498708724975586,
      "memory(GiB)": 66.02,
      "step": 16835,
      "token_acc": 0.47878787878787876,
      "train_speed(iter/s)": 1.449283
    },
    {
      "epoch": 0.7214772289105008,
      "grad_norm": 3.9724533557891846,
      "learning_rate": 9.494995731670742e-05,
      "loss": 2.7124778747558596,
      "memory(GiB)": 66.02,
      "step": 16840,
      "token_acc": 0.445859872611465,
      "train_speed(iter/s)": 1.449288
    },
    {
      "epoch": 0.7216914442397498,
      "grad_norm": 5.2523651123046875,
      "learning_rate": 9.494700960290141e-05,
      "loss": 2.6246126174926756,
      "memory(GiB)": 66.02,
      "step": 16845,
      "token_acc": 0.45934959349593496,
      "train_speed(iter/s)": 1.449383
    },
    {
      "epoch": 0.7219056595689988,
      "grad_norm": 5.6819987297058105,
      "learning_rate": 9.494406107483817e-05,
      "loss": 2.6334814071655273,
      "memory(GiB)": 66.02,
      "step": 16850,
      "token_acc": 0.4360655737704918,
      "train_speed(iter/s)": 1.449414
    },
    {
      "epoch": 0.7221198748982477,
      "grad_norm": 5.335559844970703,
      "learning_rate": 9.49411117325711e-05,
      "loss": 2.5696552276611326,
      "memory(GiB)": 66.02,
      "step": 16855,
      "token_acc": 0.47520661157024796,
      "train_speed(iter/s)": 1.449446
    },
    {
      "epoch": 0.7223340902274967,
      "grad_norm": 6.168452739715576,
      "learning_rate": 9.493816157615363e-05,
      "loss": 2.649845314025879,
      "memory(GiB)": 66.02,
      "step": 16860,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.449463
    },
    {
      "epoch": 0.7225483055567457,
      "grad_norm": 4.4678239822387695,
      "learning_rate": 9.493521060563921e-05,
      "loss": 2.5617391586303713,
      "memory(GiB)": 66.02,
      "step": 16865,
      "token_acc": 0.45016077170418006,
      "train_speed(iter/s)": 1.449549
    },
    {
      "epoch": 0.7227625208859946,
      "grad_norm": 4.201544284820557,
      "learning_rate": 9.49322588210813e-05,
      "loss": 3.019285774230957,
      "memory(GiB)": 66.02,
      "step": 16870,
      "token_acc": 0.40524781341107874,
      "train_speed(iter/s)": 1.449525
    },
    {
      "epoch": 0.7229767362152436,
      "grad_norm": 3.6340491771698,
      "learning_rate": 9.492930622253336e-05,
      "loss": 2.402752685546875,
      "memory(GiB)": 66.02,
      "step": 16875,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.449615
    },
    {
      "epoch": 0.7231909515444925,
      "grad_norm": 4.638591766357422,
      "learning_rate": 9.49263528100489e-05,
      "loss": 2.734838676452637,
      "memory(GiB)": 66.02,
      "step": 16880,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.449627
    },
    {
      "epoch": 0.7234051668737415,
      "grad_norm": 7.417115211486816,
      "learning_rate": 9.492339858368141e-05,
      "loss": 3.159964370727539,
      "memory(GiB)": 66.02,
      "step": 16885,
      "token_acc": 0.3783783783783784,
      "train_speed(iter/s)": 1.44968
    },
    {
      "epoch": 0.7236193822029905,
      "grad_norm": 5.388756275177002,
      "learning_rate": 9.492044354348442e-05,
      "loss": 2.4829620361328124,
      "memory(GiB)": 66.02,
      "step": 16890,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.449608
    },
    {
      "epoch": 0.7238335975322394,
      "grad_norm": 4.924877166748047,
      "learning_rate": 9.491748768951145e-05,
      "loss": 2.8124759674072264,
      "memory(GiB)": 66.02,
      "step": 16895,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.449694
    },
    {
      "epoch": 0.7240478128614883,
      "grad_norm": 5.154012680053711,
      "learning_rate": 9.491453102181606e-05,
      "loss": 2.5469661712646485,
      "memory(GiB)": 66.02,
      "step": 16900,
      "token_acc": 0.4911660777385159,
      "train_speed(iter/s)": 1.4498
    },
    {
      "epoch": 0.7242620281907374,
      "grad_norm": 4.461989879608154,
      "learning_rate": 9.49115735404518e-05,
      "loss": 2.574415588378906,
      "memory(GiB)": 66.02,
      "step": 16905,
      "token_acc": 0.44410876132930516,
      "train_speed(iter/s)": 1.449766
    },
    {
      "epoch": 0.7244762435199863,
      "grad_norm": 4.099794387817383,
      "learning_rate": 9.490861524547225e-05,
      "loss": 2.5401105880737305,
      "memory(GiB)": 66.02,
      "step": 16910,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.449788
    },
    {
      "epoch": 0.7246904588492352,
      "grad_norm": 4.441971302032471,
      "learning_rate": 9.490565613693102e-05,
      "loss": 2.5891605377197267,
      "memory(GiB)": 66.02,
      "step": 16915,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.449874
    },
    {
      "epoch": 0.7249046741784843,
      "grad_norm": 4.345794677734375,
      "learning_rate": 9.49026962148817e-05,
      "loss": 2.6970636367797853,
      "memory(GiB)": 66.02,
      "step": 16920,
      "token_acc": 0.44505494505494503,
      "train_speed(iter/s)": 1.449987
    },
    {
      "epoch": 0.7251188895077332,
      "grad_norm": 4.45413875579834,
      "learning_rate": 9.489973547937792e-05,
      "loss": 2.512656402587891,
      "memory(GiB)": 66.02,
      "step": 16925,
      "token_acc": 0.48598130841121495,
      "train_speed(iter/s)": 1.450104
    },
    {
      "epoch": 0.7253331048369821,
      "grad_norm": 4.909793853759766,
      "learning_rate": 9.48967739304733e-05,
      "loss": 2.256917190551758,
      "memory(GiB)": 66.02,
      "step": 16930,
      "token_acc": 0.47653429602888087,
      "train_speed(iter/s)": 1.450177
    },
    {
      "epoch": 0.7255473201662311,
      "grad_norm": 3.5779850482940674,
      "learning_rate": 9.489381156822152e-05,
      "loss": 2.7278827667236327,
      "memory(GiB)": 66.02,
      "step": 16935,
      "token_acc": 0.4740740740740741,
      "train_speed(iter/s)": 1.450192
    },
    {
      "epoch": 0.7257615354954801,
      "grad_norm": 5.764263153076172,
      "learning_rate": 9.489084839267621e-05,
      "loss": 2.571573257446289,
      "memory(GiB)": 66.02,
      "step": 16940,
      "token_acc": 0.48031496062992124,
      "train_speed(iter/s)": 1.450253
    },
    {
      "epoch": 0.725975750824729,
      "grad_norm": 3.739387273788452,
      "learning_rate": 9.488788440389109e-05,
      "loss": 2.6918308258056642,
      "memory(GiB)": 66.02,
      "step": 16945,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.45033
    },
    {
      "epoch": 0.726189966153978,
      "grad_norm": 3.6499791145324707,
      "learning_rate": 9.488491960191984e-05,
      "loss": 2.6464052200317383,
      "memory(GiB)": 66.02,
      "step": 16950,
      "token_acc": 0.4386503067484663,
      "train_speed(iter/s)": 1.45033
    },
    {
      "epoch": 0.7264041814832269,
      "grad_norm": 5.935890197753906,
      "learning_rate": 9.488195398681614e-05,
      "loss": 2.749914360046387,
      "memory(GiB)": 66.02,
      "step": 16955,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.450322
    },
    {
      "epoch": 0.7266183968124759,
      "grad_norm": 3.368055582046509,
      "learning_rate": 9.487898755863376e-05,
      "loss": 2.6027194976806642,
      "memory(GiB)": 66.02,
      "step": 16960,
      "token_acc": 0.4602076124567474,
      "train_speed(iter/s)": 1.450311
    },
    {
      "epoch": 0.7268326121417249,
      "grad_norm": 5.101498603820801,
      "learning_rate": 9.487602031742642e-05,
      "loss": 2.441142272949219,
      "memory(GiB)": 66.02,
      "step": 16965,
      "token_acc": 0.5049833887043189,
      "train_speed(iter/s)": 1.450289
    },
    {
      "epoch": 0.7270468274709738,
      "grad_norm": 4.607686996459961,
      "learning_rate": 9.487305226324786e-05,
      "loss": 2.597370147705078,
      "memory(GiB)": 66.02,
      "step": 16970,
      "token_acc": 0.45018450184501846,
      "train_speed(iter/s)": 1.450263
    },
    {
      "epoch": 0.7272610428002227,
      "grad_norm": 3.64021635055542,
      "learning_rate": 9.487008339615187e-05,
      "loss": 2.4405391693115233,
      "memory(GiB)": 66.02,
      "step": 16975,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.45027
    },
    {
      "epoch": 0.7274752581294718,
      "grad_norm": 4.465456962585449,
      "learning_rate": 9.486711371619224e-05,
      "loss": 2.839253234863281,
      "memory(GiB)": 66.02,
      "step": 16980,
      "token_acc": 0.4214046822742475,
      "train_speed(iter/s)": 1.450409
    },
    {
      "epoch": 0.7276894734587207,
      "grad_norm": 3.530322790145874,
      "learning_rate": 9.486414322342274e-05,
      "loss": 2.679307556152344,
      "memory(GiB)": 66.02,
      "step": 16985,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.450452
    },
    {
      "epoch": 0.7279036887879696,
      "grad_norm": 10.375099182128906,
      "learning_rate": 9.486117191789721e-05,
      "loss": 2.8104286193847656,
      "memory(GiB)": 66.02,
      "step": 16990,
      "token_acc": 0.4384057971014493,
      "train_speed(iter/s)": 1.450459
    },
    {
      "epoch": 0.7281179041172187,
      "grad_norm": 3.8432939052581787,
      "learning_rate": 9.485819979966946e-05,
      "loss": 2.687255859375,
      "memory(GiB)": 66.02,
      "step": 16995,
      "token_acc": 0.43103448275862066,
      "train_speed(iter/s)": 1.450417
    },
    {
      "epoch": 0.7283321194464676,
      "grad_norm": 5.823300361633301,
      "learning_rate": 9.485522686879336e-05,
      "loss": 2.5549243927001952,
      "memory(GiB)": 66.02,
      "step": 17000,
      "token_acc": 0.47305389221556887,
      "train_speed(iter/s)": 1.450509
    },
    {
      "epoch": 0.7283321194464676,
      "eval_loss": 2.220410108566284,
      "eval_runtime": 13.2803,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "eval_token_acc": 0.466484268125855,
      "step": 17000
    },
    {
      "epoch": 0.7285463347757165,
      "grad_norm": 3.686988115310669,
      "learning_rate": 9.485225312532274e-05,
      "loss": 2.4599887847900392,
      "memory(GiB)": 66.02,
      "step": 17005,
      "token_acc": 0.45675413022351796,
      "train_speed(iter/s)": 1.448857
    },
    {
      "epoch": 0.7287605501049655,
      "grad_norm": 4.44323205947876,
      "learning_rate": 9.484927856931146e-05,
      "loss": 2.3029518127441406,
      "memory(GiB)": 66.02,
      "step": 17010,
      "token_acc": 0.49767441860465117,
      "train_speed(iter/s)": 1.448929
    },
    {
      "epoch": 0.7289747654342145,
      "grad_norm": 6.396110534667969,
      "learning_rate": 9.484630320081343e-05,
      "loss": 2.4460918426513674,
      "memory(GiB)": 66.02,
      "step": 17015,
      "token_acc": 0.5052264808362369,
      "train_speed(iter/s)": 1.448915
    },
    {
      "epoch": 0.7291889807634634,
      "grad_norm": 3.8762753009796143,
      "learning_rate": 9.484332701988257e-05,
      "loss": 2.427947425842285,
      "memory(GiB)": 66.02,
      "step": 17020,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.448997
    },
    {
      "epoch": 0.7294031960927124,
      "grad_norm": 4.288548469543457,
      "learning_rate": 9.484035002657276e-05,
      "loss": 2.561084747314453,
      "memory(GiB)": 66.02,
      "step": 17025,
      "token_acc": 0.46706586826347307,
      "train_speed(iter/s)": 1.448909
    },
    {
      "epoch": 0.7296174114219613,
      "grad_norm": 4.329801559448242,
      "learning_rate": 9.483737222093794e-05,
      "loss": 2.4620681762695313,
      "memory(GiB)": 66.02,
      "step": 17030,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.448988
    },
    {
      "epoch": 0.7298316267512103,
      "grad_norm": 5.658819675445557,
      "learning_rate": 9.483439360303208e-05,
      "loss": 2.3278966903686524,
      "memory(GiB)": 66.02,
      "step": 17035,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.4491
    },
    {
      "epoch": 0.7300458420804593,
      "grad_norm": 5.306597709655762,
      "learning_rate": 9.48314141729091e-05,
      "loss": 2.765092468261719,
      "memory(GiB)": 66.02,
      "step": 17040,
      "token_acc": 0.45564516129032256,
      "train_speed(iter/s)": 1.449075
    },
    {
      "epoch": 0.7302600574097082,
      "grad_norm": 13.803184509277344,
      "learning_rate": 9.482843393062302e-05,
      "loss": 2.764059638977051,
      "memory(GiB)": 66.02,
      "step": 17045,
      "token_acc": 0.4228395061728395,
      "train_speed(iter/s)": 1.449114
    },
    {
      "epoch": 0.7304742727389572,
      "grad_norm": 3.409217357635498,
      "learning_rate": 9.482545287622779e-05,
      "loss": 2.5887920379638674,
      "memory(GiB)": 66.02,
      "step": 17050,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.449142
    },
    {
      "epoch": 0.7306884880682062,
      "grad_norm": 3.3520493507385254,
      "learning_rate": 9.482247100977744e-05,
      "loss": 2.419658088684082,
      "memory(GiB)": 66.02,
      "step": 17055,
      "token_acc": 0.4786885245901639,
      "train_speed(iter/s)": 1.449056
    },
    {
      "epoch": 0.7309027033974551,
      "grad_norm": 7.660039901733398,
      "learning_rate": 9.481948833132597e-05,
      "loss": 2.8795650482177733,
      "memory(GiB)": 66.02,
      "step": 17060,
      "token_acc": 0.41825095057034223,
      "train_speed(iter/s)": 1.449113
    },
    {
      "epoch": 0.731116918726704,
      "grad_norm": 3.8919930458068848,
      "learning_rate": 9.481650484092744e-05,
      "loss": 2.5224071502685548,
      "memory(GiB)": 66.02,
      "step": 17065,
      "token_acc": 0.5021645021645021,
      "train_speed(iter/s)": 1.449143
    },
    {
      "epoch": 0.7313311340559531,
      "grad_norm": 3.0317535400390625,
      "learning_rate": 9.481352053863585e-05,
      "loss": 2.791655921936035,
      "memory(GiB)": 66.02,
      "step": 17070,
      "token_acc": 0.4423076923076923,
      "train_speed(iter/s)": 1.449188
    },
    {
      "epoch": 0.731545349385202,
      "grad_norm": 5.536355972290039,
      "learning_rate": 9.481053542450533e-05,
      "loss": 2.79862003326416,
      "memory(GiB)": 66.02,
      "step": 17075,
      "token_acc": 0.43130990415335463,
      "train_speed(iter/s)": 1.449188
    },
    {
      "epoch": 0.7317595647144509,
      "grad_norm": 4.253134727478027,
      "learning_rate": 9.480754949858992e-05,
      "loss": 2.510112762451172,
      "memory(GiB)": 66.02,
      "step": 17080,
      "token_acc": 0.4689922480620155,
      "train_speed(iter/s)": 1.449218
    },
    {
      "epoch": 0.7319737800437,
      "grad_norm": 5.482084274291992,
      "learning_rate": 9.480456276094372e-05,
      "loss": 2.4189802169799806,
      "memory(GiB)": 66.02,
      "step": 17085,
      "token_acc": 0.4541984732824427,
      "train_speed(iter/s)": 1.449333
    },
    {
      "epoch": 0.7321879953729489,
      "grad_norm": 4.540184020996094,
      "learning_rate": 9.480157521162084e-05,
      "loss": 2.663313293457031,
      "memory(GiB)": 66.02,
      "step": 17090,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.449373
    },
    {
      "epoch": 0.7324022107021978,
      "grad_norm": 4.649466514587402,
      "learning_rate": 9.479858685067538e-05,
      "loss": 2.7879491806030274,
      "memory(GiB)": 66.02,
      "step": 17095,
      "token_acc": 0.4125,
      "train_speed(iter/s)": 1.449412
    },
    {
      "epoch": 0.7326164260314468,
      "grad_norm": 3.7161691188812256,
      "learning_rate": 9.479559767816151e-05,
      "loss": 2.9131710052490236,
      "memory(GiB)": 66.02,
      "step": 17100,
      "token_acc": 0.44036697247706424,
      "train_speed(iter/s)": 1.449286
    },
    {
      "epoch": 0.7328306413606958,
      "grad_norm": 4.126161098480225,
      "learning_rate": 9.479260769413335e-05,
      "loss": 2.775347900390625,
      "memory(GiB)": 66.02,
      "step": 17105,
      "token_acc": 0.44625407166123776,
      "train_speed(iter/s)": 1.44938
    },
    {
      "epoch": 0.7330448566899447,
      "grad_norm": 3.6814563274383545,
      "learning_rate": 9.47896168986451e-05,
      "loss": 2.5244625091552733,
      "memory(GiB)": 66.02,
      "step": 17110,
      "token_acc": 0.4601449275362319,
      "train_speed(iter/s)": 1.449404
    },
    {
      "epoch": 0.7332590720191937,
      "grad_norm": 5.023619174957275,
      "learning_rate": 9.478662529175091e-05,
      "loss": 2.7300262451171875,
      "memory(GiB)": 66.02,
      "step": 17115,
      "token_acc": 0.45544554455445546,
      "train_speed(iter/s)": 1.44941
    },
    {
      "epoch": 0.7334732873484426,
      "grad_norm": 3.727031707763672,
      "learning_rate": 9.4783632873505e-05,
      "loss": 2.1511747360229494,
      "memory(GiB)": 66.02,
      "step": 17120,
      "token_acc": 0.5038759689922481,
      "train_speed(iter/s)": 1.449482
    },
    {
      "epoch": 0.7336875026776916,
      "grad_norm": 6.000275135040283,
      "learning_rate": 9.478063964396156e-05,
      "loss": 2.6481086730957033,
      "memory(GiB)": 66.02,
      "step": 17125,
      "token_acc": 0.4396887159533074,
      "train_speed(iter/s)": 1.449574
    },
    {
      "epoch": 0.7339017180069406,
      "grad_norm": 3.840193033218384,
      "learning_rate": 9.477764560317483e-05,
      "loss": 2.6058633804321287,
      "memory(GiB)": 66.02,
      "step": 17130,
      "token_acc": 0.4669421487603306,
      "train_speed(iter/s)": 1.449719
    },
    {
      "epoch": 0.7341159333361895,
      "grad_norm": 4.2880754470825195,
      "learning_rate": 9.477465075119905e-05,
      "loss": 2.7650218963623048,
      "memory(GiB)": 66.02,
      "step": 17135,
      "token_acc": 0.43086816720257237,
      "train_speed(iter/s)": 1.449769
    },
    {
      "epoch": 0.7343301486654386,
      "grad_norm": 4.753639221191406,
      "learning_rate": 9.477165508808847e-05,
      "loss": 2.8226192474365233,
      "memory(GiB)": 66.02,
      "step": 17140,
      "token_acc": 0.43050847457627117,
      "train_speed(iter/s)": 1.449841
    },
    {
      "epoch": 0.7345443639946875,
      "grad_norm": 3.7705907821655273,
      "learning_rate": 9.476865861389735e-05,
      "loss": 2.580490303039551,
      "memory(GiB)": 66.02,
      "step": 17145,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.449804
    },
    {
      "epoch": 0.7347585793239364,
      "grad_norm": 4.285682678222656,
      "learning_rate": 9.476566132867999e-05,
      "loss": 2.20565242767334,
      "memory(GiB)": 66.02,
      "step": 17150,
      "token_acc": 0.5225225225225225,
      "train_speed(iter/s)": 1.449832
    },
    {
      "epoch": 0.7349727946531854,
      "grad_norm": 3.546729564666748,
      "learning_rate": 9.476266323249068e-05,
      "loss": 2.4750633239746094,
      "memory(GiB)": 66.02,
      "step": 17155,
      "token_acc": 0.5203761755485894,
      "train_speed(iter/s)": 1.449815
    },
    {
      "epoch": 0.7351870099824344,
      "grad_norm": 4.114686489105225,
      "learning_rate": 9.475966432538372e-05,
      "loss": 2.786367988586426,
      "memory(GiB)": 66.02,
      "step": 17160,
      "token_acc": 0.3859060402684564,
      "train_speed(iter/s)": 1.449752
    },
    {
      "epoch": 0.7354012253116833,
      "grad_norm": 4.056107044219971,
      "learning_rate": 9.475666460741348e-05,
      "loss": 2.5953720092773436,
      "memory(GiB)": 66.02,
      "step": 17165,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.449773
    },
    {
      "epoch": 0.7356154406409323,
      "grad_norm": 4.376646518707275,
      "learning_rate": 9.475366407863427e-05,
      "loss": 2.645228385925293,
      "memory(GiB)": 66.02,
      "step": 17170,
      "token_acc": 0.47230320699708456,
      "train_speed(iter/s)": 1.449742
    },
    {
      "epoch": 0.7358296559701812,
      "grad_norm": 5.003983974456787,
      "learning_rate": 9.475066273910043e-05,
      "loss": 2.7561511993408203,
      "memory(GiB)": 66.02,
      "step": 17175,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.449773
    },
    {
      "epoch": 0.7360438712994302,
      "grad_norm": 3.858750343322754,
      "learning_rate": 9.474766058886637e-05,
      "loss": 2.612289047241211,
      "memory(GiB)": 66.02,
      "step": 17180,
      "token_acc": 0.49085365853658536,
      "train_speed(iter/s)": 1.449818
    },
    {
      "epoch": 0.7362580866286792,
      "grad_norm": 5.102039337158203,
      "learning_rate": 9.474465762798646e-05,
      "loss": 2.579855728149414,
      "memory(GiB)": 66.02,
      "step": 17185,
      "token_acc": 0.46099290780141844,
      "train_speed(iter/s)": 1.449892
    },
    {
      "epoch": 0.7364723019579281,
      "grad_norm": 4.245791912078857,
      "learning_rate": 9.474165385651511e-05,
      "loss": 2.4613508224487304,
      "memory(GiB)": 66.02,
      "step": 17190,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.449868
    },
    {
      "epoch": 0.736686517287177,
      "grad_norm": 5.47557258605957,
      "learning_rate": 9.473864927450673e-05,
      "loss": 2.4753299713134767,
      "memory(GiB)": 66.02,
      "step": 17195,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.449868
    },
    {
      "epoch": 0.7369007326164261,
      "grad_norm": 4.8811869621276855,
      "learning_rate": 9.473564388201575e-05,
      "loss": 2.5692346572875975,
      "memory(GiB)": 66.02,
      "step": 17200,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.449913
    },
    {
      "epoch": 0.737114947945675,
      "grad_norm": 4.696002006530762,
      "learning_rate": 9.473263767909659e-05,
      "loss": 2.8190792083740233,
      "memory(GiB)": 66.02,
      "step": 17205,
      "token_acc": 0.4504792332268371,
      "train_speed(iter/s)": 1.449948
    },
    {
      "epoch": 0.7373291632749239,
      "grad_norm": 4.203622341156006,
      "learning_rate": 9.472963066580375e-05,
      "loss": 3.2341758728027346,
      "memory(GiB)": 66.02,
      "step": 17210,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.450074
    },
    {
      "epoch": 0.737543378604173,
      "grad_norm": 4.111051559448242,
      "learning_rate": 9.472662284219171e-05,
      "loss": 2.8425106048583983,
      "memory(GiB)": 66.02,
      "step": 17215,
      "token_acc": 0.43260188087774293,
      "train_speed(iter/s)": 1.450163
    },
    {
      "epoch": 0.7377575939334219,
      "grad_norm": 4.058981418609619,
      "learning_rate": 9.472361420831493e-05,
      "loss": 2.6882980346679686,
      "memory(GiB)": 66.02,
      "step": 17220,
      "token_acc": 0.46273291925465837,
      "train_speed(iter/s)": 1.450247
    },
    {
      "epoch": 0.7379718092626708,
      "grad_norm": 4.370203495025635,
      "learning_rate": 9.472060476422791e-05,
      "loss": 2.318088722229004,
      "memory(GiB)": 66.02,
      "step": 17225,
      "token_acc": 0.4954682779456193,
      "train_speed(iter/s)": 1.450205
    },
    {
      "epoch": 0.7381860245919198,
      "grad_norm": 4.204536437988281,
      "learning_rate": 9.471759450998521e-05,
      "loss": 2.496956634521484,
      "memory(GiB)": 66.02,
      "step": 17230,
      "token_acc": 0.4309210526315789,
      "train_speed(iter/s)": 1.450172
    },
    {
      "epoch": 0.7384002399211688,
      "grad_norm": 4.768651008605957,
      "learning_rate": 9.471458344564132e-05,
      "loss": 2.6269115447998046,
      "memory(GiB)": 66.02,
      "step": 17235,
      "token_acc": 0.4641638225255973,
      "train_speed(iter/s)": 1.450207
    },
    {
      "epoch": 0.7386144552504177,
      "grad_norm": 4.9118781089782715,
      "learning_rate": 9.47115715712508e-05,
      "loss": 2.758819007873535,
      "memory(GiB)": 66.02,
      "step": 17240,
      "token_acc": 0.43042071197411,
      "train_speed(iter/s)": 1.450172
    },
    {
      "epoch": 0.7388286705796667,
      "grad_norm": 4.1345744132995605,
      "learning_rate": 9.470855888686822e-05,
      "loss": 2.638313102722168,
      "memory(GiB)": 66.02,
      "step": 17245,
      "token_acc": 0.4491017964071856,
      "train_speed(iter/s)": 1.450197
    },
    {
      "epoch": 0.7390428859089156,
      "grad_norm": 4.762982368469238,
      "learning_rate": 9.470554539254816e-05,
      "loss": 2.7447622299194334,
      "memory(GiB)": 66.02,
      "step": 17250,
      "token_acc": 0.44281524926686217,
      "train_speed(iter/s)": 1.450241
    },
    {
      "epoch": 0.7392571012381646,
      "grad_norm": 4.486371040344238,
      "learning_rate": 9.470253108834522e-05,
      "loss": 2.7322099685668944,
      "memory(GiB)": 66.02,
      "step": 17255,
      "token_acc": 0.41968911917098445,
      "train_speed(iter/s)": 1.450174
    },
    {
      "epoch": 0.7394713165674136,
      "grad_norm": 3.7400050163269043,
      "learning_rate": 9.469951597431398e-05,
      "loss": 2.816219711303711,
      "memory(GiB)": 66.02,
      "step": 17260,
      "token_acc": 0.4208860759493671,
      "train_speed(iter/s)": 1.450182
    },
    {
      "epoch": 0.7396855318966625,
      "grad_norm": 6.051055431365967,
      "learning_rate": 9.469650005050907e-05,
      "loss": 2.57449951171875,
      "memory(GiB)": 66.02,
      "step": 17265,
      "token_acc": 0.47470817120622566,
      "train_speed(iter/s)": 1.450168
    },
    {
      "epoch": 0.7398997472259115,
      "grad_norm": 5.436326503753662,
      "learning_rate": 9.469348331698516e-05,
      "loss": 2.6801521301269533,
      "memory(GiB)": 66.02,
      "step": 17270,
      "token_acc": 0.4186746987951807,
      "train_speed(iter/s)": 1.450144
    },
    {
      "epoch": 0.7401139625551605,
      "grad_norm": 4.679691314697266,
      "learning_rate": 9.469046577379685e-05,
      "loss": 2.7199474334716798,
      "memory(GiB)": 66.02,
      "step": 17275,
      "token_acc": 0.4339622641509434,
      "train_speed(iter/s)": 1.450258
    },
    {
      "epoch": 0.7403281778844094,
      "grad_norm": 5.3936686515808105,
      "learning_rate": 9.468744742099885e-05,
      "loss": 2.459560012817383,
      "memory(GiB)": 66.02,
      "step": 17280,
      "token_acc": 0.5078125,
      "train_speed(iter/s)": 1.45026
    },
    {
      "epoch": 0.7405423932136583,
      "grad_norm": 4.026109218597412,
      "learning_rate": 9.468442825864578e-05,
      "loss": 2.6350627899169923,
      "memory(GiB)": 66.02,
      "step": 17285,
      "token_acc": 0.45724907063197023,
      "train_speed(iter/s)": 1.450334
    },
    {
      "epoch": 0.7407566085429074,
      "grad_norm": 3.9744529724121094,
      "learning_rate": 9.468140828679242e-05,
      "loss": 2.5128152847290037,
      "memory(GiB)": 66.02,
      "step": 17290,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.450349
    },
    {
      "epoch": 0.7409708238721563,
      "grad_norm": 4.821835041046143,
      "learning_rate": 9.467838750549341e-05,
      "loss": 2.7504316329956056,
      "memory(GiB)": 66.02,
      "step": 17295,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.45032
    },
    {
      "epoch": 0.7411850392014052,
      "grad_norm": 4.159036159515381,
      "learning_rate": 9.46753659148035e-05,
      "loss": 2.8426998138427733,
      "memory(GiB)": 66.02,
      "step": 17300,
      "token_acc": 0.4271523178807947,
      "train_speed(iter/s)": 1.450319
    },
    {
      "epoch": 0.7413992545306543,
      "grad_norm": 5.1360554695129395,
      "learning_rate": 9.467234351477743e-05,
      "loss": 2.6242515563964846,
      "memory(GiB)": 66.02,
      "step": 17305,
      "token_acc": 0.42452830188679247,
      "train_speed(iter/s)": 1.45043
    },
    {
      "epoch": 0.7416134698599032,
      "grad_norm": 4.638542175292969,
      "learning_rate": 9.466932030546996e-05,
      "loss": 2.698273849487305,
      "memory(GiB)": 66.02,
      "step": 17310,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.450486
    },
    {
      "epoch": 0.7418276851891521,
      "grad_norm": 3.99564266204834,
      "learning_rate": 9.466629628693584e-05,
      "loss": 2.454903984069824,
      "memory(GiB)": 66.02,
      "step": 17315,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.45057
    },
    {
      "epoch": 0.7420419005184011,
      "grad_norm": 4.749660968780518,
      "learning_rate": 9.466327145922986e-05,
      "loss": 2.6469778060913085,
      "memory(GiB)": 66.02,
      "step": 17320,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.45059
    },
    {
      "epoch": 0.74225611584765,
      "grad_norm": 4.526060581207275,
      "learning_rate": 9.466024582240683e-05,
      "loss": 2.284941482543945,
      "memory(GiB)": 66.02,
      "step": 17325,
      "token_acc": 0.5019607843137255,
      "train_speed(iter/s)": 1.450661
    },
    {
      "epoch": 0.742470331176899,
      "grad_norm": 7.708323001861572,
      "learning_rate": 9.465721937652155e-05,
      "loss": 2.456673812866211,
      "memory(GiB)": 66.02,
      "step": 17330,
      "token_acc": 0.4979919678714859,
      "train_speed(iter/s)": 1.450554
    },
    {
      "epoch": 0.742684546506148,
      "grad_norm": 4.336303234100342,
      "learning_rate": 9.465419212162883e-05,
      "loss": 2.5639923095703123,
      "memory(GiB)": 66.02,
      "step": 17335,
      "token_acc": 0.47468354430379744,
      "train_speed(iter/s)": 1.450559
    },
    {
      "epoch": 0.7428987618353969,
      "grad_norm": 3.728372812271118,
      "learning_rate": 9.465116405778357e-05,
      "loss": 2.7540443420410154,
      "memory(GiB)": 66.02,
      "step": 17340,
      "token_acc": 0.47410358565737054,
      "train_speed(iter/s)": 1.450645
    },
    {
      "epoch": 0.7431129771646459,
      "grad_norm": 5.069708824157715,
      "learning_rate": 9.464813518504055e-05,
      "loss": 2.604481506347656,
      "memory(GiB)": 66.02,
      "step": 17345,
      "token_acc": 0.44140625,
      "train_speed(iter/s)": 1.450759
    },
    {
      "epoch": 0.7433271924938949,
      "grad_norm": 4.775428771972656,
      "learning_rate": 9.464510550345469e-05,
      "loss": 2.732996368408203,
      "memory(GiB)": 66.02,
      "step": 17350,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.450796
    },
    {
      "epoch": 0.7435414078231438,
      "grad_norm": 4.489157199859619,
      "learning_rate": 9.464207501308086e-05,
      "loss": 2.366771697998047,
      "memory(GiB)": 66.02,
      "step": 17355,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.450912
    },
    {
      "epoch": 0.7437556231523927,
      "grad_norm": 4.524479389190674,
      "learning_rate": 9.463904371397397e-05,
      "loss": 2.513582229614258,
      "memory(GiB)": 66.02,
      "step": 17360,
      "token_acc": 0.48249027237354086,
      "train_speed(iter/s)": 1.450915
    },
    {
      "epoch": 0.7439698384816418,
      "grad_norm": 4.527760982513428,
      "learning_rate": 9.463601160618892e-05,
      "loss": 2.7093578338623048,
      "memory(GiB)": 66.02,
      "step": 17365,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.450911
    },
    {
      "epoch": 0.7441840538108907,
      "grad_norm": 5.473022937774658,
      "learning_rate": 9.463297868978066e-05,
      "loss": 2.577850341796875,
      "memory(GiB)": 66.02,
      "step": 17370,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.450963
    },
    {
      "epoch": 0.7443982691401396,
      "grad_norm": 5.752939701080322,
      "learning_rate": 9.462994496480411e-05,
      "loss": 2.5562368392944337,
      "memory(GiB)": 66.02,
      "step": 17375,
      "token_acc": 0.4707792207792208,
      "train_speed(iter/s)": 1.450881
    },
    {
      "epoch": 0.7446124844693887,
      "grad_norm": 4.581209659576416,
      "learning_rate": 9.462691043131424e-05,
      "loss": 2.479447364807129,
      "memory(GiB)": 66.02,
      "step": 17380,
      "token_acc": 0.5126353790613718,
      "train_speed(iter/s)": 1.450983
    },
    {
      "epoch": 0.7448266997986376,
      "grad_norm": 5.816927909851074,
      "learning_rate": 9.462387508936605e-05,
      "loss": 2.7149803161621096,
      "memory(GiB)": 66.02,
      "step": 17385,
      "token_acc": 0.4235294117647059,
      "train_speed(iter/s)": 1.450978
    },
    {
      "epoch": 0.7450409151278865,
      "grad_norm": 4.131303310394287,
      "learning_rate": 9.462083893901448e-05,
      "loss": 2.7552743911743165,
      "memory(GiB)": 66.02,
      "step": 17390,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.450967
    },
    {
      "epoch": 0.7452551304571355,
      "grad_norm": 4.87317419052124,
      "learning_rate": 9.461780198031456e-05,
      "loss": 2.588461685180664,
      "memory(GiB)": 66.02,
      "step": 17395,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.450928
    },
    {
      "epoch": 0.7454693457863845,
      "grad_norm": 6.666911602020264,
      "learning_rate": 9.46147642133213e-05,
      "loss": 2.7245845794677734,
      "memory(GiB)": 66.02,
      "step": 17400,
      "token_acc": 0.43389830508474575,
      "train_speed(iter/s)": 1.451009
    },
    {
      "epoch": 0.7456835611156334,
      "grad_norm": 3.6350271701812744,
      "learning_rate": 9.461172563808973e-05,
      "loss": 2.5686161041259767,
      "memory(GiB)": 66.02,
      "step": 17405,
      "token_acc": 0.4646153846153846,
      "train_speed(iter/s)": 1.451058
    },
    {
      "epoch": 0.7458977764448824,
      "grad_norm": 4.5575666427612305,
      "learning_rate": 9.46086862546749e-05,
      "loss": 2.672595977783203,
      "memory(GiB)": 66.02,
      "step": 17410,
      "token_acc": 0.45317220543806647,
      "train_speed(iter/s)": 1.451049
    },
    {
      "epoch": 0.7461119917741313,
      "grad_norm": 3.5397253036499023,
      "learning_rate": 9.460564606313189e-05,
      "loss": 2.6991245269775392,
      "memory(GiB)": 66.02,
      "step": 17415,
      "token_acc": 0.45930232558139533,
      "train_speed(iter/s)": 1.451117
    },
    {
      "epoch": 0.7463262071033803,
      "grad_norm": 3.950761556625366,
      "learning_rate": 9.460260506351573e-05,
      "loss": 2.5688545227050783,
      "memory(GiB)": 66.02,
      "step": 17420,
      "token_acc": 0.43582089552238806,
      "train_speed(iter/s)": 1.451096
    },
    {
      "epoch": 0.7465404224326293,
      "grad_norm": 4.27651834487915,
      "learning_rate": 9.459956325588155e-05,
      "loss": 2.6067344665527346,
      "memory(GiB)": 66.02,
      "step": 17425,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.451137
    },
    {
      "epoch": 0.7467546377618782,
      "grad_norm": 4.900177478790283,
      "learning_rate": 9.459652064028445e-05,
      "loss": 2.91363468170166,
      "memory(GiB)": 66.02,
      "step": 17430,
      "token_acc": 0.4246987951807229,
      "train_speed(iter/s)": 1.451214
    },
    {
      "epoch": 0.7469688530911271,
      "grad_norm": 4.552701473236084,
      "learning_rate": 9.459347721677954e-05,
      "loss": 2.4086576461791993,
      "memory(GiB)": 66.02,
      "step": 17435,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.451154
    },
    {
      "epoch": 0.7471830684203762,
      "grad_norm": 6.436448574066162,
      "learning_rate": 9.459043298542196e-05,
      "loss": 2.7376712799072265,
      "memory(GiB)": 66.02,
      "step": 17440,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.451254
    },
    {
      "epoch": 0.7473972837496251,
      "grad_norm": 4.306189060211182,
      "learning_rate": 9.458738794626687e-05,
      "loss": 2.7037752151489256,
      "memory(GiB)": 66.02,
      "step": 17445,
      "token_acc": 0.5109717868338558,
      "train_speed(iter/s)": 1.451335
    },
    {
      "epoch": 0.747611499078874,
      "grad_norm": 3.639421224594116,
      "learning_rate": 9.45843420993694e-05,
      "loss": 2.3604948043823244,
      "memory(GiB)": 66.02,
      "step": 17450,
      "token_acc": 0.4721189591078067,
      "train_speed(iter/s)": 1.451412
    },
    {
      "epoch": 0.7478257144081231,
      "grad_norm": 3.6510393619537354,
      "learning_rate": 9.458129544478476e-05,
      "loss": 2.79281063079834,
      "memory(GiB)": 66.02,
      "step": 17455,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.451391
    },
    {
      "epoch": 0.748039929737372,
      "grad_norm": 4.624139308929443,
      "learning_rate": 9.457824798256813e-05,
      "loss": 2.5438098907470703,
      "memory(GiB)": 66.02,
      "step": 17460,
      "token_acc": 0.4315068493150685,
      "train_speed(iter/s)": 1.451249
    },
    {
      "epoch": 0.7482541450666209,
      "grad_norm": 3.588677406311035,
      "learning_rate": 9.457519971277471e-05,
      "loss": 2.483653259277344,
      "memory(GiB)": 66.02,
      "step": 17465,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.451298
    },
    {
      "epoch": 0.74846836039587,
      "grad_norm": 3.9264121055603027,
      "learning_rate": 9.457215063545974e-05,
      "loss": 2.7714941024780275,
      "memory(GiB)": 66.02,
      "step": 17470,
      "token_acc": 0.4421052631578947,
      "train_speed(iter/s)": 1.4513
    },
    {
      "epoch": 0.7486825757251189,
      "grad_norm": 3.972663164138794,
      "learning_rate": 9.456910075067847e-05,
      "loss": 2.0708208084106445,
      "memory(GiB)": 66.02,
      "step": 17475,
      "token_acc": 0.544,
      "train_speed(iter/s)": 1.451334
    },
    {
      "epoch": 0.7488967910543679,
      "grad_norm": 6.843869209289551,
      "learning_rate": 9.45660500584861e-05,
      "loss": 2.442848014831543,
      "memory(GiB)": 66.02,
      "step": 17480,
      "token_acc": 0.544,
      "train_speed(iter/s)": 1.451375
    },
    {
      "epoch": 0.7491110063836168,
      "grad_norm": 4.040668487548828,
      "learning_rate": 9.456299855893794e-05,
      "loss": 2.41268310546875,
      "memory(GiB)": 66.02,
      "step": 17485,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.451335
    },
    {
      "epoch": 0.7493252217128658,
      "grad_norm": 4.642605781555176,
      "learning_rate": 9.455994625208926e-05,
      "loss": 2.4285951614379884,
      "memory(GiB)": 66.02,
      "step": 17490,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.451439
    },
    {
      "epoch": 0.7495394370421148,
      "grad_norm": 4.087214946746826,
      "learning_rate": 9.455689313799535e-05,
      "loss": 2.445493698120117,
      "memory(GiB)": 66.02,
      "step": 17495,
      "token_acc": 0.46932515337423314,
      "train_speed(iter/s)": 1.451506
    },
    {
      "epoch": 0.7497536523713637,
      "grad_norm": 4.687131404876709,
      "learning_rate": 9.455383921671153e-05,
      "loss": 2.786838912963867,
      "memory(GiB)": 66.02,
      "step": 17500,
      "token_acc": 0.43252595155709345,
      "train_speed(iter/s)": 1.451542
    },
    {
      "epoch": 0.7497536523713637,
      "eval_loss": 2.231870174407959,
      "eval_runtime": 14.8081,
      "eval_samples_per_second": 6.753,
      "eval_steps_per_second": 6.753,
      "eval_token_acc": 0.47354497354497355,
      "step": 17500
    },
    {
      "epoch": 0.7499678677006126,
      "grad_norm": 5.118946552276611,
      "learning_rate": 9.45507844882931e-05,
      "loss": 2.565242385864258,
      "memory(GiB)": 66.02,
      "step": 17505,
      "token_acc": 0.473633748801534,
      "train_speed(iter/s)": 1.449738
    },
    {
      "epoch": 0.7501820830298617,
      "grad_norm": 6.066102027893066,
      "learning_rate": 9.454772895279543e-05,
      "loss": 2.4543426513671873,
      "memory(GiB)": 66.02,
      "step": 17510,
      "token_acc": 0.5101214574898786,
      "train_speed(iter/s)": 1.449789
    },
    {
      "epoch": 0.7503962983591106,
      "grad_norm": 3.9736335277557373,
      "learning_rate": 9.454467261027386e-05,
      "loss": 2.4278087615966797,
      "memory(GiB)": 66.02,
      "step": 17515,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.449808
    },
    {
      "epoch": 0.7506105136883595,
      "grad_norm": 4.78377628326416,
      "learning_rate": 9.454161546078375e-05,
      "loss": 2.2966567993164064,
      "memory(GiB)": 66.02,
      "step": 17520,
      "token_acc": 0.444015444015444,
      "train_speed(iter/s)": 1.449754
    },
    {
      "epoch": 0.7508247290176085,
      "grad_norm": 4.362670421600342,
      "learning_rate": 9.453855750438049e-05,
      "loss": 2.4804916381835938,
      "memory(GiB)": 66.02,
      "step": 17525,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.44969
    },
    {
      "epoch": 0.7510389443468575,
      "grad_norm": 4.259242057800293,
      "learning_rate": 9.45354987411195e-05,
      "loss": 2.792049789428711,
      "memory(GiB)": 66.02,
      "step": 17530,
      "token_acc": 0.45564516129032256,
      "train_speed(iter/s)": 1.449794
    },
    {
      "epoch": 0.7512531596761064,
      "grad_norm": 5.366889476776123,
      "learning_rate": 9.453243917105617e-05,
      "loss": 2.934235954284668,
      "memory(GiB)": 66.02,
      "step": 17535,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.449786
    },
    {
      "epoch": 0.7514673750053554,
      "grad_norm": 4.995511054992676,
      "learning_rate": 9.452937879424592e-05,
      "loss": 2.5973579406738283,
      "memory(GiB)": 66.02,
      "step": 17540,
      "token_acc": 0.44554455445544555,
      "train_speed(iter/s)": 1.449902
    },
    {
      "epoch": 0.7516815903346044,
      "grad_norm": 3.5699355602264404,
      "learning_rate": 9.452631761074419e-05,
      "loss": 2.5971807479858398,
      "memory(GiB)": 66.02,
      "step": 17545,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.449844
    },
    {
      "epoch": 0.7518958056638533,
      "grad_norm": 4.320559024810791,
      "learning_rate": 9.452325562060647e-05,
      "loss": 2.797038269042969,
      "memory(GiB)": 66.02,
      "step": 17550,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.449845
    },
    {
      "epoch": 0.7521100209931023,
      "grad_norm": 4.72542667388916,
      "learning_rate": 9.45201928238882e-05,
      "loss": 2.6849876403808595,
      "memory(GiB)": 66.02,
      "step": 17555,
      "token_acc": 0.4404332129963899,
      "train_speed(iter/s)": 1.449862
    },
    {
      "epoch": 0.7523242363223512,
      "grad_norm": 7.0616936683654785,
      "learning_rate": 9.451712922064488e-05,
      "loss": 2.8545696258544924,
      "memory(GiB)": 66.02,
      "step": 17560,
      "token_acc": 0.3981191222570533,
      "train_speed(iter/s)": 1.449808
    },
    {
      "epoch": 0.7525384516516002,
      "grad_norm": 4.490034103393555,
      "learning_rate": 9.4514064810932e-05,
      "loss": 2.809182357788086,
      "memory(GiB)": 66.02,
      "step": 17565,
      "token_acc": 0.4652567975830816,
      "train_speed(iter/s)": 1.449811
    },
    {
      "epoch": 0.7527526669808492,
      "grad_norm": 3.929772138595581,
      "learning_rate": 9.451099959480507e-05,
      "loss": 2.648844909667969,
      "memory(GiB)": 66.02,
      "step": 17570,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.449763
    },
    {
      "epoch": 0.7529668823100981,
      "grad_norm": 3.9892163276672363,
      "learning_rate": 9.450793357231962e-05,
      "loss": 2.5053421020507813,
      "memory(GiB)": 66.02,
      "step": 17575,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.449791
    },
    {
      "epoch": 0.753181097639347,
      "grad_norm": 4.415933132171631,
      "learning_rate": 9.450486674353122e-05,
      "loss": 2.777522087097168,
      "memory(GiB)": 66.02,
      "step": 17580,
      "token_acc": 0.45514950166112955,
      "train_speed(iter/s)": 1.449786
    },
    {
      "epoch": 0.7533953129685961,
      "grad_norm": 4.361843109130859,
      "learning_rate": 9.450179910849541e-05,
      "loss": 2.814442825317383,
      "memory(GiB)": 66.02,
      "step": 17585,
      "token_acc": 0.43130990415335463,
      "train_speed(iter/s)": 1.449747
    },
    {
      "epoch": 0.753609528297845,
      "grad_norm": 4.309413433074951,
      "learning_rate": 9.449873066726775e-05,
      "loss": 2.5275644302368163,
      "memory(GiB)": 66.02,
      "step": 17590,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.449846
    },
    {
      "epoch": 0.7538237436270939,
      "grad_norm": 4.586771011352539,
      "learning_rate": 9.449566141990384e-05,
      "loss": 2.312491607666016,
      "memory(GiB)": 66.02,
      "step": 17595,
      "token_acc": 0.4847328244274809,
      "train_speed(iter/s)": 1.44991
    },
    {
      "epoch": 0.754037958956343,
      "grad_norm": 4.46777868270874,
      "learning_rate": 9.449259136645929e-05,
      "loss": 2.7186101913452148,
      "memory(GiB)": 66.02,
      "step": 17600,
      "token_acc": 0.45666666666666667,
      "train_speed(iter/s)": 1.44994
    },
    {
      "epoch": 0.7542521742855919,
      "grad_norm": 4.810561180114746,
      "learning_rate": 9.448952050698972e-05,
      "loss": 2.4807811737060548,
      "memory(GiB)": 66.02,
      "step": 17605,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.45005
    },
    {
      "epoch": 0.7544663896148408,
      "grad_norm": 6.472301959991455,
      "learning_rate": 9.448644884155075e-05,
      "loss": 2.7250158309936525,
      "memory(GiB)": 66.02,
      "step": 17610,
      "token_acc": 0.4765840220385675,
      "train_speed(iter/s)": 1.450145
    },
    {
      "epoch": 0.7546806049440898,
      "grad_norm": 3.7796828746795654,
      "learning_rate": 9.448337637019801e-05,
      "loss": 2.7196197509765625,
      "memory(GiB)": 66.02,
      "step": 17615,
      "token_acc": 0.45132743362831856,
      "train_speed(iter/s)": 1.450192
    },
    {
      "epoch": 0.7548948202733388,
      "grad_norm": 4.772680282592773,
      "learning_rate": 9.44803030929872e-05,
      "loss": 2.7956024169921876,
      "memory(GiB)": 66.02,
      "step": 17620,
      "token_acc": 0.4255952380952381,
      "train_speed(iter/s)": 1.450212
    },
    {
      "epoch": 0.7551090356025877,
      "grad_norm": 5.025880336761475,
      "learning_rate": 9.447722900997396e-05,
      "loss": 2.634055328369141,
      "memory(GiB)": 66.02,
      "step": 17625,
      "token_acc": 0.4357142857142857,
      "train_speed(iter/s)": 1.450299
    },
    {
      "epoch": 0.7553232509318367,
      "grad_norm": 4.180508613586426,
      "learning_rate": 9.447415412121399e-05,
      "loss": 2.4728137969970705,
      "memory(GiB)": 66.02,
      "step": 17630,
      "token_acc": 0.467680608365019,
      "train_speed(iter/s)": 1.450384
    },
    {
      "epoch": 0.7555374662610856,
      "grad_norm": 7.413858413696289,
      "learning_rate": 9.4471078426763e-05,
      "loss": 2.812285232543945,
      "memory(GiB)": 66.02,
      "step": 17635,
      "token_acc": 0.4222972972972973,
      "train_speed(iter/s)": 1.450437
    },
    {
      "epoch": 0.7557516815903346,
      "grad_norm": 3.3859031200408936,
      "learning_rate": 9.44680019266767e-05,
      "loss": 2.370113563537598,
      "memory(GiB)": 66.02,
      "step": 17640,
      "token_acc": 0.5382165605095541,
      "train_speed(iter/s)": 1.45046
    },
    {
      "epoch": 0.7559658969195836,
      "grad_norm": 6.244542121887207,
      "learning_rate": 9.446492462101087e-05,
      "loss": 2.546002006530762,
      "memory(GiB)": 66.02,
      "step": 17645,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.450616
    },
    {
      "epoch": 0.7561801122488325,
      "grad_norm": 3.882312297821045,
      "learning_rate": 9.446184650982117e-05,
      "loss": 2.5146961212158203,
      "memory(GiB)": 66.02,
      "step": 17650,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.45068
    },
    {
      "epoch": 0.7563943275780814,
      "grad_norm": 5.0457940101623535,
      "learning_rate": 9.445876759316344e-05,
      "loss": 2.7844831466674806,
      "memory(GiB)": 66.02,
      "step": 17655,
      "token_acc": 0.43434343434343436,
      "train_speed(iter/s)": 1.450762
    },
    {
      "epoch": 0.7566085429073305,
      "grad_norm": 6.0084943771362305,
      "learning_rate": 9.445568787109342e-05,
      "loss": 2.819187545776367,
      "memory(GiB)": 66.02,
      "step": 17660,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.450809
    },
    {
      "epoch": 0.7568227582365794,
      "grad_norm": 4.842067241668701,
      "learning_rate": 9.445260734366692e-05,
      "loss": 2.751005172729492,
      "memory(GiB)": 66.02,
      "step": 17665,
      "token_acc": 0.4509090909090909,
      "train_speed(iter/s)": 1.450824
    },
    {
      "epoch": 0.7570369735658283,
      "grad_norm": 3.4147021770477295,
      "learning_rate": 9.444952601093972e-05,
      "loss": 2.756491470336914,
      "memory(GiB)": 66.02,
      "step": 17670,
      "token_acc": 0.4440993788819876,
      "train_speed(iter/s)": 1.450816
    },
    {
      "epoch": 0.7572511888950774,
      "grad_norm": 5.491716384887695,
      "learning_rate": 9.444644387296767e-05,
      "loss": 2.3792045593261717,
      "memory(GiB)": 66.02,
      "step": 17675,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.450758
    },
    {
      "epoch": 0.7574654042243263,
      "grad_norm": 3.91731333732605,
      "learning_rate": 9.444336092980661e-05,
      "loss": 2.4315948486328125,
      "memory(GiB)": 66.02,
      "step": 17680,
      "token_acc": 0.5109034267912772,
      "train_speed(iter/s)": 1.450777
    },
    {
      "epoch": 0.7576796195535752,
      "grad_norm": 4.7529401779174805,
      "learning_rate": 9.444027718151235e-05,
      "loss": 2.8342124938964846,
      "memory(GiB)": 66.02,
      "step": 17685,
      "token_acc": 0.4432624113475177,
      "train_speed(iter/s)": 1.450729
    },
    {
      "epoch": 0.7578938348828242,
      "grad_norm": 4.680537223815918,
      "learning_rate": 9.443719262814079e-05,
      "loss": 2.5661392211914062,
      "memory(GiB)": 66.02,
      "step": 17690,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.450765
    },
    {
      "epoch": 0.7581080502120732,
      "grad_norm": 3.670809268951416,
      "learning_rate": 9.44341072697478e-05,
      "loss": 2.448692512512207,
      "memory(GiB)": 66.02,
      "step": 17695,
      "token_acc": 0.47384615384615386,
      "train_speed(iter/s)": 1.450862
    },
    {
      "epoch": 0.7583222655413221,
      "grad_norm": 3.857788324356079,
      "learning_rate": 9.443102110638928e-05,
      "loss": 3.052908515930176,
      "memory(GiB)": 66.02,
      "step": 17700,
      "token_acc": 0.39751552795031053,
      "train_speed(iter/s)": 1.450919
    },
    {
      "epoch": 0.7585364808705711,
      "grad_norm": 5.057369709014893,
      "learning_rate": 9.442793413812113e-05,
      "loss": 2.669346809387207,
      "memory(GiB)": 66.02,
      "step": 17705,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.450841
    },
    {
      "epoch": 0.75875069619982,
      "grad_norm": 4.74763298034668,
      "learning_rate": 9.442484636499927e-05,
      "loss": 2.49759521484375,
      "memory(GiB)": 66.02,
      "step": 17710,
      "token_acc": 0.4380664652567976,
      "train_speed(iter/s)": 1.45092
    },
    {
      "epoch": 0.758964911529069,
      "grad_norm": 4.796550273895264,
      "learning_rate": 9.442175778707965e-05,
      "loss": 2.521480941772461,
      "memory(GiB)": 66.02,
      "step": 17715,
      "token_acc": 0.45126353790613716,
      "train_speed(iter/s)": 1.450918
    },
    {
      "epoch": 0.759179126858318,
      "grad_norm": 4.271181106567383,
      "learning_rate": 9.441866840441822e-05,
      "loss": 2.573577880859375,
      "memory(GiB)": 66.02,
      "step": 17720,
      "token_acc": 0.4794007490636704,
      "train_speed(iter/s)": 1.450977
    },
    {
      "epoch": 0.7593933421875669,
      "grad_norm": 5.223025798797607,
      "learning_rate": 9.441557821707094e-05,
      "loss": 2.578267478942871,
      "memory(GiB)": 66.02,
      "step": 17725,
      "token_acc": 0.44727272727272727,
      "train_speed(iter/s)": 1.451081
    },
    {
      "epoch": 0.7596075575168159,
      "grad_norm": 4.418719291687012,
      "learning_rate": 9.44124872250938e-05,
      "loss": 2.6255107879638673,
      "memory(GiB)": 66.02,
      "step": 17730,
      "token_acc": 0.4494047619047619,
      "train_speed(iter/s)": 1.451117
    },
    {
      "epoch": 0.7598217728460649,
      "grad_norm": 3.666797637939453,
      "learning_rate": 9.440939542854278e-05,
      "loss": 2.6461490631103515,
      "memory(GiB)": 66.02,
      "step": 17735,
      "token_acc": 0.4421364985163205,
      "train_speed(iter/s)": 1.451094
    },
    {
      "epoch": 0.7600359881753138,
      "grad_norm": 4.563016891479492,
      "learning_rate": 9.440630282747392e-05,
      "loss": 2.5761444091796877,
      "memory(GiB)": 66.02,
      "step": 17740,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.451114
    },
    {
      "epoch": 0.7602502035045627,
      "grad_norm": 4.366426944732666,
      "learning_rate": 9.44032094219432e-05,
      "loss": 2.7161720275878904,
      "memory(GiB)": 66.02,
      "step": 17745,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.451125
    },
    {
      "epoch": 0.7604644188338118,
      "grad_norm": 5.017696380615234,
      "learning_rate": 9.440011521200671e-05,
      "loss": 2.853174591064453,
      "memory(GiB)": 66.02,
      "step": 17750,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.451034
    },
    {
      "epoch": 0.7606786341630607,
      "grad_norm": 4.0472235679626465,
      "learning_rate": 9.439702019772047e-05,
      "loss": 2.7683490753173827,
      "memory(GiB)": 66.02,
      "step": 17755,
      "token_acc": 0.4542857142857143,
      "train_speed(iter/s)": 1.451026
    },
    {
      "epoch": 0.7608928494923096,
      "grad_norm": 6.4518256187438965,
      "learning_rate": 9.439392437914058e-05,
      "loss": 2.53245849609375,
      "memory(GiB)": 66.02,
      "step": 17760,
      "token_acc": 0.4222222222222222,
      "train_speed(iter/s)": 1.451057
    },
    {
      "epoch": 0.7611070648215587,
      "grad_norm": 4.387604236602783,
      "learning_rate": 9.439082775632309e-05,
      "loss": 2.4662708282470702,
      "memory(GiB)": 66.02,
      "step": 17765,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.451129
    },
    {
      "epoch": 0.7613212801508076,
      "grad_norm": 6.294944763183594,
      "learning_rate": 9.43877303293241e-05,
      "loss": 2.5766059875488283,
      "memory(GiB)": 66.02,
      "step": 17770,
      "token_acc": 0.42452830188679247,
      "train_speed(iter/s)": 1.451235
    },
    {
      "epoch": 0.7615354954800565,
      "grad_norm": 5.089325428009033,
      "learning_rate": 9.438463209819976e-05,
      "loss": 2.781337356567383,
      "memory(GiB)": 66.02,
      "step": 17775,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.451227
    },
    {
      "epoch": 0.7617497108093055,
      "grad_norm": 5.062034606933594,
      "learning_rate": 9.438153306300616e-05,
      "loss": 2.726021575927734,
      "memory(GiB)": 66.02,
      "step": 17780,
      "token_acc": 0.4509090909090909,
      "train_speed(iter/s)": 1.451221
    },
    {
      "epoch": 0.7619639261385545,
      "grad_norm": 5.949607849121094,
      "learning_rate": 9.437843322379947e-05,
      "loss": 2.318865966796875,
      "memory(GiB)": 66.02,
      "step": 17785,
      "token_acc": 0.5179282868525896,
      "train_speed(iter/s)": 1.451305
    },
    {
      "epoch": 0.7621781414678034,
      "grad_norm": 3.746591091156006,
      "learning_rate": 9.437533258063582e-05,
      "loss": 3.1004989624023436,
      "memory(GiB)": 66.02,
      "step": 17790,
      "token_acc": 0.4119496855345912,
      "train_speed(iter/s)": 1.451344
    },
    {
      "epoch": 0.7623923567970524,
      "grad_norm": 5.227212429046631,
      "learning_rate": 9.437223113357141e-05,
      "loss": 2.436220169067383,
      "memory(GiB)": 66.02,
      "step": 17795,
      "token_acc": 0.4600760456273764,
      "train_speed(iter/s)": 1.4514
    },
    {
      "epoch": 0.7626065721263013,
      "grad_norm": 3.2719600200653076,
      "learning_rate": 9.436912888266239e-05,
      "loss": 2.4398998260498046,
      "memory(GiB)": 66.02,
      "step": 17800,
      "token_acc": 0.5064516129032258,
      "train_speed(iter/s)": 1.451242
    },
    {
      "epoch": 0.7628207874555503,
      "grad_norm": 3.56967830657959,
      "learning_rate": 9.436602582796498e-05,
      "loss": 2.569136619567871,
      "memory(GiB)": 66.02,
      "step": 17805,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.451348
    },
    {
      "epoch": 0.7630350027847993,
      "grad_norm": 4.502561569213867,
      "learning_rate": 9.43629219695354e-05,
      "loss": 2.598386764526367,
      "memory(GiB)": 66.02,
      "step": 17810,
      "token_acc": 0.4483985765124555,
      "train_speed(iter/s)": 1.451262
    },
    {
      "epoch": 0.7632492181140482,
      "grad_norm": 3.789357900619507,
      "learning_rate": 9.435981730742986e-05,
      "loss": 2.718033027648926,
      "memory(GiB)": 66.02,
      "step": 17815,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.451329
    },
    {
      "epoch": 0.7634634334432973,
      "grad_norm": 3.2228078842163086,
      "learning_rate": 9.435671184170463e-05,
      "loss": 2.681374168395996,
      "memory(GiB)": 66.02,
      "step": 17820,
      "token_acc": 0.467680608365019,
      "train_speed(iter/s)": 1.451369
    },
    {
      "epoch": 0.7636776487725462,
      "grad_norm": 3.3303074836730957,
      "learning_rate": 9.435360557241595e-05,
      "loss": 2.9132007598876952,
      "memory(GiB)": 66.02,
      "step": 17825,
      "token_acc": 0.4143646408839779,
      "train_speed(iter/s)": 1.451362
    },
    {
      "epoch": 0.7638918641017951,
      "grad_norm": 5.304351806640625,
      "learning_rate": 9.43504984996201e-05,
      "loss": 2.518760108947754,
      "memory(GiB)": 66.02,
      "step": 17830,
      "token_acc": 0.5041322314049587,
      "train_speed(iter/s)": 1.451438
    },
    {
      "epoch": 0.7641060794310441,
      "grad_norm": 3.8242204189300537,
      "learning_rate": 9.434739062337337e-05,
      "loss": 2.5196044921875,
      "memory(GiB)": 66.02,
      "step": 17835,
      "token_acc": 0.46285714285714286,
      "train_speed(iter/s)": 1.451449
    },
    {
      "epoch": 0.7643202947602931,
      "grad_norm": 3.9230220317840576,
      "learning_rate": 9.434428194373205e-05,
      "loss": 2.6513303756713866,
      "memory(GiB)": 66.02,
      "step": 17840,
      "token_acc": 0.4728682170542636,
      "train_speed(iter/s)": 1.451502
    },
    {
      "epoch": 0.764534510089542,
      "grad_norm": 4.761760711669922,
      "learning_rate": 9.434117246075246e-05,
      "loss": 2.5037139892578124,
      "memory(GiB)": 66.02,
      "step": 17845,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.451526
    },
    {
      "epoch": 0.764748725418791,
      "grad_norm": 4.526192665100098,
      "learning_rate": 9.433806217449094e-05,
      "loss": 2.534004974365234,
      "memory(GiB)": 66.02,
      "step": 17850,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.451488
    },
    {
      "epoch": 0.7649629407480399,
      "grad_norm": 4.827754974365234,
      "learning_rate": 9.433495108500384e-05,
      "loss": 2.4859790802001953,
      "memory(GiB)": 66.02,
      "step": 17855,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.451514
    },
    {
      "epoch": 0.7651771560772889,
      "grad_norm": 3.4157497882843018,
      "learning_rate": 9.43318391923475e-05,
      "loss": 2.5476917266845702,
      "memory(GiB)": 66.02,
      "step": 17860,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.451488
    },
    {
      "epoch": 0.7653913714065379,
      "grad_norm": 4.01407527923584,
      "learning_rate": 9.432872649657832e-05,
      "loss": 2.730925750732422,
      "memory(GiB)": 66.02,
      "step": 17865,
      "token_acc": 0.46963562753036436,
      "train_speed(iter/s)": 1.451435
    },
    {
      "epoch": 0.7656055867357868,
      "grad_norm": 3.7629337310791016,
      "learning_rate": 9.432561299775267e-05,
      "loss": 2.6077495574951173,
      "memory(GiB)": 66.02,
      "step": 17870,
      "token_acc": 0.4548736462093863,
      "train_speed(iter/s)": 1.451443
    },
    {
      "epoch": 0.7658198020650357,
      "grad_norm": 3.8664438724517822,
      "learning_rate": 9.432249869592696e-05,
      "loss": 2.6030914306640627,
      "memory(GiB)": 66.02,
      "step": 17875,
      "token_acc": 0.4808743169398907,
      "train_speed(iter/s)": 1.451476
    },
    {
      "epoch": 0.7660340173942848,
      "grad_norm": 4.5877156257629395,
      "learning_rate": 9.431938359115759e-05,
      "loss": 2.7656469345092773,
      "memory(GiB)": 66.02,
      "step": 17880,
      "token_acc": 0.4427710843373494,
      "train_speed(iter/s)": 1.451538
    },
    {
      "epoch": 0.7662482327235337,
      "grad_norm": 4.128017902374268,
      "learning_rate": 9.431626768350104e-05,
      "loss": 2.3668609619140626,
      "memory(GiB)": 66.02,
      "step": 17885,
      "token_acc": 0.49407114624505927,
      "train_speed(iter/s)": 1.451567
    },
    {
      "epoch": 0.7664624480527826,
      "grad_norm": 3.8643641471862793,
      "learning_rate": 9.43131509730137e-05,
      "loss": 2.262523078918457,
      "memory(GiB)": 66.02,
      "step": 17890,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.451522
    },
    {
      "epoch": 0.7666766633820317,
      "grad_norm": 4.605358123779297,
      "learning_rate": 9.43100334597521e-05,
      "loss": 2.3411001205444335,
      "memory(GiB)": 66.02,
      "step": 17895,
      "token_acc": 0.5140562248995983,
      "train_speed(iter/s)": 1.451475
    },
    {
      "epoch": 0.7668908787112806,
      "grad_norm": 4.67607307434082,
      "learning_rate": 9.430691514377264e-05,
      "loss": 2.4089115142822264,
      "memory(GiB)": 66.02,
      "step": 17900,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.451525
    },
    {
      "epoch": 0.7671050940405295,
      "grad_norm": 3.2710061073303223,
      "learning_rate": 9.430379602513186e-05,
      "loss": 2.5764934539794924,
      "memory(GiB)": 66.02,
      "step": 17905,
      "token_acc": 0.4768211920529801,
      "train_speed(iter/s)": 1.451583
    },
    {
      "epoch": 0.7673193093697785,
      "grad_norm": 5.5393195152282715,
      "learning_rate": 9.430067610388626e-05,
      "loss": 2.6153459548950195,
      "memory(GiB)": 66.02,
      "step": 17910,
      "token_acc": 0.4828897338403042,
      "train_speed(iter/s)": 1.451527
    },
    {
      "epoch": 0.7675335246990275,
      "grad_norm": 3.900489568710327,
      "learning_rate": 9.429755538009235e-05,
      "loss": 2.8840137481689454,
      "memory(GiB)": 66.02,
      "step": 17915,
      "token_acc": 0.4584837545126354,
      "train_speed(iter/s)": 1.4516
    },
    {
      "epoch": 0.7677477400282764,
      "grad_norm": 5.313302040100098,
      "learning_rate": 9.429443385380667e-05,
      "loss": 2.4606693267822264,
      "memory(GiB)": 66.02,
      "step": 17920,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.451684
    },
    {
      "epoch": 0.7679619553575254,
      "grad_norm": 4.444176197052002,
      "learning_rate": 9.429131152508577e-05,
      "loss": 2.484004592895508,
      "memory(GiB)": 66.02,
      "step": 17925,
      "token_acc": 0.4867924528301887,
      "train_speed(iter/s)": 1.451739
    },
    {
      "epoch": 0.7681761706867744,
      "grad_norm": 4.170411586761475,
      "learning_rate": 9.42881883939862e-05,
      "loss": 2.614640808105469,
      "memory(GiB)": 66.02,
      "step": 17930,
      "token_acc": 0.46504559270516715,
      "train_speed(iter/s)": 1.451818
    },
    {
      "epoch": 0.7683903860160233,
      "grad_norm": 5.1936750411987305,
      "learning_rate": 9.428506446056458e-05,
      "loss": 2.537836456298828,
      "memory(GiB)": 66.02,
      "step": 17935,
      "token_acc": 0.5362318840579711,
      "train_speed(iter/s)": 1.451722
    },
    {
      "epoch": 0.7686046013452723,
      "grad_norm": 4.254932403564453,
      "learning_rate": 9.428193972487746e-05,
      "loss": 2.4651468276977537,
      "memory(GiB)": 66.02,
      "step": 17940,
      "token_acc": 0.5412186379928315,
      "train_speed(iter/s)": 1.451836
    },
    {
      "epoch": 0.7688188166745212,
      "grad_norm": 4.5974555015563965,
      "learning_rate": 9.427881418698147e-05,
      "loss": 2.428760528564453,
      "memory(GiB)": 66.02,
      "step": 17945,
      "token_acc": 0.502092050209205,
      "train_speed(iter/s)": 1.451855
    },
    {
      "epoch": 0.7690330320037702,
      "grad_norm": 4.533402442932129,
      "learning_rate": 9.427568784693321e-05,
      "loss": 2.4003820419311523,
      "memory(GiB)": 66.02,
      "step": 17950,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.451893
    },
    {
      "epoch": 0.7692472473330192,
      "grad_norm": 5.776178359985352,
      "learning_rate": 9.427256070478934e-05,
      "loss": 2.593052291870117,
      "memory(GiB)": 66.02,
      "step": 17955,
      "token_acc": 0.4265232974910394,
      "train_speed(iter/s)": 1.451927
    },
    {
      "epoch": 0.7694614626622681,
      "grad_norm": 4.446044921875,
      "learning_rate": 9.426943276060649e-05,
      "loss": 2.844767761230469,
      "memory(GiB)": 66.02,
      "step": 17960,
      "token_acc": 0.4355400696864111,
      "train_speed(iter/s)": 1.451976
    },
    {
      "epoch": 0.769675677991517,
      "grad_norm": 6.065513610839844,
      "learning_rate": 9.426630401444136e-05,
      "loss": 2.5938388824462892,
      "memory(GiB)": 66.02,
      "step": 17965,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.452019
    },
    {
      "epoch": 0.7698898933207661,
      "grad_norm": 4.160604476928711,
      "learning_rate": 9.426317446635059e-05,
      "loss": 2.8272281646728517,
      "memory(GiB)": 66.02,
      "step": 17970,
      "token_acc": 0.4328767123287671,
      "train_speed(iter/s)": 1.452031
    },
    {
      "epoch": 0.770104108650015,
      "grad_norm": 4.037982940673828,
      "learning_rate": 9.426004411639089e-05,
      "loss": 2.7478988647460936,
      "memory(GiB)": 66.02,
      "step": 17975,
      "token_acc": 0.4676470588235294,
      "train_speed(iter/s)": 1.452143
    },
    {
      "epoch": 0.7703183239792639,
      "grad_norm": 3.574831247329712,
      "learning_rate": 9.425691296461898e-05,
      "loss": 2.6368877410888674,
      "memory(GiB)": 66.02,
      "step": 17980,
      "token_acc": 0.4551083591331269,
      "train_speed(iter/s)": 1.452149
    },
    {
      "epoch": 0.770532539308513,
      "grad_norm": 4.331003665924072,
      "learning_rate": 9.425378101109158e-05,
      "loss": 2.5000373840332033,
      "memory(GiB)": 66.02,
      "step": 17985,
      "token_acc": 0.47474747474747475,
      "train_speed(iter/s)": 1.452133
    },
    {
      "epoch": 0.7707467546377619,
      "grad_norm": 4.412370681762695,
      "learning_rate": 9.425064825586541e-05,
      "loss": 2.7192708969116213,
      "memory(GiB)": 66.02,
      "step": 17990,
      "token_acc": 0.41904761904761906,
      "train_speed(iter/s)": 1.45218
    },
    {
      "epoch": 0.7709609699670108,
      "grad_norm": 3.9561266899108887,
      "learning_rate": 9.424751469899724e-05,
      "loss": 2.809209632873535,
      "memory(GiB)": 66.02,
      "step": 17995,
      "token_acc": 0.4271523178807947,
      "train_speed(iter/s)": 1.452232
    },
    {
      "epoch": 0.7711751852962598,
      "grad_norm": 3.802600622177124,
      "learning_rate": 9.424438034054385e-05,
      "loss": 2.6348880767822265,
      "memory(GiB)": 66.02,
      "step": 18000,
      "token_acc": 0.5176470588235295,
      "train_speed(iter/s)": 1.452281
    },
    {
      "epoch": 0.7711751852962598,
      "eval_loss": 2.1363372802734375,
      "eval_runtime": 14.8199,
      "eval_samples_per_second": 6.748,
      "eval_steps_per_second": 6.748,
      "eval_token_acc": 0.4845222072678331,
      "step": 18000
    },
    {
      "epoch": 0.7713894006255088,
      "grad_norm": 3.982149362564087,
      "learning_rate": 9.424124518056199e-05,
      "loss": 2.6583667755126954,
      "memory(GiB)": 66.02,
      "step": 18005,
      "token_acc": 0.48217821782178216,
      "train_speed(iter/s)": 1.450389
    },
    {
      "epoch": 0.7716036159547577,
      "grad_norm": 4.694133281707764,
      "learning_rate": 9.423810921910848e-05,
      "loss": 2.9110685348510743,
      "memory(GiB)": 66.02,
      "step": 18010,
      "token_acc": 0.4383561643835616,
      "train_speed(iter/s)": 1.450338
    },
    {
      "epoch": 0.7718178312840067,
      "grad_norm": 4.896605491638184,
      "learning_rate": 9.42349724562401e-05,
      "loss": 2.373472213745117,
      "memory(GiB)": 66.02,
      "step": 18015,
      "token_acc": 0.5056179775280899,
      "train_speed(iter/s)": 1.450378
    },
    {
      "epoch": 0.7720320466132556,
      "grad_norm": 3.073610305786133,
      "learning_rate": 9.423183489201373e-05,
      "loss": 2.56213321685791,
      "memory(GiB)": 66.02,
      "step": 18020,
      "token_acc": 0.48417721518987344,
      "train_speed(iter/s)": 1.450362
    },
    {
      "epoch": 0.7722462619425046,
      "grad_norm": 4.432365417480469,
      "learning_rate": 9.422869652648617e-05,
      "loss": 2.6012752532958983,
      "memory(GiB)": 66.02,
      "step": 18025,
      "token_acc": 0.4380664652567976,
      "train_speed(iter/s)": 1.450337
    },
    {
      "epoch": 0.7724604772717536,
      "grad_norm": 3.4521918296813965,
      "learning_rate": 9.422555735971426e-05,
      "loss": 2.4666662216186523,
      "memory(GiB)": 66.02,
      "step": 18030,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.45032
    },
    {
      "epoch": 0.7726746926010025,
      "grad_norm": 4.163490295410156,
      "learning_rate": 9.422241739175491e-05,
      "loss": 2.196506881713867,
      "memory(GiB)": 66.02,
      "step": 18035,
      "token_acc": 0.5355648535564853,
      "train_speed(iter/s)": 1.450345
    },
    {
      "epoch": 0.7728889079302514,
      "grad_norm": 4.549071311950684,
      "learning_rate": 9.421927662266499e-05,
      "loss": 2.367755889892578,
      "memory(GiB)": 66.02,
      "step": 18040,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.450314
    },
    {
      "epoch": 0.7731031232595005,
      "grad_norm": 3.2333977222442627,
      "learning_rate": 9.421613505250138e-05,
      "loss": 2.5339868545532225,
      "memory(GiB)": 66.02,
      "step": 18045,
      "token_acc": 0.4647058823529412,
      "train_speed(iter/s)": 1.450155
    },
    {
      "epoch": 0.7733173385887494,
      "grad_norm": 5.415011882781982,
      "learning_rate": 9.421299268132103e-05,
      "loss": 2.6513235092163088,
      "memory(GiB)": 66.02,
      "step": 18050,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.450173
    },
    {
      "epoch": 0.7735315539179983,
      "grad_norm": 4.054705619812012,
      "learning_rate": 9.420984950918082e-05,
      "loss": 2.5059566497802734,
      "memory(GiB)": 66.02,
      "step": 18055,
      "token_acc": 0.42517006802721086,
      "train_speed(iter/s)": 1.450235
    },
    {
      "epoch": 0.7737457692472474,
      "grad_norm": 4.358946323394775,
      "learning_rate": 9.42067055361377e-05,
      "loss": 2.851646614074707,
      "memory(GiB)": 66.02,
      "step": 18060,
      "token_acc": 0.41055718475073316,
      "train_speed(iter/s)": 1.450248
    },
    {
      "epoch": 0.7739599845764963,
      "grad_norm": 4.50553035736084,
      "learning_rate": 9.420356076224865e-05,
      "loss": 2.516449737548828,
      "memory(GiB)": 66.02,
      "step": 18065,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.450234
    },
    {
      "epoch": 0.7741741999057452,
      "grad_norm": 5.02170991897583,
      "learning_rate": 9.420041518757066e-05,
      "loss": 2.619902992248535,
      "memory(GiB)": 66.02,
      "step": 18070,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.450157
    },
    {
      "epoch": 0.7743884152349942,
      "grad_norm": 4.110943794250488,
      "learning_rate": 9.419726881216066e-05,
      "loss": 2.503653717041016,
      "memory(GiB)": 66.02,
      "step": 18075,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.45022
    },
    {
      "epoch": 0.7746026305642432,
      "grad_norm": 4.794353008270264,
      "learning_rate": 9.419412163607567e-05,
      "loss": 2.4214576721191405,
      "memory(GiB)": 66.02,
      "step": 18080,
      "token_acc": 0.504,
      "train_speed(iter/s)": 1.450294
    },
    {
      "epoch": 0.7748168458934921,
      "grad_norm": 3.4237639904022217,
      "learning_rate": 9.419097365937272e-05,
      "loss": 2.720335006713867,
      "memory(GiB)": 66.02,
      "step": 18085,
      "token_acc": 0.42071197411003236,
      "train_speed(iter/s)": 1.450338
    },
    {
      "epoch": 0.7750310612227411,
      "grad_norm": 4.554632186889648,
      "learning_rate": 9.418782488210882e-05,
      "loss": 2.496338653564453,
      "memory(GiB)": 66.02,
      "step": 18090,
      "token_acc": 0.45229681978798586,
      "train_speed(iter/s)": 1.450282
    },
    {
      "epoch": 0.77524527655199,
      "grad_norm": 3.722590684890747,
      "learning_rate": 9.418467530434103e-05,
      "loss": 2.606220245361328,
      "memory(GiB)": 66.02,
      "step": 18095,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.450305
    },
    {
      "epoch": 0.775459491881239,
      "grad_norm": 4.078646183013916,
      "learning_rate": 9.41815249261264e-05,
      "loss": 2.6520879745483397,
      "memory(GiB)": 66.02,
      "step": 18100,
      "token_acc": 0.42567567567567566,
      "train_speed(iter/s)": 1.450402
    },
    {
      "epoch": 0.775673707210488,
      "grad_norm": 4.396276950836182,
      "learning_rate": 9.417837374752199e-05,
      "loss": 2.7209846496582033,
      "memory(GiB)": 66.02,
      "step": 18105,
      "token_acc": 0.4582043343653251,
      "train_speed(iter/s)": 1.450452
    },
    {
      "epoch": 0.7758879225397369,
      "grad_norm": 4.03439474105835,
      "learning_rate": 9.417522176858491e-05,
      "loss": 2.641116905212402,
      "memory(GiB)": 66.02,
      "step": 18110,
      "token_acc": 0.46397694524495675,
      "train_speed(iter/s)": 1.45036
    },
    {
      "epoch": 0.7761021378689859,
      "grad_norm": 4.848409652709961,
      "learning_rate": 9.417206898937223e-05,
      "loss": 2.436016654968262,
      "memory(GiB)": 66.02,
      "step": 18115,
      "token_acc": 0.49767441860465117,
      "train_speed(iter/s)": 1.450261
    },
    {
      "epoch": 0.7763163531982349,
      "grad_norm": 7.31134557723999,
      "learning_rate": 9.41689154099411e-05,
      "loss": 2.5979223251342773,
      "memory(GiB)": 66.02,
      "step": 18120,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.450142
    },
    {
      "epoch": 0.7765305685274838,
      "grad_norm": 3.8297173976898193,
      "learning_rate": 9.416576103034862e-05,
      "loss": 2.53304386138916,
      "memory(GiB)": 66.02,
      "step": 18125,
      "token_acc": 0.46839080459770116,
      "train_speed(iter/s)": 1.450088
    },
    {
      "epoch": 0.7767447838567327,
      "grad_norm": 4.4571099281311035,
      "learning_rate": 9.416260585065195e-05,
      "loss": 2.6305755615234374,
      "memory(GiB)": 66.02,
      "step": 18130,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.449995
    },
    {
      "epoch": 0.7769589991859818,
      "grad_norm": 3.7489819526672363,
      "learning_rate": 9.415944987090826e-05,
      "loss": 2.425373840332031,
      "memory(GiB)": 66.02,
      "step": 18135,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.449956
    },
    {
      "epoch": 0.7771732145152307,
      "grad_norm": 3.8894479274749756,
      "learning_rate": 9.415629309117471e-05,
      "loss": 2.251003646850586,
      "memory(GiB)": 66.02,
      "step": 18140,
      "token_acc": 0.5531914893617021,
      "train_speed(iter/s)": 1.449913
    },
    {
      "epoch": 0.7773874298444796,
      "grad_norm": 4.083927154541016,
      "learning_rate": 9.415313551150847e-05,
      "loss": 2.693222236633301,
      "memory(GiB)": 66.02,
      "step": 18145,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.44986
    },
    {
      "epoch": 0.7776016451737287,
      "grad_norm": 7.487312316894531,
      "learning_rate": 9.414997713196678e-05,
      "loss": 2.689481735229492,
      "memory(GiB)": 66.02,
      "step": 18150,
      "token_acc": 0.4280936454849498,
      "train_speed(iter/s)": 1.449906
    },
    {
      "epoch": 0.7778158605029776,
      "grad_norm": 6.086834907531738,
      "learning_rate": 9.414681795260683e-05,
      "loss": 2.7687519073486326,
      "memory(GiB)": 66.02,
      "step": 18155,
      "token_acc": 0.4637223974763407,
      "train_speed(iter/s)": 1.450002
    },
    {
      "epoch": 0.7780300758322266,
      "grad_norm": 4.046126365661621,
      "learning_rate": 9.414365797348586e-05,
      "loss": 2.657626914978027,
      "memory(GiB)": 66.02,
      "step": 18160,
      "token_acc": 0.43217665615141954,
      "train_speed(iter/s)": 1.449989
    },
    {
      "epoch": 0.7782442911614755,
      "grad_norm": 3.8933658599853516,
      "learning_rate": 9.414049719466111e-05,
      "loss": 2.5921215057373046,
      "memory(GiB)": 66.02,
      "step": 18165,
      "token_acc": 0.4681647940074906,
      "train_speed(iter/s)": 1.449964
    },
    {
      "epoch": 0.7784585064907245,
      "grad_norm": 4.799526691436768,
      "learning_rate": 9.413733561618985e-05,
      "loss": 2.7638824462890623,
      "memory(GiB)": 66.02,
      "step": 18170,
      "token_acc": 0.45229681978798586,
      "train_speed(iter/s)": 1.44989
    },
    {
      "epoch": 0.7786727218199735,
      "grad_norm": 5.367789268493652,
      "learning_rate": 9.413417323812936e-05,
      "loss": 2.3000259399414062,
      "memory(GiB)": 66.02,
      "step": 18175,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.449868
    },
    {
      "epoch": 0.7788869371492224,
      "grad_norm": 4.351169109344482,
      "learning_rate": 9.41310100605369e-05,
      "loss": 2.697555732727051,
      "memory(GiB)": 66.02,
      "step": 18180,
      "token_acc": 0.4478021978021978,
      "train_speed(iter/s)": 1.449802
    },
    {
      "epoch": 0.7791011524784713,
      "grad_norm": 4.8134541511535645,
      "learning_rate": 9.412784608346983e-05,
      "loss": 2.8435754776000977,
      "memory(GiB)": 66.02,
      "step": 18185,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.44973
    },
    {
      "epoch": 0.7793153678077204,
      "grad_norm": 5.636275291442871,
      "learning_rate": 9.41246813069854e-05,
      "loss": 2.467054557800293,
      "memory(GiB)": 66.02,
      "step": 18190,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.449771
    },
    {
      "epoch": 0.7795295831369693,
      "grad_norm": 4.488534927368164,
      "learning_rate": 9.412151573114098e-05,
      "loss": 2.6613616943359375,
      "memory(GiB)": 66.02,
      "step": 18195,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.449799
    },
    {
      "epoch": 0.7797437984662182,
      "grad_norm": 4.791868209838867,
      "learning_rate": 9.411834935599393e-05,
      "loss": 2.77734489440918,
      "memory(GiB)": 66.02,
      "step": 18200,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.449803
    },
    {
      "epoch": 0.7799580137954673,
      "grad_norm": 3.9967901706695557,
      "learning_rate": 9.411518218160158e-05,
      "loss": 2.680686378479004,
      "memory(GiB)": 66.02,
      "step": 18205,
      "token_acc": 0.44089456869009586,
      "train_speed(iter/s)": 1.449883
    },
    {
      "epoch": 0.7801722291247162,
      "grad_norm": 3.768009662628174,
      "learning_rate": 9.411201420802134e-05,
      "loss": 2.5262779235839843,
      "memory(GiB)": 66.02,
      "step": 18210,
      "token_acc": 0.445859872611465,
      "train_speed(iter/s)": 1.449904
    },
    {
      "epoch": 0.7803864444539651,
      "grad_norm": 6.320119857788086,
      "learning_rate": 9.410884543531056e-05,
      "loss": 2.4924684524536134,
      "memory(GiB)": 66.02,
      "step": 18215,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.44976
    },
    {
      "epoch": 0.7806006597832141,
      "grad_norm": 3.9295129776000977,
      "learning_rate": 9.410567586352668e-05,
      "loss": 2.752115249633789,
      "memory(GiB)": 66.02,
      "step": 18220,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.449737
    },
    {
      "epoch": 0.7808148751124631,
      "grad_norm": 3.391709089279175,
      "learning_rate": 9.41025054927271e-05,
      "loss": 2.3972236633300783,
      "memory(GiB)": 66.02,
      "step": 18225,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.44969
    },
    {
      "epoch": 0.781029090441712,
      "grad_norm": 4.542713165283203,
      "learning_rate": 9.409933432296927e-05,
      "loss": 2.6938899993896483,
      "memory(GiB)": 66.02,
      "step": 18230,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.449607
    },
    {
      "epoch": 0.781243305770961,
      "grad_norm": 4.435104846954346,
      "learning_rate": 9.409616235431062e-05,
      "loss": 2.4522470474243163,
      "memory(GiB)": 66.02,
      "step": 18235,
      "token_acc": 0.47924528301886793,
      "train_speed(iter/s)": 1.449424
    },
    {
      "epoch": 0.7814575211002099,
      "grad_norm": 5.2817864418029785,
      "learning_rate": 9.409298958680864e-05,
      "loss": 2.65223388671875,
      "memory(GiB)": 66.02,
      "step": 18240,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.449496
    },
    {
      "epoch": 0.7816717364294589,
      "grad_norm": 4.501829624176025,
      "learning_rate": 9.408981602052078e-05,
      "loss": 2.47408447265625,
      "memory(GiB)": 66.02,
      "step": 18245,
      "token_acc": 0.4725274725274725,
      "train_speed(iter/s)": 1.449509
    },
    {
      "epoch": 0.7818859517587079,
      "grad_norm": 3.8733203411102295,
      "learning_rate": 9.408664165550453e-05,
      "loss": 2.434902572631836,
      "memory(GiB)": 66.02,
      "step": 18250,
      "token_acc": 0.49429657794676807,
      "train_speed(iter/s)": 1.44958
    },
    {
      "epoch": 0.7821001670879568,
      "grad_norm": 4.0289626121521,
      "learning_rate": 9.408346649181742e-05,
      "loss": 2.555371856689453,
      "memory(GiB)": 66.02,
      "step": 18255,
      "token_acc": 0.4536741214057508,
      "train_speed(iter/s)": 1.449654
    },
    {
      "epoch": 0.7823143824172057,
      "grad_norm": 3.9965853691101074,
      "learning_rate": 9.408029052951698e-05,
      "loss": 2.5941064834594725,
      "memory(GiB)": 66.02,
      "step": 18260,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.449677
    },
    {
      "epoch": 0.7825285977464548,
      "grad_norm": 4.760998249053955,
      "learning_rate": 9.407711376866071e-05,
      "loss": 2.619590187072754,
      "memory(GiB)": 66.02,
      "step": 18265,
      "token_acc": 0.45555555555555555,
      "train_speed(iter/s)": 1.449678
    },
    {
      "epoch": 0.7827428130757037,
      "grad_norm": 4.3195600509643555,
      "learning_rate": 9.407393620930618e-05,
      "loss": 2.5690332412719727,
      "memory(GiB)": 66.02,
      "step": 18270,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.449675
    },
    {
      "epoch": 0.7829570284049526,
      "grad_norm": 4.01470947265625,
      "learning_rate": 9.407075785151094e-05,
      "loss": 2.6800205230712892,
      "memory(GiB)": 66.02,
      "step": 18275,
      "token_acc": 0.41935483870967744,
      "train_speed(iter/s)": 1.449692
    },
    {
      "epoch": 0.7831712437342017,
      "grad_norm": 4.600347995758057,
      "learning_rate": 9.406757869533259e-05,
      "loss": 2.6699668884277346,
      "memory(GiB)": 66.02,
      "step": 18280,
      "token_acc": 0.4512987012987013,
      "train_speed(iter/s)": 1.449634
    },
    {
      "epoch": 0.7833854590634506,
      "grad_norm": 4.298788070678711,
      "learning_rate": 9.406439874082871e-05,
      "loss": 2.2426826477050783,
      "memory(GiB)": 66.02,
      "step": 18285,
      "token_acc": 0.5117056856187291,
      "train_speed(iter/s)": 1.449694
    },
    {
      "epoch": 0.7835996743926995,
      "grad_norm": 3.2672784328460693,
      "learning_rate": 9.406121798805692e-05,
      "loss": 2.3077823638916017,
      "memory(GiB)": 66.02,
      "step": 18290,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.449681
    },
    {
      "epoch": 0.7838138897219485,
      "grad_norm": 4.382030487060547,
      "learning_rate": 9.405803643707482e-05,
      "loss": 2.4252155303955076,
      "memory(GiB)": 66.02,
      "step": 18295,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.449665
    },
    {
      "epoch": 0.7840281050511975,
      "grad_norm": 4.850598335266113,
      "learning_rate": 9.405485408794005e-05,
      "loss": 2.634998321533203,
      "memory(GiB)": 66.02,
      "step": 18300,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.449711
    },
    {
      "epoch": 0.7842423203804464,
      "grad_norm": 4.17399263381958,
      "learning_rate": 9.40516709407103e-05,
      "loss": 2.8445547103881834,
      "memory(GiB)": 66.02,
      "step": 18305,
      "token_acc": 0.3961661341853035,
      "train_speed(iter/s)": 1.449564
    },
    {
      "epoch": 0.7844565357096954,
      "grad_norm": 4.272390842437744,
      "learning_rate": 9.40484869954432e-05,
      "loss": 2.8256629943847655,
      "memory(GiB)": 66.02,
      "step": 18310,
      "token_acc": 0.4384858044164038,
      "train_speed(iter/s)": 1.449582
    },
    {
      "epoch": 0.7846707510389443,
      "grad_norm": 4.42106819152832,
      "learning_rate": 9.404530225219643e-05,
      "loss": 2.5621002197265623,
      "memory(GiB)": 66.02,
      "step": 18315,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.449464
    },
    {
      "epoch": 0.7848849663681933,
      "grad_norm": 6.792846202850342,
      "learning_rate": 9.404211671102769e-05,
      "loss": 2.8990150451660157,
      "memory(GiB)": 66.02,
      "step": 18320,
      "token_acc": 0.4152542372881356,
      "train_speed(iter/s)": 1.449451
    },
    {
      "epoch": 0.7850991816974423,
      "grad_norm": 4.029753684997559,
      "learning_rate": 9.403893037199469e-05,
      "loss": 2.663385009765625,
      "memory(GiB)": 66.02,
      "step": 18325,
      "token_acc": 0.4228571428571429,
      "train_speed(iter/s)": 1.449505
    },
    {
      "epoch": 0.7853133970266912,
      "grad_norm": 3.72320556640625,
      "learning_rate": 9.403574323515516e-05,
      "loss": 2.673150634765625,
      "memory(GiB)": 66.02,
      "step": 18330,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.449571
    },
    {
      "epoch": 0.7855276123559402,
      "grad_norm": 4.929439544677734,
      "learning_rate": 9.403255530056682e-05,
      "loss": 2.5984115600585938,
      "memory(GiB)": 66.02,
      "step": 18335,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.449636
    },
    {
      "epoch": 0.7857418276851892,
      "grad_norm": 4.097453594207764,
      "learning_rate": 9.402936656828745e-05,
      "loss": 2.5976675033569334,
      "memory(GiB)": 66.02,
      "step": 18340,
      "token_acc": 0.46557377049180326,
      "train_speed(iter/s)": 1.44965
    },
    {
      "epoch": 0.7859560430144381,
      "grad_norm": 3.460083484649658,
      "learning_rate": 9.40261770383748e-05,
      "loss": 2.2677680969238283,
      "memory(GiB)": 66.02,
      "step": 18345,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.449617
    },
    {
      "epoch": 0.786170258343687,
      "grad_norm": 5.9269232749938965,
      "learning_rate": 9.402298671088665e-05,
      "loss": 2.5967548370361326,
      "memory(GiB)": 66.02,
      "step": 18350,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.449481
    },
    {
      "epoch": 0.7863844736729361,
      "grad_norm": 3.7726407051086426,
      "learning_rate": 9.401979558588081e-05,
      "loss": 2.911211395263672,
      "memory(GiB)": 66.02,
      "step": 18355,
      "token_acc": 0.4094292803970223,
      "train_speed(iter/s)": 1.449473
    },
    {
      "epoch": 0.786598689002185,
      "grad_norm": 3.9955241680145264,
      "learning_rate": 9.401660366341506e-05,
      "loss": 2.807758331298828,
      "memory(GiB)": 66.02,
      "step": 18360,
      "token_acc": 0.43820224719101125,
      "train_speed(iter/s)": 1.44954
    },
    {
      "epoch": 0.7868129043314339,
      "grad_norm": 7.486758708953857,
      "learning_rate": 9.401341094354725e-05,
      "loss": 2.7489816665649416,
      "memory(GiB)": 66.02,
      "step": 18365,
      "token_acc": 0.437125748502994,
      "train_speed(iter/s)": 1.449511
    },
    {
      "epoch": 0.787027119660683,
      "grad_norm": 6.179536819458008,
      "learning_rate": 9.401021742633523e-05,
      "loss": 2.3739242553710938,
      "memory(GiB)": 66.02,
      "step": 18370,
      "token_acc": 0.5255972696245734,
      "train_speed(iter/s)": 1.449468
    },
    {
      "epoch": 0.7872413349899319,
      "grad_norm": 3.909493923187256,
      "learning_rate": 9.400702311183681e-05,
      "loss": 2.706329345703125,
      "memory(GiB)": 66.02,
      "step": 18375,
      "token_acc": 0.4309210526315789,
      "train_speed(iter/s)": 1.449486
    },
    {
      "epoch": 0.7874555503191808,
      "grad_norm": 3.4717910289764404,
      "learning_rate": 9.40038280001099e-05,
      "loss": 2.4165367126464843,
      "memory(GiB)": 66.02,
      "step": 18380,
      "token_acc": 0.4789915966386555,
      "train_speed(iter/s)": 1.449526
    },
    {
      "epoch": 0.7876697656484298,
      "grad_norm": 5.798118591308594,
      "learning_rate": 9.400063209121235e-05,
      "loss": 2.2983631134033202,
      "memory(GiB)": 66.02,
      "step": 18385,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.449484
    },
    {
      "epoch": 0.7878839809776788,
      "grad_norm": 5.7551116943359375,
      "learning_rate": 9.39974353852021e-05,
      "loss": 2.595143127441406,
      "memory(GiB)": 66.02,
      "step": 18390,
      "token_acc": 0.4291497975708502,
      "train_speed(iter/s)": 1.449401
    },
    {
      "epoch": 0.7880981963069277,
      "grad_norm": 3.806825637817383,
      "learning_rate": 9.399423788213701e-05,
      "loss": 2.587324333190918,
      "memory(GiB)": 66.02,
      "step": 18395,
      "token_acc": 0.5034965034965035,
      "train_speed(iter/s)": 1.449503
    },
    {
      "epoch": 0.7883124116361767,
      "grad_norm": 4.1000213623046875,
      "learning_rate": 9.399103958207505e-05,
      "loss": 2.3451332092285155,
      "memory(GiB)": 66.02,
      "step": 18400,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.449541
    },
    {
      "epoch": 0.7885266269654256,
      "grad_norm": 4.8470940589904785,
      "learning_rate": 9.398784048507414e-05,
      "loss": 2.7898975372314454,
      "memory(GiB)": 66.02,
      "step": 18405,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.449603
    },
    {
      "epoch": 0.7887408422946746,
      "grad_norm": 3.740290880203247,
      "learning_rate": 9.398464059119222e-05,
      "loss": 2.669422912597656,
      "memory(GiB)": 66.02,
      "step": 18410,
      "token_acc": 0.4125,
      "train_speed(iter/s)": 1.449494
    },
    {
      "epoch": 0.7889550576239236,
      "grad_norm": 3.710299491882324,
      "learning_rate": 9.398143990048727e-05,
      "loss": 2.6664682388305665,
      "memory(GiB)": 66.02,
      "step": 18415,
      "token_acc": 0.44244604316546765,
      "train_speed(iter/s)": 1.4495
    },
    {
      "epoch": 0.7891692729531725,
      "grad_norm": 3.7856040000915527,
      "learning_rate": 9.397823841301729e-05,
      "loss": 2.745924949645996,
      "memory(GiB)": 66.02,
      "step": 18420,
      "token_acc": 0.4382716049382716,
      "train_speed(iter/s)": 1.449529
    },
    {
      "epoch": 0.7893834882824214,
      "grad_norm": 4.311188697814941,
      "learning_rate": 9.397503612884028e-05,
      "loss": 2.4199142456054688,
      "memory(GiB)": 66.02,
      "step": 18425,
      "token_acc": 0.4608150470219436,
      "train_speed(iter/s)": 1.449544
    },
    {
      "epoch": 0.7895977036116705,
      "grad_norm": 4.5722737312316895,
      "learning_rate": 9.397183304801422e-05,
      "loss": 2.563701629638672,
      "memory(GiB)": 66.02,
      "step": 18430,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.449632
    },
    {
      "epoch": 0.7898119189409194,
      "grad_norm": 3.772848606109619,
      "learning_rate": 9.396862917059716e-05,
      "loss": 2.603957939147949,
      "memory(GiB)": 66.02,
      "step": 18435,
      "token_acc": 0.44025157232704404,
      "train_speed(iter/s)": 1.44969
    },
    {
      "epoch": 0.7900261342701683,
      "grad_norm": 5.602076530456543,
      "learning_rate": 9.396542449664714e-05,
      "loss": 2.7424373626708984,
      "memory(GiB)": 66.02,
      "step": 18440,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.449766
    },
    {
      "epoch": 0.7902403495994174,
      "grad_norm": 4.258472442626953,
      "learning_rate": 9.396221902622221e-05,
      "loss": 2.6096662521362304,
      "memory(GiB)": 66.02,
      "step": 18445,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.449837
    },
    {
      "epoch": 0.7904545649286663,
      "grad_norm": 4.308422565460205,
      "learning_rate": 9.395901275938047e-05,
      "loss": 2.534760093688965,
      "memory(GiB)": 66.02,
      "step": 18450,
      "token_acc": 0.45652173913043476,
      "train_speed(iter/s)": 1.449884
    },
    {
      "epoch": 0.7906687802579152,
      "grad_norm": 4.1195220947265625,
      "learning_rate": 9.395580569617993e-05,
      "loss": 2.20571346282959,
      "memory(GiB)": 66.02,
      "step": 18455,
      "token_acc": 0.5427509293680297,
      "train_speed(iter/s)": 1.449898
    },
    {
      "epoch": 0.7908829955871642,
      "grad_norm": 4.290310382843018,
      "learning_rate": 9.395259783667876e-05,
      "loss": 2.574893760681152,
      "memory(GiB)": 66.02,
      "step": 18460,
      "token_acc": 0.4483985765124555,
      "train_speed(iter/s)": 1.449932
    },
    {
      "epoch": 0.7910972109164132,
      "grad_norm": 8.034736633300781,
      "learning_rate": 9.394938918093506e-05,
      "loss": 2.740865135192871,
      "memory(GiB)": 66.02,
      "step": 18465,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.450036
    },
    {
      "epoch": 0.7913114262456621,
      "grad_norm": 6.057497501373291,
      "learning_rate": 9.394617972900692e-05,
      "loss": 2.6498220443725584,
      "memory(GiB)": 66.02,
      "step": 18470,
      "token_acc": 0.4315068493150685,
      "train_speed(iter/s)": 1.450064
    },
    {
      "epoch": 0.7915256415749111,
      "grad_norm": 4.421770095825195,
      "learning_rate": 9.394296948095251e-05,
      "loss": 2.7161481857299803,
      "memory(GiB)": 66.02,
      "step": 18475,
      "token_acc": 0.5045871559633027,
      "train_speed(iter/s)": 1.450103
    },
    {
      "epoch": 0.79173985690416,
      "grad_norm": 4.770242214202881,
      "learning_rate": 9.393975843683e-05,
      "loss": 2.5783054351806642,
      "memory(GiB)": 66.02,
      "step": 18480,
      "token_acc": 0.4383954154727794,
      "train_speed(iter/s)": 1.450216
    },
    {
      "epoch": 0.791954072233409,
      "grad_norm": 5.175363540649414,
      "learning_rate": 9.393654659669755e-05,
      "loss": 2.3660795211791994,
      "memory(GiB)": 66.02,
      "step": 18485,
      "token_acc": 0.5112540192926045,
      "train_speed(iter/s)": 1.450213
    },
    {
      "epoch": 0.792168287562658,
      "grad_norm": 4.325352191925049,
      "learning_rate": 9.393333396061335e-05,
      "loss": 2.932937240600586,
      "memory(GiB)": 66.02,
      "step": 18490,
      "token_acc": 0.3987915407854985,
      "train_speed(iter/s)": 1.450181
    },
    {
      "epoch": 0.7923825028919069,
      "grad_norm": 3.7582504749298096,
      "learning_rate": 9.393012052863556e-05,
      "loss": 2.5134958267211913,
      "memory(GiB)": 66.02,
      "step": 18495,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.450093
    },
    {
      "epoch": 0.792596718221156,
      "grad_norm": 3.7133169174194336,
      "learning_rate": 9.392690630082246e-05,
      "loss": 2.4369785308837892,
      "memory(GiB)": 66.02,
      "step": 18500,
      "token_acc": 0.49262536873156343,
      "train_speed(iter/s)": 1.450082
    },
    {
      "epoch": 0.792596718221156,
      "eval_loss": 2.226203203201294,
      "eval_runtime": 15.2544,
      "eval_samples_per_second": 6.555,
      "eval_steps_per_second": 6.555,
      "eval_token_acc": 0.46776406035665297,
      "step": 18500
    },
    {
      "epoch": 0.7928109335504049,
      "grad_norm": 4.184749126434326,
      "learning_rate": 9.392369127723224e-05,
      "loss": 2.520800971984863,
      "memory(GiB)": 66.02,
      "step": 18505,
      "token_acc": 0.46406570841889117,
      "train_speed(iter/s)": 1.448273
    },
    {
      "epoch": 0.7930251488796538,
      "grad_norm": 4.034061908721924,
      "learning_rate": 9.392047545792312e-05,
      "loss": 2.63183479309082,
      "memory(GiB)": 66.02,
      "step": 18510,
      "token_acc": 0.45180722891566266,
      "train_speed(iter/s)": 1.448247
    },
    {
      "epoch": 0.7932393642089028,
      "grad_norm": 4.331101417541504,
      "learning_rate": 9.391725884295342e-05,
      "loss": 2.3379444122314452,
      "memory(GiB)": 66.02,
      "step": 18515,
      "token_acc": 0.48299319727891155,
      "train_speed(iter/s)": 1.448287
    },
    {
      "epoch": 0.7934535795381518,
      "grad_norm": 3.51999568939209,
      "learning_rate": 9.391404143238137e-05,
      "loss": 2.6013370513916017,
      "memory(GiB)": 66.02,
      "step": 18520,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.448288
    },
    {
      "epoch": 0.7936677948674007,
      "grad_norm": 4.5327534675598145,
      "learning_rate": 9.391082322626526e-05,
      "loss": 2.2977149963378904,
      "memory(GiB)": 66.02,
      "step": 18525,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.448288
    },
    {
      "epoch": 0.7938820101966497,
      "grad_norm": 3.8526196479797363,
      "learning_rate": 9.39076042246634e-05,
      "loss": 2.3243608474731445,
      "memory(GiB)": 66.02,
      "step": 18530,
      "token_acc": 0.5375939849624061,
      "train_speed(iter/s)": 1.448339
    },
    {
      "epoch": 0.7940962255258986,
      "grad_norm": 4.235838890075684,
      "learning_rate": 9.39043844276341e-05,
      "loss": 2.608317184448242,
      "memory(GiB)": 66.02,
      "step": 18535,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.448436
    },
    {
      "epoch": 0.7943104408551476,
      "grad_norm": 5.187750339508057,
      "learning_rate": 9.390116383523568e-05,
      "loss": 2.5674930572509767,
      "memory(GiB)": 66.02,
      "step": 18540,
      "token_acc": 0.45112781954887216,
      "train_speed(iter/s)": 1.448511
    },
    {
      "epoch": 0.7945246561843966,
      "grad_norm": 4.310109615325928,
      "learning_rate": 9.38979424475265e-05,
      "loss": 2.553070831298828,
      "memory(GiB)": 66.02,
      "step": 18545,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.448561
    },
    {
      "epoch": 0.7947388715136455,
      "grad_norm": 4.1276469230651855,
      "learning_rate": 9.389472026456492e-05,
      "loss": 2.384823799133301,
      "memory(GiB)": 66.02,
      "step": 18550,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.448522
    },
    {
      "epoch": 0.7949530868428945,
      "grad_norm": 4.401536464691162,
      "learning_rate": 9.38914972864093e-05,
      "loss": 2.794795799255371,
      "memory(GiB)": 66.02,
      "step": 18555,
      "token_acc": 0.41690962099125367,
      "train_speed(iter/s)": 1.448558
    },
    {
      "epoch": 0.7951673021721435,
      "grad_norm": 5.56785774230957,
      "learning_rate": 9.388827351311804e-05,
      "loss": 2.7745285034179688,
      "memory(GiB)": 66.02,
      "step": 18560,
      "token_acc": 0.44744744744744747,
      "train_speed(iter/s)": 1.44854
    },
    {
      "epoch": 0.7953815175013924,
      "grad_norm": 5.077114105224609,
      "learning_rate": 9.388504894474953e-05,
      "loss": 2.5758234024047852,
      "memory(GiB)": 66.02,
      "step": 18565,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.448563
    },
    {
      "epoch": 0.7955957328306413,
      "grad_norm": 3.3762216567993164,
      "learning_rate": 9.38818235813622e-05,
      "loss": 2.6224308013916016,
      "memory(GiB)": 66.02,
      "step": 18570,
      "token_acc": 0.47352941176470587,
      "train_speed(iter/s)": 1.448522
    },
    {
      "epoch": 0.7958099481598904,
      "grad_norm": 3.718240261077881,
      "learning_rate": 9.387859742301445e-05,
      "loss": 2.5745002746582033,
      "memory(GiB)": 66.02,
      "step": 18575,
      "token_acc": 0.484,
      "train_speed(iter/s)": 1.44847
    },
    {
      "epoch": 0.7960241634891393,
      "grad_norm": 4.881780624389648,
      "learning_rate": 9.387537046976476e-05,
      "loss": 2.486758804321289,
      "memory(GiB)": 66.02,
      "step": 18580,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.448512
    },
    {
      "epoch": 0.7962383788183882,
      "grad_norm": 4.2203593254089355,
      "learning_rate": 9.387214272167157e-05,
      "loss": 2.6436946868896483,
      "memory(GiB)": 66.02,
      "step": 18585,
      "token_acc": 0.4797507788161994,
      "train_speed(iter/s)": 1.448512
    },
    {
      "epoch": 0.7964525941476372,
      "grad_norm": 4.751131057739258,
      "learning_rate": 9.386891417879335e-05,
      "loss": 2.6586116790771483,
      "memory(GiB)": 66.02,
      "step": 18590,
      "token_acc": 0.45786516853932585,
      "train_speed(iter/s)": 1.448524
    },
    {
      "epoch": 0.7966668094768862,
      "grad_norm": 4.591014862060547,
      "learning_rate": 9.386568484118862e-05,
      "loss": 2.6276678085327148,
      "memory(GiB)": 66.02,
      "step": 18595,
      "token_acc": 0.4393939393939394,
      "train_speed(iter/s)": 1.448492
    },
    {
      "epoch": 0.7968810248061351,
      "grad_norm": 3.6758246421813965,
      "learning_rate": 9.386245470891584e-05,
      "loss": 2.7444358825683595,
      "memory(GiB)": 66.02,
      "step": 18600,
      "token_acc": 0.44376899696048633,
      "train_speed(iter/s)": 1.448533
    },
    {
      "epoch": 0.7970952401353841,
      "grad_norm": 4.360864639282227,
      "learning_rate": 9.385922378203356e-05,
      "loss": 2.509359359741211,
      "memory(GiB)": 66.02,
      "step": 18605,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.448577
    },
    {
      "epoch": 0.797309455464633,
      "grad_norm": 4.772853374481201,
      "learning_rate": 9.38559920606003e-05,
      "loss": 2.564883232116699,
      "memory(GiB)": 66.02,
      "step": 18610,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.448586
    },
    {
      "epoch": 0.797523670793882,
      "grad_norm": 4.266146183013916,
      "learning_rate": 9.38527595446746e-05,
      "loss": 2.61375732421875,
      "memory(GiB)": 66.02,
      "step": 18615,
      "token_acc": 0.47774480712166173,
      "train_speed(iter/s)": 1.448602
    },
    {
      "epoch": 0.797737886123131,
      "grad_norm": 6.605282306671143,
      "learning_rate": 9.384952623431502e-05,
      "loss": 2.5659912109375,
      "memory(GiB)": 66.02,
      "step": 18620,
      "token_acc": 0.44594594594594594,
      "train_speed(iter/s)": 1.448542
    },
    {
      "epoch": 0.7979521014523799,
      "grad_norm": 3.884753704071045,
      "learning_rate": 9.384629212958014e-05,
      "loss": 2.5951923370361327,
      "memory(GiB)": 66.02,
      "step": 18625,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.448572
    },
    {
      "epoch": 0.7981663167816289,
      "grad_norm": 4.864241123199463,
      "learning_rate": 9.384305723052855e-05,
      "loss": 2.782741165161133,
      "memory(GiB)": 66.02,
      "step": 18630,
      "token_acc": 0.43548387096774194,
      "train_speed(iter/s)": 1.448536
    },
    {
      "epoch": 0.7983805321108779,
      "grad_norm": 4.610132694244385,
      "learning_rate": 9.383982153721884e-05,
      "loss": 2.7461664199829103,
      "memory(GiB)": 66.02,
      "step": 18635,
      "token_acc": 0.46774193548387094,
      "train_speed(iter/s)": 1.448552
    },
    {
      "epoch": 0.7985947474401268,
      "grad_norm": 4.440431594848633,
      "learning_rate": 9.383658504970965e-05,
      "loss": 2.855160140991211,
      "memory(GiB)": 66.02,
      "step": 18640,
      "token_acc": 0.44866920152091255,
      "train_speed(iter/s)": 1.448559
    },
    {
      "epoch": 0.7988089627693757,
      "grad_norm": 4.1292405128479,
      "learning_rate": 9.383334776805958e-05,
      "loss": 2.6430744171142577,
      "memory(GiB)": 66.02,
      "step": 18645,
      "token_acc": 0.4117647058823529,
      "train_speed(iter/s)": 1.448451
    },
    {
      "epoch": 0.7990231780986248,
      "grad_norm": 4.249571323394775,
      "learning_rate": 9.383010969232731e-05,
      "loss": 2.4393062591552734,
      "memory(GiB)": 66.02,
      "step": 18650,
      "token_acc": 0.4690909090909091,
      "train_speed(iter/s)": 1.448438
    },
    {
      "epoch": 0.7992373934278737,
      "grad_norm": 5.063155174255371,
      "learning_rate": 9.382687082257149e-05,
      "loss": 2.7747039794921875,
      "memory(GiB)": 66.02,
      "step": 18655,
      "token_acc": 0.4107142857142857,
      "train_speed(iter/s)": 1.448508
    },
    {
      "epoch": 0.7994516087571226,
      "grad_norm": 4.814256191253662,
      "learning_rate": 9.38236311588508e-05,
      "loss": 2.858232116699219,
      "memory(GiB)": 66.02,
      "step": 18660,
      "token_acc": 0.44405594405594406,
      "train_speed(iter/s)": 1.448423
    },
    {
      "epoch": 0.7996658240863717,
      "grad_norm": 3.941995620727539,
      "learning_rate": 9.38203907012239e-05,
      "loss": 2.9491191864013673,
      "memory(GiB)": 66.02,
      "step": 18665,
      "token_acc": 0.3942652329749104,
      "train_speed(iter/s)": 1.448526
    },
    {
      "epoch": 0.7998800394156206,
      "grad_norm": 4.696353912353516,
      "learning_rate": 9.381714944974953e-05,
      "loss": 2.374323272705078,
      "memory(GiB)": 66.02,
      "step": 18670,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.448551
    },
    {
      "epoch": 0.8000942547448695,
      "grad_norm": 4.271224021911621,
      "learning_rate": 9.38139074044864e-05,
      "loss": 2.416745185852051,
      "memory(GiB)": 66.02,
      "step": 18675,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.448411
    },
    {
      "epoch": 0.8003084700741185,
      "grad_norm": 5.279973030090332,
      "learning_rate": 9.381066456549321e-05,
      "loss": 2.6077295303344727,
      "memory(GiB)": 66.02,
      "step": 18680,
      "token_acc": 0.4681647940074906,
      "train_speed(iter/s)": 1.448442
    },
    {
      "epoch": 0.8005226854033675,
      "grad_norm": 4.325677871704102,
      "learning_rate": 9.380742093282876e-05,
      "loss": 2.5408451080322267,
      "memory(GiB)": 66.02,
      "step": 18685,
      "token_acc": 0.45229681978798586,
      "train_speed(iter/s)": 1.448442
    },
    {
      "epoch": 0.8007369007326164,
      "grad_norm": 4.190362453460693,
      "learning_rate": 9.380417650655178e-05,
      "loss": 2.664555549621582,
      "memory(GiB)": 66.02,
      "step": 18690,
      "token_acc": 0.44366197183098594,
      "train_speed(iter/s)": 1.448489
    },
    {
      "epoch": 0.8009511160618654,
      "grad_norm": 3.899315357208252,
      "learning_rate": 9.380093128672104e-05,
      "loss": 2.51683349609375,
      "memory(GiB)": 66.02,
      "step": 18695,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.448476
    },
    {
      "epoch": 0.8011653313911143,
      "grad_norm": 4.963001251220703,
      "learning_rate": 9.379768527339534e-05,
      "loss": 2.554346466064453,
      "memory(GiB)": 66.02,
      "step": 18700,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.448524
    },
    {
      "epoch": 0.8013795467203633,
      "grad_norm": 4.530016899108887,
      "learning_rate": 9.379443846663349e-05,
      "loss": 2.560807991027832,
      "memory(GiB)": 66.02,
      "step": 18705,
      "token_acc": 0.45229681978798586,
      "train_speed(iter/s)": 1.448411
    },
    {
      "epoch": 0.8015937620496123,
      "grad_norm": 4.441322326660156,
      "learning_rate": 9.379119086649432e-05,
      "loss": 2.585462188720703,
      "memory(GiB)": 66.02,
      "step": 18710,
      "token_acc": 0.449685534591195,
      "train_speed(iter/s)": 1.448434
    },
    {
      "epoch": 0.8018079773788612,
      "grad_norm": 3.917233467102051,
      "learning_rate": 9.378794247303664e-05,
      "loss": 2.6268999099731447,
      "memory(GiB)": 66.02,
      "step": 18715,
      "token_acc": 0.4819277108433735,
      "train_speed(iter/s)": 1.448454
    },
    {
      "epoch": 0.8020221927081101,
      "grad_norm": 4.160663604736328,
      "learning_rate": 9.378469328631929e-05,
      "loss": 2.4982261657714844,
      "memory(GiB)": 66.02,
      "step": 18720,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.448509
    },
    {
      "epoch": 0.8022364080373592,
      "grad_norm": 4.181886672973633,
      "learning_rate": 9.378144330640118e-05,
      "loss": 2.4761966705322265,
      "memory(GiB)": 66.02,
      "step": 18725,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.448536
    },
    {
      "epoch": 0.8024506233666081,
      "grad_norm": 6.050583839416504,
      "learning_rate": 9.377819253334113e-05,
      "loss": 2.1962636947631835,
      "memory(GiB)": 66.02,
      "step": 18730,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.448589
    },
    {
      "epoch": 0.802664838695857,
      "grad_norm": 4.506659984588623,
      "learning_rate": 9.377494096719805e-05,
      "loss": 2.6051998138427734,
      "memory(GiB)": 66.02,
      "step": 18735,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.448622
    },
    {
      "epoch": 0.8028790540251061,
      "grad_norm": 3.6847870349884033,
      "learning_rate": 9.377168860803085e-05,
      "loss": 2.5616291046142576,
      "memory(GiB)": 66.02,
      "step": 18740,
      "token_acc": 0.44876325088339225,
      "train_speed(iter/s)": 1.448687
    },
    {
      "epoch": 0.803093269354355,
      "grad_norm": 4.197353839874268,
      "learning_rate": 9.376843545589846e-05,
      "loss": 3.0736923217773438,
      "memory(GiB)": 66.02,
      "step": 18745,
      "token_acc": 0.3853820598006645,
      "train_speed(iter/s)": 1.448734
    },
    {
      "epoch": 0.8033074846836039,
      "grad_norm": 4.323554992675781,
      "learning_rate": 9.37651815108598e-05,
      "loss": 2.567486381530762,
      "memory(GiB)": 66.02,
      "step": 18750,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.448703
    },
    {
      "epoch": 0.803521700012853,
      "grad_norm": 9.776946067810059,
      "learning_rate": 9.376192677297383e-05,
      "loss": 2.5702493667602537,
      "memory(GiB)": 66.02,
      "step": 18755,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.448759
    },
    {
      "epoch": 0.8037359153421019,
      "grad_norm": 7.530881881713867,
      "learning_rate": 9.37586712422995e-05,
      "loss": 2.0877525329589846,
      "memory(GiB)": 66.02,
      "step": 18760,
      "token_acc": 0.5382059800664452,
      "train_speed(iter/s)": 1.448722
    },
    {
      "epoch": 0.8039501306713508,
      "grad_norm": 4.051812171936035,
      "learning_rate": 9.375541491889579e-05,
      "loss": 2.8175514221191404,
      "memory(GiB)": 66.02,
      "step": 18765,
      "token_acc": 0.41605839416058393,
      "train_speed(iter/s)": 1.448692
    },
    {
      "epoch": 0.8041643460005998,
      "grad_norm": 6.542845726013184,
      "learning_rate": 9.375215780282169e-05,
      "loss": 2.9809213638305665,
      "memory(GiB)": 66.02,
      "step": 18770,
      "token_acc": 0.41901408450704225,
      "train_speed(iter/s)": 1.448647
    },
    {
      "epoch": 0.8043785613298488,
      "grad_norm": 4.400534629821777,
      "learning_rate": 9.374889989413622e-05,
      "loss": 2.3808618545532227,
      "memory(GiB)": 66.02,
      "step": 18775,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.448664
    },
    {
      "epoch": 0.8045927766590977,
      "grad_norm": 4.855630397796631,
      "learning_rate": 9.374564119289837e-05,
      "loss": 2.5283443450927736,
      "memory(GiB)": 66.02,
      "step": 18780,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.448757
    },
    {
      "epoch": 0.8048069919883467,
      "grad_norm": 4.238831043243408,
      "learning_rate": 9.37423816991672e-05,
      "loss": 2.5206716537475584,
      "memory(GiB)": 66.02,
      "step": 18785,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.448736
    },
    {
      "epoch": 0.8050212073175956,
      "grad_norm": 5.935600757598877,
      "learning_rate": 9.373912141300177e-05,
      "loss": 2.6006214141845705,
      "memory(GiB)": 66.02,
      "step": 18790,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.448722
    },
    {
      "epoch": 0.8052354226468446,
      "grad_norm": 6.571051597595215,
      "learning_rate": 9.37358603344611e-05,
      "loss": 2.515240478515625,
      "memory(GiB)": 66.02,
      "step": 18795,
      "token_acc": 0.4758364312267658,
      "train_speed(iter/s)": 1.448696
    },
    {
      "epoch": 0.8054496379760936,
      "grad_norm": 4.672883987426758,
      "learning_rate": 9.37325984636043e-05,
      "loss": 2.981559181213379,
      "memory(GiB)": 66.02,
      "step": 18800,
      "token_acc": 0.4307116104868914,
      "train_speed(iter/s)": 1.448665
    },
    {
      "epoch": 0.8056638533053425,
      "grad_norm": 5.836612701416016,
      "learning_rate": 9.372933580049047e-05,
      "loss": 2.3291561126708986,
      "memory(GiB)": 66.02,
      "step": 18805,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.448689
    },
    {
      "epoch": 0.8058780686345914,
      "grad_norm": 4.198070526123047,
      "learning_rate": 9.372607234517868e-05,
      "loss": 2.516164016723633,
      "memory(GiB)": 66.02,
      "step": 18810,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.448664
    },
    {
      "epoch": 0.8060922839638405,
      "grad_norm": 5.6745452880859375,
      "learning_rate": 9.372280809772808e-05,
      "loss": 2.4835153579711915,
      "memory(GiB)": 66.02,
      "step": 18815,
      "token_acc": 0.5059288537549407,
      "train_speed(iter/s)": 1.448702
    },
    {
      "epoch": 0.8063064992930894,
      "grad_norm": 3.5774919986724854,
      "learning_rate": 9.371954305819779e-05,
      "loss": 2.755037307739258,
      "memory(GiB)": 66.02,
      "step": 18820,
      "token_acc": 0.4752186588921283,
      "train_speed(iter/s)": 1.448731
    },
    {
      "epoch": 0.8065207146223383,
      "grad_norm": 4.555241107940674,
      "learning_rate": 9.371627722664698e-05,
      "loss": 2.492181396484375,
      "memory(GiB)": 66.02,
      "step": 18825,
      "token_acc": 0.4402730375426621,
      "train_speed(iter/s)": 1.448703
    },
    {
      "epoch": 0.8067349299515874,
      "grad_norm": 3.933934450149536,
      "learning_rate": 9.371301060313477e-05,
      "loss": 2.0711223602294924,
      "memory(GiB)": 66.02,
      "step": 18830,
      "token_acc": 0.5508771929824562,
      "train_speed(iter/s)": 1.448724
    },
    {
      "epoch": 0.8069491452808363,
      "grad_norm": 4.408028602600098,
      "learning_rate": 9.370974318772038e-05,
      "loss": 2.677082061767578,
      "memory(GiB)": 66.02,
      "step": 18835,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.448778
    },
    {
      "epoch": 0.8071633606100853,
      "grad_norm": 12.37784194946289,
      "learning_rate": 9.370647498046302e-05,
      "loss": 2.6479148864746094,
      "memory(GiB)": 66.02,
      "step": 18840,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.448802
    },
    {
      "epoch": 0.8073775759393342,
      "grad_norm": 4.033270359039307,
      "learning_rate": 9.370320598142183e-05,
      "loss": 2.695201301574707,
      "memory(GiB)": 66.02,
      "step": 18845,
      "token_acc": 0.43309859154929575,
      "train_speed(iter/s)": 1.448749
    },
    {
      "epoch": 0.8075917912685832,
      "grad_norm": 4.585914611816406,
      "learning_rate": 9.369993619065608e-05,
      "loss": 2.5486961364746095,
      "memory(GiB)": 66.02,
      "step": 18850,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.448808
    },
    {
      "epoch": 0.8078060065978322,
      "grad_norm": 4.526068687438965,
      "learning_rate": 9.369666560822498e-05,
      "loss": 2.6846969604492186,
      "memory(GiB)": 66.02,
      "step": 18855,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.448773
    },
    {
      "epoch": 0.8080202219270811,
      "grad_norm": 4.248173713684082,
      "learning_rate": 9.36933942341878e-05,
      "loss": 2.5881731033325197,
      "memory(GiB)": 66.02,
      "step": 18860,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.448796
    },
    {
      "epoch": 0.80823443725633,
      "grad_norm": 5.018126487731934,
      "learning_rate": 9.369012206860381e-05,
      "loss": 2.653145599365234,
      "memory(GiB)": 66.02,
      "step": 18865,
      "token_acc": 0.45353159851301117,
      "train_speed(iter/s)": 1.448831
    },
    {
      "epoch": 0.8084486525855791,
      "grad_norm": 4.510242462158203,
      "learning_rate": 9.368684911153225e-05,
      "loss": 2.7652151107788088,
      "memory(GiB)": 66.02,
      "step": 18870,
      "token_acc": 0.4169278996865204,
      "train_speed(iter/s)": 1.448807
    },
    {
      "epoch": 0.808662867914828,
      "grad_norm": 4.0523762702941895,
      "learning_rate": 9.368357536303245e-05,
      "loss": 2.5270395278930664,
      "memory(GiB)": 66.02,
      "step": 18875,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.448917
    },
    {
      "epoch": 0.8088770832440769,
      "grad_norm": 4.647110462188721,
      "learning_rate": 9.36803008231637e-05,
      "loss": 2.404300308227539,
      "memory(GiB)": 66.02,
      "step": 18880,
      "token_acc": 0.5317725752508361,
      "train_speed(iter/s)": 1.448993
    },
    {
      "epoch": 0.809091298573326,
      "grad_norm": 4.7010321617126465,
      "learning_rate": 9.367702549198533e-05,
      "loss": 2.342572021484375,
      "memory(GiB)": 66.02,
      "step": 18885,
      "token_acc": 0.5100401606425703,
      "train_speed(iter/s)": 1.449005
    },
    {
      "epoch": 0.8093055139025749,
      "grad_norm": 3.4574878215789795,
      "learning_rate": 9.367374936955666e-05,
      "loss": 2.5595396041870115,
      "memory(GiB)": 66.02,
      "step": 18890,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.448895
    },
    {
      "epoch": 0.8095197292318238,
      "grad_norm": 5.867266654968262,
      "learning_rate": 9.367047245593705e-05,
      "loss": 2.6523746490478515,
      "memory(GiB)": 66.02,
      "step": 18895,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.448933
    },
    {
      "epoch": 0.8097339445610728,
      "grad_norm": 3.5428473949432373,
      "learning_rate": 9.366719475118588e-05,
      "loss": 2.5579483032226564,
      "memory(GiB)": 66.02,
      "step": 18900,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.449043
    },
    {
      "epoch": 0.8099481598903218,
      "grad_norm": 4.248826026916504,
      "learning_rate": 9.36639162553625e-05,
      "loss": 2.466800308227539,
      "memory(GiB)": 66.02,
      "step": 18905,
      "token_acc": 0.47575757575757577,
      "train_speed(iter/s)": 1.449012
    },
    {
      "epoch": 0.8101623752195707,
      "grad_norm": 6.2241997718811035,
      "learning_rate": 9.366063696852634e-05,
      "loss": 2.4950037002563477,
      "memory(GiB)": 66.02,
      "step": 18910,
      "token_acc": 0.44565217391304346,
      "train_speed(iter/s)": 1.448962
    },
    {
      "epoch": 0.8103765905488197,
      "grad_norm": 4.128664016723633,
      "learning_rate": 9.365735689073676e-05,
      "loss": 2.2902360916137696,
      "memory(GiB)": 66.02,
      "step": 18915,
      "token_acc": 0.55,
      "train_speed(iter/s)": 1.449014
    },
    {
      "epoch": 0.8105908058780686,
      "grad_norm": 3.3307580947875977,
      "learning_rate": 9.365407602205322e-05,
      "loss": 2.375588035583496,
      "memory(GiB)": 66.02,
      "step": 18920,
      "token_acc": 0.5689655172413793,
      "train_speed(iter/s)": 1.449038
    },
    {
      "epoch": 0.8108050212073176,
      "grad_norm": 4.561521053314209,
      "learning_rate": 9.365079436253513e-05,
      "loss": 2.688313102722168,
      "memory(GiB)": 66.02,
      "step": 18925,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.44909
    },
    {
      "epoch": 0.8110192365365666,
      "grad_norm": 3.610499143600464,
      "learning_rate": 9.364751191224197e-05,
      "loss": 2.392082405090332,
      "memory(GiB)": 66.02,
      "step": 18930,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.449154
    },
    {
      "epoch": 0.8112334518658155,
      "grad_norm": 3.648470878601074,
      "learning_rate": 9.364422867123317e-05,
      "loss": 2.3470903396606446,
      "memory(GiB)": 66.02,
      "step": 18935,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.449157
    },
    {
      "epoch": 0.8114476671950644,
      "grad_norm": 4.74207067489624,
      "learning_rate": 9.364094463956823e-05,
      "loss": 2.464779281616211,
      "memory(GiB)": 66.02,
      "step": 18940,
      "token_acc": 0.47151898734177217,
      "train_speed(iter/s)": 1.449103
    },
    {
      "epoch": 0.8116618825243135,
      "grad_norm": 5.336584091186523,
      "learning_rate": 9.363765981730664e-05,
      "loss": 2.446050262451172,
      "memory(GiB)": 66.02,
      "step": 18945,
      "token_acc": 0.4749034749034749,
      "train_speed(iter/s)": 1.449055
    },
    {
      "epoch": 0.8118760978535624,
      "grad_norm": 3.5935122966766357,
      "learning_rate": 9.363437420450791e-05,
      "loss": 2.469440460205078,
      "memory(GiB)": 66.02,
      "step": 18950,
      "token_acc": 0.5031446540880503,
      "train_speed(iter/s)": 1.448962
    },
    {
      "epoch": 0.8120903131828113,
      "grad_norm": 4.4591875076293945,
      "learning_rate": 9.363108780123155e-05,
      "loss": 2.7129817962646485,
      "memory(GiB)": 66.02,
      "step": 18955,
      "token_acc": 0.48598130841121495,
      "train_speed(iter/s)": 1.448949
    },
    {
      "epoch": 0.8123045285120604,
      "grad_norm": 6.068639278411865,
      "learning_rate": 9.362780060753712e-05,
      "loss": 2.301922416687012,
      "memory(GiB)": 66.02,
      "step": 18960,
      "token_acc": 0.5280898876404494,
      "train_speed(iter/s)": 1.448954
    },
    {
      "epoch": 0.8125187438413093,
      "grad_norm": 4.075112342834473,
      "learning_rate": 9.362451262348414e-05,
      "loss": 2.8235143661499023,
      "memory(GiB)": 66.02,
      "step": 18965,
      "token_acc": 0.4716417910447761,
      "train_speed(iter/s)": 1.449072
    },
    {
      "epoch": 0.8127329591705582,
      "grad_norm": 6.5669708251953125,
      "learning_rate": 9.362122384913221e-05,
      "loss": 2.4278364181518555,
      "memory(GiB)": 66.02,
      "step": 18970,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.44893
    },
    {
      "epoch": 0.8129471744998072,
      "grad_norm": 3.743220329284668,
      "learning_rate": 9.361793428454087e-05,
      "loss": 2.736924171447754,
      "memory(GiB)": 66.02,
      "step": 18975,
      "token_acc": 0.4437299035369775,
      "train_speed(iter/s)": 1.449007
    },
    {
      "epoch": 0.8131613898290562,
      "grad_norm": 5.922600746154785,
      "learning_rate": 9.361464392976975e-05,
      "loss": 2.7900842666625976,
      "memory(GiB)": 66.02,
      "step": 18980,
      "token_acc": 0.43548387096774194,
      "train_speed(iter/s)": 1.449005
    },
    {
      "epoch": 0.8133756051583051,
      "grad_norm": 4.516504287719727,
      "learning_rate": 9.361135278487843e-05,
      "loss": 2.632344055175781,
      "memory(GiB)": 66.02,
      "step": 18985,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.448967
    },
    {
      "epoch": 0.8135898204875541,
      "grad_norm": 3.773824453353882,
      "learning_rate": 9.360806084992655e-05,
      "loss": 2.6191478729248048,
      "memory(GiB)": 66.02,
      "step": 18990,
      "token_acc": 0.45925925925925926,
      "train_speed(iter/s)": 1.448993
    },
    {
      "epoch": 0.813804035816803,
      "grad_norm": 3.8800442218780518,
      "learning_rate": 9.360476812497374e-05,
      "loss": 2.7537099838256838,
      "memory(GiB)": 66.02,
      "step": 18995,
      "token_acc": 0.4085714285714286,
      "train_speed(iter/s)": 1.448961
    },
    {
      "epoch": 0.814018251146052,
      "grad_norm": 4.386427402496338,
      "learning_rate": 9.360147461007964e-05,
      "loss": 2.56646728515625,
      "memory(GiB)": 66.02,
      "step": 19000,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.448939
    },
    {
      "epoch": 0.814018251146052,
      "eval_loss": 2.147026777267456,
      "eval_runtime": 14.6797,
      "eval_samples_per_second": 6.812,
      "eval_steps_per_second": 6.812,
      "eval_token_acc": 0.4878048780487805,
      "step": 19000
    },
    {
      "epoch": 0.814232466475301,
      "grad_norm": 4.334740161895752,
      "learning_rate": 9.359818030530394e-05,
      "loss": 2.6520919799804688,
      "memory(GiB)": 66.02,
      "step": 19005,
      "token_acc": 0.4805194805194805,
      "train_speed(iter/s)": 1.447267
    },
    {
      "epoch": 0.8144466818045499,
      "grad_norm": 4.401294231414795,
      "learning_rate": 9.359488521070629e-05,
      "loss": 2.2585668563842773,
      "memory(GiB)": 66.02,
      "step": 19010,
      "token_acc": 0.5155038759689923,
      "train_speed(iter/s)": 1.447278
    },
    {
      "epoch": 0.8146608971337989,
      "grad_norm": 3.8697566986083984,
      "learning_rate": 9.359158932634642e-05,
      "loss": 2.6603321075439452,
      "memory(GiB)": 66.02,
      "step": 19015,
      "token_acc": 0.4557377049180328,
      "train_speed(iter/s)": 1.447262
    },
    {
      "epoch": 0.8148751124630479,
      "grad_norm": 4.988037109375,
      "learning_rate": 9.3588292652284e-05,
      "loss": 2.3017856597900392,
      "memory(GiB)": 66.02,
      "step": 19020,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.447174
    },
    {
      "epoch": 0.8150893277922968,
      "grad_norm": 4.837860107421875,
      "learning_rate": 9.358499518857877e-05,
      "loss": 2.5823156356811525,
      "memory(GiB)": 66.02,
      "step": 19025,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.44721
    },
    {
      "epoch": 0.8153035431215457,
      "grad_norm": 3.803999185562134,
      "learning_rate": 9.358169693529048e-05,
      "loss": 2.7338947296142577,
      "memory(GiB)": 66.02,
      "step": 19030,
      "token_acc": 0.45768025078369906,
      "train_speed(iter/s)": 1.447154
    },
    {
      "epoch": 0.8155177584507948,
      "grad_norm": 3.702883243560791,
      "learning_rate": 9.357839789247886e-05,
      "loss": 2.5634740829467773,
      "memory(GiB)": 66.02,
      "step": 19035,
      "token_acc": 0.4536741214057508,
      "train_speed(iter/s)": 1.447201
    },
    {
      "epoch": 0.8157319737800437,
      "grad_norm": 5.722247123718262,
      "learning_rate": 9.357509806020369e-05,
      "loss": 2.6622148513793946,
      "memory(GiB)": 66.02,
      "step": 19040,
      "token_acc": 0.4620938628158845,
      "train_speed(iter/s)": 1.447189
    },
    {
      "epoch": 0.8159461891092926,
      "grad_norm": 3.2920172214508057,
      "learning_rate": 9.357179743852471e-05,
      "loss": 2.6106306076049806,
      "memory(GiB)": 66.02,
      "step": 19045,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.447124
    },
    {
      "epoch": 0.8161604044385417,
      "grad_norm": 4.404176712036133,
      "learning_rate": 9.356849602750177e-05,
      "loss": 2.526082229614258,
      "memory(GiB)": 66.02,
      "step": 19050,
      "token_acc": 0.4519230769230769,
      "train_speed(iter/s)": 1.447111
    },
    {
      "epoch": 0.8163746197677906,
      "grad_norm": 6.708415508270264,
      "learning_rate": 9.356519382719467e-05,
      "loss": 2.5774837493896485,
      "memory(GiB)": 66.02,
      "step": 19055,
      "token_acc": 0.4470284237726098,
      "train_speed(iter/s)": 1.447098
    },
    {
      "epoch": 0.8165888350970395,
      "grad_norm": 4.452574729919434,
      "learning_rate": 9.356189083766318e-05,
      "loss": 2.672476577758789,
      "memory(GiB)": 66.02,
      "step": 19060,
      "token_acc": 0.4232081911262799,
      "train_speed(iter/s)": 1.447156
    },
    {
      "epoch": 0.8168030504262885,
      "grad_norm": 7.4739484786987305,
      "learning_rate": 9.35585870589672e-05,
      "loss": 2.913471984863281,
      "memory(GiB)": 66.02,
      "step": 19065,
      "token_acc": 0.4573643410852713,
      "train_speed(iter/s)": 1.447248
    },
    {
      "epoch": 0.8170172657555375,
      "grad_norm": 4.520939350128174,
      "learning_rate": 9.355528249116653e-05,
      "loss": 2.40564079284668,
      "memory(GiB)": 66.02,
      "step": 19070,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.447187
    },
    {
      "epoch": 0.8172314810847864,
      "grad_norm": 3.5942249298095703,
      "learning_rate": 9.355197713432109e-05,
      "loss": 2.7534931182861326,
      "memory(GiB)": 66.02,
      "step": 19075,
      "token_acc": 0.459546925566343,
      "train_speed(iter/s)": 1.447116
    },
    {
      "epoch": 0.8174456964140354,
      "grad_norm": 3.7783420085906982,
      "learning_rate": 9.354867098849071e-05,
      "loss": 2.6439193725585937,
      "memory(GiB)": 66.02,
      "step": 19080,
      "token_acc": 0.474025974025974,
      "train_speed(iter/s)": 1.446914
    },
    {
      "epoch": 0.8176599117432843,
      "grad_norm": 4.424737453460693,
      "learning_rate": 9.35453640537353e-05,
      "loss": 2.5768512725830077,
      "memory(GiB)": 66.02,
      "step": 19085,
      "token_acc": 0.4558303886925795,
      "train_speed(iter/s)": 1.446766
    },
    {
      "epoch": 0.8178741270725333,
      "grad_norm": 4.611454486846924,
      "learning_rate": 9.354205633011479e-05,
      "loss": 2.604601287841797,
      "memory(GiB)": 66.02,
      "step": 19090,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.446845
    },
    {
      "epoch": 0.8180883424017823,
      "grad_norm": 3.5911543369293213,
      "learning_rate": 9.353874781768908e-05,
      "loss": 2.3541257858276365,
      "memory(GiB)": 66.02,
      "step": 19095,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.446916
    },
    {
      "epoch": 0.8183025577310312,
      "grad_norm": 4.482860088348389,
      "learning_rate": 9.353543851651809e-05,
      "loss": 2.7378005981445312,
      "memory(GiB)": 66.02,
      "step": 19100,
      "token_acc": 0.4294117647058823,
      "train_speed(iter/s)": 1.446936
    },
    {
      "epoch": 0.8185167730602801,
      "grad_norm": 5.56183385848999,
      "learning_rate": 9.353212842666181e-05,
      "loss": 2.664101791381836,
      "memory(GiB)": 66.02,
      "step": 19105,
      "token_acc": 0.48249027237354086,
      "train_speed(iter/s)": 1.446943
    },
    {
      "epoch": 0.8187309883895292,
      "grad_norm": 3.9395222663879395,
      "learning_rate": 9.352881754818019e-05,
      "loss": 2.949480438232422,
      "memory(GiB)": 66.02,
      "step": 19110,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.447034
    },
    {
      "epoch": 0.8189452037187781,
      "grad_norm": 4.044198036193848,
      "learning_rate": 9.352550588113319e-05,
      "loss": 2.3677745819091798,
      "memory(GiB)": 66.02,
      "step": 19115,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.447083
    },
    {
      "epoch": 0.819159419048027,
      "grad_norm": 4.579442977905273,
      "learning_rate": 9.352219342558083e-05,
      "loss": 2.8724830627441404,
      "memory(GiB)": 66.02,
      "step": 19120,
      "token_acc": 0.4146341463414634,
      "train_speed(iter/s)": 1.447095
    },
    {
      "epoch": 0.8193736343772761,
      "grad_norm": 3.308806896209717,
      "learning_rate": 9.351888018158312e-05,
      "loss": 2.7751276016235353,
      "memory(GiB)": 66.02,
      "step": 19125,
      "token_acc": 0.4289940828402367,
      "train_speed(iter/s)": 1.447154
    },
    {
      "epoch": 0.819587849706525,
      "grad_norm": 5.318056106567383,
      "learning_rate": 9.351556614920005e-05,
      "loss": 2.530400276184082,
      "memory(GiB)": 66.02,
      "step": 19130,
      "token_acc": 0.48497854077253216,
      "train_speed(iter/s)": 1.44714
    },
    {
      "epoch": 0.8198020650357739,
      "grad_norm": 4.2228522300720215,
      "learning_rate": 9.351225132849169e-05,
      "loss": 2.9578628540039062,
      "memory(GiB)": 66.02,
      "step": 19135,
      "token_acc": 0.40773809523809523,
      "train_speed(iter/s)": 1.447156
    },
    {
      "epoch": 0.8200162803650229,
      "grad_norm": 4.344516754150391,
      "learning_rate": 9.350893571951808e-05,
      "loss": 2.6064916610717774,
      "memory(GiB)": 66.02,
      "step": 19140,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.447135
    },
    {
      "epoch": 0.8202304956942719,
      "grad_norm": 5.083049774169922,
      "learning_rate": 9.350561932233928e-05,
      "loss": 2.6279670715332033,
      "memory(GiB)": 66.02,
      "step": 19145,
      "token_acc": 0.5247524752475248,
      "train_speed(iter/s)": 1.447138
    },
    {
      "epoch": 0.8204447110235208,
      "grad_norm": 7.079374313354492,
      "learning_rate": 9.350230213701537e-05,
      "loss": 2.7200139999389648,
      "memory(GiB)": 66.02,
      "step": 19150,
      "token_acc": 0.445578231292517,
      "train_speed(iter/s)": 1.447093
    },
    {
      "epoch": 0.8206589263527698,
      "grad_norm": 4.158751487731934,
      "learning_rate": 9.349898416360646e-05,
      "loss": 2.7096811294555665,
      "memory(GiB)": 66.02,
      "step": 19155,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.447078
    },
    {
      "epoch": 0.8208731416820187,
      "grad_norm": 3.67203688621521,
      "learning_rate": 9.349566540217265e-05,
      "loss": 2.7657135009765623,
      "memory(GiB)": 66.02,
      "step": 19160,
      "token_acc": 0.4511784511784512,
      "train_speed(iter/s)": 1.44708
    },
    {
      "epoch": 0.8210873570112677,
      "grad_norm": 5.364586353302002,
      "learning_rate": 9.349234585277404e-05,
      "loss": 2.5590272903442384,
      "memory(GiB)": 66.02,
      "step": 19165,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.447039
    },
    {
      "epoch": 0.8213015723405167,
      "grad_norm": 4.690365791320801,
      "learning_rate": 9.348902551547081e-05,
      "loss": 2.5245126724243163,
      "memory(GiB)": 66.02,
      "step": 19170,
      "token_acc": 0.4868913857677903,
      "train_speed(iter/s)": 1.447057
    },
    {
      "epoch": 0.8215157876697656,
      "grad_norm": 4.606238842010498,
      "learning_rate": 9.348570439032306e-05,
      "loss": 2.4096059799194336,
      "memory(GiB)": 66.02,
      "step": 19175,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.447084
    },
    {
      "epoch": 0.8217300029990147,
      "grad_norm": 5.548707008361816,
      "learning_rate": 9.348238247739101e-05,
      "loss": 2.740232467651367,
      "memory(GiB)": 66.02,
      "step": 19180,
      "token_acc": 0.46218487394957986,
      "train_speed(iter/s)": 1.44709
    },
    {
      "epoch": 0.8219442183282636,
      "grad_norm": 3.649252414703369,
      "learning_rate": 9.34790597767348e-05,
      "loss": 2.513283538818359,
      "memory(GiB)": 66.02,
      "step": 19185,
      "token_acc": 0.483695652173913,
      "train_speed(iter/s)": 1.44716
    },
    {
      "epoch": 0.8221584336575125,
      "grad_norm": 4.202706336975098,
      "learning_rate": 9.347573628841462e-05,
      "loss": 2.3912729263305663,
      "memory(GiB)": 66.02,
      "step": 19190,
      "token_acc": 0.4595375722543353,
      "train_speed(iter/s)": 1.447188
    },
    {
      "epoch": 0.8223726489867615,
      "grad_norm": 5.691158771514893,
      "learning_rate": 9.347241201249071e-05,
      "loss": 2.98565559387207,
      "memory(GiB)": 66.02,
      "step": 19195,
      "token_acc": 0.432258064516129,
      "train_speed(iter/s)": 1.447173
    },
    {
      "epoch": 0.8225868643160105,
      "grad_norm": 4.348105430603027,
      "learning_rate": 9.346908694902327e-05,
      "loss": 2.437234306335449,
      "memory(GiB)": 66.02,
      "step": 19200,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.447066
    },
    {
      "epoch": 0.8228010796452594,
      "grad_norm": 3.9248456954956055,
      "learning_rate": 9.346576109807255e-05,
      "loss": 2.6050180435180663,
      "memory(GiB)": 66.02,
      "step": 19205,
      "token_acc": 0.4743202416918429,
      "train_speed(iter/s)": 1.44715
    },
    {
      "epoch": 0.8230152949745084,
      "grad_norm": 5.578907012939453,
      "learning_rate": 9.346243445969877e-05,
      "loss": 2.5511499404907227,
      "memory(GiB)": 66.02,
      "step": 19210,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.447174
    },
    {
      "epoch": 0.8232295103037574,
      "grad_norm": 4.100949287414551,
      "learning_rate": 9.345910703396225e-05,
      "loss": 2.5610687255859377,
      "memory(GiB)": 66.02,
      "step": 19215,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.447117
    },
    {
      "epoch": 0.8234437256330063,
      "grad_norm": 4.5237531661987305,
      "learning_rate": 9.345577882092322e-05,
      "loss": 2.7338815689086915,
      "memory(GiB)": 66.02,
      "step": 19220,
      "token_acc": 0.43042071197411,
      "train_speed(iter/s)": 1.447062
    },
    {
      "epoch": 0.8236579409622553,
      "grad_norm": 4.642475605010986,
      "learning_rate": 9.345244982064201e-05,
      "loss": 2.573164939880371,
      "memory(GiB)": 66.02,
      "step": 19225,
      "token_acc": 0.4899598393574297,
      "train_speed(iter/s)": 1.447108
    },
    {
      "epoch": 0.8238721562915042,
      "grad_norm": 5.348715782165527,
      "learning_rate": 9.344912003317888e-05,
      "loss": 2.4919498443603514,
      "memory(GiB)": 66.02,
      "step": 19230,
      "token_acc": 0.483739837398374,
      "train_speed(iter/s)": 1.447074
    },
    {
      "epoch": 0.8240863716207532,
      "grad_norm": 3.156010150909424,
      "learning_rate": 9.344578945859421e-05,
      "loss": 2.637017822265625,
      "memory(GiB)": 66.02,
      "step": 19235,
      "token_acc": 0.4520123839009288,
      "train_speed(iter/s)": 1.447122
    },
    {
      "epoch": 0.8243005869500022,
      "grad_norm": 3.584594488143921,
      "learning_rate": 9.344245809694829e-05,
      "loss": 2.5854602813720704,
      "memory(GiB)": 66.02,
      "step": 19240,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.446998
    },
    {
      "epoch": 0.8245148022792511,
      "grad_norm": 5.1024322509765625,
      "learning_rate": 9.34391259483015e-05,
      "loss": 2.683881378173828,
      "memory(GiB)": 66.02,
      "step": 19245,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.44699
    },
    {
      "epoch": 0.8247290176085,
      "grad_norm": 4.292627811431885,
      "learning_rate": 9.343579301271419e-05,
      "loss": 2.5055210113525392,
      "memory(GiB)": 66.02,
      "step": 19250,
      "token_acc": 0.47941176470588237,
      "train_speed(iter/s)": 1.447044
    },
    {
      "epoch": 0.8249432329377491,
      "grad_norm": 6.777864933013916,
      "learning_rate": 9.343245929024674e-05,
      "loss": 2.869306755065918,
      "memory(GiB)": 66.02,
      "step": 19255,
      "token_acc": 0.4592833876221498,
      "train_speed(iter/s)": 1.447
    },
    {
      "epoch": 0.825157448266998,
      "grad_norm": 6.696408271789551,
      "learning_rate": 9.342912478095955e-05,
      "loss": 2.294254493713379,
      "memory(GiB)": 66.02,
      "step": 19260,
      "token_acc": 0.5486381322957199,
      "train_speed(iter/s)": 1.447093
    },
    {
      "epoch": 0.8253716635962469,
      "grad_norm": 4.6820173263549805,
      "learning_rate": 9.342578948491303e-05,
      "loss": 2.6426769256591798,
      "memory(GiB)": 66.02,
      "step": 19265,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.447136
    },
    {
      "epoch": 0.825585878925496,
      "grad_norm": 4.027639389038086,
      "learning_rate": 9.342245340216759e-05,
      "loss": 2.5612937927246096,
      "memory(GiB)": 66.02,
      "step": 19270,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.447137
    },
    {
      "epoch": 0.8258000942547449,
      "grad_norm": 3.4892492294311523,
      "learning_rate": 9.341911653278367e-05,
      "loss": 2.5044734954833983,
      "memory(GiB)": 66.02,
      "step": 19275,
      "token_acc": 0.5150501672240803,
      "train_speed(iter/s)": 1.447244
    },
    {
      "epoch": 0.8260143095839938,
      "grad_norm": 4.824902057647705,
      "learning_rate": 9.341577887682173e-05,
      "loss": 2.8290802001953126,
      "memory(GiB)": 66.02,
      "step": 19280,
      "token_acc": 0.44807121661721067,
      "train_speed(iter/s)": 1.447198
    },
    {
      "epoch": 0.8262285249132428,
      "grad_norm": 4.836545467376709,
      "learning_rate": 9.341244043434224e-05,
      "loss": 2.495344352722168,
      "memory(GiB)": 66.02,
      "step": 19285,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.447192
    },
    {
      "epoch": 0.8264427402424918,
      "grad_norm": 5.338891506195068,
      "learning_rate": 9.340910120540564e-05,
      "loss": 2.9964624404907227,
      "memory(GiB)": 66.02,
      "step": 19290,
      "token_acc": 0.4300341296928328,
      "train_speed(iter/s)": 1.447221
    },
    {
      "epoch": 0.8266569555717407,
      "grad_norm": 4.3757524490356445,
      "learning_rate": 9.340576119007249e-05,
      "loss": 2.865975189208984,
      "memory(GiB)": 66.02,
      "step": 19295,
      "token_acc": 0.4702194357366771,
      "train_speed(iter/s)": 1.447223
    },
    {
      "epoch": 0.8268711709009897,
      "grad_norm": 4.286296844482422,
      "learning_rate": 9.340242038840322e-05,
      "loss": 2.495265007019043,
      "memory(GiB)": 66.02,
      "step": 19300,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.447216
    },
    {
      "epoch": 0.8270853862302386,
      "grad_norm": 3.7346765995025635,
      "learning_rate": 9.33990788004584e-05,
      "loss": 2.980963706970215,
      "memory(GiB)": 66.02,
      "step": 19305,
      "token_acc": 0.4045584045584046,
      "train_speed(iter/s)": 1.447161
    },
    {
      "epoch": 0.8272996015594876,
      "grad_norm": 4.579444885253906,
      "learning_rate": 9.339573642629857e-05,
      "loss": 2.602779579162598,
      "memory(GiB)": 66.02,
      "step": 19310,
      "token_acc": 0.45323741007194246,
      "train_speed(iter/s)": 1.447242
    },
    {
      "epoch": 0.8275138168887366,
      "grad_norm": 4.895353317260742,
      "learning_rate": 9.339239326598426e-05,
      "loss": 2.724811553955078,
      "memory(GiB)": 66.02,
      "step": 19315,
      "token_acc": 0.4398826979472141,
      "train_speed(iter/s)": 1.44729
    },
    {
      "epoch": 0.8277280322179855,
      "grad_norm": 4.600613594055176,
      "learning_rate": 9.338904931957603e-05,
      "loss": 2.6580682754516602,
      "memory(GiB)": 66.02,
      "step": 19320,
      "token_acc": 0.4249201277955272,
      "train_speed(iter/s)": 1.44728
    },
    {
      "epoch": 0.8279422475472344,
      "grad_norm": 4.4494099617004395,
      "learning_rate": 9.338570458713447e-05,
      "loss": 2.9500267028808596,
      "memory(GiB)": 66.02,
      "step": 19325,
      "token_acc": 0.4427710843373494,
      "train_speed(iter/s)": 1.447383
    },
    {
      "epoch": 0.8281564628764835,
      "grad_norm": 4.115397930145264,
      "learning_rate": 9.338235906872019e-05,
      "loss": 2.7892601013183596,
      "memory(GiB)": 66.02,
      "step": 19330,
      "token_acc": 0.4559748427672956,
      "train_speed(iter/s)": 1.447289
    },
    {
      "epoch": 0.8283706782057324,
      "grad_norm": 5.089390754699707,
      "learning_rate": 9.337901276439376e-05,
      "loss": 2.7787311553955076,
      "memory(GiB)": 66.02,
      "step": 19335,
      "token_acc": 0.4550898203592814,
      "train_speed(iter/s)": 1.447272
    },
    {
      "epoch": 0.8285848935349813,
      "grad_norm": 4.719737529754639,
      "learning_rate": 9.337566567421583e-05,
      "loss": 2.5675708770751955,
      "memory(GiB)": 66.02,
      "step": 19340,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.4473
    },
    {
      "epoch": 0.8287991088642304,
      "grad_norm": 4.357607841491699,
      "learning_rate": 9.337231779824703e-05,
      "loss": 2.4175018310546874,
      "memory(GiB)": 66.02,
      "step": 19345,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.447294
    },
    {
      "epoch": 0.8290133241934793,
      "grad_norm": 3.693588972091675,
      "learning_rate": 9.3368969136548e-05,
      "loss": 2.6972537994384767,
      "memory(GiB)": 66.02,
      "step": 19350,
      "token_acc": 0.4397590361445783,
      "train_speed(iter/s)": 1.447209
    },
    {
      "epoch": 0.8292275395227282,
      "grad_norm": 4.481620788574219,
      "learning_rate": 9.33656196891794e-05,
      "loss": 2.5331802368164062,
      "memory(GiB)": 66.02,
      "step": 19355,
      "token_acc": 0.43854748603351956,
      "train_speed(iter/s)": 1.447161
    },
    {
      "epoch": 0.8294417548519772,
      "grad_norm": 3.9050045013427734,
      "learning_rate": 9.336226945620194e-05,
      "loss": 2.417461395263672,
      "memory(GiB)": 66.02,
      "step": 19360,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.44718
    },
    {
      "epoch": 0.8296559701812262,
      "grad_norm": 4.633638858795166,
      "learning_rate": 9.335891843767629e-05,
      "loss": 2.4926803588867186,
      "memory(GiB)": 66.02,
      "step": 19365,
      "token_acc": 0.46774193548387094,
      "train_speed(iter/s)": 1.447125
    },
    {
      "epoch": 0.8298701855104751,
      "grad_norm": 5.955918312072754,
      "learning_rate": 9.335556663366314e-05,
      "loss": 2.575380325317383,
      "memory(GiB)": 66.02,
      "step": 19370,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.447101
    },
    {
      "epoch": 0.8300844008397241,
      "grad_norm": 3.949324607849121,
      "learning_rate": 9.335221404422325e-05,
      "loss": 2.6844150543212892,
      "memory(GiB)": 66.02,
      "step": 19375,
      "token_acc": 0.4247787610619469,
      "train_speed(iter/s)": 1.447212
    },
    {
      "epoch": 0.830298616168973,
      "grad_norm": 5.545995712280273,
      "learning_rate": 9.334886066941733e-05,
      "loss": 2.6630107879638674,
      "memory(GiB)": 66.02,
      "step": 19380,
      "token_acc": 0.468503937007874,
      "train_speed(iter/s)": 1.447275
    },
    {
      "epoch": 0.830512831498222,
      "grad_norm": 5.249592304229736,
      "learning_rate": 9.334550650930613e-05,
      "loss": 2.828268051147461,
      "memory(GiB)": 66.02,
      "step": 19385,
      "token_acc": 0.4006734006734007,
      "train_speed(iter/s)": 1.447229
    },
    {
      "epoch": 0.830727046827471,
      "grad_norm": 4.565025329589844,
      "learning_rate": 9.334215156395042e-05,
      "loss": 2.4273197174072267,
      "memory(GiB)": 66.02,
      "step": 19390,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.447295
    },
    {
      "epoch": 0.8309412621567199,
      "grad_norm": 3.8825125694274902,
      "learning_rate": 9.333879583341098e-05,
      "loss": 2.393206024169922,
      "memory(GiB)": 66.02,
      "step": 19395,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.447316
    },
    {
      "epoch": 0.8311554774859689,
      "grad_norm": 4.470056533813477,
      "learning_rate": 9.33354393177486e-05,
      "loss": 3.019732666015625,
      "memory(GiB)": 66.02,
      "step": 19400,
      "token_acc": 0.42507645259938837,
      "train_speed(iter/s)": 1.447326
    },
    {
      "epoch": 0.8313696928152179,
      "grad_norm": 5.510180950164795,
      "learning_rate": 9.333208201702407e-05,
      "loss": 2.6127593994140623,
      "memory(GiB)": 66.02,
      "step": 19405,
      "token_acc": 0.4339622641509434,
      "train_speed(iter/s)": 1.447254
    },
    {
      "epoch": 0.8315839081444668,
      "grad_norm": 4.819840908050537,
      "learning_rate": 9.332872393129823e-05,
      "loss": 2.9776878356933594,
      "memory(GiB)": 66.02,
      "step": 19410,
      "token_acc": 0.3870967741935484,
      "train_speed(iter/s)": 1.44722
    },
    {
      "epoch": 0.8317981234737157,
      "grad_norm": 3.3434503078460693,
      "learning_rate": 9.332536506063192e-05,
      "loss": 2.2912576675415037,
      "memory(GiB)": 66.02,
      "step": 19415,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.447159
    },
    {
      "epoch": 0.8320123388029648,
      "grad_norm": 4.7845540046691895,
      "learning_rate": 9.332200540508598e-05,
      "loss": 2.5155660629272463,
      "memory(GiB)": 66.02,
      "step": 19420,
      "token_acc": 0.44274809160305345,
      "train_speed(iter/s)": 1.447186
    },
    {
      "epoch": 0.8322265541322137,
      "grad_norm": 4.668683052062988,
      "learning_rate": 9.331864496472125e-05,
      "loss": 2.446110153198242,
      "memory(GiB)": 66.02,
      "step": 19425,
      "token_acc": 0.4862068965517241,
      "train_speed(iter/s)": 1.447191
    },
    {
      "epoch": 0.8324407694614626,
      "grad_norm": 4.651973724365234,
      "learning_rate": 9.331528373959865e-05,
      "loss": 2.647796630859375,
      "memory(GiB)": 66.02,
      "step": 19430,
      "token_acc": 0.43209876543209874,
      "train_speed(iter/s)": 1.447187
    },
    {
      "epoch": 0.8326549847907116,
      "grad_norm": 4.906210899353027,
      "learning_rate": 9.331192172977905e-05,
      "loss": 2.9209243774414064,
      "memory(GiB)": 66.02,
      "step": 19435,
      "token_acc": 0.4377104377104377,
      "train_speed(iter/s)": 1.447161
    },
    {
      "epoch": 0.8328692001199606,
      "grad_norm": 4.209266662597656,
      "learning_rate": 9.330855893532335e-05,
      "loss": 2.649152374267578,
      "memory(GiB)": 66.02,
      "step": 19440,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.447224
    },
    {
      "epoch": 0.8330834154492095,
      "grad_norm": 4.533636569976807,
      "learning_rate": 9.330519535629249e-05,
      "loss": 2.589357376098633,
      "memory(GiB)": 66.02,
      "step": 19445,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.447291
    },
    {
      "epoch": 0.8332976307784585,
      "grad_norm": 3.3974220752716064,
      "learning_rate": 9.330183099274739e-05,
      "loss": 2.4823240280151366,
      "memory(GiB)": 66.02,
      "step": 19450,
      "token_acc": 0.4681647940074906,
      "train_speed(iter/s)": 1.447261
    },
    {
      "epoch": 0.8335118461077075,
      "grad_norm": 3.6404998302459717,
      "learning_rate": 9.329846584474899e-05,
      "loss": 2.297855758666992,
      "memory(GiB)": 66.02,
      "step": 19455,
      "token_acc": 0.5048231511254019,
      "train_speed(iter/s)": 1.447331
    },
    {
      "epoch": 0.8337260614369564,
      "grad_norm": 6.674795627593994,
      "learning_rate": 9.329509991235829e-05,
      "loss": 2.7794775009155273,
      "memory(GiB)": 66.02,
      "step": 19460,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.447463
    },
    {
      "epoch": 0.8339402767662054,
      "grad_norm": 4.150593280792236,
      "learning_rate": 9.329173319563622e-05,
      "loss": 2.540719985961914,
      "memory(GiB)": 66.02,
      "step": 19465,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.447492
    },
    {
      "epoch": 0.8341544920954543,
      "grad_norm": 4.216850757598877,
      "learning_rate": 9.328836569464379e-05,
      "loss": 2.5613162994384764,
      "memory(GiB)": 66.02,
      "step": 19470,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.447413
    },
    {
      "epoch": 0.8343687074247033,
      "grad_norm": 4.351722717285156,
      "learning_rate": 9.328499740944201e-05,
      "loss": 2.4275341033935547,
      "memory(GiB)": 66.02,
      "step": 19475,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.447499
    },
    {
      "epoch": 0.8345829227539523,
      "grad_norm": 5.566346645355225,
      "learning_rate": 9.328162834009192e-05,
      "loss": 2.4238985061645506,
      "memory(GiB)": 66.02,
      "step": 19480,
      "token_acc": 0.4959349593495935,
      "train_speed(iter/s)": 1.447538
    },
    {
      "epoch": 0.8347971380832012,
      "grad_norm": 4.994141578674316,
      "learning_rate": 9.32782584866545e-05,
      "loss": 2.710413932800293,
      "memory(GiB)": 66.02,
      "step": 19485,
      "token_acc": 0.4261744966442953,
      "train_speed(iter/s)": 1.447632
    },
    {
      "epoch": 0.8350113534124501,
      "grad_norm": 4.252994537353516,
      "learning_rate": 9.327488784919084e-05,
      "loss": 2.8690027236938476,
      "memory(GiB)": 66.02,
      "step": 19490,
      "token_acc": 0.4316109422492401,
      "train_speed(iter/s)": 1.447531
    },
    {
      "epoch": 0.8352255687416992,
      "grad_norm": 3.779324769973755,
      "learning_rate": 9.3271516427762e-05,
      "loss": 2.803751754760742,
      "memory(GiB)": 66.02,
      "step": 19495,
      "token_acc": 0.440625,
      "train_speed(iter/s)": 1.447599
    },
    {
      "epoch": 0.8354397840709481,
      "grad_norm": 3.6907694339752197,
      "learning_rate": 9.326814422242905e-05,
      "loss": 2.467353630065918,
      "memory(GiB)": 66.02,
      "step": 19500,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.447611
    },
    {
      "epoch": 0.8354397840709481,
      "eval_loss": 2.274688482284546,
      "eval_runtime": 14.081,
      "eval_samples_per_second": 7.102,
      "eval_steps_per_second": 7.102,
      "eval_token_acc": 0.47075208913649025,
      "step": 19500
    },
    {
      "epoch": 0.835653999400197,
      "grad_norm": 4.500802516937256,
      "learning_rate": 9.326477123325306e-05,
      "loss": 2.769436073303223,
      "memory(GiB)": 66.02,
      "step": 19505,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.446073
    },
    {
      "epoch": 0.8358682147294461,
      "grad_norm": 4.522373199462891,
      "learning_rate": 9.326139746029516e-05,
      "loss": 2.3506784439086914,
      "memory(GiB)": 66.02,
      "step": 19510,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.446172
    },
    {
      "epoch": 0.836082430058695,
      "grad_norm": 4.026937961578369,
      "learning_rate": 9.325802290361647e-05,
      "loss": 2.4716814041137694,
      "memory(GiB)": 66.02,
      "step": 19515,
      "token_acc": 0.439873417721519,
      "train_speed(iter/s)": 1.446215
    },
    {
      "epoch": 0.836296645387944,
      "grad_norm": 6.328344821929932,
      "learning_rate": 9.32546475632781e-05,
      "loss": 2.758492660522461,
      "memory(GiB)": 66.02,
      "step": 19520,
      "token_acc": 0.44106463878326996,
      "train_speed(iter/s)": 1.446229
    },
    {
      "epoch": 0.8365108607171929,
      "grad_norm": 4.819122791290283,
      "learning_rate": 9.325127143934122e-05,
      "loss": 2.472138214111328,
      "memory(GiB)": 66.02,
      "step": 19525,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.446279
    },
    {
      "epoch": 0.8367250760464419,
      "grad_norm": 4.229304790496826,
      "learning_rate": 9.324789453186699e-05,
      "loss": 2.786085319519043,
      "memory(GiB)": 66.02,
      "step": 19530,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.446258
    },
    {
      "epoch": 0.8369392913756909,
      "grad_norm": 3.706948757171631,
      "learning_rate": 9.32445168409166e-05,
      "loss": 2.5545644760131836,
      "memory(GiB)": 66.02,
      "step": 19535,
      "token_acc": 0.5035211267605634,
      "train_speed(iter/s)": 1.446341
    },
    {
      "epoch": 0.8371535067049398,
      "grad_norm": 5.0965986251831055,
      "learning_rate": 9.324113836655119e-05,
      "loss": 2.3366153717041014,
      "memory(GiB)": 66.02,
      "step": 19540,
      "token_acc": 0.5020242914979757,
      "train_speed(iter/s)": 1.446421
    },
    {
      "epoch": 0.8373677220341887,
      "grad_norm": 3.7695391178131104,
      "learning_rate": 9.323775910883202e-05,
      "loss": 2.6092723846435546,
      "memory(GiB)": 66.02,
      "step": 19545,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.446457
    },
    {
      "epoch": 0.8375819373634378,
      "grad_norm": 3.9775147438049316,
      "learning_rate": 9.323437906782026e-05,
      "loss": 2.418645668029785,
      "memory(GiB)": 66.02,
      "step": 19550,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.446459
    },
    {
      "epoch": 0.8377961526926867,
      "grad_norm": 6.02524471282959,
      "learning_rate": 9.323099824357717e-05,
      "loss": 2.733505630493164,
      "memory(GiB)": 66.02,
      "step": 19555,
      "token_acc": 0.4402730375426621,
      "train_speed(iter/s)": 1.446475
    },
    {
      "epoch": 0.8380103680219356,
      "grad_norm": 4.156128883361816,
      "learning_rate": 9.3227616636164e-05,
      "loss": 2.5118431091308593,
      "memory(GiB)": 66.02,
      "step": 19560,
      "token_acc": 0.5037878787878788,
      "train_speed(iter/s)": 1.44652
    },
    {
      "epoch": 0.8382245833511847,
      "grad_norm": 4.817605972290039,
      "learning_rate": 9.322423424564201e-05,
      "loss": 2.7719282150268554,
      "memory(GiB)": 66.02,
      "step": 19565,
      "token_acc": 0.4382716049382716,
      "train_speed(iter/s)": 1.446475
    },
    {
      "epoch": 0.8384387986804336,
      "grad_norm": 4.123440265655518,
      "learning_rate": 9.322085107207245e-05,
      "loss": 2.5501338958740236,
      "memory(GiB)": 66.02,
      "step": 19570,
      "token_acc": 0.49570200573065903,
      "train_speed(iter/s)": 1.446514
    },
    {
      "epoch": 0.8386530140096825,
      "grad_norm": 4.149228096008301,
      "learning_rate": 9.321746711551665e-05,
      "loss": 2.702737045288086,
      "memory(GiB)": 66.02,
      "step": 19575,
      "token_acc": 0.4609375,
      "train_speed(iter/s)": 1.4465
    },
    {
      "epoch": 0.8388672293389315,
      "grad_norm": 5.247049808502197,
      "learning_rate": 9.321408237603589e-05,
      "loss": 2.679691696166992,
      "memory(GiB)": 66.02,
      "step": 19580,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.446498
    },
    {
      "epoch": 0.8390814446681805,
      "grad_norm": 6.044321537017822,
      "learning_rate": 9.321069685369147e-05,
      "loss": 2.8071435928344726,
      "memory(GiB)": 66.02,
      "step": 19585,
      "token_acc": 0.46530612244897956,
      "train_speed(iter/s)": 1.446551
    },
    {
      "epoch": 0.8392956599974294,
      "grad_norm": 3.886892080307007,
      "learning_rate": 9.320731054854474e-05,
      "loss": 2.3325725555419923,
      "memory(GiB)": 66.02,
      "step": 19590,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.446637
    },
    {
      "epoch": 0.8395098753266784,
      "grad_norm": 3.09775447845459,
      "learning_rate": 9.320392346065707e-05,
      "loss": 2.4970985412597657,
      "memory(GiB)": 66.02,
      "step": 19595,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.446757
    },
    {
      "epoch": 0.8397240906559273,
      "grad_norm": 4.860101222991943,
      "learning_rate": 9.320053559008979e-05,
      "loss": 2.8612951278686523,
      "memory(GiB)": 66.02,
      "step": 19600,
      "token_acc": 0.46394984326018807,
      "train_speed(iter/s)": 1.44684
    },
    {
      "epoch": 0.8399383059851763,
      "grad_norm": 5.01551628112793,
      "learning_rate": 9.319714693690429e-05,
      "loss": 2.6764516830444336,
      "memory(GiB)": 66.02,
      "step": 19605,
      "token_acc": 0.44144144144144143,
      "train_speed(iter/s)": 1.446925
    },
    {
      "epoch": 0.8401525213144253,
      "grad_norm": 6.008364200592041,
      "learning_rate": 9.319375750116194e-05,
      "loss": 2.810929870605469,
      "memory(GiB)": 66.02,
      "step": 19610,
      "token_acc": 0.4401294498381877,
      "train_speed(iter/s)": 1.446874
    },
    {
      "epoch": 0.8403667366436742,
      "grad_norm": 5.056024551391602,
      "learning_rate": 9.319036728292415e-05,
      "loss": 2.5516054153442385,
      "memory(GiB)": 66.02,
      "step": 19615,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.446906
    },
    {
      "epoch": 0.8405809519729232,
      "grad_norm": 4.122622966766357,
      "learning_rate": 9.318697628225235e-05,
      "loss": 2.4294466018676757,
      "memory(GiB)": 66.02,
      "step": 19620,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.446917
    },
    {
      "epoch": 0.8407951673021722,
      "grad_norm": 4.909486293792725,
      "learning_rate": 9.318358449920795e-05,
      "loss": 2.573863983154297,
      "memory(GiB)": 66.02,
      "step": 19625,
      "token_acc": 0.46440677966101696,
      "train_speed(iter/s)": 1.447004
    },
    {
      "epoch": 0.8410093826314211,
      "grad_norm": 3.7781527042388916,
      "learning_rate": 9.318019193385242e-05,
      "loss": 2.692704772949219,
      "memory(GiB)": 66.02,
      "step": 19630,
      "token_acc": 0.4472049689440994,
      "train_speed(iter/s)": 1.447017
    },
    {
      "epoch": 0.84122359796067,
      "grad_norm": 4.423552989959717,
      "learning_rate": 9.317679858624721e-05,
      "loss": 2.7754079818725588,
      "memory(GiB)": 66.02,
      "step": 19635,
      "token_acc": 0.4146341463414634,
      "train_speed(iter/s)": 1.446982
    },
    {
      "epoch": 0.8414378132899191,
      "grad_norm": 6.33167028427124,
      "learning_rate": 9.317340445645377e-05,
      "loss": 2.6974552154541014,
      "memory(GiB)": 66.02,
      "step": 19640,
      "token_acc": 0.45918367346938777,
      "train_speed(iter/s)": 1.447068
    },
    {
      "epoch": 0.841652028619168,
      "grad_norm": 4.291542053222656,
      "learning_rate": 9.317000954453364e-05,
      "loss": 2.6369171142578125,
      "memory(GiB)": 66.02,
      "step": 19645,
      "token_acc": 0.4492307692307692,
      "train_speed(iter/s)": 1.446968
    },
    {
      "epoch": 0.8418662439484169,
      "grad_norm": 4.765768051147461,
      "learning_rate": 9.316661385054825e-05,
      "loss": 2.4904666900634767,
      "memory(GiB)": 66.02,
      "step": 19650,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.446999
    },
    {
      "epoch": 0.842080459277666,
      "grad_norm": 4.3831467628479,
      "learning_rate": 9.316321737455919e-05,
      "loss": 2.6797822952270507,
      "memory(GiB)": 66.02,
      "step": 19655,
      "token_acc": 0.4471299093655589,
      "train_speed(iter/s)": 1.447075
    },
    {
      "epoch": 0.8422946746069149,
      "grad_norm": 3.520472288131714,
      "learning_rate": 9.315982011662794e-05,
      "loss": 2.318328285217285,
      "memory(GiB)": 66.02,
      "step": 19660,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.447112
    },
    {
      "epoch": 0.8425088899361638,
      "grad_norm": 3.961155414581299,
      "learning_rate": 9.315642207681607e-05,
      "loss": 2.645852470397949,
      "memory(GiB)": 66.02,
      "step": 19665,
      "token_acc": 0.45857988165680474,
      "train_speed(iter/s)": 1.447191
    },
    {
      "epoch": 0.8427231052654128,
      "grad_norm": 5.533195495605469,
      "learning_rate": 9.315302325518514e-05,
      "loss": 2.5814552307128906,
      "memory(GiB)": 66.02,
      "step": 19670,
      "token_acc": 0.4641509433962264,
      "train_speed(iter/s)": 1.447185
    },
    {
      "epoch": 0.8429373205946618,
      "grad_norm": 3.973841667175293,
      "learning_rate": 9.31496236517967e-05,
      "loss": 2.647699546813965,
      "memory(GiB)": 66.02,
      "step": 19675,
      "token_acc": 0.45646437994722955,
      "train_speed(iter/s)": 1.447228
    },
    {
      "epoch": 0.8431515359239107,
      "grad_norm": 6.171427249908447,
      "learning_rate": 9.314622326671233e-05,
      "loss": 2.9970001220703124,
      "memory(GiB)": 66.02,
      "step": 19680,
      "token_acc": 0.39759036144578314,
      "train_speed(iter/s)": 1.447201
    },
    {
      "epoch": 0.8433657512531597,
      "grad_norm": 6.479748249053955,
      "learning_rate": 9.314282209999368e-05,
      "loss": 2.4964162826538088,
      "memory(GiB)": 66.02,
      "step": 19685,
      "token_acc": 0.5168539325842697,
      "train_speed(iter/s)": 1.447259
    },
    {
      "epoch": 0.8435799665824086,
      "grad_norm": 4.9075236320495605,
      "learning_rate": 9.313942015170233e-05,
      "loss": 2.711712646484375,
      "memory(GiB)": 66.02,
      "step": 19690,
      "token_acc": 0.45493562231759654,
      "train_speed(iter/s)": 1.447262
    },
    {
      "epoch": 0.8437941819116576,
      "grad_norm": 4.661203861236572,
      "learning_rate": 9.31360174218999e-05,
      "loss": 2.6397701263427735,
      "memory(GiB)": 66.02,
      "step": 19695,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.447214
    },
    {
      "epoch": 0.8440083972409066,
      "grad_norm": 4.87868595123291,
      "learning_rate": 9.313261391064807e-05,
      "loss": 2.3757896423339844,
      "memory(GiB)": 66.02,
      "step": 19700,
      "token_acc": 0.4859437751004016,
      "train_speed(iter/s)": 1.447264
    },
    {
      "epoch": 0.8442226125701555,
      "grad_norm": 3.4998130798339844,
      "learning_rate": 9.312920961800847e-05,
      "loss": 2.357621765136719,
      "memory(GiB)": 66.02,
      "step": 19705,
      "token_acc": 0.5083612040133779,
      "train_speed(iter/s)": 1.447322
    },
    {
      "epoch": 0.8444368278994044,
      "grad_norm": 4.68348503112793,
      "learning_rate": 9.312580454404278e-05,
      "loss": 2.655163383483887,
      "memory(GiB)": 66.02,
      "step": 19710,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.447273
    },
    {
      "epoch": 0.8446510432286535,
      "grad_norm": 6.707536697387695,
      "learning_rate": 9.312239868881268e-05,
      "loss": 2.594804382324219,
      "memory(GiB)": 66.02,
      "step": 19715,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.447261
    },
    {
      "epoch": 0.8448652585579024,
      "grad_norm": 5.359857082366943,
      "learning_rate": 9.311899205237989e-05,
      "loss": 2.7529628753662108,
      "memory(GiB)": 66.02,
      "step": 19720,
      "token_acc": 0.42748091603053434,
      "train_speed(iter/s)": 1.447237
    },
    {
      "epoch": 0.8450794738871513,
      "grad_norm": 4.478527069091797,
      "learning_rate": 9.31155846348061e-05,
      "loss": 2.4507400512695314,
      "memory(GiB)": 66.02,
      "step": 19725,
      "token_acc": 0.4696485623003195,
      "train_speed(iter/s)": 1.44725
    },
    {
      "epoch": 0.8452936892164004,
      "grad_norm": 4.161308288574219,
      "learning_rate": 9.311217643615304e-05,
      "loss": 2.9077001571655274,
      "memory(GiB)": 66.02,
      "step": 19730,
      "token_acc": 0.46006389776357826,
      "train_speed(iter/s)": 1.447214
    },
    {
      "epoch": 0.8455079045456493,
      "grad_norm": 3.512089490890503,
      "learning_rate": 9.310876745648247e-05,
      "loss": 2.738383483886719,
      "memory(GiB)": 66.02,
      "step": 19735,
      "token_acc": 0.41847826086956524,
      "train_speed(iter/s)": 1.447308
    },
    {
      "epoch": 0.8457221198748982,
      "grad_norm": 6.097390651702881,
      "learning_rate": 9.310535769585615e-05,
      "loss": 2.6992351531982424,
      "memory(GiB)": 66.02,
      "step": 19740,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.447323
    },
    {
      "epoch": 0.8459363352041472,
      "grad_norm": 6.157765865325928,
      "learning_rate": 9.310194715433583e-05,
      "loss": 2.6572113037109375,
      "memory(GiB)": 66.02,
      "step": 19745,
      "token_acc": 0.4525316455696203,
      "train_speed(iter/s)": 1.447411
    },
    {
      "epoch": 0.8461505505333962,
      "grad_norm": 5.736446380615234,
      "learning_rate": 9.309853583198328e-05,
      "loss": 2.7184785842895507,
      "memory(GiB)": 66.02,
      "step": 19750,
      "token_acc": 0.43283582089552236,
      "train_speed(iter/s)": 1.447482
    },
    {
      "epoch": 0.8463647658626451,
      "grad_norm": 6.251994609832764,
      "learning_rate": 9.309512372886036e-05,
      "loss": 2.4597482681274414,
      "memory(GiB)": 66.02,
      "step": 19755,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.447497
    },
    {
      "epoch": 0.8465789811918941,
      "grad_norm": 5.289851665496826,
      "learning_rate": 9.309171084502883e-05,
      "loss": 2.600275421142578,
      "memory(GiB)": 66.02,
      "step": 19760,
      "token_acc": 0.44516129032258067,
      "train_speed(iter/s)": 1.447482
    },
    {
      "epoch": 0.846793196521143,
      "grad_norm": 4.868996620178223,
      "learning_rate": 9.308829718055054e-05,
      "loss": 3.076642608642578,
      "memory(GiB)": 66.02,
      "step": 19765,
      "token_acc": 0.3829787234042553,
      "train_speed(iter/s)": 1.447487
    },
    {
      "epoch": 0.847007411850392,
      "grad_norm": 5.199337482452393,
      "learning_rate": 9.308488273548732e-05,
      "loss": 2.368132781982422,
      "memory(GiB)": 66.02,
      "step": 19770,
      "token_acc": 0.5127272727272727,
      "train_speed(iter/s)": 1.447488
    },
    {
      "epoch": 0.847221627179641,
      "grad_norm": 5.29536247253418,
      "learning_rate": 9.308146750990106e-05,
      "loss": 2.670939636230469,
      "memory(GiB)": 66.02,
      "step": 19775,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.44754
    },
    {
      "epoch": 0.8474358425088899,
      "grad_norm": 3.8321127891540527,
      "learning_rate": 9.307805150385356e-05,
      "loss": 2.5518962860107424,
      "memory(GiB)": 66.02,
      "step": 19780,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.447599
    },
    {
      "epoch": 0.8476500578381388,
      "grad_norm": 5.568121433258057,
      "learning_rate": 9.307463471740678e-05,
      "loss": 2.5398372650146483,
      "memory(GiB)": 66.02,
      "step": 19785,
      "token_acc": 0.45564516129032256,
      "train_speed(iter/s)": 1.447645
    },
    {
      "epoch": 0.8478642731673879,
      "grad_norm": 3.0147359371185303,
      "learning_rate": 9.307121715062257e-05,
      "loss": 2.749627876281738,
      "memory(GiB)": 66.02,
      "step": 19790,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.447588
    },
    {
      "epoch": 0.8480784884966368,
      "grad_norm": 5.127834796905518,
      "learning_rate": 9.306779880356284e-05,
      "loss": 2.801944351196289,
      "memory(GiB)": 66.02,
      "step": 19795,
      "token_acc": 0.448,
      "train_speed(iter/s)": 1.447674
    },
    {
      "epoch": 0.8482927038258857,
      "grad_norm": 4.423109531402588,
      "learning_rate": 9.306437967628956e-05,
      "loss": 2.5503236770629885,
      "memory(GiB)": 66.02,
      "step": 19800,
      "token_acc": 0.47950819672131145,
      "train_speed(iter/s)": 1.447673
    },
    {
      "epoch": 0.8485069191551348,
      "grad_norm": 3.6239840984344482,
      "learning_rate": 9.306095976886464e-05,
      "loss": 2.49837646484375,
      "memory(GiB)": 66.02,
      "step": 19805,
      "token_acc": 0.4789156626506024,
      "train_speed(iter/s)": 1.447689
    },
    {
      "epoch": 0.8487211344843837,
      "grad_norm": 4.444400310516357,
      "learning_rate": 9.305753908135003e-05,
      "loss": 2.423139190673828,
      "memory(GiB)": 66.02,
      "step": 19810,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.447665
    },
    {
      "epoch": 0.8489353498136326,
      "grad_norm": 4.946013450622559,
      "learning_rate": 9.305411761380771e-05,
      "loss": 2.5663604736328125,
      "memory(GiB)": 66.02,
      "step": 19815,
      "token_acc": 0.48659003831417624,
      "train_speed(iter/s)": 1.447677
    },
    {
      "epoch": 0.8491495651428816,
      "grad_norm": 3.828543186187744,
      "learning_rate": 9.305069536629967e-05,
      "loss": 2.6054725646972656,
      "memory(GiB)": 66.02,
      "step": 19820,
      "token_acc": 0.4405144694533762,
      "train_speed(iter/s)": 1.447681
    },
    {
      "epoch": 0.8493637804721306,
      "grad_norm": 4.285463809967041,
      "learning_rate": 9.30472723388879e-05,
      "loss": 2.7441246032714846,
      "memory(GiB)": 66.02,
      "step": 19825,
      "token_acc": 0.4254658385093168,
      "train_speed(iter/s)": 1.447649
    },
    {
      "epoch": 0.8495779958013795,
      "grad_norm": 4.809802532196045,
      "learning_rate": 9.30438485316344e-05,
      "loss": 2.611124801635742,
      "memory(GiB)": 66.02,
      "step": 19830,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.447651
    },
    {
      "epoch": 0.8497922111306285,
      "grad_norm": 3.7109873294830322,
      "learning_rate": 9.30404239446012e-05,
      "loss": 2.4324920654296873,
      "memory(GiB)": 66.02,
      "step": 19835,
      "token_acc": 0.5134328358208955,
      "train_speed(iter/s)": 1.447713
    },
    {
      "epoch": 0.8500064264598775,
      "grad_norm": 4.709730625152588,
      "learning_rate": 9.303699857785035e-05,
      "loss": 2.7048797607421875,
      "memory(GiB)": 66.02,
      "step": 19840,
      "token_acc": 0.46437994722955145,
      "train_speed(iter/s)": 1.447795
    },
    {
      "epoch": 0.8502206417891264,
      "grad_norm": 3.773698329925537,
      "learning_rate": 9.303357243144392e-05,
      "loss": 2.5571128845214846,
      "memory(GiB)": 66.02,
      "step": 19845,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.447745
    },
    {
      "epoch": 0.8504348571183754,
      "grad_norm": 4.688603401184082,
      "learning_rate": 9.303014550544394e-05,
      "loss": 2.304994583129883,
      "memory(GiB)": 66.02,
      "step": 19850,
      "token_acc": 0.5211267605633803,
      "train_speed(iter/s)": 1.447684
    },
    {
      "epoch": 0.8506490724476243,
      "grad_norm": 6.75828218460083,
      "learning_rate": 9.30267177999125e-05,
      "loss": 2.6469951629638673,
      "memory(GiB)": 66.02,
      "step": 19855,
      "token_acc": 0.44805194805194803,
      "train_speed(iter/s)": 1.447527
    },
    {
      "epoch": 0.8508632877768734,
      "grad_norm": 5.713206768035889,
      "learning_rate": 9.302328931491172e-05,
      "loss": 2.735860252380371,
      "memory(GiB)": 66.02,
      "step": 19860,
      "token_acc": 0.43846153846153846,
      "train_speed(iter/s)": 1.447528
    },
    {
      "epoch": 0.8510775031061223,
      "grad_norm": 4.005903244018555,
      "learning_rate": 9.301986005050369e-05,
      "loss": 2.430007743835449,
      "memory(GiB)": 66.02,
      "step": 19865,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.447545
    },
    {
      "epoch": 0.8512917184353712,
      "grad_norm": 4.3053741455078125,
      "learning_rate": 9.301643000675054e-05,
      "loss": 2.3170469284057615,
      "memory(GiB)": 66.02,
      "step": 19870,
      "token_acc": 0.48328267477203646,
      "train_speed(iter/s)": 1.447449
    },
    {
      "epoch": 0.8515059337646202,
      "grad_norm": 4.132357120513916,
      "learning_rate": 9.30129991837144e-05,
      "loss": 2.483839416503906,
      "memory(GiB)": 66.02,
      "step": 19875,
      "token_acc": 0.4732510288065844,
      "train_speed(iter/s)": 1.44744
    },
    {
      "epoch": 0.8517201490938692,
      "grad_norm": 4.457902908325195,
      "learning_rate": 9.300956758145742e-05,
      "loss": 2.625859260559082,
      "memory(GiB)": 66.02,
      "step": 19880,
      "token_acc": 0.4574898785425101,
      "train_speed(iter/s)": 1.447348
    },
    {
      "epoch": 0.8519343644231181,
      "grad_norm": 5.308292865753174,
      "learning_rate": 9.30061352000418e-05,
      "loss": 2.663136291503906,
      "memory(GiB)": 66.02,
      "step": 19885,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.447301
    },
    {
      "epoch": 0.8521485797523671,
      "grad_norm": 4.78318977355957,
      "learning_rate": 9.30027020395297e-05,
      "loss": 2.8430709838867188,
      "memory(GiB)": 66.02,
      "step": 19890,
      "token_acc": 0.44363636363636366,
      "train_speed(iter/s)": 1.447231
    },
    {
      "epoch": 0.852362795081616,
      "grad_norm": 4.209441661834717,
      "learning_rate": 9.299926809998329e-05,
      "loss": 2.816679763793945,
      "memory(GiB)": 66.02,
      "step": 19895,
      "token_acc": 0.43018867924528303,
      "train_speed(iter/s)": 1.447296
    },
    {
      "epoch": 0.852577010410865,
      "grad_norm": 7.264736652374268,
      "learning_rate": 9.299583338146483e-05,
      "loss": 2.572256851196289,
      "memory(GiB)": 66.02,
      "step": 19900,
      "token_acc": 0.43174603174603177,
      "train_speed(iter/s)": 1.447234
    },
    {
      "epoch": 0.852791225740114,
      "grad_norm": 4.424783229827881,
      "learning_rate": 9.299239788403649e-05,
      "loss": 2.16674690246582,
      "memory(GiB)": 66.02,
      "step": 19905,
      "token_acc": 0.54296875,
      "train_speed(iter/s)": 1.447311
    },
    {
      "epoch": 0.8530054410693629,
      "grad_norm": 4.210756778717041,
      "learning_rate": 9.298896160776054e-05,
      "loss": 2.7489221572875975,
      "memory(GiB)": 66.02,
      "step": 19910,
      "token_acc": 0.40752351097178685,
      "train_speed(iter/s)": 1.44725
    },
    {
      "epoch": 0.8532196563986119,
      "grad_norm": 5.202207088470459,
      "learning_rate": 9.298552455269923e-05,
      "loss": 2.7239757537841798,
      "memory(GiB)": 66.02,
      "step": 19915,
      "token_acc": 0.40502793296089384,
      "train_speed(iter/s)": 1.447245
    },
    {
      "epoch": 0.8534338717278609,
      "grad_norm": 3.9920990467071533,
      "learning_rate": 9.298208671891482e-05,
      "loss": 2.497640609741211,
      "memory(GiB)": 66.02,
      "step": 19920,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.447258
    },
    {
      "epoch": 0.8536480870571098,
      "grad_norm": 3.311734914779663,
      "learning_rate": 9.297864810646958e-05,
      "loss": 2.72473087310791,
      "memory(GiB)": 66.02,
      "step": 19925,
      "token_acc": 0.42138364779874216,
      "train_speed(iter/s)": 1.447371
    },
    {
      "epoch": 0.8538623023863587,
      "grad_norm": 4.473320960998535,
      "learning_rate": 9.297520871542583e-05,
      "loss": 2.7096553802490235,
      "memory(GiB)": 66.02,
      "step": 19930,
      "token_acc": 0.45544554455445546,
      "train_speed(iter/s)": 1.447454
    },
    {
      "epoch": 0.8540765177156078,
      "grad_norm": 4.602941513061523,
      "learning_rate": 9.297176854584582e-05,
      "loss": 2.5323307037353517,
      "memory(GiB)": 66.02,
      "step": 19935,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.44742
    },
    {
      "epoch": 0.8542907330448567,
      "grad_norm": 4.637984752655029,
      "learning_rate": 9.296832759779195e-05,
      "loss": 2.4924673080444335,
      "memory(GiB)": 66.02,
      "step": 19940,
      "token_acc": 0.4245614035087719,
      "train_speed(iter/s)": 1.447434
    },
    {
      "epoch": 0.8545049483741056,
      "grad_norm": 3.401311159133911,
      "learning_rate": 9.29648858713265e-05,
      "loss": 2.800104522705078,
      "memory(GiB)": 66.02,
      "step": 19945,
      "token_acc": 0.4602076124567474,
      "train_speed(iter/s)": 1.447359
    },
    {
      "epoch": 0.8547191637033547,
      "grad_norm": 5.960352420806885,
      "learning_rate": 9.296144336651185e-05,
      "loss": 2.698061943054199,
      "memory(GiB)": 66.02,
      "step": 19950,
      "token_acc": 0.44573643410852715,
      "train_speed(iter/s)": 1.447358
    },
    {
      "epoch": 0.8549333790326036,
      "grad_norm": 4.572206497192383,
      "learning_rate": 9.295800008341033e-05,
      "loss": 2.777944564819336,
      "memory(GiB)": 66.02,
      "step": 19955,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.447419
    },
    {
      "epoch": 0.8551475943618525,
      "grad_norm": 4.109676837921143,
      "learning_rate": 9.295455602208438e-05,
      "loss": 2.779833221435547,
      "memory(GiB)": 66.02,
      "step": 19960,
      "token_acc": 0.4680232558139535,
      "train_speed(iter/s)": 1.447392
    },
    {
      "epoch": 0.8553618096911015,
      "grad_norm": 4.0725860595703125,
      "learning_rate": 9.295111118259632e-05,
      "loss": 2.788290786743164,
      "memory(GiB)": 66.02,
      "step": 19965,
      "token_acc": 0.445578231292517,
      "train_speed(iter/s)": 1.447476
    },
    {
      "epoch": 0.8555760250203505,
      "grad_norm": 4.575146198272705,
      "learning_rate": 9.29476655650086e-05,
      "loss": 3.0003833770751953,
      "memory(GiB)": 66.02,
      "step": 19970,
      "token_acc": 0.4171779141104294,
      "train_speed(iter/s)": 1.447481
    },
    {
      "epoch": 0.8557902403495994,
      "grad_norm": 3.6619763374328613,
      "learning_rate": 9.294421916938363e-05,
      "loss": 2.3414730072021483,
      "memory(GiB)": 66.02,
      "step": 19975,
      "token_acc": 0.49508196721311476,
      "train_speed(iter/s)": 1.447553
    },
    {
      "epoch": 0.8560044556788484,
      "grad_norm": 3.6070966720581055,
      "learning_rate": 9.294077199578384e-05,
      "loss": 2.7277811050415037,
      "memory(GiB)": 66.02,
      "step": 19980,
      "token_acc": 0.45934959349593496,
      "train_speed(iter/s)": 1.447548
    },
    {
      "epoch": 0.8562186710080973,
      "grad_norm": 4.308478355407715,
      "learning_rate": 9.293732404427169e-05,
      "loss": 2.609332275390625,
      "memory(GiB)": 66.02,
      "step": 19985,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.447563
    },
    {
      "epoch": 0.8564328863373463,
      "grad_norm": 3.5091500282287598,
      "learning_rate": 9.293387531490964e-05,
      "loss": 2.626960563659668,
      "memory(GiB)": 66.02,
      "step": 19990,
      "token_acc": 0.4633431085043988,
      "train_speed(iter/s)": 1.447647
    },
    {
      "epoch": 0.8566471016665953,
      "grad_norm": 4.095282077789307,
      "learning_rate": 9.293042580776015e-05,
      "loss": 2.5017480850219727,
      "memory(GiB)": 66.02,
      "step": 19995,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.447702
    },
    {
      "epoch": 0.8568613169958442,
      "grad_norm": 7.032236099243164,
      "learning_rate": 9.292697552288574e-05,
      "loss": 2.6451932907104494,
      "memory(GiB)": 66.02,
      "step": 20000,
      "token_acc": 0.47019867549668876,
      "train_speed(iter/s)": 1.447715
    },
    {
      "epoch": 0.8568613169958442,
      "eval_loss": 2.3184609413146973,
      "eval_runtime": 14.0948,
      "eval_samples_per_second": 7.095,
      "eval_steps_per_second": 7.095,
      "eval_token_acc": 0.45674300254452926,
      "step": 20000
    },
    {
      "epoch": 0.8570755323250931,
      "grad_norm": 5.316519737243652,
      "learning_rate": 9.292352446034889e-05,
      "loss": 2.8579620361328124,
      "memory(GiB)": 66.02,
      "step": 20005,
      "token_acc": 0.44265232974910396,
      "train_speed(iter/s)": 1.446228
    },
    {
      "epoch": 0.8572897476543422,
      "grad_norm": 3.648613452911377,
      "learning_rate": 9.292007262021213e-05,
      "loss": 2.344171905517578,
      "memory(GiB)": 66.02,
      "step": 20010,
      "token_acc": 0.5119047619047619,
      "train_speed(iter/s)": 1.446273
    },
    {
      "epoch": 0.8575039629835911,
      "grad_norm": 4.44399881362915,
      "learning_rate": 9.291662000253799e-05,
      "loss": 2.589639663696289,
      "memory(GiB)": 66.02,
      "step": 20015,
      "token_acc": 0.44814814814814813,
      "train_speed(iter/s)": 1.446266
    },
    {
      "epoch": 0.85771817831284,
      "grad_norm": 4.210684299468994,
      "learning_rate": 9.291316660738903e-05,
      "loss": 2.7232810974121096,
      "memory(GiB)": 66.02,
      "step": 20020,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.446296
    },
    {
      "epoch": 0.8579323936420891,
      "grad_norm": 3.345099925994873,
      "learning_rate": 9.290971243482778e-05,
      "loss": 3.020463562011719,
      "memory(GiB)": 66.02,
      "step": 20025,
      "token_acc": 0.3972972972972973,
      "train_speed(iter/s)": 1.446287
    },
    {
      "epoch": 0.858146608971338,
      "grad_norm": 5.379486083984375,
      "learning_rate": 9.290625748491686e-05,
      "loss": 2.8090381622314453,
      "memory(GiB)": 66.02,
      "step": 20030,
      "token_acc": 0.41114982578397213,
      "train_speed(iter/s)": 1.446346
    },
    {
      "epoch": 0.8583608243005869,
      "grad_norm": 5.432685852050781,
      "learning_rate": 9.290280175771882e-05,
      "loss": 2.428244781494141,
      "memory(GiB)": 66.02,
      "step": 20035,
      "token_acc": 0.47035573122529645,
      "train_speed(iter/s)": 1.44632
    },
    {
      "epoch": 0.858575039629836,
      "grad_norm": 3.5571537017822266,
      "learning_rate": 9.289934525329628e-05,
      "loss": 2.463956832885742,
      "memory(GiB)": 66.02,
      "step": 20040,
      "token_acc": 0.5032258064516129,
      "train_speed(iter/s)": 1.44638
    },
    {
      "epoch": 0.8587892549590849,
      "grad_norm": 4.1907148361206055,
      "learning_rate": 9.289588797171186e-05,
      "loss": 2.8935022354125977,
      "memory(GiB)": 66.02,
      "step": 20045,
      "token_acc": 0.4551282051282051,
      "train_speed(iter/s)": 1.446488
    },
    {
      "epoch": 0.8590034702883338,
      "grad_norm": 6.904212474822998,
      "learning_rate": 9.289242991302821e-05,
      "loss": 2.7568593978881837,
      "memory(GiB)": 66.02,
      "step": 20050,
      "token_acc": 0.4647058823529412,
      "train_speed(iter/s)": 1.446448
    },
    {
      "epoch": 0.8592176856175828,
      "grad_norm": 5.2837700843811035,
      "learning_rate": 9.288897107730794e-05,
      "loss": 2.474040985107422,
      "memory(GiB)": 66.02,
      "step": 20055,
      "token_acc": 0.47843137254901963,
      "train_speed(iter/s)": 1.446472
    },
    {
      "epoch": 0.8594319009468318,
      "grad_norm": 4.97775411605835,
      "learning_rate": 9.288551146461373e-05,
      "loss": 2.7760700225830077,
      "memory(GiB)": 66.02,
      "step": 20060,
      "token_acc": 0.4383116883116883,
      "train_speed(iter/s)": 1.446522
    },
    {
      "epoch": 0.8596461162760807,
      "grad_norm": 4.208581447601318,
      "learning_rate": 9.288205107500825e-05,
      "loss": 2.504689025878906,
      "memory(GiB)": 66.02,
      "step": 20065,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.446542
    },
    {
      "epoch": 0.8598603316053297,
      "grad_norm": 4.59130859375,
      "learning_rate": 9.287858990855418e-05,
      "loss": 2.6606418609619142,
      "memory(GiB)": 66.02,
      "step": 20070,
      "token_acc": 0.42704626334519574,
      "train_speed(iter/s)": 1.446585
    },
    {
      "epoch": 0.8600745469345786,
      "grad_norm": 4.684729099273682,
      "learning_rate": 9.287512796531423e-05,
      "loss": 2.469797134399414,
      "memory(GiB)": 66.02,
      "step": 20075,
      "token_acc": 0.4623287671232877,
      "train_speed(iter/s)": 1.446597
    },
    {
      "epoch": 0.8602887622638276,
      "grad_norm": 4.1291584968566895,
      "learning_rate": 9.287166524535111e-05,
      "loss": 2.669881820678711,
      "memory(GiB)": 66.02,
      "step": 20080,
      "token_acc": 0.42105263157894735,
      "train_speed(iter/s)": 1.446541
    },
    {
      "epoch": 0.8605029775930766,
      "grad_norm": 3.9851222038269043,
      "learning_rate": 9.286820174872758e-05,
      "loss": 2.527341842651367,
      "memory(GiB)": 66.02,
      "step": 20085,
      "token_acc": 0.4625,
      "train_speed(iter/s)": 1.446552
    },
    {
      "epoch": 0.8607171929223255,
      "grad_norm": 4.214713096618652,
      "learning_rate": 9.286473747550635e-05,
      "loss": 2.6485807418823244,
      "memory(GiB)": 66.02,
      "step": 20090,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.446583
    },
    {
      "epoch": 0.8609314082515744,
      "grad_norm": 5.188320636749268,
      "learning_rate": 9.286127242575019e-05,
      "loss": 2.5529800415039063,
      "memory(GiB)": 66.02,
      "step": 20095,
      "token_acc": 0.46788990825688076,
      "train_speed(iter/s)": 1.446659
    },
    {
      "epoch": 0.8611456235808235,
      "grad_norm": 4.187458515167236,
      "learning_rate": 9.285780659952188e-05,
      "loss": 2.537496566772461,
      "memory(GiB)": 66.02,
      "step": 20100,
      "token_acc": 0.4752851711026616,
      "train_speed(iter/s)": 1.446528
    },
    {
      "epoch": 0.8613598389100724,
      "grad_norm": 5.257472038269043,
      "learning_rate": 9.285433999688419e-05,
      "loss": 2.6902347564697267,
      "memory(GiB)": 66.02,
      "step": 20105,
      "token_acc": 0.488,
      "train_speed(iter/s)": 1.446573
    },
    {
      "epoch": 0.8615740542393213,
      "grad_norm": 4.594825744628906,
      "learning_rate": 9.285087261789993e-05,
      "loss": 2.7862216949462892,
      "memory(GiB)": 66.02,
      "step": 20110,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.446581
    },
    {
      "epoch": 0.8617882695685704,
      "grad_norm": 5.086245059967041,
      "learning_rate": 9.284740446263191e-05,
      "loss": 2.489419937133789,
      "memory(GiB)": 66.02,
      "step": 20115,
      "token_acc": 0.48338368580060426,
      "train_speed(iter/s)": 1.446644
    },
    {
      "epoch": 0.8620024848978193,
      "grad_norm": 4.404093265533447,
      "learning_rate": 9.284393553114298e-05,
      "loss": 2.7474906921386717,
      "memory(GiB)": 66.02,
      "step": 20120,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.446516
    },
    {
      "epoch": 0.8622167002270682,
      "grad_norm": 3.989957094192505,
      "learning_rate": 9.284046582349596e-05,
      "loss": 2.380805778503418,
      "memory(GiB)": 66.02,
      "step": 20125,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.44659
    },
    {
      "epoch": 0.8624309155563172,
      "grad_norm": 7.043115615844727,
      "learning_rate": 9.283699533975372e-05,
      "loss": 2.6302469253540037,
      "memory(GiB)": 66.02,
      "step": 20130,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.446553
    },
    {
      "epoch": 0.8626451308855662,
      "grad_norm": 3.9640233516693115,
      "learning_rate": 9.283352407997912e-05,
      "loss": 2.661854934692383,
      "memory(GiB)": 66.02,
      "step": 20135,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.446512
    },
    {
      "epoch": 0.8628593462148151,
      "grad_norm": 4.779181957244873,
      "learning_rate": 9.283005204423504e-05,
      "loss": 2.5626220703125,
      "memory(GiB)": 66.02,
      "step": 20140,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.446531
    },
    {
      "epoch": 0.8630735615440641,
      "grad_norm": 5.667942523956299,
      "learning_rate": 9.28265792325844e-05,
      "loss": 2.440044403076172,
      "memory(GiB)": 66.02,
      "step": 20145,
      "token_acc": 0.521311475409836,
      "train_speed(iter/s)": 1.446511
    },
    {
      "epoch": 0.863287776873313,
      "grad_norm": 3.9496004581451416,
      "learning_rate": 9.282310564509009e-05,
      "loss": 2.5329425811767576,
      "memory(GiB)": 66.02,
      "step": 20150,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.446515
    },
    {
      "epoch": 0.863501992202562,
      "grad_norm": 4.588701248168945,
      "learning_rate": 9.281963128181508e-05,
      "loss": 2.47724609375,
      "memory(GiB)": 66.02,
      "step": 20155,
      "token_acc": 0.4728682170542636,
      "train_speed(iter/s)": 1.446567
    },
    {
      "epoch": 0.863716207531811,
      "grad_norm": 4.740305423736572,
      "learning_rate": 9.281615614282225e-05,
      "loss": 3.1929759979248047,
      "memory(GiB)": 66.02,
      "step": 20160,
      "token_acc": 0.4197183098591549,
      "train_speed(iter/s)": 1.44654
    },
    {
      "epoch": 0.8639304228610599,
      "grad_norm": 4.869343280792236,
      "learning_rate": 9.28126802281746e-05,
      "loss": 2.6420190811157225,
      "memory(GiB)": 66.02,
      "step": 20165,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.446501
    },
    {
      "epoch": 0.8641446381903088,
      "grad_norm": 5.131339073181152,
      "learning_rate": 9.280920353793508e-05,
      "loss": 2.423465919494629,
      "memory(GiB)": 66.02,
      "step": 20170,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.446434
    },
    {
      "epoch": 0.8643588535195579,
      "grad_norm": 3.910404682159424,
      "learning_rate": 9.280572607216668e-05,
      "loss": 2.235536575317383,
      "memory(GiB)": 66.02,
      "step": 20175,
      "token_acc": 0.5066225165562914,
      "train_speed(iter/s)": 1.446502
    },
    {
      "epoch": 0.8645730688488068,
      "grad_norm": 4.464833736419678,
      "learning_rate": 9.280224783093241e-05,
      "loss": 2.7030141830444334,
      "memory(GiB)": 66.02,
      "step": 20180,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.446532
    },
    {
      "epoch": 0.8647872841780557,
      "grad_norm": 4.7071852684021,
      "learning_rate": 9.279876881429526e-05,
      "loss": 2.5194549560546875,
      "memory(GiB)": 66.02,
      "step": 20185,
      "token_acc": 0.47435897435897434,
      "train_speed(iter/s)": 1.446573
    },
    {
      "epoch": 0.8650014995073048,
      "grad_norm": 4.958098888397217,
      "learning_rate": 9.279528902231827e-05,
      "loss": 2.520359420776367,
      "memory(GiB)": 66.02,
      "step": 20190,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.446588
    },
    {
      "epoch": 0.8652157148365537,
      "grad_norm": 6.080878734588623,
      "learning_rate": 9.279180845506446e-05,
      "loss": 2.768878173828125,
      "memory(GiB)": 66.02,
      "step": 20195,
      "token_acc": 0.4233128834355828,
      "train_speed(iter/s)": 1.446666
    },
    {
      "epoch": 0.8654299301658027,
      "grad_norm": 3.969670295715332,
      "learning_rate": 9.27883271125969e-05,
      "loss": 2.7485870361328124,
      "memory(GiB)": 66.02,
      "step": 20200,
      "token_acc": 0.44982698961937717,
      "train_speed(iter/s)": 1.446722
    },
    {
      "epoch": 0.8656441454950516,
      "grad_norm": 5.300868511199951,
      "learning_rate": 9.278484499497868e-05,
      "loss": 2.477072334289551,
      "memory(GiB)": 66.02,
      "step": 20205,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.446586
    },
    {
      "epoch": 0.8658583608243006,
      "grad_norm": 4.885673522949219,
      "learning_rate": 9.278136210227284e-05,
      "loss": 2.606531524658203,
      "memory(GiB)": 66.02,
      "step": 20210,
      "token_acc": 0.46381578947368424,
      "train_speed(iter/s)": 1.446598
    },
    {
      "epoch": 0.8660725761535496,
      "grad_norm": 4.389687538146973,
      "learning_rate": 9.277787843454248e-05,
      "loss": 2.5676076889038084,
      "memory(GiB)": 66.02,
      "step": 20215,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.446609
    },
    {
      "epoch": 0.8662867914827985,
      "grad_norm": 7.971065998077393,
      "learning_rate": 9.277439399185074e-05,
      "loss": 2.6598133087158202,
      "memory(GiB)": 66.02,
      "step": 20220,
      "token_acc": 0.46825396825396826,
      "train_speed(iter/s)": 1.446616
    },
    {
      "epoch": 0.8665010068120474,
      "grad_norm": 4.387348651885986,
      "learning_rate": 9.277090877426074e-05,
      "loss": 2.55792350769043,
      "memory(GiB)": 66.02,
      "step": 20225,
      "token_acc": 0.4432624113475177,
      "train_speed(iter/s)": 1.44661
    },
    {
      "epoch": 0.8667152221412965,
      "grad_norm": 4.705660820007324,
      "learning_rate": 9.276742278183558e-05,
      "loss": 2.6393453598022463,
      "memory(GiB)": 66.02,
      "step": 20230,
      "token_acc": 0.48427672955974843,
      "train_speed(iter/s)": 1.446603
    },
    {
      "epoch": 0.8669294374705454,
      "grad_norm": 3.941145420074463,
      "learning_rate": 9.276393601463844e-05,
      "loss": 2.630137825012207,
      "memory(GiB)": 66.02,
      "step": 20235,
      "token_acc": 0.5077881619937694,
      "train_speed(iter/s)": 1.446611
    },
    {
      "epoch": 0.8671436527997943,
      "grad_norm": 5.056817054748535,
      "learning_rate": 9.276044847273249e-05,
      "loss": 2.3690935134887696,
      "memory(GiB)": 66.02,
      "step": 20240,
      "token_acc": 0.4900398406374502,
      "train_speed(iter/s)": 1.446654
    },
    {
      "epoch": 0.8673578681290434,
      "grad_norm": 4.330813407897949,
      "learning_rate": 9.27569601561809e-05,
      "loss": 2.8303770065307616,
      "memory(GiB)": 66.02,
      "step": 20245,
      "token_acc": 0.46037735849056605,
      "train_speed(iter/s)": 1.446683
    },
    {
      "epoch": 0.8675720834582923,
      "grad_norm": 6.021437644958496,
      "learning_rate": 9.275347106504689e-05,
      "loss": 2.4435911178588867,
      "memory(GiB)": 66.02,
      "step": 20250,
      "token_acc": 0.47183098591549294,
      "train_speed(iter/s)": 1.446776
    },
    {
      "epoch": 0.8677862987875412,
      "grad_norm": 4.0443572998046875,
      "learning_rate": 9.274998119939362e-05,
      "loss": 2.717410659790039,
      "memory(GiB)": 66.02,
      "step": 20255,
      "token_acc": 0.48703170028818443,
      "train_speed(iter/s)": 1.446728
    },
    {
      "epoch": 0.8680005141167902,
      "grad_norm": 4.352334976196289,
      "learning_rate": 9.274649055928434e-05,
      "loss": 2.6219051361083983,
      "memory(GiB)": 66.02,
      "step": 20260,
      "token_acc": 0.4208955223880597,
      "train_speed(iter/s)": 1.446756
    },
    {
      "epoch": 0.8682147294460392,
      "grad_norm": 4.7858781814575195,
      "learning_rate": 9.274299914478228e-05,
      "loss": 2.683307075500488,
      "memory(GiB)": 66.02,
      "step": 20265,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.446761
    },
    {
      "epoch": 0.8684289447752881,
      "grad_norm": 5.566839218139648,
      "learning_rate": 9.273950695595071e-05,
      "loss": 2.4615032196044924,
      "memory(GiB)": 66.02,
      "step": 20270,
      "token_acc": 0.49201277955271566,
      "train_speed(iter/s)": 1.446721
    },
    {
      "epoch": 0.8686431601045371,
      "grad_norm": 6.246694087982178,
      "learning_rate": 9.273601399285287e-05,
      "loss": 2.8437292098999025,
      "memory(GiB)": 66.02,
      "step": 20275,
      "token_acc": 0.43283582089552236,
      "train_speed(iter/s)": 1.44675
    },
    {
      "epoch": 0.868857375433786,
      "grad_norm": 3.2209200859069824,
      "learning_rate": 9.273252025555205e-05,
      "loss": 2.403131294250488,
      "memory(GiB)": 66.02,
      "step": 20280,
      "token_acc": 0.5177304964539007,
      "train_speed(iter/s)": 1.44675
    },
    {
      "epoch": 0.869071590763035,
      "grad_norm": 6.783487319946289,
      "learning_rate": 9.272902574411153e-05,
      "loss": 2.703940200805664,
      "memory(GiB)": 66.02,
      "step": 20285,
      "token_acc": 0.4478114478114478,
      "train_speed(iter/s)": 1.446795
    },
    {
      "epoch": 0.869285806092284,
      "grad_norm": 5.519693374633789,
      "learning_rate": 9.272553045859464e-05,
      "loss": 2.6275714874267577,
      "memory(GiB)": 66.02,
      "step": 20290,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.446831
    },
    {
      "epoch": 0.8695000214215329,
      "grad_norm": 4.119229793548584,
      "learning_rate": 9.272203439906469e-05,
      "loss": 2.6545949935913087,
      "memory(GiB)": 66.02,
      "step": 20295,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.4468
    },
    {
      "epoch": 0.8697142367507819,
      "grad_norm": 5.105564117431641,
      "learning_rate": 9.271853756558497e-05,
      "loss": 2.4343557357788086,
      "memory(GiB)": 66.02,
      "step": 20300,
      "token_acc": 0.4664429530201342,
      "train_speed(iter/s)": 1.446767
    },
    {
      "epoch": 0.8699284520800309,
      "grad_norm": 4.066102981567383,
      "learning_rate": 9.271503995821891e-05,
      "loss": 2.8815902709960937,
      "memory(GiB)": 66.02,
      "step": 20305,
      "token_acc": 0.43812709030100333,
      "train_speed(iter/s)": 1.446776
    },
    {
      "epoch": 0.8701426674092798,
      "grad_norm": 4.553707599639893,
      "learning_rate": 9.271154157702982e-05,
      "loss": 2.451702880859375,
      "memory(GiB)": 66.02,
      "step": 20310,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.446836
    },
    {
      "epoch": 0.8703568827385287,
      "grad_norm": 3.8057193756103516,
      "learning_rate": 9.270804242208109e-05,
      "loss": 2.717175102233887,
      "memory(GiB)": 66.02,
      "step": 20315,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.446861
    },
    {
      "epoch": 0.8705710980677778,
      "grad_norm": 4.666566371917725,
      "learning_rate": 9.27045424934361e-05,
      "loss": 3.0016950607299804,
      "memory(GiB)": 66.02,
      "step": 20320,
      "token_acc": 0.42517006802721086,
      "train_speed(iter/s)": 1.446776
    },
    {
      "epoch": 0.8707853133970267,
      "grad_norm": 4.413992404937744,
      "learning_rate": 9.270104179115825e-05,
      "loss": 2.8052406311035156,
      "memory(GiB)": 66.02,
      "step": 20325,
      "token_acc": 0.40878378378378377,
      "train_speed(iter/s)": 1.446853
    },
    {
      "epoch": 0.8709995287262756,
      "grad_norm": 4.927813529968262,
      "learning_rate": 9.2697540315311e-05,
      "loss": 2.657339096069336,
      "memory(GiB)": 66.02,
      "step": 20330,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.446831
    },
    {
      "epoch": 0.8712137440555247,
      "grad_norm": 4.384943008422852,
      "learning_rate": 9.269403806595775e-05,
      "loss": 2.370720100402832,
      "memory(GiB)": 66.02,
      "step": 20335,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.446817
    },
    {
      "epoch": 0.8714279593847736,
      "grad_norm": 4.819328784942627,
      "learning_rate": 9.269053504316194e-05,
      "loss": 2.832560729980469,
      "memory(GiB)": 66.02,
      "step": 20340,
      "token_acc": 0.43911439114391143,
      "train_speed(iter/s)": 1.446826
    },
    {
      "epoch": 0.8716421747140225,
      "grad_norm": 5.835651874542236,
      "learning_rate": 9.268703124698704e-05,
      "loss": 2.793199920654297,
      "memory(GiB)": 66.02,
      "step": 20345,
      "token_acc": 0.45051194539249145,
      "train_speed(iter/s)": 1.446845
    },
    {
      "epoch": 0.8718563900432715,
      "grad_norm": 6.481338977813721,
      "learning_rate": 9.268352667749653e-05,
      "loss": 2.480810356140137,
      "memory(GiB)": 66.02,
      "step": 20350,
      "token_acc": 0.44947735191637633,
      "train_speed(iter/s)": 1.446899
    },
    {
      "epoch": 0.8720706053725205,
      "grad_norm": 3.819230556488037,
      "learning_rate": 9.268002133475388e-05,
      "loss": 2.4810722351074217,
      "memory(GiB)": 66.02,
      "step": 20355,
      "token_acc": 0.4588235294117647,
      "train_speed(iter/s)": 1.446964
    },
    {
      "epoch": 0.8722848207017694,
      "grad_norm": 5.878935813903809,
      "learning_rate": 9.267651521882263e-05,
      "loss": 2.5962139129638673,
      "memory(GiB)": 66.02,
      "step": 20360,
      "token_acc": 0.48828125,
      "train_speed(iter/s)": 1.446997
    },
    {
      "epoch": 0.8724990360310184,
      "grad_norm": 4.918642520904541,
      "learning_rate": 9.267300832976626e-05,
      "loss": 2.899725914001465,
      "memory(GiB)": 66.02,
      "step": 20365,
      "token_acc": 0.3926174496644295,
      "train_speed(iter/s)": 1.446972
    },
    {
      "epoch": 0.8727132513602673,
      "grad_norm": 4.57191276550293,
      "learning_rate": 9.266950066764832e-05,
      "loss": 2.340840530395508,
      "memory(GiB)": 66.02,
      "step": 20370,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.446966
    },
    {
      "epoch": 0.8729274666895163,
      "grad_norm": 4.200679779052734,
      "learning_rate": 9.266599223253234e-05,
      "loss": 2.6033843994140624,
      "memory(GiB)": 66.02,
      "step": 20375,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.446956
    },
    {
      "epoch": 0.8731416820187653,
      "grad_norm": 3.744687557220459,
      "learning_rate": 9.266248302448188e-05,
      "loss": 2.5803451538085938,
      "memory(GiB)": 66.02,
      "step": 20380,
      "token_acc": 0.4452054794520548,
      "train_speed(iter/s)": 1.447061
    },
    {
      "epoch": 0.8733558973480142,
      "grad_norm": 5.624828815460205,
      "learning_rate": 9.265897304356055e-05,
      "loss": 2.7657032012939453,
      "memory(GiB)": 66.02,
      "step": 20385,
      "token_acc": 0.42962962962962964,
      "train_speed(iter/s)": 1.447105
    },
    {
      "epoch": 0.8735701126772631,
      "grad_norm": 5.167118549346924,
      "learning_rate": 9.265546228983189e-05,
      "loss": 2.6485198974609374,
      "memory(GiB)": 66.02,
      "step": 20390,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.447051
    },
    {
      "epoch": 0.8737843280065122,
      "grad_norm": 4.024702548980713,
      "learning_rate": 9.26519507633595e-05,
      "loss": 2.53060302734375,
      "memory(GiB)": 66.02,
      "step": 20395,
      "token_acc": 0.45514950166112955,
      "train_speed(iter/s)": 1.447025
    },
    {
      "epoch": 0.8739985433357611,
      "grad_norm": 4.909636974334717,
      "learning_rate": 9.264843846420702e-05,
      "loss": 2.6260467529296876,
      "memory(GiB)": 66.02,
      "step": 20400,
      "token_acc": 0.4520547945205479,
      "train_speed(iter/s)": 1.44697
    },
    {
      "epoch": 0.87421275866501,
      "grad_norm": 4.184407711029053,
      "learning_rate": 9.264492539243808e-05,
      "loss": 2.6808338165283203,
      "memory(GiB)": 66.02,
      "step": 20405,
      "token_acc": 0.44285714285714284,
      "train_speed(iter/s)": 1.446941
    },
    {
      "epoch": 0.8744269739942591,
      "grad_norm": 4.637807369232178,
      "learning_rate": 9.26414115481163e-05,
      "loss": 2.5490180969238283,
      "memory(GiB)": 66.02,
      "step": 20410,
      "token_acc": 0.46558704453441296,
      "train_speed(iter/s)": 1.446984
    },
    {
      "epoch": 0.874641189323508,
      "grad_norm": 5.338601589202881,
      "learning_rate": 9.263789693130535e-05,
      "loss": 2.607616996765137,
      "memory(GiB)": 66.02,
      "step": 20415,
      "token_acc": 0.45307443365695793,
      "train_speed(iter/s)": 1.446926
    },
    {
      "epoch": 0.8748554046527569,
      "grad_norm": 5.417995452880859,
      "learning_rate": 9.263438154206892e-05,
      "loss": 2.6932701110839843,
      "memory(GiB)": 66.02,
      "step": 20420,
      "token_acc": 0.4602076124567474,
      "train_speed(iter/s)": 1.446931
    },
    {
      "epoch": 0.8750696199820059,
      "grad_norm": 3.8684165477752686,
      "learning_rate": 9.263086538047064e-05,
      "loss": 2.5578914642333985,
      "memory(GiB)": 66.02,
      "step": 20425,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.446889
    },
    {
      "epoch": 0.8752838353112549,
      "grad_norm": 3.9209108352661133,
      "learning_rate": 9.262734844657425e-05,
      "loss": 2.691782760620117,
      "memory(GiB)": 66.02,
      "step": 20430,
      "token_acc": 0.479108635097493,
      "train_speed(iter/s)": 1.446835
    },
    {
      "epoch": 0.8754980506405038,
      "grad_norm": 5.319073677062988,
      "learning_rate": 9.262383074044347e-05,
      "loss": 2.475202751159668,
      "memory(GiB)": 66.02,
      "step": 20435,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.44692
    },
    {
      "epoch": 0.8757122659697528,
      "grad_norm": 4.863601207733154,
      "learning_rate": 9.262031226214201e-05,
      "loss": 2.626394271850586,
      "memory(GiB)": 66.02,
      "step": 20440,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.446953
    },
    {
      "epoch": 0.8759264812990017,
      "grad_norm": 6.365195274353027,
      "learning_rate": 9.261679301173361e-05,
      "loss": 2.4839855194091798,
      "memory(GiB)": 66.02,
      "step": 20445,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.446977
    },
    {
      "epoch": 0.8761406966282507,
      "grad_norm": 6.444025039672852,
      "learning_rate": 9.261327298928203e-05,
      "loss": 2.427138328552246,
      "memory(GiB)": 66.02,
      "step": 20450,
      "token_acc": 0.4805194805194805,
      "train_speed(iter/s)": 1.447008
    },
    {
      "epoch": 0.8763549119574997,
      "grad_norm": 3.8496649265289307,
      "learning_rate": 9.260975219485104e-05,
      "loss": 2.5181474685668945,
      "memory(GiB)": 66.02,
      "step": 20455,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.44706
    },
    {
      "epoch": 0.8765691272867486,
      "grad_norm": 4.388014793395996,
      "learning_rate": 9.26062306285044e-05,
      "loss": 2.391971969604492,
      "memory(GiB)": 66.02,
      "step": 20460,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.447112
    },
    {
      "epoch": 0.8767833426159976,
      "grad_norm": 3.3938064575195312,
      "learning_rate": 9.260270829030594e-05,
      "loss": 2.385320281982422,
      "memory(GiB)": 66.02,
      "step": 20465,
      "token_acc": 0.5310344827586206,
      "train_speed(iter/s)": 1.447117
    },
    {
      "epoch": 0.8769975579452466,
      "grad_norm": 4.626340389251709,
      "learning_rate": 9.259918518031944e-05,
      "loss": 2.3799556732177733,
      "memory(GiB)": 66.02,
      "step": 20470,
      "token_acc": 0.5155038759689923,
      "train_speed(iter/s)": 1.447124
    },
    {
      "epoch": 0.8772117732744955,
      "grad_norm": 4.65280294418335,
      "learning_rate": 9.259566129860874e-05,
      "loss": 2.9183542251586916,
      "memory(GiB)": 66.02,
      "step": 20475,
      "token_acc": 0.396078431372549,
      "train_speed(iter/s)": 1.447123
    },
    {
      "epoch": 0.8774259886037444,
      "grad_norm": 4.689497470855713,
      "learning_rate": 9.25921366452377e-05,
      "loss": 2.5879859924316406,
      "memory(GiB)": 66.02,
      "step": 20480,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.447129
    },
    {
      "epoch": 0.8776402039329935,
      "grad_norm": 7.428919792175293,
      "learning_rate": 9.258861122027013e-05,
      "loss": 2.4727554321289062,
      "memory(GiB)": 66.02,
      "step": 20485,
      "token_acc": 0.463768115942029,
      "train_speed(iter/s)": 1.447223
    },
    {
      "epoch": 0.8778544192622424,
      "grad_norm": 5.427867889404297,
      "learning_rate": 9.258508502376992e-05,
      "loss": 2.6388469696044923,
      "memory(GiB)": 66.02,
      "step": 20490,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.447279
    },
    {
      "epoch": 0.8780686345914913,
      "grad_norm": 4.802516937255859,
      "learning_rate": 9.258155805580095e-05,
      "loss": 2.357114791870117,
      "memory(GiB)": 66.02,
      "step": 20495,
      "token_acc": 0.4775510204081633,
      "train_speed(iter/s)": 1.447317
    },
    {
      "epoch": 0.8782828499207403,
      "grad_norm": 4.4269795417785645,
      "learning_rate": 9.257803031642711e-05,
      "loss": 2.624612045288086,
      "memory(GiB)": 66.02,
      "step": 20500,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.447265
    },
    {
      "epoch": 0.8782828499207403,
      "eval_loss": 2.3010377883911133,
      "eval_runtime": 14.7646,
      "eval_samples_per_second": 6.773,
      "eval_steps_per_second": 6.773,
      "eval_token_acc": 0.46108490566037735,
      "step": 20500
    },
    {
      "epoch": 0.8784970652499893,
      "grad_norm": 4.085290431976318,
      "learning_rate": 9.257450180571232e-05,
      "loss": 2.4852075576782227,
      "memory(GiB)": 66.02,
      "step": 20505,
      "token_acc": 0.4701159678858162,
      "train_speed(iter/s)": 1.445707
    },
    {
      "epoch": 0.8787112805792382,
      "grad_norm": 3.283400535583496,
      "learning_rate": 9.257097252372047e-05,
      "loss": 2.671334457397461,
      "memory(GiB)": 66.02,
      "step": 20510,
      "token_acc": 0.4066265060240964,
      "train_speed(iter/s)": 1.44566
    },
    {
      "epoch": 0.8789254959084872,
      "grad_norm": 4.380463123321533,
      "learning_rate": 9.256744247051553e-05,
      "loss": 2.5247406005859374,
      "memory(GiB)": 66.02,
      "step": 20515,
      "token_acc": 0.45564516129032256,
      "train_speed(iter/s)": 1.445746
    },
    {
      "epoch": 0.8791397112377362,
      "grad_norm": 4.257956027984619,
      "learning_rate": 9.256391164616144e-05,
      "loss": 2.5347347259521484,
      "memory(GiB)": 66.02,
      "step": 20520,
      "token_acc": 0.47633136094674555,
      "train_speed(iter/s)": 1.445807
    },
    {
      "epoch": 0.8793539265669851,
      "grad_norm": 4.6171417236328125,
      "learning_rate": 9.256038005072216e-05,
      "loss": 2.715629005432129,
      "memory(GiB)": 66.02,
      "step": 20525,
      "token_acc": 0.4206896551724138,
      "train_speed(iter/s)": 1.445852
    },
    {
      "epoch": 0.8795681418962341,
      "grad_norm": 4.8964152336120605,
      "learning_rate": 9.255684768426168e-05,
      "loss": 2.3374588012695314,
      "memory(GiB)": 66.02,
      "step": 20530,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.445886
    },
    {
      "epoch": 0.879782357225483,
      "grad_norm": 4.855588912963867,
      "learning_rate": 9.255331454684395e-05,
      "loss": 2.7803354263305664,
      "memory(GiB)": 66.02,
      "step": 20535,
      "token_acc": 0.4492753623188406,
      "train_speed(iter/s)": 1.445909
    },
    {
      "epoch": 0.8799965725547321,
      "grad_norm": 4.195810317993164,
      "learning_rate": 9.254978063853303e-05,
      "loss": 2.497849464416504,
      "memory(GiB)": 66.02,
      "step": 20540,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.445918
    },
    {
      "epoch": 0.880210787883981,
      "grad_norm": 5.927244663238525,
      "learning_rate": 9.254624595939293e-05,
      "loss": 2.4883745193481444,
      "memory(GiB)": 66.02,
      "step": 20545,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.44584
    },
    {
      "epoch": 0.8804250032132299,
      "grad_norm": 4.772204875946045,
      "learning_rate": 9.254271050948767e-05,
      "loss": 2.6938032150268554,
      "memory(GiB)": 66.02,
      "step": 20550,
      "token_acc": 0.4258064516129032,
      "train_speed(iter/s)": 1.445834
    },
    {
      "epoch": 0.880639218542479,
      "grad_norm": 4.112085342407227,
      "learning_rate": 9.253917428888129e-05,
      "loss": 2.398793411254883,
      "memory(GiB)": 66.02,
      "step": 20555,
      "token_acc": 0.4267515923566879,
      "train_speed(iter/s)": 1.445884
    },
    {
      "epoch": 0.8808534338717279,
      "grad_norm": 4.269628047943115,
      "learning_rate": 9.253563729763786e-05,
      "loss": 2.4429323196411135,
      "memory(GiB)": 66.02,
      "step": 20560,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.445941
    },
    {
      "epoch": 0.8810676492009768,
      "grad_norm": 3.393285036087036,
      "learning_rate": 9.253209953582147e-05,
      "loss": 2.6097085952758787,
      "memory(GiB)": 66.02,
      "step": 20565,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.446012
    },
    {
      "epoch": 0.8812818645302258,
      "grad_norm": 4.838438034057617,
      "learning_rate": 9.25285610034962e-05,
      "loss": 2.8534555435180664,
      "memory(GiB)": 66.02,
      "step": 20570,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.44581
    },
    {
      "epoch": 0.8814960798594748,
      "grad_norm": 4.8219685554504395,
      "learning_rate": 9.252502170072615e-05,
      "loss": 2.448726272583008,
      "memory(GiB)": 66.02,
      "step": 20575,
      "token_acc": 0.516728624535316,
      "train_speed(iter/s)": 1.445895
    },
    {
      "epoch": 0.8817102951887237,
      "grad_norm": 5.260610580444336,
      "learning_rate": 9.252148162757542e-05,
      "loss": 2.6347301483154295,
      "memory(GiB)": 66.02,
      "step": 20580,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.445927
    },
    {
      "epoch": 0.8819245105179727,
      "grad_norm": 4.973104000091553,
      "learning_rate": 9.251794078410818e-05,
      "loss": 2.560206985473633,
      "memory(GiB)": 66.02,
      "step": 20585,
      "token_acc": 0.4779874213836478,
      "train_speed(iter/s)": 1.445975
    },
    {
      "epoch": 0.8821387258472216,
      "grad_norm": 5.514646053314209,
      "learning_rate": 9.251439917038856e-05,
      "loss": 2.6100704193115236,
      "memory(GiB)": 66.02,
      "step": 20590,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.445961
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 2.9874496459960938,
      "learning_rate": 9.251085678648072e-05,
      "loss": 2.350851058959961,
      "memory(GiB)": 66.02,
      "step": 20595,
      "token_acc": 0.5015673981191222,
      "train_speed(iter/s)": 1.445936
    },
    {
      "epoch": 0.8825671565057196,
      "grad_norm": 4.191380977630615,
      "learning_rate": 9.250731363244882e-05,
      "loss": 2.6914560317993166,
      "memory(GiB)": 66.02,
      "step": 20600,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.445953
    },
    {
      "epoch": 0.8827813718349685,
      "grad_norm": 6.398646354675293,
      "learning_rate": 9.250376970835706e-05,
      "loss": 2.7345439910888674,
      "memory(GiB)": 66.02,
      "step": 20605,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.445932
    },
    {
      "epoch": 0.8829955871642174,
      "grad_norm": 4.889020919799805,
      "learning_rate": 9.250022501426965e-05,
      "loss": 2.6503929138183593,
      "memory(GiB)": 66.02,
      "step": 20610,
      "token_acc": 0.43157894736842106,
      "train_speed(iter/s)": 1.445941
    },
    {
      "epoch": 0.8832098024934665,
      "grad_norm": 4.379112720489502,
      "learning_rate": 9.249667955025079e-05,
      "loss": 2.8273773193359375,
      "memory(GiB)": 66.02,
      "step": 20615,
      "token_acc": 0.4281524926686217,
      "train_speed(iter/s)": 1.445993
    },
    {
      "epoch": 0.8834240178227154,
      "grad_norm": 3.438961982727051,
      "learning_rate": 9.24931333163647e-05,
      "loss": 2.3543167114257812,
      "memory(GiB)": 66.02,
      "step": 20620,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.445977
    },
    {
      "epoch": 0.8836382331519643,
      "grad_norm": 4.19992733001709,
      "learning_rate": 9.248958631267566e-05,
      "loss": 2.9201812744140625,
      "memory(GiB)": 66.02,
      "step": 20625,
      "token_acc": 0.3770491803278688,
      "train_speed(iter/s)": 1.445957
    },
    {
      "epoch": 0.8838524484812134,
      "grad_norm": 4.5692009925842285,
      "learning_rate": 9.248603853924789e-05,
      "loss": 2.5510080337524412,
      "memory(GiB)": 66.02,
      "step": 20630,
      "token_acc": 0.4664429530201342,
      "train_speed(iter/s)": 1.446031
    },
    {
      "epoch": 0.8840666638104623,
      "grad_norm": 5.530004024505615,
      "learning_rate": 9.248248999614568e-05,
      "loss": 2.80223388671875,
      "memory(GiB)": 66.02,
      "step": 20635,
      "token_acc": 0.41947565543071164,
      "train_speed(iter/s)": 1.446146
    },
    {
      "epoch": 0.8842808791397112,
      "grad_norm": 4.1893463134765625,
      "learning_rate": 9.247894068343332e-05,
      "loss": 2.6047887802124023,
      "memory(GiB)": 66.02,
      "step": 20640,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.446189
    },
    {
      "epoch": 0.8844950944689602,
      "grad_norm": 4.4396820068359375,
      "learning_rate": 9.24753906011751e-05,
      "loss": 2.528820610046387,
      "memory(GiB)": 66.02,
      "step": 20645,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.446279
    },
    {
      "epoch": 0.8847093097982092,
      "grad_norm": 5.101334095001221,
      "learning_rate": 9.247183974943532e-05,
      "loss": 2.7585237503051756,
      "memory(GiB)": 66.02,
      "step": 20650,
      "token_acc": 0.44912280701754387,
      "train_speed(iter/s)": 1.446325
    },
    {
      "epoch": 0.8849235251274581,
      "grad_norm": 3.763733148574829,
      "learning_rate": 9.246828812827834e-05,
      "loss": 2.4597408294677736,
      "memory(GiB)": 66.02,
      "step": 20655,
      "token_acc": 0.45323741007194246,
      "train_speed(iter/s)": 1.446286
    },
    {
      "epoch": 0.8851377404567071,
      "grad_norm": 5.465420246124268,
      "learning_rate": 9.246473573776845e-05,
      "loss": 2.6887969970703125,
      "memory(GiB)": 66.02,
      "step": 20660,
      "token_acc": 0.45481927710843373,
      "train_speed(iter/s)": 1.446292
    },
    {
      "epoch": 0.885351955785956,
      "grad_norm": 4.568000316619873,
      "learning_rate": 9.246118257797007e-05,
      "loss": 2.639945220947266,
      "memory(GiB)": 66.02,
      "step": 20665,
      "token_acc": 0.4492753623188406,
      "train_speed(iter/s)": 1.446346
    },
    {
      "epoch": 0.885566171115205,
      "grad_norm": 4.150634765625,
      "learning_rate": 9.245762864894754e-05,
      "loss": 2.996517562866211,
      "memory(GiB)": 66.02,
      "step": 20670,
      "token_acc": 0.41194029850746267,
      "train_speed(iter/s)": 1.446471
    },
    {
      "epoch": 0.885780386444454,
      "grad_norm": 5.055774688720703,
      "learning_rate": 9.245407395076522e-05,
      "loss": 2.773493766784668,
      "memory(GiB)": 66.02,
      "step": 20675,
      "token_acc": 0.4493927125506073,
      "train_speed(iter/s)": 1.446528
    },
    {
      "epoch": 0.8859946017737029,
      "grad_norm": 5.247165679931641,
      "learning_rate": 9.245051848348753e-05,
      "loss": 2.7089481353759766,
      "memory(GiB)": 66.02,
      "step": 20680,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.44646
    },
    {
      "epoch": 0.8862088171029519,
      "grad_norm": 4.205036640167236,
      "learning_rate": 9.244696224717889e-05,
      "loss": 2.763844108581543,
      "memory(GiB)": 66.02,
      "step": 20685,
      "token_acc": 0.45555555555555555,
      "train_speed(iter/s)": 1.446553
    },
    {
      "epoch": 0.8864230324322009,
      "grad_norm": 4.1134033203125,
      "learning_rate": 9.244340524190368e-05,
      "loss": 2.6551750183105467,
      "memory(GiB)": 66.02,
      "step": 20690,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.446561
    },
    {
      "epoch": 0.8866372477614498,
      "grad_norm": 3.585674285888672,
      "learning_rate": 9.243984746772638e-05,
      "loss": 2.7241180419921873,
      "memory(GiB)": 66.02,
      "step": 20695,
      "token_acc": 0.4591194968553459,
      "train_speed(iter/s)": 1.446469
    },
    {
      "epoch": 0.8868514630906987,
      "grad_norm": 4.635261535644531,
      "learning_rate": 9.243628892471145e-05,
      "loss": 2.594484329223633,
      "memory(GiB)": 66.02,
      "step": 20700,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.446361
    },
    {
      "epoch": 0.8870656784199478,
      "grad_norm": 4.641414165496826,
      "learning_rate": 9.243272961292331e-05,
      "loss": 2.5772079467773437,
      "memory(GiB)": 66.02,
      "step": 20705,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.446347
    },
    {
      "epoch": 0.8872798937491967,
      "grad_norm": 4.592565059661865,
      "learning_rate": 9.242916953242649e-05,
      "loss": 2.693921661376953,
      "memory(GiB)": 66.02,
      "step": 20710,
      "token_acc": 0.44534412955465585,
      "train_speed(iter/s)": 1.446347
    },
    {
      "epoch": 0.8874941090784456,
      "grad_norm": 4.181694984436035,
      "learning_rate": 9.242560868328545e-05,
      "loss": 2.5933982849121096,
      "memory(GiB)": 66.02,
      "step": 20715,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.446386
    },
    {
      "epoch": 0.8877083244076946,
      "grad_norm": 3.5672190189361572,
      "learning_rate": 9.242204706556472e-05,
      "loss": 2.6774166107177733,
      "memory(GiB)": 66.02,
      "step": 20720,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.446359
    },
    {
      "epoch": 0.8879225397369436,
      "grad_norm": 4.95294189453125,
      "learning_rate": 9.241848467932881e-05,
      "loss": 2.7005905151367187,
      "memory(GiB)": 66.02,
      "step": 20725,
      "token_acc": 0.41114982578397213,
      "train_speed(iter/s)": 1.44637
    },
    {
      "epoch": 0.8881367550661925,
      "grad_norm": 4.768760681152344,
      "learning_rate": 9.241492152464225e-05,
      "loss": 2.5643665313720705,
      "memory(GiB)": 66.02,
      "step": 20730,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.446301
    },
    {
      "epoch": 0.8883509703954415,
      "grad_norm": 3.995396375656128,
      "learning_rate": 9.24113576015696e-05,
      "loss": 2.3804933547973635,
      "memory(GiB)": 66.02,
      "step": 20735,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.446308
    },
    {
      "epoch": 0.8885651857246905,
      "grad_norm": 4.0110697746276855,
      "learning_rate": 9.240779291017542e-05,
      "loss": 2.4697078704833983,
      "memory(GiB)": 66.02,
      "step": 20740,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.446309
    },
    {
      "epoch": 0.8887794010539394,
      "grad_norm": 5.562434673309326,
      "learning_rate": 9.240422745052431e-05,
      "loss": 2.7784284591674804,
      "memory(GiB)": 66.02,
      "step": 20745,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.446312
    },
    {
      "epoch": 0.8889936163831884,
      "grad_norm": 6.544858932495117,
      "learning_rate": 9.240066122268081e-05,
      "loss": 2.5637544631958007,
      "memory(GiB)": 66.02,
      "step": 20750,
      "token_acc": 0.4751552795031056,
      "train_speed(iter/s)": 1.446312
    },
    {
      "epoch": 0.8892078317124373,
      "grad_norm": 4.335330009460449,
      "learning_rate": 9.239709422670958e-05,
      "loss": 2.6714353561401367,
      "memory(GiB)": 66.02,
      "step": 20755,
      "token_acc": 0.4119601328903654,
      "train_speed(iter/s)": 1.446367
    },
    {
      "epoch": 0.8894220470416863,
      "grad_norm": 4.983273029327393,
      "learning_rate": 9.23935264626752e-05,
      "loss": 2.6674016952514648,
      "memory(GiB)": 66.02,
      "step": 20760,
      "token_acc": 0.4369230769230769,
      "train_speed(iter/s)": 1.446408
    },
    {
      "epoch": 0.8896362623709353,
      "grad_norm": 4.446269989013672,
      "learning_rate": 9.238995793064233e-05,
      "loss": 2.5938438415527343,
      "memory(GiB)": 66.02,
      "step": 20765,
      "token_acc": 0.43874643874643876,
      "train_speed(iter/s)": 1.44652
    },
    {
      "epoch": 0.8898504777001842,
      "grad_norm": 4.286979675292969,
      "learning_rate": 9.238638863067561e-05,
      "loss": 2.5131658554077148,
      "memory(GiB)": 66.02,
      "step": 20770,
      "token_acc": 0.47592067988668557,
      "train_speed(iter/s)": 1.446522
    },
    {
      "epoch": 0.8900646930294331,
      "grad_norm": 6.719756126403809,
      "learning_rate": 9.23828185628397e-05,
      "loss": 2.4843597412109375,
      "memory(GiB)": 66.02,
      "step": 20775,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.446584
    },
    {
      "epoch": 0.8902789083586822,
      "grad_norm": 3.612745761871338,
      "learning_rate": 9.237924772719927e-05,
      "loss": 2.49188175201416,
      "memory(GiB)": 66.02,
      "step": 20780,
      "token_acc": 0.45689655172413796,
      "train_speed(iter/s)": 1.446616
    },
    {
      "epoch": 0.8904931236879311,
      "grad_norm": 4.251318454742432,
      "learning_rate": 9.237567612381901e-05,
      "loss": 2.505125045776367,
      "memory(GiB)": 66.02,
      "step": 20785,
      "token_acc": 0.5020080321285141,
      "train_speed(iter/s)": 1.446671
    },
    {
      "epoch": 0.89070733901718,
      "grad_norm": 4.86512565612793,
      "learning_rate": 9.237210375276361e-05,
      "loss": 2.444071960449219,
      "memory(GiB)": 66.02,
      "step": 20790,
      "token_acc": 0.4711111111111111,
      "train_speed(iter/s)": 1.446678
    },
    {
      "epoch": 0.8909215543464291,
      "grad_norm": 6.499415874481201,
      "learning_rate": 9.236853061409783e-05,
      "loss": 2.8627498626708983,
      "memory(GiB)": 66.02,
      "step": 20795,
      "token_acc": 0.4297872340425532,
      "train_speed(iter/s)": 1.446707
    },
    {
      "epoch": 0.891135769675678,
      "grad_norm": 3.7318472862243652,
      "learning_rate": 9.236495670788636e-05,
      "loss": 2.509539985656738,
      "memory(GiB)": 66.02,
      "step": 20800,
      "token_acc": 0.5227272727272727,
      "train_speed(iter/s)": 1.446716
    },
    {
      "epoch": 0.8913499850049269,
      "grad_norm": 5.019067764282227,
      "learning_rate": 9.236138203419396e-05,
      "loss": 2.483055305480957,
      "memory(GiB)": 66.02,
      "step": 20805,
      "token_acc": 0.4707792207792208,
      "train_speed(iter/s)": 1.446643
    },
    {
      "epoch": 0.8915642003341759,
      "grad_norm": 5.08335018157959,
      "learning_rate": 9.23578065930854e-05,
      "loss": 2.661589813232422,
      "memory(GiB)": 66.02,
      "step": 20810,
      "token_acc": 0.4478114478114478,
      "train_speed(iter/s)": 1.446676
    },
    {
      "epoch": 0.8917784156634249,
      "grad_norm": 4.337501525878906,
      "learning_rate": 9.235423038462542e-05,
      "loss": 2.251332664489746,
      "memory(GiB)": 66.02,
      "step": 20815,
      "token_acc": 0.4858490566037736,
      "train_speed(iter/s)": 1.44666
    },
    {
      "epoch": 0.8919926309926738,
      "grad_norm": 5.828814506530762,
      "learning_rate": 9.235065340887882e-05,
      "loss": 2.5227634429931642,
      "memory(GiB)": 66.02,
      "step": 20820,
      "token_acc": 0.4700460829493088,
      "train_speed(iter/s)": 1.446713
    },
    {
      "epoch": 0.8922068463219228,
      "grad_norm": 4.543949604034424,
      "learning_rate": 9.234707566591042e-05,
      "loss": 2.909804916381836,
      "memory(GiB)": 66.02,
      "step": 20825,
      "token_acc": 0.429042904290429,
      "train_speed(iter/s)": 1.446725
    },
    {
      "epoch": 0.8924210616511717,
      "grad_norm": 3.9868791103363037,
      "learning_rate": 9.234349715578501e-05,
      "loss": 2.3819196701049803,
      "memory(GiB)": 66.02,
      "step": 20830,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.44667
    },
    {
      "epoch": 0.8926352769804207,
      "grad_norm": 5.507164001464844,
      "learning_rate": 9.233991787856743e-05,
      "loss": 2.607301139831543,
      "memory(GiB)": 66.02,
      "step": 20835,
      "token_acc": 0.46616541353383456,
      "train_speed(iter/s)": 1.446722
    },
    {
      "epoch": 0.8928494923096697,
      "grad_norm": 6.382336139678955,
      "learning_rate": 9.233633783432252e-05,
      "loss": 2.650271987915039,
      "memory(GiB)": 66.02,
      "step": 20840,
      "token_acc": 0.4494047619047619,
      "train_speed(iter/s)": 1.446704
    },
    {
      "epoch": 0.8930637076389186,
      "grad_norm": 4.958181858062744,
      "learning_rate": 9.233275702311514e-05,
      "loss": 2.453841972351074,
      "memory(GiB)": 66.02,
      "step": 20845,
      "token_acc": 0.5241635687732342,
      "train_speed(iter/s)": 1.446667
    },
    {
      "epoch": 0.8932779229681675,
      "grad_norm": 6.049378871917725,
      "learning_rate": 9.232917544501015e-05,
      "loss": 2.4255767822265626,
      "memory(GiB)": 66.02,
      "step": 20850,
      "token_acc": 0.553030303030303,
      "train_speed(iter/s)": 1.446618
    },
    {
      "epoch": 0.8934921382974166,
      "grad_norm": 3.888660430908203,
      "learning_rate": 9.232559310007245e-05,
      "loss": 2.5009271621704103,
      "memory(GiB)": 66.02,
      "step": 20855,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.446662
    },
    {
      "epoch": 0.8937063536266655,
      "grad_norm": 4.2253265380859375,
      "learning_rate": 9.23220099883669e-05,
      "loss": 2.607834815979004,
      "memory(GiB)": 66.02,
      "step": 20860,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.44668
    },
    {
      "epoch": 0.8939205689559145,
      "grad_norm": 3.559831142425537,
      "learning_rate": 9.231842610995847e-05,
      "loss": 2.653416633605957,
      "memory(GiB)": 66.02,
      "step": 20865,
      "token_acc": 0.4328767123287671,
      "train_speed(iter/s)": 1.44663
    },
    {
      "epoch": 0.8941347842851635,
      "grad_norm": 5.760776042938232,
      "learning_rate": 9.231484146491203e-05,
      "loss": 2.746925163269043,
      "memory(GiB)": 66.02,
      "step": 20870,
      "token_acc": 0.45136186770428016,
      "train_speed(iter/s)": 1.446614
    },
    {
      "epoch": 0.8943489996144124,
      "grad_norm": 3.818387508392334,
      "learning_rate": 9.231125605329256e-05,
      "loss": 2.5633007049560548,
      "memory(GiB)": 66.02,
      "step": 20875,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.446561
    },
    {
      "epoch": 0.8945632149436614,
      "grad_norm": 4.665847301483154,
      "learning_rate": 9.230766987516498e-05,
      "loss": 2.7993341445922852,
      "memory(GiB)": 66.02,
      "step": 20880,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.446619
    },
    {
      "epoch": 0.8947774302729103,
      "grad_norm": 4.266956329345703,
      "learning_rate": 9.230408293059427e-05,
      "loss": 2.537516784667969,
      "memory(GiB)": 66.02,
      "step": 20885,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.446641
    },
    {
      "epoch": 0.8949916456021593,
      "grad_norm": 5.539259433746338,
      "learning_rate": 9.230049521964544e-05,
      "loss": 2.1880725860595702,
      "memory(GiB)": 66.02,
      "step": 20890,
      "token_acc": 0.5635593220338984,
      "train_speed(iter/s)": 1.446634
    },
    {
      "epoch": 0.8952058609314083,
      "grad_norm": 5.82073450088501,
      "learning_rate": 9.229690674238343e-05,
      "loss": 2.5165029525756837,
      "memory(GiB)": 66.02,
      "step": 20895,
      "token_acc": 0.4472049689440994,
      "train_speed(iter/s)": 1.446635
    },
    {
      "epoch": 0.8954200762606572,
      "grad_norm": 4.196836471557617,
      "learning_rate": 9.22933174988733e-05,
      "loss": 2.6166725158691406,
      "memory(GiB)": 66.02,
      "step": 20900,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.446632
    },
    {
      "epoch": 0.8956342915899062,
      "grad_norm": 4.747260093688965,
      "learning_rate": 9.228972748918002e-05,
      "loss": 2.535319519042969,
      "memory(GiB)": 66.02,
      "step": 20905,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.446703
    },
    {
      "epoch": 0.8958485069191552,
      "grad_norm": 3.4472787380218506,
      "learning_rate": 9.228613671336867e-05,
      "loss": 2.454991340637207,
      "memory(GiB)": 66.02,
      "step": 20910,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.446698
    },
    {
      "epoch": 0.8960627222484041,
      "grad_norm": 3.922443151473999,
      "learning_rate": 9.228254517150429e-05,
      "loss": 2.7372570037841797,
      "memory(GiB)": 66.02,
      "step": 20915,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.4467
    },
    {
      "epoch": 0.896276937577653,
      "grad_norm": 6.497615337371826,
      "learning_rate": 9.227895286365194e-05,
      "loss": 2.6493576049804686,
      "memory(GiB)": 66.02,
      "step": 20920,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.446715
    },
    {
      "epoch": 0.8964911529069021,
      "grad_norm": 6.321664810180664,
      "learning_rate": 9.227535978987669e-05,
      "loss": 2.916431999206543,
      "memory(GiB)": 66.02,
      "step": 20925,
      "token_acc": 0.41964285714285715,
      "train_speed(iter/s)": 1.446792
    },
    {
      "epoch": 0.896705368236151,
      "grad_norm": 4.5049238204956055,
      "learning_rate": 9.227176595024364e-05,
      "loss": 2.614213562011719,
      "memory(GiB)": 66.02,
      "step": 20930,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.446823
    },
    {
      "epoch": 0.8969195835653999,
      "grad_norm": 6.199278831481934,
      "learning_rate": 9.22681713448179e-05,
      "loss": 2.6324741363525392,
      "memory(GiB)": 66.02,
      "step": 20935,
      "token_acc": 0.4732824427480916,
      "train_speed(iter/s)": 1.446876
    },
    {
      "epoch": 0.897133798894649,
      "grad_norm": 3.929837465286255,
      "learning_rate": 9.226457597366459e-05,
      "loss": 2.565410614013672,
      "memory(GiB)": 66.02,
      "step": 20940,
      "token_acc": 0.45918367346938777,
      "train_speed(iter/s)": 1.446925
    },
    {
      "epoch": 0.8973480142238979,
      "grad_norm": 4.632893085479736,
      "learning_rate": 9.226097983684883e-05,
      "loss": 2.125894546508789,
      "memory(GiB)": 66.02,
      "step": 20945,
      "token_acc": 0.5269230769230769,
      "train_speed(iter/s)": 1.446942
    },
    {
      "epoch": 0.8975622295531468,
      "grad_norm": 4.071660995483398,
      "learning_rate": 9.225738293443578e-05,
      "loss": 2.6710739135742188,
      "memory(GiB)": 66.02,
      "step": 20950,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.447016
    },
    {
      "epoch": 0.8977764448823958,
      "grad_norm": 4.499776840209961,
      "learning_rate": 9.22537852664906e-05,
      "loss": 2.9659128189086914,
      "memory(GiB)": 66.02,
      "step": 20955,
      "token_acc": 0.38181818181818183,
      "train_speed(iter/s)": 1.446973
    },
    {
      "epoch": 0.8979906602116448,
      "grad_norm": 4.787110805511475,
      "learning_rate": 9.225018683307846e-05,
      "loss": 2.8772682189941405,
      "memory(GiB)": 66.02,
      "step": 20960,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.446918
    },
    {
      "epoch": 0.8982048755408937,
      "grad_norm": 5.112154006958008,
      "learning_rate": 9.224658763426454e-05,
      "loss": 2.6717370986938476,
      "memory(GiB)": 66.02,
      "step": 20965,
      "token_acc": 0.4668769716088328,
      "train_speed(iter/s)": 1.446982
    },
    {
      "epoch": 0.8984190908701427,
      "grad_norm": 3.814579486846924,
      "learning_rate": 9.224298767011407e-05,
      "loss": 2.573265266418457,
      "memory(GiB)": 66.02,
      "step": 20970,
      "token_acc": 0.45390070921985815,
      "train_speed(iter/s)": 1.447048
    },
    {
      "epoch": 0.8986333061993916,
      "grad_norm": 4.138620376586914,
      "learning_rate": 9.223938694069227e-05,
      "loss": 2.760302734375,
      "memory(GiB)": 66.02,
      "step": 20975,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.447107
    },
    {
      "epoch": 0.8988475215286406,
      "grad_norm": 3.342966079711914,
      "learning_rate": 9.223578544606432e-05,
      "loss": 2.569215202331543,
      "memory(GiB)": 66.02,
      "step": 20980,
      "token_acc": 0.4768211920529801,
      "train_speed(iter/s)": 1.447115
    },
    {
      "epoch": 0.8990617368578896,
      "grad_norm": 5.396997451782227,
      "learning_rate": 9.223218318629551e-05,
      "loss": 2.7148935317993166,
      "memory(GiB)": 66.02,
      "step": 20985,
      "token_acc": 0.437984496124031,
      "train_speed(iter/s)": 1.447138
    },
    {
      "epoch": 0.8992759521871385,
      "grad_norm": 6.1343793869018555,
      "learning_rate": 9.222858016145108e-05,
      "loss": 2.3639318466186525,
      "memory(GiB)": 66.02,
      "step": 20990,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.447066
    },
    {
      "epoch": 0.8994901675163874,
      "grad_norm": 4.446055889129639,
      "learning_rate": 9.22249763715963e-05,
      "loss": 2.5438253402709963,
      "memory(GiB)": 66.02,
      "step": 20995,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.447081
    },
    {
      "epoch": 0.8997043828456365,
      "grad_norm": 3.5794870853424072,
      "learning_rate": 9.222137181679648e-05,
      "loss": 2.6343250274658203,
      "memory(GiB)": 66.02,
      "step": 21000,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.447092
    },
    {
      "epoch": 0.8997043828456365,
      "eval_loss": 2.172961950302124,
      "eval_runtime": 14.0706,
      "eval_samples_per_second": 7.107,
      "eval_steps_per_second": 7.107,
      "eval_token_acc": 0.47564469914040114,
      "step": 21000
    },
    {
      "epoch": 0.8999185981748854,
      "grad_norm": 4.638935565948486,
      "learning_rate": 9.221776649711689e-05,
      "loss": 2.503186798095703,
      "memory(GiB)": 66.02,
      "step": 21005,
      "token_acc": 0.4863731656184486,
      "train_speed(iter/s)": 1.445655
    },
    {
      "epoch": 0.9001328135041343,
      "grad_norm": 6.466837406158447,
      "learning_rate": 9.221416041262285e-05,
      "loss": 2.6534194946289062,
      "memory(GiB)": 66.02,
      "step": 21010,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.445751
    },
    {
      "epoch": 0.9003470288333834,
      "grad_norm": 3.8257768154144287,
      "learning_rate": 9.22105535633797e-05,
      "loss": 2.412603759765625,
      "memory(GiB)": 66.02,
      "step": 21015,
      "token_acc": 0.4890282131661442,
      "train_speed(iter/s)": 1.44583
    },
    {
      "epoch": 0.9005612441626323,
      "grad_norm": 4.2280659675598145,
      "learning_rate": 9.220694594945278e-05,
      "loss": 2.7420732498168947,
      "memory(GiB)": 66.02,
      "step": 21020,
      "token_acc": 0.43304843304843305,
      "train_speed(iter/s)": 1.445807
    },
    {
      "epoch": 0.9007754594918812,
      "grad_norm": 4.948489665985107,
      "learning_rate": 9.220333757090745e-05,
      "loss": 2.318073272705078,
      "memory(GiB)": 66.02,
      "step": 21025,
      "token_acc": 0.5164319248826291,
      "train_speed(iter/s)": 1.445727
    },
    {
      "epoch": 0.9009896748211302,
      "grad_norm": 5.825639724731445,
      "learning_rate": 9.219972842780907e-05,
      "loss": 2.5972923278808593,
      "memory(GiB)": 66.02,
      "step": 21030,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.445663
    },
    {
      "epoch": 0.9012038901503792,
      "grad_norm": 4.015948295593262,
      "learning_rate": 9.219611852022301e-05,
      "loss": 2.6458385467529295,
      "memory(GiB)": 66.02,
      "step": 21035,
      "token_acc": 0.46060606060606063,
      "train_speed(iter/s)": 1.445659
    },
    {
      "epoch": 0.9014181054796281,
      "grad_norm": 5.295448303222656,
      "learning_rate": 9.219250784821467e-05,
      "loss": 2.5657411575317384,
      "memory(GiB)": 66.02,
      "step": 21040,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.445702
    },
    {
      "epoch": 0.9016323208088771,
      "grad_norm": 3.9318182468414307,
      "learning_rate": 9.218889641184949e-05,
      "loss": 2.386408233642578,
      "memory(GiB)": 66.02,
      "step": 21045,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.44568
    },
    {
      "epoch": 0.901846536138126,
      "grad_norm": 4.602927207946777,
      "learning_rate": 9.218528421119287e-05,
      "loss": 2.8008617401123046,
      "memory(GiB)": 66.02,
      "step": 21050,
      "token_acc": 0.4690265486725664,
      "train_speed(iter/s)": 1.445743
    },
    {
      "epoch": 0.902060751467375,
      "grad_norm": 4.676817893981934,
      "learning_rate": 9.218167124631025e-05,
      "loss": 2.843794250488281,
      "memory(GiB)": 66.02,
      "step": 21055,
      "token_acc": 0.45878136200716846,
      "train_speed(iter/s)": 1.445844
    },
    {
      "epoch": 0.902274966796624,
      "grad_norm": 9.051002502441406,
      "learning_rate": 9.21780575172671e-05,
      "loss": 2.697144317626953,
      "memory(GiB)": 66.02,
      "step": 21060,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.44583
    },
    {
      "epoch": 0.9024891821258729,
      "grad_norm": 5.728804111480713,
      "learning_rate": 9.217444302412886e-05,
      "loss": 2.633591079711914,
      "memory(GiB)": 66.02,
      "step": 21065,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.445785
    },
    {
      "epoch": 0.9027033974551218,
      "grad_norm": 6.295378684997559,
      "learning_rate": 9.217082776696101e-05,
      "loss": 2.569627571105957,
      "memory(GiB)": 66.02,
      "step": 21070,
      "token_acc": 0.45041322314049587,
      "train_speed(iter/s)": 1.445853
    },
    {
      "epoch": 0.9029176127843709,
      "grad_norm": 6.806280612945557,
      "learning_rate": 9.216721174582907e-05,
      "loss": 2.5210811614990236,
      "memory(GiB)": 66.02,
      "step": 21075,
      "token_acc": 0.4735099337748344,
      "train_speed(iter/s)": 1.445877
    },
    {
      "epoch": 0.9031318281136198,
      "grad_norm": 4.269497871398926,
      "learning_rate": 9.216359496079851e-05,
      "loss": 2.9506881713867186,
      "memory(GiB)": 66.02,
      "step": 21080,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.445862
    },
    {
      "epoch": 0.9033460434428687,
      "grad_norm": 4.604902267456055,
      "learning_rate": 9.215997741193491e-05,
      "loss": 2.4697088241577148,
      "memory(GiB)": 66.02,
      "step": 21085,
      "token_acc": 0.4927007299270073,
      "train_speed(iter/s)": 1.445918
    },
    {
      "epoch": 0.9035602587721178,
      "grad_norm": 6.149738788604736,
      "learning_rate": 9.215635909930376e-05,
      "loss": 2.571335220336914,
      "memory(GiB)": 66.02,
      "step": 21090,
      "token_acc": 0.4397163120567376,
      "train_speed(iter/s)": 1.44592
    },
    {
      "epoch": 0.9037744741013667,
      "grad_norm": 4.23390531539917,
      "learning_rate": 9.21527400229706e-05,
      "loss": 2.5584247589111326,
      "memory(GiB)": 66.02,
      "step": 21095,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.445925
    },
    {
      "epoch": 0.9039886894306156,
      "grad_norm": 5.736716270446777,
      "learning_rate": 9.214912018300103e-05,
      "loss": 2.7630359649658205,
      "memory(GiB)": 66.02,
      "step": 21100,
      "token_acc": 0.4230769230769231,
      "train_speed(iter/s)": 1.445877
    },
    {
      "epoch": 0.9042029047598646,
      "grad_norm": 5.614219665527344,
      "learning_rate": 9.214549957946061e-05,
      "loss": 2.31036262512207,
      "memory(GiB)": 66.02,
      "step": 21105,
      "token_acc": 0.4830508474576271,
      "train_speed(iter/s)": 1.4459
    },
    {
      "epoch": 0.9044171200891136,
      "grad_norm": 6.216885089874268,
      "learning_rate": 9.214187821241492e-05,
      "loss": 2.558477592468262,
      "memory(GiB)": 66.02,
      "step": 21110,
      "token_acc": 0.483271375464684,
      "train_speed(iter/s)": 1.445946
    },
    {
      "epoch": 0.9046313354183625,
      "grad_norm": 3.8255560398101807,
      "learning_rate": 9.213825608192959e-05,
      "loss": 2.6169076919555665,
      "memory(GiB)": 66.02,
      "step": 21115,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.445972
    },
    {
      "epoch": 0.9048455507476115,
      "grad_norm": 4.41001558303833,
      "learning_rate": 9.213463318807021e-05,
      "loss": 2.5118234634399412,
      "memory(GiB)": 66.02,
      "step": 21120,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.446026
    },
    {
      "epoch": 0.9050597660768605,
      "grad_norm": 4.655299186706543,
      "learning_rate": 9.213100953090241e-05,
      "loss": 2.426678466796875,
      "memory(GiB)": 66.02,
      "step": 21125,
      "token_acc": 0.501432664756447,
      "train_speed(iter/s)": 1.445969
    },
    {
      "epoch": 0.9052739814061094,
      "grad_norm": 4.1010823249816895,
      "learning_rate": 9.212738511049187e-05,
      "loss": 2.468875503540039,
      "memory(GiB)": 66.02,
      "step": 21130,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.445903
    },
    {
      "epoch": 0.9054881967353584,
      "grad_norm": 3.8986330032348633,
      "learning_rate": 9.212375992690423e-05,
      "loss": 2.4613273620605467,
      "memory(GiB)": 66.02,
      "step": 21135,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.445981
    },
    {
      "epoch": 0.9057024120646073,
      "grad_norm": 5.170470714569092,
      "learning_rate": 9.212013398020516e-05,
      "loss": 2.6442886352539063,
      "memory(GiB)": 66.02,
      "step": 21140,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.445955
    },
    {
      "epoch": 0.9059166273938563,
      "grad_norm": 5.627382755279541,
      "learning_rate": 9.211650727046033e-05,
      "loss": 2.7149894714355467,
      "memory(GiB)": 66.02,
      "step": 21145,
      "token_acc": 0.45209580838323354,
      "train_speed(iter/s)": 1.445963
    },
    {
      "epoch": 0.9061308427231053,
      "grad_norm": 3.8872225284576416,
      "learning_rate": 9.211287979773548e-05,
      "loss": 2.677489471435547,
      "memory(GiB)": 66.02,
      "step": 21150,
      "token_acc": 0.476038338658147,
      "train_speed(iter/s)": 1.445963
    },
    {
      "epoch": 0.9063450580523542,
      "grad_norm": 5.081066131591797,
      "learning_rate": 9.210925156209632e-05,
      "loss": 2.3350507736206056,
      "memory(GiB)": 66.02,
      "step": 21155,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.445995
    },
    {
      "epoch": 0.9065592733816031,
      "grad_norm": 6.163038730621338,
      "learning_rate": 9.210562256360855e-05,
      "loss": 2.7246273040771483,
      "memory(GiB)": 66.02,
      "step": 21160,
      "token_acc": 0.4921259842519685,
      "train_speed(iter/s)": 1.446069
    },
    {
      "epoch": 0.9067734887108522,
      "grad_norm": 4.30300760269165,
      "learning_rate": 9.210199280233794e-05,
      "loss": 2.771466827392578,
      "memory(GiB)": 66.02,
      "step": 21165,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.446118
    },
    {
      "epoch": 0.9069877040401011,
      "grad_norm": 3.454042673110962,
      "learning_rate": 9.209836227835022e-05,
      "loss": 2.6793468475341795,
      "memory(GiB)": 66.02,
      "step": 21170,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.446189
    },
    {
      "epoch": 0.90720191936935,
      "grad_norm": 3.537275552749634,
      "learning_rate": 9.20947309917112e-05,
      "loss": 2.6368431091308593,
      "memory(GiB)": 66.02,
      "step": 21175,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.446225
    },
    {
      "epoch": 0.907416134698599,
      "grad_norm": 4.747765064239502,
      "learning_rate": 9.209109894248662e-05,
      "loss": 2.5317012786865236,
      "memory(GiB)": 66.02,
      "step": 21180,
      "token_acc": 0.4389438943894389,
      "train_speed(iter/s)": 1.446225
    },
    {
      "epoch": 0.907630350027848,
      "grad_norm": 3.6141278743743896,
      "learning_rate": 9.208746613074231e-05,
      "loss": 2.5407249450683596,
      "memory(GiB)": 66.02,
      "step": 21185,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.446258
    },
    {
      "epoch": 0.9078445653570969,
      "grad_norm": 4.709258079528809,
      "learning_rate": 9.208383255654406e-05,
      "loss": 2.942336654663086,
      "memory(GiB)": 66.02,
      "step": 21190,
      "token_acc": 0.42618384401114207,
      "train_speed(iter/s)": 1.446252
    },
    {
      "epoch": 0.9080587806863459,
      "grad_norm": 4.7240071296691895,
      "learning_rate": 9.20801982199577e-05,
      "loss": 2.747900390625,
      "memory(GiB)": 66.02,
      "step": 21195,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.446281
    },
    {
      "epoch": 0.9082729960155949,
      "grad_norm": 5.564262866973877,
      "learning_rate": 9.20765631210491e-05,
      "loss": 2.5454097747802735,
      "memory(GiB)": 66.02,
      "step": 21200,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.446352
    },
    {
      "epoch": 0.9084872113448439,
      "grad_norm": 4.614888668060303,
      "learning_rate": 9.207292725988405e-05,
      "loss": 2.7390678405761717,
      "memory(GiB)": 66.02,
      "step": 21205,
      "token_acc": 0.4117647058823529,
      "train_speed(iter/s)": 1.44636
    },
    {
      "epoch": 0.9087014266740928,
      "grad_norm": 5.1803154945373535,
      "learning_rate": 9.206929063652849e-05,
      "loss": 2.8002925872802735,
      "memory(GiB)": 66.02,
      "step": 21210,
      "token_acc": 0.42258064516129035,
      "train_speed(iter/s)": 1.44632
    },
    {
      "epoch": 0.9089156420033417,
      "grad_norm": 5.064484119415283,
      "learning_rate": 9.206565325104826e-05,
      "loss": 2.586495590209961,
      "memory(GiB)": 66.02,
      "step": 21215,
      "token_acc": 0.49794238683127573,
      "train_speed(iter/s)": 1.446354
    },
    {
      "epoch": 0.9091298573325908,
      "grad_norm": 4.219216823577881,
      "learning_rate": 9.206201510350925e-05,
      "loss": 2.4654720306396483,
      "memory(GiB)": 66.02,
      "step": 21220,
      "token_acc": 0.5202702702702703,
      "train_speed(iter/s)": 1.446434
    },
    {
      "epoch": 0.9093440726618397,
      "grad_norm": 3.962348222732544,
      "learning_rate": 9.205837619397738e-05,
      "loss": 2.737915802001953,
      "memory(GiB)": 66.02,
      "step": 21225,
      "token_acc": 0.4201388888888889,
      "train_speed(iter/s)": 1.446406
    },
    {
      "epoch": 0.9095582879910886,
      "grad_norm": 3.7598958015441895,
      "learning_rate": 9.205473652251858e-05,
      "loss": 2.337178039550781,
      "memory(GiB)": 66.02,
      "step": 21230,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.44643
    },
    {
      "epoch": 0.9097725033203377,
      "grad_norm": 4.624716758728027,
      "learning_rate": 9.205109608919878e-05,
      "loss": 2.201300048828125,
      "memory(GiB)": 66.02,
      "step": 21235,
      "token_acc": 0.527972027972028,
      "train_speed(iter/s)": 1.446495
    },
    {
      "epoch": 0.9099867186495866,
      "grad_norm": 4.461903095245361,
      "learning_rate": 9.20474548940839e-05,
      "loss": 2.646737289428711,
      "memory(GiB)": 66.02,
      "step": 21240,
      "token_acc": 0.4326923076923077,
      "train_speed(iter/s)": 1.446513
    },
    {
      "epoch": 0.9102009339788355,
      "grad_norm": 4.8072357177734375,
      "learning_rate": 9.204381293723996e-05,
      "loss": 3.1139034271240233,
      "memory(GiB)": 66.02,
      "step": 21245,
      "token_acc": 0.4115942028985507,
      "train_speed(iter/s)": 1.44649
    },
    {
      "epoch": 0.9104151493080845,
      "grad_norm": 5.154967308044434,
      "learning_rate": 9.204017021873289e-05,
      "loss": 2.5804229736328126,
      "memory(GiB)": 66.02,
      "step": 21250,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.446539
    },
    {
      "epoch": 0.9106293646373335,
      "grad_norm": 4.2552666664123535,
      "learning_rate": 9.20365267386287e-05,
      "loss": 2.462293243408203,
      "memory(GiB)": 66.02,
      "step": 21255,
      "token_acc": 0.4866920152091255,
      "train_speed(iter/s)": 1.446631
    },
    {
      "epoch": 0.9108435799665824,
      "grad_norm": 4.117024898529053,
      "learning_rate": 9.203288249699341e-05,
      "loss": 2.6193473815917967,
      "memory(GiB)": 66.02,
      "step": 21260,
      "token_acc": 0.41901408450704225,
      "train_speed(iter/s)": 1.446677
    },
    {
      "epoch": 0.9110577952958314,
      "grad_norm": 4.155416488647461,
      "learning_rate": 9.202923749389302e-05,
      "loss": 2.5048810958862306,
      "memory(GiB)": 66.02,
      "step": 21265,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.446696
    },
    {
      "epoch": 0.9112720106250803,
      "grad_norm": 5.3150129318237305,
      "learning_rate": 9.202559172939355e-05,
      "loss": 2.827397346496582,
      "memory(GiB)": 66.02,
      "step": 21270,
      "token_acc": 0.444015444015444,
      "train_speed(iter/s)": 1.446715
    },
    {
      "epoch": 0.9114862259543293,
      "grad_norm": 4.588735103607178,
      "learning_rate": 9.202194520356108e-05,
      "loss": 2.4561126708984373,
      "memory(GiB)": 66.02,
      "step": 21275,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.446786
    },
    {
      "epoch": 0.9117004412835783,
      "grad_norm": 6.2010111808776855,
      "learning_rate": 9.201829791646165e-05,
      "loss": 2.461725044250488,
      "memory(GiB)": 66.02,
      "step": 21280,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.446777
    },
    {
      "epoch": 0.9119146566128272,
      "grad_norm": 5.03751277923584,
      "learning_rate": 9.201464986816132e-05,
      "loss": 2.651559829711914,
      "memory(GiB)": 66.02,
      "step": 21285,
      "token_acc": 0.42953020134228187,
      "train_speed(iter/s)": 1.446811
    },
    {
      "epoch": 0.9121288719420761,
      "grad_norm": 4.223094940185547,
      "learning_rate": 9.201100105872622e-05,
      "loss": 2.8223472595214845,
      "memory(GiB)": 66.02,
      "step": 21290,
      "token_acc": 0.44648318042813456,
      "train_speed(iter/s)": 1.446818
    },
    {
      "epoch": 0.9123430872713252,
      "grad_norm": 4.478932857513428,
      "learning_rate": 9.200735148822241e-05,
      "loss": 2.398407554626465,
      "memory(GiB)": 66.02,
      "step": 21295,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.446879
    },
    {
      "epoch": 0.9125573026005741,
      "grad_norm": 5.171123504638672,
      "learning_rate": 9.200370115671604e-05,
      "loss": 2.5137500762939453,
      "memory(GiB)": 66.02,
      "step": 21300,
      "token_acc": 0.43630573248407645,
      "train_speed(iter/s)": 1.446885
    },
    {
      "epoch": 0.912771517929823,
      "grad_norm": 7.187526226043701,
      "learning_rate": 9.20000500642732e-05,
      "loss": 2.4938621520996094,
      "memory(GiB)": 66.02,
      "step": 21305,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.446738
    },
    {
      "epoch": 0.9129857332590721,
      "grad_norm": 4.40461540222168,
      "learning_rate": 9.199639821096006e-05,
      "loss": 2.840202522277832,
      "memory(GiB)": 66.02,
      "step": 21310,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.446705
    },
    {
      "epoch": 0.913199948588321,
      "grad_norm": 4.7580742835998535,
      "learning_rate": 9.199274559684277e-05,
      "loss": 2.9117860794067383,
      "memory(GiB)": 66.02,
      "step": 21315,
      "token_acc": 0.4181184668989547,
      "train_speed(iter/s)": 1.44676
    },
    {
      "epoch": 0.9134141639175699,
      "grad_norm": 5.559549808502197,
      "learning_rate": 9.198909222198751e-05,
      "loss": 2.766831970214844,
      "memory(GiB)": 66.02,
      "step": 21320,
      "token_acc": 0.41292134831460675,
      "train_speed(iter/s)": 1.446827
    },
    {
      "epoch": 0.913628379246819,
      "grad_norm": 5.4706926345825195,
      "learning_rate": 9.198543808646045e-05,
      "loss": 2.6354143142700197,
      "memory(GiB)": 66.02,
      "step": 21325,
      "token_acc": 0.4980694980694981,
      "train_speed(iter/s)": 1.446787
    },
    {
      "epoch": 0.9138425945760679,
      "grad_norm": 4.737473964691162,
      "learning_rate": 9.19817831903278e-05,
      "loss": 2.52447395324707,
      "memory(GiB)": 66.02,
      "step": 21330,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.446824
    },
    {
      "epoch": 0.9140568099053168,
      "grad_norm": 3.8506369590759277,
      "learning_rate": 9.197812753365575e-05,
      "loss": 2.699715805053711,
      "memory(GiB)": 66.02,
      "step": 21335,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.44685
    },
    {
      "epoch": 0.9142710252345658,
      "grad_norm": 4.3350396156311035,
      "learning_rate": 9.197447111651055e-05,
      "loss": 2.677760124206543,
      "memory(GiB)": 66.02,
      "step": 21340,
      "token_acc": 0.4608695652173913,
      "train_speed(iter/s)": 1.446875
    },
    {
      "epoch": 0.9144852405638147,
      "grad_norm": 3.9501583576202393,
      "learning_rate": 9.197081393895843e-05,
      "loss": 2.7794120788574217,
      "memory(GiB)": 66.02,
      "step": 21345,
      "token_acc": 0.43103448275862066,
      "train_speed(iter/s)": 1.446925
    },
    {
      "epoch": 0.9146994558930637,
      "grad_norm": 4.029372692108154,
      "learning_rate": 9.196715600106564e-05,
      "loss": 2.1019163131713867,
      "memory(GiB)": 66.02,
      "step": 21350,
      "token_acc": 0.5378787878787878,
      "train_speed(iter/s)": 1.446988
    },
    {
      "epoch": 0.9149136712223127,
      "grad_norm": 4.662520885467529,
      "learning_rate": 9.196349730289845e-05,
      "loss": 2.593465042114258,
      "memory(GiB)": 66.02,
      "step": 21355,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.446987
    },
    {
      "epoch": 0.9151278865515616,
      "grad_norm": 3.5103843212127686,
      "learning_rate": 9.195983784452315e-05,
      "loss": 2.5835250854492187,
      "memory(GiB)": 66.02,
      "step": 21360,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.446917
    },
    {
      "epoch": 0.9153421018808106,
      "grad_norm": 5.815867900848389,
      "learning_rate": 9.195617762600601e-05,
      "loss": 2.379450798034668,
      "memory(GiB)": 66.02,
      "step": 21365,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.446917
    },
    {
      "epoch": 0.9155563172100596,
      "grad_norm": 3.733875036239624,
      "learning_rate": 9.195251664741337e-05,
      "loss": 2.4395851135253905,
      "memory(GiB)": 66.02,
      "step": 21370,
      "token_acc": 0.48923076923076925,
      "train_speed(iter/s)": 1.446972
    },
    {
      "epoch": 0.9157705325393085,
      "grad_norm": 5.514097690582275,
      "learning_rate": 9.194885490881153e-05,
      "loss": 2.9035226821899416,
      "memory(GiB)": 66.02,
      "step": 21375,
      "token_acc": 0.43729903536977494,
      "train_speed(iter/s)": 1.447042
    },
    {
      "epoch": 0.9159847478685574,
      "grad_norm": 5.24214506149292,
      "learning_rate": 9.194519241026684e-05,
      "loss": 2.7846353530883787,
      "memory(GiB)": 66.02,
      "step": 21380,
      "token_acc": 0.42207792207792205,
      "train_speed(iter/s)": 1.447045
    },
    {
      "epoch": 0.9161989631978065,
      "grad_norm": 4.489255905151367,
      "learning_rate": 9.194152915184564e-05,
      "loss": 2.8071937561035156,
      "memory(GiB)": 66.02,
      "step": 21385,
      "token_acc": 0.43023255813953487,
      "train_speed(iter/s)": 1.44711
    },
    {
      "epoch": 0.9164131785270554,
      "grad_norm": 3.3107810020446777,
      "learning_rate": 9.193786513361428e-05,
      "loss": 2.5622711181640625,
      "memory(GiB)": 66.02,
      "step": 21390,
      "token_acc": 0.48036253776435045,
      "train_speed(iter/s)": 1.447167
    },
    {
      "epoch": 0.9166273938563043,
      "grad_norm": 5.04474401473999,
      "learning_rate": 9.193420035563916e-05,
      "loss": 2.6250728607177733,
      "memory(GiB)": 66.02,
      "step": 21395,
      "token_acc": 0.4437299035369775,
      "train_speed(iter/s)": 1.447203
    },
    {
      "epoch": 0.9168416091855534,
      "grad_norm": 4.405237197875977,
      "learning_rate": 9.193053481798667e-05,
      "loss": 2.6750654220581054,
      "memory(GiB)": 66.02,
      "step": 21400,
      "token_acc": 0.42011834319526625,
      "train_speed(iter/s)": 1.447245
    },
    {
      "epoch": 0.9170558245148023,
      "grad_norm": 6.6985321044921875,
      "learning_rate": 9.192686852072321e-05,
      "loss": 2.8048528671264648,
      "memory(GiB)": 66.02,
      "step": 21405,
      "token_acc": 0.4553191489361702,
      "train_speed(iter/s)": 1.447324
    },
    {
      "epoch": 0.9172700398440512,
      "grad_norm": 4.291528701782227,
      "learning_rate": 9.192320146391518e-05,
      "loss": 2.6468124389648438,
      "memory(GiB)": 66.02,
      "step": 21410,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.447399
    },
    {
      "epoch": 0.9174842551733002,
      "grad_norm": 3.8271381855010986,
      "learning_rate": 9.191953364762904e-05,
      "loss": 2.7598636627197264,
      "memory(GiB)": 66.02,
      "step": 21415,
      "token_acc": 0.44814814814814813,
      "train_speed(iter/s)": 1.447329
    },
    {
      "epoch": 0.9176984705025492,
      "grad_norm": 5.290842533111572,
      "learning_rate": 9.191586507193122e-05,
      "loss": 2.4327934265136717,
      "memory(GiB)": 66.02,
      "step": 21420,
      "token_acc": 0.5211267605633803,
      "train_speed(iter/s)": 1.447226
    },
    {
      "epoch": 0.9179126858317981,
      "grad_norm": 4.74054479598999,
      "learning_rate": 9.191219573688819e-05,
      "loss": 2.803517532348633,
      "memory(GiB)": 66.02,
      "step": 21425,
      "token_acc": 0.4360655737704918,
      "train_speed(iter/s)": 1.447203
    },
    {
      "epoch": 0.9181269011610471,
      "grad_norm": 3.487269163131714,
      "learning_rate": 9.190852564256641e-05,
      "loss": 2.4647871017456056,
      "memory(GiB)": 66.02,
      "step": 21430,
      "token_acc": 0.4851190476190476,
      "train_speed(iter/s)": 1.447273
    },
    {
      "epoch": 0.918341116490296,
      "grad_norm": 4.492441654205322,
      "learning_rate": 9.190485478903238e-05,
      "loss": 2.5874406814575197,
      "memory(GiB)": 66.02,
      "step": 21435,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.44729
    },
    {
      "epoch": 0.918555331819545,
      "grad_norm": 4.885055065155029,
      "learning_rate": 9.190118317635259e-05,
      "loss": 2.6426555633544924,
      "memory(GiB)": 66.02,
      "step": 21440,
      "token_acc": 0.4603658536585366,
      "train_speed(iter/s)": 1.447309
    },
    {
      "epoch": 0.918769547148794,
      "grad_norm": 4.512348175048828,
      "learning_rate": 9.189751080459357e-05,
      "loss": 2.3507299423217773,
      "memory(GiB)": 66.02,
      "step": 21445,
      "token_acc": 0.5252918287937743,
      "train_speed(iter/s)": 1.447285
    },
    {
      "epoch": 0.9189837624780429,
      "grad_norm": 5.213791370391846,
      "learning_rate": 9.189383767382182e-05,
      "loss": 2.8945537567138673,
      "memory(GiB)": 66.02,
      "step": 21450,
      "token_acc": 0.4525316455696203,
      "train_speed(iter/s)": 1.447331
    },
    {
      "epoch": 0.9191979778072918,
      "grad_norm": 4.647881507873535,
      "learning_rate": 9.189016378410393e-05,
      "loss": 2.7044212341308596,
      "memory(GiB)": 66.02,
      "step": 21455,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.447384
    },
    {
      "epoch": 0.9194121931365409,
      "grad_norm": 4.2234392166137695,
      "learning_rate": 9.188648913550641e-05,
      "loss": 2.5259857177734375,
      "memory(GiB)": 66.02,
      "step": 21460,
      "token_acc": 0.4817813765182186,
      "train_speed(iter/s)": 1.447414
    },
    {
      "epoch": 0.9196264084657898,
      "grad_norm": 5.436551570892334,
      "learning_rate": 9.188281372809584e-05,
      "loss": 2.7624088287353517,
      "memory(GiB)": 66.02,
      "step": 21465,
      "token_acc": 0.44545454545454544,
      "train_speed(iter/s)": 1.447394
    },
    {
      "epoch": 0.9198406237950387,
      "grad_norm": 4.727115631103516,
      "learning_rate": 9.187913756193882e-05,
      "loss": 2.9652545928955076,
      "memory(GiB)": 66.02,
      "step": 21470,
      "token_acc": 0.4406779661016949,
      "train_speed(iter/s)": 1.44747
    },
    {
      "epoch": 0.9200548391242878,
      "grad_norm": 4.309323310852051,
      "learning_rate": 9.187546063710193e-05,
      "loss": 2.5118219375610353,
      "memory(GiB)": 66.02,
      "step": 21475,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.447523
    },
    {
      "epoch": 0.9202690544535367,
      "grad_norm": 3.8176162242889404,
      "learning_rate": 9.18717829536518e-05,
      "loss": 2.592088317871094,
      "memory(GiB)": 66.02,
      "step": 21480,
      "token_acc": 0.42902208201892744,
      "train_speed(iter/s)": 1.447546
    },
    {
      "epoch": 0.9204832697827856,
      "grad_norm": 4.594817161560059,
      "learning_rate": 9.186810451165502e-05,
      "loss": 2.4093158721923826,
      "memory(GiB)": 66.02,
      "step": 21485,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.447583
    },
    {
      "epoch": 0.9206974851120346,
      "grad_norm": 4.779289722442627,
      "learning_rate": 9.186442531117828e-05,
      "loss": 2.9425094604492186,
      "memory(GiB)": 66.02,
      "step": 21490,
      "token_acc": 0.43130990415335463,
      "train_speed(iter/s)": 1.447542
    },
    {
      "epoch": 0.9209117004412836,
      "grad_norm": 4.596897602081299,
      "learning_rate": 9.18607453522882e-05,
      "loss": 2.5867481231689453,
      "memory(GiB)": 66.02,
      "step": 21495,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.44754
    },
    {
      "epoch": 0.9211259157705325,
      "grad_norm": 4.5871195793151855,
      "learning_rate": 9.185706463505143e-05,
      "loss": 2.416180229187012,
      "memory(GiB)": 66.02,
      "step": 21500,
      "token_acc": 0.5065359477124183,
      "train_speed(iter/s)": 1.447429
    },
    {
      "epoch": 0.9211259157705325,
      "eval_loss": 2.2096166610717773,
      "eval_runtime": 14.5991,
      "eval_samples_per_second": 6.85,
      "eval_steps_per_second": 6.85,
      "eval_token_acc": 0.46794871794871795,
      "step": 21500
    },
    {
      "epoch": 0.9213401310997815,
      "grad_norm": 4.842560291290283,
      "learning_rate": 9.185338315953468e-05,
      "loss": 2.567431831359863,
      "memory(GiB)": 66.02,
      "step": 21505,
      "token_acc": 0.4762790697674419,
      "train_speed(iter/s)": 1.445855
    },
    {
      "epoch": 0.9215543464290304,
      "grad_norm": 5.74186372756958,
      "learning_rate": 9.184970092580463e-05,
      "loss": 2.3399642944335937,
      "memory(GiB)": 66.02,
      "step": 21510,
      "token_acc": 0.47346938775510206,
      "train_speed(iter/s)": 1.445877
    },
    {
      "epoch": 0.9217685617582794,
      "grad_norm": 3.550483226776123,
      "learning_rate": 9.1846017933928e-05,
      "loss": 2.7218713760375977,
      "memory(GiB)": 66.02,
      "step": 21515,
      "token_acc": 0.43466666666666665,
      "train_speed(iter/s)": 1.44587
    },
    {
      "epoch": 0.9219827770875284,
      "grad_norm": 3.7454476356506348,
      "learning_rate": 9.184233418397148e-05,
      "loss": 2.5545284271240236,
      "memory(GiB)": 66.02,
      "step": 21520,
      "token_acc": 0.4669260700389105,
      "train_speed(iter/s)": 1.445847
    },
    {
      "epoch": 0.9221969924167773,
      "grad_norm": 4.751285076141357,
      "learning_rate": 9.183864967600184e-05,
      "loss": 2.7277864456176757,
      "memory(GiB)": 66.02,
      "step": 21525,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.445797
    },
    {
      "epoch": 0.9224112077460263,
      "grad_norm": 4.8122782707214355,
      "learning_rate": 9.183496441008581e-05,
      "loss": 2.2724990844726562,
      "memory(GiB)": 66.02,
      "step": 21530,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.445813
    },
    {
      "epoch": 0.9226254230752753,
      "grad_norm": 4.041903018951416,
      "learning_rate": 9.183127838629016e-05,
      "loss": 2.561549186706543,
      "memory(GiB)": 66.02,
      "step": 21535,
      "token_acc": 0.4756554307116105,
      "train_speed(iter/s)": 1.445824
    },
    {
      "epoch": 0.9228396384045242,
      "grad_norm": 4.776011943817139,
      "learning_rate": 9.182759160468164e-05,
      "loss": 2.621981620788574,
      "memory(GiB)": 66.02,
      "step": 21540,
      "token_acc": 0.4717741935483871,
      "train_speed(iter/s)": 1.445721
    },
    {
      "epoch": 0.9230538537337732,
      "grad_norm": 3.802180767059326,
      "learning_rate": 9.182390406532708e-05,
      "loss": 2.591941261291504,
      "memory(GiB)": 66.02,
      "step": 21545,
      "token_acc": 0.4699248120300752,
      "train_speed(iter/s)": 1.445778
    },
    {
      "epoch": 0.9232680690630222,
      "grad_norm": 4.105423450469971,
      "learning_rate": 9.182021576829326e-05,
      "loss": 2.516992378234863,
      "memory(GiB)": 66.02,
      "step": 21550,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.445774
    },
    {
      "epoch": 0.9234822843922711,
      "grad_norm": 4.851661682128906,
      "learning_rate": 9.1816526713647e-05,
      "loss": 2.7617799758911135,
      "memory(GiB)": 66.02,
      "step": 21555,
      "token_acc": 0.4753521126760563,
      "train_speed(iter/s)": 1.445757
    },
    {
      "epoch": 0.9236964997215201,
      "grad_norm": 4.442677021026611,
      "learning_rate": 9.181283690145514e-05,
      "loss": 2.4971900939941407,
      "memory(GiB)": 66.02,
      "step": 21560,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.445826
    },
    {
      "epoch": 0.923910715050769,
      "grad_norm": 3.7416703701019287,
      "learning_rate": 9.18091463317845e-05,
      "loss": 2.5543373107910154,
      "memory(GiB)": 66.02,
      "step": 21565,
      "token_acc": 0.4734848484848485,
      "train_speed(iter/s)": 1.445855
    },
    {
      "epoch": 0.924124930380018,
      "grad_norm": 4.9060540199279785,
      "learning_rate": 9.180545500470197e-05,
      "loss": 2.5978450775146484,
      "memory(GiB)": 66.02,
      "step": 21570,
      "token_acc": 0.4837662337662338,
      "train_speed(iter/s)": 1.44587
    },
    {
      "epoch": 0.924339145709267,
      "grad_norm": 6.2970967292785645,
      "learning_rate": 9.18017629202744e-05,
      "loss": 2.798686981201172,
      "memory(GiB)": 66.02,
      "step": 21575,
      "token_acc": 0.4714828897338403,
      "train_speed(iter/s)": 1.445929
    },
    {
      "epoch": 0.9245533610385159,
      "grad_norm": 5.021624565124512,
      "learning_rate": 9.179807007856867e-05,
      "loss": 2.6187065124511717,
      "memory(GiB)": 66.02,
      "step": 21580,
      "token_acc": 0.45774647887323944,
      "train_speed(iter/s)": 1.446013
    },
    {
      "epoch": 0.9247675763677649,
      "grad_norm": 6.220231056213379,
      "learning_rate": 9.179437647965172e-05,
      "loss": 2.4467517852783205,
      "memory(GiB)": 66.02,
      "step": 21585,
      "token_acc": 0.4728682170542636,
      "train_speed(iter/s)": 1.445955
    },
    {
      "epoch": 0.9249817916970139,
      "grad_norm": 5.461313724517822,
      "learning_rate": 9.179068212359041e-05,
      "loss": 2.8383514404296877,
      "memory(GiB)": 66.02,
      "step": 21590,
      "token_acc": 0.4256198347107438,
      "train_speed(iter/s)": 1.445924
    },
    {
      "epoch": 0.9251960070262628,
      "grad_norm": 5.336484909057617,
      "learning_rate": 9.178698701045169e-05,
      "loss": 2.770197296142578,
      "memory(GiB)": 66.02,
      "step": 21595,
      "token_acc": 0.40594059405940597,
      "train_speed(iter/s)": 1.445919
    },
    {
      "epoch": 0.9254102223555117,
      "grad_norm": 10.905637741088867,
      "learning_rate": 9.178329114030251e-05,
      "loss": 2.51379280090332,
      "memory(GiB)": 66.02,
      "step": 21600,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.445878
    },
    {
      "epoch": 0.9256244376847608,
      "grad_norm": 4.201478958129883,
      "learning_rate": 9.177959451320981e-05,
      "loss": 2.6344709396362305,
      "memory(GiB)": 66.02,
      "step": 21605,
      "token_acc": 0.45018450184501846,
      "train_speed(iter/s)": 1.445953
    },
    {
      "epoch": 0.9258386530140097,
      "grad_norm": 5.29052209854126,
      "learning_rate": 9.177589712924055e-05,
      "loss": 2.8175079345703127,
      "memory(GiB)": 66.02,
      "step": 21610,
      "token_acc": 0.436950146627566,
      "train_speed(iter/s)": 1.445989
    },
    {
      "epoch": 0.9260528683432586,
      "grad_norm": 4.763354778289795,
      "learning_rate": 9.177219898846175e-05,
      "loss": 2.378049850463867,
      "memory(GiB)": 66.02,
      "step": 21615,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.445973
    },
    {
      "epoch": 0.9262670836725077,
      "grad_norm": 3.910534620285034,
      "learning_rate": 9.176850009094037e-05,
      "loss": 2.36367130279541,
      "memory(GiB)": 66.02,
      "step": 21620,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.446045
    },
    {
      "epoch": 0.9264812990017566,
      "grad_norm": 6.240082263946533,
      "learning_rate": 9.176480043674343e-05,
      "loss": 2.765350341796875,
      "memory(GiB)": 66.02,
      "step": 21625,
      "token_acc": 0.5073529411764706,
      "train_speed(iter/s)": 1.446028
    },
    {
      "epoch": 0.9266955143310055,
      "grad_norm": 4.599140167236328,
      "learning_rate": 9.176110002593794e-05,
      "loss": 2.3109182357788085,
      "memory(GiB)": 66.02,
      "step": 21630,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.446011
    },
    {
      "epoch": 0.9269097296602545,
      "grad_norm": 4.406301021575928,
      "learning_rate": 9.175739885859095e-05,
      "loss": 2.5080699920654297,
      "memory(GiB)": 66.02,
      "step": 21635,
      "token_acc": 0.47794117647058826,
      "train_speed(iter/s)": 1.445968
    },
    {
      "epoch": 0.9271239449895035,
      "grad_norm": 3.972447156906128,
      "learning_rate": 9.175369693476951e-05,
      "loss": 2.4904502868652343,
      "memory(GiB)": 66.02,
      "step": 21640,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.445882
    },
    {
      "epoch": 0.9273381603187524,
      "grad_norm": 3.660184144973755,
      "learning_rate": 9.17499942545407e-05,
      "loss": 2.281089973449707,
      "memory(GiB)": 66.02,
      "step": 21645,
      "token_acc": 0.5310077519379846,
      "train_speed(iter/s)": 1.445867
    },
    {
      "epoch": 0.9275523756480014,
      "grad_norm": 5.657915115356445,
      "learning_rate": 9.174629081797156e-05,
      "loss": 2.4793445587158205,
      "memory(GiB)": 66.02,
      "step": 21650,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.445868
    },
    {
      "epoch": 0.9277665909772503,
      "grad_norm": 4.37311315536499,
      "learning_rate": 9.174258662512921e-05,
      "loss": 2.641525459289551,
      "memory(GiB)": 66.02,
      "step": 21655,
      "token_acc": 0.46616541353383456,
      "train_speed(iter/s)": 1.445913
    },
    {
      "epoch": 0.9279808063064993,
      "grad_norm": 4.937234401702881,
      "learning_rate": 9.173888167608074e-05,
      "loss": 3.104800987243652,
      "memory(GiB)": 66.02,
      "step": 21660,
      "token_acc": 0.422680412371134,
      "train_speed(iter/s)": 1.445913
    },
    {
      "epoch": 0.9281950216357483,
      "grad_norm": 4.419463157653809,
      "learning_rate": 9.173517597089328e-05,
      "loss": 2.9346687316894533,
      "memory(GiB)": 66.02,
      "step": 21665,
      "token_acc": 0.41924398625429554,
      "train_speed(iter/s)": 1.44588
    },
    {
      "epoch": 0.9284092369649972,
      "grad_norm": 4.571208477020264,
      "learning_rate": 9.173146950963396e-05,
      "loss": 2.823246192932129,
      "memory(GiB)": 66.02,
      "step": 21670,
      "token_acc": 0.4160839160839161,
      "train_speed(iter/s)": 1.445909
    },
    {
      "epoch": 0.9286234522942461,
      "grad_norm": 4.280839920043945,
      "learning_rate": 9.17277622923699e-05,
      "loss": 2.496957206726074,
      "memory(GiB)": 66.02,
      "step": 21675,
      "token_acc": 0.4485981308411215,
      "train_speed(iter/s)": 1.445946
    },
    {
      "epoch": 0.9288376676234952,
      "grad_norm": 3.9365475177764893,
      "learning_rate": 9.172405431916831e-05,
      "loss": 2.644087791442871,
      "memory(GiB)": 66.02,
      "step": 21680,
      "token_acc": 0.42433234421364985,
      "train_speed(iter/s)": 1.445969
    },
    {
      "epoch": 0.9290518829527441,
      "grad_norm": 19.35648536682129,
      "learning_rate": 9.172034559009632e-05,
      "loss": 2.5395315170288084,
      "memory(GiB)": 66.02,
      "step": 21685,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.445946
    },
    {
      "epoch": 0.929266098281993,
      "grad_norm": 3.9599685668945312,
      "learning_rate": 9.171663610522114e-05,
      "loss": 2.814164924621582,
      "memory(GiB)": 66.02,
      "step": 21690,
      "token_acc": 0.42574257425742573,
      "train_speed(iter/s)": 1.445997
    },
    {
      "epoch": 0.9294803136112421,
      "grad_norm": 3.758268117904663,
      "learning_rate": 9.171292586460996e-05,
      "loss": 2.6677173614501952,
      "memory(GiB)": 66.02,
      "step": 21695,
      "token_acc": 0.4560810810810811,
      "train_speed(iter/s)": 1.446046
    },
    {
      "epoch": 0.929694528940491,
      "grad_norm": 4.4717631340026855,
      "learning_rate": 9.170921486833e-05,
      "loss": 2.7518178939819338,
      "memory(GiB)": 66.02,
      "step": 21700,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.446002
    },
    {
      "epoch": 0.9299087442697399,
      "grad_norm": 4.0885748863220215,
      "learning_rate": 9.170550311644848e-05,
      "loss": 2.6211666107177733,
      "memory(GiB)": 66.02,
      "step": 21705,
      "token_acc": 0.39144736842105265,
      "train_speed(iter/s)": 1.445995
    },
    {
      "epoch": 0.9301229595989889,
      "grad_norm": 4.248121738433838,
      "learning_rate": 9.170179060903265e-05,
      "loss": 2.4145360946655274,
      "memory(GiB)": 66.02,
      "step": 21710,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.446016
    },
    {
      "epoch": 0.9303371749282379,
      "grad_norm": 3.6612682342529297,
      "learning_rate": 9.169807734614976e-05,
      "loss": 2.58431453704834,
      "memory(GiB)": 66.02,
      "step": 21715,
      "token_acc": 0.4434782608695652,
      "train_speed(iter/s)": 1.446114
    },
    {
      "epoch": 0.9305513902574868,
      "grad_norm": 4.904190540313721,
      "learning_rate": 9.16943633278671e-05,
      "loss": 2.574121856689453,
      "memory(GiB)": 66.02,
      "step": 21720,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.446115
    },
    {
      "epoch": 0.9307656055867358,
      "grad_norm": 4.228428363800049,
      "learning_rate": 9.169064855425191e-05,
      "loss": 2.6321155548095705,
      "memory(GiB)": 66.02,
      "step": 21725,
      "token_acc": 0.4786885245901639,
      "train_speed(iter/s)": 1.446165
    },
    {
      "epoch": 0.9309798209159847,
      "grad_norm": 3.748889207839966,
      "learning_rate": 9.168693302537155e-05,
      "loss": 2.486474609375,
      "memory(GiB)": 66.02,
      "step": 21730,
      "token_acc": 0.4487534626038781,
      "train_speed(iter/s)": 1.446141
    },
    {
      "epoch": 0.9311940362452337,
      "grad_norm": 4.557462692260742,
      "learning_rate": 9.168321674129326e-05,
      "loss": 2.7568225860595703,
      "memory(GiB)": 66.02,
      "step": 21735,
      "token_acc": 0.4578313253012048,
      "train_speed(iter/s)": 1.44616
    },
    {
      "epoch": 0.9314082515744827,
      "grad_norm": 4.188618183135986,
      "learning_rate": 9.16794997020844e-05,
      "loss": 2.7182632446289063,
      "memory(GiB)": 66.02,
      "step": 21740,
      "token_acc": 0.46794871794871795,
      "train_speed(iter/s)": 1.446213
    },
    {
      "epoch": 0.9316224669037316,
      "grad_norm": 4.227839469909668,
      "learning_rate": 9.167578190781232e-05,
      "loss": 2.254042053222656,
      "memory(GiB)": 66.02,
      "step": 21745,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.446253
    },
    {
      "epoch": 0.9318366822329806,
      "grad_norm": 4.528618812561035,
      "learning_rate": 9.167206335854435e-05,
      "loss": 2.4962574005126954,
      "memory(GiB)": 66.02,
      "step": 21750,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.446287
    },
    {
      "epoch": 0.9320508975622296,
      "grad_norm": 3.9008610248565674,
      "learning_rate": 9.166834405434785e-05,
      "loss": 2.6359542846679687,
      "memory(GiB)": 66.02,
      "step": 21755,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.446314
    },
    {
      "epoch": 0.9322651128914785,
      "grad_norm": 6.49666166305542,
      "learning_rate": 9.166462399529021e-05,
      "loss": 2.1529220581054687,
      "memory(GiB)": 66.02,
      "step": 21760,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.44622
    },
    {
      "epoch": 0.9324793282207274,
      "grad_norm": 6.105951309204102,
      "learning_rate": 9.166090318143883e-05,
      "loss": 2.214730644226074,
      "memory(GiB)": 66.02,
      "step": 21765,
      "token_acc": 0.524390243902439,
      "train_speed(iter/s)": 1.446259
    },
    {
      "epoch": 0.9326935435499765,
      "grad_norm": 5.192975044250488,
      "learning_rate": 9.165718161286111e-05,
      "loss": 2.997336578369141,
      "memory(GiB)": 66.02,
      "step": 21770,
      "token_acc": 0.42,
      "train_speed(iter/s)": 1.446281
    },
    {
      "epoch": 0.9329077588792254,
      "grad_norm": 4.810291767120361,
      "learning_rate": 9.165345928962446e-05,
      "loss": 2.4880397796630858,
      "memory(GiB)": 66.02,
      "step": 21775,
      "token_acc": 0.4984520123839009,
      "train_speed(iter/s)": 1.4463
    },
    {
      "epoch": 0.9331219742084743,
      "grad_norm": 4.143027305603027,
      "learning_rate": 9.164973621179634e-05,
      "loss": 2.485639953613281,
      "memory(GiB)": 66.02,
      "step": 21780,
      "token_acc": 0.49809885931558934,
      "train_speed(iter/s)": 1.446237
    },
    {
      "epoch": 0.9333361895377233,
      "grad_norm": 4.098141193389893,
      "learning_rate": 9.164601237944415e-05,
      "loss": 2.7792768478393555,
      "memory(GiB)": 66.02,
      "step": 21785,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.446199
    },
    {
      "epoch": 0.9335504048669723,
      "grad_norm": 6.270641803741455,
      "learning_rate": 9.16422877926354e-05,
      "loss": 2.7639646530151367,
      "memory(GiB)": 66.02,
      "step": 21790,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.446166
    },
    {
      "epoch": 0.9337646201962212,
      "grad_norm": 4.690643787384033,
      "learning_rate": 9.163856245143752e-05,
      "loss": 2.5041481018066407,
      "memory(GiB)": 66.02,
      "step": 21795,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.446094
    },
    {
      "epoch": 0.9339788355254702,
      "grad_norm": 4.700472831726074,
      "learning_rate": 9.163483635591804e-05,
      "loss": 2.362548828125,
      "memory(GiB)": 66.02,
      "step": 21800,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.446031
    },
    {
      "epoch": 0.9341930508547192,
      "grad_norm": 4.3396759033203125,
      "learning_rate": 9.163110950614445e-05,
      "loss": 2.463043212890625,
      "memory(GiB)": 66.02,
      "step": 21805,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.446055
    },
    {
      "epoch": 0.9344072661839681,
      "grad_norm": 3.9828310012817383,
      "learning_rate": 9.162738190218424e-05,
      "loss": 3.0610742568969727,
      "memory(GiB)": 66.02,
      "step": 21810,
      "token_acc": 0.37755102040816324,
      "train_speed(iter/s)": 1.446132
    },
    {
      "epoch": 0.9346214815132171,
      "grad_norm": 3.7744083404541016,
      "learning_rate": 9.162365354410496e-05,
      "loss": 2.868472862243652,
      "memory(GiB)": 66.02,
      "step": 21815,
      "token_acc": 0.430635838150289,
      "train_speed(iter/s)": 1.446164
    },
    {
      "epoch": 0.934835696842466,
      "grad_norm": 3.9515891075134277,
      "learning_rate": 9.161992443197416e-05,
      "loss": 2.5441120147705076,
      "memory(GiB)": 66.02,
      "step": 21820,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.446156
    },
    {
      "epoch": 0.935049912171715,
      "grad_norm": 3.456465482711792,
      "learning_rate": 9.161619456585937e-05,
      "loss": 2.526171875,
      "memory(GiB)": 66.02,
      "step": 21825,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.446213
    },
    {
      "epoch": 0.935264127500964,
      "grad_norm": 3.599581718444824,
      "learning_rate": 9.161246394582818e-05,
      "loss": 2.8087135314941407,
      "memory(GiB)": 66.02,
      "step": 21830,
      "token_acc": 0.4310850439882698,
      "train_speed(iter/s)": 1.446215
    },
    {
      "epoch": 0.9354783428302129,
      "grad_norm": 5.116060733795166,
      "learning_rate": 9.160873257194818e-05,
      "loss": 2.4573360443115235,
      "memory(GiB)": 66.02,
      "step": 21835,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.446302
    },
    {
      "epoch": 0.9356925581594618,
      "grad_norm": 3.930436372756958,
      "learning_rate": 9.160500044428696e-05,
      "loss": 2.4878616333007812,
      "memory(GiB)": 66.02,
      "step": 21840,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.446285
    },
    {
      "epoch": 0.9359067734887109,
      "grad_norm": 5.1010284423828125,
      "learning_rate": 9.160126756291211e-05,
      "loss": 2.496533203125,
      "memory(GiB)": 66.02,
      "step": 21845,
      "token_acc": 0.44983818770226536,
      "train_speed(iter/s)": 1.44628
    },
    {
      "epoch": 0.9361209888179598,
      "grad_norm": 3.3701560497283936,
      "learning_rate": 9.15975339278913e-05,
      "loss": 2.364858627319336,
      "memory(GiB)": 66.02,
      "step": 21850,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.44625
    },
    {
      "epoch": 0.9363352041472087,
      "grad_norm": 3.9150311946868896,
      "learning_rate": 9.159379953929213e-05,
      "loss": 2.6046993255615236,
      "memory(GiB)": 66.02,
      "step": 21855,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.446238
    },
    {
      "epoch": 0.9365494194764578,
      "grad_norm": 4.447291851043701,
      "learning_rate": 9.159006439718226e-05,
      "loss": 2.5557540893554687,
      "memory(GiB)": 66.02,
      "step": 21860,
      "token_acc": 0.4895397489539749,
      "train_speed(iter/s)": 1.446226
    },
    {
      "epoch": 0.9367636348057067,
      "grad_norm": 4.724329948425293,
      "learning_rate": 9.158632850162935e-05,
      "loss": 2.7024940490722655,
      "memory(GiB)": 66.02,
      "step": 21865,
      "token_acc": 0.4527687296416938,
      "train_speed(iter/s)": 1.446314
    },
    {
      "epoch": 0.9369778501349556,
      "grad_norm": 4.627659797668457,
      "learning_rate": 9.158259185270108e-05,
      "loss": 2.3267560958862306,
      "memory(GiB)": 66.02,
      "step": 21870,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.446319
    },
    {
      "epoch": 0.9371920654642046,
      "grad_norm": 6.513284683227539,
      "learning_rate": 9.157885445046519e-05,
      "loss": 2.784957504272461,
      "memory(GiB)": 66.02,
      "step": 21875,
      "token_acc": 0.4378698224852071,
      "train_speed(iter/s)": 1.446313
    },
    {
      "epoch": 0.9374062807934536,
      "grad_norm": 4.638032913208008,
      "learning_rate": 9.157511629498932e-05,
      "loss": 2.5153589248657227,
      "memory(GiB)": 66.02,
      "step": 21880,
      "token_acc": 0.521551724137931,
      "train_speed(iter/s)": 1.446416
    },
    {
      "epoch": 0.9376204961227026,
      "grad_norm": 4.4747843742370605,
      "learning_rate": 9.157137738634122e-05,
      "loss": 3.0156442642211916,
      "memory(GiB)": 66.02,
      "step": 21885,
      "token_acc": 0.41134751773049644,
      "train_speed(iter/s)": 1.446529
    },
    {
      "epoch": 0.9378347114519515,
      "grad_norm": 4.044059753417969,
      "learning_rate": 9.156763772458862e-05,
      "loss": 2.6388324737548827,
      "memory(GiB)": 66.02,
      "step": 21890,
      "token_acc": 0.46905537459283386,
      "train_speed(iter/s)": 1.44663
    },
    {
      "epoch": 0.9380489267812004,
      "grad_norm": 5.029946327209473,
      "learning_rate": 9.156389730979928e-05,
      "loss": 2.612639617919922,
      "memory(GiB)": 66.02,
      "step": 21895,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.446742
    },
    {
      "epoch": 0.9382631421104495,
      "grad_norm": 5.445373058319092,
      "learning_rate": 9.156015614204094e-05,
      "loss": 2.3072505950927735,
      "memory(GiB)": 66.02,
      "step": 21900,
      "token_acc": 0.46987951807228917,
      "train_speed(iter/s)": 1.446735
    },
    {
      "epoch": 0.9384773574396984,
      "grad_norm": 5.270296573638916,
      "learning_rate": 9.155641422138139e-05,
      "loss": 2.8074317932128907,
      "memory(GiB)": 66.02,
      "step": 21905,
      "token_acc": 0.44107744107744107,
      "train_speed(iter/s)": 1.446748
    },
    {
      "epoch": 0.9386915727689473,
      "grad_norm": 3.0650320053100586,
      "learning_rate": 9.15526715478884e-05,
      "loss": 2.3497812271118166,
      "memory(GiB)": 66.02,
      "step": 21910,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.446781
    },
    {
      "epoch": 0.9389057880981964,
      "grad_norm": 5.169348239898682,
      "learning_rate": 9.15489281216298e-05,
      "loss": 2.7509584426879883,
      "memory(GiB)": 66.02,
      "step": 21915,
      "token_acc": 0.40794223826714804,
      "train_speed(iter/s)": 1.446736
    },
    {
      "epoch": 0.9391200034274453,
      "grad_norm": 5.620747089385986,
      "learning_rate": 9.154518394267338e-05,
      "loss": 2.5814395904541017,
      "memory(GiB)": 66.02,
      "step": 21920,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.446671
    },
    {
      "epoch": 0.9393342187566942,
      "grad_norm": 6.354038715362549,
      "learning_rate": 9.154143901108699e-05,
      "loss": 2.5864822387695314,
      "memory(GiB)": 66.02,
      "step": 21925,
      "token_acc": 0.46887966804979253,
      "train_speed(iter/s)": 1.446552
    },
    {
      "epoch": 0.9395484340859432,
      "grad_norm": 4.962425708770752,
      "learning_rate": 9.153769332693847e-05,
      "loss": 2.545681953430176,
      "memory(GiB)": 66.02,
      "step": 21930,
      "token_acc": 0.44485294117647056,
      "train_speed(iter/s)": 1.446519
    },
    {
      "epoch": 0.9397626494151922,
      "grad_norm": 5.091123580932617,
      "learning_rate": 9.153394689029566e-05,
      "loss": 2.5958648681640626,
      "memory(GiB)": 66.02,
      "step": 21935,
      "token_acc": 0.44483985765124556,
      "train_speed(iter/s)": 1.44649
    },
    {
      "epoch": 0.9399768647444411,
      "grad_norm": 4.598596572875977,
      "learning_rate": 9.153019970122643e-05,
      "loss": 2.4702308654785154,
      "memory(GiB)": 66.02,
      "step": 21940,
      "token_acc": 0.44964028776978415,
      "train_speed(iter/s)": 1.446473
    },
    {
      "epoch": 0.9401910800736901,
      "grad_norm": 5.330958843231201,
      "learning_rate": 9.15264517597987e-05,
      "loss": 2.559444808959961,
      "memory(GiB)": 66.02,
      "step": 21945,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.446481
    },
    {
      "epoch": 0.940405295402939,
      "grad_norm": 5.53353214263916,
      "learning_rate": 9.152270306608031e-05,
      "loss": 2.5792699813842774,
      "memory(GiB)": 66.02,
      "step": 21950,
      "token_acc": 0.4540229885057471,
      "train_speed(iter/s)": 1.446447
    },
    {
      "epoch": 0.940619510732188,
      "grad_norm": 3.1102919578552246,
      "learning_rate": 9.151895362013922e-05,
      "loss": 2.269911193847656,
      "memory(GiB)": 66.02,
      "step": 21955,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.446492
    },
    {
      "epoch": 0.940833726061437,
      "grad_norm": 4.358954906463623,
      "learning_rate": 9.151520342204334e-05,
      "loss": 2.7471052169799806,
      "memory(GiB)": 66.02,
      "step": 21960,
      "token_acc": 0.4570552147239264,
      "train_speed(iter/s)": 1.446576
    },
    {
      "epoch": 0.9410479413906859,
      "grad_norm": 3.5152010917663574,
      "learning_rate": 9.151145247186061e-05,
      "loss": 2.645145797729492,
      "memory(GiB)": 66.02,
      "step": 21965,
      "token_acc": 0.43934426229508194,
      "train_speed(iter/s)": 1.446525
    },
    {
      "epoch": 0.9412621567199349,
      "grad_norm": 3.773735284805298,
      "learning_rate": 9.150770076965895e-05,
      "loss": 2.8553417205810545,
      "memory(GiB)": 66.02,
      "step": 21970,
      "token_acc": 0.3977272727272727,
      "train_speed(iter/s)": 1.446521
    },
    {
      "epoch": 0.9414763720491839,
      "grad_norm": 3.603227376937866,
      "learning_rate": 9.15039483155064e-05,
      "loss": 2.641327667236328,
      "memory(GiB)": 66.02,
      "step": 21975,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.446526
    },
    {
      "epoch": 0.9416905873784328,
      "grad_norm": 4.119307518005371,
      "learning_rate": 9.150019510947086e-05,
      "loss": 2.4307092666625976,
      "memory(GiB)": 66.02,
      "step": 21980,
      "token_acc": 0.476038338658147,
      "train_speed(iter/s)": 1.446412
    },
    {
      "epoch": 0.9419048027076817,
      "grad_norm": 5.00947904586792,
      "learning_rate": 9.149644115162035e-05,
      "loss": 2.4045654296875,
      "memory(GiB)": 66.02,
      "step": 21985,
      "token_acc": 0.4823943661971831,
      "train_speed(iter/s)": 1.44629
    },
    {
      "epoch": 0.9421190180369308,
      "grad_norm": 4.431349277496338,
      "learning_rate": 9.149268644202289e-05,
      "loss": 2.8693355560302733,
      "memory(GiB)": 66.02,
      "step": 21990,
      "token_acc": 0.4368932038834951,
      "train_speed(iter/s)": 1.446285
    },
    {
      "epoch": 0.9423332333661797,
      "grad_norm": 4.919655799865723,
      "learning_rate": 9.148893098074649e-05,
      "loss": 2.2277074813842774,
      "memory(GiB)": 66.02,
      "step": 21995,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.446275
    },
    {
      "epoch": 0.9425474486954286,
      "grad_norm": 4.579711437225342,
      "learning_rate": 9.148517476785918e-05,
      "loss": 2.4744726181030274,
      "memory(GiB)": 66.02,
      "step": 22000,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.446332
    },
    {
      "epoch": 0.9425474486954286,
      "eval_loss": 2.200505256652832,
      "eval_runtime": 15.1137,
      "eval_samples_per_second": 6.617,
      "eval_steps_per_second": 6.617,
      "eval_token_acc": 0.46959896507115134,
      "step": 22000
    },
    {
      "epoch": 0.9427616640246776,
      "grad_norm": 3.9472317695617676,
      "learning_rate": 9.148141780342903e-05,
      "loss": 2.727242279052734,
      "memory(GiB)": 66.02,
      "step": 22005,
      "token_acc": 0.47702834799608995,
      "train_speed(iter/s)": 1.444886
    },
    {
      "epoch": 0.9429758793539266,
      "grad_norm": 7.040859699249268,
      "learning_rate": 9.147766008752407e-05,
      "loss": 2.6169862747192383,
      "memory(GiB)": 66.02,
      "step": 22010,
      "token_acc": 0.47653429602888087,
      "train_speed(iter/s)": 1.444895
    },
    {
      "epoch": 0.9431900946831755,
      "grad_norm": 5.663588523864746,
      "learning_rate": 9.14739016202124e-05,
      "loss": 3.0661840438842773,
      "memory(GiB)": 66.02,
      "step": 22015,
      "token_acc": 0.39365079365079364,
      "train_speed(iter/s)": 1.444907
    },
    {
      "epoch": 0.9434043100124245,
      "grad_norm": 4.596031665802002,
      "learning_rate": 9.14701424015621e-05,
      "loss": 2.503927230834961,
      "memory(GiB)": 66.02,
      "step": 22020,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.444992
    },
    {
      "epoch": 0.9436185253416735,
      "grad_norm": 4.302864074707031,
      "learning_rate": 9.146638243164125e-05,
      "loss": 2.3167238235473633,
      "memory(GiB)": 66.02,
      "step": 22025,
      "token_acc": 0.5040650406504065,
      "train_speed(iter/s)": 1.445031
    },
    {
      "epoch": 0.9438327406709224,
      "grad_norm": 3.596513032913208,
      "learning_rate": 9.1462621710518e-05,
      "loss": 2.5464473724365235,
      "memory(GiB)": 66.02,
      "step": 22030,
      "token_acc": 0.44074074074074077,
      "train_speed(iter/s)": 1.445038
    },
    {
      "epoch": 0.9440469560001714,
      "grad_norm": 4.43658447265625,
      "learning_rate": 9.145886023826044e-05,
      "loss": 2.6728584289550783,
      "memory(GiB)": 66.02,
      "step": 22035,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.445066
    },
    {
      "epoch": 0.9442611713294203,
      "grad_norm": 3.598688840866089,
      "learning_rate": 9.145509801493677e-05,
      "loss": 2.5031652450561523,
      "memory(GiB)": 66.02,
      "step": 22040,
      "token_acc": 0.44061302681992337,
      "train_speed(iter/s)": 1.445092
    },
    {
      "epoch": 0.9444753866586693,
      "grad_norm": 5.328486919403076,
      "learning_rate": 9.145133504061509e-05,
      "loss": 2.5981777191162108,
      "memory(GiB)": 66.02,
      "step": 22045,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.445123
    },
    {
      "epoch": 0.9446896019879183,
      "grad_norm": 4.462075710296631,
      "learning_rate": 9.14475713153636e-05,
      "loss": 2.7924291610717775,
      "memory(GiB)": 66.02,
      "step": 22050,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.445136
    },
    {
      "epoch": 0.9449038173171672,
      "grad_norm": 3.7059483528137207,
      "learning_rate": 9.144380683925044e-05,
      "loss": 2.7857032775878907,
      "memory(GiB)": 66.02,
      "step": 22055,
      "token_acc": 0.4758364312267658,
      "train_speed(iter/s)": 1.445128
    },
    {
      "epoch": 0.9451180326464161,
      "grad_norm": 4.5063276290893555,
      "learning_rate": 9.144004161234388e-05,
      "loss": 2.778232955932617,
      "memory(GiB)": 66.02,
      "step": 22060,
      "token_acc": 0.4548611111111111,
      "train_speed(iter/s)": 1.445168
    },
    {
      "epoch": 0.9453322479756652,
      "grad_norm": 5.610306262969971,
      "learning_rate": 9.143627563471209e-05,
      "loss": 2.6039690017700194,
      "memory(GiB)": 66.02,
      "step": 22065,
      "token_acc": 0.44876325088339225,
      "train_speed(iter/s)": 1.445047
    },
    {
      "epoch": 0.9455464633049141,
      "grad_norm": 3.989807605743408,
      "learning_rate": 9.143250890642327e-05,
      "loss": 2.428289031982422,
      "memory(GiB)": 66.02,
      "step": 22070,
      "token_acc": 0.5278810408921933,
      "train_speed(iter/s)": 1.44505
    },
    {
      "epoch": 0.945760678634163,
      "grad_norm": 4.690921306610107,
      "learning_rate": 9.142874142754572e-05,
      "loss": 2.592865562438965,
      "memory(GiB)": 66.02,
      "step": 22075,
      "token_acc": 0.4627831715210356,
      "train_speed(iter/s)": 1.44508
    },
    {
      "epoch": 0.9459748939634121,
      "grad_norm": 3.8750624656677246,
      "learning_rate": 9.142497319814764e-05,
      "loss": 2.4155059814453126,
      "memory(GiB)": 66.02,
      "step": 22080,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.445113
    },
    {
      "epoch": 0.946189109292661,
      "grad_norm": 5.399001598358154,
      "learning_rate": 9.142120421829729e-05,
      "loss": 2.9077178955078127,
      "memory(GiB)": 66.02,
      "step": 22085,
      "token_acc": 0.4295774647887324,
      "train_speed(iter/s)": 1.445153
    },
    {
      "epoch": 0.9464033246219099,
      "grad_norm": 4.626400470733643,
      "learning_rate": 9.141743448806301e-05,
      "loss": 2.6723724365234376,
      "memory(GiB)": 66.02,
      "step": 22090,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.445083
    },
    {
      "epoch": 0.9466175399511589,
      "grad_norm": 4.274816989898682,
      "learning_rate": 9.141366400751301e-05,
      "loss": 2.886031723022461,
      "memory(GiB)": 66.02,
      "step": 22095,
      "token_acc": 0.4578313253012048,
      "train_speed(iter/s)": 1.44517
    },
    {
      "epoch": 0.9468317552804079,
      "grad_norm": 5.222473621368408,
      "learning_rate": 9.140989277671567e-05,
      "loss": 2.282111930847168,
      "memory(GiB)": 66.02,
      "step": 22100,
      "token_acc": 0.5101351351351351,
      "train_speed(iter/s)": 1.44522
    },
    {
      "epoch": 0.9470459706096568,
      "grad_norm": 4.098766803741455,
      "learning_rate": 9.140612079573927e-05,
      "loss": 2.4320705413818358,
      "memory(GiB)": 66.02,
      "step": 22105,
      "token_acc": 0.5193798449612403,
      "train_speed(iter/s)": 1.445214
    },
    {
      "epoch": 0.9472601859389058,
      "grad_norm": 6.469631195068359,
      "learning_rate": 9.140234806465214e-05,
      "loss": 2.3476051330566405,
      "memory(GiB)": 66.02,
      "step": 22110,
      "token_acc": 0.5482456140350878,
      "train_speed(iter/s)": 1.445237
    },
    {
      "epoch": 0.9474744012681547,
      "grad_norm": 3.9835705757141113,
      "learning_rate": 9.139857458352263e-05,
      "loss": 2.6543209075927736,
      "memory(GiB)": 66.02,
      "step": 22115,
      "token_acc": 0.43506493506493504,
      "train_speed(iter/s)": 1.445175
    },
    {
      "epoch": 0.9476886165974037,
      "grad_norm": 5.18387508392334,
      "learning_rate": 9.139480035241912e-05,
      "loss": 2.807154083251953,
      "memory(GiB)": 66.02,
      "step": 22120,
      "token_acc": 0.41580756013745707,
      "train_speed(iter/s)": 1.445154
    },
    {
      "epoch": 0.9479028319266527,
      "grad_norm": 3.8181638717651367,
      "learning_rate": 9.139102537140996e-05,
      "loss": 2.4215265274047852,
      "memory(GiB)": 66.02,
      "step": 22125,
      "token_acc": 0.5152542372881356,
      "train_speed(iter/s)": 1.445089
    },
    {
      "epoch": 0.9481170472559016,
      "grad_norm": 5.602406024932861,
      "learning_rate": 9.138724964056355e-05,
      "loss": 2.3998691558837892,
      "memory(GiB)": 66.02,
      "step": 22130,
      "token_acc": 0.4568345323741007,
      "train_speed(iter/s)": 1.445098
    },
    {
      "epoch": 0.9483312625851505,
      "grad_norm": 3.582289934158325,
      "learning_rate": 9.13834731599483e-05,
      "loss": 2.4179683685302735,
      "memory(GiB)": 66.02,
      "step": 22135,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.445197
    },
    {
      "epoch": 0.9485454779143996,
      "grad_norm": 5.022667407989502,
      "learning_rate": 9.13796959296326e-05,
      "loss": 2.2809337615966796,
      "memory(GiB)": 66.02,
      "step": 22140,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.445135
    },
    {
      "epoch": 0.9487596932436485,
      "grad_norm": 4.084569931030273,
      "learning_rate": 9.137591794968489e-05,
      "loss": 2.864162826538086,
      "memory(GiB)": 66.02,
      "step": 22145,
      "token_acc": 0.44476744186046513,
      "train_speed(iter/s)": 1.445142
    },
    {
      "epoch": 0.9489739085728974,
      "grad_norm": 4.691052436828613,
      "learning_rate": 9.137213922017363e-05,
      "loss": 2.634572982788086,
      "memory(GiB)": 66.02,
      "step": 22150,
      "token_acc": 0.5458515283842795,
      "train_speed(iter/s)": 1.445185
    },
    {
      "epoch": 0.9491881239021465,
      "grad_norm": 4.898622512817383,
      "learning_rate": 9.136835974116724e-05,
      "loss": 2.829137992858887,
      "memory(GiB)": 66.02,
      "step": 22155,
      "token_acc": 0.47115384615384615,
      "train_speed(iter/s)": 1.445161
    },
    {
      "epoch": 0.9494023392313954,
      "grad_norm": 4.631893634796143,
      "learning_rate": 9.136457951273423e-05,
      "loss": 2.6244916915893555,
      "memory(GiB)": 66.02,
      "step": 22160,
      "token_acc": 0.4595375722543353,
      "train_speed(iter/s)": 1.445163
    },
    {
      "epoch": 0.9496165545606443,
      "grad_norm": 5.193511962890625,
      "learning_rate": 9.136079853494304e-05,
      "loss": 2.705002212524414,
      "memory(GiB)": 66.02,
      "step": 22165,
      "token_acc": 0.48249027237354086,
      "train_speed(iter/s)": 1.445204
    },
    {
      "epoch": 0.9498307698898933,
      "grad_norm": 4.0157880783081055,
      "learning_rate": 9.135701680786218e-05,
      "loss": 2.684200477600098,
      "memory(GiB)": 66.02,
      "step": 22170,
      "token_acc": 0.4194528875379939,
      "train_speed(iter/s)": 1.445195
    },
    {
      "epoch": 0.9500449852191423,
      "grad_norm": 5.729152679443359,
      "learning_rate": 9.135323433156018e-05,
      "loss": 2.731343460083008,
      "memory(GiB)": 66.02,
      "step": 22175,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.445248
    },
    {
      "epoch": 0.9502592005483912,
      "grad_norm": 5.193756580352783,
      "learning_rate": 9.134945110610554e-05,
      "loss": 2.280427169799805,
      "memory(GiB)": 66.02,
      "step": 22180,
      "token_acc": 0.5175438596491229,
      "train_speed(iter/s)": 1.445218
    },
    {
      "epoch": 0.9504734158776402,
      "grad_norm": 4.542730331420898,
      "learning_rate": 9.134566713156679e-05,
      "loss": 2.470453453063965,
      "memory(GiB)": 66.02,
      "step": 22185,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.445295
    },
    {
      "epoch": 0.9506876312068891,
      "grad_norm": 4.4969329833984375,
      "learning_rate": 9.134188240801251e-05,
      "loss": 2.6251392364501953,
      "memory(GiB)": 66.02,
      "step": 22190,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.445361
    },
    {
      "epoch": 0.9509018465361381,
      "grad_norm": 14.010857582092285,
      "learning_rate": 9.133809693551125e-05,
      "loss": 2.4005346298217773,
      "memory(GiB)": 66.02,
      "step": 22195,
      "token_acc": 0.46886446886446886,
      "train_speed(iter/s)": 1.445393
    },
    {
      "epoch": 0.9511160618653871,
      "grad_norm": 3.6862452030181885,
      "learning_rate": 9.133431071413158e-05,
      "loss": 2.5876438140869142,
      "memory(GiB)": 66.02,
      "step": 22200,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.445356
    },
    {
      "epoch": 0.951330277194636,
      "grad_norm": 4.559037685394287,
      "learning_rate": 9.13305237439421e-05,
      "loss": 2.7423818588256834,
      "memory(GiB)": 66.02,
      "step": 22205,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.445339
    },
    {
      "epoch": 0.951544492523885,
      "grad_norm": 5.095754623413086,
      "learning_rate": 9.13267360250114e-05,
      "loss": 2.498956871032715,
      "memory(GiB)": 66.02,
      "step": 22210,
      "token_acc": 0.47854785478547857,
      "train_speed(iter/s)": 1.445377
    },
    {
      "epoch": 0.951758707853134,
      "grad_norm": 4.862742900848389,
      "learning_rate": 9.132294755740814e-05,
      "loss": 2.6450281143188477,
      "memory(GiB)": 66.02,
      "step": 22215,
      "token_acc": 0.4580152671755725,
      "train_speed(iter/s)": 1.445392
    },
    {
      "epoch": 0.9519729231823829,
      "grad_norm": 5.515480995178223,
      "learning_rate": 9.131915834120088e-05,
      "loss": 2.463967514038086,
      "memory(GiB)": 66.02,
      "step": 22220,
      "token_acc": 0.47843137254901963,
      "train_speed(iter/s)": 1.445391
    },
    {
      "epoch": 0.952187138511632,
      "grad_norm": 4.148614883422852,
      "learning_rate": 9.131536837645833e-05,
      "loss": 2.6365741729736327,
      "memory(GiB)": 66.02,
      "step": 22225,
      "token_acc": 0.4367469879518072,
      "train_speed(iter/s)": 1.445428
    },
    {
      "epoch": 0.9524013538408809,
      "grad_norm": 4.757924556732178,
      "learning_rate": 9.131157766324912e-05,
      "loss": 2.507137107849121,
      "memory(GiB)": 66.02,
      "step": 22230,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.445467
    },
    {
      "epoch": 0.9526155691701298,
      "grad_norm": 4.670322895050049,
      "learning_rate": 9.130778620164193e-05,
      "loss": 2.5537620544433595,
      "memory(GiB)": 66.02,
      "step": 22235,
      "token_acc": 0.4395973154362416,
      "train_speed(iter/s)": 1.445559
    },
    {
      "epoch": 0.9528297844993788,
      "grad_norm": 4.049413204193115,
      "learning_rate": 9.130399399170544e-05,
      "loss": 2.618539237976074,
      "memory(GiB)": 66.02,
      "step": 22240,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.445635
    },
    {
      "epoch": 0.9530439998286278,
      "grad_norm": 5.538731098175049,
      "learning_rate": 9.130020103350836e-05,
      "loss": 2.821795654296875,
      "memory(GiB)": 66.02,
      "step": 22245,
      "token_acc": 0.42592592592592593,
      "train_speed(iter/s)": 1.445661
    },
    {
      "epoch": 0.9532582151578767,
      "grad_norm": 4.345365047454834,
      "learning_rate": 9.12964073271194e-05,
      "loss": 2.3884326934814455,
      "memory(GiB)": 66.02,
      "step": 22250,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.445637
    },
    {
      "epoch": 0.9534724304871257,
      "grad_norm": 4.030776500701904,
      "learning_rate": 9.129261287260726e-05,
      "loss": 2.64312686920166,
      "memory(GiB)": 66.02,
      "step": 22255,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.44562
    },
    {
      "epoch": 0.9536866458163746,
      "grad_norm": 5.273192405700684,
      "learning_rate": 9.128881767004072e-05,
      "loss": 2.433702278137207,
      "memory(GiB)": 69.34,
      "step": 22260,
      "token_acc": 0.5107296137339056,
      "train_speed(iter/s)": 1.445562
    },
    {
      "epoch": 0.9539008611456236,
      "grad_norm": 3.850247383117676,
      "learning_rate": 9.12850217194885e-05,
      "loss": 2.754536819458008,
      "memory(GiB)": 69.34,
      "step": 22265,
      "token_acc": 0.4146341463414634,
      "train_speed(iter/s)": 1.44558
    },
    {
      "epoch": 0.9541150764748726,
      "grad_norm": 5.440518856048584,
      "learning_rate": 9.12812250210194e-05,
      "loss": 2.704586410522461,
      "memory(GiB)": 69.34,
      "step": 22270,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.445585
    },
    {
      "epoch": 0.9543292918041215,
      "grad_norm": 3.62768816947937,
      "learning_rate": 9.127742757470217e-05,
      "loss": 2.175482749938965,
      "memory(GiB)": 69.34,
      "step": 22275,
      "token_acc": 0.5115384615384615,
      "train_speed(iter/s)": 1.445628
    },
    {
      "epoch": 0.9545435071333704,
      "grad_norm": 4.364960193634033,
      "learning_rate": 9.127362938060563e-05,
      "loss": 2.784740447998047,
      "memory(GiB)": 69.34,
      "step": 22280,
      "token_acc": 0.4266666666666667,
      "train_speed(iter/s)": 1.445673
    },
    {
      "epoch": 0.9547577224626195,
      "grad_norm": 3.407343864440918,
      "learning_rate": 9.126983043879857e-05,
      "loss": 2.602976608276367,
      "memory(GiB)": 69.34,
      "step": 22285,
      "token_acc": 0.4887459807073955,
      "train_speed(iter/s)": 1.445699
    },
    {
      "epoch": 0.9549719377918684,
      "grad_norm": 4.121309280395508,
      "learning_rate": 9.126603074934982e-05,
      "loss": 2.6240062713623047,
      "memory(GiB)": 69.34,
      "step": 22290,
      "token_acc": 0.4453125,
      "train_speed(iter/s)": 1.44571
    },
    {
      "epoch": 0.9551861531211173,
      "grad_norm": 5.251585960388184,
      "learning_rate": 9.126223031232822e-05,
      "loss": 2.4973318099975588,
      "memory(GiB)": 69.34,
      "step": 22295,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.445701
    },
    {
      "epoch": 0.9554003684503664,
      "grad_norm": 3.9109556674957275,
      "learning_rate": 9.125842912780259e-05,
      "loss": 2.4879005432128904,
      "memory(GiB)": 69.34,
      "step": 22300,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.445701
    },
    {
      "epoch": 0.9556145837796153,
      "grad_norm": 4.672306537628174,
      "learning_rate": 9.125462719584183e-05,
      "loss": 2.45849723815918,
      "memory(GiB)": 69.34,
      "step": 22305,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.445789
    },
    {
      "epoch": 0.9558287991088642,
      "grad_norm": 5.2371134757995605,
      "learning_rate": 9.125082451651479e-05,
      "loss": 2.801739311218262,
      "memory(GiB)": 69.34,
      "step": 22310,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.445817
    },
    {
      "epoch": 0.9560430144381132,
      "grad_norm": 4.777816295623779,
      "learning_rate": 9.124702108989036e-05,
      "loss": 2.7504138946533203,
      "memory(GiB)": 69.34,
      "step": 22315,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.445853
    },
    {
      "epoch": 0.9562572297673622,
      "grad_norm": 4.661057472229004,
      "learning_rate": 9.124321691603747e-05,
      "loss": 2.6545795440673827,
      "memory(GiB)": 69.34,
      "step": 22320,
      "token_acc": 0.4830188679245283,
      "train_speed(iter/s)": 1.445885
    },
    {
      "epoch": 0.9564714450966111,
      "grad_norm": 4.55148983001709,
      "learning_rate": 9.123941199502501e-05,
      "loss": 2.667533302307129,
      "memory(GiB)": 69.34,
      "step": 22325,
      "token_acc": 0.4143835616438356,
      "train_speed(iter/s)": 1.445795
    },
    {
      "epoch": 0.9566856604258601,
      "grad_norm": 3.683835983276367,
      "learning_rate": 9.12356063269219e-05,
      "loss": 2.6578132629394533,
      "memory(GiB)": 69.34,
      "step": 22330,
      "token_acc": 0.4316109422492401,
      "train_speed(iter/s)": 1.445744
    },
    {
      "epoch": 0.956899875755109,
      "grad_norm": 5.03511381149292,
      "learning_rate": 9.123179991179711e-05,
      "loss": 2.63839111328125,
      "memory(GiB)": 69.34,
      "step": 22335,
      "token_acc": 0.4630225080385852,
      "train_speed(iter/s)": 1.445709
    },
    {
      "epoch": 0.957114091084358,
      "grad_norm": 5.707028388977051,
      "learning_rate": 9.122799274971959e-05,
      "loss": 2.4655242919921876,
      "memory(GiB)": 69.34,
      "step": 22340,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.445644
    },
    {
      "epoch": 0.957328306413607,
      "grad_norm": 5.44482421875,
      "learning_rate": 9.12241848407583e-05,
      "loss": 2.7348758697509767,
      "memory(GiB)": 69.34,
      "step": 22345,
      "token_acc": 0.42953020134228187,
      "train_speed(iter/s)": 1.44569
    },
    {
      "epoch": 0.9575425217428559,
      "grad_norm": 5.052228927612305,
      "learning_rate": 9.122037618498225e-05,
      "loss": 2.2934436798095703,
      "memory(GiB)": 69.34,
      "step": 22350,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.445688
    },
    {
      "epoch": 0.9577567370721048,
      "grad_norm": 4.518186092376709,
      "learning_rate": 9.12165667824604e-05,
      "loss": 2.6791284561157225,
      "memory(GiB)": 69.34,
      "step": 22355,
      "token_acc": 0.4451219512195122,
      "train_speed(iter/s)": 1.445751
    },
    {
      "epoch": 0.9579709524013539,
      "grad_norm": 3.880202054977417,
      "learning_rate": 9.121275663326178e-05,
      "loss": 2.5310874938964845,
      "memory(GiB)": 69.34,
      "step": 22360,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.445707
    },
    {
      "epoch": 0.9581851677306028,
      "grad_norm": 4.373251438140869,
      "learning_rate": 9.120894573745542e-05,
      "loss": 2.5756412506103517,
      "memory(GiB)": 69.34,
      "step": 22365,
      "token_acc": 0.44339622641509435,
      "train_speed(iter/s)": 1.445716
    },
    {
      "epoch": 0.9583993830598517,
      "grad_norm": 4.539257049560547,
      "learning_rate": 9.120513409511033e-05,
      "loss": 2.525067138671875,
      "memory(GiB)": 69.34,
      "step": 22370,
      "token_acc": 0.5210727969348659,
      "train_speed(iter/s)": 1.445738
    },
    {
      "epoch": 0.9586135983891008,
      "grad_norm": 4.783735275268555,
      "learning_rate": 9.12013217062956e-05,
      "loss": 2.5327564239501954,
      "memory(GiB)": 69.34,
      "step": 22375,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.445749
    },
    {
      "epoch": 0.9588278137183497,
      "grad_norm": 4.265169620513916,
      "learning_rate": 9.119750857108027e-05,
      "loss": 2.7773130416870115,
      "memory(GiB)": 69.34,
      "step": 22380,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.445732
    },
    {
      "epoch": 0.9590420290475986,
      "grad_norm": 5.149278163909912,
      "learning_rate": 9.119369468953344e-05,
      "loss": 2.4701839447021485,
      "memory(GiB)": 75.22,
      "step": 22385,
      "token_acc": 0.47019867549668876,
      "train_speed(iter/s)": 1.445693
    },
    {
      "epoch": 0.9592562443768476,
      "grad_norm": 3.8407633304595947,
      "learning_rate": 9.118988006172418e-05,
      "loss": 2.675972747802734,
      "memory(GiB)": 75.22,
      "step": 22390,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.445635
    },
    {
      "epoch": 0.9594704597060966,
      "grad_norm": 5.794604301452637,
      "learning_rate": 9.11860646877216e-05,
      "loss": 2.8507226943969726,
      "memory(GiB)": 75.22,
      "step": 22395,
      "token_acc": 0.4098939929328622,
      "train_speed(iter/s)": 1.445606
    },
    {
      "epoch": 0.9596846750353455,
      "grad_norm": 4.639246463775635,
      "learning_rate": 9.118224856759482e-05,
      "loss": 2.7039623260498047,
      "memory(GiB)": 75.22,
      "step": 22400,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.4456
    },
    {
      "epoch": 0.9598988903645945,
      "grad_norm": 4.491499423980713,
      "learning_rate": 9.117843170141297e-05,
      "loss": 2.843472480773926,
      "memory(GiB)": 75.22,
      "step": 22405,
      "token_acc": 0.4267515923566879,
      "train_speed(iter/s)": 1.445542
    },
    {
      "epoch": 0.9601131056938434,
      "grad_norm": 4.32130241394043,
      "learning_rate": 9.117461408924521e-05,
      "loss": 2.566757392883301,
      "memory(GiB)": 75.22,
      "step": 22410,
      "token_acc": 0.47038327526132406,
      "train_speed(iter/s)": 1.445491
    },
    {
      "epoch": 0.9603273210230924,
      "grad_norm": 4.164856910705566,
      "learning_rate": 9.11707957311607e-05,
      "loss": 2.1697620391845702,
      "memory(GiB)": 75.22,
      "step": 22415,
      "token_acc": 0.5335570469798657,
      "train_speed(iter/s)": 1.445409
    },
    {
      "epoch": 0.9605415363523414,
      "grad_norm": 3.553744077682495,
      "learning_rate": 9.116697662722859e-05,
      "loss": 2.7220294952392576,
      "memory(GiB)": 75.22,
      "step": 22420,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.44548
    },
    {
      "epoch": 0.9607557516815903,
      "grad_norm": 4.6254191398620605,
      "learning_rate": 9.116315677751807e-05,
      "loss": 2.646860122680664,
      "memory(GiB)": 75.22,
      "step": 22425,
      "token_acc": 0.478125,
      "train_speed(iter/s)": 1.445515
    },
    {
      "epoch": 0.9609699670108393,
      "grad_norm": 3.833604335784912,
      "learning_rate": 9.115933618209838e-05,
      "loss": 2.6391712188720704,
      "memory(GiB)": 75.22,
      "step": 22430,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.445509
    },
    {
      "epoch": 0.9611841823400883,
      "grad_norm": 3.3966407775878906,
      "learning_rate": 9.115551484103869e-05,
      "loss": 2.586381721496582,
      "memory(GiB)": 75.22,
      "step": 22435,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.445533
    },
    {
      "epoch": 0.9613983976693372,
      "grad_norm": 5.000818252563477,
      "learning_rate": 9.115169275440825e-05,
      "loss": 2.6876317977905275,
      "memory(GiB)": 75.22,
      "step": 22440,
      "token_acc": 0.4844961240310077,
      "train_speed(iter/s)": 1.445543
    },
    {
      "epoch": 0.9616126129985861,
      "grad_norm": 3.923110246658325,
      "learning_rate": 9.114786992227629e-05,
      "loss": 2.6653194427490234,
      "memory(GiB)": 75.22,
      "step": 22445,
      "token_acc": 0.4608150470219436,
      "train_speed(iter/s)": 1.445546
    },
    {
      "epoch": 0.9618268283278352,
      "grad_norm": 3.779022455215454,
      "learning_rate": 9.114404634471205e-05,
      "loss": 2.6949960708618166,
      "memory(GiB)": 75.22,
      "step": 22450,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.445523
    },
    {
      "epoch": 0.9620410436570841,
      "grad_norm": 4.776429176330566,
      "learning_rate": 9.114022202178483e-05,
      "loss": 2.4824108123779296,
      "memory(GiB)": 75.22,
      "step": 22455,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.44558
    },
    {
      "epoch": 0.962255258986333,
      "grad_norm": 5.741626739501953,
      "learning_rate": 9.113639695356388e-05,
      "loss": 2.4363405227661135,
      "memory(GiB)": 75.22,
      "step": 22460,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.445542
    },
    {
      "epoch": 0.962469474315582,
      "grad_norm": 4.263327121734619,
      "learning_rate": 9.113257114011852e-05,
      "loss": 2.2999876022338865,
      "memory(GiB)": 75.22,
      "step": 22465,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.44548
    },
    {
      "epoch": 0.962683689644831,
      "grad_norm": 5.492435932159424,
      "learning_rate": 9.112874458151805e-05,
      "loss": 2.3961204528808593,
      "memory(GiB)": 75.22,
      "step": 22470,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.445533
    },
    {
      "epoch": 0.9628979049740799,
      "grad_norm": 4.865903854370117,
      "learning_rate": 9.112491727783179e-05,
      "loss": 2.5525184631347657,
      "memory(GiB)": 75.22,
      "step": 22475,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.44555
    },
    {
      "epoch": 0.9631121203033289,
      "grad_norm": 4.546634197235107,
      "learning_rate": 9.112108922912907e-05,
      "loss": 2.5844451904296877,
      "memory(GiB)": 75.22,
      "step": 22480,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.445618
    },
    {
      "epoch": 0.9633263356325779,
      "grad_norm": 4.71919584274292,
      "learning_rate": 9.111726043547926e-05,
      "loss": 2.9415866851806642,
      "memory(GiB)": 75.22,
      "step": 22485,
      "token_acc": 0.4155844155844156,
      "train_speed(iter/s)": 1.44567
    },
    {
      "epoch": 0.9635405509618268,
      "grad_norm": 5.9818034172058105,
      "learning_rate": 9.111343089695168e-05,
      "loss": 2.68177490234375,
      "memory(GiB)": 75.22,
      "step": 22490,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.44573
    },
    {
      "epoch": 0.9637547662910758,
      "grad_norm": 5.77147102355957,
      "learning_rate": 9.110960061361575e-05,
      "loss": 2.570528984069824,
      "memory(GiB)": 75.22,
      "step": 22495,
      "token_acc": 0.525,
      "train_speed(iter/s)": 1.445747
    },
    {
      "epoch": 0.9639689816203247,
      "grad_norm": 7.428699016571045,
      "learning_rate": 9.110576958554085e-05,
      "loss": 2.5904220581054687,
      "memory(GiB)": 77.56,
      "step": 22500,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.445607
    },
    {
      "epoch": 0.9639689816203247,
      "eval_loss": 2.4378061294555664,
      "eval_runtime": 12.8503,
      "eval_samples_per_second": 7.782,
      "eval_steps_per_second": 7.782,
      "eval_token_acc": 0.4587280108254398,
      "step": 22500
    },
    {
      "epoch": 0.9641831969495737,
      "grad_norm": 4.519054412841797,
      "learning_rate": 9.110193781279635e-05,
      "loss": 3.037515640258789,
      "memory(GiB)": 77.56,
      "step": 22505,
      "token_acc": 0.45545545545545546,
      "train_speed(iter/s)": 1.444433
    },
    {
      "epoch": 0.9643974122788227,
      "grad_norm": 4.015757083892822,
      "learning_rate": 9.109810529545171e-05,
      "loss": 2.4090030670166014,
      "memory(GiB)": 77.56,
      "step": 22510,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.44448
    },
    {
      "epoch": 0.9646116276080716,
      "grad_norm": 3.304832935333252,
      "learning_rate": 9.109427203357632e-05,
      "loss": 2.8798879623413085,
      "memory(GiB)": 77.56,
      "step": 22515,
      "token_acc": 0.4437869822485207,
      "train_speed(iter/s)": 1.444493
    },
    {
      "epoch": 0.9648258429373205,
      "grad_norm": 4.7493133544921875,
      "learning_rate": 9.109043802723967e-05,
      "loss": 2.4731964111328124,
      "memory(GiB)": 77.56,
      "step": 22520,
      "token_acc": 0.47924528301886793,
      "train_speed(iter/s)": 1.44452
    },
    {
      "epoch": 0.9650400582665696,
      "grad_norm": 4.440759658813477,
      "learning_rate": 9.108660327651116e-05,
      "loss": 2.348245620727539,
      "memory(GiB)": 77.56,
      "step": 22525,
      "token_acc": 0.5313653136531366,
      "train_speed(iter/s)": 1.444468
    },
    {
      "epoch": 0.9652542735958185,
      "grad_norm": 4.804719924926758,
      "learning_rate": 9.10827677814603e-05,
      "loss": 2.5432621002197267,
      "memory(GiB)": 77.56,
      "step": 22530,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.444469
    },
    {
      "epoch": 0.9654684889250674,
      "grad_norm": 4.316156387329102,
      "learning_rate": 9.107893154215656e-05,
      "loss": 2.8198347091674805,
      "memory(GiB)": 77.56,
      "step": 22535,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.44457
    },
    {
      "epoch": 0.9656827042543165,
      "grad_norm": 5.392191410064697,
      "learning_rate": 9.107509455866945e-05,
      "loss": 2.413319778442383,
      "memory(GiB)": 77.56,
      "step": 22540,
      "token_acc": 0.5310077519379846,
      "train_speed(iter/s)": 1.444569
    },
    {
      "epoch": 0.9658969195835654,
      "grad_norm": 3.2190346717834473,
      "learning_rate": 9.107125683106848e-05,
      "loss": 2.6461687088012695,
      "memory(GiB)": 77.56,
      "step": 22545,
      "token_acc": 0.4584450402144772,
      "train_speed(iter/s)": 1.444529
    },
    {
      "epoch": 0.9661111349128143,
      "grad_norm": 4.319345474243164,
      "learning_rate": 9.106741835942314e-05,
      "loss": 2.6277385711669923,
      "memory(GiB)": 77.56,
      "step": 22550,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.444508
    },
    {
      "epoch": 0.9663253502420633,
      "grad_norm": 5.934285640716553,
      "learning_rate": 9.106357914380299e-05,
      "loss": 2.2906621932983398,
      "memory(GiB)": 77.56,
      "step": 22555,
      "token_acc": 0.5222672064777328,
      "train_speed(iter/s)": 1.444561
    },
    {
      "epoch": 0.9665395655713123,
      "grad_norm": 4.374237060546875,
      "learning_rate": 9.105973918427759e-05,
      "loss": 2.664159393310547,
      "memory(GiB)": 77.56,
      "step": 22560,
      "token_acc": 0.4186851211072664,
      "train_speed(iter/s)": 1.444569
    },
    {
      "epoch": 0.9667537809005613,
      "grad_norm": 3.712407350540161,
      "learning_rate": 9.105589848091651e-05,
      "loss": 2.111763763427734,
      "memory(GiB)": 77.56,
      "step": 22565,
      "token_acc": 0.5313531353135313,
      "train_speed(iter/s)": 1.444573
    },
    {
      "epoch": 0.9669679962298102,
      "grad_norm": 4.662269115447998,
      "learning_rate": 9.105205703378931e-05,
      "loss": 2.396389961242676,
      "memory(GiB)": 77.56,
      "step": 22570,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.444454
    },
    {
      "epoch": 0.9671822115590591,
      "grad_norm": 4.030180931091309,
      "learning_rate": 9.104821484296559e-05,
      "loss": 2.530960464477539,
      "memory(GiB)": 77.56,
      "step": 22575,
      "token_acc": 0.4953560371517028,
      "train_speed(iter/s)": 1.44452
    },
    {
      "epoch": 0.9673964268883082,
      "grad_norm": 5.760007858276367,
      "learning_rate": 9.104437190851493e-05,
      "loss": 2.733035659790039,
      "memory(GiB)": 77.56,
      "step": 22580,
      "token_acc": 0.4379310344827586,
      "train_speed(iter/s)": 1.444507
    },
    {
      "epoch": 0.9676106422175571,
      "grad_norm": 3.886164426803589,
      "learning_rate": 9.104052823050699e-05,
      "loss": 2.715729331970215,
      "memory(GiB)": 77.56,
      "step": 22585,
      "token_acc": 0.45180722891566266,
      "train_speed(iter/s)": 1.444442
    },
    {
      "epoch": 0.967824857546806,
      "grad_norm": 4.08539342880249,
      "learning_rate": 9.103668380901138e-05,
      "loss": 2.5099754333496094,
      "memory(GiB)": 77.56,
      "step": 22590,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.444443
    },
    {
      "epoch": 0.9680390728760551,
      "grad_norm": 4.854295253753662,
      "learning_rate": 9.103283864409775e-05,
      "loss": 2.5760751724243165,
      "memory(GiB)": 77.56,
      "step": 22595,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.444479
    },
    {
      "epoch": 0.968253288205304,
      "grad_norm": 3.448655605316162,
      "learning_rate": 9.102899273583575e-05,
      "loss": 2.6189884185791015,
      "memory(GiB)": 77.56,
      "step": 22600,
      "token_acc": 0.44660194174757284,
      "train_speed(iter/s)": 1.444528
    },
    {
      "epoch": 0.9684675035345529,
      "grad_norm": 5.7773003578186035,
      "learning_rate": 9.102514608429507e-05,
      "loss": 2.601335334777832,
      "memory(GiB)": 77.56,
      "step": 22605,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.44453
    },
    {
      "epoch": 0.9686817188638019,
      "grad_norm": 5.030820846557617,
      "learning_rate": 9.102129868954537e-05,
      "loss": 2.4951377868652345,
      "memory(GiB)": 77.56,
      "step": 22610,
      "token_acc": 0.45054945054945056,
      "train_speed(iter/s)": 1.44451
    },
    {
      "epoch": 0.9688959341930509,
      "grad_norm": 4.682309150695801,
      "learning_rate": 9.101745055165635e-05,
      "loss": 2.450288009643555,
      "memory(GiB)": 77.56,
      "step": 22615,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.444423
    },
    {
      "epoch": 0.9691101495222998,
      "grad_norm": 3.8503715991973877,
      "learning_rate": 9.101360167069777e-05,
      "loss": 2.4555179595947267,
      "memory(GiB)": 77.56,
      "step": 22620,
      "token_acc": 0.49508196721311476,
      "train_speed(iter/s)": 1.444385
    },
    {
      "epoch": 0.9693243648515488,
      "grad_norm": 3.836268186569214,
      "learning_rate": 9.100975204673929e-05,
      "loss": 2.758441925048828,
      "memory(GiB)": 77.56,
      "step": 22625,
      "token_acc": 0.4748427672955975,
      "train_speed(iter/s)": 1.444451
    },
    {
      "epoch": 0.9695385801807977,
      "grad_norm": 4.451447486877441,
      "learning_rate": 9.10059016798507e-05,
      "loss": 2.996836853027344,
      "memory(GiB)": 77.56,
      "step": 22630,
      "token_acc": 0.43776824034334766,
      "train_speed(iter/s)": 1.444482
    },
    {
      "epoch": 0.9697527955100467,
      "grad_norm": 4.166262626647949,
      "learning_rate": 9.100205057010174e-05,
      "loss": 2.6655994415283204,
      "memory(GiB)": 77.56,
      "step": 22635,
      "token_acc": 0.44370860927152317,
      "train_speed(iter/s)": 1.444527
    },
    {
      "epoch": 0.9699670108392957,
      "grad_norm": 4.432879447937012,
      "learning_rate": 9.099819871756215e-05,
      "loss": 2.710140609741211,
      "memory(GiB)": 77.56,
      "step": 22640,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.444564
    },
    {
      "epoch": 0.9701812261685446,
      "grad_norm": 4.089960098266602,
      "learning_rate": 9.099434612230175e-05,
      "loss": 2.690087890625,
      "memory(GiB)": 77.56,
      "step": 22645,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.444639
    },
    {
      "epoch": 0.9703954414977936,
      "grad_norm": 4.260512351989746,
      "learning_rate": 9.099049278439029e-05,
      "loss": 2.7790201187133787,
      "memory(GiB)": 77.56,
      "step": 22650,
      "token_acc": 0.4153846153846154,
      "train_speed(iter/s)": 1.44467
    },
    {
      "epoch": 0.9706096568270426,
      "grad_norm": 6.258110523223877,
      "learning_rate": 9.098663870389763e-05,
      "loss": 2.717209053039551,
      "memory(GiB)": 77.56,
      "step": 22655,
      "token_acc": 0.41739130434782606,
      "train_speed(iter/s)": 1.444758
    },
    {
      "epoch": 0.9708238721562915,
      "grad_norm": 4.986264705657959,
      "learning_rate": 9.098278388089354e-05,
      "loss": 2.4684144973754885,
      "memory(GiB)": 77.56,
      "step": 22660,
      "token_acc": 0.4734848484848485,
      "train_speed(iter/s)": 1.444767
    },
    {
      "epoch": 0.9710380874855404,
      "grad_norm": 5.7784247398376465,
      "learning_rate": 9.097892831544789e-05,
      "loss": 2.8189533233642576,
      "memory(GiB)": 77.56,
      "step": 22665,
      "token_acc": 0.46785714285714286,
      "train_speed(iter/s)": 1.444787
    },
    {
      "epoch": 0.9712523028147895,
      "grad_norm": 4.401434421539307,
      "learning_rate": 9.097507200763052e-05,
      "loss": 2.5944896697998048,
      "memory(GiB)": 77.56,
      "step": 22670,
      "token_acc": 0.4391891891891892,
      "train_speed(iter/s)": 1.444849
    },
    {
      "epoch": 0.9714665181440384,
      "grad_norm": 3.987140655517578,
      "learning_rate": 9.097121495751126e-05,
      "loss": 2.5924224853515625,
      "memory(GiB)": 77.56,
      "step": 22675,
      "token_acc": 0.49612403100775193,
      "train_speed(iter/s)": 1.444811
    },
    {
      "epoch": 0.9716807334732873,
      "grad_norm": 5.120415687561035,
      "learning_rate": 9.096735716516001e-05,
      "loss": 2.543726348876953,
      "memory(GiB)": 77.56,
      "step": 22680,
      "token_acc": 0.4434250764525994,
      "train_speed(iter/s)": 1.444818
    },
    {
      "epoch": 0.9718949488025364,
      "grad_norm": 4.902359485626221,
      "learning_rate": 9.096349863064666e-05,
      "loss": 2.5126115798950197,
      "memory(GiB)": 77.56,
      "step": 22685,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.444883
    },
    {
      "epoch": 0.9721091641317853,
      "grad_norm": 5.709869384765625,
      "learning_rate": 9.09596393540411e-05,
      "loss": 2.651871109008789,
      "memory(GiB)": 77.56,
      "step": 22690,
      "token_acc": 0.4383561643835616,
      "train_speed(iter/s)": 1.444957
    },
    {
      "epoch": 0.9723233794610342,
      "grad_norm": 4.62658166885376,
      "learning_rate": 9.095577933541326e-05,
      "loss": 2.5991443634033202,
      "memory(GiB)": 77.56,
      "step": 22695,
      "token_acc": 0.4778481012658228,
      "train_speed(iter/s)": 1.444979
    },
    {
      "epoch": 0.9725375947902832,
      "grad_norm": 7.195178031921387,
      "learning_rate": 9.095191857483305e-05,
      "loss": 2.71396484375,
      "memory(GiB)": 77.56,
      "step": 22700,
      "token_acc": 0.45112781954887216,
      "train_speed(iter/s)": 1.444978
    },
    {
      "epoch": 0.9727518101195322,
      "grad_norm": 9.005720138549805,
      "learning_rate": 9.094805707237041e-05,
      "loss": 2.6929790496826174,
      "memory(GiB)": 77.56,
      "step": 22705,
      "token_acc": 0.4573170731707317,
      "train_speed(iter/s)": 1.444991
    },
    {
      "epoch": 0.9729660254487811,
      "grad_norm": 3.981722593307495,
      "learning_rate": 9.094419482809534e-05,
      "loss": 2.5417993545532225,
      "memory(GiB)": 77.56,
      "step": 22710,
      "token_acc": 0.4695121951219512,
      "train_speed(iter/s)": 1.445001
    },
    {
      "epoch": 0.9731802407780301,
      "grad_norm": 4.128677845001221,
      "learning_rate": 9.094033184207774e-05,
      "loss": 2.4229068756103516,
      "memory(GiB)": 77.56,
      "step": 22715,
      "token_acc": 0.4908424908424908,
      "train_speed(iter/s)": 1.44502
    },
    {
      "epoch": 0.973394456107279,
      "grad_norm": 3.5569841861724854,
      "learning_rate": 9.093646811438762e-05,
      "loss": 2.188077926635742,
      "memory(GiB)": 77.56,
      "step": 22720,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.44503
    },
    {
      "epoch": 0.973608671436528,
      "grad_norm": 4.370412826538086,
      "learning_rate": 9.0932603645095e-05,
      "loss": 2.7317127227783202,
      "memory(GiB)": 77.56,
      "step": 22725,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.444961
    },
    {
      "epoch": 0.973822886765777,
      "grad_norm": 4.974038600921631,
      "learning_rate": 9.092873843426986e-05,
      "loss": 2.7294788360595703,
      "memory(GiB)": 77.56,
      "step": 22730,
      "token_acc": 0.4597315436241611,
      "train_speed(iter/s)": 1.444971
    },
    {
      "epoch": 0.9740371020950259,
      "grad_norm": 6.928504943847656,
      "learning_rate": 9.092487248198222e-05,
      "loss": 2.795309066772461,
      "memory(GiB)": 77.56,
      "step": 22735,
      "token_acc": 0.4342105263157895,
      "train_speed(iter/s)": 1.445
    },
    {
      "epoch": 0.9742513174242748,
      "grad_norm": 4.414449214935303,
      "learning_rate": 9.092100578830214e-05,
      "loss": 2.5382831573486326,
      "memory(GiB)": 77.56,
      "step": 22740,
      "token_acc": 0.45625,
      "train_speed(iter/s)": 1.44499
    },
    {
      "epoch": 0.9744655327535239,
      "grad_norm": 4.456477642059326,
      "learning_rate": 9.091713835329964e-05,
      "loss": 2.560292053222656,
      "memory(GiB)": 77.56,
      "step": 22745,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.444983
    },
    {
      "epoch": 0.9746797480827728,
      "grad_norm": 8.5945405960083,
      "learning_rate": 9.091327017704479e-05,
      "loss": 2.475649833679199,
      "memory(GiB)": 77.56,
      "step": 22750,
      "token_acc": 0.5139442231075697,
      "train_speed(iter/s)": 1.444872
    },
    {
      "epoch": 0.9748939634120217,
      "grad_norm": 4.095578670501709,
      "learning_rate": 9.090940125960769e-05,
      "loss": 2.6476024627685546,
      "memory(GiB)": 77.56,
      "step": 22755,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.444803
    },
    {
      "epoch": 0.9751081787412708,
      "grad_norm": 4.861550807952881,
      "learning_rate": 9.090553160105839e-05,
      "loss": 2.380379486083984,
      "memory(GiB)": 77.56,
      "step": 22760,
      "token_acc": 0.42955326460481097,
      "train_speed(iter/s)": 1.444867
    },
    {
      "epoch": 0.9753223940705197,
      "grad_norm": 4.669407367706299,
      "learning_rate": 9.090166120146702e-05,
      "loss": 2.7587869644165037,
      "memory(GiB)": 77.56,
      "step": 22765,
      "token_acc": 0.5155038759689923,
      "train_speed(iter/s)": 1.444875
    },
    {
      "epoch": 0.9755366093997686,
      "grad_norm": 4.914674282073975,
      "learning_rate": 9.08977900609037e-05,
      "loss": 2.7206090927124023,
      "memory(GiB)": 77.56,
      "step": 22770,
      "token_acc": 0.43630573248407645,
      "train_speed(iter/s)": 1.444953
    },
    {
      "epoch": 0.9757508247290176,
      "grad_norm": 7.4558258056640625,
      "learning_rate": 9.089391817943853e-05,
      "loss": 2.3912635803222657,
      "memory(GiB)": 77.56,
      "step": 22775,
      "token_acc": 0.5655737704918032,
      "train_speed(iter/s)": 1.445006
    },
    {
      "epoch": 0.9759650400582666,
      "grad_norm": 4.739561557769775,
      "learning_rate": 9.089004555714168e-05,
      "loss": 2.5282665252685548,
      "memory(GiB)": 77.56,
      "step": 22780,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.44506
    },
    {
      "epoch": 0.9761792553875155,
      "grad_norm": 4.167211055755615,
      "learning_rate": 9.08861721940833e-05,
      "loss": 2.6400732040405273,
      "memory(GiB)": 77.56,
      "step": 22785,
      "token_acc": 0.4740061162079511,
      "train_speed(iter/s)": 1.445067
    },
    {
      "epoch": 0.9763934707167645,
      "grad_norm": 4.274940490722656,
      "learning_rate": 9.088229809033355e-05,
      "loss": 2.4364160537719726,
      "memory(GiB)": 77.56,
      "step": 22790,
      "token_acc": 0.4770992366412214,
      "train_speed(iter/s)": 1.445105
    },
    {
      "epoch": 0.9766076860460134,
      "grad_norm": 3.6536030769348145,
      "learning_rate": 9.087842324596262e-05,
      "loss": 2.3567413330078124,
      "memory(GiB)": 77.56,
      "step": 22795,
      "token_acc": 0.4307228915662651,
      "train_speed(iter/s)": 1.445098
    },
    {
      "epoch": 0.9768219013752624,
      "grad_norm": 4.775113105773926,
      "learning_rate": 9.087454766104071e-05,
      "loss": 2.8382440567016602,
      "memory(GiB)": 77.56,
      "step": 22800,
      "token_acc": 0.43234323432343236,
      "train_speed(iter/s)": 1.445129
    },
    {
      "epoch": 0.9770361167045114,
      "grad_norm": 4.350823402404785,
      "learning_rate": 9.087067133563803e-05,
      "loss": 2.4608108520507814,
      "memory(GiB)": 77.56,
      "step": 22805,
      "token_acc": 0.4552238805970149,
      "train_speed(iter/s)": 1.445189
    },
    {
      "epoch": 0.9772503320337603,
      "grad_norm": 3.461327075958252,
      "learning_rate": 9.086679426982479e-05,
      "loss": 2.4220829010009766,
      "memory(GiB)": 77.56,
      "step": 22810,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.44509
    },
    {
      "epoch": 0.9774645473630093,
      "grad_norm": 4.952081203460693,
      "learning_rate": 9.086291646367123e-05,
      "loss": 2.6396175384521485,
      "memory(GiB)": 77.56,
      "step": 22815,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.445064
    },
    {
      "epoch": 0.9776787626922583,
      "grad_norm": 4.0555195808410645,
      "learning_rate": 9.085903791724761e-05,
      "loss": 2.960065460205078,
      "memory(GiB)": 77.56,
      "step": 22820,
      "token_acc": 0.4635036496350365,
      "train_speed(iter/s)": 1.44515
    },
    {
      "epoch": 0.9778929780215072,
      "grad_norm": 4.5373945236206055,
      "learning_rate": 9.085515863062419e-05,
      "loss": 2.6047840118408203,
      "memory(GiB)": 77.56,
      "step": 22825,
      "token_acc": 0.45051194539249145,
      "train_speed(iter/s)": 1.445167
    },
    {
      "epoch": 0.9781071933507561,
      "grad_norm": 4.878065586090088,
      "learning_rate": 9.085127860387126e-05,
      "loss": 2.434050941467285,
      "memory(GiB)": 77.56,
      "step": 22830,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.445219
    },
    {
      "epoch": 0.9783214086800052,
      "grad_norm": 4.08262825012207,
      "learning_rate": 9.084739783705909e-05,
      "loss": 2.9012475967407227,
      "memory(GiB)": 77.56,
      "step": 22835,
      "token_acc": 0.4270833333333333,
      "train_speed(iter/s)": 1.445105
    },
    {
      "epoch": 0.9785356240092541,
      "grad_norm": 3.815702199935913,
      "learning_rate": 9.084351633025798e-05,
      "loss": 2.516925239562988,
      "memory(GiB)": 77.56,
      "step": 22840,
      "token_acc": 0.4862068965517241,
      "train_speed(iter/s)": 1.445117
    },
    {
      "epoch": 0.978749839338503,
      "grad_norm": 4.985994338989258,
      "learning_rate": 9.083963408353825e-05,
      "loss": 2.653405952453613,
      "memory(GiB)": 77.56,
      "step": 22845,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.445185
    },
    {
      "epoch": 0.978964054667752,
      "grad_norm": 4.019278049468994,
      "learning_rate": 9.083575109697027e-05,
      "loss": 2.7412076950073243,
      "memory(GiB)": 77.56,
      "step": 22850,
      "token_acc": 0.42443729903536975,
      "train_speed(iter/s)": 1.445137
    },
    {
      "epoch": 0.979178269997001,
      "grad_norm": 12.105887413024902,
      "learning_rate": 9.083186737062432e-05,
      "loss": 2.926678466796875,
      "memory(GiB)": 77.56,
      "step": 22855,
      "token_acc": 0.41297935103244837,
      "train_speed(iter/s)": 1.44508
    },
    {
      "epoch": 0.9793924853262499,
      "grad_norm": 3.4400057792663574,
      "learning_rate": 9.082798290457081e-05,
      "loss": 2.459037017822266,
      "memory(GiB)": 77.56,
      "step": 22860,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.445095
    },
    {
      "epoch": 0.9796067006554989,
      "grad_norm": 3.8603885173797607,
      "learning_rate": 9.082409769888008e-05,
      "loss": 2.6857845306396486,
      "memory(GiB)": 77.56,
      "step": 22865,
      "token_acc": 0.4585492227979275,
      "train_speed(iter/s)": 1.445136
    },
    {
      "epoch": 0.9798209159847479,
      "grad_norm": 3.772304058074951,
      "learning_rate": 9.082021175362252e-05,
      "loss": 2.4039142608642576,
      "memory(GiB)": 77.56,
      "step": 22870,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.445133
    },
    {
      "epoch": 0.9800351313139968,
      "grad_norm": 5.362132549285889,
      "learning_rate": 9.081632506886854e-05,
      "loss": 2.4876777648925783,
      "memory(GiB)": 77.56,
      "step": 22875,
      "token_acc": 0.4362934362934363,
      "train_speed(iter/s)": 1.445188
    },
    {
      "epoch": 0.9802493466432458,
      "grad_norm": 4.6251702308654785,
      "learning_rate": 9.081243764468854e-05,
      "loss": 2.8020580291748045,
      "memory(GiB)": 77.56,
      "step": 22880,
      "token_acc": 0.4393939393939394,
      "train_speed(iter/s)": 1.445219
    },
    {
      "epoch": 0.9804635619724947,
      "grad_norm": 5.647506237030029,
      "learning_rate": 9.080854948115295e-05,
      "loss": 2.6524225234985352,
      "memory(GiB)": 77.56,
      "step": 22885,
      "token_acc": 0.45481927710843373,
      "train_speed(iter/s)": 1.445251
    },
    {
      "epoch": 0.9806777773017437,
      "grad_norm": 4.38006591796875,
      "learning_rate": 9.080466057833221e-05,
      "loss": 2.3606569290161135,
      "memory(GiB)": 77.56,
      "step": 22890,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.445305
    },
    {
      "epoch": 0.9808919926309927,
      "grad_norm": 3.4390718936920166,
      "learning_rate": 9.080077093629675e-05,
      "loss": 2.8349235534667967,
      "memory(GiB)": 77.56,
      "step": 22895,
      "token_acc": 0.4199395770392749,
      "train_speed(iter/s)": 1.445333
    },
    {
      "epoch": 0.9811062079602416,
      "grad_norm": 4.981522083282471,
      "learning_rate": 9.079688055511707e-05,
      "loss": 2.579480743408203,
      "memory(GiB)": 77.56,
      "step": 22900,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.445347
    },
    {
      "epoch": 0.9813204232894907,
      "grad_norm": 4.231136322021484,
      "learning_rate": 9.079298943486361e-05,
      "loss": 2.410441207885742,
      "memory(GiB)": 77.56,
      "step": 22905,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.445391
    },
    {
      "epoch": 0.9815346386187396,
      "grad_norm": 4.782087326049805,
      "learning_rate": 9.078909757560687e-05,
      "loss": 2.3499324798583983,
      "memory(GiB)": 77.56,
      "step": 22910,
      "token_acc": 0.5451388888888888,
      "train_speed(iter/s)": 1.445423
    },
    {
      "epoch": 0.9817488539479885,
      "grad_norm": 7.973390102386475,
      "learning_rate": 9.07852049774174e-05,
      "loss": 2.193668174743652,
      "memory(GiB)": 77.56,
      "step": 22915,
      "token_acc": 0.49407114624505927,
      "train_speed(iter/s)": 1.445524
    },
    {
      "epoch": 0.9819630692772375,
      "grad_norm": 3.6842994689941406,
      "learning_rate": 9.078131164036565e-05,
      "loss": 2.497384262084961,
      "memory(GiB)": 77.56,
      "step": 22920,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.445543
    },
    {
      "epoch": 0.9821772846064865,
      "grad_norm": 5.814662933349609,
      "learning_rate": 9.07774175645222e-05,
      "loss": 2.5203853607177735,
      "memory(GiB)": 77.56,
      "step": 22925,
      "token_acc": 0.5046153846153846,
      "train_speed(iter/s)": 1.445567
    },
    {
      "epoch": 0.9823914999357354,
      "grad_norm": 4.383076190948486,
      "learning_rate": 9.077352274995757e-05,
      "loss": 2.5637889862060548,
      "memory(GiB)": 77.56,
      "step": 22930,
      "token_acc": 0.4249084249084249,
      "train_speed(iter/s)": 1.445555
    },
    {
      "epoch": 0.9826057152649844,
      "grad_norm": 4.262376308441162,
      "learning_rate": 9.076962719674233e-05,
      "loss": 2.8621448516845702,
      "memory(GiB)": 77.56,
      "step": 22935,
      "token_acc": 0.463855421686747,
      "train_speed(iter/s)": 1.445597
    },
    {
      "epoch": 0.9828199305942333,
      "grad_norm": 5.175243854522705,
      "learning_rate": 9.076573090494704e-05,
      "loss": 2.4571056365966797,
      "memory(GiB)": 77.56,
      "step": 22940,
      "token_acc": 0.4558303886925795,
      "train_speed(iter/s)": 1.445619
    },
    {
      "epoch": 0.9830341459234823,
      "grad_norm": 3.8173458576202393,
      "learning_rate": 9.076183387464232e-05,
      "loss": 2.4322261810302734,
      "memory(GiB)": 77.56,
      "step": 22945,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.445669
    },
    {
      "epoch": 0.9832483612527313,
      "grad_norm": 6.433363914489746,
      "learning_rate": 9.075793610589871e-05,
      "loss": 2.3937919616699217,
      "memory(GiB)": 77.56,
      "step": 22950,
      "token_acc": 0.48945147679324896,
      "train_speed(iter/s)": 1.445723
    },
    {
      "epoch": 0.9834625765819802,
      "grad_norm": 7.985085964202881,
      "learning_rate": 9.075403759878687e-05,
      "loss": 2.673828887939453,
      "memory(GiB)": 77.56,
      "step": 22955,
      "token_acc": 0.48031496062992124,
      "train_speed(iter/s)": 1.445751
    },
    {
      "epoch": 0.9836767919112291,
      "grad_norm": 5.2178215980529785,
      "learning_rate": 9.075013835337742e-05,
      "loss": 2.692142105102539,
      "memory(GiB)": 77.56,
      "step": 22960,
      "token_acc": 0.47465437788018433,
      "train_speed(iter/s)": 1.445827
    },
    {
      "epoch": 0.9838910072404782,
      "grad_norm": 3.7585647106170654,
      "learning_rate": 9.074623836974097e-05,
      "loss": 2.3891376495361327,
      "memory(GiB)": 77.56,
      "step": 22965,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.445857
    },
    {
      "epoch": 0.9841052225697271,
      "grad_norm": 6.010097980499268,
      "learning_rate": 9.074233764794818e-05,
      "loss": 2.4187442779541017,
      "memory(GiB)": 77.56,
      "step": 22970,
      "token_acc": 0.4900662251655629,
      "train_speed(iter/s)": 1.445895
    },
    {
      "epoch": 0.984319437898976,
      "grad_norm": 3.8825864791870117,
      "learning_rate": 9.073843618806974e-05,
      "loss": 2.5002138137817385,
      "memory(GiB)": 77.56,
      "step": 22975,
      "token_acc": 0.45791245791245794,
      "train_speed(iter/s)": 1.445944
    },
    {
      "epoch": 0.9845336532282251,
      "grad_norm": 4.150593280792236,
      "learning_rate": 9.073453399017631e-05,
      "loss": 2.698598098754883,
      "memory(GiB)": 77.56,
      "step": 22980,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.445926
    },
    {
      "epoch": 0.984747868557474,
      "grad_norm": 7.992836952209473,
      "learning_rate": 9.073063105433859e-05,
      "loss": 2.6483861923217775,
      "memory(GiB)": 77.56,
      "step": 22985,
      "token_acc": 0.46496815286624205,
      "train_speed(iter/s)": 1.445888
    },
    {
      "epoch": 0.9849620838867229,
      "grad_norm": 6.339759826660156,
      "learning_rate": 9.072672738062726e-05,
      "loss": 2.6745561599731444,
      "memory(GiB)": 77.56,
      "step": 22990,
      "token_acc": 0.47863247863247865,
      "train_speed(iter/s)": 1.44598
    },
    {
      "epoch": 0.9851762992159719,
      "grad_norm": 4.41357946395874,
      "learning_rate": 9.072282296911308e-05,
      "loss": 2.3496932983398438,
      "memory(GiB)": 77.56,
      "step": 22995,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.445981
    },
    {
      "epoch": 0.9853905145452209,
      "grad_norm": 4.551157474517822,
      "learning_rate": 9.071891781986675e-05,
      "loss": 2.695197105407715,
      "memory(GiB)": 77.56,
      "step": 23000,
      "token_acc": 0.43125,
      "train_speed(iter/s)": 1.44598
    },
    {
      "epoch": 0.9853905145452209,
      "eval_loss": 2.008117914199829,
      "eval_runtime": 14.9281,
      "eval_samples_per_second": 6.699,
      "eval_steps_per_second": 6.699,
      "eval_token_acc": 0.46785225718194257,
      "step": 23000
    },
    {
      "epoch": 0.9856047298744698,
      "grad_norm": 4.310020923614502,
      "learning_rate": 9.071501193295903e-05,
      "loss": 2.8055076599121094,
      "memory(GiB)": 77.56,
      "step": 23005,
      "token_acc": 0.4521224086870681,
      "train_speed(iter/s)": 1.444534
    },
    {
      "epoch": 0.9858189452037188,
      "grad_norm": 3.9575483798980713,
      "learning_rate": 9.071110530846067e-05,
      "loss": 2.4324247360229494,
      "memory(GiB)": 77.56,
      "step": 23010,
      "token_acc": 0.49117647058823527,
      "train_speed(iter/s)": 1.444589
    },
    {
      "epoch": 0.9860331605329677,
      "grad_norm": 5.247930526733398,
      "learning_rate": 9.070719794644245e-05,
      "loss": 2.610725212097168,
      "memory(GiB)": 77.56,
      "step": 23015,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.444618
    },
    {
      "epoch": 0.9862473758622167,
      "grad_norm": 4.458913803100586,
      "learning_rate": 9.070328984697516e-05,
      "loss": 2.361677360534668,
      "memory(GiB)": 77.56,
      "step": 23020,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.444574
    },
    {
      "epoch": 0.9864615911914657,
      "grad_norm": 6.627816677093506,
      "learning_rate": 9.069938101012958e-05,
      "loss": 2.5770853042602537,
      "memory(GiB)": 77.56,
      "step": 23025,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.444516
    },
    {
      "epoch": 0.9866758065207146,
      "grad_norm": 4.926617622375488,
      "learning_rate": 9.069547143597655e-05,
      "loss": 2.5908576965332033,
      "memory(GiB)": 77.56,
      "step": 23030,
      "token_acc": 0.4478114478114478,
      "train_speed(iter/s)": 1.444564
    },
    {
      "epoch": 0.9868900218499636,
      "grad_norm": 4.594491004943848,
      "learning_rate": 9.069156112458685e-05,
      "loss": 2.565652847290039,
      "memory(GiB)": 77.56,
      "step": 23035,
      "token_acc": 0.46726190476190477,
      "train_speed(iter/s)": 1.444485
    },
    {
      "epoch": 0.9871042371792126,
      "grad_norm": 4.847806453704834,
      "learning_rate": 9.068765007603137e-05,
      "loss": 2.6189918518066406,
      "memory(GiB)": 77.56,
      "step": 23040,
      "token_acc": 0.42805755395683454,
      "train_speed(iter/s)": 1.444437
    },
    {
      "epoch": 0.9873184525084615,
      "grad_norm": 5.198618412017822,
      "learning_rate": 9.068373829038095e-05,
      "loss": 2.7961402893066407,
      "memory(GiB)": 77.56,
      "step": 23045,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.444406
    },
    {
      "epoch": 0.9875326678377104,
      "grad_norm": 3.8865277767181396,
      "learning_rate": 9.067982576770644e-05,
      "loss": 2.4614585876464843,
      "memory(GiB)": 77.56,
      "step": 23050,
      "token_acc": 0.48265895953757226,
      "train_speed(iter/s)": 1.44438
    },
    {
      "epoch": 0.9877468831669595,
      "grad_norm": 4.218410968780518,
      "learning_rate": 9.067591250807872e-05,
      "loss": 2.5644222259521485,
      "memory(GiB)": 77.56,
      "step": 23055,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.444476
    },
    {
      "epoch": 0.9879610984962084,
      "grad_norm": 6.073484420776367,
      "learning_rate": 9.067199851156869e-05,
      "loss": 2.5710912704467774,
      "memory(GiB)": 77.56,
      "step": 23060,
      "token_acc": 0.4468864468864469,
      "train_speed(iter/s)": 1.44456
    },
    {
      "epoch": 0.9881753138254573,
      "grad_norm": 3.766753673553467,
      "learning_rate": 9.066808377824725e-05,
      "loss": 2.7560733795166015,
      "memory(GiB)": 77.56,
      "step": 23065,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.444613
    },
    {
      "epoch": 0.9883895291547063,
      "grad_norm": 5.844721794128418,
      "learning_rate": 9.066416830818531e-05,
      "loss": 3.010381507873535,
      "memory(GiB)": 77.56,
      "step": 23070,
      "token_acc": 0.4789272030651341,
      "train_speed(iter/s)": 1.44468
    },
    {
      "epoch": 0.9886037444839553,
      "grad_norm": 3.5435407161712646,
      "learning_rate": 9.066025210145384e-05,
      "loss": 2.3852956771850584,
      "memory(GiB)": 77.56,
      "step": 23075,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.44466
    },
    {
      "epoch": 0.9888179598132042,
      "grad_norm": 3.66520619392395,
      "learning_rate": 9.065633515812376e-05,
      "loss": 2.4725925445556642,
      "memory(GiB)": 77.56,
      "step": 23080,
      "token_acc": 0.5089820359281437,
      "train_speed(iter/s)": 1.44473
    },
    {
      "epoch": 0.9890321751424532,
      "grad_norm": 3.9173736572265625,
      "learning_rate": 9.0652417478266e-05,
      "loss": 2.9407861709594725,
      "memory(GiB)": 77.56,
      "step": 23085,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.444754
    },
    {
      "epoch": 0.9892463904717022,
      "grad_norm": 4.830893516540527,
      "learning_rate": 9.064849906195159e-05,
      "loss": 2.4900604248046876,
      "memory(GiB)": 77.56,
      "step": 23090,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.444707
    },
    {
      "epoch": 0.9894606058009511,
      "grad_norm": 6.276354789733887,
      "learning_rate": 9.064457990925149e-05,
      "loss": 2.5824522018432616,
      "memory(GiB)": 77.56,
      "step": 23095,
      "token_acc": 0.45934959349593496,
      "train_speed(iter/s)": 1.444654
    },
    {
      "epoch": 0.9896748211302001,
      "grad_norm": 3.8143110275268555,
      "learning_rate": 9.064066002023668e-05,
      "loss": 2.5432191848754884,
      "memory(GiB)": 77.56,
      "step": 23100,
      "token_acc": 0.45666666666666667,
      "train_speed(iter/s)": 1.444589
    },
    {
      "epoch": 0.989889036459449,
      "grad_norm": 4.650741100311279,
      "learning_rate": 9.06367393949782e-05,
      "loss": 2.656488227844238,
      "memory(GiB)": 77.56,
      "step": 23105,
      "token_acc": 0.4837662337662338,
      "train_speed(iter/s)": 1.444601
    },
    {
      "epoch": 0.990103251788698,
      "grad_norm": 4.140844821929932,
      "learning_rate": 9.063281803354707e-05,
      "loss": 2.60064811706543,
      "memory(GiB)": 77.56,
      "step": 23110,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.444602
    },
    {
      "epoch": 0.990317467117947,
      "grad_norm": 4.256901741027832,
      "learning_rate": 9.062889593601432e-05,
      "loss": 2.6823265075683596,
      "memory(GiB)": 77.56,
      "step": 23115,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.44459
    },
    {
      "epoch": 0.9905316824471959,
      "grad_norm": 4.084400177001953,
      "learning_rate": 9.0624973102451e-05,
      "loss": 2.900673675537109,
      "memory(GiB)": 77.56,
      "step": 23120,
      "token_acc": 0.4378531073446328,
      "train_speed(iter/s)": 1.444632
    },
    {
      "epoch": 0.9907458977764448,
      "grad_norm": 5.528582572937012,
      "learning_rate": 9.062104953292819e-05,
      "loss": 2.6865842819213865,
      "memory(GiB)": 77.56,
      "step": 23125,
      "token_acc": 0.46484375,
      "train_speed(iter/s)": 1.444671
    },
    {
      "epoch": 0.9909601131056939,
      "grad_norm": 3.866848945617676,
      "learning_rate": 9.061712522751696e-05,
      "loss": 2.5581182479858398,
      "memory(GiB)": 77.56,
      "step": 23130,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.444633
    },
    {
      "epoch": 0.9911743284349428,
      "grad_norm": 5.471525192260742,
      "learning_rate": 9.06132001862884e-05,
      "loss": 2.503224563598633,
      "memory(GiB)": 77.56,
      "step": 23135,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.444684
    },
    {
      "epoch": 0.9913885437641917,
      "grad_norm": 5.09264612197876,
      "learning_rate": 9.060927440931362e-05,
      "loss": 2.767590141296387,
      "memory(GiB)": 77.56,
      "step": 23140,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.444734
    },
    {
      "epoch": 0.9916027590934408,
      "grad_norm": 4.136631965637207,
      "learning_rate": 9.060534789666374e-05,
      "loss": 2.459285926818848,
      "memory(GiB)": 77.56,
      "step": 23145,
      "token_acc": 0.5053763440860215,
      "train_speed(iter/s)": 1.444673
    },
    {
      "epoch": 0.9918169744226897,
      "grad_norm": 5.8304338455200195,
      "learning_rate": 9.06014206484099e-05,
      "loss": 2.5782939910888674,
      "memory(GiB)": 77.56,
      "step": 23150,
      "token_acc": 0.4902597402597403,
      "train_speed(iter/s)": 1.444676
    },
    {
      "epoch": 0.9920311897519386,
      "grad_norm": 5.046599388122559,
      "learning_rate": 9.059749266462324e-05,
      "loss": 2.475712013244629,
      "memory(GiB)": 77.56,
      "step": 23155,
      "token_acc": 0.46195652173913043,
      "train_speed(iter/s)": 1.444766
    },
    {
      "epoch": 0.9922454050811876,
      "grad_norm": 8.552024841308594,
      "learning_rate": 9.05935639453749e-05,
      "loss": 2.714438247680664,
      "memory(GiB)": 77.56,
      "step": 23160,
      "token_acc": 0.4339622641509434,
      "train_speed(iter/s)": 1.444795
    },
    {
      "epoch": 0.9924596204104366,
      "grad_norm": 4.3296427726745605,
      "learning_rate": 9.058963449073607e-05,
      "loss": 2.878192901611328,
      "memory(GiB)": 77.56,
      "step": 23165,
      "token_acc": 0.4128113879003559,
      "train_speed(iter/s)": 1.444837
    },
    {
      "epoch": 0.9926738357396855,
      "grad_norm": 4.40519905090332,
      "learning_rate": 9.058570430077795e-05,
      "loss": 2.5774843215942385,
      "memory(GiB)": 77.56,
      "step": 23170,
      "token_acc": 0.4794007490636704,
      "train_speed(iter/s)": 1.444823
    },
    {
      "epoch": 0.9928880510689345,
      "grad_norm": 4.012173175811768,
      "learning_rate": 9.058177337557172e-05,
      "loss": 2.802840995788574,
      "memory(GiB)": 77.56,
      "step": 23175,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.444816
    },
    {
      "epoch": 0.9931022663981834,
      "grad_norm": 4.722075939178467,
      "learning_rate": 9.057784171518861e-05,
      "loss": 2.7361244201660155,
      "memory(GiB)": 77.56,
      "step": 23180,
      "token_acc": 0.4610169491525424,
      "train_speed(iter/s)": 1.444708
    },
    {
      "epoch": 0.9933164817274324,
      "grad_norm": 4.343459606170654,
      "learning_rate": 9.057390931969981e-05,
      "loss": 2.681656837463379,
      "memory(GiB)": 77.56,
      "step": 23185,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.444693
    },
    {
      "epoch": 0.9935306970566814,
      "grad_norm": 5.151376247406006,
      "learning_rate": 9.056997618917659e-05,
      "loss": 2.448335647583008,
      "memory(GiB)": 77.56,
      "step": 23190,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.444681
    },
    {
      "epoch": 0.9937449123859303,
      "grad_norm": 4.584866046905518,
      "learning_rate": 9.056604232369019e-05,
      "loss": 2.314210319519043,
      "memory(GiB)": 77.56,
      "step": 23195,
      "token_acc": 0.48134328358208955,
      "train_speed(iter/s)": 1.444695
    },
    {
      "epoch": 0.9939591277151792,
      "grad_norm": 4.461553573608398,
      "learning_rate": 9.056210772331188e-05,
      "loss": 2.7964332580566404,
      "memory(GiB)": 77.56,
      "step": 23200,
      "token_acc": 0.45686900958466453,
      "train_speed(iter/s)": 1.444655
    },
    {
      "epoch": 0.9941733430444283,
      "grad_norm": 4.269008636474609,
      "learning_rate": 9.055817238811295e-05,
      "loss": 2.3951398849487306,
      "memory(GiB)": 77.56,
      "step": 23205,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.444709
    },
    {
      "epoch": 0.9943875583736772,
      "grad_norm": 4.945583820343018,
      "learning_rate": 9.055423631816466e-05,
      "loss": 2.5814277648925783,
      "memory(GiB)": 77.56,
      "step": 23210,
      "token_acc": 0.5064935064935064,
      "train_speed(iter/s)": 1.444708
    },
    {
      "epoch": 0.9946017737029261,
      "grad_norm": 5.2736358642578125,
      "learning_rate": 9.055029951353835e-05,
      "loss": 2.34924259185791,
      "memory(GiB)": 77.56,
      "step": 23215,
      "token_acc": 0.5202020202020202,
      "train_speed(iter/s)": 1.444764
    },
    {
      "epoch": 0.9948159890321752,
      "grad_norm": 4.148223876953125,
      "learning_rate": 9.054636197430533e-05,
      "loss": 2.7521587371826173,
      "memory(GiB)": 77.56,
      "step": 23220,
      "token_acc": 0.43119266055045874,
      "train_speed(iter/s)": 1.444778
    },
    {
      "epoch": 0.9950302043614241,
      "grad_norm": 4.876492023468018,
      "learning_rate": 9.054242370053691e-05,
      "loss": 2.670973777770996,
      "memory(GiB)": 77.56,
      "step": 23225,
      "token_acc": 0.440625,
      "train_speed(iter/s)": 1.444687
    },
    {
      "epoch": 0.995244419690673,
      "grad_norm": 3.71134352684021,
      "learning_rate": 9.053848469230446e-05,
      "loss": 2.806547164916992,
      "memory(GiB)": 77.56,
      "step": 23230,
      "token_acc": 0.4394904458598726,
      "train_speed(iter/s)": 1.444707
    },
    {
      "epoch": 0.995458635019922,
      "grad_norm": 5.050687313079834,
      "learning_rate": 9.053454494967935e-05,
      "loss": 2.493272399902344,
      "memory(GiB)": 77.56,
      "step": 23235,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.444714
    },
    {
      "epoch": 0.995672850349171,
      "grad_norm": 4.164772033691406,
      "learning_rate": 9.053060447273291e-05,
      "loss": 2.5831829071044923,
      "memory(GiB)": 77.56,
      "step": 23240,
      "token_acc": 0.4537037037037037,
      "train_speed(iter/s)": 1.444643
    },
    {
      "epoch": 0.99588706567842,
      "grad_norm": 6.227268695831299,
      "learning_rate": 9.052666326153656e-05,
      "loss": 2.7408420562744142,
      "memory(GiB)": 77.56,
      "step": 23245,
      "token_acc": 0.44072948328267475,
      "train_speed(iter/s)": 1.444681
    },
    {
      "epoch": 0.9961012810076689,
      "grad_norm": 4.395718574523926,
      "learning_rate": 9.052272131616168e-05,
      "loss": 2.68499698638916,
      "memory(GiB)": 77.56,
      "step": 23250,
      "token_acc": 0.44224422442244227,
      "train_speed(iter/s)": 1.444717
    },
    {
      "epoch": 0.9963154963369178,
      "grad_norm": 7.666308879852295,
      "learning_rate": 9.051877863667969e-05,
      "loss": 2.57991886138916,
      "memory(GiB)": 77.56,
      "step": 23255,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.444781
    },
    {
      "epoch": 0.9965297116661669,
      "grad_norm": 5.391880512237549,
      "learning_rate": 9.051483522316202e-05,
      "loss": 2.4933521270751955,
      "memory(GiB)": 77.56,
      "step": 23260,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.444799
    },
    {
      "epoch": 0.9967439269954158,
      "grad_norm": 7.581659317016602,
      "learning_rate": 9.05108910756801e-05,
      "loss": 2.64639892578125,
      "memory(GiB)": 77.56,
      "step": 23265,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.444839
    },
    {
      "epoch": 0.9969581423246647,
      "grad_norm": 4.262197971343994,
      "learning_rate": 9.050694619430539e-05,
      "loss": 2.3680606842041017,
      "memory(GiB)": 77.56,
      "step": 23270,
      "token_acc": 0.4962121212121212,
      "train_speed(iter/s)": 1.444799
    },
    {
      "epoch": 0.9971723576539138,
      "grad_norm": 4.928798198699951,
      "learning_rate": 9.050300057910936e-05,
      "loss": 2.792987823486328,
      "memory(GiB)": 77.56,
      "step": 23275,
      "token_acc": 0.4405144694533762,
      "train_speed(iter/s)": 1.44484
    },
    {
      "epoch": 0.9973865729831627,
      "grad_norm": 3.818899154663086,
      "learning_rate": 9.049905423016347e-05,
      "loss": 2.4279727935791016,
      "memory(GiB)": 77.56,
      "step": 23280,
      "token_acc": 0.4405594405594406,
      "train_speed(iter/s)": 1.444897
    },
    {
      "epoch": 0.9976007883124116,
      "grad_norm": 4.312445640563965,
      "learning_rate": 9.049510714753922e-05,
      "loss": 2.9236495971679686,
      "memory(GiB)": 77.56,
      "step": 23285,
      "token_acc": 0.41002949852507375,
      "train_speed(iter/s)": 1.44495
    },
    {
      "epoch": 0.9978150036416606,
      "grad_norm": 4.444835662841797,
      "learning_rate": 9.049115933130811e-05,
      "loss": 2.9291763305664062,
      "memory(GiB)": 77.56,
      "step": 23290,
      "token_acc": 0.4312977099236641,
      "train_speed(iter/s)": 1.44486
    },
    {
      "epoch": 0.9980292189709096,
      "grad_norm": 8.014284133911133,
      "learning_rate": 9.048721078154168e-05,
      "loss": 2.5835525512695314,
      "memory(GiB)": 77.56,
      "step": 23295,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.4449
    },
    {
      "epoch": 0.9982434343001585,
      "grad_norm": 4.911098957061768,
      "learning_rate": 9.048326149831143e-05,
      "loss": 2.60485782623291,
      "memory(GiB)": 77.56,
      "step": 23300,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.444913
    },
    {
      "epoch": 0.9984576496294075,
      "grad_norm": 4.6003499031066895,
      "learning_rate": 9.047931148168894e-05,
      "loss": 2.5792760848999023,
      "memory(GiB)": 77.56,
      "step": 23305,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.444905
    },
    {
      "epoch": 0.9986718649586565,
      "grad_norm": 3.4633169174194336,
      "learning_rate": 9.047536073174573e-05,
      "loss": 2.8468677520751955,
      "memory(GiB)": 77.56,
      "step": 23310,
      "token_acc": 0.4322916666666667,
      "train_speed(iter/s)": 1.444853
    },
    {
      "epoch": 0.9988860802879054,
      "grad_norm": 4.67794942855835,
      "learning_rate": 9.047140924855342e-05,
      "loss": 2.48193244934082,
      "memory(GiB)": 77.56,
      "step": 23315,
      "token_acc": 0.511864406779661,
      "train_speed(iter/s)": 1.444759
    },
    {
      "epoch": 0.9991002956171544,
      "grad_norm": 5.053409099578857,
      "learning_rate": 9.046745703218356e-05,
      "loss": 2.7321683883666994,
      "memory(GiB)": 77.56,
      "step": 23320,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.444813
    },
    {
      "epoch": 0.9993145109464033,
      "grad_norm": 4.058961868286133,
      "learning_rate": 9.046350408270772e-05,
      "loss": 2.5489475250244142,
      "memory(GiB)": 77.56,
      "step": 23325,
      "token_acc": 0.5157894736842106,
      "train_speed(iter/s)": 1.444769
    },
    {
      "epoch": 0.9995287262756523,
      "grad_norm": 4.068192481994629,
      "learning_rate": 9.045955040019758e-05,
      "loss": 2.542393684387207,
      "memory(GiB)": 77.56,
      "step": 23330,
      "token_acc": 0.44483985765124556,
      "train_speed(iter/s)": 1.444758
    },
    {
      "epoch": 0.9997429416049013,
      "grad_norm": 6.695660591125488,
      "learning_rate": 9.045559598472472e-05,
      "loss": 2.796040153503418,
      "memory(GiB)": 77.56,
      "step": 23335,
      "token_acc": 0.4225352112676056,
      "train_speed(iter/s)": 1.444775
    },
    {
      "epoch": 0.9999571569341502,
      "grad_norm": 3.9384238719940186,
      "learning_rate": 9.045164083636079e-05,
      "loss": 2.6224626541137694,
      "memory(GiB)": 77.56,
      "step": 23340,
      "token_acc": 0.4293628808864266,
      "train_speed(iter/s)": 1.444831
    },
    {
      "epoch": 1.0001713722633991,
      "grad_norm": 4.683394908905029,
      "learning_rate": 9.044768495517744e-05,
      "loss": 2.376451110839844,
      "memory(GiB)": 77.56,
      "step": 23345,
      "token_acc": 0.4957983193277311,
      "train_speed(iter/s)": 1.444882
    },
    {
      "epoch": 1.0003855875926482,
      "grad_norm": 5.288326263427734,
      "learning_rate": 9.044372834124632e-05,
      "loss": 2.3113418579101563,
      "memory(GiB)": 77.56,
      "step": 23350,
      "token_acc": 0.5021097046413502,
      "train_speed(iter/s)": 1.444906
    },
    {
      "epoch": 1.000599802921897,
      "grad_norm": 5.382696628570557,
      "learning_rate": 9.043977099463914e-05,
      "loss": 2.6199560165405273,
      "memory(GiB)": 77.56,
      "step": 23355,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.444945
    },
    {
      "epoch": 1.000814018251146,
      "grad_norm": 4.5582780838012695,
      "learning_rate": 9.043581291542757e-05,
      "loss": 2.6732173919677735,
      "memory(GiB)": 77.56,
      "step": 23360,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.444919
    },
    {
      "epoch": 1.001028233580395,
      "grad_norm": 3.6138248443603516,
      "learning_rate": 9.043185410368332e-05,
      "loss": 2.5210849761962892,
      "memory(GiB)": 77.56,
      "step": 23365,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.444948
    },
    {
      "epoch": 1.0012424489096439,
      "grad_norm": 6.489718437194824,
      "learning_rate": 9.042789455947808e-05,
      "loss": 2.3586307525634767,
      "memory(GiB)": 77.56,
      "step": 23370,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.445018
    },
    {
      "epoch": 1.001456664238893,
      "grad_norm": 5.141478538513184,
      "learning_rate": 9.042393428288363e-05,
      "loss": 2.3916748046875,
      "memory(GiB)": 77.56,
      "step": 23375,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.445023
    },
    {
      "epoch": 1.001670879568142,
      "grad_norm": 4.6214728355407715,
      "learning_rate": 9.041997327397169e-05,
      "loss": 2.7524709701538086,
      "memory(GiB)": 77.56,
      "step": 23380,
      "token_acc": 0.483271375464684,
      "train_speed(iter/s)": 1.445068
    },
    {
      "epoch": 1.0018850948973907,
      "grad_norm": 5.004777908325195,
      "learning_rate": 9.0416011532814e-05,
      "loss": 2.4133243560791016,
      "memory(GiB)": 77.56,
      "step": 23385,
      "token_acc": 0.4866666666666667,
      "train_speed(iter/s)": 1.445081
    },
    {
      "epoch": 1.0020993102266398,
      "grad_norm": 3.792548179626465,
      "learning_rate": 9.041204905948236e-05,
      "loss": 2.499088096618652,
      "memory(GiB)": 77.56,
      "step": 23390,
      "token_acc": 0.5070821529745042,
      "train_speed(iter/s)": 1.445155
    },
    {
      "epoch": 1.0023135255558888,
      "grad_norm": 3.9695231914520264,
      "learning_rate": 9.040808585404854e-05,
      "loss": 2.4528261184692384,
      "memory(GiB)": 77.56,
      "step": 23395,
      "token_acc": 0.4902597402597403,
      "train_speed(iter/s)": 1.44521
    },
    {
      "epoch": 1.0025277408851376,
      "grad_norm": 4.173301696777344,
      "learning_rate": 9.040412191658434e-05,
      "loss": 2.3430273056030275,
      "memory(GiB)": 77.56,
      "step": 23400,
      "token_acc": 0.490625,
      "train_speed(iter/s)": 1.44522
    },
    {
      "epoch": 1.0027419562143867,
      "grad_norm": 5.720931529998779,
      "learning_rate": 9.040015724716157e-05,
      "loss": 2.5962554931640627,
      "memory(GiB)": 77.56,
      "step": 23405,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.445283
    },
    {
      "epoch": 1.0029561715436357,
      "grad_norm": 4.31567907333374,
      "learning_rate": 9.039619184585204e-05,
      "loss": 2.31009521484375,
      "memory(GiB)": 77.56,
      "step": 23410,
      "token_acc": 0.5316455696202531,
      "train_speed(iter/s)": 1.445276
    },
    {
      "epoch": 1.0031703868728847,
      "grad_norm": 4.946566581726074,
      "learning_rate": 9.039222571272763e-05,
      "loss": 2.402265739440918,
      "memory(GiB)": 77.56,
      "step": 23415,
      "token_acc": 0.5245283018867924,
      "train_speed(iter/s)": 1.445287
    },
    {
      "epoch": 1.0033846022021335,
      "grad_norm": 5.9220709800720215,
      "learning_rate": 9.038825884786013e-05,
      "loss": 2.526499557495117,
      "memory(GiB)": 77.56,
      "step": 23420,
      "token_acc": 0.4849624060150376,
      "train_speed(iter/s)": 1.445301
    },
    {
      "epoch": 1.0035988175313826,
      "grad_norm": 6.5811767578125,
      "learning_rate": 9.038429125132143e-05,
      "loss": 2.2615240097045897,
      "memory(GiB)": 77.56,
      "step": 23425,
      "token_acc": 0.5503875968992248,
      "train_speed(iter/s)": 1.445273
    },
    {
      "epoch": 1.0038130328606316,
      "grad_norm": 6.96791934967041,
      "learning_rate": 9.038032292318343e-05,
      "loss": 1.8306434631347657,
      "memory(GiB)": 77.56,
      "step": 23430,
      "token_acc": 0.6028708133971292,
      "train_speed(iter/s)": 1.445289
    },
    {
      "epoch": 1.0040272481898804,
      "grad_norm": 5.037871837615967,
      "learning_rate": 9.037635386351801e-05,
      "loss": 2.6830322265625,
      "memory(GiB)": 77.56,
      "step": 23435,
      "token_acc": 0.47560975609756095,
      "train_speed(iter/s)": 1.445322
    },
    {
      "epoch": 1.0042414635191295,
      "grad_norm": 4.981171607971191,
      "learning_rate": 9.037238407239705e-05,
      "loss": 2.4497539520263674,
      "memory(GiB)": 77.56,
      "step": 23440,
      "token_acc": 0.4607142857142857,
      "train_speed(iter/s)": 1.445317
    },
    {
      "epoch": 1.0044556788483785,
      "grad_norm": 4.907400608062744,
      "learning_rate": 9.036841354989248e-05,
      "loss": 2.5852592468261717,
      "memory(GiB)": 77.56,
      "step": 23445,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.445399
    },
    {
      "epoch": 1.0046698941776273,
      "grad_norm": 6.677310943603516,
      "learning_rate": 9.036444229607623e-05,
      "loss": 2.505696487426758,
      "memory(GiB)": 77.56,
      "step": 23450,
      "token_acc": 0.45634920634920634,
      "train_speed(iter/s)": 1.445407
    },
    {
      "epoch": 1.0048841095068763,
      "grad_norm": 5.097861289978027,
      "learning_rate": 9.036047031102024e-05,
      "loss": 2.3147756576538088,
      "memory(GiB)": 77.56,
      "step": 23455,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.445447
    },
    {
      "epoch": 1.0050983248361254,
      "grad_norm": 4.4812493324279785,
      "learning_rate": 9.035649759479648e-05,
      "loss": 2.6580783843994142,
      "memory(GiB)": 77.56,
      "step": 23460,
      "token_acc": 0.4979919678714859,
      "train_speed(iter/s)": 1.44545
    },
    {
      "epoch": 1.0053125401653742,
      "grad_norm": 4.0916290283203125,
      "learning_rate": 9.03525241474769e-05,
      "loss": 2.3689361572265626,
      "memory(GiB)": 77.56,
      "step": 23465,
      "token_acc": 0.49557522123893805,
      "train_speed(iter/s)": 1.445439
    },
    {
      "epoch": 1.0055267554946232,
      "grad_norm": 4.03123664855957,
      "learning_rate": 9.034854996913349e-05,
      "loss": 2.78851203918457,
      "memory(GiB)": 77.56,
      "step": 23470,
      "token_acc": 0.4696485623003195,
      "train_speed(iter/s)": 1.445535
    },
    {
      "epoch": 1.0057409708238723,
      "grad_norm": 4.383652210235596,
      "learning_rate": 9.034457505983825e-05,
      "loss": 2.697202682495117,
      "memory(GiB)": 77.56,
      "step": 23475,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.445527
    },
    {
      "epoch": 1.005955186153121,
      "grad_norm": 4.011081695556641,
      "learning_rate": 9.034059941966318e-05,
      "loss": 2.5129276275634767,
      "memory(GiB)": 77.56,
      "step": 23480,
      "token_acc": 0.47440273037542663,
      "train_speed(iter/s)": 1.445491
    },
    {
      "epoch": 1.00616940148237,
      "grad_norm": 5.929107666015625,
      "learning_rate": 9.033662304868031e-05,
      "loss": 2.607911491394043,
      "memory(GiB)": 77.56,
      "step": 23485,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.445454
    },
    {
      "epoch": 1.0063836168116191,
      "grad_norm": 5.403196811676025,
      "learning_rate": 9.033264594696169e-05,
      "loss": 2.611578178405762,
      "memory(GiB)": 77.56,
      "step": 23490,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.445462
    },
    {
      "epoch": 1.006597832140868,
      "grad_norm": 5.1242828369140625,
      "learning_rate": 9.032866811457935e-05,
      "loss": 2.7485353469848635,
      "memory(GiB)": 77.56,
      "step": 23495,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.4455
    },
    {
      "epoch": 1.006812047470117,
      "grad_norm": 4.291436195373535,
      "learning_rate": 9.032468955160533e-05,
      "loss": 2.426784706115723,
      "memory(GiB)": 77.56,
      "step": 23500,
      "token_acc": 0.5054945054945055,
      "train_speed(iter/s)": 1.445468
    },
    {
      "epoch": 1.006812047470117,
      "eval_loss": 2.2164549827575684,
      "eval_runtime": 14.6719,
      "eval_samples_per_second": 6.816,
      "eval_steps_per_second": 6.816,
      "eval_token_acc": 0.4899598393574297,
      "step": 23500
    },
    {
      "epoch": 1.007026262799366,
      "grad_norm": 4.435128211975098,
      "learning_rate": 9.032071025811175e-05,
      "loss": 2.2500232696533202,
      "memory(GiB)": 77.56,
      "step": 23505,
      "token_acc": 0.5024925224327019,
      "train_speed(iter/s)": 1.444143
    },
    {
      "epoch": 1.0072404781286148,
      "grad_norm": 3.8093514442443848,
      "learning_rate": 9.031673023417069e-05,
      "loss": 2.5141700744628905,
      "memory(GiB)": 77.56,
      "step": 23510,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.444128
    },
    {
      "epoch": 1.0074546934578639,
      "grad_norm": 4.975738525390625,
      "learning_rate": 9.031274947985422e-05,
      "loss": 2.4633569717407227,
      "memory(GiB)": 77.56,
      "step": 23515,
      "token_acc": 0.5132450331125827,
      "train_speed(iter/s)": 1.444133
    },
    {
      "epoch": 1.007668908787113,
      "grad_norm": 4.458843231201172,
      "learning_rate": 9.03087679952345e-05,
      "loss": 2.706062126159668,
      "memory(GiB)": 77.56,
      "step": 23520,
      "token_acc": 0.4412811387900356,
      "train_speed(iter/s)": 1.444104
    },
    {
      "epoch": 1.0078831241163617,
      "grad_norm": 3.636833906173706,
      "learning_rate": 9.030478578038361e-05,
      "loss": 2.2278390884399415,
      "memory(GiB)": 77.56,
      "step": 23525,
      "token_acc": 0.5256410256410257,
      "train_speed(iter/s)": 1.444151
    },
    {
      "epoch": 1.0080973394456108,
      "grad_norm": 5.378984451293945,
      "learning_rate": 9.030080283537374e-05,
      "loss": 2.7690994262695314,
      "memory(GiB)": 77.56,
      "step": 23530,
      "token_acc": 0.5210727969348659,
      "train_speed(iter/s)": 1.444155
    },
    {
      "epoch": 1.0083115547748598,
      "grad_norm": 5.114414691925049,
      "learning_rate": 9.029681916027701e-05,
      "loss": 2.727855110168457,
      "memory(GiB)": 77.56,
      "step": 23535,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.444231
    },
    {
      "epoch": 1.0085257701041086,
      "grad_norm": 4.054899215698242,
      "learning_rate": 9.029283475516561e-05,
      "loss": 2.854404830932617,
      "memory(GiB)": 77.56,
      "step": 23540,
      "token_acc": 0.4425087108013937,
      "train_speed(iter/s)": 1.44417
    },
    {
      "epoch": 1.0087399854333576,
      "grad_norm": 5.455842971801758,
      "learning_rate": 9.028884962011169e-05,
      "loss": 2.477247428894043,
      "memory(GiB)": 77.56,
      "step": 23545,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.444217
    },
    {
      "epoch": 1.0089542007626067,
      "grad_norm": 4.675895690917969,
      "learning_rate": 9.028486375518748e-05,
      "loss": 2.4009916305541994,
      "memory(GiB)": 77.56,
      "step": 23550,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.444203
    },
    {
      "epoch": 1.0091684160918555,
      "grad_norm": 4.928582191467285,
      "learning_rate": 9.028087716046516e-05,
      "loss": 2.26422004699707,
      "memory(GiB)": 77.56,
      "step": 23555,
      "token_acc": 0.5283687943262412,
      "train_speed(iter/s)": 1.444107
    },
    {
      "epoch": 1.0093826314211045,
      "grad_norm": 4.196980953216553,
      "learning_rate": 9.027688983601699e-05,
      "loss": 2.4747299194335937,
      "memory(GiB)": 77.56,
      "step": 23560,
      "token_acc": 0.4783950617283951,
      "train_speed(iter/s)": 1.44415
    },
    {
      "epoch": 1.0095968467503535,
      "grad_norm": 4.74194860458374,
      "learning_rate": 9.027290178191515e-05,
      "loss": 2.5284299850463867,
      "memory(GiB)": 77.56,
      "step": 23565,
      "token_acc": 0.47435897435897434,
      "train_speed(iter/s)": 1.444116
    },
    {
      "epoch": 1.0098110620796024,
      "grad_norm": 4.887885570526123,
      "learning_rate": 9.026891299823192e-05,
      "loss": 2.711962890625,
      "memory(GiB)": 77.56,
      "step": 23570,
      "token_acc": 0.47183098591549294,
      "train_speed(iter/s)": 1.4441
    },
    {
      "epoch": 1.0100252774088514,
      "grad_norm": 4.193269729614258,
      "learning_rate": 9.026492348503957e-05,
      "loss": 2.5940530776977537,
      "memory(GiB)": 77.56,
      "step": 23575,
      "token_acc": 0.47039473684210525,
      "train_speed(iter/s)": 1.444108
    },
    {
      "epoch": 1.0102394927381004,
      "grad_norm": 3.5875673294067383,
      "learning_rate": 9.026093324241035e-05,
      "loss": 2.2336780548095705,
      "memory(GiB)": 77.56,
      "step": 23580,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.444137
    },
    {
      "epoch": 1.0104537080673492,
      "grad_norm": 4.68424129486084,
      "learning_rate": 9.025694227041656e-05,
      "loss": 2.7033918380737303,
      "memory(GiB)": 77.56,
      "step": 23585,
      "token_acc": 0.4483870967741935,
      "train_speed(iter/s)": 1.444116
    },
    {
      "epoch": 1.0106679233965983,
      "grad_norm": 6.216456413269043,
      "learning_rate": 9.025295056913049e-05,
      "loss": 2.254861831665039,
      "memory(GiB)": 77.56,
      "step": 23590,
      "token_acc": 0.5274725274725275,
      "train_speed(iter/s)": 1.444148
    },
    {
      "epoch": 1.0108821387258473,
      "grad_norm": 4.2182698249816895,
      "learning_rate": 9.024895813862446e-05,
      "loss": 2.3986690521240233,
      "memory(GiB)": 77.56,
      "step": 23595,
      "token_acc": 0.5133333333333333,
      "train_speed(iter/s)": 1.444234
    },
    {
      "epoch": 1.0110963540550961,
      "grad_norm": 3.751181125640869,
      "learning_rate": 9.024496497897082e-05,
      "loss": 2.772421646118164,
      "memory(GiB)": 77.56,
      "step": 23600,
      "token_acc": 0.49640287769784175,
      "train_speed(iter/s)": 1.444263
    },
    {
      "epoch": 1.0113105693843452,
      "grad_norm": 5.261978626251221,
      "learning_rate": 9.024097109024186e-05,
      "loss": 2.661305618286133,
      "memory(GiB)": 77.56,
      "step": 23605,
      "token_acc": 0.4582043343653251,
      "train_speed(iter/s)": 1.444172
    },
    {
      "epoch": 1.0115247847135942,
      "grad_norm": 4.343251705169678,
      "learning_rate": 9.023697647250995e-05,
      "loss": 2.5762212753295897,
      "memory(GiB)": 77.56,
      "step": 23610,
      "token_acc": 0.45267489711934156,
      "train_speed(iter/s)": 1.444111
    },
    {
      "epoch": 1.011739000042843,
      "grad_norm": 4.401554584503174,
      "learning_rate": 9.023298112584749e-05,
      "loss": 2.5073280334472656,
      "memory(GiB)": 77.56,
      "step": 23615,
      "token_acc": 0.47232472324723246,
      "train_speed(iter/s)": 1.444144
    },
    {
      "epoch": 1.011953215372092,
      "grad_norm": 5.015374660491943,
      "learning_rate": 9.022898505032685e-05,
      "loss": 2.145386505126953,
      "memory(GiB)": 77.56,
      "step": 23620,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.444123
    },
    {
      "epoch": 1.012167430701341,
      "grad_norm": 4.877620220184326,
      "learning_rate": 9.022498824602037e-05,
      "loss": 2.8943077087402345,
      "memory(GiB)": 77.56,
      "step": 23625,
      "token_acc": 0.42803030303030304,
      "train_speed(iter/s)": 1.44421
    },
    {
      "epoch": 1.0123816460305899,
      "grad_norm": 4.435857772827148,
      "learning_rate": 9.022099071300052e-05,
      "loss": 2.5356782913208007,
      "memory(GiB)": 77.56,
      "step": 23630,
      "token_acc": 0.4548022598870056,
      "train_speed(iter/s)": 1.444255
    },
    {
      "epoch": 1.012595861359839,
      "grad_norm": 3.3223369121551514,
      "learning_rate": 9.021699245133967e-05,
      "loss": 2.3703386306762697,
      "memory(GiB)": 77.56,
      "step": 23635,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.444251
    },
    {
      "epoch": 1.012810076689088,
      "grad_norm": 5.082808971405029,
      "learning_rate": 9.02129934611103e-05,
      "loss": 2.4239749908447266,
      "memory(GiB)": 77.56,
      "step": 23640,
      "token_acc": 0.49673202614379086,
      "train_speed(iter/s)": 1.444226
    },
    {
      "epoch": 1.0130242920183368,
      "grad_norm": 4.389357566833496,
      "learning_rate": 9.020899374238481e-05,
      "loss": 2.378102111816406,
      "memory(GiB)": 77.56,
      "step": 23645,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.444253
    },
    {
      "epoch": 1.0132385073475858,
      "grad_norm": 4.6339521408081055,
      "learning_rate": 9.020499329523569e-05,
      "loss": 2.746384620666504,
      "memory(GiB)": 77.56,
      "step": 23650,
      "token_acc": 0.4817813765182186,
      "train_speed(iter/s)": 1.444261
    },
    {
      "epoch": 1.0134527226768348,
      "grad_norm": 4.083985328674316,
      "learning_rate": 9.02009921197354e-05,
      "loss": 2.3197113037109376,
      "memory(GiB)": 77.56,
      "step": 23655,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.444324
    },
    {
      "epoch": 1.0136669380060837,
      "grad_norm": 5.729962348937988,
      "learning_rate": 9.019699021595642e-05,
      "loss": 2.289439392089844,
      "memory(GiB)": 77.56,
      "step": 23660,
      "token_acc": 0.5285171102661597,
      "train_speed(iter/s)": 1.444364
    },
    {
      "epoch": 1.0138811533353327,
      "grad_norm": 5.57902193069458,
      "learning_rate": 9.019298758397127e-05,
      "loss": 2.804506301879883,
      "memory(GiB)": 77.56,
      "step": 23665,
      "token_acc": 0.39361702127659576,
      "train_speed(iter/s)": 1.444296
    },
    {
      "epoch": 1.0140953686645817,
      "grad_norm": 3.6773295402526855,
      "learning_rate": 9.018898422385243e-05,
      "loss": 2.359520149230957,
      "memory(GiB)": 77.56,
      "step": 23670,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.44425
    },
    {
      "epoch": 1.0143095839938305,
      "grad_norm": 5.6231770515441895,
      "learning_rate": 9.018498013567244e-05,
      "loss": 2.3890830993652346,
      "memory(GiB)": 77.56,
      "step": 23675,
      "token_acc": 0.5138339920948617,
      "train_speed(iter/s)": 1.444231
    },
    {
      "epoch": 1.0145237993230796,
      "grad_norm": 5.198417663574219,
      "learning_rate": 9.018097531950385e-05,
      "loss": 2.439471435546875,
      "memory(GiB)": 77.56,
      "step": 23680,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.444241
    },
    {
      "epoch": 1.0147380146523286,
      "grad_norm": 4.018439292907715,
      "learning_rate": 9.01769697754192e-05,
      "loss": 2.27854061126709,
      "memory(GiB)": 77.56,
      "step": 23685,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.444187
    },
    {
      "epoch": 1.0149522299815774,
      "grad_norm": 7.3545451164245605,
      "learning_rate": 9.017296350349105e-05,
      "loss": 2.6238658905029295,
      "memory(GiB)": 77.56,
      "step": 23690,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.444262
    },
    {
      "epoch": 1.0151664453108264,
      "grad_norm": 4.56335973739624,
      "learning_rate": 9.016895650379198e-05,
      "loss": 2.372761917114258,
      "memory(GiB)": 77.56,
      "step": 23695,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.444253
    },
    {
      "epoch": 1.0153806606400755,
      "grad_norm": 6.182762145996094,
      "learning_rate": 9.016494877639457e-05,
      "loss": 2.6362482070922852,
      "memory(GiB)": 77.56,
      "step": 23700,
      "token_acc": 0.43154761904761907,
      "train_speed(iter/s)": 1.444202
    },
    {
      "epoch": 1.0155948759693243,
      "grad_norm": 3.9759950637817383,
      "learning_rate": 9.016094032137144e-05,
      "loss": 2.288136291503906,
      "memory(GiB)": 77.56,
      "step": 23705,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.444228
    },
    {
      "epoch": 1.0158090912985733,
      "grad_norm": 4.441798686981201,
      "learning_rate": 9.015693113879521e-05,
      "loss": 2.769272232055664,
      "memory(GiB)": 77.56,
      "step": 23710,
      "token_acc": 0.4262295081967213,
      "train_speed(iter/s)": 1.444223
    },
    {
      "epoch": 1.0160233066278224,
      "grad_norm": 4.964745044708252,
      "learning_rate": 9.01529212287385e-05,
      "loss": 2.537934112548828,
      "memory(GiB)": 77.56,
      "step": 23715,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.444312
    },
    {
      "epoch": 1.0162375219570712,
      "grad_norm": 6.218685626983643,
      "learning_rate": 9.014891059127395e-05,
      "loss": 2.3481212615966798,
      "memory(GiB)": 77.56,
      "step": 23720,
      "token_acc": 0.5186440677966102,
      "train_speed(iter/s)": 1.444349
    },
    {
      "epoch": 1.0164517372863202,
      "grad_norm": 5.976919174194336,
      "learning_rate": 9.014489922647423e-05,
      "loss": 2.3539093017578123,
      "memory(GiB)": 77.56,
      "step": 23725,
      "token_acc": 0.5119047619047619,
      "train_speed(iter/s)": 1.444384
    },
    {
      "epoch": 1.0166659526155692,
      "grad_norm": 6.729485988616943,
      "learning_rate": 9.014088713441199e-05,
      "loss": 2.693385124206543,
      "memory(GiB)": 77.56,
      "step": 23730,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.444488
    },
    {
      "epoch": 1.016880167944818,
      "grad_norm": 4.621358394622803,
      "learning_rate": 9.013687431515994e-05,
      "loss": 2.6795597076416016,
      "memory(GiB)": 77.56,
      "step": 23735,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.444445
    },
    {
      "epoch": 1.017094383274067,
      "grad_norm": 6.505117416381836,
      "learning_rate": 9.013286076879075e-05,
      "loss": 2.6510812759399416,
      "memory(GiB)": 77.56,
      "step": 23740,
      "token_acc": 0.4440789473684211,
      "train_speed(iter/s)": 1.44448
    },
    {
      "epoch": 1.0173085986033161,
      "grad_norm": 4.350959777832031,
      "learning_rate": 9.012884649537715e-05,
      "loss": 2.6561456680297852,
      "memory(GiB)": 77.56,
      "step": 23745,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.444471
    },
    {
      "epoch": 1.017522813932565,
      "grad_norm": 4.7133002281188965,
      "learning_rate": 9.012483149499184e-05,
      "loss": 2.5441621780395507,
      "memory(GiB)": 77.56,
      "step": 23750,
      "token_acc": 0.5089605734767025,
      "train_speed(iter/s)": 1.44447
    },
    {
      "epoch": 1.017737029261814,
      "grad_norm": 4.532026767730713,
      "learning_rate": 9.012081576770757e-05,
      "loss": 2.4489009857177733,
      "memory(GiB)": 77.56,
      "step": 23755,
      "token_acc": 0.49032258064516127,
      "train_speed(iter/s)": 1.444491
    },
    {
      "epoch": 1.017951244591063,
      "grad_norm": 5.434869289398193,
      "learning_rate": 9.011679931359708e-05,
      "loss": 2.395196723937988,
      "memory(GiB)": 77.56,
      "step": 23760,
      "token_acc": 0.47435897435897434,
      "train_speed(iter/s)": 1.444524
    },
    {
      "epoch": 1.0181654599203118,
      "grad_norm": 4.805933475494385,
      "learning_rate": 9.011278213273315e-05,
      "loss": 2.60504264831543,
      "memory(GiB)": 77.56,
      "step": 23765,
      "token_acc": 0.43103448275862066,
      "train_speed(iter/s)": 1.444568
    },
    {
      "epoch": 1.0183796752495609,
      "grad_norm": 4.878146648406982,
      "learning_rate": 9.010876422518854e-05,
      "loss": 2.6618135452270506,
      "memory(GiB)": 77.56,
      "step": 23770,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.444585
    },
    {
      "epoch": 1.01859389057881,
      "grad_norm": 3.9071240425109863,
      "learning_rate": 9.010474559103604e-05,
      "loss": 2.258107566833496,
      "memory(GiB)": 77.56,
      "step": 23775,
      "token_acc": 0.511326860841424,
      "train_speed(iter/s)": 1.444679
    },
    {
      "epoch": 1.0188081059080587,
      "grad_norm": 4.440838813781738,
      "learning_rate": 9.010072623034845e-05,
      "loss": 2.464816093444824,
      "memory(GiB)": 77.56,
      "step": 23780,
      "token_acc": 0.4875,
      "train_speed(iter/s)": 1.444688
    },
    {
      "epoch": 1.0190223212373077,
      "grad_norm": 3.7575438022613525,
      "learning_rate": 9.00967061431986e-05,
      "loss": 2.516987609863281,
      "memory(GiB)": 77.56,
      "step": 23785,
      "token_acc": 0.47039473684210525,
      "train_speed(iter/s)": 1.44467
    },
    {
      "epoch": 1.0192365365665568,
      "grad_norm": 4.53010368347168,
      "learning_rate": 9.009268532965929e-05,
      "loss": 2.556178092956543,
      "memory(GiB)": 77.56,
      "step": 23790,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.444594
    },
    {
      "epoch": 1.0194507518958056,
      "grad_norm": 4.7371602058410645,
      "learning_rate": 9.008866378980338e-05,
      "loss": 2.479389190673828,
      "memory(GiB)": 77.56,
      "step": 23795,
      "token_acc": 0.4885057471264368,
      "train_speed(iter/s)": 1.444564
    },
    {
      "epoch": 1.0196649672250546,
      "grad_norm": 4.864514350891113,
      "learning_rate": 9.008464152370371e-05,
      "loss": 2.3049648284912108,
      "memory(GiB)": 77.56,
      "step": 23800,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.444415
    },
    {
      "epoch": 1.0198791825543037,
      "grad_norm": 4.705252170562744,
      "learning_rate": 9.008061853143318e-05,
      "loss": 2.311351776123047,
      "memory(GiB)": 77.56,
      "step": 23805,
      "token_acc": 0.49624060150375937,
      "train_speed(iter/s)": 1.444413
    },
    {
      "epoch": 1.0200933978835525,
      "grad_norm": 4.033809661865234,
      "learning_rate": 9.00765948130646e-05,
      "loss": 2.330838203430176,
      "memory(GiB)": 77.56,
      "step": 23810,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.44437
    },
    {
      "epoch": 1.0203076132128015,
      "grad_norm": 3.6460118293762207,
      "learning_rate": 9.007257036867095e-05,
      "loss": 2.457318115234375,
      "memory(GiB)": 77.56,
      "step": 23815,
      "token_acc": 0.47774480712166173,
      "train_speed(iter/s)": 1.44434
    },
    {
      "epoch": 1.0205218285420505,
      "grad_norm": 5.905128479003906,
      "learning_rate": 9.006854519832509e-05,
      "loss": 2.585536575317383,
      "memory(GiB)": 77.56,
      "step": 23820,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.444305
    },
    {
      "epoch": 1.0207360438712993,
      "grad_norm": 6.288656711578369,
      "learning_rate": 9.006451930209995e-05,
      "loss": 2.3825435638427734,
      "memory(GiB)": 77.56,
      "step": 23825,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.444299
    },
    {
      "epoch": 1.0209502592005484,
      "grad_norm": 4.655263423919678,
      "learning_rate": 9.006049268006844e-05,
      "loss": 2.332324981689453,
      "memory(GiB)": 77.56,
      "step": 23830,
      "token_acc": 0.4803921568627451,
      "train_speed(iter/s)": 1.444371
    },
    {
      "epoch": 1.0211644745297974,
      "grad_norm": 4.464681625366211,
      "learning_rate": 9.005646533230354e-05,
      "loss": 2.542655372619629,
      "memory(GiB)": 77.56,
      "step": 23835,
      "token_acc": 0.5110132158590308,
      "train_speed(iter/s)": 1.444393
    },
    {
      "epoch": 1.0213786898590462,
      "grad_norm": 5.661242485046387,
      "learning_rate": 9.005243725887819e-05,
      "loss": 2.388264465332031,
      "memory(GiB)": 77.56,
      "step": 23840,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.444441
    },
    {
      "epoch": 1.0215929051882953,
      "grad_norm": 5.917185306549072,
      "learning_rate": 9.004840845986538e-05,
      "loss": 2.3632976531982424,
      "memory(GiB)": 77.56,
      "step": 23845,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.444459
    },
    {
      "epoch": 1.0218071205175443,
      "grad_norm": 4.519986152648926,
      "learning_rate": 9.004437893533807e-05,
      "loss": 2.4975162506103517,
      "memory(GiB)": 77.56,
      "step": 23850,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.444394
    },
    {
      "epoch": 1.022021335846793,
      "grad_norm": 6.680171012878418,
      "learning_rate": 9.004034868536929e-05,
      "loss": 2.448757553100586,
      "memory(GiB)": 77.56,
      "step": 23855,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.444377
    },
    {
      "epoch": 1.0222355511760421,
      "grad_norm": 4.50553560256958,
      "learning_rate": 9.0036317710032e-05,
      "loss": 2.5293701171875,
      "memory(GiB)": 77.56,
      "step": 23860,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.444341
    },
    {
      "epoch": 1.0224497665052912,
      "grad_norm": 4.118928909301758,
      "learning_rate": 9.003228600939926e-05,
      "loss": 2.4206512451171873,
      "memory(GiB)": 77.56,
      "step": 23865,
      "token_acc": 0.4891640866873065,
      "train_speed(iter/s)": 1.444328
    },
    {
      "epoch": 1.02266398183454,
      "grad_norm": 5.375980377197266,
      "learning_rate": 9.002825358354414e-05,
      "loss": 2.9208450317382812,
      "memory(GiB)": 77.56,
      "step": 23870,
      "token_acc": 0.4066666666666667,
      "train_speed(iter/s)": 1.444306
    },
    {
      "epoch": 1.022878197163789,
      "grad_norm": 4.370960235595703,
      "learning_rate": 9.002422043253962e-05,
      "loss": 2.751597595214844,
      "memory(GiB)": 77.56,
      "step": 23875,
      "token_acc": 0.46458923512747874,
      "train_speed(iter/s)": 1.444348
    },
    {
      "epoch": 1.023092412493038,
      "grad_norm": 5.483879566192627,
      "learning_rate": 9.002018655645882e-05,
      "loss": 2.7617847442626955,
      "memory(GiB)": 77.56,
      "step": 23880,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.444382
    },
    {
      "epoch": 1.0233066278222869,
      "grad_norm": 4.41718053817749,
      "learning_rate": 9.00161519553748e-05,
      "loss": 2.233142280578613,
      "memory(GiB)": 77.56,
      "step": 23885,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.444315
    },
    {
      "epoch": 1.023520843151536,
      "grad_norm": 5.621650695800781,
      "learning_rate": 9.001211662936065e-05,
      "loss": 2.3641443252563477,
      "memory(GiB)": 77.56,
      "step": 23890,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.44436
    },
    {
      "epoch": 1.023735058480785,
      "grad_norm": 3.5997681617736816,
      "learning_rate": 9.000808057848946e-05,
      "loss": 2.4787971496582033,
      "memory(GiB)": 77.56,
      "step": 23895,
      "token_acc": 0.4696969696969697,
      "train_speed(iter/s)": 1.444375
    },
    {
      "epoch": 1.0239492738100338,
      "grad_norm": 4.5780839920043945,
      "learning_rate": 9.000404380283435e-05,
      "loss": 2.3750240325927736,
      "memory(GiB)": 77.56,
      "step": 23900,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.444328
    },
    {
      "epoch": 1.0241634891392828,
      "grad_norm": 3.27915358543396,
      "learning_rate": 9.000000630246848e-05,
      "loss": 2.408158302307129,
      "memory(GiB)": 77.56,
      "step": 23905,
      "token_acc": 0.46449704142011833,
      "train_speed(iter/s)": 1.444378
    },
    {
      "epoch": 1.0243777044685318,
      "grad_norm": 4.563897609710693,
      "learning_rate": 8.999596807746497e-05,
      "loss": 2.3405872344970704,
      "memory(GiB)": 77.56,
      "step": 23910,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.444343
    },
    {
      "epoch": 1.0245919197977806,
      "grad_norm": 5.107802391052246,
      "learning_rate": 8.999192912789697e-05,
      "loss": 2.611890983581543,
      "memory(GiB)": 77.56,
      "step": 23915,
      "token_acc": 0.4297752808988764,
      "train_speed(iter/s)": 1.444325
    },
    {
      "epoch": 1.0248061351270297,
      "grad_norm": 4.760217189788818,
      "learning_rate": 8.998788945383768e-05,
      "loss": 2.41015510559082,
      "memory(GiB)": 77.56,
      "step": 23920,
      "token_acc": 0.4892966360856269,
      "train_speed(iter/s)": 1.444407
    },
    {
      "epoch": 1.0250203504562787,
      "grad_norm": 4.066064834594727,
      "learning_rate": 8.998384905536025e-05,
      "loss": 2.623775100708008,
      "memory(GiB)": 77.56,
      "step": 23925,
      "token_acc": 0.48348348348348347,
      "train_speed(iter/s)": 1.444404
    },
    {
      "epoch": 1.0252345657855275,
      "grad_norm": 3.8480799198150635,
      "learning_rate": 8.997980793253789e-05,
      "loss": 2.5957548141479494,
      "memory(GiB)": 77.56,
      "step": 23930,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.444381
    },
    {
      "epoch": 1.0254487811147766,
      "grad_norm": 9.379072189331055,
      "learning_rate": 8.99757660854438e-05,
      "loss": 2.2241458892822266,
      "memory(GiB)": 77.56,
      "step": 23935,
      "token_acc": 0.5703703703703704,
      "train_speed(iter/s)": 1.444317
    },
    {
      "epoch": 1.0256629964440256,
      "grad_norm": 5.639321327209473,
      "learning_rate": 8.99717235141512e-05,
      "loss": 2.548636817932129,
      "memory(GiB)": 77.56,
      "step": 23940,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.444376
    },
    {
      "epoch": 1.0258772117732744,
      "grad_norm": 5.260157585144043,
      "learning_rate": 8.996768021873334e-05,
      "loss": 2.483246421813965,
      "memory(GiB)": 77.56,
      "step": 23945,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.444272
    },
    {
      "epoch": 1.0260914271025234,
      "grad_norm": 5.605607509613037,
      "learning_rate": 8.996363619926346e-05,
      "loss": 2.5595935821533202,
      "memory(GiB)": 77.56,
      "step": 23950,
      "token_acc": 0.48627450980392156,
      "train_speed(iter/s)": 1.444336
    },
    {
      "epoch": 1.0263056424317725,
      "grad_norm": 4.457199573516846,
      "learning_rate": 8.995959145581482e-05,
      "loss": 2.798406219482422,
      "memory(GiB)": 77.56,
      "step": 23955,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.444318
    },
    {
      "epoch": 1.0265198577610213,
      "grad_norm": 4.708622455596924,
      "learning_rate": 8.995554598846069e-05,
      "loss": 2.197768211364746,
      "memory(GiB)": 77.56,
      "step": 23960,
      "token_acc": 0.5282392026578073,
      "train_speed(iter/s)": 1.444219
    },
    {
      "epoch": 1.0267340730902703,
      "grad_norm": 5.235738277435303,
      "learning_rate": 8.995149979727437e-05,
      "loss": 2.715723991394043,
      "memory(GiB)": 77.56,
      "step": 23965,
      "token_acc": 0.47398843930635837,
      "train_speed(iter/s)": 1.444218
    },
    {
      "epoch": 1.0269482884195194,
      "grad_norm": 4.115472793579102,
      "learning_rate": 8.994745288232916e-05,
      "loss": 2.5359683990478517,
      "memory(GiB)": 77.56,
      "step": 23970,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.444185
    },
    {
      "epoch": 1.0271625037487682,
      "grad_norm": 3.066713809967041,
      "learning_rate": 8.994340524369836e-05,
      "loss": 2.552429962158203,
      "memory(GiB)": 77.56,
      "step": 23975,
      "token_acc": 0.42755344418052255,
      "train_speed(iter/s)": 1.444213
    },
    {
      "epoch": 1.0273767190780172,
      "grad_norm": 4.575443744659424,
      "learning_rate": 8.993935688145529e-05,
      "loss": 2.6501426696777344,
      "memory(GiB)": 77.56,
      "step": 23980,
      "token_acc": 0.4550561797752809,
      "train_speed(iter/s)": 1.444278
    },
    {
      "epoch": 1.0275909344072662,
      "grad_norm": 4.8414998054504395,
      "learning_rate": 8.993530779567332e-05,
      "loss": 2.350907897949219,
      "memory(GiB)": 77.56,
      "step": 23985,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.444298
    },
    {
      "epoch": 1.027805149736515,
      "grad_norm": 4.459887504577637,
      "learning_rate": 8.993125798642579e-05,
      "loss": 2.5612808227539063,
      "memory(GiB)": 77.56,
      "step": 23990,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.444328
    },
    {
      "epoch": 1.028019365065764,
      "grad_norm": 4.45468282699585,
      "learning_rate": 8.992720745378605e-05,
      "loss": 2.2972211837768555,
      "memory(GiB)": 77.56,
      "step": 23995,
      "token_acc": 0.5228070175438596,
      "train_speed(iter/s)": 1.444271
    },
    {
      "epoch": 1.0282335803950131,
      "grad_norm": 5.335805892944336,
      "learning_rate": 8.99231561978275e-05,
      "loss": 2.455869674682617,
      "memory(GiB)": 77.56,
      "step": 24000,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.444212
    },
    {
      "epoch": 1.0282335803950131,
      "eval_loss": 2.3627328872680664,
      "eval_runtime": 14.1403,
      "eval_samples_per_second": 7.072,
      "eval_steps_per_second": 7.072,
      "eval_token_acc": 0.44138755980861244,
      "step": 24000
    },
    {
      "epoch": 1.028447795724262,
      "grad_norm": 5.489840984344482,
      "learning_rate": 8.991910421862352e-05,
      "loss": 2.632350730895996,
      "memory(GiB)": 77.56,
      "step": 24005,
      "token_acc": 0.44190140845070425,
      "train_speed(iter/s)": 1.442874
    },
    {
      "epoch": 1.028662011053511,
      "grad_norm": 4.241428852081299,
      "learning_rate": 8.991505151624752e-05,
      "loss": 2.454863166809082,
      "memory(GiB)": 77.56,
      "step": 24010,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.442893
    },
    {
      "epoch": 1.02887622638276,
      "grad_norm": 4.7103776931762695,
      "learning_rate": 8.991099809077292e-05,
      "loss": 2.3978784561157225,
      "memory(GiB)": 77.56,
      "step": 24015,
      "token_acc": 0.44368600682593856,
      "train_speed(iter/s)": 1.442851
    },
    {
      "epoch": 1.0290904417120088,
      "grad_norm": 8.375214576721191,
      "learning_rate": 8.990694394227317e-05,
      "loss": 2.35728759765625,
      "memory(GiB)": 77.56,
      "step": 24020,
      "token_acc": 0.5047169811320755,
      "train_speed(iter/s)": 1.442866
    },
    {
      "epoch": 1.0293046570412578,
      "grad_norm": 5.615808010101318,
      "learning_rate": 8.990288907082168e-05,
      "loss": 2.4102575302124025,
      "memory(GiB)": 77.56,
      "step": 24025,
      "token_acc": 0.5098684210526315,
      "train_speed(iter/s)": 1.44277
    },
    {
      "epoch": 1.0295188723705069,
      "grad_norm": 4.144071578979492,
      "learning_rate": 8.989883347649191e-05,
      "loss": 2.474465560913086,
      "memory(GiB)": 77.56,
      "step": 24030,
      "token_acc": 0.4715447154471545,
      "train_speed(iter/s)": 1.442729
    },
    {
      "epoch": 1.0297330876997557,
      "grad_norm": 6.063302040100098,
      "learning_rate": 8.989477715935735e-05,
      "loss": 2.8166954040527346,
      "memory(GiB)": 77.56,
      "step": 24035,
      "token_acc": 0.42662116040955633,
      "train_speed(iter/s)": 1.442752
    },
    {
      "epoch": 1.0299473030290047,
      "grad_norm": 6.846219062805176,
      "learning_rate": 8.989072011949148e-05,
      "loss": 2.705123710632324,
      "memory(GiB)": 77.56,
      "step": 24040,
      "token_acc": 0.4357142857142857,
      "train_speed(iter/s)": 1.442761
    },
    {
      "epoch": 1.0301615183582538,
      "grad_norm": 4.94010591506958,
      "learning_rate": 8.988666235696779e-05,
      "loss": 2.6302959442138674,
      "memory(GiB)": 77.56,
      "step": 24045,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.442813
    },
    {
      "epoch": 1.0303757336875026,
      "grad_norm": 5.746603012084961,
      "learning_rate": 8.98826038718598e-05,
      "loss": 2.5352577209472655,
      "memory(GiB)": 77.56,
      "step": 24050,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.442859
    },
    {
      "epoch": 1.0305899490167516,
      "grad_norm": 4.987161159515381,
      "learning_rate": 8.987854466424103e-05,
      "loss": 2.4934181213378905,
      "memory(GiB)": 77.56,
      "step": 24055,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.442855
    },
    {
      "epoch": 1.0308041643460006,
      "grad_norm": 4.284734725952148,
      "learning_rate": 8.987448473418502e-05,
      "loss": 2.2293792724609376,
      "memory(GiB)": 77.56,
      "step": 24060,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.442885
    },
    {
      "epoch": 1.0310183796752495,
      "grad_norm": 6.619974136352539,
      "learning_rate": 8.98704240817653e-05,
      "loss": 2.5899566650390624,
      "memory(GiB)": 77.56,
      "step": 24065,
      "token_acc": 0.4749034749034749,
      "train_speed(iter/s)": 1.442843
    },
    {
      "epoch": 1.0312325950044985,
      "grad_norm": 4.957780838012695,
      "learning_rate": 8.986636270705545e-05,
      "loss": 2.292130470275879,
      "memory(GiB)": 77.56,
      "step": 24070,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.442884
    },
    {
      "epoch": 1.0314468103337475,
      "grad_norm": 4.451689720153809,
      "learning_rate": 8.986230061012905e-05,
      "loss": 2.4193578720092774,
      "memory(GiB)": 77.56,
      "step": 24075,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.442871
    },
    {
      "epoch": 1.0316610256629963,
      "grad_norm": 5.041201114654541,
      "learning_rate": 8.985823779105968e-05,
      "loss": 2.5141883850097657,
      "memory(GiB)": 77.56,
      "step": 24080,
      "token_acc": 0.4880952380952381,
      "train_speed(iter/s)": 1.442961
    },
    {
      "epoch": 1.0318752409922454,
      "grad_norm": 5.4844536781311035,
      "learning_rate": 8.985417424992093e-05,
      "loss": 2.4432388305664063,
      "memory(GiB)": 77.56,
      "step": 24085,
      "token_acc": 0.5099337748344371,
      "train_speed(iter/s)": 1.443022
    },
    {
      "epoch": 1.0320894563214944,
      "grad_norm": 4.77213191986084,
      "learning_rate": 8.985010998678642e-05,
      "loss": 2.46121826171875,
      "memory(GiB)": 77.56,
      "step": 24090,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.443072
    },
    {
      "epoch": 1.0323036716507432,
      "grad_norm": 3.9187328815460205,
      "learning_rate": 8.984604500172982e-05,
      "loss": 2.5874811172485352,
      "memory(GiB)": 77.56,
      "step": 24095,
      "token_acc": 0.46176470588235297,
      "train_speed(iter/s)": 1.443101
    },
    {
      "epoch": 1.0325178869799922,
      "grad_norm": 4.982126712799072,
      "learning_rate": 8.984197929482471e-05,
      "loss": 2.457944869995117,
      "memory(GiB)": 77.56,
      "step": 24100,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.443097
    },
    {
      "epoch": 1.0327321023092413,
      "grad_norm": 4.571814060211182,
      "learning_rate": 8.983791286614476e-05,
      "loss": 2.7963314056396484,
      "memory(GiB)": 77.56,
      "step": 24105,
      "token_acc": 0.45871559633027525,
      "train_speed(iter/s)": 1.443159
    },
    {
      "epoch": 1.03294631763849,
      "grad_norm": 4.690274715423584,
      "learning_rate": 8.983384571576367e-05,
      "loss": 2.499724197387695,
      "memory(GiB)": 77.56,
      "step": 24110,
      "token_acc": 0.47653429602888087,
      "train_speed(iter/s)": 1.443229
    },
    {
      "epoch": 1.0331605329677391,
      "grad_norm": 5.090278148651123,
      "learning_rate": 8.98297778437551e-05,
      "loss": 2.5692739486694336,
      "memory(GiB)": 77.56,
      "step": 24115,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.443311
    },
    {
      "epoch": 1.0333747482969882,
      "grad_norm": 4.321795463562012,
      "learning_rate": 8.982570925019273e-05,
      "loss": 2.5763378143310547,
      "memory(GiB)": 77.56,
      "step": 24120,
      "token_acc": 0.4610169491525424,
      "train_speed(iter/s)": 1.443317
    },
    {
      "epoch": 1.0335889636262372,
      "grad_norm": 5.195075035095215,
      "learning_rate": 8.982163993515027e-05,
      "loss": 2.4982065200805663,
      "memory(GiB)": 77.56,
      "step": 24125,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.443351
    },
    {
      "epoch": 1.033803178955486,
      "grad_norm": 4.9751691818237305,
      "learning_rate": 8.981756989870147e-05,
      "loss": 2.4682788848876953,
      "memory(GiB)": 77.56,
      "step": 24130,
      "token_acc": 0.4261168384879725,
      "train_speed(iter/s)": 1.44334
    },
    {
      "epoch": 1.034017394284735,
      "grad_norm": 5.988290309906006,
      "learning_rate": 8.981349914092002e-05,
      "loss": 2.4929752349853516,
      "memory(GiB)": 77.56,
      "step": 24135,
      "token_acc": 0.4979757085020243,
      "train_speed(iter/s)": 1.443408
    },
    {
      "epoch": 1.034231609613984,
      "grad_norm": 5.711163520812988,
      "learning_rate": 8.98094276618797e-05,
      "loss": 2.6479726791381837,
      "memory(GiB)": 77.56,
      "step": 24140,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.443473
    },
    {
      "epoch": 1.034445824943233,
      "grad_norm": 4.263439655303955,
      "learning_rate": 8.980535546165422e-05,
      "loss": 2.4716259002685548,
      "memory(GiB)": 77.56,
      "step": 24145,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.443487
    },
    {
      "epoch": 1.034660040272482,
      "grad_norm": 6.227019309997559,
      "learning_rate": 8.980128254031743e-05,
      "loss": 2.538905715942383,
      "memory(GiB)": 77.56,
      "step": 24150,
      "token_acc": 0.4560669456066946,
      "train_speed(iter/s)": 1.443547
    },
    {
      "epoch": 1.034874255601731,
      "grad_norm": 4.804160118103027,
      "learning_rate": 8.979720889794305e-05,
      "loss": 2.8498817443847657,
      "memory(GiB)": 77.56,
      "step": 24155,
      "token_acc": 0.47244094488188976,
      "train_speed(iter/s)": 1.443538
    },
    {
      "epoch": 1.0350884709309798,
      "grad_norm": 3.615837335586548,
      "learning_rate": 8.979313453460489e-05,
      "loss": 2.546941375732422,
      "memory(GiB)": 77.56,
      "step": 24160,
      "token_acc": 0.4727272727272727,
      "train_speed(iter/s)": 1.443586
    },
    {
      "epoch": 1.0353026862602288,
      "grad_norm": 5.576261043548584,
      "learning_rate": 8.978905945037678e-05,
      "loss": 2.5293659210205077,
      "memory(GiB)": 77.56,
      "step": 24165,
      "token_acc": 0.4968553459119497,
      "train_speed(iter/s)": 1.443642
    },
    {
      "epoch": 1.0355169015894778,
      "grad_norm": 5.04969596862793,
      "learning_rate": 8.978498364533252e-05,
      "loss": 2.2223846435546877,
      "memory(GiB)": 77.56,
      "step": 24170,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.443587
    },
    {
      "epoch": 1.0357311169187267,
      "grad_norm": 3.6998376846313477,
      "learning_rate": 8.978090711954598e-05,
      "loss": 2.5067092895507814,
      "memory(GiB)": 77.56,
      "step": 24175,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.443563
    },
    {
      "epoch": 1.0359453322479757,
      "grad_norm": 5.043964385986328,
      "learning_rate": 8.977682987309097e-05,
      "loss": 2.8575565338134767,
      "memory(GiB)": 77.56,
      "step": 24180,
      "token_acc": 0.4264705882352941,
      "train_speed(iter/s)": 1.443562
    },
    {
      "epoch": 1.0361595475772247,
      "grad_norm": 4.713266849517822,
      "learning_rate": 8.977275190604138e-05,
      "loss": 2.1276174545288087,
      "memory(GiB)": 77.56,
      "step": 24185,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.443596
    },
    {
      "epoch": 1.0363737629064735,
      "grad_norm": 4.897609233856201,
      "learning_rate": 8.97686732184711e-05,
      "loss": 2.3154813766479494,
      "memory(GiB)": 77.56,
      "step": 24190,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.443635
    },
    {
      "epoch": 1.0365879782357226,
      "grad_norm": 4.5343146324157715,
      "learning_rate": 8.976459381045396e-05,
      "loss": 2.355120849609375,
      "memory(GiB)": 77.56,
      "step": 24195,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.443658
    },
    {
      "epoch": 1.0368021935649716,
      "grad_norm": 5.203270435333252,
      "learning_rate": 8.976051368206393e-05,
      "loss": 2.558268356323242,
      "memory(GiB)": 77.56,
      "step": 24200,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.443719
    },
    {
      "epoch": 1.0370164088942204,
      "grad_norm": 5.331142425537109,
      "learning_rate": 8.97564328333749e-05,
      "loss": 2.3839046478271486,
      "memory(GiB)": 77.56,
      "step": 24205,
      "token_acc": 0.49377593360995853,
      "train_speed(iter/s)": 1.443713
    },
    {
      "epoch": 1.0372306242234695,
      "grad_norm": 5.468891143798828,
      "learning_rate": 8.975235126446077e-05,
      "loss": 2.7484081268310545,
      "memory(GiB)": 77.56,
      "step": 24210,
      "token_acc": 0.4253731343283582,
      "train_speed(iter/s)": 1.44375
    },
    {
      "epoch": 1.0374448395527185,
      "grad_norm": 3.892148017883301,
      "learning_rate": 8.974826897539553e-05,
      "loss": 2.881919288635254,
      "memory(GiB)": 77.56,
      "step": 24215,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.443796
    },
    {
      "epoch": 1.0376590548819673,
      "grad_norm": 7.222329616546631,
      "learning_rate": 8.97441859662531e-05,
      "loss": 2.2086765289306642,
      "memory(GiB)": 77.56,
      "step": 24220,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.443854
    },
    {
      "epoch": 1.0378732702112163,
      "grad_norm": 7.090409755706787,
      "learning_rate": 8.974010223710744e-05,
      "loss": 2.5318830490112303,
      "memory(GiB)": 77.56,
      "step": 24225,
      "token_acc": 0.5143769968051118,
      "train_speed(iter/s)": 1.443819
    },
    {
      "epoch": 1.0380874855404654,
      "grad_norm": 4.4375128746032715,
      "learning_rate": 8.973601778803257e-05,
      "loss": 2.4187496185302733,
      "memory(GiB)": 77.56,
      "step": 24230,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.443821
    },
    {
      "epoch": 1.0383017008697142,
      "grad_norm": 5.027166843414307,
      "learning_rate": 8.973193261910247e-05,
      "loss": 2.5862619400024416,
      "memory(GiB)": 77.56,
      "step": 24235,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.44381
    },
    {
      "epoch": 1.0385159161989632,
      "grad_norm": 4.532248497009277,
      "learning_rate": 8.972784673039111e-05,
      "loss": 2.365619087219238,
      "memory(GiB)": 77.56,
      "step": 24240,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.443818
    },
    {
      "epoch": 1.0387301315282123,
      "grad_norm": 5.359230995178223,
      "learning_rate": 8.972376012197256e-05,
      "loss": 2.5995391845703124,
      "memory(GiB)": 77.56,
      "step": 24245,
      "token_acc": 0.44755244755244755,
      "train_speed(iter/s)": 1.443856
    },
    {
      "epoch": 1.038944346857461,
      "grad_norm": 4.708656311035156,
      "learning_rate": 8.971967279392083e-05,
      "loss": 2.489506721496582,
      "memory(GiB)": 77.56,
      "step": 24250,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.443856
    },
    {
      "epoch": 1.03915856218671,
      "grad_norm": 5.532922744750977,
      "learning_rate": 8.971558474630997e-05,
      "loss": 2.671093559265137,
      "memory(GiB)": 77.56,
      "step": 24255,
      "token_acc": 0.490272373540856,
      "train_speed(iter/s)": 1.443923
    },
    {
      "epoch": 1.0393727775159591,
      "grad_norm": 4.693142414093018,
      "learning_rate": 8.971149597921402e-05,
      "loss": 2.520188331604004,
      "memory(GiB)": 77.56,
      "step": 24260,
      "token_acc": 0.46757679180887374,
      "train_speed(iter/s)": 1.443914
    },
    {
      "epoch": 1.039586992845208,
      "grad_norm": 4.718460559844971,
      "learning_rate": 8.970740649270708e-05,
      "loss": 2.177897644042969,
      "memory(GiB)": 77.56,
      "step": 24265,
      "token_acc": 0.5157894736842106,
      "train_speed(iter/s)": 1.443918
    },
    {
      "epoch": 1.039801208174457,
      "grad_norm": 5.367030620574951,
      "learning_rate": 8.970331628686323e-05,
      "loss": 2.6302547454833984,
      "memory(GiB)": 77.56,
      "step": 24270,
      "token_acc": 0.44947735191637633,
      "train_speed(iter/s)": 1.443966
    },
    {
      "epoch": 1.040015423503706,
      "grad_norm": 4.055055141448975,
      "learning_rate": 8.969922536175654e-05,
      "loss": 2.6908987045288084,
      "memory(GiB)": 77.56,
      "step": 24275,
      "token_acc": 0.5131578947368421,
      "train_speed(iter/s)": 1.443992
    },
    {
      "epoch": 1.0402296388329548,
      "grad_norm": 4.85711145401001,
      "learning_rate": 8.969513371746116e-05,
      "loss": 2.621944808959961,
      "memory(GiB)": 77.56,
      "step": 24280,
      "token_acc": 0.4844290657439446,
      "train_speed(iter/s)": 1.444023
    },
    {
      "epoch": 1.0404438541622039,
      "grad_norm": 4.508506774902344,
      "learning_rate": 8.96910413540512e-05,
      "loss": 2.5978229522705076,
      "memory(GiB)": 77.56,
      "step": 24285,
      "token_acc": 0.453416149068323,
      "train_speed(iter/s)": 1.44407
    },
    {
      "epoch": 1.040658069491453,
      "grad_norm": 5.204745769500732,
      "learning_rate": 8.968694827160078e-05,
      "loss": 2.6731306076049806,
      "memory(GiB)": 77.56,
      "step": 24290,
      "token_acc": 0.46226415094339623,
      "train_speed(iter/s)": 1.444053
    },
    {
      "epoch": 1.0408722848207017,
      "grad_norm": 4.502432823181152,
      "learning_rate": 8.968285447018406e-05,
      "loss": 2.2701461791992186,
      "memory(GiB)": 77.56,
      "step": 24295,
      "token_acc": 0.4826254826254826,
      "train_speed(iter/s)": 1.444084
    },
    {
      "epoch": 1.0410865001499507,
      "grad_norm": 5.828651428222656,
      "learning_rate": 8.967875994987521e-05,
      "loss": 2.8625324249267576,
      "memory(GiB)": 77.56,
      "step": 24300,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.444172
    },
    {
      "epoch": 1.0413007154791998,
      "grad_norm": 4.078756809234619,
      "learning_rate": 8.967466471074841e-05,
      "loss": 2.425030517578125,
      "memory(GiB)": 77.56,
      "step": 24305,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.444239
    },
    {
      "epoch": 1.0415149308084486,
      "grad_norm": 6.113180637359619,
      "learning_rate": 8.967056875287783e-05,
      "loss": 2.4928035736083984,
      "memory(GiB)": 77.56,
      "step": 24310,
      "token_acc": 0.4796511627906977,
      "train_speed(iter/s)": 1.444285
    },
    {
      "epoch": 1.0417291461376976,
      "grad_norm": 5.307697296142578,
      "learning_rate": 8.966647207633769e-05,
      "loss": 2.570161056518555,
      "memory(GiB)": 77.56,
      "step": 24315,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.444336
    },
    {
      "epoch": 1.0419433614669467,
      "grad_norm": 5.691319942474365,
      "learning_rate": 8.966237468120218e-05,
      "loss": 2.6208337783813476,
      "memory(GiB)": 77.56,
      "step": 24320,
      "token_acc": 0.440625,
      "train_speed(iter/s)": 1.444402
    },
    {
      "epoch": 1.0421575767961955,
      "grad_norm": 12.893251419067383,
      "learning_rate": 8.965827656754557e-05,
      "loss": 2.638018035888672,
      "memory(GiB)": 77.56,
      "step": 24325,
      "token_acc": 0.4483870967741935,
      "train_speed(iter/s)": 1.444417
    },
    {
      "epoch": 1.0423717921254445,
      "grad_norm": 5.798929214477539,
      "learning_rate": 8.965417773544207e-05,
      "loss": 2.61107177734375,
      "memory(GiB)": 77.56,
      "step": 24330,
      "token_acc": 0.44974874371859297,
      "train_speed(iter/s)": 1.444426
    },
    {
      "epoch": 1.0425860074546935,
      "grad_norm": 6.788809776306152,
      "learning_rate": 8.965007818496593e-05,
      "loss": 2.549547576904297,
      "memory(GiB)": 77.56,
      "step": 24335,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.444387
    },
    {
      "epoch": 1.0428002227839424,
      "grad_norm": 4.759422779083252,
      "learning_rate": 8.964597791619145e-05,
      "loss": 2.246558952331543,
      "memory(GiB)": 77.56,
      "step": 24340,
      "token_acc": 0.5192307692307693,
      "train_speed(iter/s)": 1.444361
    },
    {
      "epoch": 1.0430144381131914,
      "grad_norm": 7.374278545379639,
      "learning_rate": 8.964187692919288e-05,
      "loss": 2.699236297607422,
      "memory(GiB)": 77.56,
      "step": 24345,
      "token_acc": 0.4305555555555556,
      "train_speed(iter/s)": 1.444368
    },
    {
      "epoch": 1.0432286534424404,
      "grad_norm": 5.6786885261535645,
      "learning_rate": 8.963777522404451e-05,
      "loss": 2.6421829223632813,
      "memory(GiB)": 77.56,
      "step": 24350,
      "token_acc": 0.4651898734177215,
      "train_speed(iter/s)": 1.44443
    },
    {
      "epoch": 1.0434428687716892,
      "grad_norm": 4.032026290893555,
      "learning_rate": 8.963367280082067e-05,
      "loss": 2.542629623413086,
      "memory(GiB)": 77.56,
      "step": 24355,
      "token_acc": 0.4820846905537459,
      "train_speed(iter/s)": 1.444406
    },
    {
      "epoch": 1.0436570841009383,
      "grad_norm": 4.686112403869629,
      "learning_rate": 8.962956965959568e-05,
      "loss": 2.6689868927001954,
      "memory(GiB)": 77.56,
      "step": 24360,
      "token_acc": 0.46688741721854304,
      "train_speed(iter/s)": 1.444383
    },
    {
      "epoch": 1.0438712994301873,
      "grad_norm": 4.627802848815918,
      "learning_rate": 8.962546580044384e-05,
      "loss": 2.7924173355102537,
      "memory(GiB)": 77.56,
      "step": 24365,
      "token_acc": 0.459375,
      "train_speed(iter/s)": 1.44439
    },
    {
      "epoch": 1.0440855147594361,
      "grad_norm": 5.952791213989258,
      "learning_rate": 8.962136122343952e-05,
      "loss": 2.4013442993164062,
      "memory(GiB)": 77.56,
      "step": 24370,
      "token_acc": 0.48638132295719844,
      "train_speed(iter/s)": 1.444436
    },
    {
      "epoch": 1.0442997300886852,
      "grad_norm": 6.701351642608643,
      "learning_rate": 8.961725592865708e-05,
      "loss": 2.376656341552734,
      "memory(GiB)": 77.56,
      "step": 24375,
      "token_acc": 0.5379061371841155,
      "train_speed(iter/s)": 1.444398
    },
    {
      "epoch": 1.0445139454179342,
      "grad_norm": 4.986407279968262,
      "learning_rate": 8.961314991617089e-05,
      "loss": 2.6158611297607424,
      "memory(GiB)": 77.56,
      "step": 24380,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.444423
    },
    {
      "epoch": 1.044728160747183,
      "grad_norm": 5.211883068084717,
      "learning_rate": 8.960904318605532e-05,
      "loss": 2.3534502029418944,
      "memory(GiB)": 77.56,
      "step": 24385,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.444372
    },
    {
      "epoch": 1.044942376076432,
      "grad_norm": 4.04542350769043,
      "learning_rate": 8.960493573838477e-05,
      "loss": 2.2348995208740234,
      "memory(GiB)": 77.56,
      "step": 24390,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.444369
    },
    {
      "epoch": 1.045156591405681,
      "grad_norm": 7.3191680908203125,
      "learning_rate": 8.960082757323367e-05,
      "loss": 2.8703903198242187,
      "memory(GiB)": 77.56,
      "step": 24395,
      "token_acc": 0.44329896907216493,
      "train_speed(iter/s)": 1.444309
    },
    {
      "epoch": 1.0453708067349299,
      "grad_norm": 5.509588241577148,
      "learning_rate": 8.95967186906764e-05,
      "loss": 2.4106178283691406,
      "memory(GiB)": 77.56,
      "step": 24400,
      "token_acc": 0.5019305019305019,
      "train_speed(iter/s)": 1.444368
    },
    {
      "epoch": 1.045585022064179,
      "grad_norm": 4.54595422744751,
      "learning_rate": 8.959260909078746e-05,
      "loss": 2.410841941833496,
      "memory(GiB)": 77.56,
      "step": 24405,
      "token_acc": 0.47794117647058826,
      "train_speed(iter/s)": 1.444406
    },
    {
      "epoch": 1.045799237393428,
      "grad_norm": 5.7470245361328125,
      "learning_rate": 8.958849877364125e-05,
      "loss": 2.4927711486816406,
      "memory(GiB)": 77.56,
      "step": 24410,
      "token_acc": 0.5241635687732342,
      "train_speed(iter/s)": 1.444334
    },
    {
      "epoch": 1.0460134527226768,
      "grad_norm": 6.178435325622559,
      "learning_rate": 8.958438773931226e-05,
      "loss": 2.7552379608154296,
      "memory(GiB)": 77.56,
      "step": 24415,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.444271
    },
    {
      "epoch": 1.0462276680519258,
      "grad_norm": 4.371453285217285,
      "learning_rate": 8.958027598787495e-05,
      "loss": 2.5143932342529296,
      "memory(GiB)": 77.56,
      "step": 24420,
      "token_acc": 0.48656716417910445,
      "train_speed(iter/s)": 1.444289
    },
    {
      "epoch": 1.0464418833811748,
      "grad_norm": 4.511281490325928,
      "learning_rate": 8.957616351940381e-05,
      "loss": 2.380167770385742,
      "memory(GiB)": 77.56,
      "step": 24425,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.444221
    },
    {
      "epoch": 1.0466560987104236,
      "grad_norm": 6.007720470428467,
      "learning_rate": 8.957205033397333e-05,
      "loss": 2.4257295608520506,
      "memory(GiB)": 77.56,
      "step": 24430,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.444288
    },
    {
      "epoch": 1.0468703140396727,
      "grad_norm": 5.955282688140869,
      "learning_rate": 8.956793643165805e-05,
      "loss": 2.946666145324707,
      "memory(GiB)": 77.56,
      "step": 24435,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.444285
    },
    {
      "epoch": 1.0470845293689217,
      "grad_norm": 4.111509323120117,
      "learning_rate": 8.956382181253248e-05,
      "loss": 2.391428565979004,
      "memory(GiB)": 77.56,
      "step": 24440,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.444325
    },
    {
      "epoch": 1.0472987446981705,
      "grad_norm": 6.420551776885986,
      "learning_rate": 8.955970647667115e-05,
      "loss": 2.5946741104125977,
      "memory(GiB)": 77.56,
      "step": 24445,
      "token_acc": 0.45703125,
      "train_speed(iter/s)": 1.444409
    },
    {
      "epoch": 1.0475129600274196,
      "grad_norm": 5.487607002258301,
      "learning_rate": 8.955559042414865e-05,
      "loss": 2.655599021911621,
      "memory(GiB)": 77.56,
      "step": 24450,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.444335
    },
    {
      "epoch": 1.0477271753566686,
      "grad_norm": 4.227954864501953,
      "learning_rate": 8.955147365503952e-05,
      "loss": 2.5085376739501952,
      "memory(GiB)": 77.56,
      "step": 24455,
      "token_acc": 0.4944649446494465,
      "train_speed(iter/s)": 1.444386
    },
    {
      "epoch": 1.0479413906859174,
      "grad_norm": 6.8962531089782715,
      "learning_rate": 8.954735616941834e-05,
      "loss": 2.7214939117431642,
      "memory(GiB)": 77.56,
      "step": 24460,
      "token_acc": 0.4671814671814672,
      "train_speed(iter/s)": 1.444304
    },
    {
      "epoch": 1.0481556060151664,
      "grad_norm": 5.080885410308838,
      "learning_rate": 8.954323796735969e-05,
      "loss": 2.4229759216308593,
      "memory(GiB)": 77.56,
      "step": 24465,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.444313
    },
    {
      "epoch": 1.0483698213444155,
      "grad_norm": 5.274581432342529,
      "learning_rate": 8.953911904893821e-05,
      "loss": 2.6455875396728517,
      "memory(GiB)": 77.56,
      "step": 24470,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.444352
    },
    {
      "epoch": 1.0485840366736643,
      "grad_norm": 4.01211404800415,
      "learning_rate": 8.953499941422849e-05,
      "loss": 2.657146453857422,
      "memory(GiB)": 77.56,
      "step": 24475,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.444376
    },
    {
      "epoch": 1.0487982520029133,
      "grad_norm": 4.258591651916504,
      "learning_rate": 8.953087906330514e-05,
      "loss": 2.4957908630371093,
      "memory(GiB)": 77.56,
      "step": 24480,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.444377
    },
    {
      "epoch": 1.0490124673321624,
      "grad_norm": 7.305221080780029,
      "learning_rate": 8.952675799624288e-05,
      "loss": 2.6496051788330077,
      "memory(GiB)": 77.56,
      "step": 24485,
      "token_acc": 0.4859437751004016,
      "train_speed(iter/s)": 1.444378
    },
    {
      "epoch": 1.0492266826614112,
      "grad_norm": 5.413460731506348,
      "learning_rate": 8.952263621311629e-05,
      "loss": 2.6764875411987306,
      "memory(GiB)": 77.56,
      "step": 24490,
      "token_acc": 0.46994535519125685,
      "train_speed(iter/s)": 1.444334
    },
    {
      "epoch": 1.0494408979906602,
      "grad_norm": 4.029872417449951,
      "learning_rate": 8.951851371400008e-05,
      "loss": 2.7449920654296873,
      "memory(GiB)": 77.56,
      "step": 24495,
      "token_acc": 0.4066265060240964,
      "train_speed(iter/s)": 1.444372
    },
    {
      "epoch": 1.0496551133199092,
      "grad_norm": 5.372546672821045,
      "learning_rate": 8.951439049896892e-05,
      "loss": 2.267522430419922,
      "memory(GiB)": 77.56,
      "step": 24500,
      "token_acc": 0.525691699604743,
      "train_speed(iter/s)": 1.444442
    },
    {
      "epoch": 1.0496551133199092,
      "eval_loss": 2.305422306060791,
      "eval_runtime": 14.6397,
      "eval_samples_per_second": 6.831,
      "eval_steps_per_second": 6.831,
      "eval_token_acc": 0.45632333767926986,
      "step": 24500
    },
    {
      "epoch": 1.049869328649158,
      "grad_norm": 4.398870468139648,
      "learning_rate": 8.951026656809753e-05,
      "loss": 2.537317657470703,
      "memory(GiB)": 77.56,
      "step": 24505,
      "token_acc": 0.4588785046728972,
      "train_speed(iter/s)": 1.443118
    },
    {
      "epoch": 1.050083543978407,
      "grad_norm": 8.509885787963867,
      "learning_rate": 8.950614192146058e-05,
      "loss": 2.567396545410156,
      "memory(GiB)": 77.56,
      "step": 24510,
      "token_acc": 0.4341736694677871,
      "train_speed(iter/s)": 1.443114
    },
    {
      "epoch": 1.0502977593076561,
      "grad_norm": 4.17772912979126,
      "learning_rate": 8.950201655913279e-05,
      "loss": 2.3661081314086916,
      "memory(GiB)": 77.56,
      "step": 24515,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.443044
    },
    {
      "epoch": 1.050511974636905,
      "grad_norm": 5.490793704986572,
      "learning_rate": 8.949789048118894e-05,
      "loss": 2.191828155517578,
      "memory(GiB)": 77.56,
      "step": 24520,
      "token_acc": 0.5168539325842697,
      "train_speed(iter/s)": 1.443072
    },
    {
      "epoch": 1.050726189966154,
      "grad_norm": 4.027836799621582,
      "learning_rate": 8.949376368770375e-05,
      "loss": 2.703683090209961,
      "memory(GiB)": 77.56,
      "step": 24525,
      "token_acc": 0.43214285714285716,
      "train_speed(iter/s)": 1.443033
    },
    {
      "epoch": 1.050940405295403,
      "grad_norm": 4.876866817474365,
      "learning_rate": 8.948963617875197e-05,
      "loss": 2.6705379486083984,
      "memory(GiB)": 77.56,
      "step": 24530,
      "token_acc": 0.4027027027027027,
      "train_speed(iter/s)": 1.442934
    },
    {
      "epoch": 1.0511546206246518,
      "grad_norm": 5.685788154602051,
      "learning_rate": 8.94855079544084e-05,
      "loss": 2.6842384338378906,
      "memory(GiB)": 77.56,
      "step": 24535,
      "token_acc": 0.4358974358974359,
      "train_speed(iter/s)": 1.442982
    },
    {
      "epoch": 1.0513688359539008,
      "grad_norm": 4.552136421203613,
      "learning_rate": 8.94813790147478e-05,
      "loss": 2.4423492431640623,
      "memory(GiB)": 77.56,
      "step": 24540,
      "token_acc": 0.4618181818181818,
      "train_speed(iter/s)": 1.442892
    },
    {
      "epoch": 1.0515830512831499,
      "grad_norm": 4.318337440490723,
      "learning_rate": 8.947724935984499e-05,
      "loss": 2.3700738906860352,
      "memory(GiB)": 77.56,
      "step": 24545,
      "token_acc": 0.5468164794007491,
      "train_speed(iter/s)": 1.442868
    },
    {
      "epoch": 1.0517972666123987,
      "grad_norm": 3.6966564655303955,
      "learning_rate": 8.947311898977477e-05,
      "loss": 2.5516988754272463,
      "memory(GiB)": 77.56,
      "step": 24550,
      "token_acc": 0.4629080118694362,
      "train_speed(iter/s)": 1.442871
    },
    {
      "epoch": 1.0520114819416477,
      "grad_norm": 9.808146476745605,
      "learning_rate": 8.946898790461197e-05,
      "loss": 2.388142395019531,
      "memory(GiB)": 77.56,
      "step": 24555,
      "token_acc": 0.5086505190311419,
      "train_speed(iter/s)": 1.442914
    },
    {
      "epoch": 1.0522256972708968,
      "grad_norm": 4.863648414611816,
      "learning_rate": 8.946485610443144e-05,
      "loss": 2.265351676940918,
      "memory(GiB)": 77.56,
      "step": 24560,
      "token_acc": 0.5105740181268882,
      "train_speed(iter/s)": 1.442929
    },
    {
      "epoch": 1.0524399126001456,
      "grad_norm": 8.26724624633789,
      "learning_rate": 8.946072358930802e-05,
      "loss": 2.5950172424316404,
      "memory(GiB)": 77.56,
      "step": 24565,
      "token_acc": 0.49828178694158076,
      "train_speed(iter/s)": 1.442957
    },
    {
      "epoch": 1.0526541279293946,
      "grad_norm": 5.3287506103515625,
      "learning_rate": 8.945659035931656e-05,
      "loss": 2.8080123901367187,
      "memory(GiB)": 77.56,
      "step": 24570,
      "token_acc": 0.41007194244604317,
      "train_speed(iter/s)": 1.442987
    },
    {
      "epoch": 1.0528683432586436,
      "grad_norm": 4.792881965637207,
      "learning_rate": 8.945245641453197e-05,
      "loss": 2.601020622253418,
      "memory(GiB)": 77.56,
      "step": 24575,
      "token_acc": 0.4560810810810811,
      "train_speed(iter/s)": 1.442946
    },
    {
      "epoch": 1.0530825585878925,
      "grad_norm": 3.9182608127593994,
      "learning_rate": 8.94483217550291e-05,
      "loss": 2.4217052459716797,
      "memory(GiB)": 77.56,
      "step": 24580,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.442966
    },
    {
      "epoch": 1.0532967739171415,
      "grad_norm": 4.0062642097473145,
      "learning_rate": 8.94441863808829e-05,
      "loss": 2.3591335296630858,
      "memory(GiB)": 77.56,
      "step": 24585,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.442958
    },
    {
      "epoch": 1.0535109892463905,
      "grad_norm": 4.292083263397217,
      "learning_rate": 8.944005029216824e-05,
      "loss": 2.4691034317016602,
      "memory(GiB)": 77.56,
      "step": 24590,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.442978
    },
    {
      "epoch": 1.0537252045756393,
      "grad_norm": 6.970109939575195,
      "learning_rate": 8.94359134889601e-05,
      "loss": 2.562367057800293,
      "memory(GiB)": 77.56,
      "step": 24595,
      "token_acc": 0.4866920152091255,
      "train_speed(iter/s)": 1.442999
    },
    {
      "epoch": 1.0539394199048884,
      "grad_norm": 4.795201778411865,
      "learning_rate": 8.943177597133336e-05,
      "loss": 2.4861894607543946,
      "memory(GiB)": 77.56,
      "step": 24600,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.443044
    },
    {
      "epoch": 1.0541536352341374,
      "grad_norm": 5.581934452056885,
      "learning_rate": 8.942763773936304e-05,
      "loss": 2.726054382324219,
      "memory(GiB)": 77.56,
      "step": 24605,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.443053
    },
    {
      "epoch": 1.0543678505633862,
      "grad_norm": 5.570984363555908,
      "learning_rate": 8.942349879312406e-05,
      "loss": 2.555591011047363,
      "memory(GiB)": 77.56,
      "step": 24610,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.443051
    },
    {
      "epoch": 1.0545820658926353,
      "grad_norm": 14.814002990722656,
      "learning_rate": 8.941935913269142e-05,
      "loss": 2.2190454483032225,
      "memory(GiB)": 77.56,
      "step": 24615,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.44301
    },
    {
      "epoch": 1.0547962812218843,
      "grad_norm": 5.424521446228027,
      "learning_rate": 8.94152187581401e-05,
      "loss": 2.433231163024902,
      "memory(GiB)": 77.56,
      "step": 24620,
      "token_acc": 0.45980707395498394,
      "train_speed(iter/s)": 1.442938
    },
    {
      "epoch": 1.055010496551133,
      "grad_norm": 4.3254899978637695,
      "learning_rate": 8.941107766954513e-05,
      "loss": 2.5708267211914064,
      "memory(GiB)": 77.56,
      "step": 24625,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.442888
    },
    {
      "epoch": 1.0552247118803821,
      "grad_norm": 4.32937479019165,
      "learning_rate": 8.940693586698152e-05,
      "loss": 2.3701637268066404,
      "memory(GiB)": 77.56,
      "step": 24630,
      "token_acc": 0.5152838427947598,
      "train_speed(iter/s)": 1.442936
    },
    {
      "epoch": 1.0554389272096312,
      "grad_norm": 6.3104448318481445,
      "learning_rate": 8.940279335052428e-05,
      "loss": 2.2762271881103517,
      "memory(GiB)": 77.56,
      "step": 24635,
      "token_acc": 0.5366795366795367,
      "train_speed(iter/s)": 1.442905
    },
    {
      "epoch": 1.05565314253888,
      "grad_norm": 6.276554107666016,
      "learning_rate": 8.939865012024849e-05,
      "loss": 2.366049575805664,
      "memory(GiB)": 77.56,
      "step": 24640,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.442926
    },
    {
      "epoch": 1.055867357868129,
      "grad_norm": 5.921716690063477,
      "learning_rate": 8.939450617622919e-05,
      "loss": 2.9845218658447266,
      "memory(GiB)": 77.56,
      "step": 24645,
      "token_acc": 0.4523076923076923,
      "train_speed(iter/s)": 1.44301
    },
    {
      "epoch": 1.056081573197378,
      "grad_norm": 5.03089714050293,
      "learning_rate": 8.939036151854147e-05,
      "loss": 2.241654968261719,
      "memory(GiB)": 77.56,
      "step": 24650,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.442991
    },
    {
      "epoch": 1.0562957885266269,
      "grad_norm": 5.07546854019165,
      "learning_rate": 8.93862161472604e-05,
      "loss": 2.3906084060668946,
      "memory(GiB)": 77.56,
      "step": 24655,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.442992
    },
    {
      "epoch": 1.056510003855876,
      "grad_norm": 4.440848350524902,
      "learning_rate": 8.938207006246106e-05,
      "loss": 2.414600944519043,
      "memory(GiB)": 77.56,
      "step": 24660,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.442964
    },
    {
      "epoch": 1.056724219185125,
      "grad_norm": 4.9508056640625,
      "learning_rate": 8.937792326421859e-05,
      "loss": 2.9339134216308596,
      "memory(GiB)": 77.56,
      "step": 24665,
      "token_acc": 0.42356687898089174,
      "train_speed(iter/s)": 1.442945
    },
    {
      "epoch": 1.0569384345143737,
      "grad_norm": 5.229337692260742,
      "learning_rate": 8.937377575260809e-05,
      "loss": 2.421574592590332,
      "memory(GiB)": 77.56,
      "step": 24670,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.442975
    },
    {
      "epoch": 1.0571526498436228,
      "grad_norm": 4.360829830169678,
      "learning_rate": 8.93696275277047e-05,
      "loss": 2.466918182373047,
      "memory(GiB)": 77.56,
      "step": 24675,
      "token_acc": 0.4425770308123249,
      "train_speed(iter/s)": 1.442959
    },
    {
      "epoch": 1.0573668651728718,
      "grad_norm": 3.5781688690185547,
      "learning_rate": 8.93654785895836e-05,
      "loss": 2.5756629943847655,
      "memory(GiB)": 77.56,
      "step": 24680,
      "token_acc": 0.456973293768546,
      "train_speed(iter/s)": 1.443027
    },
    {
      "epoch": 1.0575810805021206,
      "grad_norm": 4.3807902336120605,
      "learning_rate": 8.936132893831992e-05,
      "loss": 2.5636844635009766,
      "memory(GiB)": 77.56,
      "step": 24685,
      "token_acc": 0.44072948328267475,
      "train_speed(iter/s)": 1.443009
    },
    {
      "epoch": 1.0577952958313697,
      "grad_norm": 3.8716936111450195,
      "learning_rate": 8.935717857398883e-05,
      "loss": 2.277414321899414,
      "memory(GiB)": 77.56,
      "step": 24690,
      "token_acc": 0.5284280936454849,
      "train_speed(iter/s)": 1.443041
    },
    {
      "epoch": 1.0580095111606187,
      "grad_norm": 5.007645130157471,
      "learning_rate": 8.935302749666554e-05,
      "loss": 2.5836578369140626,
      "memory(GiB)": 77.56,
      "step": 24695,
      "token_acc": 0.4591194968553459,
      "train_speed(iter/s)": 1.443027
    },
    {
      "epoch": 1.0582237264898675,
      "grad_norm": 3.8855621814727783,
      "learning_rate": 8.934887570642523e-05,
      "loss": 2.437145233154297,
      "memory(GiB)": 77.56,
      "step": 24700,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.443057
    },
    {
      "epoch": 1.0584379418191165,
      "grad_norm": 3.954162836074829,
      "learning_rate": 8.934472320334312e-05,
      "loss": 2.5538406372070312,
      "memory(GiB)": 77.56,
      "step": 24705,
      "token_acc": 0.4541984732824427,
      "train_speed(iter/s)": 1.443058
    },
    {
      "epoch": 1.0586521571483656,
      "grad_norm": 5.068463325500488,
      "learning_rate": 8.934056998749447e-05,
      "loss": 2.665746307373047,
      "memory(GiB)": 77.56,
      "step": 24710,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.443003
    },
    {
      "epoch": 1.0588663724776144,
      "grad_norm": 4.683329105377197,
      "learning_rate": 8.933641605895447e-05,
      "loss": 2.548792266845703,
      "memory(GiB)": 77.56,
      "step": 24715,
      "token_acc": 0.4307228915662651,
      "train_speed(iter/s)": 1.443016
    },
    {
      "epoch": 1.0590805878068634,
      "grad_norm": 5.163784503936768,
      "learning_rate": 8.93322614177984e-05,
      "loss": 2.2101043701171874,
      "memory(GiB)": 77.56,
      "step": 24720,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.443062
    },
    {
      "epoch": 1.0592948031361125,
      "grad_norm": 5.771228313446045,
      "learning_rate": 8.932810606410151e-05,
      "loss": 2.5359218597412108,
      "memory(GiB)": 77.56,
      "step": 24725,
      "token_acc": 0.46099290780141844,
      "train_speed(iter/s)": 1.443003
    },
    {
      "epoch": 1.0595090184653613,
      "grad_norm": 5.5019097328186035,
      "learning_rate": 8.932394999793909e-05,
      "loss": 2.2854108810424805,
      "memory(GiB)": 77.56,
      "step": 24730,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.443001
    },
    {
      "epoch": 1.0597232337946103,
      "grad_norm": 4.975171089172363,
      "learning_rate": 8.931979321938643e-05,
      "loss": 2.670800971984863,
      "memory(GiB)": 77.56,
      "step": 24735,
      "token_acc": 0.43260188087774293,
      "train_speed(iter/s)": 1.443022
    },
    {
      "epoch": 1.0599374491238593,
      "grad_norm": 5.276051998138428,
      "learning_rate": 8.931563572851883e-05,
      "loss": 2.4296491622924803,
      "memory(GiB)": 77.56,
      "step": 24740,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.443028
    },
    {
      "epoch": 1.0601516644531084,
      "grad_norm": 4.11514949798584,
      "learning_rate": 8.93114775254116e-05,
      "loss": 2.6960317611694338,
      "memory(GiB)": 77.56,
      "step": 24745,
      "token_acc": 0.4251497005988024,
      "train_speed(iter/s)": 1.443002
    },
    {
      "epoch": 1.0603658797823572,
      "grad_norm": 6.036252021789551,
      "learning_rate": 8.930731861014009e-05,
      "loss": 2.7282039642333986,
      "memory(GiB)": 77.56,
      "step": 24750,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.442955
    },
    {
      "epoch": 1.0605800951116062,
      "grad_norm": 4.531759738922119,
      "learning_rate": 8.930315898277961e-05,
      "loss": 2.4699546813964846,
      "memory(GiB)": 77.56,
      "step": 24755,
      "token_acc": 0.44516129032258067,
      "train_speed(iter/s)": 1.442947
    },
    {
      "epoch": 1.0607943104408553,
      "grad_norm": 3.531190872192383,
      "learning_rate": 8.929899864340556e-05,
      "loss": 2.17702579498291,
      "memory(GiB)": 77.56,
      "step": 24760,
      "token_acc": 0.45294117647058824,
      "train_speed(iter/s)": 1.44294
    },
    {
      "epoch": 1.061008525770104,
      "grad_norm": 4.449081897735596,
      "learning_rate": 8.929483759209327e-05,
      "loss": 2.5497406005859373,
      "memory(GiB)": 77.56,
      "step": 24765,
      "token_acc": 0.4649122807017544,
      "train_speed(iter/s)": 1.442972
    },
    {
      "epoch": 1.061222741099353,
      "grad_norm": 4.878509998321533,
      "learning_rate": 8.929067582891813e-05,
      "loss": 2.5043004989624023,
      "memory(GiB)": 77.56,
      "step": 24770,
      "token_acc": 0.4724137931034483,
      "train_speed(iter/s)": 1.443028
    },
    {
      "epoch": 1.0614369564286021,
      "grad_norm": 5.615623474121094,
      "learning_rate": 8.928651335395556e-05,
      "loss": 2.7244335174560548,
      "memory(GiB)": 77.56,
      "step": 24775,
      "token_acc": 0.45294117647058824,
      "train_speed(iter/s)": 1.44308
    },
    {
      "epoch": 1.061651171757851,
      "grad_norm": 5.27921199798584,
      "learning_rate": 8.928235016728093e-05,
      "loss": 2.484927177429199,
      "memory(GiB)": 77.56,
      "step": 24780,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.443092
    },
    {
      "epoch": 1.0618653870871,
      "grad_norm": 4.431709289550781,
      "learning_rate": 8.927818626896968e-05,
      "loss": 2.606571578979492,
      "memory(GiB)": 77.56,
      "step": 24785,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.443143
    },
    {
      "epoch": 1.062079602416349,
      "grad_norm": 4.658729553222656,
      "learning_rate": 8.927402165909724e-05,
      "loss": 2.493944549560547,
      "memory(GiB)": 77.56,
      "step": 24790,
      "token_acc": 0.4879032258064516,
      "train_speed(iter/s)": 1.443176
    },
    {
      "epoch": 1.0622938177455978,
      "grad_norm": 4.7193074226379395,
      "learning_rate": 8.926985633773906e-05,
      "loss": 2.551185417175293,
      "memory(GiB)": 77.56,
      "step": 24795,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.443219
    },
    {
      "epoch": 1.0625080330748469,
      "grad_norm": 4.909816265106201,
      "learning_rate": 8.92656903049706e-05,
      "loss": 2.3404747009277345,
      "memory(GiB)": 77.56,
      "step": 24800,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.44317
    },
    {
      "epoch": 1.062722248404096,
      "grad_norm": 6.403280258178711,
      "learning_rate": 8.92615235608673e-05,
      "loss": 2.562880516052246,
      "memory(GiB)": 77.56,
      "step": 24805,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.443188
    },
    {
      "epoch": 1.0629364637333447,
      "grad_norm": 4.773261070251465,
      "learning_rate": 8.925735610550469e-05,
      "loss": 2.6108932495117188,
      "memory(GiB)": 77.56,
      "step": 24810,
      "token_acc": 0.4860335195530726,
      "train_speed(iter/s)": 1.44323
    },
    {
      "epoch": 1.0631506790625938,
      "grad_norm": 8.39279556274414,
      "learning_rate": 8.925318793895825e-05,
      "loss": 2.3992229461669923,
      "memory(GiB)": 77.56,
      "step": 24815,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.443239
    },
    {
      "epoch": 1.0633648943918428,
      "grad_norm": 6.312216281890869,
      "learning_rate": 8.924901906130348e-05,
      "loss": 2.4384105682373045,
      "memory(GiB)": 77.56,
      "step": 24820,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.443176
    },
    {
      "epoch": 1.0635791097210916,
      "grad_norm": 5.021091461181641,
      "learning_rate": 8.924484947261592e-05,
      "loss": 2.6059885025024414,
      "memory(GiB)": 77.56,
      "step": 24825,
      "token_acc": 0.445141065830721,
      "train_speed(iter/s)": 1.443204
    },
    {
      "epoch": 1.0637933250503406,
      "grad_norm": 4.153078079223633,
      "learning_rate": 8.92406791729711e-05,
      "loss": 2.1656402587890624,
      "memory(GiB)": 77.56,
      "step": 24830,
      "token_acc": 0.4952978056426332,
      "train_speed(iter/s)": 1.443096
    },
    {
      "epoch": 1.0640075403795897,
      "grad_norm": 5.095820426940918,
      "learning_rate": 8.923650816244455e-05,
      "loss": 2.668680763244629,
      "memory(GiB)": 77.56,
      "step": 24835,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.443129
    },
    {
      "epoch": 1.0642217557088385,
      "grad_norm": 4.86871337890625,
      "learning_rate": 8.923233644111187e-05,
      "loss": 2.8476835250854493,
      "memory(GiB)": 77.56,
      "step": 24840,
      "token_acc": 0.44554455445544555,
      "train_speed(iter/s)": 1.443135
    },
    {
      "epoch": 1.0644359710380875,
      "grad_norm": 6.836292743682861,
      "learning_rate": 8.922816400904859e-05,
      "loss": 2.5800466537475586,
      "memory(GiB)": 77.56,
      "step": 24845,
      "token_acc": 0.46441947565543074,
      "train_speed(iter/s)": 1.443094
    },
    {
      "epoch": 1.0646501863673365,
      "grad_norm": 5.393304824829102,
      "learning_rate": 8.922399086633032e-05,
      "loss": 2.926360321044922,
      "memory(GiB)": 77.56,
      "step": 24850,
      "token_acc": 0.4045307443365696,
      "train_speed(iter/s)": 1.443084
    },
    {
      "epoch": 1.0648644016965854,
      "grad_norm": 5.908994197845459,
      "learning_rate": 8.921981701303267e-05,
      "loss": 2.4802446365356445,
      "memory(GiB)": 77.56,
      "step": 24855,
      "token_acc": 0.46311475409836067,
      "train_speed(iter/s)": 1.443036
    },
    {
      "epoch": 1.0650786170258344,
      "grad_norm": 3.7969794273376465,
      "learning_rate": 8.921564244923124e-05,
      "loss": 2.3489316940307616,
      "memory(GiB)": 77.56,
      "step": 24860,
      "token_acc": 0.5183823529411765,
      "train_speed(iter/s)": 1.442975
    },
    {
      "epoch": 1.0652928323550834,
      "grad_norm": 4.436620712280273,
      "learning_rate": 8.921146717500166e-05,
      "loss": 2.749288558959961,
      "memory(GiB)": 77.56,
      "step": 24865,
      "token_acc": 0.4529616724738676,
      "train_speed(iter/s)": 1.442914
    },
    {
      "epoch": 1.0655070476843322,
      "grad_norm": 4.570539474487305,
      "learning_rate": 8.920729119041957e-05,
      "loss": 2.4371414184570312,
      "memory(GiB)": 77.56,
      "step": 24870,
      "token_acc": 0.4849624060150376,
      "train_speed(iter/s)": 1.44289
    },
    {
      "epoch": 1.0657212630135813,
      "grad_norm": 4.441818714141846,
      "learning_rate": 8.920311449556062e-05,
      "loss": 2.522395896911621,
      "memory(GiB)": 77.56,
      "step": 24875,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.442891
    },
    {
      "epoch": 1.0659354783428303,
      "grad_norm": 6.656431674957275,
      "learning_rate": 8.919893709050049e-05,
      "loss": 2.601426696777344,
      "memory(GiB)": 77.56,
      "step": 24880,
      "token_acc": 0.4554794520547945,
      "train_speed(iter/s)": 1.442903
    },
    {
      "epoch": 1.0661496936720791,
      "grad_norm": 7.790807723999023,
      "learning_rate": 8.919475897531482e-05,
      "loss": 2.493697929382324,
      "memory(GiB)": 77.56,
      "step": 24885,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.442963
    },
    {
      "epoch": 1.0663639090013282,
      "grad_norm": 4.048851490020752,
      "learning_rate": 8.919058015007934e-05,
      "loss": 2.5156993865966797,
      "memory(GiB)": 77.56,
      "step": 24890,
      "token_acc": 0.478134110787172,
      "train_speed(iter/s)": 1.442968
    },
    {
      "epoch": 1.0665781243305772,
      "grad_norm": 6.370020866394043,
      "learning_rate": 8.918640061486974e-05,
      "loss": 2.289066505432129,
      "memory(GiB)": 77.56,
      "step": 24895,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.442983
    },
    {
      "epoch": 1.066792339659826,
      "grad_norm": 5.541647911071777,
      "learning_rate": 8.918222036976172e-05,
      "loss": 2.518605422973633,
      "memory(GiB)": 77.56,
      "step": 24900,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.442936
    },
    {
      "epoch": 1.067006554989075,
      "grad_norm": 6.655233860015869,
      "learning_rate": 8.917803941483101e-05,
      "loss": 2.372197914123535,
      "memory(GiB)": 77.56,
      "step": 24905,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.442965
    },
    {
      "epoch": 1.067220770318324,
      "grad_norm": 5.027594089508057,
      "learning_rate": 8.917385775015338e-05,
      "loss": 2.3086753845214845,
      "memory(GiB)": 77.56,
      "step": 24910,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.442902
    },
    {
      "epoch": 1.0674349856475729,
      "grad_norm": 4.518784046173096,
      "learning_rate": 8.916967537580457e-05,
      "loss": 2.430476760864258,
      "memory(GiB)": 77.56,
      "step": 24915,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.442933
    },
    {
      "epoch": 1.067649200976822,
      "grad_norm": 4.362011909484863,
      "learning_rate": 8.916549229186036e-05,
      "loss": 2.2647964477539064,
      "memory(GiB)": 77.56,
      "step": 24920,
      "token_acc": 0.49615384615384617,
      "train_speed(iter/s)": 1.442951
    },
    {
      "epoch": 1.067863416306071,
      "grad_norm": 6.057953357696533,
      "learning_rate": 8.916130849839649e-05,
      "loss": 2.3241344451904298,
      "memory(GiB)": 77.56,
      "step": 24925,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.442922
    },
    {
      "epoch": 1.0680776316353198,
      "grad_norm": 4.8255534172058105,
      "learning_rate": 8.915712399548879e-05,
      "loss": 2.385122299194336,
      "memory(GiB)": 77.56,
      "step": 24930,
      "token_acc": 0.49038461538461536,
      "train_speed(iter/s)": 1.442985
    },
    {
      "epoch": 1.0682918469645688,
      "grad_norm": 4.3298869132995605,
      "learning_rate": 8.915293878321308e-05,
      "loss": 2.2749584197998045,
      "memory(GiB)": 77.56,
      "step": 24935,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.443013
    },
    {
      "epoch": 1.0685060622938178,
      "grad_norm": 5.672133445739746,
      "learning_rate": 8.914875286164512e-05,
      "loss": 2.6646970748901366,
      "memory(GiB)": 77.56,
      "step": 24940,
      "token_acc": 0.45686900958466453,
      "train_speed(iter/s)": 1.443017
    },
    {
      "epoch": 1.0687202776230667,
      "grad_norm": 5.998624801635742,
      "learning_rate": 8.914456623086078e-05,
      "loss": 2.453325653076172,
      "memory(GiB)": 77.56,
      "step": 24945,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.443088
    },
    {
      "epoch": 1.0689344929523157,
      "grad_norm": 5.41029691696167,
      "learning_rate": 8.914037889093591e-05,
      "loss": 2.615672302246094,
      "memory(GiB)": 77.56,
      "step": 24950,
      "token_acc": 0.45625,
      "train_speed(iter/s)": 1.443145
    },
    {
      "epoch": 1.0691487082815647,
      "grad_norm": 4.725058078765869,
      "learning_rate": 8.913619084194638e-05,
      "loss": 2.575868034362793,
      "memory(GiB)": 77.56,
      "step": 24955,
      "token_acc": 0.5067114093959731,
      "train_speed(iter/s)": 1.443191
    },
    {
      "epoch": 1.0693629236108135,
      "grad_norm": 4.950165748596191,
      "learning_rate": 8.913200208396801e-05,
      "loss": 2.7590538024902345,
      "memory(GiB)": 77.56,
      "step": 24960,
      "token_acc": 0.42513368983957217,
      "train_speed(iter/s)": 1.443046
    },
    {
      "epoch": 1.0695771389400626,
      "grad_norm": 3.6775786876678467,
      "learning_rate": 8.912781261707674e-05,
      "loss": 2.4041927337646483,
      "memory(GiB)": 77.56,
      "step": 24965,
      "token_acc": 0.47039473684210525,
      "train_speed(iter/s)": 1.443077
    },
    {
      "epoch": 1.0697913542693116,
      "grad_norm": 5.825639724731445,
      "learning_rate": 8.912362244134842e-05,
      "loss": 2.311003875732422,
      "memory(GiB)": 77.56,
      "step": 24970,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.443123
    },
    {
      "epoch": 1.0700055695985604,
      "grad_norm": 5.18943452835083,
      "learning_rate": 8.911943155685897e-05,
      "loss": 2.6163257598876952,
      "memory(GiB)": 77.56,
      "step": 24975,
      "token_acc": 0.4575757575757576,
      "train_speed(iter/s)": 1.44317
    },
    {
      "epoch": 1.0702197849278094,
      "grad_norm": 6.513099670410156,
      "learning_rate": 8.911523996368434e-05,
      "loss": 2.6651309967041015,
      "memory(GiB)": 77.56,
      "step": 24980,
      "token_acc": 0.45808383233532934,
      "train_speed(iter/s)": 1.443177
    },
    {
      "epoch": 1.0704340002570585,
      "grad_norm": 5.062175273895264,
      "learning_rate": 8.911104766190045e-05,
      "loss": 2.5427068710327148,
      "memory(GiB)": 77.56,
      "step": 24985,
      "token_acc": 0.4609375,
      "train_speed(iter/s)": 1.443271
    },
    {
      "epoch": 1.0706482155863073,
      "grad_norm": 4.503647804260254,
      "learning_rate": 8.910685465158324e-05,
      "loss": 2.5308393478393554,
      "memory(GiB)": 77.56,
      "step": 24990,
      "token_acc": 0.4823943661971831,
      "train_speed(iter/s)": 1.443328
    },
    {
      "epoch": 1.0708624309155563,
      "grad_norm": 4.3264079093933105,
      "learning_rate": 8.910266093280866e-05,
      "loss": 2.391660690307617,
      "memory(GiB)": 77.56,
      "step": 24995,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.443337
    },
    {
      "epoch": 1.0710766462448054,
      "grad_norm": 5.4959940910339355,
      "learning_rate": 8.90984665056527e-05,
      "loss": 2.481093978881836,
      "memory(GiB)": 77.56,
      "step": 25000,
      "token_acc": 0.4744525547445255,
      "train_speed(iter/s)": 1.443315
    },
    {
      "epoch": 1.0710766462448054,
      "eval_loss": 2.2381017208099365,
      "eval_runtime": 13.5216,
      "eval_samples_per_second": 7.396,
      "eval_steps_per_second": 7.396,
      "eval_token_acc": 0.49591280653950953,
      "step": 25000
    },
    {
      "epoch": 1.0712908615740542,
      "grad_norm": 5.898514270782471,
      "learning_rate": 8.909427137019136e-05,
      "loss": 2.5043663024902343,
      "memory(GiB)": 77.56,
      "step": 25005,
      "token_acc": 0.48971596474045054,
      "train_speed(iter/s)": 1.442074
    },
    {
      "epoch": 1.0715050769033032,
      "grad_norm": 4.643527030944824,
      "learning_rate": 8.90900755265006e-05,
      "loss": 2.3257822036743163,
      "memory(GiB)": 77.56,
      "step": 25010,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.442073
    },
    {
      "epoch": 1.0717192922325522,
      "grad_norm": 4.901183605194092,
      "learning_rate": 8.908587897465644e-05,
      "loss": 2.6565052032470704,
      "memory(GiB)": 77.56,
      "step": 25015,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.442126
    },
    {
      "epoch": 1.071933507561801,
      "grad_norm": 3.8918936252593994,
      "learning_rate": 8.908168171473494e-05,
      "loss": 2.39459228515625,
      "memory(GiB)": 77.56,
      "step": 25020,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.442094
    },
    {
      "epoch": 1.07214772289105,
      "grad_norm": 4.803125858306885,
      "learning_rate": 8.907748374681211e-05,
      "loss": 2.627385902404785,
      "memory(GiB)": 77.56,
      "step": 25025,
      "token_acc": 0.4204946996466431,
      "train_speed(iter/s)": 1.442145
    },
    {
      "epoch": 1.0723619382202991,
      "grad_norm": 3.630722761154175,
      "learning_rate": 8.907328507096399e-05,
      "loss": 2.4388004302978517,
      "memory(GiB)": 77.56,
      "step": 25030,
      "token_acc": 0.4980392156862745,
      "train_speed(iter/s)": 1.442229
    },
    {
      "epoch": 1.072576153549548,
      "grad_norm": 4.244997501373291,
      "learning_rate": 8.906908568726669e-05,
      "loss": 2.653009605407715,
      "memory(GiB)": 77.56,
      "step": 25035,
      "token_acc": 0.48424068767908307,
      "train_speed(iter/s)": 1.442197
    },
    {
      "epoch": 1.072790368878797,
      "grad_norm": 5.776773452758789,
      "learning_rate": 8.906488559579623e-05,
      "loss": 2.4952442169189455,
      "memory(GiB)": 77.56,
      "step": 25040,
      "token_acc": 0.5057915057915058,
      "train_speed(iter/s)": 1.442185
    },
    {
      "epoch": 1.073004584208046,
      "grad_norm": 7.8657941818237305,
      "learning_rate": 8.906068479662871e-05,
      "loss": 2.5570497512817383,
      "memory(GiB)": 77.56,
      "step": 25045,
      "token_acc": 0.5035211267605634,
      "train_speed(iter/s)": 1.442214
    },
    {
      "epoch": 1.0732187995372948,
      "grad_norm": 5.0104875564575195,
      "learning_rate": 8.905648328984026e-05,
      "loss": 2.180466079711914,
      "memory(GiB)": 77.56,
      "step": 25050,
      "token_acc": 0.5018181818181818,
      "train_speed(iter/s)": 1.442212
    },
    {
      "epoch": 1.0734330148665439,
      "grad_norm": 5.082650661468506,
      "learning_rate": 8.905228107550697e-05,
      "loss": 2.3480173110961915,
      "memory(GiB)": 77.56,
      "step": 25055,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.442217
    },
    {
      "epoch": 1.073647230195793,
      "grad_norm": 4.033272743225098,
      "learning_rate": 8.904807815370499e-05,
      "loss": 2.507711410522461,
      "memory(GiB)": 77.56,
      "step": 25060,
      "token_acc": 0.43465045592705165,
      "train_speed(iter/s)": 1.442239
    },
    {
      "epoch": 1.0738614455250417,
      "grad_norm": 3.5127065181732178,
      "learning_rate": 8.904387452451042e-05,
      "loss": 2.6224090576171877,
      "memory(GiB)": 77.56,
      "step": 25065,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.442214
    },
    {
      "epoch": 1.0740756608542907,
      "grad_norm": 4.23453950881958,
      "learning_rate": 8.903967018799946e-05,
      "loss": 2.6901920318603514,
      "memory(GiB)": 77.56,
      "step": 25070,
      "token_acc": 0.4408284023668639,
      "train_speed(iter/s)": 1.442252
    },
    {
      "epoch": 1.0742898761835398,
      "grad_norm": 3.6714584827423096,
      "learning_rate": 8.903546514424824e-05,
      "loss": 2.7261823654174804,
      "memory(GiB)": 77.56,
      "step": 25075,
      "token_acc": 0.44857142857142857,
      "train_speed(iter/s)": 1.442268
    },
    {
      "epoch": 1.0745040915127886,
      "grad_norm": 3.8848061561584473,
      "learning_rate": 8.903125939333294e-05,
      "loss": 2.4167057037353517,
      "memory(GiB)": 77.56,
      "step": 25080,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.442259
    },
    {
      "epoch": 1.0747183068420376,
      "grad_norm": 4.971715927124023,
      "learning_rate": 8.902705293532978e-05,
      "loss": 2.7110219955444337,
      "memory(GiB)": 77.56,
      "step": 25085,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.442213
    },
    {
      "epoch": 1.0749325221712867,
      "grad_norm": 4.538780689239502,
      "learning_rate": 8.902284577031494e-05,
      "loss": 2.5083343505859377,
      "memory(GiB)": 77.56,
      "step": 25090,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.44222
    },
    {
      "epoch": 1.0751467375005355,
      "grad_norm": 5.06449556350708,
      "learning_rate": 8.901863789836464e-05,
      "loss": 2.4191352844238283,
      "memory(GiB)": 77.56,
      "step": 25095,
      "token_acc": 0.5342960288808665,
      "train_speed(iter/s)": 1.442257
    },
    {
      "epoch": 1.0753609528297845,
      "grad_norm": 3.889514684677124,
      "learning_rate": 8.901442931955512e-05,
      "loss": 2.728623390197754,
      "memory(GiB)": 77.56,
      "step": 25100,
      "token_acc": 0.43103448275862066,
      "train_speed(iter/s)": 1.442268
    },
    {
      "epoch": 1.0755751681590335,
      "grad_norm": 6.102291584014893,
      "learning_rate": 8.90102200339626e-05,
      "loss": 2.3905193328857424,
      "memory(GiB)": 77.56,
      "step": 25105,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.442319
    },
    {
      "epoch": 1.0757893834882823,
      "grad_norm": 4.236095428466797,
      "learning_rate": 8.900601004166335e-05,
      "loss": 2.690999984741211,
      "memory(GiB)": 77.56,
      "step": 25110,
      "token_acc": 0.45217391304347826,
      "train_speed(iter/s)": 1.442302
    },
    {
      "epoch": 1.0760035988175314,
      "grad_norm": 4.288787364959717,
      "learning_rate": 8.900179934273366e-05,
      "loss": 2.4521663665771483,
      "memory(GiB)": 77.56,
      "step": 25115,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.442289
    },
    {
      "epoch": 1.0762178141467804,
      "grad_norm": 11.823760032653809,
      "learning_rate": 8.899758793724978e-05,
      "loss": 2.724221420288086,
      "memory(GiB)": 77.56,
      "step": 25120,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.442362
    },
    {
      "epoch": 1.0764320294760292,
      "grad_norm": 5.338200092315674,
      "learning_rate": 8.8993375825288e-05,
      "loss": 2.414128875732422,
      "memory(GiB)": 77.56,
      "step": 25125,
      "token_acc": 0.4627450980392157,
      "train_speed(iter/s)": 1.442328
    },
    {
      "epoch": 1.0766462448052783,
      "grad_norm": 4.276544570922852,
      "learning_rate": 8.898916300692463e-05,
      "loss": 2.6601654052734376,
      "memory(GiB)": 77.56,
      "step": 25130,
      "token_acc": 0.444,
      "train_speed(iter/s)": 1.442319
    },
    {
      "epoch": 1.0768604601345273,
      "grad_norm": 4.820889949798584,
      "learning_rate": 8.898494948223604e-05,
      "loss": 2.3188581466674805,
      "memory(GiB)": 77.56,
      "step": 25135,
      "token_acc": 0.5155038759689923,
      "train_speed(iter/s)": 1.442306
    },
    {
      "epoch": 1.077074675463776,
      "grad_norm": 4.7706522941589355,
      "learning_rate": 8.898073525129848e-05,
      "loss": 2.429790496826172,
      "memory(GiB)": 77.56,
      "step": 25140,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.442319
    },
    {
      "epoch": 1.0772888907930251,
      "grad_norm": 5.803450107574463,
      "learning_rate": 8.897652031418834e-05,
      "loss": 2.4875125885009766,
      "memory(GiB)": 77.56,
      "step": 25145,
      "token_acc": 0.4714828897338403,
      "train_speed(iter/s)": 1.442373
    },
    {
      "epoch": 1.0775031061222742,
      "grad_norm": 4.465826988220215,
      "learning_rate": 8.8972304670982e-05,
      "loss": 2.676397514343262,
      "memory(GiB)": 77.56,
      "step": 25150,
      "token_acc": 0.43376623376623374,
      "train_speed(iter/s)": 1.442419
    },
    {
      "epoch": 1.077717321451523,
      "grad_norm": 4.993764877319336,
      "learning_rate": 8.896808832175576e-05,
      "loss": 2.623885726928711,
      "memory(GiB)": 77.56,
      "step": 25155,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.442421
    },
    {
      "epoch": 1.077931536780772,
      "grad_norm": 4.131471157073975,
      "learning_rate": 8.896387126658605e-05,
      "loss": 2.592173385620117,
      "memory(GiB)": 77.56,
      "step": 25160,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.442489
    },
    {
      "epoch": 1.078145752110021,
      "grad_norm": 5.068206787109375,
      "learning_rate": 8.895965350554929e-05,
      "loss": 2.6493844985961914,
      "memory(GiB)": 77.56,
      "step": 25165,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.442505
    },
    {
      "epoch": 1.0783599674392699,
      "grad_norm": 5.090224266052246,
      "learning_rate": 8.895543503872183e-05,
      "loss": 2.3927188873291017,
      "memory(GiB)": 77.56,
      "step": 25170,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.442485
    },
    {
      "epoch": 1.078574182768519,
      "grad_norm": 8.322251319885254,
      "learning_rate": 8.895121586618014e-05,
      "loss": 2.4706796646118163,
      "memory(GiB)": 77.56,
      "step": 25175,
      "token_acc": 0.5473251028806584,
      "train_speed(iter/s)": 1.442471
    },
    {
      "epoch": 1.078788398097768,
      "grad_norm": 3.7725141048431396,
      "learning_rate": 8.894699598800064e-05,
      "loss": 2.457138442993164,
      "memory(GiB)": 77.56,
      "step": 25180,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.442367
    },
    {
      "epoch": 1.0790026134270168,
      "grad_norm": 4.180278778076172,
      "learning_rate": 8.894277540425977e-05,
      "loss": 2.7426902770996096,
      "memory(GiB)": 77.56,
      "step": 25185,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.442405
    },
    {
      "epoch": 1.0792168287562658,
      "grad_norm": 5.424924850463867,
      "learning_rate": 8.893855411503398e-05,
      "loss": 2.2191679000854494,
      "memory(GiB)": 77.56,
      "step": 25190,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.442394
    },
    {
      "epoch": 1.0794310440855148,
      "grad_norm": 5.147541046142578,
      "learning_rate": 8.893433212039974e-05,
      "loss": 2.6832679748535155,
      "memory(GiB)": 77.56,
      "step": 25195,
      "token_acc": 0.42955326460481097,
      "train_speed(iter/s)": 1.442419
    },
    {
      "epoch": 1.0796452594147636,
      "grad_norm": 4.844173908233643,
      "learning_rate": 8.893010942043359e-05,
      "loss": 2.839406204223633,
      "memory(GiB)": 77.56,
      "step": 25200,
      "token_acc": 0.43506493506493504,
      "train_speed(iter/s)": 1.442488
    },
    {
      "epoch": 1.0798594747440127,
      "grad_norm": 3.2959585189819336,
      "learning_rate": 8.892588601521197e-05,
      "loss": 2.4778619766235352,
      "memory(GiB)": 77.56,
      "step": 25205,
      "token_acc": 0.48520710059171596,
      "train_speed(iter/s)": 1.442461
    },
    {
      "epoch": 1.0800736900732617,
      "grad_norm": 7.255958080291748,
      "learning_rate": 8.89216619048114e-05,
      "loss": 2.7225162506103517,
      "memory(GiB)": 77.56,
      "step": 25210,
      "token_acc": 0.43416370106761565,
      "train_speed(iter/s)": 1.442471
    },
    {
      "epoch": 1.0802879054025105,
      "grad_norm": 4.925755500793457,
      "learning_rate": 8.891743708930842e-05,
      "loss": 2.397993469238281,
      "memory(GiB)": 77.56,
      "step": 25215,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.442447
    },
    {
      "epoch": 1.0805021207317596,
      "grad_norm": 4.483622074127197,
      "learning_rate": 8.891321156877957e-05,
      "loss": 2.415365982055664,
      "memory(GiB)": 77.56,
      "step": 25220,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.442361
    },
    {
      "epoch": 1.0807163360610086,
      "grad_norm": 5.890489101409912,
      "learning_rate": 8.890898534330136e-05,
      "loss": 2.3566139221191404,
      "memory(GiB)": 77.56,
      "step": 25225,
      "token_acc": 0.49508196721311476,
      "train_speed(iter/s)": 1.442366
    },
    {
      "epoch": 1.0809305513902574,
      "grad_norm": 3.569340229034424,
      "learning_rate": 8.89047584129504e-05,
      "loss": 2.535544586181641,
      "memory(GiB)": 77.56,
      "step": 25230,
      "token_acc": 0.467680608365019,
      "train_speed(iter/s)": 1.442323
    },
    {
      "epoch": 1.0811447667195064,
      "grad_norm": 5.1981916427612305,
      "learning_rate": 8.890053077780325e-05,
      "loss": 2.487428665161133,
      "memory(GiB)": 77.56,
      "step": 25235,
      "token_acc": 0.4602076124567474,
      "train_speed(iter/s)": 1.442362
    },
    {
      "epoch": 1.0813589820487555,
      "grad_norm": 6.219172477722168,
      "learning_rate": 8.88963024379365e-05,
      "loss": 2.3975128173828124,
      "memory(GiB)": 77.56,
      "step": 25240,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.442363
    },
    {
      "epoch": 1.0815731973780043,
      "grad_norm": 3.575761556625366,
      "learning_rate": 8.889207339342673e-05,
      "loss": 2.8026546478271483,
      "memory(GiB)": 77.56,
      "step": 25245,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.442427
    },
    {
      "epoch": 1.0817874127072533,
      "grad_norm": 5.14598274230957,
      "learning_rate": 8.888784364435056e-05,
      "loss": 2.301034927368164,
      "memory(GiB)": 77.56,
      "step": 25250,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.442425
    },
    {
      "epoch": 1.0820016280365023,
      "grad_norm": 4.888864994049072,
      "learning_rate": 8.888361319078464e-05,
      "loss": 2.527876091003418,
      "memory(GiB)": 77.56,
      "step": 25255,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.442499
    },
    {
      "epoch": 1.0822158433657512,
      "grad_norm": 5.543771266937256,
      "learning_rate": 8.887938203280559e-05,
      "loss": 2.743634033203125,
      "memory(GiB)": 77.56,
      "step": 25260,
      "token_acc": 0.4362017804154303,
      "train_speed(iter/s)": 1.442459
    },
    {
      "epoch": 1.0824300586950002,
      "grad_norm": 4.484138488769531,
      "learning_rate": 8.887515017049006e-05,
      "loss": 2.4278846740722657,
      "memory(GiB)": 77.56,
      "step": 25265,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.442484
    },
    {
      "epoch": 1.0826442740242492,
      "grad_norm": 3.929466485977173,
      "learning_rate": 8.887091760391471e-05,
      "loss": 2.311697769165039,
      "memory(GiB)": 77.56,
      "step": 25270,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.442436
    },
    {
      "epoch": 1.082858489353498,
      "grad_norm": 5.253222465515137,
      "learning_rate": 8.886668433315622e-05,
      "loss": 2.537200164794922,
      "memory(GiB)": 77.56,
      "step": 25275,
      "token_acc": 0.4786885245901639,
      "train_speed(iter/s)": 1.442435
    },
    {
      "epoch": 1.083072704682747,
      "grad_norm": 5.537881851196289,
      "learning_rate": 8.88624503582913e-05,
      "loss": 2.906534194946289,
      "memory(GiB)": 77.56,
      "step": 25280,
      "token_acc": 0.43214285714285716,
      "train_speed(iter/s)": 1.442448
    },
    {
      "epoch": 1.0832869200119961,
      "grad_norm": 4.92050313949585,
      "learning_rate": 8.885821567939663e-05,
      "loss": 2.659988021850586,
      "memory(GiB)": 77.56,
      "step": 25285,
      "token_acc": 0.4188679245283019,
      "train_speed(iter/s)": 1.442461
    },
    {
      "epoch": 1.083501135341245,
      "grad_norm": 6.284702777862549,
      "learning_rate": 8.885398029654892e-05,
      "loss": 2.5327856063842775,
      "memory(GiB)": 77.56,
      "step": 25290,
      "token_acc": 0.48014440433212996,
      "train_speed(iter/s)": 1.442492
    },
    {
      "epoch": 1.083715350670494,
      "grad_norm": 5.071280479431152,
      "learning_rate": 8.884974420982493e-05,
      "loss": 2.2914306640625,
      "memory(GiB)": 77.56,
      "step": 25295,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.442449
    },
    {
      "epoch": 1.083929565999743,
      "grad_norm": 4.764655113220215,
      "learning_rate": 8.884550741930135e-05,
      "loss": 2.6236648559570312,
      "memory(GiB)": 77.56,
      "step": 25300,
      "token_acc": 0.4551083591331269,
      "train_speed(iter/s)": 1.442407
    },
    {
      "epoch": 1.0841437813289918,
      "grad_norm": 4.0674567222595215,
      "learning_rate": 8.884126992505498e-05,
      "loss": 2.7162118911743165,
      "memory(GiB)": 77.56,
      "step": 25305,
      "token_acc": 0.4489795918367347,
      "train_speed(iter/s)": 1.442393
    },
    {
      "epoch": 1.0843579966582408,
      "grad_norm": 4.113844394683838,
      "learning_rate": 8.883703172716258e-05,
      "loss": 2.3346494674682616,
      "memory(GiB)": 77.56,
      "step": 25310,
      "token_acc": 0.5016181229773463,
      "train_speed(iter/s)": 1.442398
    },
    {
      "epoch": 1.0845722119874899,
      "grad_norm": 5.086583137512207,
      "learning_rate": 8.883279282570089e-05,
      "loss": 2.295521545410156,
      "memory(GiB)": 77.56,
      "step": 25315,
      "token_acc": 0.5133333333333333,
      "train_speed(iter/s)": 1.442369
    },
    {
      "epoch": 1.0847864273167387,
      "grad_norm": 5.205321311950684,
      "learning_rate": 8.882855322074674e-05,
      "loss": 2.199264907836914,
      "memory(GiB)": 77.56,
      "step": 25320,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.442359
    },
    {
      "epoch": 1.0850006426459877,
      "grad_norm": 5.412535667419434,
      "learning_rate": 8.882431291237693e-05,
      "loss": 2.65649356842041,
      "memory(GiB)": 77.56,
      "step": 25325,
      "token_acc": 0.45514950166112955,
      "train_speed(iter/s)": 1.44243
    },
    {
      "epoch": 1.0852148579752368,
      "grad_norm": 4.51121711730957,
      "learning_rate": 8.882007190066827e-05,
      "loss": 2.490675354003906,
      "memory(GiB)": 77.56,
      "step": 25330,
      "token_acc": 0.5096525096525096,
      "train_speed(iter/s)": 1.442491
    },
    {
      "epoch": 1.0854290733044856,
      "grad_norm": 8.112110137939453,
      "learning_rate": 8.881583018569761e-05,
      "loss": 2.6366546630859373,
      "memory(GiB)": 77.56,
      "step": 25335,
      "token_acc": 0.4607142857142857,
      "train_speed(iter/s)": 1.442421
    },
    {
      "epoch": 1.0856432886337346,
      "grad_norm": 4.496501445770264,
      "learning_rate": 8.881158776754175e-05,
      "loss": 2.4299571990966795,
      "memory(GiB)": 77.56,
      "step": 25340,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.44244
    },
    {
      "epoch": 1.0858575039629836,
      "grad_norm": 7.006977081298828,
      "learning_rate": 8.880734464627757e-05,
      "loss": 2.751093864440918,
      "memory(GiB)": 77.56,
      "step": 25345,
      "token_acc": 0.40828402366863903,
      "train_speed(iter/s)": 1.442403
    },
    {
      "epoch": 1.0860717192922325,
      "grad_norm": 6.189694404602051,
      "learning_rate": 8.880310082198196e-05,
      "loss": 2.4706138610839843,
      "memory(GiB)": 77.56,
      "step": 25350,
      "token_acc": 0.47410358565737054,
      "train_speed(iter/s)": 1.442373
    },
    {
      "epoch": 1.0862859346214815,
      "grad_norm": 4.348604202270508,
      "learning_rate": 8.879885629473176e-05,
      "loss": 2.542814826965332,
      "memory(GiB)": 77.56,
      "step": 25355,
      "token_acc": 0.49812734082397003,
      "train_speed(iter/s)": 1.442376
    },
    {
      "epoch": 1.0865001499507305,
      "grad_norm": 4.553647994995117,
      "learning_rate": 8.87946110646039e-05,
      "loss": 2.4307376861572267,
      "memory(GiB)": 77.56,
      "step": 25360,
      "token_acc": 0.47474747474747475,
      "train_speed(iter/s)": 1.442389
    },
    {
      "epoch": 1.0867143652799793,
      "grad_norm": 4.445259094238281,
      "learning_rate": 8.879036513167523e-05,
      "loss": 2.8615108489990235,
      "memory(GiB)": 77.56,
      "step": 25365,
      "token_acc": 0.40782122905027934,
      "train_speed(iter/s)": 1.442417
    },
    {
      "epoch": 1.0869285806092284,
      "grad_norm": 5.165308475494385,
      "learning_rate": 8.878611849602274e-05,
      "loss": 2.4291994094848635,
      "memory(GiB)": 77.56,
      "step": 25370,
      "token_acc": 0.4962121212121212,
      "train_speed(iter/s)": 1.442387
    },
    {
      "epoch": 1.0871427959384774,
      "grad_norm": 3.711165428161621,
      "learning_rate": 8.878187115772331e-05,
      "loss": 2.595941925048828,
      "memory(GiB)": 77.56,
      "step": 25375,
      "token_acc": 0.5058479532163743,
      "train_speed(iter/s)": 1.442363
    },
    {
      "epoch": 1.0873570112677262,
      "grad_norm": 4.4258832931518555,
      "learning_rate": 8.877762311685391e-05,
      "loss": 2.642012023925781,
      "memory(GiB)": 77.56,
      "step": 25380,
      "token_acc": 0.47962382445141066,
      "train_speed(iter/s)": 1.442347
    },
    {
      "epoch": 1.0875712265969752,
      "grad_norm": 5.371784687042236,
      "learning_rate": 8.87733743734915e-05,
      "loss": 2.484430694580078,
      "memory(GiB)": 77.56,
      "step": 25385,
      "token_acc": 0.47633136094674555,
      "train_speed(iter/s)": 1.442328
    },
    {
      "epoch": 1.0877854419262243,
      "grad_norm": 4.51119327545166,
      "learning_rate": 8.876912492771303e-05,
      "loss": 2.2747976303100588,
      "memory(GiB)": 77.56,
      "step": 25390,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.442323
    },
    {
      "epoch": 1.087999657255473,
      "grad_norm": 5.508774757385254,
      "learning_rate": 8.876487477959551e-05,
      "loss": 2.6571575164794923,
      "memory(GiB)": 77.56,
      "step": 25395,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.442392
    },
    {
      "epoch": 1.0882138725847221,
      "grad_norm": 3.836334228515625,
      "learning_rate": 8.876062392921591e-05,
      "loss": 2.3518226623535154,
      "memory(GiB)": 77.56,
      "step": 25400,
      "token_acc": 0.5230125523012552,
      "train_speed(iter/s)": 1.442411
    },
    {
      "epoch": 1.0884280879139712,
      "grad_norm": 4.80452299118042,
      "learning_rate": 8.875637237665124e-05,
      "loss": 2.603152847290039,
      "memory(GiB)": 77.56,
      "step": 25405,
      "token_acc": 0.4703703703703704,
      "train_speed(iter/s)": 1.442444
    },
    {
      "epoch": 1.08864230324322,
      "grad_norm": 3.5263288021087646,
      "learning_rate": 8.875212012197853e-05,
      "loss": 2.3496030807495116,
      "memory(GiB)": 77.56,
      "step": 25410,
      "token_acc": 0.49216300940438873,
      "train_speed(iter/s)": 1.44248
    },
    {
      "epoch": 1.088856518572469,
      "grad_norm": 5.528037071228027,
      "learning_rate": 8.874786716527482e-05,
      "loss": 2.4018138885498046,
      "memory(GiB)": 77.56,
      "step": 25415,
      "token_acc": 0.519434628975265,
      "train_speed(iter/s)": 1.442525
    },
    {
      "epoch": 1.089070733901718,
      "grad_norm": 4.728265285491943,
      "learning_rate": 8.874361350661713e-05,
      "loss": 2.4568099975585938,
      "memory(GiB)": 77.56,
      "step": 25420,
      "token_acc": 0.5073529411764706,
      "train_speed(iter/s)": 1.442467
    },
    {
      "epoch": 1.0892849492309669,
      "grad_norm": 4.832590579986572,
      "learning_rate": 8.873935914608256e-05,
      "loss": 2.357774353027344,
      "memory(GiB)": 77.56,
      "step": 25425,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.442495
    },
    {
      "epoch": 1.089499164560216,
      "grad_norm": 7.3577094078063965,
      "learning_rate": 8.873510408374814e-05,
      "loss": 2.623159408569336,
      "memory(GiB)": 77.56,
      "step": 25430,
      "token_acc": 0.45075757575757575,
      "train_speed(iter/s)": 1.442556
    },
    {
      "epoch": 1.089713379889465,
      "grad_norm": 4.762428283691406,
      "learning_rate": 8.873084831969098e-05,
      "loss": 2.5407928466796874,
      "memory(GiB)": 77.56,
      "step": 25435,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.442537
    },
    {
      "epoch": 1.0899275952187137,
      "grad_norm": 4.607973098754883,
      "learning_rate": 8.872659185398817e-05,
      "loss": 2.6841594696044924,
      "memory(GiB)": 77.56,
      "step": 25440,
      "token_acc": 0.43653250773993807,
      "train_speed(iter/s)": 1.442545
    },
    {
      "epoch": 1.0901418105479628,
      "grad_norm": 8.858938217163086,
      "learning_rate": 8.872233468671683e-05,
      "loss": 2.210304832458496,
      "memory(GiB)": 77.56,
      "step": 25445,
      "token_acc": 0.5462555066079295,
      "train_speed(iter/s)": 1.442606
    },
    {
      "epoch": 1.0903560258772118,
      "grad_norm": 4.562241077423096,
      "learning_rate": 8.871807681795406e-05,
      "loss": 2.403728485107422,
      "memory(GiB)": 77.56,
      "step": 25450,
      "token_acc": 0.4731182795698925,
      "train_speed(iter/s)": 1.442646
    },
    {
      "epoch": 1.0905702412064606,
      "grad_norm": 4.31027889251709,
      "learning_rate": 8.8713818247777e-05,
      "loss": 2.3220489501953123,
      "memory(GiB)": 77.56,
      "step": 25455,
      "token_acc": 0.5218978102189781,
      "train_speed(iter/s)": 1.442695
    },
    {
      "epoch": 1.0907844565357097,
      "grad_norm": 5.132511138916016,
      "learning_rate": 8.870955897626283e-05,
      "loss": 2.5519908905029296,
      "memory(GiB)": 77.56,
      "step": 25460,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.442773
    },
    {
      "epoch": 1.0909986718649587,
      "grad_norm": 4.796254634857178,
      "learning_rate": 8.870529900348867e-05,
      "loss": 2.466431427001953,
      "memory(GiB)": 77.56,
      "step": 25465,
      "token_acc": 0.5093167701863354,
      "train_speed(iter/s)": 1.442713
    },
    {
      "epoch": 1.0912128871942075,
      "grad_norm": 4.18263053894043,
      "learning_rate": 8.870103832953172e-05,
      "loss": 2.6662452697753904,
      "memory(GiB)": 77.56,
      "step": 25470,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.442715
    },
    {
      "epoch": 1.0914271025234565,
      "grad_norm": 4.283825397491455,
      "learning_rate": 8.869677695446914e-05,
      "loss": 2.7063230514526366,
      "memory(GiB)": 77.56,
      "step": 25475,
      "token_acc": 0.43462897526501765,
      "train_speed(iter/s)": 1.442695
    },
    {
      "epoch": 1.0916413178527056,
      "grad_norm": 5.661508560180664,
      "learning_rate": 8.869251487837816e-05,
      "loss": 2.4352169036865234,
      "memory(GiB)": 77.56,
      "step": 25480,
      "token_acc": 0.4536741214057508,
      "train_speed(iter/s)": 1.442712
    },
    {
      "epoch": 1.0918555331819544,
      "grad_norm": 4.870262622833252,
      "learning_rate": 8.868825210133598e-05,
      "loss": 2.5312122344970702,
      "memory(GiB)": 77.56,
      "step": 25485,
      "token_acc": 0.48253968253968255,
      "train_speed(iter/s)": 1.442782
    },
    {
      "epoch": 1.0920697485112034,
      "grad_norm": 5.331650257110596,
      "learning_rate": 8.86839886234198e-05,
      "loss": 2.5506799697875975,
      "memory(GiB)": 77.56,
      "step": 25490,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.442802
    },
    {
      "epoch": 1.0922839638404525,
      "grad_norm": 4.522636890411377,
      "learning_rate": 8.86797244447069e-05,
      "loss": 2.5758472442626954,
      "memory(GiB)": 77.56,
      "step": 25495,
      "token_acc": 0.4519230769230769,
      "train_speed(iter/s)": 1.442817
    },
    {
      "epoch": 1.0924981791697013,
      "grad_norm": 6.4138383865356445,
      "learning_rate": 8.86754595652745e-05,
      "loss": 2.6582805633544924,
      "memory(GiB)": 77.56,
      "step": 25500,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.442841
    },
    {
      "epoch": 1.0924981791697013,
      "eval_loss": 2.074680805206299,
      "eval_runtime": 14.4656,
      "eval_samples_per_second": 6.913,
      "eval_steps_per_second": 6.913,
      "eval_token_acc": 0.5049365303244006,
      "step": 25500
    },
    {
      "epoch": 1.0927123944989503,
      "grad_norm": 5.1362504959106445,
      "learning_rate": 8.867119398519986e-05,
      "loss": 2.3715230941772463,
      "memory(GiB)": 77.56,
      "step": 25505,
      "token_acc": 0.5031185031185031,
      "train_speed(iter/s)": 1.441601
    },
    {
      "epoch": 1.0929266098281993,
      "grad_norm": 6.287789344787598,
      "learning_rate": 8.866692770456026e-05,
      "loss": 2.3799259185791017,
      "memory(GiB)": 77.56,
      "step": 25510,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441647
    },
    {
      "epoch": 1.0931408251574484,
      "grad_norm": 8.176146507263184,
      "learning_rate": 8.866266072343301e-05,
      "loss": 2.3360300064086914,
      "memory(GiB)": 77.56,
      "step": 25515,
      "token_acc": 0.48412698412698413,
      "train_speed(iter/s)": 1.441682
    },
    {
      "epoch": 1.0933550404866972,
      "grad_norm": 5.211651802062988,
      "learning_rate": 8.865839304189538e-05,
      "loss": 2.5676237106323243,
      "memory(GiB)": 77.56,
      "step": 25520,
      "token_acc": 0.4721189591078067,
      "train_speed(iter/s)": 1.441695
    },
    {
      "epoch": 1.0935692558159462,
      "grad_norm": 5.002561092376709,
      "learning_rate": 8.865412466002472e-05,
      "loss": 2.760858917236328,
      "memory(GiB)": 77.56,
      "step": 25525,
      "token_acc": 0.43730886850152906,
      "train_speed(iter/s)": 1.441679
    },
    {
      "epoch": 1.0937834711451953,
      "grad_norm": 4.139132499694824,
      "learning_rate": 8.86498555778983e-05,
      "loss": 2.6972469329833983,
      "memory(GiB)": 77.56,
      "step": 25530,
      "token_acc": 0.42724458204334365,
      "train_speed(iter/s)": 1.441691
    },
    {
      "epoch": 1.093997686474444,
      "grad_norm": 6.844869613647461,
      "learning_rate": 8.86455857955935e-05,
      "loss": 2.4446144104003906,
      "memory(GiB)": 77.56,
      "step": 25535,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.44164
    },
    {
      "epoch": 1.094211901803693,
      "grad_norm": 4.971738815307617,
      "learning_rate": 8.864131531318766e-05,
      "loss": 2.5308584213256835,
      "memory(GiB)": 77.56,
      "step": 25540,
      "token_acc": 0.41638225255972694,
      "train_speed(iter/s)": 1.441673
    },
    {
      "epoch": 1.0944261171329421,
      "grad_norm": 4.738288879394531,
      "learning_rate": 8.863704413075816e-05,
      "loss": 2.465706443786621,
      "memory(GiB)": 77.56,
      "step": 25545,
      "token_acc": 0.4867924528301887,
      "train_speed(iter/s)": 1.441681
    },
    {
      "epoch": 1.094640332462191,
      "grad_norm": 5.577242851257324,
      "learning_rate": 8.863277224838234e-05,
      "loss": 2.2139705657958983,
      "memory(GiB)": 77.56,
      "step": 25550,
      "token_acc": 0.5603112840466926,
      "train_speed(iter/s)": 1.441675
    },
    {
      "epoch": 1.09485454779144,
      "grad_norm": 3.6750497817993164,
      "learning_rate": 8.862849966613763e-05,
      "loss": 2.676479530334473,
      "memory(GiB)": 77.56,
      "step": 25555,
      "token_acc": 0.5052631578947369,
      "train_speed(iter/s)": 1.441686
    },
    {
      "epoch": 1.095068763120689,
      "grad_norm": 4.274869918823242,
      "learning_rate": 8.862422638410139e-05,
      "loss": 2.593830680847168,
      "memory(GiB)": 77.56,
      "step": 25560,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.441701
    },
    {
      "epoch": 1.0952829784499378,
      "grad_norm": 4.485520839691162,
      "learning_rate": 8.861995240235106e-05,
      "loss": 2.6144424438476563,
      "memory(GiB)": 77.56,
      "step": 25565,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.441709
    },
    {
      "epoch": 1.0954971937791869,
      "grad_norm": 5.184383869171143,
      "learning_rate": 8.861567772096408e-05,
      "loss": 2.807928466796875,
      "memory(GiB)": 77.56,
      "step": 25570,
      "token_acc": 0.460431654676259,
      "train_speed(iter/s)": 1.441766
    },
    {
      "epoch": 1.095711409108436,
      "grad_norm": 4.967194557189941,
      "learning_rate": 8.861140234001785e-05,
      "loss": 2.6933567047119142,
      "memory(GiB)": 77.56,
      "step": 25575,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.441754
    },
    {
      "epoch": 1.0959256244376847,
      "grad_norm": 5.986224174499512,
      "learning_rate": 8.860712625958987e-05,
      "loss": 2.4381717681884765,
      "memory(GiB)": 77.56,
      "step": 25580,
      "token_acc": 0.45674740484429066,
      "train_speed(iter/s)": 1.441739
    },
    {
      "epoch": 1.0961398397669337,
      "grad_norm": 4.272314071655273,
      "learning_rate": 8.860284947975758e-05,
      "loss": 2.693319320678711,
      "memory(GiB)": 77.56,
      "step": 25585,
      "token_acc": 0.42907801418439717,
      "train_speed(iter/s)": 1.441721
    },
    {
      "epoch": 1.0963540550961828,
      "grad_norm": 4.546026706695557,
      "learning_rate": 8.859857200059845e-05,
      "loss": 2.8001968383789064,
      "memory(GiB)": 77.56,
      "step": 25590,
      "token_acc": 0.4208754208754209,
      "train_speed(iter/s)": 1.441754
    },
    {
      "epoch": 1.0965682704254316,
      "grad_norm": 6.143930435180664,
      "learning_rate": 8.859429382218998e-05,
      "loss": 2.657841682434082,
      "memory(GiB)": 77.56,
      "step": 25595,
      "token_acc": 0.4793650793650794,
      "train_speed(iter/s)": 1.441783
    },
    {
      "epoch": 1.0967824857546806,
      "grad_norm": 3.526970624923706,
      "learning_rate": 8.859001494460968e-05,
      "loss": 2.2485504150390625,
      "memory(GiB)": 77.56,
      "step": 25600,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.441715
    },
    {
      "epoch": 1.0969967010839297,
      "grad_norm": 4.2236809730529785,
      "learning_rate": 8.858573536793504e-05,
      "loss": 2.58162841796875,
      "memory(GiB)": 77.56,
      "step": 25605,
      "token_acc": 0.4421364985163205,
      "train_speed(iter/s)": 1.441716
    },
    {
      "epoch": 1.0972109164131785,
      "grad_norm": 4.2564697265625,
      "learning_rate": 8.858145509224363e-05,
      "loss": 2.5570674896240235,
      "memory(GiB)": 77.56,
      "step": 25610,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.441715
    },
    {
      "epoch": 1.0974251317424275,
      "grad_norm": 4.532004356384277,
      "learning_rate": 8.857717411761296e-05,
      "loss": 2.4169822692871095,
      "memory(GiB)": 77.56,
      "step": 25615,
      "token_acc": 0.5212355212355212,
      "train_speed(iter/s)": 1.441744
    },
    {
      "epoch": 1.0976393470716765,
      "grad_norm": 6.756443977355957,
      "learning_rate": 8.857289244412059e-05,
      "loss": 2.1476806640625,
      "memory(GiB)": 77.56,
      "step": 25620,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.441836
    },
    {
      "epoch": 1.0978535624009254,
      "grad_norm": 4.35410737991333,
      "learning_rate": 8.85686100718441e-05,
      "loss": 2.5056549072265626,
      "memory(GiB)": 77.56,
      "step": 25625,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.441846
    },
    {
      "epoch": 1.0980677777301744,
      "grad_norm": 6.6620988845825195,
      "learning_rate": 8.856432700086104e-05,
      "loss": 2.506753349304199,
      "memory(GiB)": 77.56,
      "step": 25630,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.441814
    },
    {
      "epoch": 1.0982819930594234,
      "grad_norm": 3.700556755065918,
      "learning_rate": 8.856004323124903e-05,
      "loss": 2.7397764205932615,
      "memory(GiB)": 77.56,
      "step": 25635,
      "token_acc": 0.4461538461538462,
      "train_speed(iter/s)": 1.441782
    },
    {
      "epoch": 1.0984962083886722,
      "grad_norm": 3.60774827003479,
      "learning_rate": 8.855575876308566e-05,
      "loss": 2.2539976119995115,
      "memory(GiB)": 77.56,
      "step": 25640,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.44182
    },
    {
      "epoch": 1.0987104237179213,
      "grad_norm": 6.857401371002197,
      "learning_rate": 8.855147359644856e-05,
      "loss": 2.458940124511719,
      "memory(GiB)": 77.56,
      "step": 25645,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.441756
    },
    {
      "epoch": 1.0989246390471703,
      "grad_norm": 4.297567844390869,
      "learning_rate": 8.854718773141535e-05,
      "loss": 2.6402027130126955,
      "memory(GiB)": 77.56,
      "step": 25650,
      "token_acc": 0.45857988165680474,
      "train_speed(iter/s)": 1.441758
    },
    {
      "epoch": 1.0991388543764191,
      "grad_norm": 4.711802005767822,
      "learning_rate": 8.854290116806367e-05,
      "loss": 2.456376647949219,
      "memory(GiB)": 77.56,
      "step": 25655,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.441752
    },
    {
      "epoch": 1.0993530697056682,
      "grad_norm": 5.361373424530029,
      "learning_rate": 8.853861390647118e-05,
      "loss": 2.527352714538574,
      "memory(GiB)": 77.56,
      "step": 25660,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.441699
    },
    {
      "epoch": 1.0995672850349172,
      "grad_norm": 6.213389873504639,
      "learning_rate": 8.853432594671554e-05,
      "loss": 2.4731678009033202,
      "memory(GiB)": 77.56,
      "step": 25665,
      "token_acc": 0.4834710743801653,
      "train_speed(iter/s)": 1.441746
    },
    {
      "epoch": 1.099781500364166,
      "grad_norm": 4.930428981781006,
      "learning_rate": 8.853003728887446e-05,
      "loss": 2.4466569900512694,
      "memory(GiB)": 77.56,
      "step": 25670,
      "token_acc": 0.4568345323741007,
      "train_speed(iter/s)": 1.441795
    },
    {
      "epoch": 1.099995715693415,
      "grad_norm": 4.919266700744629,
      "learning_rate": 8.85257479330256e-05,
      "loss": 2.517359161376953,
      "memory(GiB)": 77.56,
      "step": 25675,
      "token_acc": 0.46494464944649444,
      "train_speed(iter/s)": 1.441808
    },
    {
      "epoch": 1.100209931022664,
      "grad_norm": 4.77811861038208,
      "learning_rate": 8.852145787924666e-05,
      "loss": 2.562126350402832,
      "memory(GiB)": 77.56,
      "step": 25680,
      "token_acc": 0.47720364741641336,
      "train_speed(iter/s)": 1.441797
    },
    {
      "epoch": 1.1004241463519129,
      "grad_norm": 4.611969947814941,
      "learning_rate": 8.851716712761538e-05,
      "loss": 2.1323192596435545,
      "memory(GiB)": 77.56,
      "step": 25685,
      "token_acc": 0.5317460317460317,
      "train_speed(iter/s)": 1.441781
    },
    {
      "epoch": 1.100638361681162,
      "grad_norm": 5.798098087310791,
      "learning_rate": 8.851287567820949e-05,
      "loss": 2.8261680603027344,
      "memory(GiB)": 77.56,
      "step": 25690,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.441673
    },
    {
      "epoch": 1.100852577010411,
      "grad_norm": 6.622003555297852,
      "learning_rate": 8.850858353110674e-05,
      "loss": 2.2589473724365234,
      "memory(GiB)": 77.56,
      "step": 25695,
      "token_acc": 0.5257731958762887,
      "train_speed(iter/s)": 1.441634
    },
    {
      "epoch": 1.1010667923396598,
      "grad_norm": 4.319852352142334,
      "learning_rate": 8.850429068638487e-05,
      "loss": 2.1083744049072264,
      "memory(GiB)": 77.56,
      "step": 25700,
      "token_acc": 0.555956678700361,
      "train_speed(iter/s)": 1.441598
    },
    {
      "epoch": 1.1012810076689088,
      "grad_norm": 4.97047233581543,
      "learning_rate": 8.849999714412165e-05,
      "loss": 2.4280059814453123,
      "memory(GiB)": 77.56,
      "step": 25705,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.441607
    },
    {
      "epoch": 1.1014952229981578,
      "grad_norm": 4.595216274261475,
      "learning_rate": 8.849570290439486e-05,
      "loss": 2.5880908966064453,
      "memory(GiB)": 77.56,
      "step": 25710,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.441625
    },
    {
      "epoch": 1.1017094383274066,
      "grad_norm": 3.864295482635498,
      "learning_rate": 8.84914079672823e-05,
      "loss": 2.6979808807373047,
      "memory(GiB)": 77.56,
      "step": 25715,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.441656
    },
    {
      "epoch": 1.1019236536566557,
      "grad_norm": 5.658937931060791,
      "learning_rate": 8.84871123328618e-05,
      "loss": 2.458305549621582,
      "memory(GiB)": 77.56,
      "step": 25720,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.441594
    },
    {
      "epoch": 1.1021378689859047,
      "grad_norm": 6.146386623382568,
      "learning_rate": 8.848281600121114e-05,
      "loss": 2.378607177734375,
      "memory(GiB)": 77.56,
      "step": 25725,
      "token_acc": 0.49795918367346936,
      "train_speed(iter/s)": 1.441559
    },
    {
      "epoch": 1.1023520843151535,
      "grad_norm": 4.334575176239014,
      "learning_rate": 8.847851897240815e-05,
      "loss": 2.3802772521972657,
      "memory(GiB)": 77.56,
      "step": 25730,
      "token_acc": 0.5204081632653061,
      "train_speed(iter/s)": 1.441587
    },
    {
      "epoch": 1.1025662996444026,
      "grad_norm": 7.283489227294922,
      "learning_rate": 8.847422124653072e-05,
      "loss": 2.2444238662719727,
      "memory(GiB)": 77.56,
      "step": 25735,
      "token_acc": 0.524822695035461,
      "train_speed(iter/s)": 1.441641
    },
    {
      "epoch": 1.1027805149736516,
      "grad_norm": 3.8427562713623047,
      "learning_rate": 8.846992282365667e-05,
      "loss": 2.221853256225586,
      "memory(GiB)": 77.56,
      "step": 25740,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.441684
    },
    {
      "epoch": 1.1029947303029004,
      "grad_norm": 5.28739595413208,
      "learning_rate": 8.846562370386389e-05,
      "loss": 2.318505859375,
      "memory(GiB)": 77.56,
      "step": 25745,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.441737
    },
    {
      "epoch": 1.1032089456321494,
      "grad_norm": 5.43912410736084,
      "learning_rate": 8.846132388723023e-05,
      "loss": 2.2960477828979493,
      "memory(GiB)": 77.56,
      "step": 25750,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.441768
    },
    {
      "epoch": 1.1034231609613985,
      "grad_norm": 5.258962631225586,
      "learning_rate": 8.845702337383363e-05,
      "loss": 2.4772518157958983,
      "memory(GiB)": 77.56,
      "step": 25755,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441764
    },
    {
      "epoch": 1.1036373762906473,
      "grad_norm": 5.141008377075195,
      "learning_rate": 8.845272216375196e-05,
      "loss": 2.6910627365112303,
      "memory(GiB)": 77.56,
      "step": 25760,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.441791
    },
    {
      "epoch": 1.1038515916198963,
      "grad_norm": 5.876232147216797,
      "learning_rate": 8.844842025706316e-05,
      "loss": 2.3275226593017577,
      "memory(GiB)": 77.56,
      "step": 25765,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.441832
    },
    {
      "epoch": 1.1040658069491454,
      "grad_norm": 5.5710835456848145,
      "learning_rate": 8.844411765384517e-05,
      "loss": 2.2320301055908205,
      "memory(GiB)": 77.56,
      "step": 25770,
      "token_acc": 0.5355805243445693,
      "train_speed(iter/s)": 1.441865
    },
    {
      "epoch": 1.1042800222783942,
      "grad_norm": 4.297060489654541,
      "learning_rate": 8.843981435417592e-05,
      "loss": 2.644546890258789,
      "memory(GiB)": 77.56,
      "step": 25775,
      "token_acc": 0.44936708860759494,
      "train_speed(iter/s)": 1.441847
    },
    {
      "epoch": 1.1044942376076432,
      "grad_norm": 4.347912311553955,
      "learning_rate": 8.843551035813337e-05,
      "loss": 2.3509782791137694,
      "memory(GiB)": 77.56,
      "step": 25780,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.441892
    },
    {
      "epoch": 1.1047084529368922,
      "grad_norm": 5.660109043121338,
      "learning_rate": 8.84312056657955e-05,
      "loss": 2.2397388458251952,
      "memory(GiB)": 77.56,
      "step": 25785,
      "token_acc": 0.4900398406374502,
      "train_speed(iter/s)": 1.441854
    },
    {
      "epoch": 1.104922668266141,
      "grad_norm": 5.345829010009766,
      "learning_rate": 8.842690027724029e-05,
      "loss": 2.4885114669799804,
      "memory(GiB)": 77.56,
      "step": 25790,
      "token_acc": 0.47183098591549294,
      "train_speed(iter/s)": 1.44187
    },
    {
      "epoch": 1.10513688359539,
      "grad_norm": 4.77627420425415,
      "learning_rate": 8.842259419254573e-05,
      "loss": 2.8829463958740233,
      "memory(GiB)": 77.56,
      "step": 25795,
      "token_acc": 0.38848920863309355,
      "train_speed(iter/s)": 1.4419
    },
    {
      "epoch": 1.1053510989246391,
      "grad_norm": 8.10155200958252,
      "learning_rate": 8.841828741178984e-05,
      "loss": 2.7255359649658204,
      "memory(GiB)": 77.56,
      "step": 25800,
      "token_acc": 0.5041322314049587,
      "train_speed(iter/s)": 1.441914
    },
    {
      "epoch": 1.105565314253888,
      "grad_norm": 6.050825119018555,
      "learning_rate": 8.841397993505062e-05,
      "loss": 2.590948486328125,
      "memory(GiB)": 77.56,
      "step": 25805,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.441939
    },
    {
      "epoch": 1.105779529583137,
      "grad_norm": 4.583866596221924,
      "learning_rate": 8.840967176240612e-05,
      "loss": 2.4210124969482423,
      "memory(GiB)": 77.56,
      "step": 25810,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.441939
    },
    {
      "epoch": 1.105993744912386,
      "grad_norm": 5.649902820587158,
      "learning_rate": 8.840536289393439e-05,
      "loss": 2.4256710052490233,
      "memory(GiB)": 77.56,
      "step": 25815,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.441874
    },
    {
      "epoch": 1.1062079602416348,
      "grad_norm": 7.8919172286987305,
      "learning_rate": 8.840105332971348e-05,
      "loss": 2.5325475692749024,
      "memory(GiB)": 77.56,
      "step": 25820,
      "token_acc": 0.4769874476987448,
      "train_speed(iter/s)": 1.441892
    },
    {
      "epoch": 1.1064221755708838,
      "grad_norm": 5.311280250549316,
      "learning_rate": 8.839674306982148e-05,
      "loss": 2.556719970703125,
      "memory(GiB)": 77.56,
      "step": 25825,
      "token_acc": 0.4148606811145511,
      "train_speed(iter/s)": 1.441943
    },
    {
      "epoch": 1.1066363909001329,
      "grad_norm": 4.978405475616455,
      "learning_rate": 8.839243211433645e-05,
      "loss": 2.3605009078979493,
      "memory(GiB)": 77.56,
      "step": 25830,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.441958
    },
    {
      "epoch": 1.1068506062293817,
      "grad_norm": 4.0087056159973145,
      "learning_rate": 8.838812046333648e-05,
      "loss": 2.4748769760131837,
      "memory(GiB)": 77.56,
      "step": 25835,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.44195
    },
    {
      "epoch": 1.1070648215586307,
      "grad_norm": 5.239276885986328,
      "learning_rate": 8.838380811689973e-05,
      "loss": 2.437875747680664,
      "memory(GiB)": 77.56,
      "step": 25840,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.441853
    },
    {
      "epoch": 1.1072790368878798,
      "grad_norm": 5.4493889808654785,
      "learning_rate": 8.837949507510427e-05,
      "loss": 2.2718408584594725,
      "memory(GiB)": 77.56,
      "step": 25845,
      "token_acc": 0.5393258426966292,
      "train_speed(iter/s)": 1.441857
    },
    {
      "epoch": 1.1074932522171286,
      "grad_norm": 8.784475326538086,
      "learning_rate": 8.837518133802826e-05,
      "loss": 2.124480438232422,
      "memory(GiB)": 77.56,
      "step": 25850,
      "token_acc": 0.5546875,
      "train_speed(iter/s)": 1.441957
    },
    {
      "epoch": 1.1077074675463776,
      "grad_norm": 5.088685035705566,
      "learning_rate": 8.837086690574983e-05,
      "loss": 2.448760223388672,
      "memory(GiB)": 77.56,
      "step": 25855,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.441989
    },
    {
      "epoch": 1.1079216828756266,
      "grad_norm": 6.981198310852051,
      "learning_rate": 8.836655177834716e-05,
      "loss": 2.1593671798706056,
      "memory(GiB)": 77.56,
      "step": 25860,
      "token_acc": 0.5173913043478261,
      "train_speed(iter/s)": 1.442022
    },
    {
      "epoch": 1.1081358982048755,
      "grad_norm": 4.095747947692871,
      "learning_rate": 8.836223595589842e-05,
      "loss": 2.2636686325073243,
      "memory(GiB)": 77.56,
      "step": 25865,
      "token_acc": 0.5050505050505051,
      "train_speed(iter/s)": 1.442001
    },
    {
      "epoch": 1.1083501135341245,
      "grad_norm": 6.862861633300781,
      "learning_rate": 8.835791943848178e-05,
      "loss": 2.42894287109375,
      "memory(GiB)": 77.56,
      "step": 25870,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.442041
    },
    {
      "epoch": 1.1085643288633735,
      "grad_norm": 4.592063903808594,
      "learning_rate": 8.835360222617545e-05,
      "loss": 2.7197782516479494,
      "memory(GiB)": 77.56,
      "step": 25875,
      "token_acc": 0.4072948328267477,
      "train_speed(iter/s)": 1.442071
    },
    {
      "epoch": 1.1087785441926223,
      "grad_norm": 5.136329174041748,
      "learning_rate": 8.834928431905764e-05,
      "loss": 2.357573127746582,
      "memory(GiB)": 77.56,
      "step": 25880,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.442088
    },
    {
      "epoch": 1.1089927595218714,
      "grad_norm": 6.3618855476379395,
      "learning_rate": 8.834496571720657e-05,
      "loss": 2.4693374633789062,
      "memory(GiB)": 77.56,
      "step": 25885,
      "token_acc": 0.4246031746031746,
      "train_speed(iter/s)": 1.442081
    },
    {
      "epoch": 1.1092069748511204,
      "grad_norm": 4.026154518127441,
      "learning_rate": 8.834064642070045e-05,
      "loss": 2.4066232681274413,
      "memory(GiB)": 77.56,
      "step": 25890,
      "token_acc": 0.5435540069686411,
      "train_speed(iter/s)": 1.442062
    },
    {
      "epoch": 1.1094211901803692,
      "grad_norm": 4.661164283752441,
      "learning_rate": 8.833632642961759e-05,
      "loss": 2.4651983261108397,
      "memory(GiB)": 77.56,
      "step": 25895,
      "token_acc": 0.4847457627118644,
      "train_speed(iter/s)": 1.442117
    },
    {
      "epoch": 1.1096354055096183,
      "grad_norm": 4.34984827041626,
      "learning_rate": 8.83320057440362e-05,
      "loss": 2.1653614044189453,
      "memory(GiB)": 77.56,
      "step": 25900,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.442182
    },
    {
      "epoch": 1.1098496208388673,
      "grad_norm": 6.945590496063232,
      "learning_rate": 8.832768436403455e-05,
      "loss": 2.418035888671875,
      "memory(GiB)": 77.56,
      "step": 25905,
      "token_acc": 0.5021459227467812,
      "train_speed(iter/s)": 1.442229
    },
    {
      "epoch": 1.110063836168116,
      "grad_norm": 7.615697383880615,
      "learning_rate": 8.832336228969098e-05,
      "loss": 2.4347999572753904,
      "memory(GiB)": 77.56,
      "step": 25910,
      "token_acc": 0.4919093851132686,
      "train_speed(iter/s)": 1.442226
    },
    {
      "epoch": 1.1102780514973651,
      "grad_norm": 4.93060827255249,
      "learning_rate": 8.831903952108372e-05,
      "loss": 2.6519432067871094,
      "memory(GiB)": 77.56,
      "step": 25915,
      "token_acc": 0.38235294117647056,
      "train_speed(iter/s)": 1.442238
    },
    {
      "epoch": 1.1104922668266142,
      "grad_norm": 4.502770900726318,
      "learning_rate": 8.831471605829113e-05,
      "loss": 2.637506103515625,
      "memory(GiB)": 77.56,
      "step": 25920,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.442243
    },
    {
      "epoch": 1.110706482155863,
      "grad_norm": 5.694281101226807,
      "learning_rate": 8.831039190139152e-05,
      "loss": 2.467295265197754,
      "memory(GiB)": 77.56,
      "step": 25925,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.442203
    },
    {
      "epoch": 1.110920697485112,
      "grad_norm": 7.324888706207275,
      "learning_rate": 8.830606705046321e-05,
      "loss": 2.5284379959106444,
      "memory(GiB)": 77.56,
      "step": 25930,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.442205
    },
    {
      "epoch": 1.111134912814361,
      "grad_norm": 5.400707721710205,
      "learning_rate": 8.830174150558456e-05,
      "loss": 2.1361209869384767,
      "memory(GiB)": 77.56,
      "step": 25935,
      "token_acc": 0.54296875,
      "train_speed(iter/s)": 1.442123
    },
    {
      "epoch": 1.1113491281436099,
      "grad_norm": 4.79054069519043,
      "learning_rate": 8.829741526683394e-05,
      "loss": 2.491075325012207,
      "memory(GiB)": 77.56,
      "step": 25940,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.442161
    },
    {
      "epoch": 1.111563343472859,
      "grad_norm": 4.609335899353027,
      "learning_rate": 8.829308833428972e-05,
      "loss": 2.286872673034668,
      "memory(GiB)": 77.56,
      "step": 25945,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.442177
    },
    {
      "epoch": 1.111777558802108,
      "grad_norm": 6.0736308097839355,
      "learning_rate": 8.828876070803028e-05,
      "loss": 2.6855098724365236,
      "memory(GiB)": 77.56,
      "step": 25950,
      "token_acc": 0.43103448275862066,
      "train_speed(iter/s)": 1.442206
    },
    {
      "epoch": 1.1119917741313567,
      "grad_norm": 5.716984272003174,
      "learning_rate": 8.828443238813402e-05,
      "loss": 2.6547584533691406,
      "memory(GiB)": 77.56,
      "step": 25955,
      "token_acc": 0.43288590604026844,
      "train_speed(iter/s)": 1.442188
    },
    {
      "epoch": 1.1122059894606058,
      "grad_norm": 6.626274108886719,
      "learning_rate": 8.828010337467936e-05,
      "loss": 2.5188533782958986,
      "memory(GiB)": 77.56,
      "step": 25960,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.442182
    },
    {
      "epoch": 1.1124202047898548,
      "grad_norm": 4.773168563842773,
      "learning_rate": 8.82757736677447e-05,
      "loss": 2.5926979064941404,
      "memory(GiB)": 77.56,
      "step": 25965,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.442128
    },
    {
      "epoch": 1.1126344201191036,
      "grad_norm": 5.194952964782715,
      "learning_rate": 8.82714432674085e-05,
      "loss": 2.8051170349121093,
      "memory(GiB)": 77.56,
      "step": 25970,
      "token_acc": 0.41896024464831805,
      "train_speed(iter/s)": 1.442127
    },
    {
      "epoch": 1.1128486354483527,
      "grad_norm": 5.4387311935424805,
      "learning_rate": 8.826711217374921e-05,
      "loss": 2.7437679290771486,
      "memory(GiB)": 77.56,
      "step": 25975,
      "token_acc": 0.45180722891566266,
      "train_speed(iter/s)": 1.442167
    },
    {
      "epoch": 1.1130628507776017,
      "grad_norm": 5.421558380126953,
      "learning_rate": 8.826278038684529e-05,
      "loss": 2.295772171020508,
      "memory(GiB)": 77.56,
      "step": 25980,
      "token_acc": 0.5541125541125541,
      "train_speed(iter/s)": 1.442096
    },
    {
      "epoch": 1.1132770661068505,
      "grad_norm": 5.219773769378662,
      "learning_rate": 8.82584479067752e-05,
      "loss": 2.402861213684082,
      "memory(GiB)": 77.56,
      "step": 25985,
      "token_acc": 0.5071942446043165,
      "train_speed(iter/s)": 1.442062
    },
    {
      "epoch": 1.1134912814360995,
      "grad_norm": 5.060913562774658,
      "learning_rate": 8.825411473361745e-05,
      "loss": 2.620148849487305,
      "memory(GiB)": 77.56,
      "step": 25990,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.442051
    },
    {
      "epoch": 1.1137054967653486,
      "grad_norm": 3.975721597671509,
      "learning_rate": 8.824978086745051e-05,
      "loss": 2.5187034606933594,
      "memory(GiB)": 77.56,
      "step": 25995,
      "token_acc": 0.49642857142857144,
      "train_speed(iter/s)": 1.442092
    },
    {
      "epoch": 1.1139197120945974,
      "grad_norm": 4.269970893859863,
      "learning_rate": 8.824544630835293e-05,
      "loss": 2.250044059753418,
      "memory(GiB)": 77.56,
      "step": 26000,
      "token_acc": 0.5175879396984925,
      "train_speed(iter/s)": 1.442168
    },
    {
      "epoch": 1.1139197120945974,
      "eval_loss": 2.251314640045166,
      "eval_runtime": 13.7991,
      "eval_samples_per_second": 7.247,
      "eval_steps_per_second": 7.247,
      "eval_token_acc": 0.47183098591549294,
      "step": 26000
    },
    {
      "epoch": 1.1141339274238464,
      "grad_norm": 4.331987380981445,
      "learning_rate": 8.82411110564032e-05,
      "loss": 2.6966156005859374,
      "memory(GiB)": 77.56,
      "step": 26005,
      "token_acc": 0.45722713864306785,
      "train_speed(iter/s)": 1.441025
    },
    {
      "epoch": 1.1143481427530955,
      "grad_norm": 5.541085243225098,
      "learning_rate": 8.823677511167986e-05,
      "loss": 2.686792182922363,
      "memory(GiB)": 77.56,
      "step": 26010,
      "token_acc": 0.44014084507042256,
      "train_speed(iter/s)": 1.440997
    },
    {
      "epoch": 1.1145623580823443,
      "grad_norm": 4.396245002746582,
      "learning_rate": 8.823243847426148e-05,
      "loss": 2.497194290161133,
      "memory(GiB)": 77.56,
      "step": 26015,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.441058
    },
    {
      "epoch": 1.1147765734115933,
      "grad_norm": 5.339602947235107,
      "learning_rate": 8.822810114422662e-05,
      "loss": 2.597373199462891,
      "memory(GiB)": 77.56,
      "step": 26020,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.441028
    },
    {
      "epoch": 1.1149907887408423,
      "grad_norm": 5.3471574783325195,
      "learning_rate": 8.822376312165384e-05,
      "loss": 2.3846054077148438,
      "memory(GiB)": 77.56,
      "step": 26025,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.440978
    },
    {
      "epoch": 1.1152050040700912,
      "grad_norm": 3.5960137844085693,
      "learning_rate": 8.821942440662172e-05,
      "loss": 2.50716552734375,
      "memory(GiB)": 77.56,
      "step": 26030,
      "token_acc": 0.46703296703296704,
      "train_speed(iter/s)": 1.441037
    },
    {
      "epoch": 1.1154192193993402,
      "grad_norm": 4.165687084197998,
      "learning_rate": 8.821508499920889e-05,
      "loss": 2.470718002319336,
      "memory(GiB)": 77.56,
      "step": 26035,
      "token_acc": 0.45348837209302323,
      "train_speed(iter/s)": 1.441029
    },
    {
      "epoch": 1.1156334347285892,
      "grad_norm": 4.627836227416992,
      "learning_rate": 8.821074489949395e-05,
      "loss": 2.4498682022094727,
      "memory(GiB)": 77.56,
      "step": 26040,
      "token_acc": 0.46946564885496184,
      "train_speed(iter/s)": 1.440992
    },
    {
      "epoch": 1.115847650057838,
      "grad_norm": 5.555534839630127,
      "learning_rate": 8.820640410755551e-05,
      "loss": 2.49194393157959,
      "memory(GiB)": 77.56,
      "step": 26045,
      "token_acc": 0.44654088050314467,
      "train_speed(iter/s)": 1.44102
    },
    {
      "epoch": 1.116061865387087,
      "grad_norm": 4.856287956237793,
      "learning_rate": 8.820206262347221e-05,
      "loss": 2.7558454513549804,
      "memory(GiB)": 77.56,
      "step": 26050,
      "token_acc": 0.47706422018348627,
      "train_speed(iter/s)": 1.44107
    },
    {
      "epoch": 1.116276080716336,
      "grad_norm": 5.018049716949463,
      "learning_rate": 8.819772044732274e-05,
      "loss": 2.62425537109375,
      "memory(GiB)": 77.56,
      "step": 26055,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.441107
    },
    {
      "epoch": 1.1164902960455851,
      "grad_norm": 5.306628704071045,
      "learning_rate": 8.81933775791857e-05,
      "loss": 2.5087875366210937,
      "memory(GiB)": 77.56,
      "step": 26060,
      "token_acc": 0.5375494071146245,
      "train_speed(iter/s)": 1.441155
    },
    {
      "epoch": 1.116704511374834,
      "grad_norm": 5.25780725479126,
      "learning_rate": 8.818903401913983e-05,
      "loss": 2.4063173294067384,
      "memory(GiB)": 77.56,
      "step": 26065,
      "token_acc": 0.4790874524714829,
      "train_speed(iter/s)": 1.441202
    },
    {
      "epoch": 1.116918726704083,
      "grad_norm": 7.838450908660889,
      "learning_rate": 8.818468976726377e-05,
      "loss": 2.5843784332275392,
      "memory(GiB)": 77.56,
      "step": 26070,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.441274
    },
    {
      "epoch": 1.117132942033332,
      "grad_norm": 6.996995449066162,
      "learning_rate": 8.818034482363622e-05,
      "loss": 2.3615653991699217,
      "memory(GiB)": 77.56,
      "step": 26075,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.441275
    },
    {
      "epoch": 1.1173471573625808,
      "grad_norm": 4.48069429397583,
      "learning_rate": 8.817599918833593e-05,
      "loss": 2.4722389221191405,
      "memory(GiB)": 77.56,
      "step": 26080,
      "token_acc": 0.5136986301369864,
      "train_speed(iter/s)": 1.441314
    },
    {
      "epoch": 1.1175613726918299,
      "grad_norm": 6.915204048156738,
      "learning_rate": 8.817165286144158e-05,
      "loss": 2.7206195831298827,
      "memory(GiB)": 77.56,
      "step": 26085,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.441349
    },
    {
      "epoch": 1.117775588021079,
      "grad_norm": 5.575260162353516,
      "learning_rate": 8.816730584303194e-05,
      "loss": 2.3187282562255858,
      "memory(GiB)": 77.56,
      "step": 26090,
      "token_acc": 0.5344827586206896,
      "train_speed(iter/s)": 1.441359
    },
    {
      "epoch": 1.1179898033503277,
      "grad_norm": 5.173477649688721,
      "learning_rate": 8.816295813318576e-05,
      "loss": 2.7541292190551756,
      "memory(GiB)": 77.56,
      "step": 26095,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.441352
    },
    {
      "epoch": 1.1182040186795768,
      "grad_norm": 3.787165880203247,
      "learning_rate": 8.815860973198177e-05,
      "loss": 2.2994710922241213,
      "memory(GiB)": 77.56,
      "step": 26100,
      "token_acc": 0.563265306122449,
      "train_speed(iter/s)": 1.441437
    },
    {
      "epoch": 1.1184182340088258,
      "grad_norm": 5.556142330169678,
      "learning_rate": 8.81542606394988e-05,
      "loss": 2.636882209777832,
      "memory(GiB)": 77.56,
      "step": 26105,
      "token_acc": 0.4483870967741935,
      "train_speed(iter/s)": 1.441432
    },
    {
      "epoch": 1.1186324493380746,
      "grad_norm": 5.094545841217041,
      "learning_rate": 8.814991085581559e-05,
      "loss": 2.1735029220581055,
      "memory(GiB)": 77.56,
      "step": 26110,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.44151
    },
    {
      "epoch": 1.1188466646673236,
      "grad_norm": 5.514937877655029,
      "learning_rate": 8.814556038101097e-05,
      "loss": 2.409612274169922,
      "memory(GiB)": 77.56,
      "step": 26115,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.441497
    },
    {
      "epoch": 1.1190608799965727,
      "grad_norm": 8.324711799621582,
      "learning_rate": 8.814120921516372e-05,
      "loss": 2.4711477279663088,
      "memory(GiB)": 77.56,
      "step": 26120,
      "token_acc": 0.46215139442231074,
      "train_speed(iter/s)": 1.441562
    },
    {
      "epoch": 1.1192750953258215,
      "grad_norm": 4.866443634033203,
      "learning_rate": 8.81368573583527e-05,
      "loss": 2.368330383300781,
      "memory(GiB)": 77.56,
      "step": 26125,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.44152
    },
    {
      "epoch": 1.1194893106550705,
      "grad_norm": 4.037306308746338,
      "learning_rate": 8.813250481065673e-05,
      "loss": 2.1914657592773437,
      "memory(GiB)": 77.56,
      "step": 26130,
      "token_acc": 0.5343511450381679,
      "train_speed(iter/s)": 1.441549
    },
    {
      "epoch": 1.1197035259843195,
      "grad_norm": 9.574823379516602,
      "learning_rate": 8.812815157215466e-05,
      "loss": 2.874228858947754,
      "memory(GiB)": 77.56,
      "step": 26135,
      "token_acc": 0.4337748344370861,
      "train_speed(iter/s)": 1.441601
    },
    {
      "epoch": 1.1199177413135684,
      "grad_norm": 4.67606782913208,
      "learning_rate": 8.812379764292535e-05,
      "loss": 2.550762939453125,
      "memory(GiB)": 77.56,
      "step": 26140,
      "token_acc": 0.49480968858131485,
      "train_speed(iter/s)": 1.441655
    },
    {
      "epoch": 1.1201319566428174,
      "grad_norm": 5.895148754119873,
      "learning_rate": 8.81194430230477e-05,
      "loss": 2.947303581237793,
      "memory(GiB)": 77.56,
      "step": 26145,
      "token_acc": 0.42946708463949845,
      "train_speed(iter/s)": 1.441708
    },
    {
      "epoch": 1.1203461719720664,
      "grad_norm": 4.9481377601623535,
      "learning_rate": 8.811508771260058e-05,
      "loss": 2.57385368347168,
      "memory(GiB)": 77.56,
      "step": 26150,
      "token_acc": 0.4423076923076923,
      "train_speed(iter/s)": 1.441689
    },
    {
      "epoch": 1.1205603873013152,
      "grad_norm": 5.031641960144043,
      "learning_rate": 8.811073171166288e-05,
      "loss": 2.6819974899291994,
      "memory(GiB)": 77.56,
      "step": 26155,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.441711
    },
    {
      "epoch": 1.1207746026305643,
      "grad_norm": 5.014657497406006,
      "learning_rate": 8.810637502031354e-05,
      "loss": 2.2858190536499023,
      "memory(GiB)": 77.56,
      "step": 26160,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.441557
    },
    {
      "epoch": 1.1209888179598133,
      "grad_norm": 4.458937644958496,
      "learning_rate": 8.810201763863145e-05,
      "loss": 2.7890573501586915,
      "memory(GiB)": 77.56,
      "step": 26165,
      "token_acc": 0.44876325088339225,
      "train_speed(iter/s)": 1.441557
    },
    {
      "epoch": 1.1212030332890621,
      "grad_norm": 5.7796630859375,
      "learning_rate": 8.809765956669558e-05,
      "loss": 2.99735107421875,
      "memory(GiB)": 77.56,
      "step": 26170,
      "token_acc": 0.42748091603053434,
      "train_speed(iter/s)": 1.441611
    },
    {
      "epoch": 1.1214172486183112,
      "grad_norm": 4.740631103515625,
      "learning_rate": 8.809330080458487e-05,
      "loss": 2.7155364990234374,
      "memory(GiB)": 77.56,
      "step": 26175,
      "token_acc": 0.43824701195219123,
      "train_speed(iter/s)": 1.441661
    },
    {
      "epoch": 1.1216314639475602,
      "grad_norm": 4.7244038581848145,
      "learning_rate": 8.808894135237828e-05,
      "loss": 2.4301692962646486,
      "memory(GiB)": 77.56,
      "step": 26180,
      "token_acc": 0.5014164305949008,
      "train_speed(iter/s)": 1.441648
    },
    {
      "epoch": 1.121845679276809,
      "grad_norm": 6.2575178146362305,
      "learning_rate": 8.808458121015477e-05,
      "loss": 2.600922966003418,
      "memory(GiB)": 77.56,
      "step": 26185,
      "token_acc": 0.4337748344370861,
      "train_speed(iter/s)": 1.441592
    },
    {
      "epoch": 1.122059894606058,
      "grad_norm": 4.577580451965332,
      "learning_rate": 8.808022037799337e-05,
      "loss": 2.485809326171875,
      "memory(GiB)": 77.56,
      "step": 26190,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.4416
    },
    {
      "epoch": 1.122274109935307,
      "grad_norm": 3.8281726837158203,
      "learning_rate": 8.807585885597305e-05,
      "loss": 2.5327531814575197,
      "memory(GiB)": 77.56,
      "step": 26195,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.441584
    },
    {
      "epoch": 1.1224883252645559,
      "grad_norm": 4.452630043029785,
      "learning_rate": 8.807149664417283e-05,
      "loss": 2.776981735229492,
      "memory(GiB)": 77.56,
      "step": 26200,
      "token_acc": 0.4430769230769231,
      "train_speed(iter/s)": 1.441576
    },
    {
      "epoch": 1.122702540593805,
      "grad_norm": 3.939777135848999,
      "learning_rate": 8.806713374267172e-05,
      "loss": 2.3384021759033202,
      "memory(GiB)": 77.56,
      "step": 26205,
      "token_acc": 0.5186567164179104,
      "train_speed(iter/s)": 1.441603
    },
    {
      "epoch": 1.122916755923054,
      "grad_norm": 5.321191787719727,
      "learning_rate": 8.806277015154877e-05,
      "loss": 2.4105514526367187,
      "memory(GiB)": 77.56,
      "step": 26210,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.441582
    },
    {
      "epoch": 1.1231309712523028,
      "grad_norm": 5.572187423706055,
      "learning_rate": 8.805840587088305e-05,
      "loss": 2.729194450378418,
      "memory(GiB)": 77.56,
      "step": 26215,
      "token_acc": 0.44482758620689655,
      "train_speed(iter/s)": 1.441523
    },
    {
      "epoch": 1.1233451865815518,
      "grad_norm": 3.8347079753875732,
      "learning_rate": 8.80540409007536e-05,
      "loss": 2.781502342224121,
      "memory(GiB)": 77.56,
      "step": 26220,
      "token_acc": 0.45151515151515154,
      "train_speed(iter/s)": 1.441555
    },
    {
      "epoch": 1.1235594019108008,
      "grad_norm": 4.109711170196533,
      "learning_rate": 8.80496752412395e-05,
      "loss": 2.345529556274414,
      "memory(GiB)": 77.56,
      "step": 26225,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.441532
    },
    {
      "epoch": 1.1237736172400496,
      "grad_norm": 6.2818732261657715,
      "learning_rate": 8.804530889241984e-05,
      "loss": 2.4447484970092774,
      "memory(GiB)": 77.56,
      "step": 26230,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.441535
    },
    {
      "epoch": 1.1239878325692987,
      "grad_norm": 4.4610676765441895,
      "learning_rate": 8.804094185437373e-05,
      "loss": 2.7864213943481446,
      "memory(GiB)": 77.56,
      "step": 26235,
      "token_acc": 0.43492063492063493,
      "train_speed(iter/s)": 1.441561
    },
    {
      "epoch": 1.1242020478985477,
      "grad_norm": 4.987005710601807,
      "learning_rate": 8.803657412718025e-05,
      "loss": 2.395465850830078,
      "memory(GiB)": 77.56,
      "step": 26240,
      "token_acc": 0.48299319727891155,
      "train_speed(iter/s)": 1.441586
    },
    {
      "epoch": 1.1244162632277965,
      "grad_norm": 4.750020980834961,
      "learning_rate": 8.803220571091857e-05,
      "loss": 2.5770553588867187,
      "memory(GiB)": 77.56,
      "step": 26245,
      "token_acc": 0.48089171974522293,
      "train_speed(iter/s)": 1.441585
    },
    {
      "epoch": 1.1246304785570456,
      "grad_norm": 5.316093444824219,
      "learning_rate": 8.80278366056678e-05,
      "loss": 2.4066978454589845,
      "memory(GiB)": 77.56,
      "step": 26250,
      "token_acc": 0.47278911564625853,
      "train_speed(iter/s)": 1.441614
    },
    {
      "epoch": 1.1248446938862946,
      "grad_norm": 8.964237213134766,
      "learning_rate": 8.80234668115071e-05,
      "loss": 2.245305061340332,
      "memory(GiB)": 77.56,
      "step": 26255,
      "token_acc": 0.4901185770750988,
      "train_speed(iter/s)": 1.441625
    },
    {
      "epoch": 1.1250589092155434,
      "grad_norm": 3.6716768741607666,
      "learning_rate": 8.801909632851561e-05,
      "loss": 2.6376131057739256,
      "memory(GiB)": 77.56,
      "step": 26260,
      "token_acc": 0.4307692307692308,
      "train_speed(iter/s)": 1.441581
    },
    {
      "epoch": 1.1252731245447924,
      "grad_norm": 6.126744747161865,
      "learning_rate": 8.801472515677255e-05,
      "loss": 2.5930644989013674,
      "memory(GiB)": 77.56,
      "step": 26265,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.441585
    },
    {
      "epoch": 1.1254873398740415,
      "grad_norm": 4.10883903503418,
      "learning_rate": 8.801035329635707e-05,
      "loss": 2.4922943115234375,
      "memory(GiB)": 77.56,
      "step": 26270,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.441547
    },
    {
      "epoch": 1.1257015552032903,
      "grad_norm": 3.914081335067749,
      "learning_rate": 8.800598074734839e-05,
      "loss": 2.489690589904785,
      "memory(GiB)": 77.56,
      "step": 26275,
      "token_acc": 0.44662921348314605,
      "train_speed(iter/s)": 1.441582
    },
    {
      "epoch": 1.1259157705325393,
      "grad_norm": 6.201168060302734,
      "learning_rate": 8.800160750982572e-05,
      "loss": 2.7673282623291016,
      "memory(GiB)": 77.56,
      "step": 26280,
      "token_acc": 0.42813455657492355,
      "train_speed(iter/s)": 1.44164
    },
    {
      "epoch": 1.1261299858617884,
      "grad_norm": 3.8510451316833496,
      "learning_rate": 8.799723358386826e-05,
      "loss": 2.39993839263916,
      "memory(GiB)": 77.56,
      "step": 26285,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.441629
    },
    {
      "epoch": 1.1263442011910372,
      "grad_norm": 4.904791831970215,
      "learning_rate": 8.79928589695553e-05,
      "loss": 2.6010610580444338,
      "memory(GiB)": 77.56,
      "step": 26290,
      "token_acc": 0.4607142857142857,
      "train_speed(iter/s)": 1.441681
    },
    {
      "epoch": 1.1265584165202862,
      "grad_norm": 3.5493972301483154,
      "learning_rate": 8.798848366696604e-05,
      "loss": 2.854984664916992,
      "memory(GiB)": 77.56,
      "step": 26295,
      "token_acc": 0.4559748427672956,
      "train_speed(iter/s)": 1.441662
    },
    {
      "epoch": 1.1267726318495352,
      "grad_norm": 5.556482315063477,
      "learning_rate": 8.798410767617977e-05,
      "loss": 2.749562644958496,
      "memory(GiB)": 77.56,
      "step": 26300,
      "token_acc": 0.47019867549668876,
      "train_speed(iter/s)": 1.441629
    },
    {
      "epoch": 1.126986847178784,
      "grad_norm": 4.702717304229736,
      "learning_rate": 8.797973099727575e-05,
      "loss": 2.464116096496582,
      "memory(GiB)": 77.56,
      "step": 26305,
      "token_acc": 0.5225563909774437,
      "train_speed(iter/s)": 1.441607
    },
    {
      "epoch": 1.127201062508033,
      "grad_norm": 4.054986000061035,
      "learning_rate": 8.797535363033327e-05,
      "loss": 2.187481689453125,
      "memory(GiB)": 77.56,
      "step": 26310,
      "token_acc": 0.5567765567765568,
      "train_speed(iter/s)": 1.441527
    },
    {
      "epoch": 1.1274152778372821,
      "grad_norm": 4.942107677459717,
      "learning_rate": 8.797097557543166e-05,
      "loss": 2.5402469635009766,
      "memory(GiB)": 77.56,
      "step": 26315,
      "token_acc": 0.4652777777777778,
      "train_speed(iter/s)": 1.441487
    },
    {
      "epoch": 1.127629493166531,
      "grad_norm": 4.518787860870361,
      "learning_rate": 8.796659683265019e-05,
      "loss": 2.1068132400512694,
      "memory(GiB)": 77.56,
      "step": 26320,
      "token_acc": 0.5450819672131147,
      "train_speed(iter/s)": 1.441442
    },
    {
      "epoch": 1.12784370849578,
      "grad_norm": 3.804300308227539,
      "learning_rate": 8.796221740206819e-05,
      "loss": 2.612284469604492,
      "memory(GiB)": 77.56,
      "step": 26325,
      "token_acc": 0.47214076246334313,
      "train_speed(iter/s)": 1.441416
    },
    {
      "epoch": 1.128057923825029,
      "grad_norm": 5.4769086837768555,
      "learning_rate": 8.795783728376504e-05,
      "loss": 2.512421226501465,
      "memory(GiB)": 77.56,
      "step": 26330,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.441471
    },
    {
      "epoch": 1.1282721391542778,
      "grad_norm": 5.546802997589111,
      "learning_rate": 8.795345647782002e-05,
      "loss": 2.368264579772949,
      "memory(GiB)": 77.56,
      "step": 26335,
      "token_acc": 0.47307692307692306,
      "train_speed(iter/s)": 1.44141
    },
    {
      "epoch": 1.1284863544835269,
      "grad_norm": 17.33209991455078,
      "learning_rate": 8.794907498431258e-05,
      "loss": 2.432367706298828,
      "memory(GiB)": 77.56,
      "step": 26340,
      "token_acc": 0.5124555160142349,
      "train_speed(iter/s)": 1.441372
    },
    {
      "epoch": 1.128700569812776,
      "grad_norm": 5.860182762145996,
      "learning_rate": 8.794469280332203e-05,
      "loss": 2.4527870178222657,
      "memory(GiB)": 77.56,
      "step": 26345,
      "token_acc": 0.46258503401360546,
      "train_speed(iter/s)": 1.441356
    },
    {
      "epoch": 1.1289147851420247,
      "grad_norm": 4.315380573272705,
      "learning_rate": 8.794030993492777e-05,
      "loss": 2.508917808532715,
      "memory(GiB)": 77.56,
      "step": 26350,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.441335
    },
    {
      "epoch": 1.1291290004712737,
      "grad_norm": 5.9567365646362305,
      "learning_rate": 8.79359263792092e-05,
      "loss": 2.632153129577637,
      "memory(GiB)": 77.56,
      "step": 26355,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.441371
    },
    {
      "epoch": 1.1293432158005228,
      "grad_norm": 4.8569231033325195,
      "learning_rate": 8.793154213624572e-05,
      "loss": 2.5251413345336915,
      "memory(GiB)": 77.56,
      "step": 26360,
      "token_acc": 0.483271375464684,
      "train_speed(iter/s)": 1.441425
    },
    {
      "epoch": 1.1295574311297716,
      "grad_norm": 6.091347694396973,
      "learning_rate": 8.792715720611682e-05,
      "loss": 2.464490509033203,
      "memory(GiB)": 77.56,
      "step": 26365,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.441374
    },
    {
      "epoch": 1.1297716464590206,
      "grad_norm": 4.500833034515381,
      "learning_rate": 8.792277158890185e-05,
      "loss": 2.649382209777832,
      "memory(GiB)": 77.56,
      "step": 26370,
      "token_acc": 0.46504559270516715,
      "train_speed(iter/s)": 1.441346
    },
    {
      "epoch": 1.1299858617882697,
      "grad_norm": 7.4974236488342285,
      "learning_rate": 8.791838528468031e-05,
      "loss": 2.7042991638183596,
      "memory(GiB)": 77.56,
      "step": 26375,
      "token_acc": 0.45185185185185184,
      "train_speed(iter/s)": 1.441411
    },
    {
      "epoch": 1.1302000771175185,
      "grad_norm": 4.057533264160156,
      "learning_rate": 8.791399829353166e-05,
      "loss": 2.5022491455078124,
      "memory(GiB)": 77.56,
      "step": 26380,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.441476
    },
    {
      "epoch": 1.1304142924467675,
      "grad_norm": 5.182284355163574,
      "learning_rate": 8.790961061553535e-05,
      "loss": 2.775295639038086,
      "memory(GiB)": 77.56,
      "step": 26385,
      "token_acc": 0.4380664652567976,
      "train_speed(iter/s)": 1.44142
    },
    {
      "epoch": 1.1306285077760165,
      "grad_norm": 5.530815124511719,
      "learning_rate": 8.79052222507709e-05,
      "loss": 2.4753040313720702,
      "memory(GiB)": 77.56,
      "step": 26390,
      "token_acc": 0.4868913857677903,
      "train_speed(iter/s)": 1.441379
    },
    {
      "epoch": 1.1308427231052653,
      "grad_norm": 4.547403335571289,
      "learning_rate": 8.790083319931776e-05,
      "loss": 2.587689208984375,
      "memory(GiB)": 77.56,
      "step": 26395,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.44144
    },
    {
      "epoch": 1.1310569384345144,
      "grad_norm": 5.189222812652588,
      "learning_rate": 8.78964434612555e-05,
      "loss": 2.399053192138672,
      "memory(GiB)": 77.56,
      "step": 26400,
      "token_acc": 0.49466192170818507,
      "train_speed(iter/s)": 1.441443
    },
    {
      "epoch": 1.1312711537637634,
      "grad_norm": 4.877856254577637,
      "learning_rate": 8.789205303666361e-05,
      "loss": 2.452896308898926,
      "memory(GiB)": 77.56,
      "step": 26405,
      "token_acc": 0.4789272030651341,
      "train_speed(iter/s)": 1.441458
    },
    {
      "epoch": 1.1314853690930122,
      "grad_norm": 4.806341648101807,
      "learning_rate": 8.788766192562162e-05,
      "loss": 2.4968421936035154,
      "memory(GiB)": 77.56,
      "step": 26410,
      "token_acc": 0.46683673469387754,
      "train_speed(iter/s)": 1.441537
    },
    {
      "epoch": 1.1316995844222613,
      "grad_norm": 4.129325866699219,
      "learning_rate": 8.78832701282091e-05,
      "loss": 2.187816047668457,
      "memory(GiB)": 77.56,
      "step": 26415,
      "token_acc": 0.5035211267605634,
      "train_speed(iter/s)": 1.441527
    },
    {
      "epoch": 1.1319137997515103,
      "grad_norm": 5.51617431640625,
      "learning_rate": 8.787887764450561e-05,
      "loss": 2.3775583267211915,
      "memory(GiB)": 77.56,
      "step": 26420,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.441477
    },
    {
      "epoch": 1.132128015080759,
      "grad_norm": 5.2360968589782715,
      "learning_rate": 8.787448447459073e-05,
      "loss": 2.3888309478759764,
      "memory(GiB)": 77.56,
      "step": 26425,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.441508
    },
    {
      "epoch": 1.1323422304100081,
      "grad_norm": 4.695786476135254,
      "learning_rate": 8.787009061854401e-05,
      "loss": 2.682851791381836,
      "memory(GiB)": 77.56,
      "step": 26430,
      "token_acc": 0.42124542124542125,
      "train_speed(iter/s)": 1.441527
    },
    {
      "epoch": 1.1325564457392572,
      "grad_norm": 5.986169815063477,
      "learning_rate": 8.786569607644509e-05,
      "loss": 2.5935699462890627,
      "memory(GiB)": 77.56,
      "step": 26435,
      "token_acc": 0.48184818481848185,
      "train_speed(iter/s)": 1.441575
    },
    {
      "epoch": 1.132770661068506,
      "grad_norm": 6.511702060699463,
      "learning_rate": 8.786130084837355e-05,
      "loss": 2.5218067169189453,
      "memory(GiB)": 77.56,
      "step": 26440,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.441524
    },
    {
      "epoch": 1.132984876397755,
      "grad_norm": 4.406321048736572,
      "learning_rate": 8.785690493440904e-05,
      "loss": 2.4994617462158204,
      "memory(GiB)": 77.56,
      "step": 26445,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.441589
    },
    {
      "epoch": 1.133199091727004,
      "grad_norm": 4.7519001960754395,
      "learning_rate": 8.78525083346312e-05,
      "loss": 2.6419448852539062,
      "memory(GiB)": 77.56,
      "step": 26450,
      "token_acc": 0.4415204678362573,
      "train_speed(iter/s)": 1.441641
    },
    {
      "epoch": 1.1334133070562529,
      "grad_norm": 4.948915481567383,
      "learning_rate": 8.784811104911962e-05,
      "loss": 2.6675222396850584,
      "memory(GiB)": 77.56,
      "step": 26455,
      "token_acc": 0.4592833876221498,
      "train_speed(iter/s)": 1.441623
    },
    {
      "epoch": 1.133627522385502,
      "grad_norm": 6.53285026550293,
      "learning_rate": 8.784371307795403e-05,
      "loss": 2.504957580566406,
      "memory(GiB)": 77.56,
      "step": 26460,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.44161
    },
    {
      "epoch": 1.133841737714751,
      "grad_norm": 3.6881697177886963,
      "learning_rate": 8.783931442121409e-05,
      "loss": 2.490731430053711,
      "memory(GiB)": 77.56,
      "step": 26465,
      "token_acc": 0.48026315789473684,
      "train_speed(iter/s)": 1.441681
    },
    {
      "epoch": 1.1340559530439998,
      "grad_norm": 4.588105201721191,
      "learning_rate": 8.783491507897946e-05,
      "loss": 2.483615493774414,
      "memory(GiB)": 77.56,
      "step": 26470,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.441687
    },
    {
      "epoch": 1.1342701683732488,
      "grad_norm": 5.5090155601501465,
      "learning_rate": 8.783051505132984e-05,
      "loss": 2.467701721191406,
      "memory(GiB)": 77.56,
      "step": 26475,
      "token_acc": 0.4521072796934866,
      "train_speed(iter/s)": 1.441681
    },
    {
      "epoch": 1.1344843837024978,
      "grad_norm": 8.281801223754883,
      "learning_rate": 8.782611433834497e-05,
      "loss": 2.4860801696777344,
      "memory(GiB)": 77.56,
      "step": 26480,
      "token_acc": 0.47950819672131145,
      "train_speed(iter/s)": 1.441723
    },
    {
      "epoch": 1.1346985990317466,
      "grad_norm": 4.080926895141602,
      "learning_rate": 8.782171294010455e-05,
      "loss": 2.6461761474609373,
      "memory(GiB)": 77.56,
      "step": 26485,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.441722
    },
    {
      "epoch": 1.1349128143609957,
      "grad_norm": 4.6654510498046875,
      "learning_rate": 8.781731085668831e-05,
      "loss": 2.4742658615112303,
      "memory(GiB)": 77.56,
      "step": 26490,
      "token_acc": 0.47854785478547857,
      "train_speed(iter/s)": 1.441659
    },
    {
      "epoch": 1.1351270296902447,
      "grad_norm": 4.042934417724609,
      "learning_rate": 8.781290808817602e-05,
      "loss": 2.532063293457031,
      "memory(GiB)": 77.56,
      "step": 26495,
      "token_acc": 0.47076023391812866,
      "train_speed(iter/s)": 1.441712
    },
    {
      "epoch": 1.1353412450194935,
      "grad_norm": 5.348151683807373,
      "learning_rate": 8.780850463464742e-05,
      "loss": 2.4239437103271486,
      "memory(GiB)": 77.56,
      "step": 26500,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.441705
    },
    {
      "epoch": 1.1353412450194935,
      "eval_loss": 2.2603161334991455,
      "eval_runtime": 15.3761,
      "eval_samples_per_second": 6.504,
      "eval_steps_per_second": 6.504,
      "eval_token_acc": 0.4730077120822622,
      "step": 26500
    },
    {
      "epoch": 1.1355554603487426,
      "grad_norm": 3.867032527923584,
      "learning_rate": 8.78041004961823e-05,
      "loss": 2.438082695007324,
      "memory(GiB)": 77.56,
      "step": 26505,
      "token_acc": 0.47074468085106386,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.1357696756779916,
      "grad_norm": 5.7271013259887695,
      "learning_rate": 8.779969567286044e-05,
      "loss": 2.486477279663086,
      "memory(GiB)": 77.56,
      "step": 26510,
      "token_acc": 0.4784172661870504,
      "train_speed(iter/s)": 1.440518
    },
    {
      "epoch": 1.1359838910072404,
      "grad_norm": 3.9770939350128174,
      "learning_rate": 8.779529016476163e-05,
      "loss": 2.4999099731445313,
      "memory(GiB)": 77.56,
      "step": 26515,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.440579
    },
    {
      "epoch": 1.1361981063364894,
      "grad_norm": 4.614752769470215,
      "learning_rate": 8.779088397196567e-05,
      "loss": 2.5234304428100587,
      "memory(GiB)": 77.56,
      "step": 26520,
      "token_acc": 0.4326241134751773,
      "train_speed(iter/s)": 1.440643
    },
    {
      "epoch": 1.1364123216657385,
      "grad_norm": 4.708162307739258,
      "learning_rate": 8.778647709455241e-05,
      "loss": 2.6936302185058594,
      "memory(GiB)": 77.56,
      "step": 26525,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.440646
    },
    {
      "epoch": 1.1366265369949873,
      "grad_norm": 6.140153884887695,
      "learning_rate": 8.778206953260166e-05,
      "loss": 2.4234237670898438,
      "memory(GiB)": 77.56,
      "step": 26530,
      "token_acc": 0.4713375796178344,
      "train_speed(iter/s)": 1.440624
    },
    {
      "epoch": 1.1368407523242363,
      "grad_norm": 4.023498058319092,
      "learning_rate": 8.77776612861933e-05,
      "loss": 2.383848190307617,
      "memory(GiB)": 77.56,
      "step": 26535,
      "token_acc": 0.4715909090909091,
      "train_speed(iter/s)": 1.440651
    },
    {
      "epoch": 1.1370549676534853,
      "grad_norm": 4.258756160736084,
      "learning_rate": 8.777325235540715e-05,
      "loss": 2.7052000045776365,
      "memory(GiB)": 77.56,
      "step": 26540,
      "token_acc": 0.45141065830721006,
      "train_speed(iter/s)": 1.440637
    },
    {
      "epoch": 1.1372691829827342,
      "grad_norm": 4.366440296173096,
      "learning_rate": 8.776884274032309e-05,
      "loss": 2.685236930847168,
      "memory(GiB)": 77.56,
      "step": 26545,
      "token_acc": 0.42900302114803623,
      "train_speed(iter/s)": 1.440702
    },
    {
      "epoch": 1.1374833983119832,
      "grad_norm": 5.327270984649658,
      "learning_rate": 8.776443244102103e-05,
      "loss": 2.7595027923583983,
      "memory(GiB)": 77.56,
      "step": 26550,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.440691
    },
    {
      "epoch": 1.1376976136412322,
      "grad_norm": 4.49923038482666,
      "learning_rate": 8.776002145758084e-05,
      "loss": 2.661919593811035,
      "memory(GiB)": 77.56,
      "step": 26555,
      "token_acc": 0.4557377049180328,
      "train_speed(iter/s)": 1.44074
    },
    {
      "epoch": 1.137911828970481,
      "grad_norm": 3.9664995670318604,
      "learning_rate": 8.775560979008244e-05,
      "loss": 2.1485109329223633,
      "memory(GiB)": 77.56,
      "step": 26560,
      "token_acc": 0.5222672064777328,
      "train_speed(iter/s)": 1.440795
    },
    {
      "epoch": 1.13812604429973,
      "grad_norm": 5.120239734649658,
      "learning_rate": 8.775119743860576e-05,
      "loss": 2.3520294189453126,
      "memory(GiB)": 77.56,
      "step": 26565,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.440864
    },
    {
      "epoch": 1.1383402596289791,
      "grad_norm": 5.736823081970215,
      "learning_rate": 8.774678440323072e-05,
      "loss": 2.43723201751709,
      "memory(GiB)": 77.56,
      "step": 26570,
      "token_acc": 0.43724696356275305,
      "train_speed(iter/s)": 1.440864
    },
    {
      "epoch": 1.138554474958228,
      "grad_norm": 5.541073322296143,
      "learning_rate": 8.774237068403727e-05,
      "loss": 2.471674919128418,
      "memory(GiB)": 77.56,
      "step": 26575,
      "token_acc": 0.47186147186147187,
      "train_speed(iter/s)": 1.440854
    },
    {
      "epoch": 1.138768690287477,
      "grad_norm": 5.638416290283203,
      "learning_rate": 8.773795628110539e-05,
      "loss": 2.578312110900879,
      "memory(GiB)": 77.56,
      "step": 26580,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.440763
    },
    {
      "epoch": 1.138982905616726,
      "grad_norm": 4.208754062652588,
      "learning_rate": 8.7733541194515e-05,
      "loss": 2.5470470428466796,
      "memory(GiB)": 77.56,
      "step": 26585,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.440766
    },
    {
      "epoch": 1.1391971209459748,
      "grad_norm": 3.9446842670440674,
      "learning_rate": 8.772912542434613e-05,
      "loss": 2.3783805847167967,
      "memory(GiB)": 77.56,
      "step": 26590,
      "token_acc": 0.48639455782312924,
      "train_speed(iter/s)": 1.440763
    },
    {
      "epoch": 1.1394113362752238,
      "grad_norm": 4.9133782386779785,
      "learning_rate": 8.772470897067877e-05,
      "loss": 2.527267646789551,
      "memory(GiB)": 77.56,
      "step": 26595,
      "token_acc": 0.4735099337748344,
      "train_speed(iter/s)": 1.440762
    },
    {
      "epoch": 1.1396255516044729,
      "grad_norm": 4.864068031311035,
      "learning_rate": 8.772029183359289e-05,
      "loss": 2.545042610168457,
      "memory(GiB)": 77.56,
      "step": 26600,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.440747
    },
    {
      "epoch": 1.1398397669337217,
      "grad_norm": 5.588677406311035,
      "learning_rate": 8.771587401316856e-05,
      "loss": 2.5636566162109373,
      "memory(GiB)": 77.56,
      "step": 26605,
      "token_acc": 0.5110294117647058,
      "train_speed(iter/s)": 1.440728
    },
    {
      "epoch": 1.1400539822629707,
      "grad_norm": 5.388134956359863,
      "learning_rate": 8.771145550948579e-05,
      "loss": 2.35457763671875,
      "memory(GiB)": 77.56,
      "step": 26610,
      "token_acc": 0.5130718954248366,
      "train_speed(iter/s)": 1.440672
    },
    {
      "epoch": 1.1402681975922198,
      "grad_norm": 6.36678409576416,
      "learning_rate": 8.770703632262463e-05,
      "loss": 2.4657562255859373,
      "memory(GiB)": 77.56,
      "step": 26615,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.440678
    },
    {
      "epoch": 1.1404824129214686,
      "grad_norm": 4.061175346374512,
      "learning_rate": 8.770261645266513e-05,
      "loss": 2.447623062133789,
      "memory(GiB)": 77.56,
      "step": 26620,
      "token_acc": 0.48299319727891155,
      "train_speed(iter/s)": 1.440755
    },
    {
      "epoch": 1.1406966282507176,
      "grad_norm": 5.193114280700684,
      "learning_rate": 8.769819589968737e-05,
      "loss": 2.6491151809692384,
      "memory(GiB)": 77.56,
      "step": 26625,
      "token_acc": 0.45695364238410596,
      "train_speed(iter/s)": 1.440811
    },
    {
      "epoch": 1.1409108435799666,
      "grad_norm": 5.166196823120117,
      "learning_rate": 8.76937746637714e-05,
      "loss": 2.4386377334594727,
      "memory(GiB)": 77.56,
      "step": 26630,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.440814
    },
    {
      "epoch": 1.1411250589092155,
      "grad_norm": 4.047293186187744,
      "learning_rate": 8.768935274499738e-05,
      "loss": 2.8470130920410157,
      "memory(GiB)": 77.56,
      "step": 26635,
      "token_acc": 0.4230769230769231,
      "train_speed(iter/s)": 1.44085
    },
    {
      "epoch": 1.1413392742384645,
      "grad_norm": 5.059890270233154,
      "learning_rate": 8.768493014344536e-05,
      "loss": 2.4384864807128905,
      "memory(GiB)": 77.56,
      "step": 26640,
      "token_acc": 0.45514950166112955,
      "train_speed(iter/s)": 1.440834
    },
    {
      "epoch": 1.1415534895677135,
      "grad_norm": 6.320841312408447,
      "learning_rate": 8.768050685919548e-05,
      "loss": 2.3440835952758787,
      "memory(GiB)": 77.56,
      "step": 26645,
      "token_acc": 0.5259515570934256,
      "train_speed(iter/s)": 1.44085
    },
    {
      "epoch": 1.1417677048969623,
      "grad_norm": 4.210465908050537,
      "learning_rate": 8.767608289232787e-05,
      "loss": 2.3900732040405273,
      "memory(GiB)": 77.56,
      "step": 26650,
      "token_acc": 0.4907749077490775,
      "train_speed(iter/s)": 1.440875
    },
    {
      "epoch": 1.1419819202262114,
      "grad_norm": 5.062823295593262,
      "learning_rate": 8.767165824292269e-05,
      "loss": 2.9886711120605467,
      "memory(GiB)": 77.56,
      "step": 26655,
      "token_acc": 0.4206128133704735,
      "train_speed(iter/s)": 1.440859
    },
    {
      "epoch": 1.1421961355554604,
      "grad_norm": 5.273383617401123,
      "learning_rate": 8.766723291106007e-05,
      "loss": 2.667852783203125,
      "memory(GiB)": 77.56,
      "step": 26660,
      "token_acc": 0.43521594684385384,
      "train_speed(iter/s)": 1.440822
    },
    {
      "epoch": 1.1424103508847092,
      "grad_norm": 6.049825191497803,
      "learning_rate": 8.76628068968202e-05,
      "loss": 2.532472038269043,
      "memory(GiB)": 77.56,
      "step": 26665,
      "token_acc": 0.46923076923076923,
      "train_speed(iter/s)": 1.440897
    },
    {
      "epoch": 1.1426245662139582,
      "grad_norm": 5.55664587020874,
      "learning_rate": 8.765838020028322e-05,
      "loss": 2.225531578063965,
      "memory(GiB)": 77.56,
      "step": 26670,
      "token_acc": 0.5362903225806451,
      "train_speed(iter/s)": 1.440905
    },
    {
      "epoch": 1.1428387815432073,
      "grad_norm": 4.696475982666016,
      "learning_rate": 8.76539528215294e-05,
      "loss": 2.5136844635009767,
      "memory(GiB)": 77.56,
      "step": 26675,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.440913
    },
    {
      "epoch": 1.143052996872456,
      "grad_norm": 3.7514920234680176,
      "learning_rate": 8.764952476063887e-05,
      "loss": 2.6212604522705076,
      "memory(GiB)": 77.56,
      "step": 26680,
      "token_acc": 0.49201277955271566,
      "train_speed(iter/s)": 1.440954
    },
    {
      "epoch": 1.1432672122017051,
      "grad_norm": 4.394812107086182,
      "learning_rate": 8.76450960176919e-05,
      "loss": 2.597858428955078,
      "memory(GiB)": 77.56,
      "step": 26685,
      "token_acc": 0.42105263157894735,
      "train_speed(iter/s)": 1.440996
    },
    {
      "epoch": 1.1434814275309542,
      "grad_norm": 5.044491291046143,
      "learning_rate": 8.76406665927687e-05,
      "loss": 2.298162078857422,
      "memory(GiB)": 77.56,
      "step": 26690,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.44101
    },
    {
      "epoch": 1.143695642860203,
      "grad_norm": 5.354703903198242,
      "learning_rate": 8.763623648594952e-05,
      "loss": 2.4597751617431642,
      "memory(GiB)": 77.56,
      "step": 26695,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.440995
    },
    {
      "epoch": 1.143909858189452,
      "grad_norm": 4.959939479827881,
      "learning_rate": 8.76318056973146e-05,
      "loss": 2.407014083862305,
      "memory(GiB)": 77.56,
      "step": 26700,
      "token_acc": 0.4752851711026616,
      "train_speed(iter/s)": 1.44091
    },
    {
      "epoch": 1.144124073518701,
      "grad_norm": 5.766183376312256,
      "learning_rate": 8.762737422694423e-05,
      "loss": 2.5530813217163084,
      "memory(GiB)": 77.56,
      "step": 26705,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.440918
    },
    {
      "epoch": 1.1443382888479499,
      "grad_norm": 4.712148666381836,
      "learning_rate": 8.762294207491865e-05,
      "loss": 2.7044399261474608,
      "memory(GiB)": 77.56,
      "step": 26710,
      "token_acc": 0.4506578947368421,
      "train_speed(iter/s)": 1.440902
    },
    {
      "epoch": 1.144552504177199,
      "grad_norm": 5.469447135925293,
      "learning_rate": 8.761850924131821e-05,
      "loss": 2.260976028442383,
      "memory(GiB)": 77.56,
      "step": 26715,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.440906
    },
    {
      "epoch": 1.144766719506448,
      "grad_norm": 5.0596747398376465,
      "learning_rate": 8.761407572622318e-05,
      "loss": 2.7360233306884765,
      "memory(GiB)": 77.56,
      "step": 26720,
      "token_acc": 0.431438127090301,
      "train_speed(iter/s)": 1.440932
    },
    {
      "epoch": 1.1449809348356967,
      "grad_norm": 5.104813098907471,
      "learning_rate": 8.760964152971387e-05,
      "loss": 2.345060348510742,
      "memory(GiB)": 77.56,
      "step": 26725,
      "token_acc": 0.5725806451612904,
      "train_speed(iter/s)": 1.440908
    },
    {
      "epoch": 1.1451951501649458,
      "grad_norm": 3.7433674335479736,
      "learning_rate": 8.760520665187064e-05,
      "loss": 2.3104143142700195,
      "memory(GiB)": 77.56,
      "step": 26730,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.440952
    },
    {
      "epoch": 1.1454093654941948,
      "grad_norm": 4.41854190826416,
      "learning_rate": 8.760077109277381e-05,
      "loss": 2.550655174255371,
      "memory(GiB)": 77.56,
      "step": 26735,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.440994
    },
    {
      "epoch": 1.1456235808234436,
      "grad_norm": 3.809072494506836,
      "learning_rate": 8.759633485250372e-05,
      "loss": 2.51839656829834,
      "memory(GiB)": 77.56,
      "step": 26740,
      "token_acc": 0.47634069400630913,
      "train_speed(iter/s)": 1.44105
    },
    {
      "epoch": 1.1458377961526927,
      "grad_norm": 4.477509021759033,
      "learning_rate": 8.759189793114077e-05,
      "loss": 2.411824417114258,
      "memory(GiB)": 77.56,
      "step": 26745,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.441096
    },
    {
      "epoch": 1.1460520114819417,
      "grad_norm": 4.557850360870361,
      "learning_rate": 8.758746032876534e-05,
      "loss": 2.428873062133789,
      "memory(GiB)": 77.56,
      "step": 26750,
      "token_acc": 0.4652567975830816,
      "train_speed(iter/s)": 1.441131
    },
    {
      "epoch": 1.1462662268111905,
      "grad_norm": 5.25700044631958,
      "learning_rate": 8.758302204545779e-05,
      "loss": 2.540868949890137,
      "memory(GiB)": 77.56,
      "step": 26755,
      "token_acc": 0.4980392156862745,
      "train_speed(iter/s)": 1.44118
    },
    {
      "epoch": 1.1464804421404395,
      "grad_norm": 4.926212310791016,
      "learning_rate": 8.757858308129854e-05,
      "loss": 2.358903503417969,
      "memory(GiB)": 77.56,
      "step": 26760,
      "token_acc": 0.5362318840579711,
      "train_speed(iter/s)": 1.4412
    },
    {
      "epoch": 1.1466946574696886,
      "grad_norm": 3.690906286239624,
      "learning_rate": 8.7574143436368e-05,
      "loss": 2.4899175643920897,
      "memory(GiB)": 77.56,
      "step": 26765,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.441261
    },
    {
      "epoch": 1.1469088727989374,
      "grad_norm": 4.500529766082764,
      "learning_rate": 8.756970311074663e-05,
      "loss": 2.657613754272461,
      "memory(GiB)": 77.56,
      "step": 26770,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.441226
    },
    {
      "epoch": 1.1471230881281864,
      "grad_norm": 5.003650665283203,
      "learning_rate": 8.756526210451483e-05,
      "loss": 2.2685012817382812,
      "memory(GiB)": 77.56,
      "step": 26775,
      "token_acc": 0.4908424908424908,
      "train_speed(iter/s)": 1.44121
    },
    {
      "epoch": 1.1473373034574355,
      "grad_norm": 6.734185695648193,
      "learning_rate": 8.756082041775307e-05,
      "loss": 2.4928712844848633,
      "memory(GiB)": 77.56,
      "step": 26780,
      "token_acc": 0.484472049689441,
      "train_speed(iter/s)": 1.441235
    },
    {
      "epoch": 1.1475515187866843,
      "grad_norm": 4.195174217224121,
      "learning_rate": 8.75563780505418e-05,
      "loss": 2.3810924530029296,
      "memory(GiB)": 77.56,
      "step": 26785,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.441271
    },
    {
      "epoch": 1.1477657341159333,
      "grad_norm": 5.03382682800293,
      "learning_rate": 8.755193500296152e-05,
      "loss": 2.7226097106933596,
      "memory(GiB)": 77.56,
      "step": 26790,
      "token_acc": 0.4695121951219512,
      "train_speed(iter/s)": 1.441326
    },
    {
      "epoch": 1.1479799494451823,
      "grad_norm": 6.038765907287598,
      "learning_rate": 8.75474912750927e-05,
      "loss": 2.545185089111328,
      "memory(GiB)": 77.56,
      "step": 26795,
      "token_acc": 0.45482866043613707,
      "train_speed(iter/s)": 1.44134
    },
    {
      "epoch": 1.1481941647744311,
      "grad_norm": 4.757031440734863,
      "learning_rate": 8.754304686701588e-05,
      "loss": 2.3836559295654296,
      "memory(GiB)": 77.56,
      "step": 26800,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.441339
    },
    {
      "epoch": 1.1484083801036802,
      "grad_norm": 4.1283416748046875,
      "learning_rate": 8.753860177881153e-05,
      "loss": 2.3828325271606445,
      "memory(GiB)": 77.56,
      "step": 26805,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.441269
    },
    {
      "epoch": 1.1486225954329292,
      "grad_norm": 5.796810150146484,
      "learning_rate": 8.753415601056019e-05,
      "loss": 2.584481620788574,
      "memory(GiB)": 77.56,
      "step": 26810,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.441324
    },
    {
      "epoch": 1.148836810762178,
      "grad_norm": 4.63601016998291,
      "learning_rate": 8.752970956234242e-05,
      "loss": 2.8847875595092773,
      "memory(GiB)": 77.56,
      "step": 26815,
      "token_acc": 0.4427710843373494,
      "train_speed(iter/s)": 1.441333
    },
    {
      "epoch": 1.149051026091427,
      "grad_norm": 4.261287689208984,
      "learning_rate": 8.752526243423874e-05,
      "loss": 2.788716506958008,
      "memory(GiB)": 77.56,
      "step": 26820,
      "token_acc": 0.4097222222222222,
      "train_speed(iter/s)": 1.441364
    },
    {
      "epoch": 1.149265241420676,
      "grad_norm": 4.280457496643066,
      "learning_rate": 8.752081462632973e-05,
      "loss": 2.270029067993164,
      "memory(GiB)": 77.56,
      "step": 26825,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.441393
    },
    {
      "epoch": 1.149479456749925,
      "grad_norm": 4.7293853759765625,
      "learning_rate": 8.751636613869599e-05,
      "loss": 2.5516822814941404,
      "memory(GiB)": 77.56,
      "step": 26830,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.441396
    },
    {
      "epoch": 1.149693672079174,
      "grad_norm": 3.4673898220062256,
      "learning_rate": 8.751191697141805e-05,
      "loss": 2.4915645599365233,
      "memory(GiB)": 77.56,
      "step": 26835,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.44142
    },
    {
      "epoch": 1.149907887408423,
      "grad_norm": 5.59115743637085,
      "learning_rate": 8.750746712457656e-05,
      "loss": 2.4356884002685546,
      "memory(GiB)": 77.56,
      "step": 26840,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.441437
    },
    {
      "epoch": 1.1501221027376718,
      "grad_norm": 4.838622093200684,
      "learning_rate": 8.750301659825213e-05,
      "loss": 2.642098808288574,
      "memory(GiB)": 77.56,
      "step": 26845,
      "token_acc": 0.4222972972972973,
      "train_speed(iter/s)": 1.441431
    },
    {
      "epoch": 1.1503363180669208,
      "grad_norm": 3.916435718536377,
      "learning_rate": 8.749856539252537e-05,
      "loss": 2.4598834991455076,
      "memory(GiB)": 77.56,
      "step": 26850,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.441475
    },
    {
      "epoch": 1.1505505333961699,
      "grad_norm": 7.166783332824707,
      "learning_rate": 8.749411350747691e-05,
      "loss": 2.483937644958496,
      "memory(GiB)": 77.56,
      "step": 26855,
      "token_acc": 0.48338368580060426,
      "train_speed(iter/s)": 1.441487
    },
    {
      "epoch": 1.1507647487254187,
      "grad_norm": 5.853246688842773,
      "learning_rate": 8.74896609431874e-05,
      "loss": 2.5939865112304688,
      "memory(GiB)": 77.56,
      "step": 26860,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.441513
    },
    {
      "epoch": 1.1509789640546677,
      "grad_norm": 5.35648250579834,
      "learning_rate": 8.748520769973754e-05,
      "loss": 2.4845726013183596,
      "memory(GiB)": 77.56,
      "step": 26865,
      "token_acc": 0.45936395759717313,
      "train_speed(iter/s)": 1.441573
    },
    {
      "epoch": 1.1511931793839167,
      "grad_norm": 3.7814955711364746,
      "learning_rate": 8.748075377720796e-05,
      "loss": 2.400625228881836,
      "memory(GiB)": 77.56,
      "step": 26870,
      "token_acc": 0.49823321554770317,
      "train_speed(iter/s)": 1.441652
    },
    {
      "epoch": 1.1514073947131656,
      "grad_norm": 5.260591506958008,
      "learning_rate": 8.747629917567938e-05,
      "loss": 2.222347068786621,
      "memory(GiB)": 77.56,
      "step": 26875,
      "token_acc": 0.543859649122807,
      "train_speed(iter/s)": 1.44167
    },
    {
      "epoch": 1.1516216100424146,
      "grad_norm": 5.286369800567627,
      "learning_rate": 8.747184389523249e-05,
      "loss": 2.5411750793457033,
      "memory(GiB)": 77.56,
      "step": 26880,
      "token_acc": 0.4630225080385852,
      "train_speed(iter/s)": 1.441658
    },
    {
      "epoch": 1.1518358253716636,
      "grad_norm": 4.017187118530273,
      "learning_rate": 8.746738793594798e-05,
      "loss": 2.5333734512329102,
      "memory(GiB)": 77.56,
      "step": 26885,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.441655
    },
    {
      "epoch": 1.1520500407009124,
      "grad_norm": 4.664718151092529,
      "learning_rate": 8.746293129790659e-05,
      "loss": 2.5239028930664062,
      "memory(GiB)": 77.56,
      "step": 26890,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.441725
    },
    {
      "epoch": 1.1522642560301615,
      "grad_norm": 4.8622636795043945,
      "learning_rate": 8.745847398118906e-05,
      "loss": 2.7554922103881836,
      "memory(GiB)": 77.56,
      "step": 26895,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.441782
    },
    {
      "epoch": 1.1524784713594105,
      "grad_norm": 4.44446325302124,
      "learning_rate": 8.745401598587614e-05,
      "loss": 2.4982284545898437,
      "memory(GiB)": 77.56,
      "step": 26900,
      "token_acc": 0.4603658536585366,
      "train_speed(iter/s)": 1.441818
    },
    {
      "epoch": 1.1526926866886593,
      "grad_norm": 4.0920305252075195,
      "learning_rate": 8.744955731204858e-05,
      "loss": 2.5126646041870115,
      "memory(GiB)": 77.56,
      "step": 26905,
      "token_acc": 0.48068669527896996,
      "train_speed(iter/s)": 1.441804
    },
    {
      "epoch": 1.1529069020179084,
      "grad_norm": 4.622930526733398,
      "learning_rate": 8.744509795978716e-05,
      "loss": 2.499393844604492,
      "memory(GiB)": 77.56,
      "step": 26910,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.441817
    },
    {
      "epoch": 1.1531211173471574,
      "grad_norm": 4.551782608032227,
      "learning_rate": 8.744063792917267e-05,
      "loss": 2.312860298156738,
      "memory(GiB)": 77.56,
      "step": 26915,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.441846
    },
    {
      "epoch": 1.1533353326764062,
      "grad_norm": 4.52789306640625,
      "learning_rate": 8.74361772202859e-05,
      "loss": 2.397283363342285,
      "memory(GiB)": 77.56,
      "step": 26920,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.441869
    },
    {
      "epoch": 1.1535495480056552,
      "grad_norm": 5.282530784606934,
      "learning_rate": 8.743171583320765e-05,
      "loss": 2.7640592575073244,
      "memory(GiB)": 77.56,
      "step": 26925,
      "token_acc": 0.4054982817869416,
      "train_speed(iter/s)": 1.441915
    },
    {
      "epoch": 1.1537637633349043,
      "grad_norm": 4.0936174392700195,
      "learning_rate": 8.742725376801877e-05,
      "loss": 2.269746017456055,
      "memory(GiB)": 77.56,
      "step": 26930,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.441859
    },
    {
      "epoch": 1.1539779786641533,
      "grad_norm": 6.713964462280273,
      "learning_rate": 8.742279102480007e-05,
      "loss": 2.378371810913086,
      "memory(GiB)": 77.56,
      "step": 26935,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.441897
    },
    {
      "epoch": 1.1541921939934021,
      "grad_norm": 4.747091770172119,
      "learning_rate": 8.74183276036324e-05,
      "loss": 2.5782567977905275,
      "memory(GiB)": 77.56,
      "step": 26940,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.441926
    },
    {
      "epoch": 1.1544064093226512,
      "grad_norm": 4.940014362335205,
      "learning_rate": 8.741386350459663e-05,
      "loss": 2.727750778198242,
      "memory(GiB)": 77.56,
      "step": 26945,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.441966
    },
    {
      "epoch": 1.1546206246519002,
      "grad_norm": 7.991199970245361,
      "learning_rate": 8.740939872777363e-05,
      "loss": 2.41723575592041,
      "memory(GiB)": 77.56,
      "step": 26950,
      "token_acc": 0.5390070921985816,
      "train_speed(iter/s)": 1.441961
    },
    {
      "epoch": 1.154834839981149,
      "grad_norm": 5.448914527893066,
      "learning_rate": 8.740493327324425e-05,
      "loss": 2.5768009185791017,
      "memory(GiB)": 77.56,
      "step": 26955,
      "token_acc": 0.506896551724138,
      "train_speed(iter/s)": 1.441974
    },
    {
      "epoch": 1.155049055310398,
      "grad_norm": 5.368537902832031,
      "learning_rate": 8.740046714108947e-05,
      "loss": 2.632326126098633,
      "memory(GiB)": 77.56,
      "step": 26960,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.441969
    },
    {
      "epoch": 1.155263270639647,
      "grad_norm": 4.7027716636657715,
      "learning_rate": 8.739600033139011e-05,
      "loss": 2.731697654724121,
      "memory(GiB)": 77.56,
      "step": 26965,
      "token_acc": 0.46319018404907975,
      "train_speed(iter/s)": 1.441958
    },
    {
      "epoch": 1.1554774859688959,
      "grad_norm": 3.4945216178894043,
      "learning_rate": 8.739153284422712e-05,
      "loss": 2.6790306091308596,
      "memory(GiB)": 77.56,
      "step": 26970,
      "token_acc": 0.4338235294117647,
      "train_speed(iter/s)": 1.441967
    },
    {
      "epoch": 1.155691701298145,
      "grad_norm": 5.31782865524292,
      "learning_rate": 8.738706467968145e-05,
      "loss": 2.2991004943847657,
      "memory(GiB)": 77.56,
      "step": 26975,
      "token_acc": 0.5078740157480315,
      "train_speed(iter/s)": 1.442004
    },
    {
      "epoch": 1.155905916627394,
      "grad_norm": 4.354224681854248,
      "learning_rate": 8.738259583783401e-05,
      "loss": 2.6499013900756836,
      "memory(GiB)": 77.56,
      "step": 26980,
      "token_acc": 0.42810457516339867,
      "train_speed(iter/s)": 1.442091
    },
    {
      "epoch": 1.1561201319566428,
      "grad_norm": 4.768760681152344,
      "learning_rate": 8.737812631876581e-05,
      "loss": 2.5658615112304686,
      "memory(GiB)": 77.56,
      "step": 26985,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.442096
    },
    {
      "epoch": 1.1563343472858918,
      "grad_norm": 4.8821330070495605,
      "learning_rate": 8.737365612255777e-05,
      "loss": 2.4449913024902346,
      "memory(GiB)": 77.56,
      "step": 26990,
      "token_acc": 0.4724137931034483,
      "train_speed(iter/s)": 1.442148
    },
    {
      "epoch": 1.1565485626151408,
      "grad_norm": 5.709953308105469,
      "learning_rate": 8.736918524929088e-05,
      "loss": 2.568259048461914,
      "memory(GiB)": 77.56,
      "step": 26995,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.44214
    },
    {
      "epoch": 1.1567627779443896,
      "grad_norm": 6.190929412841797,
      "learning_rate": 8.736471369904616e-05,
      "loss": 2.7937015533447265,
      "memory(GiB)": 77.56,
      "step": 27000,
      "token_acc": 0.4222972972972973,
      "train_speed(iter/s)": 1.442137
    },
    {
      "epoch": 1.1567627779443896,
      "eval_loss": 2.2168946266174316,
      "eval_runtime": 14.7579,
      "eval_samples_per_second": 6.776,
      "eval_steps_per_second": 6.776,
      "eval_token_acc": 0.4817320703653586,
      "step": 27000
    },
    {
      "epoch": 1.1569769932736387,
      "grad_norm": 4.465674877166748,
      "learning_rate": 8.73602414719046e-05,
      "loss": 2.444173240661621,
      "memory(GiB)": 77.56,
      "step": 27005,
      "token_acc": 0.4771622934888241,
      "train_speed(iter/s)": 1.440907
    },
    {
      "epoch": 1.1571912086028877,
      "grad_norm": 4.241729736328125,
      "learning_rate": 8.735576856794722e-05,
      "loss": 2.522038459777832,
      "memory(GiB)": 77.56,
      "step": 27010,
      "token_acc": 0.4769736842105263,
      "train_speed(iter/s)": 1.44094
    },
    {
      "epoch": 1.1574054239321365,
      "grad_norm": 4.941840648651123,
      "learning_rate": 8.735129498725504e-05,
      "loss": 2.612467956542969,
      "memory(GiB)": 77.56,
      "step": 27015,
      "token_acc": 0.46258503401360546,
      "train_speed(iter/s)": 1.440991
    },
    {
      "epoch": 1.1576196392613856,
      "grad_norm": 4.547250270843506,
      "learning_rate": 8.734682072990912e-05,
      "loss": 2.6390592575073244,
      "memory(GiB)": 77.56,
      "step": 27020,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.441022
    },
    {
      "epoch": 1.1578338545906346,
      "grad_norm": 6.028567790985107,
      "learning_rate": 8.734234579599049e-05,
      "loss": 2.4138420104980467,
      "memory(GiB)": 77.56,
      "step": 27025,
      "token_acc": 0.44357976653696496,
      "train_speed(iter/s)": 1.441108
    },
    {
      "epoch": 1.1580480699198834,
      "grad_norm": 4.942901134490967,
      "learning_rate": 8.733787018558026e-05,
      "loss": 2.1670726776123046,
      "memory(GiB)": 77.56,
      "step": 27030,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.441179
    },
    {
      "epoch": 1.1582622852491324,
      "grad_norm": 5.489515781402588,
      "learning_rate": 8.733339389875947e-05,
      "loss": 2.4094064712524412,
      "memory(GiB)": 77.56,
      "step": 27035,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441249
    },
    {
      "epoch": 1.1584765005783815,
      "grad_norm": 5.404365062713623,
      "learning_rate": 8.732891693560926e-05,
      "loss": 2.6728492736816407,
      "memory(GiB)": 77.56,
      "step": 27040,
      "token_acc": 0.48417721518987344,
      "train_speed(iter/s)": 1.441255
    },
    {
      "epoch": 1.1586907159076303,
      "grad_norm": 5.0062174797058105,
      "learning_rate": 8.732443929621067e-05,
      "loss": 2.608126640319824,
      "memory(GiB)": 77.56,
      "step": 27045,
      "token_acc": 0.4553846153846154,
      "train_speed(iter/s)": 1.441235
    },
    {
      "epoch": 1.1589049312368793,
      "grad_norm": 6.288596153259277,
      "learning_rate": 8.731996098064485e-05,
      "loss": 2.3181735992431642,
      "memory(GiB)": 77.56,
      "step": 27050,
      "token_acc": 0.49387755102040815,
      "train_speed(iter/s)": 1.441304
    },
    {
      "epoch": 1.1591191465661284,
      "grad_norm": 6.4204630851745605,
      "learning_rate": 8.731548198899295e-05,
      "loss": 2.454659080505371,
      "memory(GiB)": 77.56,
      "step": 27055,
      "token_acc": 0.4491803278688525,
      "train_speed(iter/s)": 1.441342
    },
    {
      "epoch": 1.1593333618953772,
      "grad_norm": 5.830802917480469,
      "learning_rate": 8.731100232133607e-05,
      "loss": 2.412165069580078,
      "memory(GiB)": 77.56,
      "step": 27060,
      "token_acc": 0.5186567164179104,
      "train_speed(iter/s)": 1.441317
    },
    {
      "epoch": 1.1595475772246262,
      "grad_norm": 4.659091949462891,
      "learning_rate": 8.730652197775538e-05,
      "loss": 2.4410703659057615,
      "memory(GiB)": 77.56,
      "step": 27065,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.441329
    },
    {
      "epoch": 1.1597617925538752,
      "grad_norm": 3.963329553604126,
      "learning_rate": 8.730204095833205e-05,
      "loss": 2.262119674682617,
      "memory(GiB)": 77.56,
      "step": 27070,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.441328
    },
    {
      "epoch": 1.159976007883124,
      "grad_norm": 5.534719944000244,
      "learning_rate": 8.729755926314727e-05,
      "loss": 2.466657829284668,
      "memory(GiB)": 77.56,
      "step": 27075,
      "token_acc": 0.5222222222222223,
      "train_speed(iter/s)": 1.441345
    },
    {
      "epoch": 1.160190223212373,
      "grad_norm": 5.126330852508545,
      "learning_rate": 8.72930768922822e-05,
      "loss": 2.2823165893554687,
      "memory(GiB)": 77.56,
      "step": 27080,
      "token_acc": 0.48297213622291024,
      "train_speed(iter/s)": 1.441375
    },
    {
      "epoch": 1.1604044385416221,
      "grad_norm": 6.151721954345703,
      "learning_rate": 8.728859384581807e-05,
      "loss": 2.558652877807617,
      "memory(GiB)": 77.56,
      "step": 27085,
      "token_acc": 0.48606811145510836,
      "train_speed(iter/s)": 1.441393
    },
    {
      "epoch": 1.160618653870871,
      "grad_norm": 4.088532447814941,
      "learning_rate": 8.728411012383609e-05,
      "loss": 2.3615997314453123,
      "memory(GiB)": 77.56,
      "step": 27090,
      "token_acc": 0.48404255319148937,
      "train_speed(iter/s)": 1.441417
    },
    {
      "epoch": 1.16083286920012,
      "grad_norm": 5.527521133422852,
      "learning_rate": 8.727962572641746e-05,
      "loss": 2.6201107025146486,
      "memory(GiB)": 77.56,
      "step": 27095,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.441427
    },
    {
      "epoch": 1.161047084529369,
      "grad_norm": 7.4720845222473145,
      "learning_rate": 8.727514065364344e-05,
      "loss": 2.6456672668457033,
      "memory(GiB)": 77.56,
      "step": 27100,
      "token_acc": 0.44525547445255476,
      "train_speed(iter/s)": 1.441444
    },
    {
      "epoch": 1.1612612998586178,
      "grad_norm": 5.867828369140625,
      "learning_rate": 8.72706549055953e-05,
      "loss": 2.5866325378417967,
      "memory(GiB)": 77.56,
      "step": 27105,
      "token_acc": 0.4448669201520912,
      "train_speed(iter/s)": 1.44146
    },
    {
      "epoch": 1.1614755151878668,
      "grad_norm": 6.73767614364624,
      "learning_rate": 8.726616848235426e-05,
      "loss": 2.450385856628418,
      "memory(GiB)": 77.56,
      "step": 27110,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.441516
    },
    {
      "epoch": 1.1616897305171159,
      "grad_norm": 6.656970977783203,
      "learning_rate": 8.726168138400163e-05,
      "loss": 2.6038888931274413,
      "memory(GiB)": 77.56,
      "step": 27115,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.441488
    },
    {
      "epoch": 1.1619039458463647,
      "grad_norm": 4.685435771942139,
      "learning_rate": 8.725719361061868e-05,
      "loss": 2.834351921081543,
      "memory(GiB)": 77.56,
      "step": 27120,
      "token_acc": 0.4273255813953488,
      "train_speed(iter/s)": 1.441502
    },
    {
      "epoch": 1.1621181611756137,
      "grad_norm": 5.694653511047363,
      "learning_rate": 8.725270516228674e-05,
      "loss": 3.034100341796875,
      "memory(GiB)": 77.56,
      "step": 27125,
      "token_acc": 0.4266666666666667,
      "train_speed(iter/s)": 1.441477
    },
    {
      "epoch": 1.1623323765048628,
      "grad_norm": 4.075718402862549,
      "learning_rate": 8.724821603908708e-05,
      "loss": 2.802223968505859,
      "memory(GiB)": 77.56,
      "step": 27130,
      "token_acc": 0.4405797101449275,
      "train_speed(iter/s)": 1.441487
    },
    {
      "epoch": 1.1625465918341116,
      "grad_norm": 4.457286834716797,
      "learning_rate": 8.724372624110105e-05,
      "loss": 2.566387176513672,
      "memory(GiB)": 77.56,
      "step": 27135,
      "token_acc": 0.4432624113475177,
      "train_speed(iter/s)": 1.441472
    },
    {
      "epoch": 1.1627608071633606,
      "grad_norm": 3.494842290878296,
      "learning_rate": 8.723923576840997e-05,
      "loss": 2.4439855575561524,
      "memory(GiB)": 77.56,
      "step": 27140,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.441459
    },
    {
      "epoch": 1.1629750224926096,
      "grad_norm": 3.863755226135254,
      "learning_rate": 8.723474462109522e-05,
      "loss": 2.4159778594970702,
      "memory(GiB)": 77.56,
      "step": 27145,
      "token_acc": 0.5149253731343284,
      "train_speed(iter/s)": 1.441412
    },
    {
      "epoch": 1.1631892378218585,
      "grad_norm": 4.90539026260376,
      "learning_rate": 8.723025279923813e-05,
      "loss": 2.7178874969482423,
      "memory(GiB)": 77.56,
      "step": 27150,
      "token_acc": 0.4261168384879725,
      "train_speed(iter/s)": 1.44139
    },
    {
      "epoch": 1.1634034531511075,
      "grad_norm": 8.530177116394043,
      "learning_rate": 8.722576030292009e-05,
      "loss": 2.204135513305664,
      "memory(GiB)": 77.56,
      "step": 27155,
      "token_acc": 0.5381526104417671,
      "train_speed(iter/s)": 1.441405
    },
    {
      "epoch": 1.1636176684803565,
      "grad_norm": 5.1985602378845215,
      "learning_rate": 8.72212671322225e-05,
      "loss": 2.464513969421387,
      "memory(GiB)": 77.56,
      "step": 27160,
      "token_acc": 0.4847328244274809,
      "train_speed(iter/s)": 1.441423
    },
    {
      "epoch": 1.1638318838096053,
      "grad_norm": 4.844686031341553,
      "learning_rate": 8.72167732872267e-05,
      "loss": 2.430177688598633,
      "memory(GiB)": 77.56,
      "step": 27165,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.441383
    },
    {
      "epoch": 1.1640460991388544,
      "grad_norm": 4.972643852233887,
      "learning_rate": 8.721227876801417e-05,
      "loss": 2.4846261978149413,
      "memory(GiB)": 77.56,
      "step": 27170,
      "token_acc": 0.4536741214057508,
      "train_speed(iter/s)": 1.441342
    },
    {
      "epoch": 1.1642603144681034,
      "grad_norm": 4.513660430908203,
      "learning_rate": 8.72077835746663e-05,
      "loss": 2.504902458190918,
      "memory(GiB)": 77.56,
      "step": 27175,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.441379
    },
    {
      "epoch": 1.1644745297973522,
      "grad_norm": 5.896902084350586,
      "learning_rate": 8.720328770726452e-05,
      "loss": 2.387655258178711,
      "memory(GiB)": 77.56,
      "step": 27180,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.441328
    },
    {
      "epoch": 1.1646887451266013,
      "grad_norm": 4.575392246246338,
      "learning_rate": 8.719879116589028e-05,
      "loss": 2.243789863586426,
      "memory(GiB)": 77.56,
      "step": 27185,
      "token_acc": 0.5214521452145214,
      "train_speed(iter/s)": 1.441397
    },
    {
      "epoch": 1.1649029604558503,
      "grad_norm": 5.380704879760742,
      "learning_rate": 8.719429395062504e-05,
      "loss": 2.9992685317993164,
      "memory(GiB)": 77.56,
      "step": 27190,
      "token_acc": 0.42058823529411765,
      "train_speed(iter/s)": 1.441454
    },
    {
      "epoch": 1.165117175785099,
      "grad_norm": 3.886584758758545,
      "learning_rate": 8.718979606155029e-05,
      "loss": 2.4537332534790037,
      "memory(GiB)": 77.56,
      "step": 27195,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.441474
    },
    {
      "epoch": 1.1653313911143481,
      "grad_norm": 4.117032527923584,
      "learning_rate": 8.718529749874749e-05,
      "loss": 2.2546730041503906,
      "memory(GiB)": 77.56,
      "step": 27200,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.441529
    },
    {
      "epoch": 1.1655456064435972,
      "grad_norm": 6.254148006439209,
      "learning_rate": 8.718079826229813e-05,
      "loss": 2.5495967864990234,
      "memory(GiB)": 77.56,
      "step": 27205,
      "token_acc": 0.5114754098360655,
      "train_speed(iter/s)": 1.441546
    },
    {
      "epoch": 1.165759821772846,
      "grad_norm": 4.410702228546143,
      "learning_rate": 8.717629835228375e-05,
      "loss": 2.3235519409179686,
      "memory(GiB)": 77.56,
      "step": 27210,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.441588
    },
    {
      "epoch": 1.165974037102095,
      "grad_norm": 4.281838893890381,
      "learning_rate": 8.717179776878583e-05,
      "loss": 2.635189247131348,
      "memory(GiB)": 77.56,
      "step": 27215,
      "token_acc": 0.4691358024691358,
      "train_speed(iter/s)": 1.441576
    },
    {
      "epoch": 1.166188252431344,
      "grad_norm": 5.667890548706055,
      "learning_rate": 8.716729651188595e-05,
      "loss": 2.604692077636719,
      "memory(GiB)": 77.56,
      "step": 27220,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.441617
    },
    {
      "epoch": 1.1664024677605929,
      "grad_norm": 4.3095622062683105,
      "learning_rate": 8.71627945816656e-05,
      "loss": 2.71480770111084,
      "memory(GiB)": 77.56,
      "step": 27225,
      "token_acc": 0.4506578947368421,
      "train_speed(iter/s)": 1.44159
    },
    {
      "epoch": 1.166616683089842,
      "grad_norm": 6.5331549644470215,
      "learning_rate": 8.715829197820639e-05,
      "loss": 2.582870101928711,
      "memory(GiB)": 77.56,
      "step": 27230,
      "token_acc": 0.4626334519572954,
      "train_speed(iter/s)": 1.441538
    },
    {
      "epoch": 1.166830898419091,
      "grad_norm": 6.405350208282471,
      "learning_rate": 8.715378870158986e-05,
      "loss": 2.538889694213867,
      "memory(GiB)": 77.56,
      "step": 27235,
      "token_acc": 0.45555555555555555,
      "train_speed(iter/s)": 1.441466
    },
    {
      "epoch": 1.1670451137483397,
      "grad_norm": 4.921253204345703,
      "learning_rate": 8.714928475189759e-05,
      "loss": 2.706508445739746,
      "memory(GiB)": 77.56,
      "step": 27240,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.441489
    },
    {
      "epoch": 1.1672593290775888,
      "grad_norm": 5.2867255210876465,
      "learning_rate": 8.714478012921117e-05,
      "loss": 2.542100524902344,
      "memory(GiB)": 77.56,
      "step": 27245,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.441531
    },
    {
      "epoch": 1.1674735444068378,
      "grad_norm": 6.396088600158691,
      "learning_rate": 8.714027483361222e-05,
      "loss": 2.621960258483887,
      "memory(GiB)": 77.56,
      "step": 27250,
      "token_acc": 0.4778481012658228,
      "train_speed(iter/s)": 1.441606
    },
    {
      "epoch": 1.1676877597360866,
      "grad_norm": 4.936322212219238,
      "learning_rate": 8.713576886518235e-05,
      "loss": 2.33414306640625,
      "memory(GiB)": 77.56,
      "step": 27255,
      "token_acc": 0.5426356589147286,
      "train_speed(iter/s)": 1.441652
    },
    {
      "epoch": 1.1679019750653357,
      "grad_norm": 4.305262565612793,
      "learning_rate": 8.713126222400319e-05,
      "loss": 2.306503677368164,
      "memory(GiB)": 77.56,
      "step": 27260,
      "token_acc": 0.5214285714285715,
      "train_speed(iter/s)": 1.44166
    },
    {
      "epoch": 1.1681161903945847,
      "grad_norm": 5.110226631164551,
      "learning_rate": 8.712675491015638e-05,
      "loss": 2.513997268676758,
      "memory(GiB)": 77.56,
      "step": 27265,
      "token_acc": 0.46200607902735563,
      "train_speed(iter/s)": 1.441597
    },
    {
      "epoch": 1.1683304057238335,
      "grad_norm": 5.511990547180176,
      "learning_rate": 8.712224692372358e-05,
      "loss": 2.761313819885254,
      "memory(GiB)": 77.56,
      "step": 27270,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.441659
    },
    {
      "epoch": 1.1685446210530825,
      "grad_norm": 4.586798191070557,
      "learning_rate": 8.711773826478646e-05,
      "loss": 2.7100988388061524,
      "memory(GiB)": 77.56,
      "step": 27275,
      "token_acc": 0.4337748344370861,
      "train_speed(iter/s)": 1.441719
    },
    {
      "epoch": 1.1687588363823316,
      "grad_norm": 6.718746185302734,
      "learning_rate": 8.711322893342668e-05,
      "loss": 2.5854291915893555,
      "memory(GiB)": 77.56,
      "step": 27280,
      "token_acc": 0.4618320610687023,
      "train_speed(iter/s)": 1.44174
    },
    {
      "epoch": 1.1689730517115804,
      "grad_norm": 4.221371650695801,
      "learning_rate": 8.710871892972595e-05,
      "loss": 2.454153633117676,
      "memory(GiB)": 77.56,
      "step": 27285,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.441755
    },
    {
      "epoch": 1.1691872670408294,
      "grad_norm": 3.7791264057159424,
      "learning_rate": 8.710420825376597e-05,
      "loss": 2.414768409729004,
      "memory(GiB)": 77.56,
      "step": 27290,
      "token_acc": 0.48942598187311176,
      "train_speed(iter/s)": 1.44179
    },
    {
      "epoch": 1.1694014823700785,
      "grad_norm": 4.8100152015686035,
      "learning_rate": 8.709969690562845e-05,
      "loss": 2.9373796463012694,
      "memory(GiB)": 77.56,
      "step": 27295,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.441702
    },
    {
      "epoch": 1.1696156976993273,
      "grad_norm": 6.438279628753662,
      "learning_rate": 8.709518488539511e-05,
      "loss": 2.565231132507324,
      "memory(GiB)": 77.56,
      "step": 27300,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.441756
    },
    {
      "epoch": 1.1698299130285763,
      "grad_norm": 4.581357002258301,
      "learning_rate": 8.709067219314771e-05,
      "loss": 2.4176259994506837,
      "memory(GiB)": 77.56,
      "step": 27305,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.441798
    },
    {
      "epoch": 1.1700441283578253,
      "grad_norm": 4.517197608947754,
      "learning_rate": 8.708615882896798e-05,
      "loss": 2.441977882385254,
      "memory(GiB)": 77.56,
      "step": 27310,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.441801
    },
    {
      "epoch": 1.1702583436870742,
      "grad_norm": 4.801573276519775,
      "learning_rate": 8.708164479293767e-05,
      "loss": 2.4015045166015625,
      "memory(GiB)": 77.56,
      "step": 27315,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.441789
    },
    {
      "epoch": 1.1704725590163232,
      "grad_norm": 5.074005603790283,
      "learning_rate": 8.707713008513863e-05,
      "loss": 2.510340118408203,
      "memory(GiB)": 77.56,
      "step": 27320,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.441739
    },
    {
      "epoch": 1.1706867743455722,
      "grad_norm": 4.657685279846191,
      "learning_rate": 8.707261470565256e-05,
      "loss": 2.5617416381835936,
      "memory(GiB)": 77.56,
      "step": 27325,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.441709
    },
    {
      "epoch": 1.170900989674821,
      "grad_norm": 4.859247207641602,
      "learning_rate": 8.706809865456131e-05,
      "loss": 2.5886001586914062,
      "memory(GiB)": 77.56,
      "step": 27330,
      "token_acc": 0.5209125475285171,
      "train_speed(iter/s)": 1.441639
    },
    {
      "epoch": 1.17111520500407,
      "grad_norm": 4.961772918701172,
      "learning_rate": 8.706358193194668e-05,
      "loss": 2.3527488708496094,
      "memory(GiB)": 77.56,
      "step": 27335,
      "token_acc": 0.4724137931034483,
      "train_speed(iter/s)": 1.441579
    },
    {
      "epoch": 1.171329420333319,
      "grad_norm": 4.1337785720825195,
      "learning_rate": 8.705906453789049e-05,
      "loss": 2.345884323120117,
      "memory(GiB)": 77.56,
      "step": 27340,
      "token_acc": 0.49794238683127573,
      "train_speed(iter/s)": 1.44166
    },
    {
      "epoch": 1.171543635662568,
      "grad_norm": 4.445571422576904,
      "learning_rate": 8.705454647247458e-05,
      "loss": 2.6101245880126953,
      "memory(GiB)": 77.56,
      "step": 27345,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.441689
    },
    {
      "epoch": 1.171757850991817,
      "grad_norm": 4.270284652709961,
      "learning_rate": 8.705002773578081e-05,
      "loss": 2.3811355590820313,
      "memory(GiB)": 77.56,
      "step": 27350,
      "token_acc": 0.5310344827586206,
      "train_speed(iter/s)": 1.441676
    },
    {
      "epoch": 1.171972066321066,
      "grad_norm": 4.850458145141602,
      "learning_rate": 8.704550832789103e-05,
      "loss": 2.973872947692871,
      "memory(GiB)": 77.56,
      "step": 27355,
      "token_acc": 0.4458204334365325,
      "train_speed(iter/s)": 1.441726
    },
    {
      "epoch": 1.172186281650315,
      "grad_norm": 3.9101269245147705,
      "learning_rate": 8.704098824888711e-05,
      "loss": 2.577581596374512,
      "memory(GiB)": 77.56,
      "step": 27360,
      "token_acc": 0.48026315789473684,
      "train_speed(iter/s)": 1.441795
    },
    {
      "epoch": 1.1724004969795638,
      "grad_norm": 5.0655412673950195,
      "learning_rate": 8.703646749885095e-05,
      "loss": 2.536365509033203,
      "memory(GiB)": 77.56,
      "step": 27365,
      "token_acc": 0.46646341463414637,
      "train_speed(iter/s)": 1.441832
    },
    {
      "epoch": 1.1726147123088129,
      "grad_norm": 4.293900012969971,
      "learning_rate": 8.703194607786442e-05,
      "loss": 2.609212303161621,
      "memory(GiB)": 77.56,
      "step": 27370,
      "token_acc": 0.4893048128342246,
      "train_speed(iter/s)": 1.441793
    },
    {
      "epoch": 1.172828927638062,
      "grad_norm": 5.184179306030273,
      "learning_rate": 8.702742398600947e-05,
      "loss": 2.3663558959960938,
      "memory(GiB)": 77.56,
      "step": 27375,
      "token_acc": 0.4774193548387097,
      "train_speed(iter/s)": 1.441807
    },
    {
      "epoch": 1.1730431429673107,
      "grad_norm": 4.691805362701416,
      "learning_rate": 8.702290122336798e-05,
      "loss": 2.697550964355469,
      "memory(GiB)": 77.56,
      "step": 27380,
      "token_acc": 0.42028985507246375,
      "train_speed(iter/s)": 1.441782
    },
    {
      "epoch": 1.1732573582965597,
      "grad_norm": 4.138439178466797,
      "learning_rate": 8.701837779002192e-05,
      "loss": 2.500801849365234,
      "memory(GiB)": 77.56,
      "step": 27385,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.441721
    },
    {
      "epoch": 1.1734715736258088,
      "grad_norm": 4.723711967468262,
      "learning_rate": 8.701385368605322e-05,
      "loss": 2.8100181579589845,
      "memory(GiB)": 77.56,
      "step": 27390,
      "token_acc": 0.4140127388535032,
      "train_speed(iter/s)": 1.441806
    },
    {
      "epoch": 1.1736857889550576,
      "grad_norm": 6.031518936157227,
      "learning_rate": 8.700932891154383e-05,
      "loss": 2.2398391723632813,
      "memory(GiB)": 77.56,
      "step": 27395,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.441884
    },
    {
      "epoch": 1.1739000042843066,
      "grad_norm": 4.962797164916992,
      "learning_rate": 8.700480346657574e-05,
      "loss": 2.5756046295166017,
      "memory(GiB)": 77.56,
      "step": 27400,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.441885
    },
    {
      "epoch": 1.1741142196135557,
      "grad_norm": 5.7035722732543945,
      "learning_rate": 8.70002773512309e-05,
      "loss": 2.5211410522460938,
      "memory(GiB)": 77.56,
      "step": 27405,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.441946
    },
    {
      "epoch": 1.1743284349428045,
      "grad_norm": 4.7708892822265625,
      "learning_rate": 8.699575056559134e-05,
      "loss": 2.3003841400146485,
      "memory(GiB)": 77.56,
      "step": 27410,
      "token_acc": 0.5205992509363296,
      "train_speed(iter/s)": 1.442019
    },
    {
      "epoch": 1.1745426502720535,
      "grad_norm": 5.4351630210876465,
      "learning_rate": 8.699122310973907e-05,
      "loss": 2.4649078369140627,
      "memory(GiB)": 77.56,
      "step": 27415,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.442022
    },
    {
      "epoch": 1.1747568656013025,
      "grad_norm": 5.503630638122559,
      "learning_rate": 8.698669498375606e-05,
      "loss": 2.4473007202148436,
      "memory(GiB)": 77.56,
      "step": 27420,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.442027
    },
    {
      "epoch": 1.1749710809305514,
      "grad_norm": 3.8551297187805176,
      "learning_rate": 8.698216618772439e-05,
      "loss": 2.430198860168457,
      "memory(GiB)": 77.56,
      "step": 27425,
      "token_acc": 0.5067114093959731,
      "train_speed(iter/s)": 1.442065
    },
    {
      "epoch": 1.1751852962598004,
      "grad_norm": 3.7930805683135986,
      "learning_rate": 8.69776367217261e-05,
      "loss": 2.3508865356445314,
      "memory(GiB)": 77.56,
      "step": 27430,
      "token_acc": 0.5230263157894737,
      "train_speed(iter/s)": 1.44212
    },
    {
      "epoch": 1.1753995115890494,
      "grad_norm": 4.585013389587402,
      "learning_rate": 8.697310658584321e-05,
      "loss": 2.5842390060424805,
      "memory(GiB)": 77.56,
      "step": 27435,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.442137
    },
    {
      "epoch": 1.1756137269182982,
      "grad_norm": 4.173397541046143,
      "learning_rate": 8.696857578015783e-05,
      "loss": 3.0209892272949217,
      "memory(GiB)": 77.56,
      "step": 27440,
      "token_acc": 0.4148606811145511,
      "train_speed(iter/s)": 1.442165
    },
    {
      "epoch": 1.1758279422475473,
      "grad_norm": 4.019002437591553,
      "learning_rate": 8.696404430475201e-05,
      "loss": 2.3971990585327148,
      "memory(GiB)": 77.56,
      "step": 27445,
      "token_acc": 0.45045045045045046,
      "train_speed(iter/s)": 1.442192
    },
    {
      "epoch": 1.1760421575767963,
      "grad_norm": 5.549796104431152,
      "learning_rate": 8.695951215970785e-05,
      "loss": 2.759377288818359,
      "memory(GiB)": 77.56,
      "step": 27450,
      "token_acc": 0.42700729927007297,
      "train_speed(iter/s)": 1.442127
    },
    {
      "epoch": 1.1762563729060451,
      "grad_norm": 4.445778846740723,
      "learning_rate": 8.695497934510747e-05,
      "loss": 2.447062301635742,
      "memory(GiB)": 77.56,
      "step": 27455,
      "token_acc": 0.46060606060606063,
      "train_speed(iter/s)": 1.442181
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 8.30522632598877,
      "learning_rate": 8.695044586103296e-05,
      "loss": 2.5784292221069336,
      "memory(GiB)": 77.56,
      "step": 27460,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.442237
    },
    {
      "epoch": 1.1766848035645432,
      "grad_norm": 4.122225284576416,
      "learning_rate": 8.694591170756647e-05,
      "loss": 2.432565689086914,
      "memory(GiB)": 77.56,
      "step": 27465,
      "token_acc": 0.4984520123839009,
      "train_speed(iter/s)": 1.442197
    },
    {
      "epoch": 1.176899018893792,
      "grad_norm": 4.906227111816406,
      "learning_rate": 8.694137688479012e-05,
      "loss": 2.5537179946899413,
      "memory(GiB)": 77.56,
      "step": 27470,
      "token_acc": 0.5049833887043189,
      "train_speed(iter/s)": 1.442297
    },
    {
      "epoch": 1.177113234223041,
      "grad_norm": 4.375225067138672,
      "learning_rate": 8.693684139278609e-05,
      "loss": 2.3931270599365235,
      "memory(GiB)": 77.56,
      "step": 27475,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.442334
    },
    {
      "epoch": 1.17732744955229,
      "grad_norm": 5.097619533538818,
      "learning_rate": 8.693230523163652e-05,
      "loss": 2.782934379577637,
      "memory(GiB)": 77.56,
      "step": 27480,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.442375
    },
    {
      "epoch": 1.1775416648815389,
      "grad_norm": 5.939589500427246,
      "learning_rate": 8.692776840142359e-05,
      "loss": 2.272409439086914,
      "memory(GiB)": 77.56,
      "step": 27485,
      "token_acc": 0.5148936170212766,
      "train_speed(iter/s)": 1.442359
    },
    {
      "epoch": 1.177755880210788,
      "grad_norm": 3.857491970062256,
      "learning_rate": 8.692323090222949e-05,
      "loss": 2.517545700073242,
      "memory(GiB)": 77.56,
      "step": 27490,
      "token_acc": 0.4691689008042895,
      "train_speed(iter/s)": 1.442311
    },
    {
      "epoch": 1.177970095540037,
      "grad_norm": 4.787905693054199,
      "learning_rate": 8.691869273413645e-05,
      "loss": 2.540113830566406,
      "memory(GiB)": 77.56,
      "step": 27495,
      "token_acc": 0.5061728395061729,
      "train_speed(iter/s)": 1.442377
    },
    {
      "epoch": 1.1781843108692858,
      "grad_norm": 4.3217082023620605,
      "learning_rate": 8.691415389722663e-05,
      "loss": 2.4738290786743162,
      "memory(GiB)": 77.56,
      "step": 27500,
      "token_acc": 0.46387832699619774,
      "train_speed(iter/s)": 1.442406
    },
    {
      "epoch": 1.1781843108692858,
      "eval_loss": 2.181292772293091,
      "eval_runtime": 15.071,
      "eval_samples_per_second": 6.635,
      "eval_steps_per_second": 6.635,
      "eval_token_acc": 0.4791386271870794,
      "step": 27500
    },
    {
      "epoch": 1.1783985261985348,
      "grad_norm": 4.888856410980225,
      "learning_rate": 8.690961439158229e-05,
      "loss": 2.4217287063598634,
      "memory(GiB)": 77.56,
      "step": 27505,
      "token_acc": 0.47604790419161674,
      "train_speed(iter/s)": 1.441203
    },
    {
      "epoch": 1.1786127415277838,
      "grad_norm": 6.293729305267334,
      "learning_rate": 8.690507421728566e-05,
      "loss": 2.4635908126831056,
      "memory(GiB)": 77.56,
      "step": 27510,
      "token_acc": 0.5114503816793893,
      "train_speed(iter/s)": 1.441178
    },
    {
      "epoch": 1.1788269568570326,
      "grad_norm": 4.702425956726074,
      "learning_rate": 8.690053337441901e-05,
      "loss": 2.319647026062012,
      "memory(GiB)": 77.56,
      "step": 27515,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.441269
    },
    {
      "epoch": 1.1790411721862817,
      "grad_norm": 4.294443130493164,
      "learning_rate": 8.689599186306457e-05,
      "loss": 2.5872364044189453,
      "memory(GiB)": 77.56,
      "step": 27520,
      "token_acc": 0.48026315789473684,
      "train_speed(iter/s)": 1.441277
    },
    {
      "epoch": 1.1792553875155307,
      "grad_norm": 4.025820255279541,
      "learning_rate": 8.689144968330462e-05,
      "loss": 2.332527732849121,
      "memory(GiB)": 77.56,
      "step": 27525,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.441261
    },
    {
      "epoch": 1.1794696028447795,
      "grad_norm": 6.125864505767822,
      "learning_rate": 8.688690683522147e-05,
      "loss": 2.187384033203125,
      "memory(GiB)": 77.56,
      "step": 27530,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.441303
    },
    {
      "epoch": 1.1796838181740286,
      "grad_norm": 4.251725196838379,
      "learning_rate": 8.688236331889739e-05,
      "loss": 2.6505970001220702,
      "memory(GiB)": 77.56,
      "step": 27535,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.441336
    },
    {
      "epoch": 1.1798980335032776,
      "grad_norm": 5.752416133880615,
      "learning_rate": 8.687781913441469e-05,
      "loss": 2.47400016784668,
      "memory(GiB)": 77.56,
      "step": 27540,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.441299
    },
    {
      "epoch": 1.1801122488325264,
      "grad_norm": 6.580593585968018,
      "learning_rate": 8.687327428185573e-05,
      "loss": 2.7571815490722655,
      "memory(GiB)": 77.56,
      "step": 27545,
      "token_acc": 0.44244604316546765,
      "train_speed(iter/s)": 1.44136
    },
    {
      "epoch": 1.1803264641617754,
      "grad_norm": 8.08622932434082,
      "learning_rate": 8.686872876130279e-05,
      "loss": 2.763301467895508,
      "memory(GiB)": 77.56,
      "step": 27550,
      "token_acc": 0.45564516129032256,
      "train_speed(iter/s)": 1.441431
    },
    {
      "epoch": 1.1805406794910245,
      "grad_norm": 4.863842487335205,
      "learning_rate": 8.686418257283824e-05,
      "loss": 2.249067687988281,
      "memory(GiB)": 77.56,
      "step": 27555,
      "token_acc": 0.5378151260504201,
      "train_speed(iter/s)": 1.441452
    },
    {
      "epoch": 1.1807548948202733,
      "grad_norm": 5.426926612854004,
      "learning_rate": 8.685963571654445e-05,
      "loss": 2.4215042114257814,
      "memory(GiB)": 77.56,
      "step": 27560,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.441415
    },
    {
      "epoch": 1.1809691101495223,
      "grad_norm": 3.8409945964813232,
      "learning_rate": 8.685508819250379e-05,
      "loss": 2.6023340225219727,
      "memory(GiB)": 77.56,
      "step": 27565,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.441432
    },
    {
      "epoch": 1.1811833254787714,
      "grad_norm": 4.093282699584961,
      "learning_rate": 8.685054000079863e-05,
      "loss": 2.6001235961914064,
      "memory(GiB)": 77.56,
      "step": 27570,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.441483
    },
    {
      "epoch": 1.1813975408080202,
      "grad_norm": 5.113375663757324,
      "learning_rate": 8.684599114151137e-05,
      "loss": 2.6917396545410157,
      "memory(GiB)": 77.56,
      "step": 27575,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.441502
    },
    {
      "epoch": 1.1816117561372692,
      "grad_norm": 5.857832908630371,
      "learning_rate": 8.684144161472441e-05,
      "loss": 2.4806318283081055,
      "memory(GiB)": 77.56,
      "step": 27580,
      "token_acc": 0.4461538461538462,
      "train_speed(iter/s)": 1.441462
    },
    {
      "epoch": 1.1818259714665182,
      "grad_norm": 5.172798156738281,
      "learning_rate": 8.683689142052018e-05,
      "loss": 2.2441524505615233,
      "memory(GiB)": 77.56,
      "step": 27585,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.441514
    },
    {
      "epoch": 1.182040186795767,
      "grad_norm": 4.691322326660156,
      "learning_rate": 8.68323405589811e-05,
      "loss": 2.4120204925537108,
      "memory(GiB)": 77.56,
      "step": 27590,
      "token_acc": 0.43686006825938567,
      "train_speed(iter/s)": 1.441536
    },
    {
      "epoch": 1.182254402125016,
      "grad_norm": 5.982290744781494,
      "learning_rate": 8.682778903018962e-05,
      "loss": 2.453257369995117,
      "memory(GiB)": 77.56,
      "step": 27595,
      "token_acc": 0.46099290780141844,
      "train_speed(iter/s)": 1.441457
    },
    {
      "epoch": 1.1824686174542651,
      "grad_norm": 5.5276336669921875,
      "learning_rate": 8.68232368342282e-05,
      "loss": 2.1503772735595703,
      "memory(GiB)": 77.56,
      "step": 27600,
      "token_acc": 0.5096525096525096,
      "train_speed(iter/s)": 1.441457
    },
    {
      "epoch": 1.182682832783514,
      "grad_norm": 4.469607830047607,
      "learning_rate": 8.68186839711793e-05,
      "loss": 2.3769208908081056,
      "memory(GiB)": 77.56,
      "step": 27605,
      "token_acc": 0.5236486486486487,
      "train_speed(iter/s)": 1.441469
    },
    {
      "epoch": 1.182897048112763,
      "grad_norm": 5.754303932189941,
      "learning_rate": 8.68141304411254e-05,
      "loss": 2.381919288635254,
      "memory(GiB)": 77.56,
      "step": 27610,
      "token_acc": 0.48559670781893005,
      "train_speed(iter/s)": 1.441424
    },
    {
      "epoch": 1.183111263442012,
      "grad_norm": 7.353774547576904,
      "learning_rate": 8.680957624414901e-05,
      "loss": 2.832828140258789,
      "memory(GiB)": 77.56,
      "step": 27615,
      "token_acc": 0.45054945054945056,
      "train_speed(iter/s)": 1.441461
    },
    {
      "epoch": 1.1833254787712608,
      "grad_norm": 6.191554069519043,
      "learning_rate": 8.68050213803326e-05,
      "loss": 2.3928375244140625,
      "memory(GiB)": 77.56,
      "step": 27620,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.44149
    },
    {
      "epoch": 1.1835396941005099,
      "grad_norm": 5.659239768981934,
      "learning_rate": 8.68004658497587e-05,
      "loss": 2.41302490234375,
      "memory(GiB)": 77.56,
      "step": 27625,
      "token_acc": 0.49377593360995853,
      "train_speed(iter/s)": 1.44152
    },
    {
      "epoch": 1.1837539094297589,
      "grad_norm": 4.310657501220703,
      "learning_rate": 8.679590965250984e-05,
      "loss": 2.7027875900268556,
      "memory(GiB)": 77.56,
      "step": 27630,
      "token_acc": 0.4665012406947891,
      "train_speed(iter/s)": 1.44146
    },
    {
      "epoch": 1.1839681247590077,
      "grad_norm": 4.37050724029541,
      "learning_rate": 8.679135278866855e-05,
      "loss": 2.389889144897461,
      "memory(GiB)": 77.56,
      "step": 27635,
      "token_acc": 0.48828125,
      "train_speed(iter/s)": 1.441456
    },
    {
      "epoch": 1.1841823400882567,
      "grad_norm": 4.196377277374268,
      "learning_rate": 8.678679525831742e-05,
      "loss": 2.388215255737305,
      "memory(GiB)": 77.56,
      "step": 27640,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.441494
    },
    {
      "epoch": 1.1843965554175058,
      "grad_norm": 5.027113437652588,
      "learning_rate": 8.678223706153895e-05,
      "loss": 2.5079744338989256,
      "memory(GiB)": 77.56,
      "step": 27645,
      "token_acc": 0.5271084337349398,
      "train_speed(iter/s)": 1.441461
    },
    {
      "epoch": 1.1846107707467546,
      "grad_norm": 6.359491348266602,
      "learning_rate": 8.677767819841577e-05,
      "loss": 2.491704750061035,
      "memory(GiB)": 77.56,
      "step": 27650,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.441427
    },
    {
      "epoch": 1.1848249860760036,
      "grad_norm": 4.476621627807617,
      "learning_rate": 8.677311866903044e-05,
      "loss": 2.7572935104370115,
      "memory(GiB)": 77.56,
      "step": 27655,
      "token_acc": 0.43846153846153846,
      "train_speed(iter/s)": 1.441452
    },
    {
      "epoch": 1.1850392014052527,
      "grad_norm": 5.465272903442383,
      "learning_rate": 8.676855847346559e-05,
      "loss": 2.8429534912109373,
      "memory(GiB)": 77.56,
      "step": 27660,
      "token_acc": 0.4275618374558304,
      "train_speed(iter/s)": 1.441426
    },
    {
      "epoch": 1.1852534167345015,
      "grad_norm": 5.176022529602051,
      "learning_rate": 8.676399761180379e-05,
      "loss": 2.4429162979125976,
      "memory(GiB)": 77.56,
      "step": 27665,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.441409
    },
    {
      "epoch": 1.1854676320637505,
      "grad_norm": 4.39204216003418,
      "learning_rate": 8.67594360841277e-05,
      "loss": 2.4336334228515626,
      "memory(GiB)": 77.56,
      "step": 27670,
      "token_acc": 0.4477124183006536,
      "train_speed(iter/s)": 1.44143
    },
    {
      "epoch": 1.1856818473929995,
      "grad_norm": 4.6355133056640625,
      "learning_rate": 8.675487389051993e-05,
      "loss": 2.3863924026489256,
      "memory(GiB)": 77.56,
      "step": 27675,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.441443
    },
    {
      "epoch": 1.1858960627222483,
      "grad_norm": 6.04000186920166,
      "learning_rate": 8.675031103106315e-05,
      "loss": 2.492868423461914,
      "memory(GiB)": 77.56,
      "step": 27680,
      "token_acc": 0.4403292181069959,
      "train_speed(iter/s)": 1.441436
    },
    {
      "epoch": 1.1861102780514974,
      "grad_norm": 6.974379539489746,
      "learning_rate": 8.674574750584002e-05,
      "loss": 2.434065818786621,
      "memory(GiB)": 77.56,
      "step": 27685,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.441517
    },
    {
      "epoch": 1.1863244933807464,
      "grad_norm": 5.399651527404785,
      "learning_rate": 8.674118331493317e-05,
      "loss": 2.4655824661254884,
      "memory(GiB)": 77.56,
      "step": 27690,
      "token_acc": 0.43521594684385384,
      "train_speed(iter/s)": 1.441553
    },
    {
      "epoch": 1.1865387087099952,
      "grad_norm": 4.4834442138671875,
      "learning_rate": 8.673661845842536e-05,
      "loss": 2.3906164169311523,
      "memory(GiB)": 77.56,
      "step": 27695,
      "token_acc": 0.517799352750809,
      "train_speed(iter/s)": 1.441595
    },
    {
      "epoch": 1.1867529240392443,
      "grad_norm": 7.482815742492676,
      "learning_rate": 8.673205293639921e-05,
      "loss": 2.4204029083251952,
      "memory(GiB)": 77.56,
      "step": 27700,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.4416
    },
    {
      "epoch": 1.1869671393684933,
      "grad_norm": 7.446115493774414,
      "learning_rate": 8.672748674893749e-05,
      "loss": 2.209854507446289,
      "memory(GiB)": 77.56,
      "step": 27705,
      "token_acc": 0.5290519877675841,
      "train_speed(iter/s)": 1.441614
    },
    {
      "epoch": 1.187181354697742,
      "grad_norm": 4.794114112854004,
      "learning_rate": 8.672291989612287e-05,
      "loss": 2.4992061614990235,
      "memory(GiB)": 77.56,
      "step": 27710,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.441661
    },
    {
      "epoch": 1.1873955700269911,
      "grad_norm": 10.642241477966309,
      "learning_rate": 8.671835237803812e-05,
      "loss": 2.7294851303100587,
      "memory(GiB)": 77.56,
      "step": 27715,
      "token_acc": 0.4280155642023346,
      "train_speed(iter/s)": 1.441724
    },
    {
      "epoch": 1.1876097853562402,
      "grad_norm": 6.134695529937744,
      "learning_rate": 8.671378419476596e-05,
      "loss": 2.6088140487670897,
      "memory(GiB)": 77.56,
      "step": 27720,
      "token_acc": 0.46449704142011833,
      "train_speed(iter/s)": 1.4417
    },
    {
      "epoch": 1.187824000685489,
      "grad_norm": 6.8227009773254395,
      "learning_rate": 8.670921534638918e-05,
      "loss": 2.5411964416503907,
      "memory(GiB)": 77.56,
      "step": 27725,
      "token_acc": 0.45569620253164556,
      "train_speed(iter/s)": 1.441652
    },
    {
      "epoch": 1.188038216014738,
      "grad_norm": 5.670330047607422,
      "learning_rate": 8.670464583299052e-05,
      "loss": 2.5149412155151367,
      "memory(GiB)": 77.56,
      "step": 27730,
      "token_acc": 0.4530612244897959,
      "train_speed(iter/s)": 1.441711
    },
    {
      "epoch": 1.188252431343987,
      "grad_norm": 8.834687232971191,
      "learning_rate": 8.670007565465277e-05,
      "loss": 2.5030658721923826,
      "memory(GiB)": 77.56,
      "step": 27735,
      "token_acc": 0.4752475247524752,
      "train_speed(iter/s)": 1.441726
    },
    {
      "epoch": 1.1884666466732359,
      "grad_norm": 12.955204963684082,
      "learning_rate": 8.669550481145872e-05,
      "loss": 2.710847282409668,
      "memory(GiB)": 77.56,
      "step": 27740,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.441765
    },
    {
      "epoch": 1.188680862002485,
      "grad_norm": 4.122494220733643,
      "learning_rate": 8.669093330349118e-05,
      "loss": 2.2909990310668946,
      "memory(GiB)": 77.56,
      "step": 27745,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.44183
    },
    {
      "epoch": 1.188895077331734,
      "grad_norm": 4.734692096710205,
      "learning_rate": 8.668636113083296e-05,
      "loss": 2.3412214279174806,
      "memory(GiB)": 77.56,
      "step": 27750,
      "token_acc": 0.5323741007194245,
      "train_speed(iter/s)": 1.441805
    },
    {
      "epoch": 1.1891092926609828,
      "grad_norm": 4.5979132652282715,
      "learning_rate": 8.66817882935669e-05,
      "loss": 2.637179946899414,
      "memory(GiB)": 77.56,
      "step": 27755,
      "token_acc": 0.49216300940438873,
      "train_speed(iter/s)": 1.441782
    },
    {
      "epoch": 1.1893235079902318,
      "grad_norm": 6.253782749176025,
      "learning_rate": 8.667721479177583e-05,
      "loss": 2.268264389038086,
      "memory(GiB)": 77.56,
      "step": 27760,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.441817
    },
    {
      "epoch": 1.1895377233194808,
      "grad_norm": 7.791936874389648,
      "learning_rate": 8.66726406255426e-05,
      "loss": 2.813027191162109,
      "memory(GiB)": 77.56,
      "step": 27765,
      "token_acc": 0.4703703703703704,
      "train_speed(iter/s)": 1.441829
    },
    {
      "epoch": 1.1897519386487296,
      "grad_norm": 6.149044990539551,
      "learning_rate": 8.66680657949501e-05,
      "loss": 2.553750419616699,
      "memory(GiB)": 77.56,
      "step": 27770,
      "token_acc": 0.5139318885448917,
      "train_speed(iter/s)": 1.441846
    },
    {
      "epoch": 1.1899661539779787,
      "grad_norm": 5.280455112457275,
      "learning_rate": 8.666349030008118e-05,
      "loss": 2.625950241088867,
      "memory(GiB)": 77.56,
      "step": 27775,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.441874
    },
    {
      "epoch": 1.1901803693072277,
      "grad_norm": 4.8144049644470215,
      "learning_rate": 8.665891414101874e-05,
      "loss": 2.4960559844970702,
      "memory(GiB)": 77.56,
      "step": 27780,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.441879
    },
    {
      "epoch": 1.1903945846364765,
      "grad_norm": 5.351006031036377,
      "learning_rate": 8.665433731784568e-05,
      "loss": 2.6527963638305665,
      "memory(GiB)": 77.56,
      "step": 27785,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.441898
    },
    {
      "epoch": 1.1906087999657256,
      "grad_norm": 5.2274580001831055,
      "learning_rate": 8.664975983064491e-05,
      "loss": 2.6767696380615233,
      "memory(GiB)": 77.56,
      "step": 27790,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.441898
    },
    {
      "epoch": 1.1908230152949746,
      "grad_norm": 5.817531585693359,
      "learning_rate": 8.664518167949937e-05,
      "loss": 2.6544605255126954,
      "memory(GiB)": 77.56,
      "step": 27795,
      "token_acc": 0.476038338658147,
      "train_speed(iter/s)": 1.441875
    },
    {
      "epoch": 1.1910372306242234,
      "grad_norm": 4.971656322479248,
      "learning_rate": 8.6640602864492e-05,
      "loss": 2.581356430053711,
      "memory(GiB)": 77.56,
      "step": 27800,
      "token_acc": 0.45484949832775917,
      "train_speed(iter/s)": 1.441852
    },
    {
      "epoch": 1.1912514459534724,
      "grad_norm": 5.127297401428223,
      "learning_rate": 8.663602338570571e-05,
      "loss": 2.68997802734375,
      "memory(GiB)": 77.56,
      "step": 27805,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.44186
    },
    {
      "epoch": 1.1914656612827215,
      "grad_norm": 4.589285373687744,
      "learning_rate": 8.663144324322351e-05,
      "loss": 2.1073837280273438,
      "memory(GiB)": 77.56,
      "step": 27810,
      "token_acc": 0.5543859649122806,
      "train_speed(iter/s)": 1.441765
    },
    {
      "epoch": 1.1916798766119703,
      "grad_norm": 5.023262977600098,
      "learning_rate": 8.662686243712834e-05,
      "loss": 2.5463430404663088,
      "memory(GiB)": 77.56,
      "step": 27815,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.441737
    },
    {
      "epoch": 1.1918940919412193,
      "grad_norm": 5.036184310913086,
      "learning_rate": 8.662228096750321e-05,
      "loss": 2.344204139709473,
      "memory(GiB)": 77.56,
      "step": 27820,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.441675
    },
    {
      "epoch": 1.1921083072704683,
      "grad_norm": 5.8417463302612305,
      "learning_rate": 8.661769883443109e-05,
      "loss": 2.637803077697754,
      "memory(GiB)": 77.56,
      "step": 27825,
      "token_acc": 0.4664179104477612,
      "train_speed(iter/s)": 1.44169
    },
    {
      "epoch": 1.1923225225997172,
      "grad_norm": 5.606911659240723,
      "learning_rate": 8.661311603799503e-05,
      "loss": 2.269527626037598,
      "memory(GiB)": 77.56,
      "step": 27830,
      "token_acc": 0.5203252032520326,
      "train_speed(iter/s)": 1.44172
    },
    {
      "epoch": 1.1925367379289662,
      "grad_norm": 4.498233795166016,
      "learning_rate": 8.660853257827801e-05,
      "loss": 2.674235725402832,
      "memory(GiB)": 77.56,
      "step": 27835,
      "token_acc": 0.4236760124610592,
      "train_speed(iter/s)": 1.441748
    },
    {
      "epoch": 1.1927509532582152,
      "grad_norm": 5.204937934875488,
      "learning_rate": 8.660394845536309e-05,
      "loss": 2.1245370864868165,
      "memory(GiB)": 77.56,
      "step": 27840,
      "token_acc": 0.5382059800664452,
      "train_speed(iter/s)": 1.441746
    },
    {
      "epoch": 1.192965168587464,
      "grad_norm": 4.371767520904541,
      "learning_rate": 8.659936366933328e-05,
      "loss": 2.521686553955078,
      "memory(GiB)": 77.56,
      "step": 27845,
      "token_acc": 0.46348314606741575,
      "train_speed(iter/s)": 1.441764
    },
    {
      "epoch": 1.193179383916713,
      "grad_norm": 4.029289245605469,
      "learning_rate": 8.659477822027169e-05,
      "loss": 2.5883317947387696,
      "memory(GiB)": 77.56,
      "step": 27850,
      "token_acc": 0.43465045592705165,
      "train_speed(iter/s)": 1.441781
    },
    {
      "epoch": 1.1933935992459621,
      "grad_norm": 3.722578287124634,
      "learning_rate": 8.659019210826136e-05,
      "loss": 2.4981433868408205,
      "memory(GiB)": 77.56,
      "step": 27855,
      "token_acc": 0.4536082474226804,
      "train_speed(iter/s)": 1.441761
    },
    {
      "epoch": 1.193607814575211,
      "grad_norm": 6.433159351348877,
      "learning_rate": 8.658560533338537e-05,
      "loss": 2.2970319747924806,
      "memory(GiB)": 77.56,
      "step": 27860,
      "token_acc": 0.5139442231075697,
      "train_speed(iter/s)": 1.441797
    },
    {
      "epoch": 1.19382202990446,
      "grad_norm": 4.9400105476379395,
      "learning_rate": 8.658101789572683e-05,
      "loss": 2.425587463378906,
      "memory(GiB)": 77.56,
      "step": 27865,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.441751
    },
    {
      "epoch": 1.194036245233709,
      "grad_norm": 4.408133029937744,
      "learning_rate": 8.657642979536882e-05,
      "loss": 2.490181732177734,
      "memory(GiB)": 77.56,
      "step": 27870,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.441763
    },
    {
      "epoch": 1.1942504605629578,
      "grad_norm": 4.0465617179870605,
      "learning_rate": 8.657184103239448e-05,
      "loss": 2.584272575378418,
      "memory(GiB)": 77.56,
      "step": 27875,
      "token_acc": 0.48089171974522293,
      "train_speed(iter/s)": 1.441789
    },
    {
      "epoch": 1.1944646758922068,
      "grad_norm": 5.069238185882568,
      "learning_rate": 8.656725160688694e-05,
      "loss": 2.3231069564819338,
      "memory(GiB)": 77.56,
      "step": 27880,
      "token_acc": 0.48616600790513836,
      "train_speed(iter/s)": 1.441843
    },
    {
      "epoch": 1.1946788912214559,
      "grad_norm": 4.324473857879639,
      "learning_rate": 8.656266151892932e-05,
      "loss": 2.605357360839844,
      "memory(GiB)": 77.56,
      "step": 27885,
      "token_acc": 0.424812030075188,
      "train_speed(iter/s)": 1.441891
    },
    {
      "epoch": 1.1948931065507047,
      "grad_norm": 3.8083996772766113,
      "learning_rate": 8.655807076860481e-05,
      "loss": 3.0749267578125,
      "memory(GiB)": 77.56,
      "step": 27890,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.441974
    },
    {
      "epoch": 1.1951073218799537,
      "grad_norm": 4.634095191955566,
      "learning_rate": 8.655347935599653e-05,
      "loss": 2.557315444946289,
      "memory(GiB)": 77.56,
      "step": 27895,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.442015
    },
    {
      "epoch": 1.1953215372092028,
      "grad_norm": 3.6966121196746826,
      "learning_rate": 8.654888728118769e-05,
      "loss": 2.479615592956543,
      "memory(GiB)": 77.56,
      "step": 27900,
      "token_acc": 0.5228070175438596,
      "train_speed(iter/s)": 1.442035
    },
    {
      "epoch": 1.1955357525384516,
      "grad_norm": 4.629363536834717,
      "learning_rate": 8.654429454426148e-05,
      "loss": 2.7010009765625,
      "memory(GiB)": 77.56,
      "step": 27905,
      "token_acc": 0.49049429657794674,
      "train_speed(iter/s)": 1.442036
    },
    {
      "epoch": 1.1957499678677006,
      "grad_norm": 4.705044269561768,
      "learning_rate": 8.653970114530107e-05,
      "loss": 2.6380115509033204,
      "memory(GiB)": 77.56,
      "step": 27910,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.442064
    },
    {
      "epoch": 1.1959641831969496,
      "grad_norm": 4.2426910400390625,
      "learning_rate": 8.65351070843897e-05,
      "loss": 2.43743953704834,
      "memory(GiB)": 77.56,
      "step": 27915,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.442043
    },
    {
      "epoch": 1.1961783985261984,
      "grad_norm": 4.5766215324401855,
      "learning_rate": 8.653051236161062e-05,
      "loss": 2.4382553100585938,
      "memory(GiB)": 77.56,
      "step": 27920,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.442009
    },
    {
      "epoch": 1.1963926138554475,
      "grad_norm": 4.8841986656188965,
      "learning_rate": 8.652591697704702e-05,
      "loss": 2.796978759765625,
      "memory(GiB)": 77.56,
      "step": 27925,
      "token_acc": 0.4195583596214511,
      "train_speed(iter/s)": 1.442018
    },
    {
      "epoch": 1.1966068291846965,
      "grad_norm": 5.63884973526001,
      "learning_rate": 8.652132093078217e-05,
      "loss": 2.701239585876465,
      "memory(GiB)": 77.56,
      "step": 27930,
      "token_acc": 0.44876325088339225,
      "train_speed(iter/s)": 1.442007
    },
    {
      "epoch": 1.1968210445139453,
      "grad_norm": 6.249762535095215,
      "learning_rate": 8.651672422289934e-05,
      "loss": 2.2940149307250977,
      "memory(GiB)": 77.56,
      "step": 27935,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.441995
    },
    {
      "epoch": 1.1970352598431944,
      "grad_norm": 4.73747444152832,
      "learning_rate": 8.65121268534818e-05,
      "loss": 2.690804100036621,
      "memory(GiB)": 77.56,
      "step": 27940,
      "token_acc": 0.4342105263157895,
      "train_speed(iter/s)": 1.442011
    },
    {
      "epoch": 1.1972494751724434,
      "grad_norm": 4.564331531524658,
      "learning_rate": 8.650752882261282e-05,
      "loss": 2.754376029968262,
      "memory(GiB)": 77.56,
      "step": 27945,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.442009
    },
    {
      "epoch": 1.1974636905016922,
      "grad_norm": 5.374021053314209,
      "learning_rate": 8.65029301303757e-05,
      "loss": 2.6044281005859373,
      "memory(GiB)": 77.56,
      "step": 27950,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.441993
    },
    {
      "epoch": 1.1976779058309412,
      "grad_norm": 3.782750129699707,
      "learning_rate": 8.649833077685377e-05,
      "loss": 2.444538879394531,
      "memory(GiB)": 77.56,
      "step": 27955,
      "token_acc": 0.5476190476190477,
      "train_speed(iter/s)": 1.442025
    },
    {
      "epoch": 1.1978921211601903,
      "grad_norm": 7.202226638793945,
      "learning_rate": 8.649373076213035e-05,
      "loss": 2.836447334289551,
      "memory(GiB)": 77.56,
      "step": 27960,
      "token_acc": 0.4460431654676259,
      "train_speed(iter/s)": 1.442017
    },
    {
      "epoch": 1.198106336489439,
      "grad_norm": 5.353616237640381,
      "learning_rate": 8.648913008628874e-05,
      "loss": 2.3830245971679687,
      "memory(GiB)": 77.56,
      "step": 27965,
      "token_acc": 0.5267175572519084,
      "train_speed(iter/s)": 1.441956
    },
    {
      "epoch": 1.1983205518186881,
      "grad_norm": 4.24569034576416,
      "learning_rate": 8.648452874941232e-05,
      "loss": 2.7241838455200194,
      "memory(GiB)": 77.56,
      "step": 27970,
      "token_acc": 0.4756554307116105,
      "train_speed(iter/s)": 1.441986
    },
    {
      "epoch": 1.1985347671479372,
      "grad_norm": 6.994162082672119,
      "learning_rate": 8.647992675158443e-05,
      "loss": 2.473683547973633,
      "memory(GiB)": 77.56,
      "step": 27975,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.442001
    },
    {
      "epoch": 1.198748982477186,
      "grad_norm": 5.098709583282471,
      "learning_rate": 8.647532409288846e-05,
      "loss": 2.47924690246582,
      "memory(GiB)": 77.56,
      "step": 27980,
      "token_acc": 0.4890282131661442,
      "train_speed(iter/s)": 1.441996
    },
    {
      "epoch": 1.198963197806435,
      "grad_norm": 4.81426477432251,
      "learning_rate": 8.647072077340776e-05,
      "loss": 2.635319137573242,
      "memory(GiB)": 77.56,
      "step": 27985,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.442026
    },
    {
      "epoch": 1.199177413135684,
      "grad_norm": 4.6148881912231445,
      "learning_rate": 8.646611679322576e-05,
      "loss": 2.3667037963867186,
      "memory(GiB)": 77.56,
      "step": 27990,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.442048
    },
    {
      "epoch": 1.1993916284649329,
      "grad_norm": 5.3459601402282715,
      "learning_rate": 8.646151215242584e-05,
      "loss": 2.5292531967163088,
      "memory(GiB)": 77.56,
      "step": 27995,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.442065
    },
    {
      "epoch": 1.199605843794182,
      "grad_norm": 5.146407127380371,
      "learning_rate": 8.645690685109143e-05,
      "loss": 2.474112129211426,
      "memory(GiB)": 77.56,
      "step": 28000,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.441967
    },
    {
      "epoch": 1.199605843794182,
      "eval_loss": 2.4750900268554688,
      "eval_runtime": 13.7177,
      "eval_samples_per_second": 7.29,
      "eval_steps_per_second": 7.29,
      "eval_token_acc": 0.44711538461538464,
      "step": 28000
    },
    {
      "epoch": 1.199820059123431,
      "grad_norm": 5.654464244842529,
      "learning_rate": 8.645230088930593e-05,
      "loss": 2.449993133544922,
      "memory(GiB)": 77.56,
      "step": 28005,
      "token_acc": 0.45607808340727596,
      "train_speed(iter/s)": 1.440885
    },
    {
      "epoch": 1.2000342744526797,
      "grad_norm": 7.380916595458984,
      "learning_rate": 8.644769426715282e-05,
      "loss": 2.5308441162109374,
      "memory(GiB)": 77.56,
      "step": 28010,
      "token_acc": 0.4806201550387597,
      "train_speed(iter/s)": 1.440886
    },
    {
      "epoch": 1.2002484897819288,
      "grad_norm": 4.218809604644775,
      "learning_rate": 8.644308698471553e-05,
      "loss": 2.3638336181640627,
      "memory(GiB)": 77.56,
      "step": 28015,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.440915
    },
    {
      "epoch": 1.2004627051111778,
      "grad_norm": 4.558990955352783,
      "learning_rate": 8.643847904207755e-05,
      "loss": 2.814834213256836,
      "memory(GiB)": 77.56,
      "step": 28020,
      "token_acc": 0.43260188087774293,
      "train_speed(iter/s)": 1.441005
    },
    {
      "epoch": 1.2006769204404266,
      "grad_norm": 3.6906096935272217,
      "learning_rate": 8.643387043932232e-05,
      "loss": 2.5152759552001953,
      "memory(GiB)": 77.56,
      "step": 28025,
      "token_acc": 0.4889502762430939,
      "train_speed(iter/s)": 1.441012
    },
    {
      "epoch": 1.2008911357696757,
      "grad_norm": 5.000164031982422,
      "learning_rate": 8.642926117653336e-05,
      "loss": 2.4909133911132812,
      "memory(GiB)": 77.56,
      "step": 28030,
      "token_acc": 0.48627450980392156,
      "train_speed(iter/s)": 1.440989
    },
    {
      "epoch": 1.2011053510989247,
      "grad_norm": 4.8146586418151855,
      "learning_rate": 8.642465125379415e-05,
      "loss": 2.8136323928833007,
      "memory(GiB)": 77.56,
      "step": 28035,
      "token_acc": 0.43909348441926344,
      "train_speed(iter/s)": 1.441017
    },
    {
      "epoch": 1.2013195664281735,
      "grad_norm": 4.577399730682373,
      "learning_rate": 8.642004067118824e-05,
      "loss": 2.327980613708496,
      "memory(GiB)": 77.56,
      "step": 28040,
      "token_acc": 0.5198675496688742,
      "train_speed(iter/s)": 1.441061
    },
    {
      "epoch": 1.2015337817574225,
      "grad_norm": 5.458673000335693,
      "learning_rate": 8.64154294287991e-05,
      "loss": 2.6136566162109376,
      "memory(GiB)": 77.56,
      "step": 28045,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441085
    },
    {
      "epoch": 1.2017479970866716,
      "grad_norm": 4.3608880043029785,
      "learning_rate": 8.64108175267103e-05,
      "loss": 2.5459054946899413,
      "memory(GiB)": 77.56,
      "step": 28050,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.441053
    },
    {
      "epoch": 1.2019622124159204,
      "grad_norm": 4.632081508636475,
      "learning_rate": 8.64062049650054e-05,
      "loss": 2.4002363204956056,
      "memory(GiB)": 77.56,
      "step": 28055,
      "token_acc": 0.5358361774744027,
      "train_speed(iter/s)": 1.441042
    },
    {
      "epoch": 1.2021764277451694,
      "grad_norm": 5.0971503257751465,
      "learning_rate": 8.640159174376793e-05,
      "loss": 2.895140838623047,
      "memory(GiB)": 77.56,
      "step": 28060,
      "token_acc": 0.4037735849056604,
      "train_speed(iter/s)": 1.44107
    },
    {
      "epoch": 1.2023906430744185,
      "grad_norm": 5.100149154663086,
      "learning_rate": 8.63969778630815e-05,
      "loss": 2.2651060104370115,
      "memory(GiB)": 77.56,
      "step": 28065,
      "token_acc": 0.4944649446494465,
      "train_speed(iter/s)": 1.441088
    },
    {
      "epoch": 1.2026048584036673,
      "grad_norm": 4.2331223487854,
      "learning_rate": 8.639236332302964e-05,
      "loss": 2.599004554748535,
      "memory(GiB)": 77.56,
      "step": 28070,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.44116
    },
    {
      "epoch": 1.2028190737329163,
      "grad_norm": 5.119237422943115,
      "learning_rate": 8.638774812369599e-05,
      "loss": 2.554437255859375,
      "memory(GiB)": 77.56,
      "step": 28075,
      "token_acc": 0.4603658536585366,
      "train_speed(iter/s)": 1.441205
    },
    {
      "epoch": 1.2030332890621653,
      "grad_norm": 4.890880584716797,
      "learning_rate": 8.638313226516417e-05,
      "loss": 2.72982120513916,
      "memory(GiB)": 77.56,
      "step": 28080,
      "token_acc": 0.445993031358885,
      "train_speed(iter/s)": 1.441175
    },
    {
      "epoch": 1.2032475043914141,
      "grad_norm": 6.190558433532715,
      "learning_rate": 8.637851574751776e-05,
      "loss": 2.5672435760498047,
      "memory(GiB)": 77.56,
      "step": 28085,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.441195
    },
    {
      "epoch": 1.2034617197206632,
      "grad_norm": 3.8964924812316895,
      "learning_rate": 8.637389857084041e-05,
      "loss": 2.7974910736083984,
      "memory(GiB)": 77.56,
      "step": 28090,
      "token_acc": 0.4584615384615385,
      "train_speed(iter/s)": 1.441154
    },
    {
      "epoch": 1.2036759350499122,
      "grad_norm": 4.55295991897583,
      "learning_rate": 8.636928073521577e-05,
      "loss": 2.7204612731933593,
      "memory(GiB)": 77.56,
      "step": 28095,
      "token_acc": 0.4403183023872679,
      "train_speed(iter/s)": 1.44116
    },
    {
      "epoch": 1.203890150379161,
      "grad_norm": 4.697570323944092,
      "learning_rate": 8.63646622407275e-05,
      "loss": 2.5776172637939454,
      "memory(GiB)": 77.56,
      "step": 28100,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.441196
    },
    {
      "epoch": 1.20410436570841,
      "grad_norm": 4.943467140197754,
      "learning_rate": 8.636004308745925e-05,
      "loss": 2.399165916442871,
      "memory(GiB)": 77.56,
      "step": 28105,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.441151
    },
    {
      "epoch": 1.204318581037659,
      "grad_norm": 5.140091419219971,
      "learning_rate": 8.635542327549471e-05,
      "loss": 2.7213165283203127,
      "memory(GiB)": 77.56,
      "step": 28110,
      "token_acc": 0.4329896907216495,
      "train_speed(iter/s)": 1.441198
    },
    {
      "epoch": 1.204532796366908,
      "grad_norm": 4.847456932067871,
      "learning_rate": 8.635080280491758e-05,
      "loss": 2.42418212890625,
      "memory(GiB)": 77.56,
      "step": 28115,
      "token_acc": 0.48059701492537316,
      "train_speed(iter/s)": 1.441213
    },
    {
      "epoch": 1.204747011696157,
      "grad_norm": 4.328200340270996,
      "learning_rate": 8.634618167581156e-05,
      "loss": 2.562395477294922,
      "memory(GiB)": 77.56,
      "step": 28120,
      "token_acc": 0.4824561403508772,
      "train_speed(iter/s)": 1.441221
    },
    {
      "epoch": 1.204961227025406,
      "grad_norm": 5.6406073570251465,
      "learning_rate": 8.634155988826035e-05,
      "loss": 2.6292280197143554,
      "memory(GiB)": 77.56,
      "step": 28125,
      "token_acc": 0.4968152866242038,
      "train_speed(iter/s)": 1.441184
    },
    {
      "epoch": 1.2051754423546548,
      "grad_norm": 5.257379055023193,
      "learning_rate": 8.633693744234771e-05,
      "loss": 2.5983179092407225,
      "memory(GiB)": 77.56,
      "step": 28130,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.441208
    },
    {
      "epoch": 1.2053896576839038,
      "grad_norm": 5.054792404174805,
      "learning_rate": 8.633231433815735e-05,
      "loss": 2.593419075012207,
      "memory(GiB)": 77.56,
      "step": 28135,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.441242
    },
    {
      "epoch": 1.2056038730131529,
      "grad_norm": 4.093530178070068,
      "learning_rate": 8.632769057577304e-05,
      "loss": 2.6712366104125977,
      "memory(GiB)": 77.56,
      "step": 28140,
      "token_acc": 0.4420289855072464,
      "train_speed(iter/s)": 1.44127
    },
    {
      "epoch": 1.2058180883424017,
      "grad_norm": 4.249990463256836,
      "learning_rate": 8.632306615527853e-05,
      "loss": 2.654378128051758,
      "memory(GiB)": 77.56,
      "step": 28145,
      "token_acc": 0.4608150470219436,
      "train_speed(iter/s)": 1.441277
    },
    {
      "epoch": 1.2060323036716507,
      "grad_norm": 4.977593898773193,
      "learning_rate": 8.63184410767576e-05,
      "loss": 2.525018501281738,
      "memory(GiB)": 77.56,
      "step": 28150,
      "token_acc": 0.46407185628742514,
      "train_speed(iter/s)": 1.441286
    },
    {
      "epoch": 1.2062465190008997,
      "grad_norm": 5.1998515129089355,
      "learning_rate": 8.631381534029404e-05,
      "loss": 2.335630416870117,
      "memory(GiB)": 77.56,
      "step": 28155,
      "token_acc": 0.5059760956175299,
      "train_speed(iter/s)": 1.441202
    },
    {
      "epoch": 1.2064607343301486,
      "grad_norm": 5.370985984802246,
      "learning_rate": 8.630918894597166e-05,
      "loss": 2.2999839782714844,
      "memory(GiB)": 77.56,
      "step": 28160,
      "token_acc": 0.48627450980392156,
      "train_speed(iter/s)": 1.441192
    },
    {
      "epoch": 1.2066749496593976,
      "grad_norm": 5.30683708190918,
      "learning_rate": 8.630456189387426e-05,
      "loss": 2.421610450744629,
      "memory(GiB)": 77.56,
      "step": 28165,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.441244
    },
    {
      "epoch": 1.2068891649886466,
      "grad_norm": 12.023375511169434,
      "learning_rate": 8.629993418408566e-05,
      "loss": 2.8368083953857424,
      "memory(GiB)": 77.56,
      "step": 28170,
      "token_acc": 0.4939271255060729,
      "train_speed(iter/s)": 1.441283
    },
    {
      "epoch": 1.2071033803178954,
      "grad_norm": 5.206878662109375,
      "learning_rate": 8.62953058166897e-05,
      "loss": 2.3973886489868166,
      "memory(GiB)": 77.56,
      "step": 28175,
      "token_acc": 0.45918367346938777,
      "train_speed(iter/s)": 1.44133
    },
    {
      "epoch": 1.2073175956471445,
      "grad_norm": 4.301476955413818,
      "learning_rate": 8.629067679177023e-05,
      "loss": 2.366256523132324,
      "memory(GiB)": 77.56,
      "step": 28180,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.441318
    },
    {
      "epoch": 1.2075318109763935,
      "grad_norm": 6.912600040435791,
      "learning_rate": 8.62860471094111e-05,
      "loss": 2.543890380859375,
      "memory(GiB)": 77.56,
      "step": 28185,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.441318
    },
    {
      "epoch": 1.2077460263056423,
      "grad_norm": 4.203740119934082,
      "learning_rate": 8.62814167696962e-05,
      "loss": 2.4901100158691407,
      "memory(GiB)": 77.56,
      "step": 28190,
      "token_acc": 0.4559748427672956,
      "train_speed(iter/s)": 1.441281
    },
    {
      "epoch": 1.2079602416348914,
      "grad_norm": 6.206925392150879,
      "learning_rate": 8.627678577270939e-05,
      "loss": 2.9209819793701173,
      "memory(GiB)": 77.56,
      "step": 28195,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.441294
    },
    {
      "epoch": 1.2081744569641404,
      "grad_norm": 4.495066165924072,
      "learning_rate": 8.627215411853459e-05,
      "loss": 2.5599943161010743,
      "memory(GiB)": 77.56,
      "step": 28200,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.441317
    },
    {
      "epoch": 1.2083886722933892,
      "grad_norm": 4.902633190155029,
      "learning_rate": 8.626752180725568e-05,
      "loss": 2.6431455612182617,
      "memory(GiB)": 77.56,
      "step": 28205,
      "token_acc": 0.4751958224543081,
      "train_speed(iter/s)": 1.441363
    },
    {
      "epoch": 1.2086028876226382,
      "grad_norm": 4.8289995193481445,
      "learning_rate": 8.626288883895659e-05,
      "loss": 2.56491756439209,
      "memory(GiB)": 77.56,
      "step": 28210,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.441318
    },
    {
      "epoch": 1.2088171029518873,
      "grad_norm": 4.488157749176025,
      "learning_rate": 8.625825521372125e-05,
      "loss": 2.8338550567626952,
      "memory(GiB)": 77.56,
      "step": 28215,
      "token_acc": 0.4451219512195122,
      "train_speed(iter/s)": 1.441291
    },
    {
      "epoch": 1.209031318281136,
      "grad_norm": 5.5533552169799805,
      "learning_rate": 8.62536209316336e-05,
      "loss": 2.3425302505493164,
      "memory(GiB)": 77.56,
      "step": 28220,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441351
    },
    {
      "epoch": 1.2092455336103851,
      "grad_norm": 5.305296421051025,
      "learning_rate": 8.624898599277762e-05,
      "loss": 2.6979038238525392,
      "memory(GiB)": 77.56,
      "step": 28225,
      "token_acc": 0.44660194174757284,
      "train_speed(iter/s)": 1.441319
    },
    {
      "epoch": 1.2094597489396341,
      "grad_norm": 4.448836326599121,
      "learning_rate": 8.624435039723724e-05,
      "loss": 2.3078176498413088,
      "memory(GiB)": 77.56,
      "step": 28230,
      "token_acc": 0.5430711610486891,
      "train_speed(iter/s)": 1.441363
    },
    {
      "epoch": 1.209673964268883,
      "grad_norm": 4.726208209991455,
      "learning_rate": 8.623971414509644e-05,
      "loss": 2.6113344192504884,
      "memory(GiB)": 77.56,
      "step": 28235,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.441373
    },
    {
      "epoch": 1.209888179598132,
      "grad_norm": 4.682325839996338,
      "learning_rate": 8.623507723643924e-05,
      "loss": 2.412847709655762,
      "memory(GiB)": 77.56,
      "step": 28240,
      "token_acc": 0.4820717131474104,
      "train_speed(iter/s)": 1.44138
    },
    {
      "epoch": 1.210102394927381,
      "grad_norm": 13.473363876342773,
      "learning_rate": 8.623043967134963e-05,
      "loss": 2.4895746231079103,
      "memory(GiB)": 77.56,
      "step": 28245,
      "token_acc": 0.509493670886076,
      "train_speed(iter/s)": 1.44142
    },
    {
      "epoch": 1.2103166102566298,
      "grad_norm": 5.712949752807617,
      "learning_rate": 8.622580144991159e-05,
      "loss": 2.6608713150024412,
      "memory(GiB)": 77.56,
      "step": 28250,
      "token_acc": 0.4607142857142857,
      "train_speed(iter/s)": 1.441444
    },
    {
      "epoch": 1.2105308255858789,
      "grad_norm": 6.19139289855957,
      "learning_rate": 8.622116257220921e-05,
      "loss": 2.5623245239257812,
      "memory(GiB)": 77.56,
      "step": 28255,
      "token_acc": 0.4280701754385965,
      "train_speed(iter/s)": 1.441443
    },
    {
      "epoch": 1.210745040915128,
      "grad_norm": 5.0009074211120605,
      "learning_rate": 8.621652303832647e-05,
      "loss": 2.79967041015625,
      "memory(GiB)": 77.56,
      "step": 28260,
      "token_acc": 0.4290909090909091,
      "train_speed(iter/s)": 1.441461
    },
    {
      "epoch": 1.2109592562443767,
      "grad_norm": 5.7083563804626465,
      "learning_rate": 8.621188284834745e-05,
      "loss": 2.3252872467041015,
      "memory(GiB)": 77.56,
      "step": 28265,
      "token_acc": 0.47435897435897434,
      "train_speed(iter/s)": 1.441466
    },
    {
      "epoch": 1.2111734715736258,
      "grad_norm": 4.7395920753479,
      "learning_rate": 8.62072420023562e-05,
      "loss": 2.365593910217285,
      "memory(GiB)": 77.56,
      "step": 28270,
      "token_acc": 0.5138339920948617,
      "train_speed(iter/s)": 1.441427
    },
    {
      "epoch": 1.2113876869028748,
      "grad_norm": 5.456732749938965,
      "learning_rate": 8.620260050043678e-05,
      "loss": 2.5022214889526366,
      "memory(GiB)": 77.56,
      "step": 28275,
      "token_acc": 0.47079037800687284,
      "train_speed(iter/s)": 1.44145
    },
    {
      "epoch": 1.2116019022321236,
      "grad_norm": 4.1444172859191895,
      "learning_rate": 8.619795834267331e-05,
      "loss": 2.7523942947387696,
      "memory(GiB)": 77.56,
      "step": 28280,
      "token_acc": 0.4425087108013937,
      "train_speed(iter/s)": 1.441552
    },
    {
      "epoch": 1.2118161175613726,
      "grad_norm": 4.467493534088135,
      "learning_rate": 8.619331552914987e-05,
      "loss": 2.402785301208496,
      "memory(GiB)": 77.56,
      "step": 28285,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.441565
    },
    {
      "epoch": 1.2120303328906217,
      "grad_norm": 5.909074306488037,
      "learning_rate": 8.618867205995056e-05,
      "loss": 2.17706298828125,
      "memory(GiB)": 77.56,
      "step": 28290,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.441568
    },
    {
      "epoch": 1.2122445482198707,
      "grad_norm": 6.347830295562744,
      "learning_rate": 8.618402793515949e-05,
      "loss": 2.550565719604492,
      "memory(GiB)": 77.56,
      "step": 28295,
      "token_acc": 0.47491638795986624,
      "train_speed(iter/s)": 1.441625
    },
    {
      "epoch": 1.2124587635491195,
      "grad_norm": 4.585397720336914,
      "learning_rate": 8.617938315486084e-05,
      "loss": 2.403661346435547,
      "memory(GiB)": 77.56,
      "step": 28300,
      "token_acc": 0.4900398406374502,
      "train_speed(iter/s)": 1.441584
    },
    {
      "epoch": 1.2126729788783686,
      "grad_norm": 5.238585472106934,
      "learning_rate": 8.617473771913871e-05,
      "loss": 2.9338611602783202,
      "memory(GiB)": 77.56,
      "step": 28305,
      "token_acc": 0.421259842519685,
      "train_speed(iter/s)": 1.44154
    },
    {
      "epoch": 1.2128871942076176,
      "grad_norm": 4.562910556793213,
      "learning_rate": 8.617009162807727e-05,
      "loss": 2.6408969879150392,
      "memory(GiB)": 77.56,
      "step": 28310,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.44154
    },
    {
      "epoch": 1.2131014095368664,
      "grad_norm": 5.460917949676514,
      "learning_rate": 8.61654448817607e-05,
      "loss": 2.3609785079956054,
      "memory(GiB)": 77.56,
      "step": 28315,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.441597
    },
    {
      "epoch": 1.2133156248661154,
      "grad_norm": 5.652011394500732,
      "learning_rate": 8.616079748027317e-05,
      "loss": 2.647988128662109,
      "memory(GiB)": 77.56,
      "step": 28320,
      "token_acc": 0.4559386973180077,
      "train_speed(iter/s)": 1.441528
    },
    {
      "epoch": 1.2135298401953645,
      "grad_norm": 7.903143882751465,
      "learning_rate": 8.615614942369887e-05,
      "loss": 2.2158252716064455,
      "memory(GiB)": 77.56,
      "step": 28325,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.441444
    },
    {
      "epoch": 1.2137440555246133,
      "grad_norm": 4.761894226074219,
      "learning_rate": 8.6151500712122e-05,
      "loss": 2.592799758911133,
      "memory(GiB)": 77.56,
      "step": 28330,
      "token_acc": 0.4778761061946903,
      "train_speed(iter/s)": 1.441475
    },
    {
      "epoch": 1.2139582708538623,
      "grad_norm": 4.6801018714904785,
      "learning_rate": 8.614685134562679e-05,
      "loss": 2.719898223876953,
      "memory(GiB)": 77.56,
      "step": 28335,
      "token_acc": 0.4732824427480916,
      "train_speed(iter/s)": 1.441496
    },
    {
      "epoch": 1.2141724861831114,
      "grad_norm": 4.945274353027344,
      "learning_rate": 8.614220132429746e-05,
      "loss": 2.555727958679199,
      "memory(GiB)": 77.56,
      "step": 28340,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.441495
    },
    {
      "epoch": 1.2143867015123602,
      "grad_norm": 5.15886116027832,
      "learning_rate": 8.613755064821824e-05,
      "loss": 2.8917753219604494,
      "memory(GiB)": 77.56,
      "step": 28345,
      "token_acc": 0.40671641791044777,
      "train_speed(iter/s)": 1.441518
    },
    {
      "epoch": 1.2146009168416092,
      "grad_norm": 4.952661037445068,
      "learning_rate": 8.613289931747339e-05,
      "loss": 2.2962602615356444,
      "memory(GiB)": 77.56,
      "step": 28350,
      "token_acc": 0.5271317829457365,
      "train_speed(iter/s)": 1.441507
    },
    {
      "epoch": 1.2148151321708582,
      "grad_norm": 4.85758113861084,
      "learning_rate": 8.612824733214717e-05,
      "loss": 2.3349756240844726,
      "memory(GiB)": 77.56,
      "step": 28355,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441518
    },
    {
      "epoch": 1.215029347500107,
      "grad_norm": 5.112732887268066,
      "learning_rate": 8.612359469232387e-05,
      "loss": 2.2662685394287108,
      "memory(GiB)": 77.56,
      "step": 28360,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.441528
    },
    {
      "epoch": 1.215243562829356,
      "grad_norm": 4.784420967102051,
      "learning_rate": 8.611894139808776e-05,
      "loss": 2.8001258850097654,
      "memory(GiB)": 77.56,
      "step": 28365,
      "token_acc": 0.4318181818181818,
      "train_speed(iter/s)": 1.441469
    },
    {
      "epoch": 1.2154577781586051,
      "grad_norm": 6.794675350189209,
      "learning_rate": 8.611428744952315e-05,
      "loss": 2.2682168960571287,
      "memory(GiB)": 77.56,
      "step": 28370,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.441502
    },
    {
      "epoch": 1.215671993487854,
      "grad_norm": 4.324798583984375,
      "learning_rate": 8.610963284671433e-05,
      "loss": 2.441996383666992,
      "memory(GiB)": 77.56,
      "step": 28375,
      "token_acc": 0.4717741935483871,
      "train_speed(iter/s)": 1.441481
    },
    {
      "epoch": 1.215886208817103,
      "grad_norm": 6.754473686218262,
      "learning_rate": 8.610497758974566e-05,
      "loss": 2.261363410949707,
      "memory(GiB)": 77.56,
      "step": 28380,
      "token_acc": 0.4896265560165975,
      "train_speed(iter/s)": 1.441501
    },
    {
      "epoch": 1.216100424146352,
      "grad_norm": 4.434762954711914,
      "learning_rate": 8.610032167870144e-05,
      "loss": 2.4972591400146484,
      "memory(GiB)": 77.56,
      "step": 28385,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.441506
    },
    {
      "epoch": 1.2163146394756008,
      "grad_norm": 5.20905065536499,
      "learning_rate": 8.609566511366603e-05,
      "loss": 2.5753446578979493,
      "memory(GiB)": 77.56,
      "step": 28390,
      "token_acc": 0.47410358565737054,
      "train_speed(iter/s)": 1.44156
    },
    {
      "epoch": 1.2165288548048498,
      "grad_norm": 5.327287673950195,
      "learning_rate": 8.609100789472377e-05,
      "loss": 2.5779697418212892,
      "memory(GiB)": 77.56,
      "step": 28395,
      "token_acc": 0.49843260188087773,
      "train_speed(iter/s)": 1.441603
    },
    {
      "epoch": 1.2167430701340989,
      "grad_norm": 7.3156633377075195,
      "learning_rate": 8.608635002195908e-05,
      "loss": 2.1596099853515627,
      "memory(GiB)": 77.56,
      "step": 28400,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.441565
    },
    {
      "epoch": 1.2169572854633477,
      "grad_norm": 5.743879318237305,
      "learning_rate": 8.608169149545629e-05,
      "loss": 2.638324737548828,
      "memory(GiB)": 77.56,
      "step": 28405,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.441563
    },
    {
      "epoch": 1.2171715007925967,
      "grad_norm": 5.754978179931641,
      "learning_rate": 8.60770323152998e-05,
      "loss": 2.3123592376708983,
      "memory(GiB)": 77.56,
      "step": 28410,
      "token_acc": 0.5076452599388379,
      "train_speed(iter/s)": 1.441579
    },
    {
      "epoch": 1.2173857161218458,
      "grad_norm": 5.318662643432617,
      "learning_rate": 8.607237248157403e-05,
      "loss": 2.647915840148926,
      "memory(GiB)": 77.56,
      "step": 28415,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.441529
    },
    {
      "epoch": 1.2175999314510946,
      "grad_norm": 6.1738505363464355,
      "learning_rate": 8.606771199436341e-05,
      "loss": 2.607415771484375,
      "memory(GiB)": 77.56,
      "step": 28420,
      "token_acc": 0.425531914893617,
      "train_speed(iter/s)": 1.441522
    },
    {
      "epoch": 1.2178141467803436,
      "grad_norm": 5.154613018035889,
      "learning_rate": 8.606305085375234e-05,
      "loss": 2.6470447540283204,
      "memory(GiB)": 77.56,
      "step": 28425,
      "token_acc": 0.43956043956043955,
      "train_speed(iter/s)": 1.441481
    },
    {
      "epoch": 1.2180283621095926,
      "grad_norm": 3.8204874992370605,
      "learning_rate": 8.605838905982526e-05,
      "loss": 2.443175506591797,
      "memory(GiB)": 77.56,
      "step": 28430,
      "token_acc": 0.49393939393939396,
      "train_speed(iter/s)": 1.441441
    },
    {
      "epoch": 1.2182425774388415,
      "grad_norm": 4.8001179695129395,
      "learning_rate": 8.605372661266667e-05,
      "loss": 2.530910110473633,
      "memory(GiB)": 77.56,
      "step": 28435,
      "token_acc": 0.46616541353383456,
      "train_speed(iter/s)": 1.441458
    },
    {
      "epoch": 1.2184567927680905,
      "grad_norm": 4.321229457855225,
      "learning_rate": 8.604906351236097e-05,
      "loss": 2.264784812927246,
      "memory(GiB)": 77.56,
      "step": 28440,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.441474
    },
    {
      "epoch": 1.2186710080973395,
      "grad_norm": 5.285921096801758,
      "learning_rate": 8.604439975899269e-05,
      "loss": 2.804207611083984,
      "memory(GiB)": 77.56,
      "step": 28445,
      "token_acc": 0.44776119402985076,
      "train_speed(iter/s)": 1.441469
    },
    {
      "epoch": 1.2188852234265883,
      "grad_norm": 4.859533309936523,
      "learning_rate": 8.60397353526463e-05,
      "loss": 2.3932064056396483,
      "memory(GiB)": 77.56,
      "step": 28450,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.441479
    },
    {
      "epoch": 1.2190994387558374,
      "grad_norm": 6.09795618057251,
      "learning_rate": 8.603507029340627e-05,
      "loss": 2.565334510803223,
      "memory(GiB)": 77.56,
      "step": 28455,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.441524
    },
    {
      "epoch": 1.2193136540850864,
      "grad_norm": 6.738732814788818,
      "learning_rate": 8.603040458135715e-05,
      "loss": 2.468765640258789,
      "memory(GiB)": 77.56,
      "step": 28460,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.441507
    },
    {
      "epoch": 1.2195278694143352,
      "grad_norm": 4.478949546813965,
      "learning_rate": 8.602573821658345e-05,
      "loss": 2.242085266113281,
      "memory(GiB)": 77.56,
      "step": 28465,
      "token_acc": 0.5144694533762058,
      "train_speed(iter/s)": 1.441465
    },
    {
      "epoch": 1.2197420847435843,
      "grad_norm": 8.1392183303833,
      "learning_rate": 8.602107119916971e-05,
      "loss": 2.906325912475586,
      "memory(GiB)": 77.56,
      "step": 28470,
      "token_acc": 0.4232081911262799,
      "train_speed(iter/s)": 1.441498
    },
    {
      "epoch": 1.2199563000728333,
      "grad_norm": 4.383642196655273,
      "learning_rate": 8.601640352920049e-05,
      "loss": 2.4978002548217773,
      "memory(GiB)": 77.56,
      "step": 28475,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.441542
    },
    {
      "epoch": 1.220170515402082,
      "grad_norm": 5.771337509155273,
      "learning_rate": 8.601173520676031e-05,
      "loss": 2.406903839111328,
      "memory(GiB)": 77.56,
      "step": 28480,
      "token_acc": 0.5056179775280899,
      "train_speed(iter/s)": 1.441578
    },
    {
      "epoch": 1.2203847307313311,
      "grad_norm": 5.979637145996094,
      "learning_rate": 8.600706623193377e-05,
      "loss": 2.421849822998047,
      "memory(GiB)": 77.56,
      "step": 28485,
      "token_acc": 0.4730290456431535,
      "train_speed(iter/s)": 1.441614
    },
    {
      "epoch": 1.2205989460605802,
      "grad_norm": 5.303152084350586,
      "learning_rate": 8.600239660480546e-05,
      "loss": 2.4490163803100584,
      "memory(GiB)": 77.56,
      "step": 28490,
      "token_acc": 0.4961832061068702,
      "train_speed(iter/s)": 1.441586
    },
    {
      "epoch": 1.220813161389829,
      "grad_norm": 5.009498596191406,
      "learning_rate": 8.599772632545995e-05,
      "loss": 2.367350387573242,
      "memory(GiB)": 77.56,
      "step": 28495,
      "token_acc": 0.5148514851485149,
      "train_speed(iter/s)": 1.441588
    },
    {
      "epoch": 1.221027376719078,
      "grad_norm": 6.146751403808594,
      "learning_rate": 8.599305539398186e-05,
      "loss": 2.753386878967285,
      "memory(GiB)": 77.56,
      "step": 28500,
      "token_acc": 0.44982698961937717,
      "train_speed(iter/s)": 1.44162
    },
    {
      "epoch": 1.221027376719078,
      "eval_loss": 2.276956558227539,
      "eval_runtime": 14.061,
      "eval_samples_per_second": 7.112,
      "eval_steps_per_second": 7.112,
      "eval_token_acc": 0.45796737766624845,
      "step": 28500
    },
    {
      "epoch": 1.221241592048327,
      "grad_norm": 4.489764213562012,
      "learning_rate": 8.598838381045582e-05,
      "loss": 2.6436840057373048,
      "memory(GiB)": 77.56,
      "step": 28505,
      "token_acc": 0.4633920296570899,
      "train_speed(iter/s)": 1.440588
    },
    {
      "epoch": 1.2214558073775759,
      "grad_norm": 5.637967586517334,
      "learning_rate": 8.598371157496642e-05,
      "loss": 2.493491554260254,
      "memory(GiB)": 77.56,
      "step": 28510,
      "token_acc": 0.5039370078740157,
      "train_speed(iter/s)": 1.440562
    },
    {
      "epoch": 1.221670022706825,
      "grad_norm": 6.490506649017334,
      "learning_rate": 8.597903868759836e-05,
      "loss": 2.630386734008789,
      "memory(GiB)": 77.56,
      "step": 28515,
      "token_acc": 0.4697802197802198,
      "train_speed(iter/s)": 1.440604
    },
    {
      "epoch": 1.221884238036074,
      "grad_norm": 4.837250232696533,
      "learning_rate": 8.597436514843625e-05,
      "loss": 2.6151580810546875,
      "memory(GiB)": 77.56,
      "step": 28520,
      "token_acc": 0.4641509433962264,
      "train_speed(iter/s)": 1.440655
    },
    {
      "epoch": 1.2220984533653227,
      "grad_norm": 6.502688884735107,
      "learning_rate": 8.596969095756478e-05,
      "loss": 2.8448753356933594,
      "memory(GiB)": 77.56,
      "step": 28525,
      "token_acc": 0.39490445859872614,
      "train_speed(iter/s)": 1.440695
    },
    {
      "epoch": 1.2223126686945718,
      "grad_norm": 4.855820178985596,
      "learning_rate": 8.59650161150686e-05,
      "loss": 2.3215974807739257,
      "memory(GiB)": 77.56,
      "step": 28530,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.440581
    },
    {
      "epoch": 1.2225268840238208,
      "grad_norm": 4.5305914878845215,
      "learning_rate": 8.596034062103242e-05,
      "loss": 2.51825065612793,
      "memory(GiB)": 77.56,
      "step": 28535,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.440541
    },
    {
      "epoch": 1.2227410993530696,
      "grad_norm": 4.3391947746276855,
      "learning_rate": 8.595566447554093e-05,
      "loss": 2.5289453506469726,
      "memory(GiB)": 77.56,
      "step": 28540,
      "token_acc": 0.4801587301587302,
      "train_speed(iter/s)": 1.440539
    },
    {
      "epoch": 1.2229553146823187,
      "grad_norm": 4.805681228637695,
      "learning_rate": 8.595098767867886e-05,
      "loss": 2.638051414489746,
      "memory(GiB)": 77.56,
      "step": 28545,
      "token_acc": 0.44551282051282054,
      "train_speed(iter/s)": 1.440564
    },
    {
      "epoch": 1.2231695300115677,
      "grad_norm": 3.7796120643615723,
      "learning_rate": 8.594631023053093e-05,
      "loss": 2.52237491607666,
      "memory(GiB)": 77.56,
      "step": 28550,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.440557
    },
    {
      "epoch": 1.2233837453408165,
      "grad_norm": 3.972053050994873,
      "learning_rate": 8.594163213118185e-05,
      "loss": 2.4955272674560547,
      "memory(GiB)": 77.56,
      "step": 28555,
      "token_acc": 0.48466257668711654,
      "train_speed(iter/s)": 1.44056
    },
    {
      "epoch": 1.2235979606700655,
      "grad_norm": 5.100379467010498,
      "learning_rate": 8.593695338071639e-05,
      "loss": 2.626585578918457,
      "memory(GiB)": 77.56,
      "step": 28560,
      "token_acc": 0.4733893557422969,
      "train_speed(iter/s)": 1.440567
    },
    {
      "epoch": 1.2238121759993146,
      "grad_norm": 5.830081939697266,
      "learning_rate": 8.593227397921932e-05,
      "loss": 2.6314258575439453,
      "memory(GiB)": 77.56,
      "step": 28565,
      "token_acc": 0.47039473684210525,
      "train_speed(iter/s)": 1.440621
    },
    {
      "epoch": 1.2240263913285634,
      "grad_norm": 4.231169700622559,
      "learning_rate": 8.59275939267754e-05,
      "loss": 2.4682697296142577,
      "memory(GiB)": 77.56,
      "step": 28570,
      "token_acc": 0.5324675324675324,
      "train_speed(iter/s)": 1.440554
    },
    {
      "epoch": 1.2242406066578124,
      "grad_norm": 4.236193656921387,
      "learning_rate": 8.59229132234694e-05,
      "loss": 2.3920560836791993,
      "memory(GiB)": 77.56,
      "step": 28575,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.440555
    },
    {
      "epoch": 1.2244548219870615,
      "grad_norm": 6.2157440185546875,
      "learning_rate": 8.591823186938614e-05,
      "loss": 2.2324760437011717,
      "memory(GiB)": 77.56,
      "step": 28580,
      "token_acc": 0.513677811550152,
      "train_speed(iter/s)": 1.440593
    },
    {
      "epoch": 1.2246690373163103,
      "grad_norm": 5.721351146697998,
      "learning_rate": 8.591354986461042e-05,
      "loss": 2.4571372985839846,
      "memory(GiB)": 77.56,
      "step": 28585,
      "token_acc": 0.46062992125984253,
      "train_speed(iter/s)": 1.440644
    },
    {
      "epoch": 1.2248832526455593,
      "grad_norm": 9.207639694213867,
      "learning_rate": 8.590886720922704e-05,
      "loss": 2.6439254760742186,
      "memory(GiB)": 77.56,
      "step": 28590,
      "token_acc": 0.40540540540540543,
      "train_speed(iter/s)": 1.440612
    },
    {
      "epoch": 1.2250974679748083,
      "grad_norm": 4.636092185974121,
      "learning_rate": 8.590418390332084e-05,
      "loss": 2.6139123916625975,
      "memory(GiB)": 77.56,
      "step": 28595,
      "token_acc": 0.4523076923076923,
      "train_speed(iter/s)": 1.440659
    },
    {
      "epoch": 1.2253116833040572,
      "grad_norm": 9.474297523498535,
      "learning_rate": 8.589949994697669e-05,
      "loss": 2.553696060180664,
      "memory(GiB)": 77.56,
      "step": 28600,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.440722
    },
    {
      "epoch": 1.2255258986333062,
      "grad_norm": 4.777644634246826,
      "learning_rate": 8.589481534027941e-05,
      "loss": 2.6511844635009765,
      "memory(GiB)": 77.56,
      "step": 28605,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.440744
    },
    {
      "epoch": 1.2257401139625552,
      "grad_norm": 4.623128890991211,
      "learning_rate": 8.589013008331388e-05,
      "loss": 2.417831802368164,
      "memory(GiB)": 77.56,
      "step": 28610,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.440804
    },
    {
      "epoch": 1.225954329291804,
      "grad_norm": 5.694374084472656,
      "learning_rate": 8.588544417616497e-05,
      "loss": 2.4633270263671876,
      "memory(GiB)": 77.56,
      "step": 28615,
      "token_acc": 0.4756554307116105,
      "train_speed(iter/s)": 1.440837
    },
    {
      "epoch": 1.226168544621053,
      "grad_norm": 4.325733184814453,
      "learning_rate": 8.588075761891758e-05,
      "loss": 2.691203308105469,
      "memory(GiB)": 77.56,
      "step": 28620,
      "token_acc": 0.4980544747081712,
      "train_speed(iter/s)": 1.440783
    },
    {
      "epoch": 1.226382759950302,
      "grad_norm": 5.457977294921875,
      "learning_rate": 8.587607041165662e-05,
      "loss": 2.687010383605957,
      "memory(GiB)": 77.56,
      "step": 28625,
      "token_acc": 0.4742268041237113,
      "train_speed(iter/s)": 1.440775
    },
    {
      "epoch": 1.226596975279551,
      "grad_norm": 5.60301399230957,
      "learning_rate": 8.587138255446698e-05,
      "loss": 2.6495382308959963,
      "memory(GiB)": 77.56,
      "step": 28630,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.440725
    },
    {
      "epoch": 1.2268111906088,
      "grad_norm": 4.6819915771484375,
      "learning_rate": 8.586669404743359e-05,
      "loss": 2.4744903564453127,
      "memory(GiB)": 77.56,
      "step": 28635,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.440686
    },
    {
      "epoch": 1.227025405938049,
      "grad_norm": 6.039593696594238,
      "learning_rate": 8.586200489064138e-05,
      "loss": 2.622127151489258,
      "memory(GiB)": 77.56,
      "step": 28640,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.440728
    },
    {
      "epoch": 1.2272396212672978,
      "grad_norm": 4.093668460845947,
      "learning_rate": 8.585731508417533e-05,
      "loss": 2.571345901489258,
      "memory(GiB)": 77.56,
      "step": 28645,
      "token_acc": 0.4570552147239264,
      "train_speed(iter/s)": 1.440806
    },
    {
      "epoch": 1.2274538365965468,
      "grad_norm": 4.982674598693848,
      "learning_rate": 8.585262462812038e-05,
      "loss": 2.5796134948730467,
      "memory(GiB)": 77.56,
      "step": 28650,
      "token_acc": 0.47249190938511326,
      "train_speed(iter/s)": 1.440791
    },
    {
      "epoch": 1.2276680519257959,
      "grad_norm": 4.81125545501709,
      "learning_rate": 8.584793352256149e-05,
      "loss": 2.326946258544922,
      "memory(GiB)": 77.56,
      "step": 28655,
      "token_acc": 0.503968253968254,
      "train_speed(iter/s)": 1.440837
    },
    {
      "epoch": 1.2278822672550447,
      "grad_norm": 5.603147983551025,
      "learning_rate": 8.584324176758367e-05,
      "loss": 2.6811941146850584,
      "memory(GiB)": 77.56,
      "step": 28660,
      "token_acc": 0.43670886075949367,
      "train_speed(iter/s)": 1.440839
    },
    {
      "epoch": 1.2280964825842937,
      "grad_norm": 4.917667388916016,
      "learning_rate": 8.583854936327189e-05,
      "loss": 2.6356378555297852,
      "memory(GiB)": 77.56,
      "step": 28665,
      "token_acc": 0.501577287066246,
      "train_speed(iter/s)": 1.440775
    },
    {
      "epoch": 1.2283106979135427,
      "grad_norm": 7.304621696472168,
      "learning_rate": 8.583385630971118e-05,
      "loss": 2.55916748046875,
      "memory(GiB)": 77.56,
      "step": 28670,
      "token_acc": 0.5093167701863354,
      "train_speed(iter/s)": 1.440808
    },
    {
      "epoch": 1.2285249132427916,
      "grad_norm": 5.387117385864258,
      "learning_rate": 8.582916260698655e-05,
      "loss": 2.304153823852539,
      "memory(GiB)": 77.56,
      "step": 28675,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.44085
    },
    {
      "epoch": 1.2287391285720406,
      "grad_norm": 3.856370687484741,
      "learning_rate": 8.582446825518302e-05,
      "loss": 2.2381420135498047,
      "memory(GiB)": 77.56,
      "step": 28680,
      "token_acc": 0.524390243902439,
      "train_speed(iter/s)": 1.440874
    },
    {
      "epoch": 1.2289533439012896,
      "grad_norm": 4.070316314697266,
      "learning_rate": 8.581977325438564e-05,
      "loss": 2.48612060546875,
      "memory(GiB)": 77.56,
      "step": 28685,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.440927
    },
    {
      "epoch": 1.2291675592305384,
      "grad_norm": 4.756408214569092,
      "learning_rate": 8.581507760467945e-05,
      "loss": 2.7492162704467775,
      "memory(GiB)": 77.56,
      "step": 28690,
      "token_acc": 0.41883116883116883,
      "train_speed(iter/s)": 1.440974
    },
    {
      "epoch": 1.2293817745597875,
      "grad_norm": 4.741931915283203,
      "learning_rate": 8.581038130614957e-05,
      "loss": 2.7649917602539062,
      "memory(GiB)": 77.56,
      "step": 28695,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.440958
    },
    {
      "epoch": 1.2295959898890365,
      "grad_norm": 5.995345115661621,
      "learning_rate": 8.580568435888102e-05,
      "loss": 2.6990283966064452,
      "memory(GiB)": 77.56,
      "step": 28700,
      "token_acc": 0.42771084337349397,
      "train_speed(iter/s)": 1.44087
    },
    {
      "epoch": 1.2298102052182853,
      "grad_norm": 6.213626384735107,
      "learning_rate": 8.580098676295891e-05,
      "loss": 2.670740509033203,
      "memory(GiB)": 77.56,
      "step": 28705,
      "token_acc": 0.41543026706231456,
      "train_speed(iter/s)": 1.440881
    },
    {
      "epoch": 1.2300244205475344,
      "grad_norm": 4.130982875823975,
      "learning_rate": 8.579628851846834e-05,
      "loss": 2.2273691177368162,
      "memory(GiB)": 77.56,
      "step": 28710,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.440919
    },
    {
      "epoch": 1.2302386358767834,
      "grad_norm": 4.355827808380127,
      "learning_rate": 8.579158962549443e-05,
      "loss": 2.4503570556640626,
      "memory(GiB)": 77.56,
      "step": 28715,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.440942
    },
    {
      "epoch": 1.2304528512060324,
      "grad_norm": 4.211069107055664,
      "learning_rate": 8.578689008412229e-05,
      "loss": 2.5717063903808595,
      "memory(GiB)": 77.56,
      "step": 28720,
      "token_acc": 0.48338368580060426,
      "train_speed(iter/s)": 1.440999
    },
    {
      "epoch": 1.2306670665352812,
      "grad_norm": 5.150546073913574,
      "learning_rate": 8.578218989443706e-05,
      "loss": 2.775617790222168,
      "memory(GiB)": 77.56,
      "step": 28725,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.441009
    },
    {
      "epoch": 1.2308812818645303,
      "grad_norm": 5.9000935554504395,
      "learning_rate": 8.577748905652389e-05,
      "loss": 2.530570411682129,
      "memory(GiB)": 77.56,
      "step": 28730,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.441075
    },
    {
      "epoch": 1.2310954971937793,
      "grad_norm": 4.757238864898682,
      "learning_rate": 8.577278757046797e-05,
      "loss": 2.39785099029541,
      "memory(GiB)": 77.56,
      "step": 28735,
      "token_acc": 0.484,
      "train_speed(iter/s)": 1.441112
    },
    {
      "epoch": 1.2313097125230281,
      "grad_norm": 4.571513652801514,
      "learning_rate": 8.576808543635443e-05,
      "loss": 2.4586132049560545,
      "memory(GiB)": 77.56,
      "step": 28740,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.441096
    },
    {
      "epoch": 1.2315239278522772,
      "grad_norm": 4.509367942810059,
      "learning_rate": 8.576338265426846e-05,
      "loss": 2.420689010620117,
      "memory(GiB)": 77.56,
      "step": 28745,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.441122
    },
    {
      "epoch": 1.2317381431815262,
      "grad_norm": 4.456842422485352,
      "learning_rate": 8.575867922429529e-05,
      "loss": 2.5845626831054687,
      "memory(GiB)": 77.56,
      "step": 28750,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.441171
    },
    {
      "epoch": 1.231952358510775,
      "grad_norm": 4.6104559898376465,
      "learning_rate": 8.575397514652008e-05,
      "loss": 2.108664131164551,
      "memory(GiB)": 77.56,
      "step": 28755,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.441142
    },
    {
      "epoch": 1.232166573840024,
      "grad_norm": 7.478261470794678,
      "learning_rate": 8.574927042102807e-05,
      "loss": 2.7321355819702147,
      "memory(GiB)": 77.56,
      "step": 28760,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.441208
    },
    {
      "epoch": 1.232380789169273,
      "grad_norm": 7.4106526374816895,
      "learning_rate": 8.574456504790451e-05,
      "loss": 2.5289710998535155,
      "memory(GiB)": 77.56,
      "step": 28765,
      "token_acc": 0.45188284518828453,
      "train_speed(iter/s)": 1.441262
    },
    {
      "epoch": 1.2325950044985219,
      "grad_norm": 5.109609127044678,
      "learning_rate": 8.57398590272346e-05,
      "loss": 2.3794837951660157,
      "memory(GiB)": 77.56,
      "step": 28770,
      "token_acc": 0.48828125,
      "train_speed(iter/s)": 1.44132
    },
    {
      "epoch": 1.232809219827771,
      "grad_norm": 4.757969379425049,
      "learning_rate": 8.573515235910364e-05,
      "loss": 2.7019912719726564,
      "memory(GiB)": 77.56,
      "step": 28775,
      "token_acc": 0.4194528875379939,
      "train_speed(iter/s)": 1.44133
    },
    {
      "epoch": 1.23302343515702,
      "grad_norm": 4.6256513595581055,
      "learning_rate": 8.573044504359686e-05,
      "loss": 2.5817726135253904,
      "memory(GiB)": 77.56,
      "step": 28780,
      "token_acc": 0.46105919003115264,
      "train_speed(iter/s)": 1.441366
    },
    {
      "epoch": 1.2332376504862688,
      "grad_norm": 4.061574459075928,
      "learning_rate": 8.572573708079954e-05,
      "loss": 2.3653993606567383,
      "memory(GiB)": 77.56,
      "step": 28785,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.441366
    },
    {
      "epoch": 1.2334518658155178,
      "grad_norm": 5.002175807952881,
      "learning_rate": 8.5721028470797e-05,
      "loss": 2.7393707275390624,
      "memory(GiB)": 77.56,
      "step": 28790,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.441382
    },
    {
      "epoch": 1.2336660811447668,
      "grad_norm": 5.107406139373779,
      "learning_rate": 8.571631921367451e-05,
      "loss": 2.3718990325927733,
      "memory(GiB)": 77.56,
      "step": 28795,
      "token_acc": 0.48828125,
      "train_speed(iter/s)": 1.441412
    },
    {
      "epoch": 1.2338802964740156,
      "grad_norm": 6.161802291870117,
      "learning_rate": 8.571160930951738e-05,
      "loss": 2.694923973083496,
      "memory(GiB)": 77.56,
      "step": 28800,
      "token_acc": 0.42813455657492355,
      "train_speed(iter/s)": 1.441482
    },
    {
      "epoch": 1.2340945118032647,
      "grad_norm": 4.458662033081055,
      "learning_rate": 8.570689875841095e-05,
      "loss": 2.5193561553955077,
      "memory(GiB)": 77.56,
      "step": 28805,
      "token_acc": 0.4696969696969697,
      "train_speed(iter/s)": 1.441516
    },
    {
      "epoch": 1.2343087271325137,
      "grad_norm": 8.706774711608887,
      "learning_rate": 8.570218756044058e-05,
      "loss": 2.3342342376708984,
      "memory(GiB)": 77.56,
      "step": 28810,
      "token_acc": 0.5196850393700787,
      "train_speed(iter/s)": 1.441513
    },
    {
      "epoch": 1.2345229424617625,
      "grad_norm": 4.946343898773193,
      "learning_rate": 8.569747571569157e-05,
      "loss": 2.4420707702636717,
      "memory(GiB)": 77.56,
      "step": 28815,
      "token_acc": 0.5197368421052632,
      "train_speed(iter/s)": 1.441491
    },
    {
      "epoch": 1.2347371577910116,
      "grad_norm": 5.19757604598999,
      "learning_rate": 8.569276322424931e-05,
      "loss": 2.3729801177978516,
      "memory(GiB)": 77.56,
      "step": 28820,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.441505
    },
    {
      "epoch": 1.2349513731202606,
      "grad_norm": 4.7357707023620605,
      "learning_rate": 8.568805008619916e-05,
      "loss": 2.4615493774414063,
      "memory(GiB)": 77.56,
      "step": 28825,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.441519
    },
    {
      "epoch": 1.2351655884495094,
      "grad_norm": 6.024968147277832,
      "learning_rate": 8.568333630162649e-05,
      "loss": 2.231813430786133,
      "memory(GiB)": 77.56,
      "step": 28830,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.441516
    },
    {
      "epoch": 1.2353798037787584,
      "grad_norm": 4.668166160583496,
      "learning_rate": 8.567862187061673e-05,
      "loss": 2.5431495666503907,
      "memory(GiB)": 77.56,
      "step": 28835,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.441494
    },
    {
      "epoch": 1.2355940191080075,
      "grad_norm": 5.450082778930664,
      "learning_rate": 8.567390679325525e-05,
      "loss": 2.1610176086425783,
      "memory(GiB)": 77.56,
      "step": 28840,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.441518
    },
    {
      "epoch": 1.2358082344372563,
      "grad_norm": 5.276432037353516,
      "learning_rate": 8.56691910696275e-05,
      "loss": 2.484408950805664,
      "memory(GiB)": 77.56,
      "step": 28845,
      "token_acc": 0.5167095115681234,
      "train_speed(iter/s)": 1.441467
    },
    {
      "epoch": 1.2360224497665053,
      "grad_norm": 5.5432515144348145,
      "learning_rate": 8.566447469981888e-05,
      "loss": 2.7759578704833983,
      "memory(GiB)": 77.56,
      "step": 28850,
      "token_acc": 0.4174174174174174,
      "train_speed(iter/s)": 1.441514
    },
    {
      "epoch": 1.2362366650957544,
      "grad_norm": 7.759845733642578,
      "learning_rate": 8.565975768391484e-05,
      "loss": 2.3879852294921875,
      "memory(GiB)": 77.56,
      "step": 28855,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.441519
    },
    {
      "epoch": 1.2364508804250032,
      "grad_norm": 6.212899684906006,
      "learning_rate": 8.565504002200084e-05,
      "loss": 2.567481803894043,
      "memory(GiB)": 77.56,
      "step": 28860,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.441558
    },
    {
      "epoch": 1.2366650957542522,
      "grad_norm": 5.026437759399414,
      "learning_rate": 8.565032171416236e-05,
      "loss": 2.4668468475341796,
      "memory(GiB)": 77.56,
      "step": 28865,
      "token_acc": 0.46200607902735563,
      "train_speed(iter/s)": 1.441602
    },
    {
      "epoch": 1.2368793110835012,
      "grad_norm": 5.263489723205566,
      "learning_rate": 8.564560276048483e-05,
      "loss": 2.433339309692383,
      "memory(GiB)": 77.56,
      "step": 28870,
      "token_acc": 0.47278911564625853,
      "train_speed(iter/s)": 1.441618
    },
    {
      "epoch": 1.23709352641275,
      "grad_norm": 4.9902801513671875,
      "learning_rate": 8.56408831610538e-05,
      "loss": 2.2896575927734375,
      "memory(GiB)": 77.56,
      "step": 28875,
      "token_acc": 0.45864661654135336,
      "train_speed(iter/s)": 1.441594
    },
    {
      "epoch": 1.237307741741999,
      "grad_norm": 3.9070582389831543,
      "learning_rate": 8.563616291595473e-05,
      "loss": 2.371356964111328,
      "memory(GiB)": 77.56,
      "step": 28880,
      "token_acc": 0.4836795252225519,
      "train_speed(iter/s)": 1.441575
    },
    {
      "epoch": 1.2375219570712481,
      "grad_norm": 4.652554988861084,
      "learning_rate": 8.563144202527312e-05,
      "loss": 2.4942867279052736,
      "memory(GiB)": 77.56,
      "step": 28885,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.441575
    },
    {
      "epoch": 1.237736172400497,
      "grad_norm": 5.192071914672852,
      "learning_rate": 8.562672048909453e-05,
      "loss": 2.924012565612793,
      "memory(GiB)": 77.56,
      "step": 28890,
      "token_acc": 0.42507645259938837,
      "train_speed(iter/s)": 1.44161
    },
    {
      "epoch": 1.237950387729746,
      "grad_norm": 5.600824356079102,
      "learning_rate": 8.562199830750447e-05,
      "loss": 2.4559005737304687,
      "memory(GiB)": 77.56,
      "step": 28895,
      "token_acc": 0.4662576687116564,
      "train_speed(iter/s)": 1.441595
    },
    {
      "epoch": 1.238164603058995,
      "grad_norm": 4.9247517585754395,
      "learning_rate": 8.561727548058849e-05,
      "loss": 2.629568862915039,
      "memory(GiB)": 77.56,
      "step": 28900,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.441647
    },
    {
      "epoch": 1.2383788183882438,
      "grad_norm": 6.085575103759766,
      "learning_rate": 8.561255200843216e-05,
      "loss": 2.50693359375,
      "memory(GiB)": 77.56,
      "step": 28905,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.441647
    },
    {
      "epoch": 1.2385930337174929,
      "grad_norm": 5.520236968994141,
      "learning_rate": 8.560782789112105e-05,
      "loss": 2.460283660888672,
      "memory(GiB)": 77.56,
      "step": 28910,
      "token_acc": 0.46332046332046334,
      "train_speed(iter/s)": 1.441664
    },
    {
      "epoch": 1.2388072490467419,
      "grad_norm": 7.209451675415039,
      "learning_rate": 8.560310312874074e-05,
      "loss": 2.617263603210449,
      "memory(GiB)": 77.56,
      "step": 28915,
      "token_acc": 0.45918367346938777,
      "train_speed(iter/s)": 1.441653
    },
    {
      "epoch": 1.2390214643759907,
      "grad_norm": 4.452922344207764,
      "learning_rate": 8.55983777213768e-05,
      "loss": 2.3573495864868166,
      "memory(GiB)": 77.56,
      "step": 28920,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.44163
    },
    {
      "epoch": 1.2392356797052397,
      "grad_norm": 4.127779483795166,
      "learning_rate": 8.559365166911486e-05,
      "loss": 2.2596084594726564,
      "memory(GiB)": 77.56,
      "step": 28925,
      "token_acc": 0.4983388704318937,
      "train_speed(iter/s)": 1.44164
    },
    {
      "epoch": 1.2394498950344888,
      "grad_norm": 4.5671820640563965,
      "learning_rate": 8.558892497204052e-05,
      "loss": 2.238787078857422,
      "memory(GiB)": 77.56,
      "step": 28930,
      "token_acc": 0.49142857142857144,
      "train_speed(iter/s)": 1.441667
    },
    {
      "epoch": 1.2396641103637376,
      "grad_norm": 5.462334632873535,
      "learning_rate": 8.558419763023944e-05,
      "loss": 2.6892871856689453,
      "memory(GiB)": 77.56,
      "step": 28935,
      "token_acc": 0.43050847457627117,
      "train_speed(iter/s)": 1.441585
    },
    {
      "epoch": 1.2398783256929866,
      "grad_norm": 4.344351291656494,
      "learning_rate": 8.557946964379723e-05,
      "loss": 2.277344512939453,
      "memory(GiB)": 77.56,
      "step": 28940,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441573
    },
    {
      "epoch": 1.2400925410222357,
      "grad_norm": 4.448167324066162,
      "learning_rate": 8.557474101279955e-05,
      "loss": 2.6097690582275392,
      "memory(GiB)": 77.56,
      "step": 28945,
      "token_acc": 0.4376899696048632,
      "train_speed(iter/s)": 1.441542
    },
    {
      "epoch": 1.2403067563514845,
      "grad_norm": 4.940723896026611,
      "learning_rate": 8.557001173733206e-05,
      "loss": 2.3075437545776367,
      "memory(GiB)": 77.56,
      "step": 28950,
      "token_acc": 0.5168918918918919,
      "train_speed(iter/s)": 1.441587
    },
    {
      "epoch": 1.2405209716807335,
      "grad_norm": 4.797855854034424,
      "learning_rate": 8.556528181748044e-05,
      "loss": 2.4329502105712892,
      "memory(GiB)": 77.56,
      "step": 28955,
      "token_acc": 0.47230320699708456,
      "train_speed(iter/s)": 1.44159
    },
    {
      "epoch": 1.2407351870099825,
      "grad_norm": 5.140381813049316,
      "learning_rate": 8.556055125333039e-05,
      "loss": 2.5371442794799806,
      "memory(GiB)": 77.56,
      "step": 28960,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.441592
    },
    {
      "epoch": 1.2409494023392313,
      "grad_norm": 7.099332332611084,
      "learning_rate": 8.555582004496758e-05,
      "loss": 2.264695167541504,
      "memory(GiB)": 77.56,
      "step": 28965,
      "token_acc": 0.4609053497942387,
      "train_speed(iter/s)": 1.441584
    },
    {
      "epoch": 1.2411636176684804,
      "grad_norm": 5.5887980461120605,
      "learning_rate": 8.555108819247774e-05,
      "loss": 2.434825897216797,
      "memory(GiB)": 77.56,
      "step": 28970,
      "token_acc": 0.46794871794871795,
      "train_speed(iter/s)": 1.441543
    },
    {
      "epoch": 1.2413778329977294,
      "grad_norm": 6.623556137084961,
      "learning_rate": 8.55463556959466e-05,
      "loss": 2.302535629272461,
      "memory(GiB)": 77.56,
      "step": 28975,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441469
    },
    {
      "epoch": 1.2415920483269782,
      "grad_norm": 4.901788711547852,
      "learning_rate": 8.554162255545987e-05,
      "loss": 2.586678695678711,
      "memory(GiB)": 77.56,
      "step": 28980,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.441448
    },
    {
      "epoch": 1.2418062636562273,
      "grad_norm": 4.8584303855896,
      "learning_rate": 8.55368887711033e-05,
      "loss": 2.325054740905762,
      "memory(GiB)": 77.56,
      "step": 28985,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.441472
    },
    {
      "epoch": 1.2420204789854763,
      "grad_norm": 4.714451789855957,
      "learning_rate": 8.553215434296268e-05,
      "loss": 2.535818862915039,
      "memory(GiB)": 77.56,
      "step": 28990,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.441525
    },
    {
      "epoch": 1.242234694314725,
      "grad_norm": 8.418222427368164,
      "learning_rate": 8.552741927112373e-05,
      "loss": 2.4478073120117188,
      "memory(GiB)": 77.56,
      "step": 28995,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.441557
    },
    {
      "epoch": 1.2424489096439741,
      "grad_norm": 5.557422161102295,
      "learning_rate": 8.552268355567226e-05,
      "loss": 2.9656723022460936,
      "memory(GiB)": 77.56,
      "step": 29000,
      "token_acc": 0.4041916167664671,
      "train_speed(iter/s)": 1.441547
    },
    {
      "epoch": 1.2424489096439741,
      "eval_loss": 2.2247400283813477,
      "eval_runtime": 13.7495,
      "eval_samples_per_second": 7.273,
      "eval_steps_per_second": 7.273,
      "eval_token_acc": 0.4855072463768116,
      "step": 29000
    },
    {
      "epoch": 1.2426631249732232,
      "grad_norm": 5.335652828216553,
      "learning_rate": 8.551794719669405e-05,
      "loss": 2.480691337585449,
      "memory(GiB)": 77.56,
      "step": 29005,
      "token_acc": 0.48810754912099275,
      "train_speed(iter/s)": 1.44048
    },
    {
      "epoch": 1.242877340302472,
      "grad_norm": 6.0281267166137695,
      "learning_rate": 8.551321019427491e-05,
      "loss": 2.332222747802734,
      "memory(GiB)": 77.56,
      "step": 29010,
      "token_acc": 0.5261194029850746,
      "train_speed(iter/s)": 1.440465
    },
    {
      "epoch": 1.243091555631721,
      "grad_norm": 4.924759864807129,
      "learning_rate": 8.550847254850065e-05,
      "loss": 2.4824304580688477,
      "memory(GiB)": 77.56,
      "step": 29015,
      "token_acc": 0.4774774774774775,
      "train_speed(iter/s)": 1.440537
    },
    {
      "epoch": 1.24330577096097,
      "grad_norm": 4.872472763061523,
      "learning_rate": 8.550373425945711e-05,
      "loss": 2.276318359375,
      "memory(GiB)": 77.56,
      "step": 29020,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.440624
    },
    {
      "epoch": 1.2435199862902189,
      "grad_norm": 5.124520778656006,
      "learning_rate": 8.549899532723011e-05,
      "loss": 2.8170703887939452,
      "memory(GiB)": 77.56,
      "step": 29025,
      "token_acc": 0.4139072847682119,
      "train_speed(iter/s)": 1.440666
    },
    {
      "epoch": 1.243734201619468,
      "grad_norm": 4.6610002517700195,
      "learning_rate": 8.549425575190551e-05,
      "loss": 2.350818634033203,
      "memory(GiB)": 77.56,
      "step": 29030,
      "token_acc": 0.501628664495114,
      "train_speed(iter/s)": 1.44067
    },
    {
      "epoch": 1.243948416948717,
      "grad_norm": 5.190739154815674,
      "learning_rate": 8.548951553356917e-05,
      "loss": 2.732608413696289,
      "memory(GiB)": 77.56,
      "step": 29035,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.440715
    },
    {
      "epoch": 1.2441626322779658,
      "grad_norm": 4.973451137542725,
      "learning_rate": 8.548477467230697e-05,
      "loss": 2.4846229553222656,
      "memory(GiB)": 77.56,
      "step": 29040,
      "token_acc": 0.47244094488188976,
      "train_speed(iter/s)": 1.440732
    },
    {
      "epoch": 1.2443768476072148,
      "grad_norm": 5.122106552124023,
      "learning_rate": 8.548003316820478e-05,
      "loss": 2.3691707611083985,
      "memory(GiB)": 77.56,
      "step": 29045,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.440761
    },
    {
      "epoch": 1.2445910629364638,
      "grad_norm": 3.669872999191284,
      "learning_rate": 8.547529102134852e-05,
      "loss": 2.5579437255859374,
      "memory(GiB)": 77.56,
      "step": 29050,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.440766
    },
    {
      "epoch": 1.2448052782657126,
      "grad_norm": 5.317296981811523,
      "learning_rate": 8.547054823182408e-05,
      "loss": 2.6559097290039064,
      "memory(GiB)": 77.56,
      "step": 29055,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.440708
    },
    {
      "epoch": 1.2450194935949617,
      "grad_norm": 4.863055229187012,
      "learning_rate": 8.546580479971737e-05,
      "loss": 2.473911666870117,
      "memory(GiB)": 77.56,
      "step": 29060,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.440689
    },
    {
      "epoch": 1.2452337089242107,
      "grad_norm": 4.805677890777588,
      "learning_rate": 8.546106072511435e-05,
      "loss": 2.504608154296875,
      "memory(GiB)": 77.56,
      "step": 29065,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.440693
    },
    {
      "epoch": 1.2454479242534595,
      "grad_norm": 4.455987453460693,
      "learning_rate": 8.545631600810094e-05,
      "loss": 2.338959503173828,
      "memory(GiB)": 77.56,
      "step": 29070,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.440667
    },
    {
      "epoch": 1.2456621395827085,
      "grad_norm": 4.937370777130127,
      "learning_rate": 8.545157064876311e-05,
      "loss": 2.693893814086914,
      "memory(GiB)": 77.56,
      "step": 29075,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.440647
    },
    {
      "epoch": 1.2458763549119576,
      "grad_norm": 4.100717544555664,
      "learning_rate": 8.544682464718684e-05,
      "loss": 2.204322624206543,
      "memory(GiB)": 77.56,
      "step": 29080,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.440619
    },
    {
      "epoch": 1.2460905702412064,
      "grad_norm": 5.114598751068115,
      "learning_rate": 8.544207800345808e-05,
      "loss": 2.542263221740723,
      "memory(GiB)": 77.56,
      "step": 29085,
      "token_acc": 0.45625,
      "train_speed(iter/s)": 1.440653
    },
    {
      "epoch": 1.2463047855704554,
      "grad_norm": 5.612907409667969,
      "learning_rate": 8.54373307176628e-05,
      "loss": 2.358996772766113,
      "memory(GiB)": 77.56,
      "step": 29090,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.440663
    },
    {
      "epoch": 1.2465190008997045,
      "grad_norm": 4.823172092437744,
      "learning_rate": 8.543258278988706e-05,
      "loss": 2.5500766754150392,
      "memory(GiB)": 77.56,
      "step": 29095,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.44062
    },
    {
      "epoch": 1.2467332162289533,
      "grad_norm": 5.943562030792236,
      "learning_rate": 8.542783422021684e-05,
      "loss": 2.4312461853027343,
      "memory(GiB)": 77.56,
      "step": 29100,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.44066
    },
    {
      "epoch": 1.2469474315582023,
      "grad_norm": 6.189527988433838,
      "learning_rate": 8.542308500873817e-05,
      "loss": 3.0052860260009764,
      "memory(GiB)": 77.56,
      "step": 29105,
      "token_acc": 0.41472868217054265,
      "train_speed(iter/s)": 1.440697
    },
    {
      "epoch": 1.2471616468874513,
      "grad_norm": 4.743948936462402,
      "learning_rate": 8.541833515553707e-05,
      "loss": 2.497550201416016,
      "memory(GiB)": 77.56,
      "step": 29110,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.44067
    },
    {
      "epoch": 1.2473758622167002,
      "grad_norm": 6.074448585510254,
      "learning_rate": 8.541358466069962e-05,
      "loss": 2.5804285049438476,
      "memory(GiB)": 77.56,
      "step": 29115,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440696
    },
    {
      "epoch": 1.2475900775459492,
      "grad_norm": 5.792123317718506,
      "learning_rate": 8.540883352431186e-05,
      "loss": 2.6468175888061523,
      "memory(GiB)": 77.56,
      "step": 29120,
      "token_acc": 0.43636363636363634,
      "train_speed(iter/s)": 1.440708
    },
    {
      "epoch": 1.2478042928751982,
      "grad_norm": 5.768387317657471,
      "learning_rate": 8.540408174645986e-05,
      "loss": 2.383892059326172,
      "memory(GiB)": 77.56,
      "step": 29125,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.440733
    },
    {
      "epoch": 1.248018508204447,
      "grad_norm": 4.049323558807373,
      "learning_rate": 8.539932932722971e-05,
      "loss": 2.5525436401367188,
      "memory(GiB)": 77.56,
      "step": 29130,
      "token_acc": 0.4576923076923077,
      "train_speed(iter/s)": 1.440699
    },
    {
      "epoch": 1.248232723533696,
      "grad_norm": 4.429920673370361,
      "learning_rate": 8.539457626670752e-05,
      "loss": 2.5024272918701174,
      "memory(GiB)": 77.56,
      "step": 29135,
      "token_acc": 0.5034965034965035,
      "train_speed(iter/s)": 1.440677
    },
    {
      "epoch": 1.248446938862945,
      "grad_norm": 4.3035101890563965,
      "learning_rate": 8.538982256497937e-05,
      "loss": 2.5994089126586912,
      "memory(GiB)": 77.56,
      "step": 29140,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440682
    },
    {
      "epoch": 1.248661154192194,
      "grad_norm": 5.251198768615723,
      "learning_rate": 8.538506822213136e-05,
      "loss": 2.7291500091552736,
      "memory(GiB)": 77.56,
      "step": 29145,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.440688
    },
    {
      "epoch": 1.248875369521443,
      "grad_norm": 4.390237808227539,
      "learning_rate": 8.538031323824967e-05,
      "loss": 2.503163719177246,
      "memory(GiB)": 77.56,
      "step": 29150,
      "token_acc": 0.46062992125984253,
      "train_speed(iter/s)": 1.440744
    },
    {
      "epoch": 1.249089584850692,
      "grad_norm": 5.376598834991455,
      "learning_rate": 8.537555761342039e-05,
      "loss": 2.403537368774414,
      "memory(GiB)": 77.56,
      "step": 29155,
      "token_acc": 0.49794238683127573,
      "train_speed(iter/s)": 1.440717
    },
    {
      "epoch": 1.2493038001799408,
      "grad_norm": 5.448025226593018,
      "learning_rate": 8.537080134772973e-05,
      "loss": 2.554519462585449,
      "memory(GiB)": 77.56,
      "step": 29160,
      "token_acc": 0.5083612040133779,
      "train_speed(iter/s)": 1.440726
    },
    {
      "epoch": 1.2495180155091898,
      "grad_norm": 5.245378494262695,
      "learning_rate": 8.536604444126382e-05,
      "loss": 2.830800247192383,
      "memory(GiB)": 77.56,
      "step": 29165,
      "token_acc": 0.4597014925373134,
      "train_speed(iter/s)": 1.440802
    },
    {
      "epoch": 1.2497322308384389,
      "grad_norm": 5.468711853027344,
      "learning_rate": 8.536128689410882e-05,
      "loss": 2.648890495300293,
      "memory(GiB)": 77.56,
      "step": 29170,
      "token_acc": 0.4701492537313433,
      "train_speed(iter/s)": 1.440833
    },
    {
      "epoch": 1.2499464461676877,
      "grad_norm": 32.12142562866211,
      "learning_rate": 8.535652870635094e-05,
      "loss": 2.5189035415649412,
      "memory(GiB)": 77.56,
      "step": 29175,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.440788
    },
    {
      "epoch": 1.2501606614969367,
      "grad_norm": 6.855515956878662,
      "learning_rate": 8.535176987807639e-05,
      "loss": 2.553010368347168,
      "memory(GiB)": 77.56,
      "step": 29180,
      "token_acc": 0.5110294117647058,
      "train_speed(iter/s)": 1.440789
    },
    {
      "epoch": 1.2503748768261858,
      "grad_norm": 3.473304271697998,
      "learning_rate": 8.534701040937136e-05,
      "loss": 2.285348320007324,
      "memory(GiB)": 77.56,
      "step": 29185,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.440789
    },
    {
      "epoch": 1.2505890921554346,
      "grad_norm": 5.68897819519043,
      "learning_rate": 8.534225030032208e-05,
      "loss": 2.392068862915039,
      "memory(GiB)": 77.56,
      "step": 29190,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.440777
    },
    {
      "epoch": 1.2508033074846836,
      "grad_norm": 7.997988224029541,
      "learning_rate": 8.533748955101477e-05,
      "loss": 2.3846378326416016,
      "memory(GiB)": 77.56,
      "step": 29195,
      "token_acc": 0.48034934497816595,
      "train_speed(iter/s)": 1.440811
    },
    {
      "epoch": 1.2510175228139326,
      "grad_norm": 4.5932841300964355,
      "learning_rate": 8.533272816153571e-05,
      "loss": 2.4555004119873045,
      "memory(GiB)": 77.56,
      "step": 29200,
      "token_acc": 0.4696969696969697,
      "train_speed(iter/s)": 1.44087
    },
    {
      "epoch": 1.2512317381431814,
      "grad_norm": 5.046009540557861,
      "learning_rate": 8.532796613197112e-05,
      "loss": 2.460248565673828,
      "memory(GiB)": 77.56,
      "step": 29205,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.440881
    },
    {
      "epoch": 1.2514459534724305,
      "grad_norm": 4.072773456573486,
      "learning_rate": 8.532320346240728e-05,
      "loss": 2.6071218490600585,
      "memory(GiB)": 77.56,
      "step": 29210,
      "token_acc": 0.4228571428571429,
      "train_speed(iter/s)": 1.440893
    },
    {
      "epoch": 1.2516601688016795,
      "grad_norm": 3.9104371070861816,
      "learning_rate": 8.531844015293047e-05,
      "loss": 2.4960987091064455,
      "memory(GiB)": 77.56,
      "step": 29215,
      "token_acc": 0.45980707395498394,
      "train_speed(iter/s)": 1.440927
    },
    {
      "epoch": 1.2518743841309283,
      "grad_norm": 5.324708938598633,
      "learning_rate": 8.531367620362699e-05,
      "loss": 2.3529176712036133,
      "memory(GiB)": 77.56,
      "step": 29220,
      "token_acc": 0.5258964143426295,
      "train_speed(iter/s)": 1.440929
    },
    {
      "epoch": 1.2520885994601774,
      "grad_norm": 5.784875869750977,
      "learning_rate": 8.530891161458314e-05,
      "loss": 2.5704120635986327,
      "memory(GiB)": 77.56,
      "step": 29225,
      "token_acc": 0.4708029197080292,
      "train_speed(iter/s)": 1.440974
    },
    {
      "epoch": 1.2523028147894264,
      "grad_norm": 4.044406890869141,
      "learning_rate": 8.530414638588525e-05,
      "loss": 2.2505264282226562,
      "memory(GiB)": 77.56,
      "step": 29230,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.441012
    },
    {
      "epoch": 1.2525170301186752,
      "grad_norm": 6.830445289611816,
      "learning_rate": 8.529938051761961e-05,
      "loss": 2.8425745010375976,
      "memory(GiB)": 77.56,
      "step": 29235,
      "token_acc": 0.4155844155844156,
      "train_speed(iter/s)": 1.441034
    },
    {
      "epoch": 1.2527312454479242,
      "grad_norm": 5.800511360168457,
      "learning_rate": 8.529461400987258e-05,
      "loss": 2.721932601928711,
      "memory(GiB)": 77.56,
      "step": 29240,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.44108
    },
    {
      "epoch": 1.2529454607771733,
      "grad_norm": 6.483523845672607,
      "learning_rate": 8.528984686273051e-05,
      "loss": 2.186646270751953,
      "memory(GiB)": 77.56,
      "step": 29245,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.441125
    },
    {
      "epoch": 1.253159676106422,
      "grad_norm": 5.5024094581604,
      "learning_rate": 8.528507907627977e-05,
      "loss": 2.644719123840332,
      "memory(GiB)": 77.56,
      "step": 29250,
      "token_acc": 0.48639455782312924,
      "train_speed(iter/s)": 1.441102
    },
    {
      "epoch": 1.2533738914356711,
      "grad_norm": 4.80596923828125,
      "learning_rate": 8.528031065060669e-05,
      "loss": 2.0365554809570314,
      "memory(GiB)": 77.56,
      "step": 29255,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.441125
    },
    {
      "epoch": 1.2535881067649202,
      "grad_norm": 4.627771377563477,
      "learning_rate": 8.527554158579772e-05,
      "loss": 2.210123825073242,
      "memory(GiB)": 77.56,
      "step": 29260,
      "token_acc": 0.5145228215767634,
      "train_speed(iter/s)": 1.441147
    },
    {
      "epoch": 1.253802322094169,
      "grad_norm": 5.042235851287842,
      "learning_rate": 8.527077188193921e-05,
      "loss": 2.6837717056274415,
      "memory(GiB)": 77.56,
      "step": 29265,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.441218
    },
    {
      "epoch": 1.254016537423418,
      "grad_norm": 6.353658199310303,
      "learning_rate": 8.52660015391176e-05,
      "loss": 2.2099325180053713,
      "memory(GiB)": 77.56,
      "step": 29270,
      "token_acc": 0.516260162601626,
      "train_speed(iter/s)": 1.441262
    },
    {
      "epoch": 1.254230752752667,
      "grad_norm": 5.837703227996826,
      "learning_rate": 8.526123055741926e-05,
      "loss": 2.3983953475952147,
      "memory(GiB)": 77.56,
      "step": 29275,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.44126
    },
    {
      "epoch": 1.2544449680819159,
      "grad_norm": 4.5204315185546875,
      "learning_rate": 8.525645893693067e-05,
      "loss": 2.6883926391601562,
      "memory(GiB)": 77.56,
      "step": 29280,
      "token_acc": 0.4725609756097561,
      "train_speed(iter/s)": 1.441311
    },
    {
      "epoch": 1.254659183411165,
      "grad_norm": 5.766561508178711,
      "learning_rate": 8.525168667773824e-05,
      "loss": 2.4652259826660154,
      "memory(GiB)": 77.56,
      "step": 29285,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.441352
    },
    {
      "epoch": 1.254873398740414,
      "grad_norm": 5.017345428466797,
      "learning_rate": 8.524691377992844e-05,
      "loss": 2.6708446502685548,
      "memory(GiB)": 77.56,
      "step": 29290,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.441361
    },
    {
      "epoch": 1.2550876140696627,
      "grad_norm": 7.347275733947754,
      "learning_rate": 8.524214024358775e-05,
      "loss": 2.7507781982421875,
      "memory(GiB)": 77.56,
      "step": 29295,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441395
    },
    {
      "epoch": 1.2553018293989118,
      "grad_norm": 4.259577751159668,
      "learning_rate": 8.523736606880261e-05,
      "loss": 2.4214689254760744,
      "memory(GiB)": 77.56,
      "step": 29300,
      "token_acc": 0.5410447761194029,
      "train_speed(iter/s)": 1.441437
    },
    {
      "epoch": 1.2555160447281608,
      "grad_norm": 4.1819610595703125,
      "learning_rate": 8.523259125565954e-05,
      "loss": 2.726064682006836,
      "memory(GiB)": 77.56,
      "step": 29305,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.441436
    },
    {
      "epoch": 1.2557302600574096,
      "grad_norm": 5.189210891723633,
      "learning_rate": 8.522781580424502e-05,
      "loss": 2.473117637634277,
      "memory(GiB)": 77.56,
      "step": 29310,
      "token_acc": 0.5181159420289855,
      "train_speed(iter/s)": 1.441456
    },
    {
      "epoch": 1.2559444753866587,
      "grad_norm": 5.402461528778076,
      "learning_rate": 8.522303971464557e-05,
      "loss": 2.6044374465942384,
      "memory(GiB)": 77.56,
      "step": 29315,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.441478
    },
    {
      "epoch": 1.2561586907159077,
      "grad_norm": 5.355550289154053,
      "learning_rate": 8.521826298694773e-05,
      "loss": 2.649668884277344,
      "memory(GiB)": 77.56,
      "step": 29320,
      "token_acc": 0.4459016393442623,
      "train_speed(iter/s)": 1.441474
    },
    {
      "epoch": 1.2563729060451565,
      "grad_norm": 5.541404724121094,
      "learning_rate": 8.5213485621238e-05,
      "loss": 2.808111000061035,
      "memory(GiB)": 77.56,
      "step": 29325,
      "token_acc": 0.4798206278026906,
      "train_speed(iter/s)": 1.441497
    },
    {
      "epoch": 1.2565871213744055,
      "grad_norm": 6.287618637084961,
      "learning_rate": 8.520870761760296e-05,
      "loss": 2.5784027099609377,
      "memory(GiB)": 77.56,
      "step": 29330,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.4415
    },
    {
      "epoch": 1.2568013367036546,
      "grad_norm": 4.258848667144775,
      "learning_rate": 8.520392897612913e-05,
      "loss": 2.4917228698730467,
      "memory(GiB)": 77.56,
      "step": 29335,
      "token_acc": 0.4879032258064516,
      "train_speed(iter/s)": 1.441495
    },
    {
      "epoch": 1.2570155520329034,
      "grad_norm": 3.6803226470947266,
      "learning_rate": 8.519914969690313e-05,
      "loss": 2.457961082458496,
      "memory(GiB)": 77.56,
      "step": 29340,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441541
    },
    {
      "epoch": 1.2572297673621524,
      "grad_norm": 6.132785320281982,
      "learning_rate": 8.51943697800115e-05,
      "loss": 2.4769752502441404,
      "memory(GiB)": 77.56,
      "step": 29345,
      "token_acc": 0.4708333333333333,
      "train_speed(iter/s)": 1.441452
    },
    {
      "epoch": 1.2574439826914015,
      "grad_norm": 4.694347381591797,
      "learning_rate": 8.518958922554085e-05,
      "loss": 2.442934036254883,
      "memory(GiB)": 77.56,
      "step": 29350,
      "token_acc": 0.5031446540880503,
      "train_speed(iter/s)": 1.441451
    },
    {
      "epoch": 1.2576581980206503,
      "grad_norm": 5.689286708831787,
      "learning_rate": 8.518480803357778e-05,
      "loss": 2.617264175415039,
      "memory(GiB)": 77.56,
      "step": 29355,
      "token_acc": 0.5119453924914675,
      "train_speed(iter/s)": 1.441407
    },
    {
      "epoch": 1.2578724133498993,
      "grad_norm": 4.608502388000488,
      "learning_rate": 8.51800262042089e-05,
      "loss": 2.3587196350097654,
      "memory(GiB)": 77.56,
      "step": 29360,
      "token_acc": 0.47674418604651164,
      "train_speed(iter/s)": 1.441402
    },
    {
      "epoch": 1.2580866286791483,
      "grad_norm": 6.740025997161865,
      "learning_rate": 8.517524373752083e-05,
      "loss": 2.3277217864990236,
      "memory(GiB)": 77.56,
      "step": 29365,
      "token_acc": 0.45867768595041325,
      "train_speed(iter/s)": 1.44141
    },
    {
      "epoch": 1.2583008440083971,
      "grad_norm": 4.192903518676758,
      "learning_rate": 8.517046063360026e-05,
      "loss": 2.3574275970458984,
      "memory(GiB)": 77.56,
      "step": 29370,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.441426
    },
    {
      "epoch": 1.2585150593376462,
      "grad_norm": 5.203413009643555,
      "learning_rate": 8.516567689253378e-05,
      "loss": 2.327602767944336,
      "memory(GiB)": 77.56,
      "step": 29375,
      "token_acc": 0.528957528957529,
      "train_speed(iter/s)": 1.441448
    },
    {
      "epoch": 1.2587292746668952,
      "grad_norm": 5.421355724334717,
      "learning_rate": 8.516089251440809e-05,
      "loss": 2.5142133712768553,
      "memory(GiB)": 77.56,
      "step": 29380,
      "token_acc": 0.465625,
      "train_speed(iter/s)": 1.44142
    },
    {
      "epoch": 1.258943489996144,
      "grad_norm": 4.307124137878418,
      "learning_rate": 8.515610749930982e-05,
      "loss": 2.445865821838379,
      "memory(GiB)": 77.56,
      "step": 29385,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.44146
    },
    {
      "epoch": 1.259157705325393,
      "grad_norm": 3.540876626968384,
      "learning_rate": 8.51513218473257e-05,
      "loss": 2.5506105422973633,
      "memory(GiB)": 77.56,
      "step": 29390,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.441413
    },
    {
      "epoch": 1.259371920654642,
      "grad_norm": 4.286125659942627,
      "learning_rate": 8.514653555854242e-05,
      "loss": 2.609775161743164,
      "memory(GiB)": 77.56,
      "step": 29395,
      "token_acc": 0.4674922600619195,
      "train_speed(iter/s)": 1.441396
    },
    {
      "epoch": 1.259586135983891,
      "grad_norm": 5.016783237457275,
      "learning_rate": 8.514174863304667e-05,
      "loss": 2.781946563720703,
      "memory(GiB)": 77.56,
      "step": 29400,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.441388
    },
    {
      "epoch": 1.25980035131314,
      "grad_norm": 3.8409063816070557,
      "learning_rate": 8.513696107092517e-05,
      "loss": 2.5462841033935546,
      "memory(GiB)": 77.56,
      "step": 29405,
      "token_acc": 0.4720496894409938,
      "train_speed(iter/s)": 1.441419
    },
    {
      "epoch": 1.260014566642389,
      "grad_norm": 5.7038421630859375,
      "learning_rate": 8.513217287226466e-05,
      "loss": 2.49387149810791,
      "memory(GiB)": 77.56,
      "step": 29410,
      "token_acc": 0.5036231884057971,
      "train_speed(iter/s)": 1.441448
    },
    {
      "epoch": 1.2602287819716378,
      "grad_norm": 4.566816329956055,
      "learning_rate": 8.51273840371519e-05,
      "loss": 2.2985652923583983,
      "memory(GiB)": 77.56,
      "step": 29415,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.441456
    },
    {
      "epoch": 1.2604429973008868,
      "grad_norm": 4.351600646972656,
      "learning_rate": 8.512259456567362e-05,
      "loss": 2.575051116943359,
      "memory(GiB)": 77.56,
      "step": 29420,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441473
    },
    {
      "epoch": 1.2606572126301359,
      "grad_norm": 6.6223978996276855,
      "learning_rate": 8.511780445791659e-05,
      "loss": 2.7139156341552733,
      "memory(GiB)": 77.56,
      "step": 29425,
      "token_acc": 0.43214285714285716,
      "train_speed(iter/s)": 1.441527
    },
    {
      "epoch": 1.2608714279593847,
      "grad_norm": 7.783810138702393,
      "learning_rate": 8.511301371396758e-05,
      "loss": 2.307796859741211,
      "memory(GiB)": 77.56,
      "step": 29430,
      "token_acc": 0.4939759036144578,
      "train_speed(iter/s)": 1.441568
    },
    {
      "epoch": 1.2610856432886337,
      "grad_norm": 4.968101501464844,
      "learning_rate": 8.510822233391338e-05,
      "loss": 2.2231395721435545,
      "memory(GiB)": 77.56,
      "step": 29435,
      "token_acc": 0.5125448028673835,
      "train_speed(iter/s)": 1.441603
    },
    {
      "epoch": 1.2612998586178827,
      "grad_norm": 5.41532564163208,
      "learning_rate": 8.510343031784082e-05,
      "loss": 2.586380195617676,
      "memory(GiB)": 77.56,
      "step": 29440,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.441635
    },
    {
      "epoch": 1.2615140739471316,
      "grad_norm": 4.374895095825195,
      "learning_rate": 8.509863766583668e-05,
      "loss": 2.7054927825927733,
      "memory(GiB)": 77.56,
      "step": 29445,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.441667
    },
    {
      "epoch": 1.2617282892763806,
      "grad_norm": 4.795865535736084,
      "learning_rate": 8.50938443779878e-05,
      "loss": 2.680415153503418,
      "memory(GiB)": 77.56,
      "step": 29450,
      "token_acc": 0.43171806167400884,
      "train_speed(iter/s)": 1.441669
    },
    {
      "epoch": 1.2619425046056296,
      "grad_norm": 4.2429962158203125,
      "learning_rate": 8.5089050454381e-05,
      "loss": 2.6380331039428713,
      "memory(GiB)": 77.56,
      "step": 29455,
      "token_acc": 0.46107784431137727,
      "train_speed(iter/s)": 1.441721
    },
    {
      "epoch": 1.2621567199348784,
      "grad_norm": 4.403748512268066,
      "learning_rate": 8.508425589510314e-05,
      "loss": 2.669622039794922,
      "memory(GiB)": 77.56,
      "step": 29460,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.441747
    },
    {
      "epoch": 1.2623709352641275,
      "grad_norm": 4.5235443115234375,
      "learning_rate": 8.507946070024108e-05,
      "loss": 2.8870777130126952,
      "memory(GiB)": 77.56,
      "step": 29465,
      "token_acc": 0.39655172413793105,
      "train_speed(iter/s)": 1.441781
    },
    {
      "epoch": 1.2625851505933765,
      "grad_norm": 5.992175579071045,
      "learning_rate": 8.507466486988168e-05,
      "loss": 2.6534969329833986,
      "memory(GiB)": 77.56,
      "step": 29470,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.441802
    },
    {
      "epoch": 1.2627993659226253,
      "grad_norm": 5.435794353485107,
      "learning_rate": 8.50698684041118e-05,
      "loss": 2.5746421813964844,
      "memory(GiB)": 77.56,
      "step": 29475,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.441806
    },
    {
      "epoch": 1.2630135812518744,
      "grad_norm": 4.327826023101807,
      "learning_rate": 8.506507130301837e-05,
      "loss": 2.605919075012207,
      "memory(GiB)": 77.56,
      "step": 29480,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.441764
    },
    {
      "epoch": 1.2632277965811234,
      "grad_norm": 4.611506938934326,
      "learning_rate": 8.50602735666883e-05,
      "loss": 2.632436752319336,
      "memory(GiB)": 77.56,
      "step": 29485,
      "token_acc": 0.4251497005988024,
      "train_speed(iter/s)": 1.44178
    },
    {
      "epoch": 1.2634420119103722,
      "grad_norm": 4.321711540222168,
      "learning_rate": 8.505547519520845e-05,
      "loss": 2.5482723236083986,
      "memory(GiB)": 77.56,
      "step": 29490,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.44182
    },
    {
      "epoch": 1.2636562272396212,
      "grad_norm": 4.042707920074463,
      "learning_rate": 8.50506761886658e-05,
      "loss": 2.3914003372192383,
      "memory(GiB)": 77.56,
      "step": 29495,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.441796
    },
    {
      "epoch": 1.2638704425688703,
      "grad_norm": 4.777688503265381,
      "learning_rate": 8.504587654714727e-05,
      "loss": 2.768147277832031,
      "memory(GiB)": 77.56,
      "step": 29500,
      "token_acc": 0.4296028880866426,
      "train_speed(iter/s)": 1.441835
    },
    {
      "epoch": 1.2638704425688703,
      "eval_loss": 2.2530200481414795,
      "eval_runtime": 14.0803,
      "eval_samples_per_second": 7.102,
      "eval_steps_per_second": 7.102,
      "eval_token_acc": 0.5006765899864682,
      "step": 29500
    },
    {
      "epoch": 1.264084657898119,
      "grad_norm": 5.182931423187256,
      "learning_rate": 8.50410762707398e-05,
      "loss": 2.4805917739868164,
      "memory(GiB)": 77.56,
      "step": 29505,
      "token_acc": 0.4851024208566108,
      "train_speed(iter/s)": 1.440771
    },
    {
      "epoch": 1.2642988732273681,
      "grad_norm": 18.61556053161621,
      "learning_rate": 8.503627535953039e-05,
      "loss": 2.5214603424072264,
      "memory(GiB)": 77.56,
      "step": 29510,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.440787
    },
    {
      "epoch": 1.2645130885566171,
      "grad_norm": 5.857314586639404,
      "learning_rate": 8.503147381360595e-05,
      "loss": 2.8611879348754883,
      "memory(GiB)": 77.56,
      "step": 29515,
      "token_acc": 0.42244224422442245,
      "train_speed(iter/s)": 1.440788
    },
    {
      "epoch": 1.264727303885866,
      "grad_norm": 4.183280944824219,
      "learning_rate": 8.502667163305353e-05,
      "loss": 2.587234306335449,
      "memory(GiB)": 77.56,
      "step": 29520,
      "token_acc": 0.41317365269461076,
      "train_speed(iter/s)": 1.440824
    },
    {
      "epoch": 1.264941519215115,
      "grad_norm": 4.637208938598633,
      "learning_rate": 8.502186881796008e-05,
      "loss": 2.1354629516601564,
      "memory(GiB)": 77.56,
      "step": 29525,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.440803
    },
    {
      "epoch": 1.265155734544364,
      "grad_norm": 3.8511526584625244,
      "learning_rate": 8.501706536841263e-05,
      "loss": 2.7837804794311523,
      "memory(GiB)": 77.56,
      "step": 29530,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.440849
    },
    {
      "epoch": 1.2653699498736128,
      "grad_norm": 4.983039855957031,
      "learning_rate": 8.501226128449818e-05,
      "loss": 2.6234146118164063,
      "memory(GiB)": 77.56,
      "step": 29535,
      "token_acc": 0.4757834757834758,
      "train_speed(iter/s)": 1.440798
    },
    {
      "epoch": 1.2655841652028619,
      "grad_norm": 5.084944248199463,
      "learning_rate": 8.500745656630379e-05,
      "loss": 2.774045944213867,
      "memory(GiB)": 77.56,
      "step": 29540,
      "token_acc": 0.43252595155709345,
      "train_speed(iter/s)": 1.440785
    },
    {
      "epoch": 1.265798380532111,
      "grad_norm": 3.617826461791992,
      "learning_rate": 8.500265121391649e-05,
      "loss": 2.4186511993408204,
      "memory(GiB)": 77.56,
      "step": 29545,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.440786
    },
    {
      "epoch": 1.2660125958613597,
      "grad_norm": 4.013150215148926,
      "learning_rate": 8.499784522742329e-05,
      "loss": 2.6996299743652346,
      "memory(GiB)": 77.56,
      "step": 29550,
      "token_acc": 0.46273291925465837,
      "train_speed(iter/s)": 1.440808
    },
    {
      "epoch": 1.2662268111906088,
      "grad_norm": 5.148481369018555,
      "learning_rate": 8.499303860691132e-05,
      "loss": 2.5028945922851564,
      "memory(GiB)": 77.56,
      "step": 29555,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.440844
    },
    {
      "epoch": 1.2664410265198578,
      "grad_norm": 4.05296516418457,
      "learning_rate": 8.498823135246763e-05,
      "loss": 2.5566755294799806,
      "memory(GiB)": 77.56,
      "step": 29560,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.440865
    },
    {
      "epoch": 1.2666552418491066,
      "grad_norm": 5.525784015655518,
      "learning_rate": 8.49834234641793e-05,
      "loss": 2.2289154052734377,
      "memory(GiB)": 77.56,
      "step": 29565,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.440906
    },
    {
      "epoch": 1.2668694571783556,
      "grad_norm": 4.3178887367248535,
      "learning_rate": 8.497861494213344e-05,
      "loss": 2.4543731689453123,
      "memory(GiB)": 77.56,
      "step": 29570,
      "token_acc": 0.5318352059925093,
      "train_speed(iter/s)": 1.440951
    },
    {
      "epoch": 1.2670836725076047,
      "grad_norm": 5.241393089294434,
      "learning_rate": 8.497380578641717e-05,
      "loss": 2.5240528106689455,
      "memory(GiB)": 77.56,
      "step": 29575,
      "token_acc": 0.49044585987261147,
      "train_speed(iter/s)": 1.440992
    },
    {
      "epoch": 1.2672978878368535,
      "grad_norm": 10.083868026733398,
      "learning_rate": 8.496899599711759e-05,
      "loss": 2.7484014511108397,
      "memory(GiB)": 77.56,
      "step": 29580,
      "token_acc": 0.45768025078369906,
      "train_speed(iter/s)": 1.440965
    },
    {
      "epoch": 1.2675121031661025,
      "grad_norm": 7.475610256195068,
      "learning_rate": 8.496418557432183e-05,
      "loss": 2.37349796295166,
      "memory(GiB)": 77.56,
      "step": 29585,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440982
    },
    {
      "epoch": 1.2677263184953516,
      "grad_norm": 5.21260404586792,
      "learning_rate": 8.495937451811706e-05,
      "loss": 3.0564125061035154,
      "memory(GiB)": 77.56,
      "step": 29590,
      "token_acc": 0.40625,
      "train_speed(iter/s)": 1.440962
    },
    {
      "epoch": 1.2679405338246004,
      "grad_norm": 6.1735148429870605,
      "learning_rate": 8.495456282859043e-05,
      "loss": 2.9267141342163088,
      "memory(GiB)": 77.56,
      "step": 29595,
      "token_acc": 0.40059347181008903,
      "train_speed(iter/s)": 1.441033
    },
    {
      "epoch": 1.2681547491538494,
      "grad_norm": 5.1847968101501465,
      "learning_rate": 8.494975050582909e-05,
      "loss": 2.7431713104248048,
      "memory(GiB)": 77.56,
      "step": 29600,
      "token_acc": 0.4560669456066946,
      "train_speed(iter/s)": 1.441048
    },
    {
      "epoch": 1.2683689644830984,
      "grad_norm": 4.126093864440918,
      "learning_rate": 8.494493754992026e-05,
      "loss": 2.567129135131836,
      "memory(GiB)": 77.56,
      "step": 29605,
      "token_acc": 0.45517241379310347,
      "train_speed(iter/s)": 1.44111
    },
    {
      "epoch": 1.2685831798123473,
      "grad_norm": 4.953516006469727,
      "learning_rate": 8.49401239609511e-05,
      "loss": 2.479683685302734,
      "memory(GiB)": 77.56,
      "step": 29610,
      "token_acc": 0.4766355140186916,
      "train_speed(iter/s)": 1.441084
    },
    {
      "epoch": 1.2687973951415963,
      "grad_norm": 5.266232490539551,
      "learning_rate": 8.49353097390088e-05,
      "loss": 2.767433929443359,
      "memory(GiB)": 77.56,
      "step": 29615,
      "token_acc": 0.4,
      "train_speed(iter/s)": 1.441015
    },
    {
      "epoch": 1.2690116104708453,
      "grad_norm": 4.574163436889648,
      "learning_rate": 8.493049488418061e-05,
      "loss": 2.4078720092773436,
      "memory(GiB)": 77.56,
      "step": 29620,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.441089
    },
    {
      "epoch": 1.2692258258000941,
      "grad_norm": 5.394957065582275,
      "learning_rate": 8.492567939655371e-05,
      "loss": 2.544032859802246,
      "memory(GiB)": 77.56,
      "step": 29625,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.441139
    },
    {
      "epoch": 1.2694400411293432,
      "grad_norm": 4.767903804779053,
      "learning_rate": 8.49208632762154e-05,
      "loss": 2.6711475372314455,
      "memory(GiB)": 77.56,
      "step": 29630,
      "token_acc": 0.4624277456647399,
      "train_speed(iter/s)": 1.441114
    },
    {
      "epoch": 1.2696542564585922,
      "grad_norm": 4.421485424041748,
      "learning_rate": 8.491604652325287e-05,
      "loss": 2.718253326416016,
      "memory(GiB)": 77.56,
      "step": 29635,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.441153
    },
    {
      "epoch": 1.269868471787841,
      "grad_norm": 5.698289394378662,
      "learning_rate": 8.491122913775342e-05,
      "loss": 3.055817413330078,
      "memory(GiB)": 77.56,
      "step": 29640,
      "token_acc": 0.3904109589041096,
      "train_speed(iter/s)": 1.441116
    },
    {
      "epoch": 1.27008268711709,
      "grad_norm": 4.952931880950928,
      "learning_rate": 8.490641111980429e-05,
      "loss": 2.50198974609375,
      "memory(GiB)": 77.56,
      "step": 29645,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.441181
    },
    {
      "epoch": 1.270296902446339,
      "grad_norm": 4.448631763458252,
      "learning_rate": 8.490159246949278e-05,
      "loss": 2.729186248779297,
      "memory(GiB)": 77.56,
      "step": 29650,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.441163
    },
    {
      "epoch": 1.270511117775588,
      "grad_norm": 4.190755844116211,
      "learning_rate": 8.489677318690619e-05,
      "loss": 2.6032169342041014,
      "memory(GiB)": 77.56,
      "step": 29655,
      "token_acc": 0.45565749235474007,
      "train_speed(iter/s)": 1.441183
    },
    {
      "epoch": 1.270725333104837,
      "grad_norm": 4.64596700668335,
      "learning_rate": 8.48919532721318e-05,
      "loss": 2.600752258300781,
      "memory(GiB)": 77.56,
      "step": 29660,
      "token_acc": 0.46794871794871795,
      "train_speed(iter/s)": 1.441195
    },
    {
      "epoch": 1.270939548434086,
      "grad_norm": 4.705674171447754,
      "learning_rate": 8.488713272525696e-05,
      "loss": 2.573043441772461,
      "memory(GiB)": 77.56,
      "step": 29665,
      "token_acc": 0.4253968253968254,
      "train_speed(iter/s)": 1.441253
    },
    {
      "epoch": 1.2711537637633348,
      "grad_norm": 4.295762062072754,
      "learning_rate": 8.488231154636899e-05,
      "loss": 2.5170421600341797,
      "memory(GiB)": 77.56,
      "step": 29670,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.44131
    },
    {
      "epoch": 1.2713679790925838,
      "grad_norm": 4.633725643157959,
      "learning_rate": 8.487748973555523e-05,
      "loss": 2.4535173416137694,
      "memory(GiB)": 77.56,
      "step": 29675,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.441293
    },
    {
      "epoch": 1.2715821944218328,
      "grad_norm": 4.391819953918457,
      "learning_rate": 8.487266729290299e-05,
      "loss": 2.348721504211426,
      "memory(GiB)": 77.56,
      "step": 29680,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.441354
    },
    {
      "epoch": 1.2717964097510817,
      "grad_norm": 7.077546119689941,
      "learning_rate": 8.48678442184997e-05,
      "loss": 2.9757081985473635,
      "memory(GiB)": 77.56,
      "step": 29685,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.441361
    },
    {
      "epoch": 1.2720106250803307,
      "grad_norm": 5.971894264221191,
      "learning_rate": 8.48630205124327e-05,
      "loss": 2.3883392333984377,
      "memory(GiB)": 77.56,
      "step": 29690,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441363
    },
    {
      "epoch": 1.2722248404095797,
      "grad_norm": 5.288169860839844,
      "learning_rate": 8.485819617478936e-05,
      "loss": 2.5470142364501953,
      "memory(GiB)": 77.56,
      "step": 29695,
      "token_acc": 0.48466257668711654,
      "train_speed(iter/s)": 1.441349
    },
    {
      "epoch": 1.2724390557388285,
      "grad_norm": 5.781827449798584,
      "learning_rate": 8.485337120565712e-05,
      "loss": 2.4330787658691406,
      "memory(GiB)": 77.56,
      "step": 29700,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.441351
    },
    {
      "epoch": 1.2726532710680776,
      "grad_norm": 6.260162830352783,
      "learning_rate": 8.484854560512335e-05,
      "loss": 2.556867790222168,
      "memory(GiB)": 77.56,
      "step": 29705,
      "token_acc": 0.45018450184501846,
      "train_speed(iter/s)": 1.441306
    },
    {
      "epoch": 1.2728674863973266,
      "grad_norm": 4.832083225250244,
      "learning_rate": 8.48437193732755e-05,
      "loss": 2.403263473510742,
      "memory(GiB)": 77.56,
      "step": 29710,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.44127
    },
    {
      "epoch": 1.2730817017265754,
      "grad_norm": 4.045101642608643,
      "learning_rate": 8.483889251020097e-05,
      "loss": 2.421524429321289,
      "memory(GiB)": 77.56,
      "step": 29715,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.441242
    },
    {
      "epoch": 1.2732959170558245,
      "grad_norm": 5.044188022613525,
      "learning_rate": 8.483406501598724e-05,
      "loss": 2.3714595794677735,
      "memory(GiB)": 77.56,
      "step": 29720,
      "token_acc": 0.4746376811594203,
      "train_speed(iter/s)": 1.44126
    },
    {
      "epoch": 1.2735101323850735,
      "grad_norm": 5.459587574005127,
      "learning_rate": 8.482923689072173e-05,
      "loss": 2.447173309326172,
      "memory(GiB)": 77.56,
      "step": 29725,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.441192
    },
    {
      "epoch": 1.2737243477143223,
      "grad_norm": 4.656224250793457,
      "learning_rate": 8.482440813449193e-05,
      "loss": 2.7099254608154295,
      "memory(GiB)": 77.56,
      "step": 29730,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.441172
    },
    {
      "epoch": 1.2739385630435713,
      "grad_norm": 4.5779900550842285,
      "learning_rate": 8.481957874738529e-05,
      "loss": 2.446799468994141,
      "memory(GiB)": 77.56,
      "step": 29735,
      "token_acc": 0.47076023391812866,
      "train_speed(iter/s)": 1.441138
    },
    {
      "epoch": 1.2741527783728204,
      "grad_norm": 4.220985412597656,
      "learning_rate": 8.481474872948933e-05,
      "loss": 2.419972229003906,
      "memory(GiB)": 77.56,
      "step": 29740,
      "token_acc": 0.5049180327868853,
      "train_speed(iter/s)": 1.441127
    },
    {
      "epoch": 1.2743669937020692,
      "grad_norm": 6.136692047119141,
      "learning_rate": 8.480991808089156e-05,
      "loss": 2.4867759704589845,
      "memory(GiB)": 77.56,
      "step": 29745,
      "token_acc": 0.4789156626506024,
      "train_speed(iter/s)": 1.441114
    },
    {
      "epoch": 1.2745812090313182,
      "grad_norm": 4.943179607391357,
      "learning_rate": 8.480508680167945e-05,
      "loss": 2.5574302673339844,
      "memory(GiB)": 77.56,
      "step": 29750,
      "token_acc": 0.4954128440366973,
      "train_speed(iter/s)": 1.441157
    },
    {
      "epoch": 1.2747954243605673,
      "grad_norm": 5.318636417388916,
      "learning_rate": 8.480025489194055e-05,
      "loss": 2.1998023986816406,
      "memory(GiB)": 77.56,
      "step": 29755,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.441148
    },
    {
      "epoch": 1.275009639689816,
      "grad_norm": 5.354988098144531,
      "learning_rate": 8.479542235176236e-05,
      "loss": 2.6404611587524416,
      "memory(GiB)": 77.56,
      "step": 29760,
      "token_acc": 0.43944636678200694,
      "train_speed(iter/s)": 1.441137
    },
    {
      "epoch": 1.275223855019065,
      "grad_norm": 5.203178405761719,
      "learning_rate": 8.479058918123248e-05,
      "loss": 2.486052131652832,
      "memory(GiB)": 77.56,
      "step": 29765,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.441166
    },
    {
      "epoch": 1.2754380703483141,
      "grad_norm": 6.913606643676758,
      "learning_rate": 8.478575538043843e-05,
      "loss": 2.5961864471435545,
      "memory(GiB)": 77.56,
      "step": 29770,
      "token_acc": 0.4549019607843137,
      "train_speed(iter/s)": 1.44118
    },
    {
      "epoch": 1.2756522856775632,
      "grad_norm": 4.3769001960754395,
      "learning_rate": 8.47809209494678e-05,
      "loss": 2.586769676208496,
      "memory(GiB)": 77.56,
      "step": 29775,
      "token_acc": 0.45394736842105265,
      "train_speed(iter/s)": 1.441162
    },
    {
      "epoch": 1.275866501006812,
      "grad_norm": 4.2965569496154785,
      "learning_rate": 8.477608588840815e-05,
      "loss": 2.709580421447754,
      "memory(GiB)": 77.56,
      "step": 29780,
      "token_acc": 0.44385026737967914,
      "train_speed(iter/s)": 1.441167
    },
    {
      "epoch": 1.276080716336061,
      "grad_norm": 4.180333614349365,
      "learning_rate": 8.477125019734709e-05,
      "loss": 2.206060218811035,
      "memory(GiB)": 77.56,
      "step": 29785,
      "token_acc": 0.5541125541125541,
      "train_speed(iter/s)": 1.441115
    },
    {
      "epoch": 1.27629493166531,
      "grad_norm": 4.147204399108887,
      "learning_rate": 8.476641387637221e-05,
      "loss": 2.558124542236328,
      "memory(GiB)": 77.56,
      "step": 29790,
      "token_acc": 0.451505016722408,
      "train_speed(iter/s)": 1.441146
    },
    {
      "epoch": 1.2765091469945589,
      "grad_norm": 4.534034252166748,
      "learning_rate": 8.476157692557112e-05,
      "loss": 2.290990447998047,
      "memory(GiB)": 77.56,
      "step": 29795,
      "token_acc": 0.5144694533762058,
      "train_speed(iter/s)": 1.441105
    },
    {
      "epoch": 1.276723362323808,
      "grad_norm": 5.902621746063232,
      "learning_rate": 8.475673934503147e-05,
      "loss": 2.5384904861450197,
      "memory(GiB)": 77.56,
      "step": 29800,
      "token_acc": 0.46557377049180326,
      "train_speed(iter/s)": 1.441063
    },
    {
      "epoch": 1.276937577653057,
      "grad_norm": 5.168031692504883,
      "learning_rate": 8.475190113484088e-05,
      "loss": 2.5275623321533205,
      "memory(GiB)": 77.56,
      "step": 29805,
      "token_acc": 0.42592592592592593,
      "train_speed(iter/s)": 1.441055
    },
    {
      "epoch": 1.2771517929823057,
      "grad_norm": 4.34005069732666,
      "learning_rate": 8.4747062295087e-05,
      "loss": 2.7979345321655273,
      "memory(GiB)": 77.56,
      "step": 29810,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.441146
    },
    {
      "epoch": 1.2773660083115548,
      "grad_norm": 3.788792610168457,
      "learning_rate": 8.474222282585749e-05,
      "loss": 2.6363540649414063,
      "memory(GiB)": 77.56,
      "step": 29815,
      "token_acc": 0.45144356955380577,
      "train_speed(iter/s)": 1.441147
    },
    {
      "epoch": 1.2775802236408038,
      "grad_norm": 4.921347141265869,
      "learning_rate": 8.473738272724001e-05,
      "loss": 2.4330434799194336,
      "memory(GiB)": 77.56,
      "step": 29820,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.441114
    },
    {
      "epoch": 1.2777944389700526,
      "grad_norm": 4.742170810699463,
      "learning_rate": 8.473254199932227e-05,
      "loss": 2.4082475662231446,
      "memory(GiB)": 77.56,
      "step": 29825,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.44113
    },
    {
      "epoch": 1.2780086542993017,
      "grad_norm": 5.1656975746154785,
      "learning_rate": 8.472770064219196e-05,
      "loss": 2.3235877990722655,
      "memory(GiB)": 77.56,
      "step": 29830,
      "token_acc": 0.5313531353135313,
      "train_speed(iter/s)": 1.441116
    },
    {
      "epoch": 1.2782228696285507,
      "grad_norm": 3.8671486377716064,
      "learning_rate": 8.472285865593674e-05,
      "loss": 2.4731122970581056,
      "memory(GiB)": 77.56,
      "step": 29835,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.441104
    },
    {
      "epoch": 1.2784370849577995,
      "grad_norm": 4.296897888183594,
      "learning_rate": 8.47180160406444e-05,
      "loss": 2.556016731262207,
      "memory(GiB)": 77.56,
      "step": 29840,
      "token_acc": 0.5136986301369864,
      "train_speed(iter/s)": 1.441102
    },
    {
      "epoch": 1.2786513002870485,
      "grad_norm": 4.75508975982666,
      "learning_rate": 8.471317279640261e-05,
      "loss": 2.7080209732055662,
      "memory(GiB)": 77.56,
      "step": 29845,
      "token_acc": 0.45396825396825397,
      "train_speed(iter/s)": 1.4411
    },
    {
      "epoch": 1.2788655156162976,
      "grad_norm": 4.338958740234375,
      "learning_rate": 8.470832892329912e-05,
      "loss": 2.454414176940918,
      "memory(GiB)": 77.56,
      "step": 29850,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.44104
    },
    {
      "epoch": 1.2790797309455464,
      "grad_norm": 5.543600559234619,
      "learning_rate": 8.470348442142172e-05,
      "loss": 2.488588333129883,
      "memory(GiB)": 77.56,
      "step": 29855,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.441028
    },
    {
      "epoch": 1.2792939462747954,
      "grad_norm": 5.980303764343262,
      "learning_rate": 8.469863929085813e-05,
      "loss": 2.561494064331055,
      "memory(GiB)": 77.56,
      "step": 29860,
      "token_acc": 0.4623287671232877,
      "train_speed(iter/s)": 1.440997
    },
    {
      "epoch": 1.2795081616040445,
      "grad_norm": 4.543352127075195,
      "learning_rate": 8.469379353169615e-05,
      "loss": 2.335466003417969,
      "memory(GiB)": 77.56,
      "step": 29865,
      "token_acc": 0.5054151624548736,
      "train_speed(iter/s)": 1.440972
    },
    {
      "epoch": 1.2797223769332933,
      "grad_norm": 6.944456577301025,
      "learning_rate": 8.468894714402356e-05,
      "loss": 3.019215393066406,
      "memory(GiB)": 77.56,
      "step": 29870,
      "token_acc": 0.40390879478827363,
      "train_speed(iter/s)": 1.441014
    },
    {
      "epoch": 1.2799365922625423,
      "grad_norm": 5.229017734527588,
      "learning_rate": 8.468410012792813e-05,
      "loss": 2.2623971939086913,
      "memory(GiB)": 77.56,
      "step": 29875,
      "token_acc": 0.5098684210526315,
      "train_speed(iter/s)": 1.440983
    },
    {
      "epoch": 1.2801508075917913,
      "grad_norm": 8.276698112487793,
      "learning_rate": 8.467925248349771e-05,
      "loss": 2.4009773254394533,
      "memory(GiB)": 77.56,
      "step": 29880,
      "token_acc": 0.49830508474576274,
      "train_speed(iter/s)": 1.440973
    },
    {
      "epoch": 1.2803650229210402,
      "grad_norm": 5.859376907348633,
      "learning_rate": 8.467440421082008e-05,
      "loss": 2.3379093170166017,
      "memory(GiB)": 77.56,
      "step": 29885,
      "token_acc": 0.518796992481203,
      "train_speed(iter/s)": 1.44096
    },
    {
      "epoch": 1.2805792382502892,
      "grad_norm": 4.952724456787109,
      "learning_rate": 8.466955530998311e-05,
      "loss": 2.523929786682129,
      "memory(GiB)": 77.56,
      "step": 29890,
      "token_acc": 0.4879032258064516,
      "train_speed(iter/s)": 1.441012
    },
    {
      "epoch": 1.2807934535795382,
      "grad_norm": 5.0092878341674805,
      "learning_rate": 8.466470578107462e-05,
      "loss": 2.7064552307128906,
      "memory(GiB)": 77.56,
      "step": 29895,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.440983
    },
    {
      "epoch": 1.281007668908787,
      "grad_norm": 4.82404088973999,
      "learning_rate": 8.465985562418244e-05,
      "loss": 2.713532066345215,
      "memory(GiB)": 77.56,
      "step": 29900,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.440986
    },
    {
      "epoch": 1.281221884238036,
      "grad_norm": 6.226202964782715,
      "learning_rate": 8.465500483939449e-05,
      "loss": 2.8195823669433593,
      "memory(GiB)": 77.56,
      "step": 29905,
      "token_acc": 0.46303501945525294,
      "train_speed(iter/s)": 1.441021
    },
    {
      "epoch": 1.281436099567285,
      "grad_norm": 5.871167182922363,
      "learning_rate": 8.465015342679861e-05,
      "loss": 2.382765769958496,
      "memory(GiB)": 77.56,
      "step": 29910,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.440985
    },
    {
      "epoch": 1.281650314896534,
      "grad_norm": 5.723329067230225,
      "learning_rate": 8.46453013864827e-05,
      "loss": 2.5458908081054688,
      "memory(GiB)": 77.56,
      "step": 29915,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.441042
    },
    {
      "epoch": 1.281864530225783,
      "grad_norm": 4.599218845367432,
      "learning_rate": 8.464044871853465e-05,
      "loss": 2.552521514892578,
      "memory(GiB)": 77.56,
      "step": 29920,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.441029
    },
    {
      "epoch": 1.282078745555032,
      "grad_norm": 4.348132610321045,
      "learning_rate": 8.463559542304238e-05,
      "loss": 2.5979137420654297,
      "memory(GiB)": 77.56,
      "step": 29925,
      "token_acc": 0.4638888888888889,
      "train_speed(iter/s)": 1.441008
    },
    {
      "epoch": 1.2822929608842808,
      "grad_norm": 4.032307147979736,
      "learning_rate": 8.46307415000938e-05,
      "loss": 2.4703521728515625,
      "memory(GiB)": 77.56,
      "step": 29930,
      "token_acc": 0.4777327935222672,
      "train_speed(iter/s)": 1.441061
    },
    {
      "epoch": 1.2825071762135298,
      "grad_norm": 6.230744361877441,
      "learning_rate": 8.462588694977686e-05,
      "loss": 2.4588584899902344,
      "memory(GiB)": 77.56,
      "step": 29935,
      "token_acc": 0.4558303886925795,
      "train_speed(iter/s)": 1.441122
    },
    {
      "epoch": 1.2827213915427789,
      "grad_norm": 6.356720447540283,
      "learning_rate": 8.46210317721795e-05,
      "loss": 2.537357711791992,
      "memory(GiB)": 77.56,
      "step": 29940,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.441127
    },
    {
      "epoch": 1.282935606872028,
      "grad_norm": 4.601590633392334,
      "learning_rate": 8.461617596738967e-05,
      "loss": 2.2179203033447266,
      "memory(GiB)": 77.56,
      "step": 29945,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.441155
    },
    {
      "epoch": 1.2831498222012767,
      "grad_norm": 6.738268852233887,
      "learning_rate": 8.461131953549532e-05,
      "loss": 2.5237405776977537,
      "memory(GiB)": 77.56,
      "step": 29950,
      "token_acc": 0.4728682170542636,
      "train_speed(iter/s)": 1.441165
    },
    {
      "epoch": 1.2833640375305257,
      "grad_norm": 4.284161567687988,
      "learning_rate": 8.460646247658446e-05,
      "loss": 2.600858688354492,
      "memory(GiB)": 77.56,
      "step": 29955,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.441169
    },
    {
      "epoch": 1.2835782528597748,
      "grad_norm": 4.127976894378662,
      "learning_rate": 8.460160479074508e-05,
      "loss": 2.69942626953125,
      "memory(GiB)": 77.56,
      "step": 29960,
      "token_acc": 0.44744744744744747,
      "train_speed(iter/s)": 1.441152
    },
    {
      "epoch": 1.2837924681890236,
      "grad_norm": 5.700314521789551,
      "learning_rate": 8.459674647806517e-05,
      "loss": 2.6903078079223635,
      "memory(GiB)": 77.56,
      "step": 29965,
      "token_acc": 0.44314868804664725,
      "train_speed(iter/s)": 1.441189
    },
    {
      "epoch": 1.2840066835182726,
      "grad_norm": 3.552034616470337,
      "learning_rate": 8.459188753863273e-05,
      "loss": 2.807967758178711,
      "memory(GiB)": 77.56,
      "step": 29970,
      "token_acc": 0.445993031358885,
      "train_speed(iter/s)": 1.441236
    },
    {
      "epoch": 1.2842208988475217,
      "grad_norm": 4.367836952209473,
      "learning_rate": 8.458702797253581e-05,
      "loss": 2.3090091705322267,
      "memory(GiB)": 77.56,
      "step": 29975,
      "token_acc": 0.5310077519379846,
      "train_speed(iter/s)": 1.441271
    },
    {
      "epoch": 1.2844351141767705,
      "grad_norm": 6.73507833480835,
      "learning_rate": 8.458216777986243e-05,
      "loss": 2.1440126419067385,
      "memory(GiB)": 77.56,
      "step": 29980,
      "token_acc": 0.5061728395061729,
      "train_speed(iter/s)": 1.441337
    },
    {
      "epoch": 1.2846493295060195,
      "grad_norm": 4.563814163208008,
      "learning_rate": 8.457730696070063e-05,
      "loss": 2.487034797668457,
      "memory(GiB)": 77.56,
      "step": 29985,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.441373
    },
    {
      "epoch": 1.2848635448352685,
      "grad_norm": 5.482424736022949,
      "learning_rate": 8.457244551513848e-05,
      "loss": 2.790231132507324,
      "memory(GiB)": 77.56,
      "step": 29990,
      "token_acc": 0.44086021505376344,
      "train_speed(iter/s)": 1.441336
    },
    {
      "epoch": 1.2850777601645174,
      "grad_norm": 5.418969631195068,
      "learning_rate": 8.456758344326406e-05,
      "loss": 2.3300899505615233,
      "memory(GiB)": 77.56,
      "step": 29995,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.441335
    },
    {
      "epoch": 1.2852919754937664,
      "grad_norm": 4.195066928863525,
      "learning_rate": 8.456272074516542e-05,
      "loss": 2.653728485107422,
      "memory(GiB)": 77.56,
      "step": 30000,
      "token_acc": 0.44857142857142857,
      "train_speed(iter/s)": 1.441371
    },
    {
      "epoch": 1.2852919754937664,
      "eval_loss": 2.228975296020508,
      "eval_runtime": 14.4878,
      "eval_samples_per_second": 6.902,
      "eval_steps_per_second": 6.902,
      "eval_token_acc": 0.44813278008298757,
      "step": 30000
    },
    {
      "epoch": 1.2855061908230154,
      "grad_norm": 4.7024993896484375,
      "learning_rate": 8.45578574209307e-05,
      "loss": 2.5826650619506837,
      "memory(GiB)": 77.56,
      "step": 30005,
      "token_acc": 0.4550102249488753,
      "train_speed(iter/s)": 1.440312
    },
    {
      "epoch": 1.2857204061522642,
      "grad_norm": 6.061879634857178,
      "learning_rate": 8.455299347064795e-05,
      "loss": 2.5359498977661135,
      "memory(GiB)": 77.56,
      "step": 30010,
      "token_acc": 0.44107744107744107,
      "train_speed(iter/s)": 1.440321
    },
    {
      "epoch": 1.2859346214815133,
      "grad_norm": 6.374854564666748,
      "learning_rate": 8.454812889440533e-05,
      "loss": 2.3168535232543945,
      "memory(GiB)": 77.56,
      "step": 30015,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.440292
    },
    {
      "epoch": 1.2861488368107623,
      "grad_norm": 6.208530426025391,
      "learning_rate": 8.454326369229094e-05,
      "loss": 2.752610206604004,
      "memory(GiB)": 77.56,
      "step": 30020,
      "token_acc": 0.43989071038251365,
      "train_speed(iter/s)": 1.440312
    },
    {
      "epoch": 1.2863630521400111,
      "grad_norm": 5.622336387634277,
      "learning_rate": 8.45383978643929e-05,
      "loss": 2.6988750457763673,
      "memory(GiB)": 77.56,
      "step": 30025,
      "token_acc": 0.45544554455445546,
      "train_speed(iter/s)": 1.440287
    },
    {
      "epoch": 1.2865772674692602,
      "grad_norm": 5.669009685516357,
      "learning_rate": 8.453353141079941e-05,
      "loss": 2.32655029296875,
      "memory(GiB)": 77.56,
      "step": 30030,
      "token_acc": 0.5486381322957199,
      "train_speed(iter/s)": 1.440273
    },
    {
      "epoch": 1.2867914827985092,
      "grad_norm": 4.0814208984375,
      "learning_rate": 8.452866433159859e-05,
      "loss": 2.4197036743164064,
      "memory(GiB)": 77.56,
      "step": 30035,
      "token_acc": 0.4529616724738676,
      "train_speed(iter/s)": 1.44032
    },
    {
      "epoch": 1.287005698127758,
      "grad_norm": 5.440887928009033,
      "learning_rate": 8.452379662687864e-05,
      "loss": 2.7072935104370117,
      "memory(GiB)": 77.56,
      "step": 30040,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.440344
    },
    {
      "epoch": 1.287219913457007,
      "grad_norm": 5.3693976402282715,
      "learning_rate": 8.45189282967277e-05,
      "loss": 2.7098106384277343,
      "memory(GiB)": 77.56,
      "step": 30045,
      "token_acc": 0.43656716417910446,
      "train_speed(iter/s)": 1.440422
    },
    {
      "epoch": 1.287434128786256,
      "grad_norm": 4.537566184997559,
      "learning_rate": 8.451405934123402e-05,
      "loss": 2.397121810913086,
      "memory(GiB)": 77.56,
      "step": 30050,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.440424
    },
    {
      "epoch": 1.2876483441155049,
      "grad_norm": 4.082806587219238,
      "learning_rate": 8.450918976048577e-05,
      "loss": 2.542250061035156,
      "memory(GiB)": 77.56,
      "step": 30055,
      "token_acc": 0.43661971830985913,
      "train_speed(iter/s)": 1.440433
    },
    {
      "epoch": 1.287862559444754,
      "grad_norm": 4.772249698638916,
      "learning_rate": 8.450431955457118e-05,
      "loss": 2.2546098709106444,
      "memory(GiB)": 77.56,
      "step": 30060,
      "token_acc": 0.5110294117647058,
      "train_speed(iter/s)": 1.440471
    },
    {
      "epoch": 1.288076774774003,
      "grad_norm": 5.119551658630371,
      "learning_rate": 8.449944872357845e-05,
      "loss": 2.6211755752563475,
      "memory(GiB)": 77.56,
      "step": 30065,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.440482
    },
    {
      "epoch": 1.2882909901032518,
      "grad_norm": 4.743472576141357,
      "learning_rate": 8.449457726759586e-05,
      "loss": 2.328533172607422,
      "memory(GiB)": 77.56,
      "step": 30070,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.440466
    },
    {
      "epoch": 1.2885052054325008,
      "grad_norm": 4.37932825088501,
      "learning_rate": 8.448970518671165e-05,
      "loss": 2.489857864379883,
      "memory(GiB)": 77.56,
      "step": 30075,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.44049
    },
    {
      "epoch": 1.2887194207617498,
      "grad_norm": 5.37784481048584,
      "learning_rate": 8.448483248101408e-05,
      "loss": 2.578757476806641,
      "memory(GiB)": 77.56,
      "step": 30080,
      "token_acc": 0.46567164179104475,
      "train_speed(iter/s)": 1.440489
    },
    {
      "epoch": 1.2889336360909986,
      "grad_norm": 4.141708850860596,
      "learning_rate": 8.447995915059142e-05,
      "loss": 2.4571609497070312,
      "memory(GiB)": 77.56,
      "step": 30085,
      "token_acc": 0.46075085324232085,
      "train_speed(iter/s)": 1.440436
    },
    {
      "epoch": 1.2891478514202477,
      "grad_norm": 6.63078498840332,
      "learning_rate": 8.447508519553194e-05,
      "loss": 2.537199783325195,
      "memory(GiB)": 77.56,
      "step": 30090,
      "token_acc": 0.43343653250773995,
      "train_speed(iter/s)": 1.440451
    },
    {
      "epoch": 1.2893620667494967,
      "grad_norm": 3.774007797241211,
      "learning_rate": 8.447021061592396e-05,
      "loss": 2.522028923034668,
      "memory(GiB)": 77.56,
      "step": 30095,
      "token_acc": 0.48520710059171596,
      "train_speed(iter/s)": 1.440489
    },
    {
      "epoch": 1.2895762820787455,
      "grad_norm": 7.373660087585449,
      "learning_rate": 8.446533541185577e-05,
      "loss": 2.354354476928711,
      "memory(GiB)": 77.56,
      "step": 30100,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.440564
    },
    {
      "epoch": 1.2897904974079946,
      "grad_norm": 5.748693466186523,
      "learning_rate": 8.44604595834157e-05,
      "loss": 2.551074981689453,
      "memory(GiB)": 77.56,
      "step": 30105,
      "token_acc": 0.5130434782608696,
      "train_speed(iter/s)": 1.440572
    },
    {
      "epoch": 1.2900047127372436,
      "grad_norm": 3.907365322113037,
      "learning_rate": 8.445558313069208e-05,
      "loss": 2.5525793075561523,
      "memory(GiB)": 77.56,
      "step": 30110,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.440581
    },
    {
      "epoch": 1.2902189280664924,
      "grad_norm": 5.5799102783203125,
      "learning_rate": 8.445070605377326e-05,
      "loss": 2.7172801971435545,
      "memory(GiB)": 77.56,
      "step": 30115,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.44054
    },
    {
      "epoch": 1.2904331433957414,
      "grad_norm": 5.489712238311768,
      "learning_rate": 8.444582835274758e-05,
      "loss": 2.4466455459594725,
      "memory(GiB)": 77.56,
      "step": 30120,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.440558
    },
    {
      "epoch": 1.2906473587249905,
      "grad_norm": 4.120311737060547,
      "learning_rate": 8.444095002770341e-05,
      "loss": 2.492127227783203,
      "memory(GiB)": 77.56,
      "step": 30125,
      "token_acc": 0.4956521739130435,
      "train_speed(iter/s)": 1.4405
    },
    {
      "epoch": 1.2908615740542393,
      "grad_norm": 5.713088035583496,
      "learning_rate": 8.44360710787291e-05,
      "loss": 2.3864126205444336,
      "memory(GiB)": 77.56,
      "step": 30130,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.440494
    },
    {
      "epoch": 1.2910757893834883,
      "grad_norm": 4.283507823944092,
      "learning_rate": 8.443119150591309e-05,
      "loss": 2.364004707336426,
      "memory(GiB)": 77.56,
      "step": 30135,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.440536
    },
    {
      "epoch": 1.2912900047127374,
      "grad_norm": 4.761294364929199,
      "learning_rate": 8.442631130934372e-05,
      "loss": 2.564166450500488,
      "memory(GiB)": 77.56,
      "step": 30140,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.440507
    },
    {
      "epoch": 1.2915042200419862,
      "grad_norm": 4.866940975189209,
      "learning_rate": 8.442143048910944e-05,
      "loss": 2.268455696105957,
      "memory(GiB)": 77.56,
      "step": 30145,
      "token_acc": 0.5042372881355932,
      "train_speed(iter/s)": 1.440497
    },
    {
      "epoch": 1.2917184353712352,
      "grad_norm": 4.474341869354248,
      "learning_rate": 8.441654904529866e-05,
      "loss": 2.220620346069336,
      "memory(GiB)": 77.56,
      "step": 30150,
      "token_acc": 0.5242718446601942,
      "train_speed(iter/s)": 1.440472
    },
    {
      "epoch": 1.2919326507004842,
      "grad_norm": 5.508834362030029,
      "learning_rate": 8.441166697799981e-05,
      "loss": 2.6750415802001952,
      "memory(GiB)": 77.56,
      "step": 30155,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.440502
    },
    {
      "epoch": 1.292146866029733,
      "grad_norm": 4.944782733917236,
      "learning_rate": 8.440678428730131e-05,
      "loss": 2.588155746459961,
      "memory(GiB)": 77.56,
      "step": 30160,
      "token_acc": 0.49375,
      "train_speed(iter/s)": 1.440533
    },
    {
      "epoch": 1.292361081358982,
      "grad_norm": 5.445731163024902,
      "learning_rate": 8.440190097329165e-05,
      "loss": 2.3746435165405275,
      "memory(GiB)": 77.56,
      "step": 30165,
      "token_acc": 0.5019607843137255,
      "train_speed(iter/s)": 1.440557
    },
    {
      "epoch": 1.2925752966882311,
      "grad_norm": 6.3063554763793945,
      "learning_rate": 8.439701703605929e-05,
      "loss": 2.836256408691406,
      "memory(GiB)": 77.56,
      "step": 30170,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.44061
    },
    {
      "epoch": 1.29278951201748,
      "grad_norm": 5.162059783935547,
      "learning_rate": 8.439213247569269e-05,
      "loss": 2.5087608337402343,
      "memory(GiB)": 77.56,
      "step": 30175,
      "token_acc": 0.4669260700389105,
      "train_speed(iter/s)": 1.440571
    },
    {
      "epoch": 1.293003727346729,
      "grad_norm": 9.062597274780273,
      "learning_rate": 8.438724729228034e-05,
      "loss": 2.2947444915771484,
      "memory(GiB)": 77.56,
      "step": 30180,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440556
    },
    {
      "epoch": 1.293217942675978,
      "grad_norm": 5.128320217132568,
      "learning_rate": 8.438236148591076e-05,
      "loss": 2.704743576049805,
      "memory(GiB)": 77.56,
      "step": 30185,
      "token_acc": 0.4214876033057851,
      "train_speed(iter/s)": 1.440589
    },
    {
      "epoch": 1.2934321580052268,
      "grad_norm": 4.757826805114746,
      "learning_rate": 8.437747505667243e-05,
      "loss": 2.3504001617431642,
      "memory(GiB)": 77.56,
      "step": 30190,
      "token_acc": 0.49480968858131485,
      "train_speed(iter/s)": 1.440609
    },
    {
      "epoch": 1.2936463733344759,
      "grad_norm": 6.292047500610352,
      "learning_rate": 8.437258800465392e-05,
      "loss": 2.2857501983642576,
      "memory(GiB)": 77.56,
      "step": 30195,
      "token_acc": 0.511520737327189,
      "train_speed(iter/s)": 1.440617
    },
    {
      "epoch": 1.2938605886637249,
      "grad_norm": 5.029705047607422,
      "learning_rate": 8.436770032994372e-05,
      "loss": 2.325166702270508,
      "memory(GiB)": 77.56,
      "step": 30200,
      "token_acc": 0.5190839694656488,
      "train_speed(iter/s)": 1.440672
    },
    {
      "epoch": 1.2940748039929737,
      "grad_norm": 4.224485874176025,
      "learning_rate": 8.436281203263039e-05,
      "loss": 2.3237606048583985,
      "memory(GiB)": 77.56,
      "step": 30205,
      "token_acc": 0.45185185185185184,
      "train_speed(iter/s)": 1.440693
    },
    {
      "epoch": 1.2942890193222227,
      "grad_norm": 4.651941299438477,
      "learning_rate": 8.435792311280248e-05,
      "loss": 2.204146957397461,
      "memory(GiB)": 77.56,
      "step": 30210,
      "token_acc": 0.5401929260450161,
      "train_speed(iter/s)": 1.440679
    },
    {
      "epoch": 1.2945032346514718,
      "grad_norm": 5.682839393615723,
      "learning_rate": 8.435303357054857e-05,
      "loss": 2.5758430480957033,
      "memory(GiB)": 77.56,
      "step": 30215,
      "token_acc": 0.4601226993865031,
      "train_speed(iter/s)": 1.440665
    },
    {
      "epoch": 1.2947174499807206,
      "grad_norm": 5.664210796356201,
      "learning_rate": 8.434814340595722e-05,
      "loss": 2.5128652572631838,
      "memory(GiB)": 77.56,
      "step": 30220,
      "token_acc": 0.45569620253164556,
      "train_speed(iter/s)": 1.44067
    },
    {
      "epoch": 1.2949316653099696,
      "grad_norm": 4.011490821838379,
      "learning_rate": 8.434325261911705e-05,
      "loss": 2.680637741088867,
      "memory(GiB)": 77.56,
      "step": 30225,
      "token_acc": 0.4377224199288256,
      "train_speed(iter/s)": 1.44071
    },
    {
      "epoch": 1.2951458806392186,
      "grad_norm": 4.033299922943115,
      "learning_rate": 8.433836121011662e-05,
      "loss": 2.615276908874512,
      "memory(GiB)": 77.56,
      "step": 30230,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.440768
    },
    {
      "epoch": 1.2953600959684675,
      "grad_norm": 4.467537879943848,
      "learning_rate": 8.43334691790446e-05,
      "loss": 2.5148754119873047,
      "memory(GiB)": 77.56,
      "step": 30235,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.440784
    },
    {
      "epoch": 1.2955743112977165,
      "grad_norm": 3.937587261199951,
      "learning_rate": 8.432857652598954e-05,
      "loss": 2.328424072265625,
      "memory(GiB)": 77.56,
      "step": 30240,
      "token_acc": 0.5129032258064516,
      "train_speed(iter/s)": 1.440819
    },
    {
      "epoch": 1.2957885266269655,
      "grad_norm": 5.247279167175293,
      "learning_rate": 8.432368325104014e-05,
      "loss": 2.4510190963745115,
      "memory(GiB)": 77.56,
      "step": 30245,
      "token_acc": 0.4980544747081712,
      "train_speed(iter/s)": 1.440825
    },
    {
      "epoch": 1.2960027419562143,
      "grad_norm": 4.112699508666992,
      "learning_rate": 8.4318789354285e-05,
      "loss": 2.8035411834716797,
      "memory(GiB)": 77.56,
      "step": 30250,
      "token_acc": 0.42214532871972316,
      "train_speed(iter/s)": 1.440855
    },
    {
      "epoch": 1.2962169572854634,
      "grad_norm": 5.823759078979492,
      "learning_rate": 8.43138948358128e-05,
      "loss": 2.723115158081055,
      "memory(GiB)": 77.56,
      "step": 30255,
      "token_acc": 0.4485049833887043,
      "train_speed(iter/s)": 1.440902
    },
    {
      "epoch": 1.2964311726147124,
      "grad_norm": 4.194140434265137,
      "learning_rate": 8.43089996957122e-05,
      "loss": 2.4155887603759765,
      "memory(GiB)": 77.56,
      "step": 30260,
      "token_acc": 0.5272727272727272,
      "train_speed(iter/s)": 1.440961
    },
    {
      "epoch": 1.2966453879439612,
      "grad_norm": 4.884934425354004,
      "learning_rate": 8.43041039340719e-05,
      "loss": 2.9081209182739256,
      "memory(GiB)": 77.56,
      "step": 30265,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.441037
    },
    {
      "epoch": 1.2968596032732103,
      "grad_norm": 4.536020278930664,
      "learning_rate": 8.429920755098058e-05,
      "loss": 2.5895225524902346,
      "memory(GiB)": 77.56,
      "step": 30270,
      "token_acc": 0.438871473354232,
      "train_speed(iter/s)": 1.441042
    },
    {
      "epoch": 1.2970738186024593,
      "grad_norm": 4.42290735244751,
      "learning_rate": 8.429431054652695e-05,
      "loss": 2.05948429107666,
      "memory(GiB)": 77.56,
      "step": 30275,
      "token_acc": 0.5703422053231939,
      "train_speed(iter/s)": 1.441055
    },
    {
      "epoch": 1.297288033931708,
      "grad_norm": 5.010835647583008,
      "learning_rate": 8.428941292079968e-05,
      "loss": 2.635399055480957,
      "memory(GiB)": 77.56,
      "step": 30280,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.440991
    },
    {
      "epoch": 1.2975022492609571,
      "grad_norm": 6.494987487792969,
      "learning_rate": 8.428451467388756e-05,
      "loss": 2.6263401031494142,
      "memory(GiB)": 77.56,
      "step": 30285,
      "token_acc": 0.4481707317073171,
      "train_speed(iter/s)": 1.441006
    },
    {
      "epoch": 1.2977164645902062,
      "grad_norm": 5.413225173950195,
      "learning_rate": 8.427961580587927e-05,
      "loss": 2.60037841796875,
      "memory(GiB)": 77.56,
      "step": 30290,
      "token_acc": 0.46096654275092935,
      "train_speed(iter/s)": 1.441013
    },
    {
      "epoch": 1.297930679919455,
      "grad_norm": 4.137669086456299,
      "learning_rate": 8.427471631686359e-05,
      "loss": 2.451686477661133,
      "memory(GiB)": 77.56,
      "step": 30295,
      "token_acc": 0.5046153846153846,
      "train_speed(iter/s)": 1.440965
    },
    {
      "epoch": 1.298144895248704,
      "grad_norm": 4.886295318603516,
      "learning_rate": 8.426981620692929e-05,
      "loss": 2.62661075592041,
      "memory(GiB)": 77.56,
      "step": 30300,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.440948
    },
    {
      "epoch": 1.298359110577953,
      "grad_norm": 10.413487434387207,
      "learning_rate": 8.426491547616511e-05,
      "loss": 2.5587528228759764,
      "memory(GiB)": 77.56,
      "step": 30305,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.440991
    },
    {
      "epoch": 1.2985733259072019,
      "grad_norm": 4.705319881439209,
      "learning_rate": 8.426001412465984e-05,
      "loss": 2.501504898071289,
      "memory(GiB)": 77.56,
      "step": 30310,
      "token_acc": 0.49809885931558934,
      "train_speed(iter/s)": 1.440974
    },
    {
      "epoch": 1.298787541236451,
      "grad_norm": 6.307821750640869,
      "learning_rate": 8.425511215250226e-05,
      "loss": 2.345410919189453,
      "memory(GiB)": 77.56,
      "step": 30315,
      "token_acc": 0.46484375,
      "train_speed(iter/s)": 1.440973
    },
    {
      "epoch": 1.2990017565657,
      "grad_norm": 8.535022735595703,
      "learning_rate": 8.425020955978122e-05,
      "loss": 2.7694124221801757,
      "memory(GiB)": 77.56,
      "step": 30320,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.440985
    },
    {
      "epoch": 1.2992159718949488,
      "grad_norm": 4.92496395111084,
      "learning_rate": 8.424530634658549e-05,
      "loss": 2.417336654663086,
      "memory(GiB)": 77.56,
      "step": 30325,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.440999
    },
    {
      "epoch": 1.2994301872241978,
      "grad_norm": 5.449120044708252,
      "learning_rate": 8.424040251300392e-05,
      "loss": 2.5895715713500977,
      "memory(GiB)": 77.56,
      "step": 30330,
      "token_acc": 0.4756554307116105,
      "train_speed(iter/s)": 1.441064
    },
    {
      "epoch": 1.2996444025534468,
      "grad_norm": 5.59116268157959,
      "learning_rate": 8.423549805912532e-05,
      "loss": 2.591572570800781,
      "memory(GiB)": 77.56,
      "step": 30335,
      "token_acc": 0.44039735099337746,
      "train_speed(iter/s)": 1.441084
    },
    {
      "epoch": 1.2998586178826956,
      "grad_norm": 5.4039626121521,
      "learning_rate": 8.423059298503857e-05,
      "loss": 2.8692398071289062,
      "memory(GiB)": 77.56,
      "step": 30340,
      "token_acc": 0.4171779141104294,
      "train_speed(iter/s)": 1.441059
    },
    {
      "epoch": 1.3000728332119447,
      "grad_norm": 5.4849982261657715,
      "learning_rate": 8.422568729083251e-05,
      "loss": 2.378757858276367,
      "memory(GiB)": 77.56,
      "step": 30345,
      "token_acc": 0.489010989010989,
      "train_speed(iter/s)": 1.441063
    },
    {
      "epoch": 1.3002870485411937,
      "grad_norm": 5.328290939331055,
      "learning_rate": 8.422078097659602e-05,
      "loss": 2.5465137481689455,
      "memory(GiB)": 77.56,
      "step": 30350,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.441087
    },
    {
      "epoch": 1.3005012638704425,
      "grad_norm": 4.981322288513184,
      "learning_rate": 8.4215874042418e-05,
      "loss": 2.591132926940918,
      "memory(GiB)": 77.56,
      "step": 30355,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.441125
    },
    {
      "epoch": 1.3007154791996915,
      "grad_norm": 4.698912143707275,
      "learning_rate": 8.421096648838728e-05,
      "loss": 2.6168724060058595,
      "memory(GiB)": 77.56,
      "step": 30360,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.441066
    },
    {
      "epoch": 1.3009296945289406,
      "grad_norm": 4.894650936126709,
      "learning_rate": 8.420605831459285e-05,
      "loss": 2.451618957519531,
      "memory(GiB)": 77.56,
      "step": 30365,
      "token_acc": 0.45741324921135645,
      "train_speed(iter/s)": 1.441018
    },
    {
      "epoch": 1.3011439098581894,
      "grad_norm": 4.435378074645996,
      "learning_rate": 8.420114952112358e-05,
      "loss": 2.362919235229492,
      "memory(GiB)": 77.56,
      "step": 30370,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.441059
    },
    {
      "epoch": 1.3013581251874384,
      "grad_norm": 5.77326774597168,
      "learning_rate": 8.419624010806839e-05,
      "loss": 2.655732345581055,
      "memory(GiB)": 77.56,
      "step": 30375,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.441133
    },
    {
      "epoch": 1.3015723405166875,
      "grad_norm": 4.066826343536377,
      "learning_rate": 8.419133007551626e-05,
      "loss": 2.5782665252685546,
      "memory(GiB)": 77.56,
      "step": 30380,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.441145
    },
    {
      "epoch": 1.3017865558459363,
      "grad_norm": 7.82096004486084,
      "learning_rate": 8.418641942355608e-05,
      "loss": 2.4948854446411133,
      "memory(GiB)": 77.56,
      "step": 30385,
      "token_acc": 0.4847328244274809,
      "train_speed(iter/s)": 1.441091
    },
    {
      "epoch": 1.3020007711751853,
      "grad_norm": 5.382538318634033,
      "learning_rate": 8.418150815227686e-05,
      "loss": 2.6521175384521483,
      "memory(GiB)": 77.56,
      "step": 30390,
      "token_acc": 0.4622356495468278,
      "train_speed(iter/s)": 1.441049
    },
    {
      "epoch": 1.3022149865044343,
      "grad_norm": 4.82567834854126,
      "learning_rate": 8.417659626176754e-05,
      "loss": 2.624923324584961,
      "memory(GiB)": 77.56,
      "step": 30395,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.440993
    },
    {
      "epoch": 1.3024292018336832,
      "grad_norm": 5.103214740753174,
      "learning_rate": 8.417168375211713e-05,
      "loss": 2.4815736770629884,
      "memory(GiB)": 77.56,
      "step": 30400,
      "token_acc": 0.4984423676012461,
      "train_speed(iter/s)": 1.441065
    },
    {
      "epoch": 1.3026434171629322,
      "grad_norm": 5.527584075927734,
      "learning_rate": 8.41667706234146e-05,
      "loss": 2.3787628173828126,
      "memory(GiB)": 77.56,
      "step": 30405,
      "token_acc": 0.5103734439834025,
      "train_speed(iter/s)": 1.441112
    },
    {
      "epoch": 1.3028576324921812,
      "grad_norm": 7.315732002258301,
      "learning_rate": 8.416185687574898e-05,
      "loss": 2.5277336120605467,
      "memory(GiB)": 77.56,
      "step": 30410,
      "token_acc": 0.44912280701754387,
      "train_speed(iter/s)": 1.441159
    },
    {
      "epoch": 1.30307184782143,
      "grad_norm": 3.7005109786987305,
      "learning_rate": 8.415694250920927e-05,
      "loss": 2.5848033905029295,
      "memory(GiB)": 77.56,
      "step": 30415,
      "token_acc": 0.45244956772334294,
      "train_speed(iter/s)": 1.44111
    },
    {
      "epoch": 1.303286063150679,
      "grad_norm": 5.403597354888916,
      "learning_rate": 8.415202752388451e-05,
      "loss": 2.568196487426758,
      "memory(GiB)": 77.56,
      "step": 30420,
      "token_acc": 0.5102639296187683,
      "train_speed(iter/s)": 1.441115
    },
    {
      "epoch": 1.303500278479928,
      "grad_norm": 5.422910213470459,
      "learning_rate": 8.414711191986374e-05,
      "loss": 2.562359619140625,
      "memory(GiB)": 77.56,
      "step": 30425,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.441098
    },
    {
      "epoch": 1.303714493809177,
      "grad_norm": 8.326611518859863,
      "learning_rate": 8.4142195697236e-05,
      "loss": 2.673186683654785,
      "memory(GiB)": 77.56,
      "step": 30430,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.441134
    },
    {
      "epoch": 1.303928709138426,
      "grad_norm": 3.7338290214538574,
      "learning_rate": 8.413727885609036e-05,
      "loss": 2.2320104598999024,
      "memory(GiB)": 77.56,
      "step": 30435,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.441158
    },
    {
      "epoch": 1.304142924467675,
      "grad_norm": 4.473420143127441,
      "learning_rate": 8.413236139651588e-05,
      "loss": 2.5908416748046874,
      "memory(GiB)": 77.56,
      "step": 30440,
      "token_acc": 0.44542772861356933,
      "train_speed(iter/s)": 1.441131
    },
    {
      "epoch": 1.3043571397969238,
      "grad_norm": 7.763184547424316,
      "learning_rate": 8.412744331860167e-05,
      "loss": 2.4383451461791994,
      "memory(GiB)": 77.56,
      "step": 30445,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.441139
    },
    {
      "epoch": 1.3045713551261728,
      "grad_norm": 5.2608962059021,
      "learning_rate": 8.412252462243679e-05,
      "loss": 2.4560430526733397,
      "memory(GiB)": 77.56,
      "step": 30450,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.441092
    },
    {
      "epoch": 1.3047855704554219,
      "grad_norm": 5.97885274887085,
      "learning_rate": 8.411760530811038e-05,
      "loss": 2.6400531768798827,
      "memory(GiB)": 77.56,
      "step": 30455,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.441112
    },
    {
      "epoch": 1.3049997857846707,
      "grad_norm": 3.698084831237793,
      "learning_rate": 8.411268537571154e-05,
      "loss": 2.6130773544311525,
      "memory(GiB)": 77.56,
      "step": 30460,
      "token_acc": 0.444743935309973,
      "train_speed(iter/s)": 1.441154
    },
    {
      "epoch": 1.3052140011139197,
      "grad_norm": 6.4532270431518555,
      "learning_rate": 8.410776482532941e-05,
      "loss": 2.73227596282959,
      "memory(GiB)": 77.56,
      "step": 30465,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.441224
    },
    {
      "epoch": 1.3054282164431688,
      "grad_norm": 4.533722400665283,
      "learning_rate": 8.410284365705311e-05,
      "loss": 2.5809717178344727,
      "memory(GiB)": 77.56,
      "step": 30470,
      "token_acc": 0.4271523178807947,
      "train_speed(iter/s)": 1.441236
    },
    {
      "epoch": 1.3056424317724176,
      "grad_norm": 5.006070137023926,
      "learning_rate": 8.409792187097183e-05,
      "loss": 2.391727828979492,
      "memory(GiB)": 77.56,
      "step": 30475,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441206
    },
    {
      "epoch": 1.3058566471016666,
      "grad_norm": 3.820786714553833,
      "learning_rate": 8.409299946717469e-05,
      "loss": 2.6763145446777346,
      "memory(GiB)": 77.56,
      "step": 30480,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.441114
    },
    {
      "epoch": 1.3060708624309156,
      "grad_norm": 4.826564788818359,
      "learning_rate": 8.408807644575089e-05,
      "loss": 2.90842227935791,
      "memory(GiB)": 77.56,
      "step": 30485,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.441163
    },
    {
      "epoch": 1.3062850777601644,
      "grad_norm": 4.939174175262451,
      "learning_rate": 8.40831528067896e-05,
      "loss": 2.670140266418457,
      "memory(GiB)": 77.56,
      "step": 30490,
      "token_acc": 0.46494464944649444,
      "train_speed(iter/s)": 1.441174
    },
    {
      "epoch": 1.3064992930894135,
      "grad_norm": 4.767963409423828,
      "learning_rate": 8.407822855038004e-05,
      "loss": 2.3368053436279297,
      "memory(GiB)": 77.56,
      "step": 30495,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.441151
    },
    {
      "epoch": 1.3067135084186625,
      "grad_norm": 4.725824356079102,
      "learning_rate": 8.40733036766114e-05,
      "loss": 2.627589988708496,
      "memory(GiB)": 77.56,
      "step": 30500,
      "token_acc": 0.45878136200716846,
      "train_speed(iter/s)": 1.441137
    },
    {
      "epoch": 1.3067135084186625,
      "eval_loss": 2.191805601119995,
      "eval_runtime": 14.4793,
      "eval_samples_per_second": 6.906,
      "eval_steps_per_second": 6.906,
      "eval_token_acc": 0.498567335243553,
      "step": 30500
    },
    {
      "epoch": 1.3069277237479113,
      "grad_norm": 5.181138038635254,
      "learning_rate": 8.406837818557289e-05,
      "loss": 2.4915164947509765,
      "memory(GiB)": 77.56,
      "step": 30505,
      "token_acc": 0.49851924975320827,
      "train_speed(iter/s)": 1.440041
    },
    {
      "epoch": 1.3071419390771604,
      "grad_norm": 5.8320441246032715,
      "learning_rate": 8.406345207735375e-05,
      "loss": 2.4027130126953127,
      "memory(GiB)": 77.56,
      "step": 30510,
      "token_acc": 0.4485049833887043,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 1.3073561544064094,
      "grad_norm": 6.1275506019592285,
      "learning_rate": 8.405852535204323e-05,
      "loss": 2.534292984008789,
      "memory(GiB)": 77.56,
      "step": 30515,
      "token_acc": 0.4513888888888889,
      "train_speed(iter/s)": 1.440037
    },
    {
      "epoch": 1.3075703697356582,
      "grad_norm": 5.595847129821777,
      "learning_rate": 8.405359800973056e-05,
      "loss": 2.812118721008301,
      "memory(GiB)": 77.56,
      "step": 30520,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.440081
    },
    {
      "epoch": 1.3077845850649072,
      "grad_norm": 4.512628078460693,
      "learning_rate": 8.404867005050504e-05,
      "loss": 2.545647621154785,
      "memory(GiB)": 77.56,
      "step": 30525,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.440091
    },
    {
      "epoch": 1.3079988003941563,
      "grad_norm": 5.215756416320801,
      "learning_rate": 8.404374147445591e-05,
      "loss": 2.7159494400024413,
      "memory(GiB)": 77.56,
      "step": 30530,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.440092
    },
    {
      "epoch": 1.308213015723405,
      "grad_norm": 6.1765899658203125,
      "learning_rate": 8.403881228167245e-05,
      "loss": 2.8459245681762697,
      "memory(GiB)": 77.56,
      "step": 30535,
      "token_acc": 0.4735099337748344,
      "train_speed(iter/s)": 1.440111
    },
    {
      "epoch": 1.3084272310526541,
      "grad_norm": 5.327668190002441,
      "learning_rate": 8.4033882472244e-05,
      "loss": 2.5296817779541017,
      "memory(GiB)": 77.56,
      "step": 30540,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.440089
    },
    {
      "epoch": 1.3086414463819032,
      "grad_norm": 4.702518463134766,
      "learning_rate": 8.402895204625982e-05,
      "loss": 2.412190628051758,
      "memory(GiB)": 77.56,
      "step": 30545,
      "token_acc": 0.46438746438746437,
      "train_speed(iter/s)": 1.440101
    },
    {
      "epoch": 1.308855661711152,
      "grad_norm": 4.6268839836120605,
      "learning_rate": 8.402402100380927e-05,
      "loss": 2.1747419357299806,
      "memory(GiB)": 77.56,
      "step": 30550,
      "token_acc": 0.5510204081632653,
      "train_speed(iter/s)": 1.440075
    },
    {
      "epoch": 1.309069877040401,
      "grad_norm": 4.304233074188232,
      "learning_rate": 8.401908934498166e-05,
      "loss": 2.7301734924316405,
      "memory(GiB)": 77.56,
      "step": 30555,
      "token_acc": 0.43312101910828027,
      "train_speed(iter/s)": 1.440124
    },
    {
      "epoch": 1.30928409236965,
      "grad_norm": 4.967772006988525,
      "learning_rate": 8.401415706986634e-05,
      "loss": 2.75872917175293,
      "memory(GiB)": 77.56,
      "step": 30560,
      "token_acc": 0.4513888888888889,
      "train_speed(iter/s)": 1.44013
    },
    {
      "epoch": 1.3094983076988989,
      "grad_norm": 7.840235710144043,
      "learning_rate": 8.400922417855265e-05,
      "loss": 2.5490289688110352,
      "memory(GiB)": 77.56,
      "step": 30565,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.440087
    },
    {
      "epoch": 1.309712523028148,
      "grad_norm": 4.7686767578125,
      "learning_rate": 8.400429067112996e-05,
      "loss": 2.0024553298950196,
      "memory(GiB)": 77.56,
      "step": 30570,
      "token_acc": 0.5471014492753623,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 1.309926738357397,
      "grad_norm": 4.6255083084106445,
      "learning_rate": 8.399935654768765e-05,
      "loss": 2.648905944824219,
      "memory(GiB)": 77.56,
      "step": 30575,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.440083
    },
    {
      "epoch": 1.3101409536866457,
      "grad_norm": 4.743405342102051,
      "learning_rate": 8.399442180831509e-05,
      "loss": 2.6169843673706055,
      "memory(GiB)": 77.56,
      "step": 30580,
      "token_acc": 0.4601449275362319,
      "train_speed(iter/s)": 1.440091
    },
    {
      "epoch": 1.3103551690158948,
      "grad_norm": 5.546091079711914,
      "learning_rate": 8.398948645310169e-05,
      "loss": 2.365212821960449,
      "memory(GiB)": 77.56,
      "step": 30585,
      "token_acc": 0.4753521126760563,
      "train_speed(iter/s)": 1.440082
    },
    {
      "epoch": 1.3105693843451438,
      "grad_norm": 6.177235126495361,
      "learning_rate": 8.398455048213687e-05,
      "loss": 2.355448913574219,
      "memory(GiB)": 77.56,
      "step": 30590,
      "token_acc": 0.5079872204472844,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 1.3107835996743926,
      "grad_norm": 4.434200286865234,
      "learning_rate": 8.397961389551003e-05,
      "loss": 2.444801902770996,
      "memory(GiB)": 77.56,
      "step": 30595,
      "token_acc": 0.4629080118694362,
      "train_speed(iter/s)": 1.440144
    },
    {
      "epoch": 1.3109978150036417,
      "grad_norm": 4.852989196777344,
      "learning_rate": 8.39746766933106e-05,
      "loss": 2.4627485275268555,
      "memory(GiB)": 77.56,
      "step": 30600,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.440142
    },
    {
      "epoch": 1.3112120303328907,
      "grad_norm": 4.629029273986816,
      "learning_rate": 8.396973887562806e-05,
      "loss": 2.662624168395996,
      "memory(GiB)": 77.56,
      "step": 30605,
      "token_acc": 0.4336283185840708,
      "train_speed(iter/s)": 1.440146
    },
    {
      "epoch": 1.3114262456621395,
      "grad_norm": 5.103628158569336,
      "learning_rate": 8.396480044255181e-05,
      "loss": 2.5142938613891603,
      "memory(GiB)": 77.56,
      "step": 30610,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.440155
    },
    {
      "epoch": 1.3116404609913885,
      "grad_norm": 5.212147235870361,
      "learning_rate": 8.395986139417135e-05,
      "loss": 2.6481840133666994,
      "memory(GiB)": 77.56,
      "step": 30615,
      "token_acc": 0.44481605351170567,
      "train_speed(iter/s)": 1.440148
    },
    {
      "epoch": 1.3118546763206376,
      "grad_norm": 5.196325302124023,
      "learning_rate": 8.395492173057613e-05,
      "loss": 2.6507835388183594,
      "memory(GiB)": 77.56,
      "step": 30620,
      "token_acc": 0.5016501650165016,
      "train_speed(iter/s)": 1.440129
    },
    {
      "epoch": 1.3120688916498864,
      "grad_norm": 4.532564640045166,
      "learning_rate": 8.394998145185566e-05,
      "loss": 2.4831264495849608,
      "memory(GiB)": 77.56,
      "step": 30625,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.440154
    },
    {
      "epoch": 1.3122831069791354,
      "grad_norm": 6.058781147003174,
      "learning_rate": 8.394504055809944e-05,
      "loss": 2.5596569061279295,
      "memory(GiB)": 77.56,
      "step": 30630,
      "token_acc": 0.4517133956386293,
      "train_speed(iter/s)": 1.440196
    },
    {
      "epoch": 1.3124973223083845,
      "grad_norm": 5.254405498504639,
      "learning_rate": 8.394009904939695e-05,
      "loss": 2.311823844909668,
      "memory(GiB)": 77.56,
      "step": 30635,
      "token_acc": 0.47079037800687284,
      "train_speed(iter/s)": 1.440224
    },
    {
      "epoch": 1.3127115376376333,
      "grad_norm": 5.523976802825928,
      "learning_rate": 8.393515692583774e-05,
      "loss": 2.5997011184692385,
      "memory(GiB)": 77.56,
      "step": 30640,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.440205
    },
    {
      "epoch": 1.3129257529668823,
      "grad_norm": 6.062639236450195,
      "learning_rate": 8.393021418751132e-05,
      "loss": 2.713842582702637,
      "memory(GiB)": 77.56,
      "step": 30645,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.44027
    },
    {
      "epoch": 1.3131399682961313,
      "grad_norm": 4.327581405639648,
      "learning_rate": 8.392527083450723e-05,
      "loss": 2.296024513244629,
      "memory(GiB)": 77.56,
      "step": 30650,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.440255
    },
    {
      "epoch": 1.3133541836253801,
      "grad_norm": 4.298564434051514,
      "learning_rate": 8.392032686691504e-05,
      "loss": 2.3184110641479494,
      "memory(GiB)": 77.56,
      "step": 30655,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.440267
    },
    {
      "epoch": 1.3135683989546292,
      "grad_norm": 4.8839569091796875,
      "learning_rate": 8.391538228482432e-05,
      "loss": 2.8416332244873046,
      "memory(GiB)": 77.56,
      "step": 30660,
      "token_acc": 0.4457831325301205,
      "train_speed(iter/s)": 1.440236
    },
    {
      "epoch": 1.3137826142838782,
      "grad_norm": 5.313823223114014,
      "learning_rate": 8.391043708832463e-05,
      "loss": 2.5864084243774412,
      "memory(GiB)": 77.56,
      "step": 30665,
      "token_acc": 0.47039473684210525,
      "train_speed(iter/s)": 1.440296
    },
    {
      "epoch": 1.313996829613127,
      "grad_norm": 6.0417914390563965,
      "learning_rate": 8.390549127750558e-05,
      "loss": 2.6308626174926757,
      "memory(GiB)": 77.56,
      "step": 30670,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.440309
    },
    {
      "epoch": 1.314211044942376,
      "grad_norm": 3.9929563999176025,
      "learning_rate": 8.390054485245672e-05,
      "loss": 2.1022972106933593,
      "memory(GiB)": 77.56,
      "step": 30675,
      "token_acc": 0.5310344827586206,
      "train_speed(iter/s)": 1.440302
    },
    {
      "epoch": 1.314425260271625,
      "grad_norm": 5.207853317260742,
      "learning_rate": 8.38955978132677e-05,
      "loss": 2.3052059173583985,
      "memory(GiB)": 77.56,
      "step": 30680,
      "token_acc": 0.5226480836236934,
      "train_speed(iter/s)": 1.440262
    },
    {
      "epoch": 1.314639475600874,
      "grad_norm": 5.290105819702148,
      "learning_rate": 8.389065016002812e-05,
      "loss": 2.326431655883789,
      "memory(GiB)": 77.56,
      "step": 30685,
      "token_acc": 0.5165562913907285,
      "train_speed(iter/s)": 1.440246
    },
    {
      "epoch": 1.314853690930123,
      "grad_norm": 5.975013732910156,
      "learning_rate": 8.388570189282765e-05,
      "loss": 2.4550048828125,
      "memory(GiB)": 77.56,
      "step": 30690,
      "token_acc": 0.46096654275092935,
      "train_speed(iter/s)": 1.440226
    },
    {
      "epoch": 1.315067906259372,
      "grad_norm": 5.328584671020508,
      "learning_rate": 8.388075301175586e-05,
      "loss": 2.6362619400024414,
      "memory(GiB)": 77.56,
      "step": 30695,
      "token_acc": 0.4980237154150198,
      "train_speed(iter/s)": 1.440248
    },
    {
      "epoch": 1.3152821215886208,
      "grad_norm": 5.431981086730957,
      "learning_rate": 8.387580351690248e-05,
      "loss": 2.6616044998168946,
      "memory(GiB)": 77.56,
      "step": 30700,
      "token_acc": 0.4742268041237113,
      "train_speed(iter/s)": 1.440251
    },
    {
      "epoch": 1.3154963369178698,
      "grad_norm": 6.600261688232422,
      "learning_rate": 8.387085340835713e-05,
      "loss": 2.4584869384765624,
      "memory(GiB)": 77.56,
      "step": 30705,
      "token_acc": 0.47794117647058826,
      "train_speed(iter/s)": 1.440293
    },
    {
      "epoch": 1.3157105522471189,
      "grad_norm": 4.675459384918213,
      "learning_rate": 8.38659026862095e-05,
      "loss": 2.3372697830200195,
      "memory(GiB)": 77.56,
      "step": 30710,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.440316
    },
    {
      "epoch": 1.3159247675763677,
      "grad_norm": 5.866064548492432,
      "learning_rate": 8.386095135054927e-05,
      "loss": 2.628317642211914,
      "memory(GiB)": 77.56,
      "step": 30715,
      "token_acc": 0.496,
      "train_speed(iter/s)": 1.440316
    },
    {
      "epoch": 1.3161389829056167,
      "grad_norm": 5.0868659019470215,
      "learning_rate": 8.385599940146613e-05,
      "loss": 2.426080513000488,
      "memory(GiB)": 77.56,
      "step": 30720,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440264
    },
    {
      "epoch": 1.3163531982348657,
      "grad_norm": 4.216841220855713,
      "learning_rate": 8.385104683904983e-05,
      "loss": 2.5281440734863283,
      "memory(GiB)": 77.56,
      "step": 30725,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.440275
    },
    {
      "epoch": 1.3165674135641146,
      "grad_norm": 5.248867034912109,
      "learning_rate": 8.384609366339003e-05,
      "loss": 2.657305145263672,
      "memory(GiB)": 77.56,
      "step": 30730,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.440238
    },
    {
      "epoch": 1.3167816288933636,
      "grad_norm": 5.731428146362305,
      "learning_rate": 8.38411398745765e-05,
      "loss": 2.6855926513671875,
      "memory(GiB)": 77.56,
      "step": 30735,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.440248
    },
    {
      "epoch": 1.3169958442226126,
      "grad_norm": 5.230101108551025,
      "learning_rate": 8.383618547269901e-05,
      "loss": 2.4764209747314454,
      "memory(GiB)": 77.56,
      "step": 30740,
      "token_acc": 0.45907473309608543,
      "train_speed(iter/s)": 1.440261
    },
    {
      "epoch": 1.3172100595518614,
      "grad_norm": 3.928238868713379,
      "learning_rate": 8.383123045784724e-05,
      "loss": 2.663407897949219,
      "memory(GiB)": 77.56,
      "step": 30745,
      "token_acc": 0.4539249146757679,
      "train_speed(iter/s)": 1.440306
    },
    {
      "epoch": 1.3174242748811105,
      "grad_norm": 5.327571868896484,
      "learning_rate": 8.382627483011101e-05,
      "loss": 2.7467296600341795,
      "memory(GiB)": 77.56,
      "step": 30750,
      "token_acc": 0.4368932038834951,
      "train_speed(iter/s)": 1.440353
    },
    {
      "epoch": 1.3176384902103595,
      "grad_norm": 3.9081926345825195,
      "learning_rate": 8.382131858958009e-05,
      "loss": 2.473705291748047,
      "memory(GiB)": 77.56,
      "step": 30755,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.44033
    },
    {
      "epoch": 1.3178527055396083,
      "grad_norm": 4.930261135101318,
      "learning_rate": 8.381636173634425e-05,
      "loss": 2.481026840209961,
      "memory(GiB)": 77.56,
      "step": 30760,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.440364
    },
    {
      "epoch": 1.3180669208688574,
      "grad_norm": 4.128588676452637,
      "learning_rate": 8.381140427049328e-05,
      "loss": 2.537253570556641,
      "memory(GiB)": 77.56,
      "step": 30765,
      "token_acc": 0.4582043343653251,
      "train_speed(iter/s)": 1.440365
    },
    {
      "epoch": 1.3182811361981064,
      "grad_norm": 5.725658893585205,
      "learning_rate": 8.380644619211702e-05,
      "loss": 2.3736797332763673,
      "memory(GiB)": 77.56,
      "step": 30770,
      "token_acc": 0.5016949152542373,
      "train_speed(iter/s)": 1.440327
    },
    {
      "epoch": 1.3184953515273552,
      "grad_norm": 6.728057861328125,
      "learning_rate": 8.380148750130525e-05,
      "loss": 2.5051387786865233,
      "memory(GiB)": 77.56,
      "step": 30775,
      "token_acc": 0.4980694980694981,
      "train_speed(iter/s)": 1.44031
    },
    {
      "epoch": 1.3187095668566042,
      "grad_norm": 5.706851959228516,
      "learning_rate": 8.379652819814785e-05,
      "loss": 2.551733207702637,
      "memory(GiB)": 77.56,
      "step": 30780,
      "token_acc": 0.44694533762057875,
      "train_speed(iter/s)": 1.440297
    },
    {
      "epoch": 1.3189237821858533,
      "grad_norm": 5.804297924041748,
      "learning_rate": 8.379156828273463e-05,
      "loss": 2.484577941894531,
      "memory(GiB)": 77.56,
      "step": 30785,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.440245
    },
    {
      "epoch": 1.319137997515102,
      "grad_norm": 5.132512092590332,
      "learning_rate": 8.378660775515546e-05,
      "loss": 2.730057144165039,
      "memory(GiB)": 77.56,
      "step": 30790,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.440298
    },
    {
      "epoch": 1.3193522128443511,
      "grad_norm": 5.32224702835083,
      "learning_rate": 8.378164661550019e-05,
      "loss": 2.4412097930908203,
      "memory(GiB)": 77.56,
      "step": 30795,
      "token_acc": 0.5126353790613718,
      "train_speed(iter/s)": 1.440278
    },
    {
      "epoch": 1.3195664281736001,
      "grad_norm": 6.233191967010498,
      "learning_rate": 8.37766848638587e-05,
      "loss": 2.685907173156738,
      "memory(GiB)": 77.56,
      "step": 30800,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.4403
    },
    {
      "epoch": 1.319780643502849,
      "grad_norm": 5.858996868133545,
      "learning_rate": 8.377172250032086e-05,
      "loss": 2.5612987518310546,
      "memory(GiB)": 77.56,
      "step": 30805,
      "token_acc": 0.4752475247524752,
      "train_speed(iter/s)": 1.440257
    },
    {
      "epoch": 1.319994858832098,
      "grad_norm": 5.193116188049316,
      "learning_rate": 8.376675952497661e-05,
      "loss": 2.216260528564453,
      "memory(GiB)": 77.56,
      "step": 30810,
      "token_acc": 0.5241635687732342,
      "train_speed(iter/s)": 1.440288
    },
    {
      "epoch": 1.320209074161347,
      "grad_norm": 5.159361362457275,
      "learning_rate": 8.376179593791585e-05,
      "loss": 2.3703945159912108,
      "memory(GiB)": 77.56,
      "step": 30815,
      "token_acc": 0.4495798319327731,
      "train_speed(iter/s)": 1.440303
    },
    {
      "epoch": 1.3204232894905958,
      "grad_norm": 4.9706807136535645,
      "learning_rate": 8.375683173922846e-05,
      "loss": 2.440373992919922,
      "memory(GiB)": 77.56,
      "step": 30820,
      "token_acc": 0.5258064516129032,
      "train_speed(iter/s)": 1.440323
    },
    {
      "epoch": 1.3206375048198449,
      "grad_norm": 5.335484027862549,
      "learning_rate": 8.37518669290044e-05,
      "loss": 2.5746845245361327,
      "memory(GiB)": 77.56,
      "step": 30825,
      "token_acc": 0.4902597402597403,
      "train_speed(iter/s)": 1.440364
    },
    {
      "epoch": 1.320851720149094,
      "grad_norm": 5.414564609527588,
      "learning_rate": 8.374690150733362e-05,
      "loss": 2.8610759735107423,
      "memory(GiB)": 77.56,
      "step": 30830,
      "token_acc": 0.4088235294117647,
      "train_speed(iter/s)": 1.440363
    },
    {
      "epoch": 1.3210659354783427,
      "grad_norm": 5.210866451263428,
      "learning_rate": 8.374193547430605e-05,
      "loss": 2.3322175979614257,
      "memory(GiB)": 77.56,
      "step": 30835,
      "token_acc": 0.47794117647058826,
      "train_speed(iter/s)": 1.440411
    },
    {
      "epoch": 1.3212801508075918,
      "grad_norm": 4.217972755432129,
      "learning_rate": 8.373696883001166e-05,
      "loss": 2.41158447265625,
      "memory(GiB)": 77.56,
      "step": 30840,
      "token_acc": 0.4668769716088328,
      "train_speed(iter/s)": 1.440374
    },
    {
      "epoch": 1.3214943661368408,
      "grad_norm": 6.352040767669678,
      "learning_rate": 8.373200157454044e-05,
      "loss": 2.4626605987548826,
      "memory(GiB)": 77.56,
      "step": 30845,
      "token_acc": 0.5098684210526315,
      "train_speed(iter/s)": 1.440384
    },
    {
      "epoch": 1.3217085814660896,
      "grad_norm": 5.207592964172363,
      "learning_rate": 8.372703370798239e-05,
      "loss": 2.6127735137939454,
      "memory(GiB)": 77.56,
      "step": 30850,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.440425
    },
    {
      "epoch": 1.3219227967953386,
      "grad_norm": 5.286231994628906,
      "learning_rate": 8.372206523042746e-05,
      "loss": 2.353206825256348,
      "memory(GiB)": 77.56,
      "step": 30855,
      "token_acc": 0.511326860841424,
      "train_speed(iter/s)": 1.440439
    },
    {
      "epoch": 1.3221370121245877,
      "grad_norm": 6.365725040435791,
      "learning_rate": 8.371709614196568e-05,
      "loss": 2.554417037963867,
      "memory(GiB)": 77.56,
      "step": 30860,
      "token_acc": 0.494949494949495,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.3223512274538365,
      "grad_norm": 5.4799981117248535,
      "learning_rate": 8.371212644268709e-05,
      "loss": 2.3065364837646483,
      "memory(GiB)": 77.56,
      "step": 30865,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.440512
    },
    {
      "epoch": 1.3225654427830855,
      "grad_norm": 4.490663051605225,
      "learning_rate": 8.370715613268171e-05,
      "loss": 2.422890853881836,
      "memory(GiB)": 77.56,
      "step": 30870,
      "token_acc": 0.483271375464684,
      "train_speed(iter/s)": 1.440524
    },
    {
      "epoch": 1.3227796581123346,
      "grad_norm": 3.842900276184082,
      "learning_rate": 8.370218521203956e-05,
      "loss": 2.372048568725586,
      "memory(GiB)": 77.56,
      "step": 30875,
      "token_acc": 0.5080385852090032,
      "train_speed(iter/s)": 1.440565
    },
    {
      "epoch": 1.3229938734415834,
      "grad_norm": 5.78187894821167,
      "learning_rate": 8.369721368085072e-05,
      "loss": 2.3127971649169923,
      "memory(GiB)": 77.56,
      "step": 30880,
      "token_acc": 0.5251798561151079,
      "train_speed(iter/s)": 1.440593
    },
    {
      "epoch": 1.3232080887708324,
      "grad_norm": 5.000535488128662,
      "learning_rate": 8.369224153920523e-05,
      "loss": 2.654227066040039,
      "memory(GiB)": 77.56,
      "step": 30885,
      "token_acc": 0.44569288389513106,
      "train_speed(iter/s)": 1.440615
    },
    {
      "epoch": 1.3234223041000814,
      "grad_norm": 5.567465782165527,
      "learning_rate": 8.36872687871932e-05,
      "loss": 2.454271697998047,
      "memory(GiB)": 77.56,
      "step": 30890,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.440639
    },
    {
      "epoch": 1.3236365194293302,
      "grad_norm": 4.689681529998779,
      "learning_rate": 8.368229542490468e-05,
      "loss": 2.3774295806884767,
      "memory(GiB)": 77.56,
      "step": 30895,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.440637
    },
    {
      "epoch": 1.3238507347585793,
      "grad_norm": 5.641341209411621,
      "learning_rate": 8.36773214524298e-05,
      "loss": 2.691015434265137,
      "memory(GiB)": 77.56,
      "step": 30900,
      "token_acc": 0.4359861591695502,
      "train_speed(iter/s)": 1.44067
    },
    {
      "epoch": 1.3240649500878283,
      "grad_norm": 5.560390472412109,
      "learning_rate": 8.367234686985862e-05,
      "loss": 2.8286022186279296,
      "memory(GiB)": 77.56,
      "step": 30905,
      "token_acc": 0.4453125,
      "train_speed(iter/s)": 1.440639
    },
    {
      "epoch": 1.3242791654170771,
      "grad_norm": 4.860203742980957,
      "learning_rate": 8.366737167728132e-05,
      "loss": 2.6245336532592773,
      "memory(GiB)": 77.56,
      "step": 30910,
      "token_acc": 0.4384057971014493,
      "train_speed(iter/s)": 1.440635
    },
    {
      "epoch": 1.3244933807463262,
      "grad_norm": 4.59218692779541,
      "learning_rate": 8.366239587478797e-05,
      "loss": 2.4169998168945312,
      "memory(GiB)": 77.56,
      "step": 30915,
      "token_acc": 0.43641618497109824,
      "train_speed(iter/s)": 1.440628
    },
    {
      "epoch": 1.3247075960755752,
      "grad_norm": 4.9459757804870605,
      "learning_rate": 8.365741946246875e-05,
      "loss": 2.3696567535400392,
      "memory(GiB)": 77.56,
      "step": 30920,
      "token_acc": 0.4566929133858268,
      "train_speed(iter/s)": 1.440609
    },
    {
      "epoch": 1.324921811404824,
      "grad_norm": 6.135128974914551,
      "learning_rate": 8.365244244041382e-05,
      "loss": 2.5217370986938477,
      "memory(GiB)": 77.56,
      "step": 30925,
      "token_acc": 0.49523809523809526,
      "train_speed(iter/s)": 1.440619
    },
    {
      "epoch": 1.325136026734073,
      "grad_norm": 4.655352592468262,
      "learning_rate": 8.364746480871331e-05,
      "loss": 2.4324872970581053,
      "memory(GiB)": 77.56,
      "step": 30930,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.440619
    },
    {
      "epoch": 1.325350242063322,
      "grad_norm": 9.142215728759766,
      "learning_rate": 8.364248656745741e-05,
      "loss": 2.2758420944213866,
      "memory(GiB)": 77.56,
      "step": 30935,
      "token_acc": 0.49795918367346936,
      "train_speed(iter/s)": 1.440624
    },
    {
      "epoch": 1.325564457392571,
      "grad_norm": 4.344929218292236,
      "learning_rate": 8.36375077167363e-05,
      "loss": 2.3214689254760743,
      "memory(GiB)": 77.56,
      "step": 30940,
      "token_acc": 0.5257731958762887,
      "train_speed(iter/s)": 1.440613
    },
    {
      "epoch": 1.32577867272182,
      "grad_norm": 5.71087121963501,
      "learning_rate": 8.36325282566402e-05,
      "loss": 2.453940200805664,
      "memory(GiB)": 77.56,
      "step": 30945,
      "token_acc": 0.5186567164179104,
      "train_speed(iter/s)": 1.440659
    },
    {
      "epoch": 1.325992888051069,
      "grad_norm": 5.22177267074585,
      "learning_rate": 8.36275481872593e-05,
      "loss": 2.5214260101318358,
      "memory(GiB)": 77.56,
      "step": 30950,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.440644
    },
    {
      "epoch": 1.3262071033803178,
      "grad_norm": 4.714686870574951,
      "learning_rate": 8.362256750868382e-05,
      "loss": 2.5031557083129883,
      "memory(GiB)": 77.56,
      "step": 30955,
      "token_acc": 0.5050167224080268,
      "train_speed(iter/s)": 1.440648
    },
    {
      "epoch": 1.3264213187095668,
      "grad_norm": 3.841874361038208,
      "learning_rate": 8.361758622100399e-05,
      "loss": 2.238236999511719,
      "memory(GiB)": 77.56,
      "step": 30960,
      "token_acc": 0.5240793201133145,
      "train_speed(iter/s)": 1.440693
    },
    {
      "epoch": 1.3266355340388158,
      "grad_norm": 5.512101173400879,
      "learning_rate": 8.361260432431003e-05,
      "loss": 2.4659755706787108,
      "memory(GiB)": 77.56,
      "step": 30965,
      "token_acc": 0.47678018575851394,
      "train_speed(iter/s)": 1.440676
    },
    {
      "epoch": 1.3268497493680647,
      "grad_norm": 3.927079677581787,
      "learning_rate": 8.360762181869224e-05,
      "loss": 2.7166528701782227,
      "memory(GiB)": 77.56,
      "step": 30970,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.440705
    },
    {
      "epoch": 1.3270639646973137,
      "grad_norm": 4.834190845489502,
      "learning_rate": 8.360263870424083e-05,
      "loss": 2.770062255859375,
      "memory(GiB)": 77.56,
      "step": 30975,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.440701
    },
    {
      "epoch": 1.3272781800265627,
      "grad_norm": 4.216663360595703,
      "learning_rate": 8.359765498104612e-05,
      "loss": 2.662898635864258,
      "memory(GiB)": 77.56,
      "step": 30980,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.440663
    },
    {
      "epoch": 1.3274923953558115,
      "grad_norm": 5.50461483001709,
      "learning_rate": 8.359267064919836e-05,
      "loss": 2.662904167175293,
      "memory(GiB)": 77.56,
      "step": 30985,
      "token_acc": 0.48286604361370716,
      "train_speed(iter/s)": 1.440637
    },
    {
      "epoch": 1.3277066106850606,
      "grad_norm": 4.458168983459473,
      "learning_rate": 8.358768570878786e-05,
      "loss": 2.64044132232666,
      "memory(GiB)": 77.56,
      "step": 30990,
      "token_acc": 0.46546546546546547,
      "train_speed(iter/s)": 1.440681
    },
    {
      "epoch": 1.3279208260143096,
      "grad_norm": 4.690812110900879,
      "learning_rate": 8.358270015990492e-05,
      "loss": 2.4562482833862305,
      "memory(GiB)": 77.56,
      "step": 30995,
      "token_acc": 0.5111940298507462,
      "train_speed(iter/s)": 1.440666
    },
    {
      "epoch": 1.3281350413435584,
      "grad_norm": 7.401129722595215,
      "learning_rate": 8.357771400263988e-05,
      "loss": 2.774020767211914,
      "memory(GiB)": 77.56,
      "step": 31000,
      "token_acc": 0.4338235294117647,
      "train_speed(iter/s)": 1.4407
    },
    {
      "epoch": 1.3281350413435584,
      "eval_loss": 2.1634511947631836,
      "eval_runtime": 13.8855,
      "eval_samples_per_second": 7.202,
      "eval_steps_per_second": 7.202,
      "eval_token_acc": 0.48216833095577744,
      "step": 31000
    },
    {
      "epoch": 1.3283492566728075,
      "grad_norm": 6.538474082946777,
      "learning_rate": 8.357272723708303e-05,
      "loss": 2.881578826904297,
      "memory(GiB)": 77.56,
      "step": 31005,
      "token_acc": 0.4655674102812803,
      "train_speed(iter/s)": 1.439742
    },
    {
      "epoch": 1.3285634720020565,
      "grad_norm": 4.092084884643555,
      "learning_rate": 8.356773986332475e-05,
      "loss": 2.7259519577026365,
      "memory(GiB)": 77.56,
      "step": 31010,
      "token_acc": 0.4402332361516035,
      "train_speed(iter/s)": 1.439772
    },
    {
      "epoch": 1.3287776873313053,
      "grad_norm": 5.630678653717041,
      "learning_rate": 8.356275188145537e-05,
      "loss": 2.4747346878051757,
      "memory(GiB)": 77.56,
      "step": 31015,
      "token_acc": 0.4919093851132686,
      "train_speed(iter/s)": 1.439716
    },
    {
      "epoch": 1.3289919026605543,
      "grad_norm": 4.750728130340576,
      "learning_rate": 8.355776329156525e-05,
      "loss": 2.626193618774414,
      "memory(GiB)": 77.56,
      "step": 31020,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.439717
    },
    {
      "epoch": 1.3292061179898034,
      "grad_norm": 4.294435977935791,
      "learning_rate": 8.355277409374477e-05,
      "loss": 2.298265838623047,
      "memory(GiB)": 77.56,
      "step": 31025,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.439788
    },
    {
      "epoch": 1.3294203333190522,
      "grad_norm": 8.400869369506836,
      "learning_rate": 8.354778428808431e-05,
      "loss": 2.7615495681762696,
      "memory(GiB)": 77.56,
      "step": 31030,
      "token_acc": 0.44089456869009586,
      "train_speed(iter/s)": 1.439843
    },
    {
      "epoch": 1.3296345486483012,
      "grad_norm": 4.415702819824219,
      "learning_rate": 8.354279387467427e-05,
      "loss": 2.6714847564697264,
      "memory(GiB)": 77.56,
      "step": 31035,
      "token_acc": 0.44126074498567336,
      "train_speed(iter/s)": 1.439927
    },
    {
      "epoch": 1.3298487639775503,
      "grad_norm": 4.608646869659424,
      "learning_rate": 8.353780285360505e-05,
      "loss": 2.430695915222168,
      "memory(GiB)": 77.56,
      "step": 31040,
      "token_acc": 0.5103857566765578,
      "train_speed(iter/s)": 1.439964
    },
    {
      "epoch": 1.330062979306799,
      "grad_norm": 4.7237725257873535,
      "learning_rate": 8.353281122496708e-05,
      "loss": 2.490535926818848,
      "memory(GiB)": 77.56,
      "step": 31045,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.439924
    },
    {
      "epoch": 1.330277194636048,
      "grad_norm": 4.348401069641113,
      "learning_rate": 8.352781898885076e-05,
      "loss": 2.503744697570801,
      "memory(GiB)": 77.56,
      "step": 31050,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.4399
    },
    {
      "epoch": 1.3304914099652971,
      "grad_norm": 5.597479343414307,
      "learning_rate": 8.352282614534657e-05,
      "loss": 2.297476577758789,
      "memory(GiB)": 77.56,
      "step": 31055,
      "token_acc": 0.48068669527896996,
      "train_speed(iter/s)": 1.439909
    },
    {
      "epoch": 1.330705625294546,
      "grad_norm": 4.203425884246826,
      "learning_rate": 8.351783269454493e-05,
      "loss": 2.5553943634033205,
      "memory(GiB)": 77.56,
      "step": 31060,
      "token_acc": 0.46283783783783783,
      "train_speed(iter/s)": 1.439911
    },
    {
      "epoch": 1.330919840623795,
      "grad_norm": 5.708987712860107,
      "learning_rate": 8.35128386365363e-05,
      "loss": 2.604556083679199,
      "memory(GiB)": 77.56,
      "step": 31065,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439913
    },
    {
      "epoch": 1.331134055953044,
      "grad_norm": 4.4014129638671875,
      "learning_rate": 8.350784397141116e-05,
      "loss": 2.885649871826172,
      "memory(GiB)": 77.56,
      "step": 31070,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.439968
    },
    {
      "epoch": 1.3313482712822928,
      "grad_norm": 6.514688014984131,
      "learning_rate": 8.350284869926e-05,
      "loss": 2.5280813217163085,
      "memory(GiB)": 77.56,
      "step": 31075,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.439979
    },
    {
      "epoch": 1.3315624866115419,
      "grad_norm": 5.240673542022705,
      "learning_rate": 8.349785282017332e-05,
      "loss": 2.4918291091918947,
      "memory(GiB)": 77.56,
      "step": 31080,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.439985
    },
    {
      "epoch": 1.331776701940791,
      "grad_norm": 4.627394199371338,
      "learning_rate": 8.34928563342416e-05,
      "loss": 2.419125556945801,
      "memory(GiB)": 77.56,
      "step": 31085,
      "token_acc": 0.5080906148867314,
      "train_speed(iter/s)": 1.440046
    },
    {
      "epoch": 1.3319909172700397,
      "grad_norm": 6.271434783935547,
      "learning_rate": 8.348785924155536e-05,
      "loss": 2.270029067993164,
      "memory(GiB)": 77.56,
      "step": 31090,
      "token_acc": 0.501577287066246,
      "train_speed(iter/s)": 1.440052
    },
    {
      "epoch": 1.3322051325992887,
      "grad_norm": 4.423927307128906,
      "learning_rate": 8.348286154220514e-05,
      "loss": 2.1954986572265627,
      "memory(GiB)": 77.56,
      "step": 31095,
      "token_acc": 0.5508196721311476,
      "train_speed(iter/s)": 1.440094
    },
    {
      "epoch": 1.3324193479285378,
      "grad_norm": 4.762854099273682,
      "learning_rate": 8.347786323628149e-05,
      "loss": 2.633818817138672,
      "memory(GiB)": 77.56,
      "step": 31100,
      "token_acc": 0.4805194805194805,
      "train_speed(iter/s)": 1.440139
    },
    {
      "epoch": 1.3326335632577866,
      "grad_norm": 4.1464619636535645,
      "learning_rate": 8.347286432387493e-05,
      "loss": 2.3386953353881834,
      "memory(GiB)": 77.56,
      "step": 31105,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.440128
    },
    {
      "epoch": 1.3328477785870356,
      "grad_norm": 4.485672473907471,
      "learning_rate": 8.346786480507603e-05,
      "loss": 2.4894718170166015,
      "memory(GiB)": 77.56,
      "step": 31110,
      "token_acc": 0.5014925373134328,
      "train_speed(iter/s)": 1.440113
    },
    {
      "epoch": 1.3330619939162847,
      "grad_norm": 3.9842441082000732,
      "learning_rate": 8.346286467997536e-05,
      "loss": 2.7668119430541993,
      "memory(GiB)": 77.56,
      "step": 31115,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.440154
    },
    {
      "epoch": 1.3332762092455335,
      "grad_norm": 4.314476013183594,
      "learning_rate": 8.345786394866352e-05,
      "loss": 2.3545547485351563,
      "memory(GiB)": 77.56,
      "step": 31120,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.440184
    },
    {
      "epoch": 1.3334904245747825,
      "grad_norm": 5.358580589294434,
      "learning_rate": 8.345286261123108e-05,
      "loss": 2.2670669555664062,
      "memory(GiB)": 77.56,
      "step": 31125,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.440219
    },
    {
      "epoch": 1.3337046399040315,
      "grad_norm": 4.57749080657959,
      "learning_rate": 8.344786066776865e-05,
      "loss": 2.4415740966796875,
      "memory(GiB)": 77.56,
      "step": 31130,
      "token_acc": 0.4697986577181208,
      "train_speed(iter/s)": 1.440265
    },
    {
      "epoch": 1.3339188552332806,
      "grad_norm": 4.659688949584961,
      "learning_rate": 8.344285811836686e-05,
      "loss": 2.528257942199707,
      "memory(GiB)": 77.56,
      "step": 31135,
      "token_acc": 0.4823943661971831,
      "train_speed(iter/s)": 1.440296
    },
    {
      "epoch": 1.3341330705625294,
      "grad_norm": 4.4153008460998535,
      "learning_rate": 8.34378549631163e-05,
      "loss": 2.794023895263672,
      "memory(GiB)": 77.56,
      "step": 31140,
      "token_acc": 0.4129692832764505,
      "train_speed(iter/s)": 1.440228
    },
    {
      "epoch": 1.3343472858917784,
      "grad_norm": 5.269262790679932,
      "learning_rate": 8.343285120210764e-05,
      "loss": 2.438818168640137,
      "memory(GiB)": 77.56,
      "step": 31145,
      "token_acc": 0.48184818481848185,
      "train_speed(iter/s)": 1.440265
    },
    {
      "epoch": 1.3345615012210275,
      "grad_norm": 4.74795389175415,
      "learning_rate": 8.342784683543152e-05,
      "loss": 2.4837642669677735,
      "memory(GiB)": 77.56,
      "step": 31150,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.4403
    },
    {
      "epoch": 1.3347757165502763,
      "grad_norm": 7.848414897918701,
      "learning_rate": 8.342284186317862e-05,
      "loss": 2.4641300201416017,
      "memory(GiB)": 77.56,
      "step": 31155,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.440292
    },
    {
      "epoch": 1.3349899318795253,
      "grad_norm": 4.578596115112305,
      "learning_rate": 8.341783628543956e-05,
      "loss": 2.5964681625366213,
      "memory(GiB)": 77.56,
      "step": 31160,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.440277
    },
    {
      "epoch": 1.3352041472087743,
      "grad_norm": 5.1937174797058105,
      "learning_rate": 8.341283010230506e-05,
      "loss": 2.5030731201171874,
      "memory(GiB)": 77.56,
      "step": 31165,
      "token_acc": 0.4437299035369775,
      "train_speed(iter/s)": 1.44029
    },
    {
      "epoch": 1.3354183625380232,
      "grad_norm": 5.432137489318848,
      "learning_rate": 8.340782331386578e-05,
      "loss": 2.3957469940185545,
      "memory(GiB)": 77.56,
      "step": 31170,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.440284
    },
    {
      "epoch": 1.3356325778672722,
      "grad_norm": 5.077791690826416,
      "learning_rate": 8.340281592021246e-05,
      "loss": 2.755970001220703,
      "memory(GiB)": 77.56,
      "step": 31175,
      "token_acc": 0.415625,
      "train_speed(iter/s)": 1.440301
    },
    {
      "epoch": 1.3358467931965212,
      "grad_norm": 3.7372591495513916,
      "learning_rate": 8.339780792143578e-05,
      "loss": 2.623807907104492,
      "memory(GiB)": 77.56,
      "step": 31180,
      "token_acc": 0.47112462006079026,
      "train_speed(iter/s)": 1.44026
    },
    {
      "epoch": 1.33606100852577,
      "grad_norm": 6.986464977264404,
      "learning_rate": 8.33927993176265e-05,
      "loss": 2.3935569763183593,
      "memory(GiB)": 77.56,
      "step": 31185,
      "token_acc": 0.5020408163265306,
      "train_speed(iter/s)": 1.440326
    },
    {
      "epoch": 1.336275223855019,
      "grad_norm": 5.754143714904785,
      "learning_rate": 8.338779010887532e-05,
      "loss": 2.685280990600586,
      "memory(GiB)": 77.56,
      "step": 31190,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.440249
    },
    {
      "epoch": 1.336489439184268,
      "grad_norm": 5.6976752281188965,
      "learning_rate": 8.3382780295273e-05,
      "loss": 2.590597152709961,
      "memory(GiB)": 77.56,
      "step": 31195,
      "token_acc": 0.44329896907216493,
      "train_speed(iter/s)": 1.440265
    },
    {
      "epoch": 1.336703654513517,
      "grad_norm": 7.17396879196167,
      "learning_rate": 8.337776987691031e-05,
      "loss": 2.7046672821044924,
      "memory(GiB)": 77.56,
      "step": 31200,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.440273
    },
    {
      "epoch": 1.336917869842766,
      "grad_norm": 4.85002326965332,
      "learning_rate": 8.3372758853878e-05,
      "loss": 2.267042541503906,
      "memory(GiB)": 77.56,
      "step": 31205,
      "token_acc": 0.5138339920948617,
      "train_speed(iter/s)": 1.440339
    },
    {
      "epoch": 1.337132085172015,
      "grad_norm": 8.528756141662598,
      "learning_rate": 8.336774722626685e-05,
      "loss": 2.5631269454956054,
      "memory(GiB)": 77.56,
      "step": 31210,
      "token_acc": 0.453781512605042,
      "train_speed(iter/s)": 1.440348
    },
    {
      "epoch": 1.3373463005012638,
      "grad_norm": 4.431490898132324,
      "learning_rate": 8.336273499416767e-05,
      "loss": 2.7103233337402344,
      "memory(GiB)": 77.56,
      "step": 31215,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.440391
    },
    {
      "epoch": 1.3375605158305128,
      "grad_norm": 4.47111177444458,
      "learning_rate": 8.335772215767125e-05,
      "loss": 2.5284324645996095,
      "memory(GiB)": 77.56,
      "step": 31220,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.44047
    },
    {
      "epoch": 1.3377747311597619,
      "grad_norm": 5.26793098449707,
      "learning_rate": 8.33527087168684e-05,
      "loss": 2.1152149200439454,
      "memory(GiB)": 77.56,
      "step": 31225,
      "token_acc": 0.5303030303030303,
      "train_speed(iter/s)": 1.440465
    },
    {
      "epoch": 1.3379889464890107,
      "grad_norm": 3.017547845840454,
      "learning_rate": 8.334769467184992e-05,
      "loss": 2.404205894470215,
      "memory(GiB)": 77.56,
      "step": 31230,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.440497
    },
    {
      "epoch": 1.3382031618182597,
      "grad_norm": 4.037070274353027,
      "learning_rate": 8.334268002270671e-05,
      "loss": 2.6690139770507812,
      "memory(GiB)": 77.56,
      "step": 31235,
      "token_acc": 0.4783950617283951,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.3384173771475087,
      "grad_norm": 3.340653419494629,
      "learning_rate": 8.333766476952955e-05,
      "loss": 2.2853366851806642,
      "memory(GiB)": 77.56,
      "step": 31240,
      "token_acc": 0.5190311418685121,
      "train_speed(iter/s)": 1.440526
    },
    {
      "epoch": 1.3386315924767576,
      "grad_norm": 4.789028644561768,
      "learning_rate": 8.333264891240933e-05,
      "loss": 2.578736686706543,
      "memory(GiB)": 77.56,
      "step": 31245,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440584
    },
    {
      "epoch": 1.3388458078060066,
      "grad_norm": 5.581756591796875,
      "learning_rate": 8.33276324514369e-05,
      "loss": 2.5892478942871096,
      "memory(GiB)": 77.56,
      "step": 31250,
      "token_acc": 0.47384615384615386,
      "train_speed(iter/s)": 1.4406
    },
    {
      "epoch": 1.3390600231352556,
      "grad_norm": 4.968892574310303,
      "learning_rate": 8.332261538670313e-05,
      "loss": 2.659806251525879,
      "memory(GiB)": 77.56,
      "step": 31255,
      "token_acc": 0.4740061162079511,
      "train_speed(iter/s)": 1.440609
    },
    {
      "epoch": 1.3392742384645044,
      "grad_norm": 5.407258033752441,
      "learning_rate": 8.331759771829893e-05,
      "loss": 2.676869201660156,
      "memory(GiB)": 77.56,
      "step": 31260,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.440628
    },
    {
      "epoch": 1.3394884537937535,
      "grad_norm": 4.9764533042907715,
      "learning_rate": 8.331257944631521e-05,
      "loss": 2.3965240478515626,
      "memory(GiB)": 77.56,
      "step": 31265,
      "token_acc": 0.4971751412429379,
      "train_speed(iter/s)": 1.440591
    },
    {
      "epoch": 1.3397026691230025,
      "grad_norm": 6.027521133422852,
      "learning_rate": 8.330756057084284e-05,
      "loss": 2.6725067138671874,
      "memory(GiB)": 77.56,
      "step": 31270,
      "token_acc": 0.4349442379182156,
      "train_speed(iter/s)": 1.440582
    },
    {
      "epoch": 1.3399168844522513,
      "grad_norm": 4.616183280944824,
      "learning_rate": 8.330254109197279e-05,
      "loss": 2.5812007904052736,
      "memory(GiB)": 77.56,
      "step": 31275,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.440636
    },
    {
      "epoch": 1.3401310997815004,
      "grad_norm": 4.795625686645508,
      "learning_rate": 8.329752100979595e-05,
      "loss": 2.7311847686767576,
      "memory(GiB)": 77.56,
      "step": 31280,
      "token_acc": 0.46229508196721314,
      "train_speed(iter/s)": 1.440696
    },
    {
      "epoch": 1.3403453151107494,
      "grad_norm": 4.979424476623535,
      "learning_rate": 8.329250032440329e-05,
      "loss": 2.5712642669677734,
      "memory(GiB)": 77.56,
      "step": 31285,
      "token_acc": 0.43853820598006643,
      "train_speed(iter/s)": 1.440716
    },
    {
      "epoch": 1.3405595304399982,
      "grad_norm": 5.80098295211792,
      "learning_rate": 8.328747903588575e-05,
      "loss": 2.408363914489746,
      "memory(GiB)": 77.56,
      "step": 31290,
      "token_acc": 0.4618181818181818,
      "train_speed(iter/s)": 1.440706
    },
    {
      "epoch": 1.3407737457692472,
      "grad_norm": 5.018841743469238,
      "learning_rate": 8.32824571443343e-05,
      "loss": 2.5181486129760744,
      "memory(GiB)": 77.56,
      "step": 31295,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.44077
    },
    {
      "epoch": 1.3409879610984963,
      "grad_norm": 4.644965648651123,
      "learning_rate": 8.327743464983992e-05,
      "loss": 2.5715700149536134,
      "memory(GiB)": 77.56,
      "step": 31300,
      "token_acc": 0.42507645259938837,
      "train_speed(iter/s)": 1.440799
    },
    {
      "epoch": 1.3412021764277453,
      "grad_norm": 6.5134782791137695,
      "learning_rate": 8.327241155249361e-05,
      "loss": 2.4259647369384765,
      "memory(GiB)": 77.56,
      "step": 31305,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.440825
    },
    {
      "epoch": 1.3414163917569941,
      "grad_norm": 4.758327007293701,
      "learning_rate": 8.326738785238633e-05,
      "loss": 2.514888381958008,
      "memory(GiB)": 77.56,
      "step": 31310,
      "token_acc": 0.480225988700565,
      "train_speed(iter/s)": 1.440911
    },
    {
      "epoch": 1.3416306070862432,
      "grad_norm": 3.7796146869659424,
      "learning_rate": 8.326236354960912e-05,
      "loss": 2.4515625,
      "memory(GiB)": 77.56,
      "step": 31315,
      "token_acc": 0.4713896457765668,
      "train_speed(iter/s)": 1.440893
    },
    {
      "epoch": 1.3418448224154922,
      "grad_norm": 4.210323810577393,
      "learning_rate": 8.325733864425299e-05,
      "loss": 2.2994077682495115,
      "memory(GiB)": 77.56,
      "step": 31320,
      "token_acc": 0.532051282051282,
      "train_speed(iter/s)": 1.440919
    },
    {
      "epoch": 1.342059037744741,
      "grad_norm": 5.029773235321045,
      "learning_rate": 8.325231313640898e-05,
      "loss": 2.553785705566406,
      "memory(GiB)": 77.56,
      "step": 31325,
      "token_acc": 0.47648902821316613,
      "train_speed(iter/s)": 1.440931
    },
    {
      "epoch": 1.34227325307399,
      "grad_norm": 4.7754902839660645,
      "learning_rate": 8.32472870261681e-05,
      "loss": 2.460570526123047,
      "memory(GiB)": 77.56,
      "step": 31330,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.440938
    },
    {
      "epoch": 1.342487468403239,
      "grad_norm": 5.723888874053955,
      "learning_rate": 8.324226031362146e-05,
      "loss": 2.6780044555664064,
      "memory(GiB)": 77.56,
      "step": 31335,
      "token_acc": 0.4558303886925795,
      "train_speed(iter/s)": 1.44097
    },
    {
      "epoch": 1.3427016837324879,
      "grad_norm": 4.395916938781738,
      "learning_rate": 8.323723299886008e-05,
      "loss": 2.5784963607788085,
      "memory(GiB)": 77.56,
      "step": 31340,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.441018
    },
    {
      "epoch": 1.342915899061737,
      "grad_norm": 4.826033592224121,
      "learning_rate": 8.323220508197505e-05,
      "loss": 2.4980697631835938,
      "memory(GiB)": 77.56,
      "step": 31345,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.441074
    },
    {
      "epoch": 1.343130114390986,
      "grad_norm": 4.233396530151367,
      "learning_rate": 8.322717656305744e-05,
      "loss": 2.4995664596557616,
      "memory(GiB)": 77.56,
      "step": 31350,
      "token_acc": 0.4764705882352941,
      "train_speed(iter/s)": 1.441102
    },
    {
      "epoch": 1.3433443297202348,
      "grad_norm": 5.334184646606445,
      "learning_rate": 8.322214744219837e-05,
      "loss": 2.3861522674560547,
      "memory(GiB)": 77.56,
      "step": 31355,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.441115
    },
    {
      "epoch": 1.3435585450494838,
      "grad_norm": 5.229465484619141,
      "learning_rate": 8.32171177194889e-05,
      "loss": 2.5869619369506838,
      "memory(GiB)": 77.56,
      "step": 31360,
      "token_acc": 0.4744525547445255,
      "train_speed(iter/s)": 1.441161
    },
    {
      "epoch": 1.3437727603787328,
      "grad_norm": 6.362223148345947,
      "learning_rate": 8.32120873950202e-05,
      "loss": 2.609721565246582,
      "memory(GiB)": 77.56,
      "step": 31365,
      "token_acc": 0.4608150470219436,
      "train_speed(iter/s)": 1.441183
    },
    {
      "epoch": 1.3439869757079816,
      "grad_norm": 4.201914310455322,
      "learning_rate": 8.320705646888341e-05,
      "loss": 2.675884819030762,
      "memory(GiB)": 77.56,
      "step": 31370,
      "token_acc": 0.41025641025641024,
      "train_speed(iter/s)": 1.441198
    },
    {
      "epoch": 1.3442011910372307,
      "grad_norm": 4.556771755218506,
      "learning_rate": 8.320202494116961e-05,
      "loss": 2.5075345993041993,
      "memory(GiB)": 77.56,
      "step": 31375,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.441211
    },
    {
      "epoch": 1.3444154063664797,
      "grad_norm": 6.09708833694458,
      "learning_rate": 8.319699281196999e-05,
      "loss": 2.4620147705078126,
      "memory(GiB)": 77.56,
      "step": 31380,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.441208
    },
    {
      "epoch": 1.3446296216957285,
      "grad_norm": 5.378604888916016,
      "learning_rate": 8.319196008137571e-05,
      "loss": 2.4251192092895506,
      "memory(GiB)": 77.56,
      "step": 31385,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.441224
    },
    {
      "epoch": 1.3448438370249776,
      "grad_norm": 5.7608819007873535,
      "learning_rate": 8.318692674947793e-05,
      "loss": 2.4822433471679686,
      "memory(GiB)": 77.56,
      "step": 31390,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.441251
    },
    {
      "epoch": 1.3450580523542266,
      "grad_norm": 6.302339553833008,
      "learning_rate": 8.318189281636784e-05,
      "loss": 2.61944522857666,
      "memory(GiB)": 77.56,
      "step": 31395,
      "token_acc": 0.46715328467153283,
      "train_speed(iter/s)": 1.441292
    },
    {
      "epoch": 1.3452722676834754,
      "grad_norm": 4.51309061050415,
      "learning_rate": 8.317685828213664e-05,
      "loss": 2.6544521331787108,
      "memory(GiB)": 77.56,
      "step": 31400,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.441346
    },
    {
      "epoch": 1.3454864830127244,
      "grad_norm": 4.59455680847168,
      "learning_rate": 8.317182314687551e-05,
      "loss": 2.6286258697509766,
      "memory(GiB)": 77.56,
      "step": 31405,
      "token_acc": 0.4742268041237113,
      "train_speed(iter/s)": 1.441355
    },
    {
      "epoch": 1.3457006983419735,
      "grad_norm": 5.485098838806152,
      "learning_rate": 8.316678741067571e-05,
      "loss": 2.4383277893066406,
      "memory(GiB)": 77.56,
      "step": 31410,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.441402
    },
    {
      "epoch": 1.3459149136712223,
      "grad_norm": 5.019629955291748,
      "learning_rate": 8.316175107362842e-05,
      "loss": 2.1067447662353516,
      "memory(GiB)": 77.56,
      "step": 31415,
      "token_acc": 0.5359712230215827,
      "train_speed(iter/s)": 1.441425
    },
    {
      "epoch": 1.3461291290004713,
      "grad_norm": 5.156033039093018,
      "learning_rate": 8.315671413582493e-05,
      "loss": 2.608934211730957,
      "memory(GiB)": 77.56,
      "step": 31420,
      "token_acc": 0.4900662251655629,
      "train_speed(iter/s)": 1.44146
    },
    {
      "epoch": 1.3463433443297204,
      "grad_norm": 4.857155799865723,
      "learning_rate": 8.315167659735646e-05,
      "loss": 2.121651840209961,
      "memory(GiB)": 77.56,
      "step": 31425,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441471
    },
    {
      "epoch": 1.3465575596589692,
      "grad_norm": 5.442330837249756,
      "learning_rate": 8.314663845831425e-05,
      "loss": 2.4756095886230467,
      "memory(GiB)": 77.56,
      "step": 31430,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.44148
    },
    {
      "epoch": 1.3467717749882182,
      "grad_norm": 4.46121072769165,
      "learning_rate": 8.314159971878958e-05,
      "loss": 2.612537956237793,
      "memory(GiB)": 77.56,
      "step": 31435,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.441459
    },
    {
      "epoch": 1.3469859903174672,
      "grad_norm": 4.219089984893799,
      "learning_rate": 8.313656037887377e-05,
      "loss": 2.809228706359863,
      "memory(GiB)": 77.56,
      "step": 31440,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.441452
    },
    {
      "epoch": 1.347200205646716,
      "grad_norm": 6.721188068389893,
      "learning_rate": 8.313152043865806e-05,
      "loss": 2.7048713684082033,
      "memory(GiB)": 77.56,
      "step": 31445,
      "token_acc": 0.46325878594249204,
      "train_speed(iter/s)": 1.441488
    },
    {
      "epoch": 1.347414420975965,
      "grad_norm": 7.163428783416748,
      "learning_rate": 8.312647989823379e-05,
      "loss": 2.533897590637207,
      "memory(GiB)": 77.56,
      "step": 31450,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.441524
    },
    {
      "epoch": 1.3476286363052141,
      "grad_norm": 4.878605842590332,
      "learning_rate": 8.312143875769227e-05,
      "loss": 2.642001152038574,
      "memory(GiB)": 77.56,
      "step": 31455,
      "token_acc": 0.420863309352518,
      "train_speed(iter/s)": 1.441565
    },
    {
      "epoch": 1.347842851634463,
      "grad_norm": 4.353795528411865,
      "learning_rate": 8.31163970171248e-05,
      "loss": 2.7180862426757812,
      "memory(GiB)": 77.56,
      "step": 31460,
      "token_acc": 0.43843843843843844,
      "train_speed(iter/s)": 1.441612
    },
    {
      "epoch": 1.348057066963712,
      "grad_norm": 5.946815490722656,
      "learning_rate": 8.311135467662275e-05,
      "loss": 2.5462614059448243,
      "memory(GiB)": 77.56,
      "step": 31465,
      "token_acc": 0.5080645161290323,
      "train_speed(iter/s)": 1.441666
    },
    {
      "epoch": 1.348271282292961,
      "grad_norm": 4.680395603179932,
      "learning_rate": 8.310631173627743e-05,
      "loss": 2.780935287475586,
      "memory(GiB)": 77.56,
      "step": 31470,
      "token_acc": 0.4493927125506073,
      "train_speed(iter/s)": 1.441695
    },
    {
      "epoch": 1.3484854976222098,
      "grad_norm": 4.769514083862305,
      "learning_rate": 8.310126819618023e-05,
      "loss": 2.515646743774414,
      "memory(GiB)": 77.56,
      "step": 31475,
      "token_acc": 0.4744525547445255,
      "train_speed(iter/s)": 1.441694
    },
    {
      "epoch": 1.3486997129514589,
      "grad_norm": 6.272392749786377,
      "learning_rate": 8.30962240564225e-05,
      "loss": 2.8053808212280273,
      "memory(GiB)": 77.56,
      "step": 31480,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.441701
    },
    {
      "epoch": 1.3489139282807079,
      "grad_norm": 5.002593517303467,
      "learning_rate": 8.309117931709563e-05,
      "loss": 2.5735185623168944,
      "memory(GiB)": 77.56,
      "step": 31485,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.441669
    },
    {
      "epoch": 1.3491281436099567,
      "grad_norm": 10.882744789123535,
      "learning_rate": 8.3086133978291e-05,
      "loss": 2.4281139373779297,
      "memory(GiB)": 77.56,
      "step": 31490,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.441608
    },
    {
      "epoch": 1.3493423589392057,
      "grad_norm": 5.653684616088867,
      "learning_rate": 8.308108804009999e-05,
      "loss": 2.6841964721679688,
      "memory(GiB)": 77.56,
      "step": 31495,
      "token_acc": 0.43125,
      "train_speed(iter/s)": 1.441571
    },
    {
      "epoch": 1.3495565742684548,
      "grad_norm": 5.4192399978637695,
      "learning_rate": 8.307604150261408e-05,
      "loss": 2.6143789291381836,
      "memory(GiB)": 77.56,
      "step": 31500,
      "token_acc": 0.46105919003115264,
      "train_speed(iter/s)": 1.441559
    },
    {
      "epoch": 1.3495565742684548,
      "eval_loss": 2.2718944549560547,
      "eval_runtime": 14.2635,
      "eval_samples_per_second": 7.011,
      "eval_steps_per_second": 7.011,
      "eval_token_acc": 0.453646477132262,
      "step": 31500
    },
    {
      "epoch": 1.3497707895977036,
      "grad_norm": 4.021576881408691,
      "learning_rate": 8.307099436592463e-05,
      "loss": 2.451487350463867,
      "memory(GiB)": 77.56,
      "step": 31505,
      "token_acc": 0.47302158273381295,
      "train_speed(iter/s)": 1.440563
    },
    {
      "epoch": 1.3499850049269526,
      "grad_norm": 4.678742408752441,
      "learning_rate": 8.306594663012308e-05,
      "loss": 2.7054092407226564,
      "memory(GiB)": 77.56,
      "step": 31510,
      "token_acc": 0.44554455445544555,
      "train_speed(iter/s)": 1.440572
    },
    {
      "epoch": 1.3501992202562016,
      "grad_norm": 5.106549263000488,
      "learning_rate": 8.306089829530092e-05,
      "loss": 2.810398483276367,
      "memory(GiB)": 77.56,
      "step": 31515,
      "token_acc": 0.4407894736842105,
      "train_speed(iter/s)": 1.440588
    },
    {
      "epoch": 1.3504134355854505,
      "grad_norm": 5.233591079711914,
      "learning_rate": 8.305584936154956e-05,
      "loss": 2.6204063415527346,
      "memory(GiB)": 77.56,
      "step": 31520,
      "token_acc": 0.42136498516320475,
      "train_speed(iter/s)": 1.440582
    },
    {
      "epoch": 1.3506276509146995,
      "grad_norm": 3.6382761001586914,
      "learning_rate": 8.305079982896047e-05,
      "loss": 2.325724792480469,
      "memory(GiB)": 77.56,
      "step": 31525,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.440537
    },
    {
      "epoch": 1.3508418662439485,
      "grad_norm": 6.607351779937744,
      "learning_rate": 8.304574969762515e-05,
      "loss": 2.378327178955078,
      "memory(GiB)": 77.56,
      "step": 31530,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.440544
    },
    {
      "epoch": 1.3510560815731973,
      "grad_norm": 4.294212818145752,
      "learning_rate": 8.304069896763506e-05,
      "loss": 2.536751556396484,
      "memory(GiB)": 77.56,
      "step": 31535,
      "token_acc": 0.4153846153846154,
      "train_speed(iter/s)": 1.44058
    },
    {
      "epoch": 1.3512702969024464,
      "grad_norm": 7.424275875091553,
      "learning_rate": 8.303564763908173e-05,
      "loss": 2.9064138412475584,
      "memory(GiB)": 77.56,
      "step": 31540,
      "token_acc": 0.43703703703703706,
      "train_speed(iter/s)": 1.440601
    },
    {
      "epoch": 1.3514845122316954,
      "grad_norm": 5.487633228302002,
      "learning_rate": 8.303059571205665e-05,
      "loss": 2.692577934265137,
      "memory(GiB)": 77.56,
      "step": 31545,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.440595
    },
    {
      "epoch": 1.3516987275609442,
      "grad_norm": 5.942775249481201,
      "learning_rate": 8.302554318665135e-05,
      "loss": 2.3590694427490235,
      "memory(GiB)": 77.56,
      "step": 31550,
      "token_acc": 0.5048231511254019,
      "train_speed(iter/s)": 1.440581
    },
    {
      "epoch": 1.3519129428901933,
      "grad_norm": 5.499608516693115,
      "learning_rate": 8.302049006295734e-05,
      "loss": 2.209274673461914,
      "memory(GiB)": 77.56,
      "step": 31555,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.440565
    },
    {
      "epoch": 1.3521271582194423,
      "grad_norm": 3.7635302543640137,
      "learning_rate": 8.301543634106617e-05,
      "loss": 2.4393642425537108,
      "memory(GiB)": 77.56,
      "step": 31560,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.440563
    },
    {
      "epoch": 1.352341373548691,
      "grad_norm": 4.146806240081787,
      "learning_rate": 8.301038202106942e-05,
      "loss": 2.742236518859863,
      "memory(GiB)": 77.56,
      "step": 31565,
      "token_acc": 0.4358288770053476,
      "train_speed(iter/s)": 1.440581
    },
    {
      "epoch": 1.3525555888779401,
      "grad_norm": 4.635458469390869,
      "learning_rate": 8.300532710305862e-05,
      "loss": 2.612195587158203,
      "memory(GiB)": 77.56,
      "step": 31570,
      "token_acc": 0.4483695652173913,
      "train_speed(iter/s)": 1.440554
    },
    {
      "epoch": 1.3527698042071892,
      "grad_norm": 5.260425567626953,
      "learning_rate": 8.300027158712537e-05,
      "loss": 2.4457496643066405,
      "memory(GiB)": 77.56,
      "step": 31575,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.44057
    },
    {
      "epoch": 1.352984019536438,
      "grad_norm": 4.975513458251953,
      "learning_rate": 8.299521547336123e-05,
      "loss": 2.519771766662598,
      "memory(GiB)": 77.56,
      "step": 31580,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.440547
    },
    {
      "epoch": 1.353198234865687,
      "grad_norm": 5.17201566696167,
      "learning_rate": 8.299015876185782e-05,
      "loss": 2.3570817947387694,
      "memory(GiB)": 77.56,
      "step": 31585,
      "token_acc": 0.5016393442622951,
      "train_speed(iter/s)": 1.440504
    },
    {
      "epoch": 1.353412450194936,
      "grad_norm": 5.8714213371276855,
      "learning_rate": 8.298510145270673e-05,
      "loss": 2.204788017272949,
      "memory(GiB)": 77.56,
      "step": 31590,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.440544
    },
    {
      "epoch": 1.3536266655241849,
      "grad_norm": 3.987436532974243,
      "learning_rate": 8.298004354599959e-05,
      "loss": 2.542467498779297,
      "memory(GiB)": 77.56,
      "step": 31595,
      "token_acc": 0.4551282051282051,
      "train_speed(iter/s)": 1.440516
    },
    {
      "epoch": 1.353840880853434,
      "grad_norm": 7.022082805633545,
      "learning_rate": 8.297498504182802e-05,
      "loss": 2.42038631439209,
      "memory(GiB)": 77.56,
      "step": 31600,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.440563
    },
    {
      "epoch": 1.354055096182683,
      "grad_norm": 3.850215196609497,
      "learning_rate": 8.296992594028366e-05,
      "loss": 2.5717395782470702,
      "memory(GiB)": 77.56,
      "step": 31605,
      "token_acc": 0.46449704142011833,
      "train_speed(iter/s)": 1.440531
    },
    {
      "epoch": 1.3542693115119318,
      "grad_norm": 6.579049110412598,
      "learning_rate": 8.296486624145819e-05,
      "loss": 2.7226057052612305,
      "memory(GiB)": 77.56,
      "step": 31610,
      "token_acc": 0.40059347181008903,
      "train_speed(iter/s)": 1.440532
    },
    {
      "epoch": 1.3544835268411808,
      "grad_norm": 4.3547139167785645,
      "learning_rate": 8.295980594544321e-05,
      "loss": 2.7584354400634767,
      "memory(GiB)": 77.56,
      "step": 31615,
      "token_acc": 0.414985590778098,
      "train_speed(iter/s)": 1.440592
    },
    {
      "epoch": 1.3546977421704298,
      "grad_norm": 6.598477363586426,
      "learning_rate": 8.295474505233042e-05,
      "loss": 2.7339330673217774,
      "memory(GiB)": 77.56,
      "step": 31620,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.440617
    },
    {
      "epoch": 1.3549119574996786,
      "grad_norm": 6.258572578430176,
      "learning_rate": 8.294968356221155e-05,
      "loss": 2.359242630004883,
      "memory(GiB)": 77.56,
      "step": 31625,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.440607
    },
    {
      "epoch": 1.3551261728289277,
      "grad_norm": 3.8502566814422607,
      "learning_rate": 8.294462147517823e-05,
      "loss": 2.6734107971191405,
      "memory(GiB)": 77.56,
      "step": 31630,
      "token_acc": 0.46112600536193027,
      "train_speed(iter/s)": 1.440638
    },
    {
      "epoch": 1.3553403881581767,
      "grad_norm": 4.248429775238037,
      "learning_rate": 8.293955879132219e-05,
      "loss": 2.321911430358887,
      "memory(GiB)": 77.56,
      "step": 31635,
      "token_acc": 0.5016949152542373,
      "train_speed(iter/s)": 1.440628
    },
    {
      "epoch": 1.3555546034874255,
      "grad_norm": 4.681156635284424,
      "learning_rate": 8.293449551073513e-05,
      "loss": 2.7754085540771483,
      "memory(GiB)": 77.56,
      "step": 31640,
      "token_acc": 0.4452296819787986,
      "train_speed(iter/s)": 1.440572
    },
    {
      "epoch": 1.3557688188166745,
      "grad_norm": 6.744724273681641,
      "learning_rate": 8.29294316335088e-05,
      "loss": 2.5335769653320312,
      "memory(GiB)": 77.56,
      "step": 31645,
      "token_acc": 0.4551282051282051,
      "train_speed(iter/s)": 1.440555
    },
    {
      "epoch": 1.3559830341459236,
      "grad_norm": 4.318502426147461,
      "learning_rate": 8.292436715973493e-05,
      "loss": 2.8196269989013674,
      "memory(GiB)": 77.56,
      "step": 31650,
      "token_acc": 0.4304635761589404,
      "train_speed(iter/s)": 1.440608
    },
    {
      "epoch": 1.3561972494751724,
      "grad_norm": 7.155418395996094,
      "learning_rate": 8.291930208950525e-05,
      "loss": 2.180579948425293,
      "memory(GiB)": 77.56,
      "step": 31655,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.440645
    },
    {
      "epoch": 1.3564114648044214,
      "grad_norm": 4.491590976715088,
      "learning_rate": 8.291423642291153e-05,
      "loss": 2.1374509811401365,
      "memory(GiB)": 77.56,
      "step": 31660,
      "token_acc": 0.5587044534412956,
      "train_speed(iter/s)": 1.440631
    },
    {
      "epoch": 1.3566256801336705,
      "grad_norm": 4.511011123657227,
      "learning_rate": 8.290917016004556e-05,
      "loss": 2.2907419204711914,
      "memory(GiB)": 77.56,
      "step": 31665,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.44062
    },
    {
      "epoch": 1.3568398954629193,
      "grad_norm": 7.064165115356445,
      "learning_rate": 8.29041033009991e-05,
      "loss": 2.507394027709961,
      "memory(GiB)": 77.56,
      "step": 31670,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.44066
    },
    {
      "epoch": 1.3570541107921683,
      "grad_norm": 6.257872581481934,
      "learning_rate": 8.289903584586394e-05,
      "loss": 2.676230621337891,
      "memory(GiB)": 77.56,
      "step": 31675,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.440599
    },
    {
      "epoch": 1.3572683261214173,
      "grad_norm": 4.7414422035217285,
      "learning_rate": 8.289396779473188e-05,
      "loss": 2.6043365478515623,
      "memory(GiB)": 77.56,
      "step": 31680,
      "token_acc": 0.48297213622291024,
      "train_speed(iter/s)": 1.440598
    },
    {
      "epoch": 1.3574825414506662,
      "grad_norm": 3.660566806793213,
      "learning_rate": 8.288889914769473e-05,
      "loss": 2.528450775146484,
      "memory(GiB)": 77.56,
      "step": 31685,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.440644
    },
    {
      "epoch": 1.3576967567799152,
      "grad_norm": 5.024524211883545,
      "learning_rate": 8.288382990484432e-05,
      "loss": 2.401573371887207,
      "memory(GiB)": 77.56,
      "step": 31690,
      "token_acc": 0.527027027027027,
      "train_speed(iter/s)": 1.440644
    },
    {
      "epoch": 1.3579109721091642,
      "grad_norm": 6.42276668548584,
      "learning_rate": 8.287876006627248e-05,
      "loss": 2.7882152557373048,
      "memory(GiB)": 77.56,
      "step": 31695,
      "token_acc": 0.4234527687296417,
      "train_speed(iter/s)": 1.440642
    },
    {
      "epoch": 1.358125187438413,
      "grad_norm": 6.427697658538818,
      "learning_rate": 8.287368963207107e-05,
      "loss": 2.370273399353027,
      "memory(GiB)": 77.56,
      "step": 31700,
      "token_acc": 0.47962382445141066,
      "train_speed(iter/s)": 1.440681
    },
    {
      "epoch": 1.358339402767662,
      "grad_norm": 5.093068599700928,
      "learning_rate": 8.286861860233194e-05,
      "loss": 3.053812789916992,
      "memory(GiB)": 77.56,
      "step": 31705,
      "token_acc": 0.4094955489614243,
      "train_speed(iter/s)": 1.440685
    },
    {
      "epoch": 1.358553618096911,
      "grad_norm": 4.350356101989746,
      "learning_rate": 8.286354697714695e-05,
      "loss": 2.5586700439453125,
      "memory(GiB)": 77.56,
      "step": 31710,
      "token_acc": 0.5302491103202847,
      "train_speed(iter/s)": 1.440624
    },
    {
      "epoch": 1.35876783342616,
      "grad_norm": 5.138138294219971,
      "learning_rate": 8.285847475660797e-05,
      "loss": 2.552874755859375,
      "memory(GiB)": 77.56,
      "step": 31715,
      "token_acc": 0.47183098591549294,
      "train_speed(iter/s)": 1.440609
    },
    {
      "epoch": 1.358982048755409,
      "grad_norm": 6.08490514755249,
      "learning_rate": 8.285340194080688e-05,
      "loss": 2.8570438385009767,
      "memory(GiB)": 77.56,
      "step": 31720,
      "token_acc": 0.41118421052631576,
      "train_speed(iter/s)": 1.440617
    },
    {
      "epoch": 1.359196264084658,
      "grad_norm": 4.018260478973389,
      "learning_rate": 8.284832852983562e-05,
      "loss": 2.285373497009277,
      "memory(GiB)": 77.56,
      "step": 31725,
      "token_acc": 0.5407725321888412,
      "train_speed(iter/s)": 1.440635
    },
    {
      "epoch": 1.3594104794139068,
      "grad_norm": 6.264819145202637,
      "learning_rate": 8.284325452378606e-05,
      "loss": 2.553983688354492,
      "memory(GiB)": 77.56,
      "step": 31730,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.440691
    },
    {
      "epoch": 1.3596246947431558,
      "grad_norm": 6.526613712310791,
      "learning_rate": 8.283817992275014e-05,
      "loss": 2.6119489669799805,
      "memory(GiB)": 77.56,
      "step": 31735,
      "token_acc": 0.4780058651026393,
      "train_speed(iter/s)": 1.440719
    },
    {
      "epoch": 1.3598389100724049,
      "grad_norm": 5.595158576965332,
      "learning_rate": 8.28331047268198e-05,
      "loss": 2.2213531494140626,
      "memory(GiB)": 77.56,
      "step": 31740,
      "token_acc": 0.5040322580645161,
      "train_speed(iter/s)": 1.440744
    },
    {
      "epoch": 1.3600531254016537,
      "grad_norm": 4.940277099609375,
      "learning_rate": 8.282802893608694e-05,
      "loss": 2.478802490234375,
      "memory(GiB)": 77.56,
      "step": 31745,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.440704
    },
    {
      "epoch": 1.3602673407309027,
      "grad_norm": 6.406253337860107,
      "learning_rate": 8.282295255064356e-05,
      "loss": 2.4300052642822267,
      "memory(GiB)": 77.56,
      "step": 31750,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.440751
    },
    {
      "epoch": 1.3604815560601518,
      "grad_norm": 4.3721208572387695,
      "learning_rate": 8.28178755705816e-05,
      "loss": 2.2829425811767576,
      "memory(GiB)": 77.56,
      "step": 31755,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.440701
    },
    {
      "epoch": 1.3606957713894006,
      "grad_norm": 4.818572521209717,
      "learning_rate": 8.281279799599303e-05,
      "loss": 2.547805404663086,
      "memory(GiB)": 77.56,
      "step": 31760,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.440724
    },
    {
      "epoch": 1.3609099867186496,
      "grad_norm": 5.698460578918457,
      "learning_rate": 8.280771982696985e-05,
      "loss": 2.5330101013183595,
      "memory(GiB)": 77.56,
      "step": 31765,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.440696
    },
    {
      "epoch": 1.3611242020478986,
      "grad_norm": 4.337852478027344,
      "learning_rate": 8.280264106360405e-05,
      "loss": 2.480868911743164,
      "memory(GiB)": 77.56,
      "step": 31770,
      "token_acc": 0.4621212121212121,
      "train_speed(iter/s)": 1.440718
    },
    {
      "epoch": 1.3613384173771474,
      "grad_norm": 6.176379203796387,
      "learning_rate": 8.279756170598764e-05,
      "loss": 2.332061195373535,
      "memory(GiB)": 77.56,
      "step": 31775,
      "token_acc": 0.5284810126582279,
      "train_speed(iter/s)": 1.440725
    },
    {
      "epoch": 1.3615526327063965,
      "grad_norm": 7.281599521636963,
      "learning_rate": 8.279248175421264e-05,
      "loss": 2.780141067504883,
      "memory(GiB)": 77.56,
      "step": 31780,
      "token_acc": 0.4119496855345912,
      "train_speed(iter/s)": 1.440735
    },
    {
      "epoch": 1.3617668480356455,
      "grad_norm": 5.348024845123291,
      "learning_rate": 8.278740120837106e-05,
      "loss": 2.6644657135009764,
      "memory(GiB)": 77.56,
      "step": 31785,
      "token_acc": 0.43322475570032576,
      "train_speed(iter/s)": 1.440758
    },
    {
      "epoch": 1.3619810633648943,
      "grad_norm": 5.615641117095947,
      "learning_rate": 8.278232006855495e-05,
      "loss": 2.578746223449707,
      "memory(GiB)": 77.56,
      "step": 31790,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.440798
    },
    {
      "epoch": 1.3621952786941434,
      "grad_norm": 4.55649471282959,
      "learning_rate": 8.277723833485635e-05,
      "loss": 2.334025764465332,
      "memory(GiB)": 77.56,
      "step": 31795,
      "token_acc": 0.47346938775510206,
      "train_speed(iter/s)": 1.440785
    },
    {
      "epoch": 1.3624094940233924,
      "grad_norm": 5.463602066040039,
      "learning_rate": 8.277215600736734e-05,
      "loss": 2.733611488342285,
      "memory(GiB)": 77.56,
      "step": 31800,
      "token_acc": 0.4290657439446367,
      "train_speed(iter/s)": 1.440815
    },
    {
      "epoch": 1.3626237093526412,
      "grad_norm": 5.173390865325928,
      "learning_rate": 8.276707308617999e-05,
      "loss": 2.8185863494873047,
      "memory(GiB)": 77.56,
      "step": 31805,
      "token_acc": 0.4429530201342282,
      "train_speed(iter/s)": 1.440855
    },
    {
      "epoch": 1.3628379246818902,
      "grad_norm": 3.8689751625061035,
      "learning_rate": 8.276198957138636e-05,
      "loss": 2.4822025299072266,
      "memory(GiB)": 77.56,
      "step": 31810,
      "token_acc": 0.43006993006993005,
      "train_speed(iter/s)": 1.440893
    },
    {
      "epoch": 1.3630521400111393,
      "grad_norm": 5.2507405281066895,
      "learning_rate": 8.275690546307854e-05,
      "loss": 2.4450321197509766,
      "memory(GiB)": 77.56,
      "step": 31815,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.440902
    },
    {
      "epoch": 1.363266355340388,
      "grad_norm": 5.465854644775391,
      "learning_rate": 8.275182076134869e-05,
      "loss": 2.6255542755126955,
      "memory(GiB)": 77.56,
      "step": 31820,
      "token_acc": 0.43258426966292135,
      "train_speed(iter/s)": 1.44095
    },
    {
      "epoch": 1.3634805706696371,
      "grad_norm": 8.459576606750488,
      "learning_rate": 8.274673546628886e-05,
      "loss": 2.4949108123779298,
      "memory(GiB)": 77.56,
      "step": 31825,
      "token_acc": 0.4952978056426332,
      "train_speed(iter/s)": 1.440972
    },
    {
      "epoch": 1.3636947859988862,
      "grad_norm": 5.377734661102295,
      "learning_rate": 8.27416495779912e-05,
      "loss": 2.4399099349975586,
      "memory(GiB)": 77.56,
      "step": 31830,
      "token_acc": 0.43812709030100333,
      "train_speed(iter/s)": 1.440939
    },
    {
      "epoch": 1.363909001328135,
      "grad_norm": 5.124488353729248,
      "learning_rate": 8.273656309654785e-05,
      "loss": 2.7032752990722657,
      "memory(GiB)": 77.56,
      "step": 31835,
      "token_acc": 0.47112462006079026,
      "train_speed(iter/s)": 1.440876
    },
    {
      "epoch": 1.364123216657384,
      "grad_norm": 6.172043323516846,
      "learning_rate": 8.273147602205094e-05,
      "loss": 2.3730249404907227,
      "memory(GiB)": 77.56,
      "step": 31840,
      "token_acc": 0.5089285714285714,
      "train_speed(iter/s)": 1.440928
    },
    {
      "epoch": 1.364337431986633,
      "grad_norm": 4.208598613739014,
      "learning_rate": 8.272638835459263e-05,
      "loss": 2.2262176513671874,
      "memory(GiB)": 77.56,
      "step": 31845,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440898
    },
    {
      "epoch": 1.3645516473158819,
      "grad_norm": 4.994443893432617,
      "learning_rate": 8.272130009426511e-05,
      "loss": 2.438750457763672,
      "memory(GiB)": 77.56,
      "step": 31850,
      "token_acc": 0.4591439688715953,
      "train_speed(iter/s)": 1.440904
    },
    {
      "epoch": 1.364765862645131,
      "grad_norm": 4.585249423980713,
      "learning_rate": 8.271621124116055e-05,
      "loss": 2.61342830657959,
      "memory(GiB)": 77.56,
      "step": 31855,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.440854
    },
    {
      "epoch": 1.36498007797438,
      "grad_norm": 6.497950077056885,
      "learning_rate": 8.271112179537113e-05,
      "loss": 2.631338119506836,
      "memory(GiB)": 77.56,
      "step": 31860,
      "token_acc": 0.4577777777777778,
      "train_speed(iter/s)": 1.440831
    },
    {
      "epoch": 1.3651942933036287,
      "grad_norm": 4.362639427185059,
      "learning_rate": 8.270603175698904e-05,
      "loss": 2.476442337036133,
      "memory(GiB)": 77.56,
      "step": 31865,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440769
    },
    {
      "epoch": 1.3654085086328778,
      "grad_norm": 4.356795787811279,
      "learning_rate": 8.270094112610651e-05,
      "loss": 2.068539047241211,
      "memory(GiB)": 77.56,
      "step": 31870,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.440764
    },
    {
      "epoch": 1.3656227239621268,
      "grad_norm": 3.883218288421631,
      "learning_rate": 8.269584990281575e-05,
      "loss": 2.4610416412353517,
      "memory(GiB)": 77.56,
      "step": 31875,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.440735
    },
    {
      "epoch": 1.3658369392913756,
      "grad_norm": 4.142423152923584,
      "learning_rate": 8.269075808720901e-05,
      "loss": 2.468758964538574,
      "memory(GiB)": 77.56,
      "step": 31880,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.440681
    },
    {
      "epoch": 1.3660511546206247,
      "grad_norm": 5.313052654266357,
      "learning_rate": 8.268566567937851e-05,
      "loss": 2.5808610916137695,
      "memory(GiB)": 77.56,
      "step": 31885,
      "token_acc": 0.49337748344370863,
      "train_speed(iter/s)": 1.440758
    },
    {
      "epoch": 1.3662653699498737,
      "grad_norm": 4.263774394989014,
      "learning_rate": 8.268057267941652e-05,
      "loss": 2.6729047775268553,
      "memory(GiB)": 77.56,
      "step": 31890,
      "token_acc": 0.4834710743801653,
      "train_speed(iter/s)": 1.440725
    },
    {
      "epoch": 1.3664795852791225,
      "grad_norm": 4.951992034912109,
      "learning_rate": 8.267547908741529e-05,
      "loss": 2.513370704650879,
      "memory(GiB)": 77.56,
      "step": 31895,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.440762
    },
    {
      "epoch": 1.3666938006083715,
      "grad_norm": 6.280810356140137,
      "learning_rate": 8.267038490346712e-05,
      "loss": 2.89864501953125,
      "memory(GiB)": 77.56,
      "step": 31900,
      "token_acc": 0.3862815884476534,
      "train_speed(iter/s)": 1.440807
    },
    {
      "epoch": 1.3669080159376206,
      "grad_norm": 6.688979625701904,
      "learning_rate": 8.266529012766428e-05,
      "loss": 2.2513904571533203,
      "memory(GiB)": 77.56,
      "step": 31905,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.440804
    },
    {
      "epoch": 1.3671222312668694,
      "grad_norm": 4.840042591094971,
      "learning_rate": 8.266019476009905e-05,
      "loss": 2.280426788330078,
      "memory(GiB)": 77.56,
      "step": 31910,
      "token_acc": 0.5139442231075697,
      "train_speed(iter/s)": 1.440755
    },
    {
      "epoch": 1.3673364465961184,
      "grad_norm": 4.748971462249756,
      "learning_rate": 8.265509880086376e-05,
      "loss": 2.5299015045166016,
      "memory(GiB)": 77.56,
      "step": 31915,
      "token_acc": 0.4527687296416938,
      "train_speed(iter/s)": 1.440802
    },
    {
      "epoch": 1.3675506619253675,
      "grad_norm": 3.980222225189209,
      "learning_rate": 8.265000225005073e-05,
      "loss": 2.3797998428344727,
      "memory(GiB)": 77.56,
      "step": 31920,
      "token_acc": 0.5074183976261127,
      "train_speed(iter/s)": 1.440833
    },
    {
      "epoch": 1.3677648772546163,
      "grad_norm": 3.3712213039398193,
      "learning_rate": 8.264490510775227e-05,
      "loss": 2.6415313720703124,
      "memory(GiB)": 77.56,
      "step": 31925,
      "token_acc": 0.4474327628361858,
      "train_speed(iter/s)": 1.440814
    },
    {
      "epoch": 1.3679790925838653,
      "grad_norm": 5.953074932098389,
      "learning_rate": 8.263980737406073e-05,
      "loss": 2.5097558975219725,
      "memory(GiB)": 77.56,
      "step": 31930,
      "token_acc": 0.4517133956386293,
      "train_speed(iter/s)": 1.440751
    },
    {
      "epoch": 1.3681933079131143,
      "grad_norm": 5.452756881713867,
      "learning_rate": 8.263470904906849e-05,
      "loss": 2.7983936309814452,
      "memory(GiB)": 77.56,
      "step": 31935,
      "token_acc": 0.4458204334365325,
      "train_speed(iter/s)": 1.440788
    },
    {
      "epoch": 1.3684075232423631,
      "grad_norm": 4.6167802810668945,
      "learning_rate": 8.262961013286785e-05,
      "loss": 2.6030508041381837,
      "memory(GiB)": 77.56,
      "step": 31940,
      "token_acc": 0.45132743362831856,
      "train_speed(iter/s)": 1.440785
    },
    {
      "epoch": 1.3686217385716122,
      "grad_norm": 4.188957691192627,
      "learning_rate": 8.262451062555123e-05,
      "loss": 2.559660720825195,
      "memory(GiB)": 77.56,
      "step": 31945,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.440781
    },
    {
      "epoch": 1.3688359539008612,
      "grad_norm": 6.492400646209717,
      "learning_rate": 8.261941052721098e-05,
      "loss": 2.6256488800048827,
      "memory(GiB)": 77.56,
      "step": 31950,
      "token_acc": 0.4956268221574344,
      "train_speed(iter/s)": 1.440781
    },
    {
      "epoch": 1.36905016923011,
      "grad_norm": 5.4128899574279785,
      "learning_rate": 8.261430983793952e-05,
      "loss": 2.4580759048461913,
      "memory(GiB)": 77.56,
      "step": 31955,
      "token_acc": 0.49185667752442996,
      "train_speed(iter/s)": 1.440833
    },
    {
      "epoch": 1.369264384559359,
      "grad_norm": 4.393882751464844,
      "learning_rate": 8.260920855782924e-05,
      "loss": 2.257247543334961,
      "memory(GiB)": 77.56,
      "step": 31960,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.440872
    },
    {
      "epoch": 1.369478599888608,
      "grad_norm": 5.233603477478027,
      "learning_rate": 8.260410668697255e-05,
      "loss": 2.361859893798828,
      "memory(GiB)": 77.56,
      "step": 31965,
      "token_acc": 0.5227272727272727,
      "train_speed(iter/s)": 1.440922
    },
    {
      "epoch": 1.369692815217857,
      "grad_norm": 5.287154197692871,
      "learning_rate": 8.25990042254619e-05,
      "loss": 2.5765869140625,
      "memory(GiB)": 77.56,
      "step": 31970,
      "token_acc": 0.46096654275092935,
      "train_speed(iter/s)": 1.440892
    },
    {
      "epoch": 1.369907030547106,
      "grad_norm": 5.994129657745361,
      "learning_rate": 8.25939011733897e-05,
      "loss": 2.6727779388427733,
      "memory(GiB)": 77.56,
      "step": 31975,
      "token_acc": 0.467680608365019,
      "train_speed(iter/s)": 1.440849
    },
    {
      "epoch": 1.370121245876355,
      "grad_norm": 4.072033405303955,
      "learning_rate": 8.25887975308484e-05,
      "loss": 2.5022659301757812,
      "memory(GiB)": 77.56,
      "step": 31980,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.4408
    },
    {
      "epoch": 1.3703354612056038,
      "grad_norm": 4.850002288818359,
      "learning_rate": 8.258369329793046e-05,
      "loss": 2.3202320098876954,
      "memory(GiB)": 77.56,
      "step": 31985,
      "token_acc": 0.5427350427350427,
      "train_speed(iter/s)": 1.440843
    },
    {
      "epoch": 1.3705496765348528,
      "grad_norm": 9.01506233215332,
      "learning_rate": 8.257858847472836e-05,
      "loss": 2.5688522338867186,
      "memory(GiB)": 77.56,
      "step": 31990,
      "token_acc": 0.4628975265017668,
      "train_speed(iter/s)": 1.440856
    },
    {
      "epoch": 1.3707638918641019,
      "grad_norm": 5.303791046142578,
      "learning_rate": 8.257348306133457e-05,
      "loss": 1.904374885559082,
      "memory(GiB)": 77.56,
      "step": 31995,
      "token_acc": 0.5689655172413793,
      "train_speed(iter/s)": 1.440887
    },
    {
      "epoch": 1.3709781071933507,
      "grad_norm": 10.566582679748535,
      "learning_rate": 8.256837705784157e-05,
      "loss": 3.1371002197265625,
      "memory(GiB)": 77.56,
      "step": 32000,
      "token_acc": 0.44140625,
      "train_speed(iter/s)": 1.440863
    },
    {
      "epoch": 1.3709781071933507,
      "eval_loss": 2.3098573684692383,
      "eval_runtime": 14.5627,
      "eval_samples_per_second": 6.867,
      "eval_steps_per_second": 6.867,
      "eval_token_acc": 0.4304556354916067,
      "step": 32000
    },
    {
      "epoch": 1.3711923225225997,
      "grad_norm": 5.158335208892822,
      "learning_rate": 8.256327046434187e-05,
      "loss": 2.698360633850098,
      "memory(GiB)": 77.56,
      "step": 32005,
      "token_acc": 0.43577981651376146,
      "train_speed(iter/s)": 1.439853
    },
    {
      "epoch": 1.3714065378518487,
      "grad_norm": 5.739282131195068,
      "learning_rate": 8.255816328092798e-05,
      "loss": 2.6402584075927735,
      "memory(GiB)": 77.56,
      "step": 32010,
      "token_acc": 0.4405594405594406,
      "train_speed(iter/s)": 1.439903
    },
    {
      "epoch": 1.3716207531810976,
      "grad_norm": 6.183213710784912,
      "learning_rate": 8.255305550769242e-05,
      "loss": 2.5954963684082033,
      "memory(GiB)": 77.56,
      "step": 32015,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.439934
    },
    {
      "epoch": 1.3718349685103466,
      "grad_norm": 5.9565110206604,
      "learning_rate": 8.254794714472771e-05,
      "loss": 2.6600975036621093,
      "memory(GiB)": 77.56,
      "step": 32020,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.43991
    },
    {
      "epoch": 1.3720491838395956,
      "grad_norm": 6.393780708312988,
      "learning_rate": 8.254283819212641e-05,
      "loss": 2.565863037109375,
      "memory(GiB)": 77.56,
      "step": 32025,
      "token_acc": 0.436,
      "train_speed(iter/s)": 1.439965
    },
    {
      "epoch": 1.3722633991688444,
      "grad_norm": 4.991652488708496,
      "learning_rate": 8.253772864998108e-05,
      "loss": 2.6734058380126955,
      "memory(GiB)": 77.56,
      "step": 32030,
      "token_acc": 0.42342342342342343,
      "train_speed(iter/s)": 1.439969
    },
    {
      "epoch": 1.3724776144980935,
      "grad_norm": 5.401498317718506,
      "learning_rate": 8.253261851838426e-05,
      "loss": 2.618927764892578,
      "memory(GiB)": 77.56,
      "step": 32035,
      "token_acc": 0.40797546012269936,
      "train_speed(iter/s)": 1.439938
    },
    {
      "epoch": 1.3726918298273425,
      "grad_norm": 5.256174087524414,
      "learning_rate": 8.252750779742855e-05,
      "loss": 2.5999101638793944,
      "memory(GiB)": 77.56,
      "step": 32040,
      "token_acc": 0.48466257668711654,
      "train_speed(iter/s)": 1.439991
    },
    {
      "epoch": 1.3729060451565913,
      "grad_norm": 4.309775352478027,
      "learning_rate": 8.252239648720652e-05,
      "loss": 2.523781585693359,
      "memory(GiB)": 77.56,
      "step": 32045,
      "token_acc": 0.45821325648414984,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 1.3731202604858403,
      "grad_norm": 5.225777626037598,
      "learning_rate": 8.251728458781077e-05,
      "loss": 2.583746910095215,
      "memory(GiB)": 77.56,
      "step": 32050,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.44008
    },
    {
      "epoch": 1.3733344758150894,
      "grad_norm": 5.951074123382568,
      "learning_rate": 8.25121720993339e-05,
      "loss": 2.4495475769042967,
      "memory(GiB)": 77.56,
      "step": 32055,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.440102
    },
    {
      "epoch": 1.3735486911443382,
      "grad_norm": 5.441335201263428,
      "learning_rate": 8.250705902186853e-05,
      "loss": 2.4885532379150392,
      "memory(GiB)": 77.56,
      "step": 32060,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.440132
    },
    {
      "epoch": 1.3737629064735872,
      "grad_norm": 4.366138458251953,
      "learning_rate": 8.250194535550731e-05,
      "loss": 2.512118911743164,
      "memory(GiB)": 77.56,
      "step": 32065,
      "token_acc": 0.4735099337748344,
      "train_speed(iter/s)": 1.440163
    },
    {
      "epoch": 1.3739771218028363,
      "grad_norm": 3.9534366130828857,
      "learning_rate": 8.249683110034283e-05,
      "loss": 2.3270990371704103,
      "memory(GiB)": 77.56,
      "step": 32070,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.440148
    },
    {
      "epoch": 1.374191337132085,
      "grad_norm": 3.901322364807129,
      "learning_rate": 8.249171625646779e-05,
      "loss": 2.7472261428833007,
      "memory(GiB)": 77.56,
      "step": 32075,
      "token_acc": 0.4126984126984127,
      "train_speed(iter/s)": 1.440082
    },
    {
      "epoch": 1.3744055524613341,
      "grad_norm": 4.305636882781982,
      "learning_rate": 8.248660082397484e-05,
      "loss": 2.2574813842773436,
      "memory(GiB)": 77.56,
      "step": 32080,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440114
    },
    {
      "epoch": 1.3746197677905831,
      "grad_norm": 4.8445515632629395,
      "learning_rate": 8.248148480295662e-05,
      "loss": 2.6581447601318358,
      "memory(GiB)": 77.56,
      "step": 32085,
      "token_acc": 0.45317220543806647,
      "train_speed(iter/s)": 1.440171
    },
    {
      "epoch": 1.374833983119832,
      "grad_norm": 4.6322760581970215,
      "learning_rate": 8.247636819350584e-05,
      "loss": 2.5092159271240235,
      "memory(GiB)": 77.56,
      "step": 32090,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.440181
    },
    {
      "epoch": 1.375048198449081,
      "grad_norm": 4.44414758682251,
      "learning_rate": 8.24712509957152e-05,
      "loss": 2.5538555145263673,
      "memory(GiB)": 77.56,
      "step": 32095,
      "token_acc": 0.4793650793650794,
      "train_speed(iter/s)": 1.440196
    },
    {
      "epoch": 1.37526241377833,
      "grad_norm": 5.26290225982666,
      "learning_rate": 8.246613320967737e-05,
      "loss": 2.5447097778320313,
      "memory(GiB)": 77.56,
      "step": 32100,
      "token_acc": 0.45180722891566266,
      "train_speed(iter/s)": 1.440247
    },
    {
      "epoch": 1.3754766291075788,
      "grad_norm": 6.584343910217285,
      "learning_rate": 8.246101483548508e-05,
      "loss": 2.485315704345703,
      "memory(GiB)": 77.56,
      "step": 32105,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.440302
    },
    {
      "epoch": 1.3756908444368279,
      "grad_norm": 5.275211811065674,
      "learning_rate": 8.245589587323106e-05,
      "loss": 2.497344970703125,
      "memory(GiB)": 77.56,
      "step": 32110,
      "token_acc": 0.42641509433962266,
      "train_speed(iter/s)": 1.440355
    },
    {
      "epoch": 1.375905059766077,
      "grad_norm": 4.924277305603027,
      "learning_rate": 8.245077632300803e-05,
      "loss": 2.176496124267578,
      "memory(GiB)": 77.56,
      "step": 32115,
      "token_acc": 0.4813664596273292,
      "train_speed(iter/s)": 1.440368
    },
    {
      "epoch": 1.3761192750953257,
      "grad_norm": 4.762798309326172,
      "learning_rate": 8.244565618490876e-05,
      "loss": 2.8513729095458986,
      "memory(GiB)": 77.56,
      "step": 32120,
      "token_acc": 0.4355300859598854,
      "train_speed(iter/s)": 1.440389
    },
    {
      "epoch": 1.3763334904245748,
      "grad_norm": 4.052890777587891,
      "learning_rate": 8.2440535459026e-05,
      "loss": 2.328522872924805,
      "memory(GiB)": 77.56,
      "step": 32125,
      "token_acc": 0.5115511551155115,
      "train_speed(iter/s)": 1.440352
    },
    {
      "epoch": 1.3765477057538238,
      "grad_norm": 3.617204189300537,
      "learning_rate": 8.24354141454525e-05,
      "loss": 3.036127471923828,
      "memory(GiB)": 77.56,
      "step": 32130,
      "token_acc": 0.4251497005988024,
      "train_speed(iter/s)": 1.440358
    },
    {
      "epoch": 1.3767619210830726,
      "grad_norm": 4.177921295166016,
      "learning_rate": 8.243029224428104e-05,
      "loss": 2.887272262573242,
      "memory(GiB)": 77.56,
      "step": 32135,
      "token_acc": 0.44722222222222224,
      "train_speed(iter/s)": 1.440365
    },
    {
      "epoch": 1.3769761364123216,
      "grad_norm": 5.442276477813721,
      "learning_rate": 8.242516975560441e-05,
      "loss": 2.3198795318603516,
      "memory(GiB)": 77.56,
      "step": 32140,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.440331
    },
    {
      "epoch": 1.3771903517415707,
      "grad_norm": 5.068624496459961,
      "learning_rate": 8.242004667951542e-05,
      "loss": 2.5350936889648437,
      "memory(GiB)": 77.56,
      "step": 32145,
      "token_acc": 0.4560810810810811,
      "train_speed(iter/s)": 1.440311
    },
    {
      "epoch": 1.3774045670708195,
      "grad_norm": 4.242217540740967,
      "learning_rate": 8.241492301610688e-05,
      "loss": 2.698986053466797,
      "memory(GiB)": 77.56,
      "step": 32150,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.440327
    },
    {
      "epoch": 1.3776187824000685,
      "grad_norm": 4.06760835647583,
      "learning_rate": 8.240979876547159e-05,
      "loss": 2.568186378479004,
      "memory(GiB)": 77.56,
      "step": 32155,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.440355
    },
    {
      "epoch": 1.3778329977293176,
      "grad_norm": 4.250998020172119,
      "learning_rate": 8.240467392770239e-05,
      "loss": 2.386155700683594,
      "memory(GiB)": 77.56,
      "step": 32160,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.440397
    },
    {
      "epoch": 1.3780472130585664,
      "grad_norm": 5.284581661224365,
      "learning_rate": 8.239954850289215e-05,
      "loss": 2.4334136962890627,
      "memory(GiB)": 77.56,
      "step": 32165,
      "token_acc": 0.4323308270676692,
      "train_speed(iter/s)": 1.440392
    },
    {
      "epoch": 1.3782614283878154,
      "grad_norm": 3.9874377250671387,
      "learning_rate": 8.239442249113366e-05,
      "loss": 2.5073711395263674,
      "memory(GiB)": 77.56,
      "step": 32170,
      "token_acc": 0.45569620253164556,
      "train_speed(iter/s)": 1.440387
    },
    {
      "epoch": 1.3784756437170644,
      "grad_norm": 4.134307384490967,
      "learning_rate": 8.238929589251984e-05,
      "loss": 2.5441864013671873,
      "memory(GiB)": 77.56,
      "step": 32175,
      "token_acc": 0.4713375796178344,
      "train_speed(iter/s)": 1.440449
    },
    {
      "epoch": 1.3786898590463132,
      "grad_norm": 4.906061172485352,
      "learning_rate": 8.238416870714354e-05,
      "loss": 2.6196819305419923,
      "memory(GiB)": 77.56,
      "step": 32180,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.440486
    },
    {
      "epoch": 1.3789040743755623,
      "grad_norm": 4.560486316680908,
      "learning_rate": 8.237904093509763e-05,
      "loss": 2.4726566314697265,
      "memory(GiB)": 77.56,
      "step": 32185,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.3791182897048113,
      "grad_norm": 3.7004616260528564,
      "learning_rate": 8.237391257647503e-05,
      "loss": 2.184162139892578,
      "memory(GiB)": 77.56,
      "step": 32190,
      "token_acc": 0.5254777070063694,
      "train_speed(iter/s)": 1.440515
    },
    {
      "epoch": 1.3793325050340601,
      "grad_norm": 4.5986127853393555,
      "learning_rate": 8.236878363136864e-05,
      "loss": 2.652243423461914,
      "memory(GiB)": 77.56,
      "step": 32195,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.440554
    },
    {
      "epoch": 1.3795467203633092,
      "grad_norm": 5.741354465484619,
      "learning_rate": 8.236365409987136e-05,
      "loss": 2.4384607315063476,
      "memory(GiB)": 77.56,
      "step": 32200,
      "token_acc": 0.45849802371541504,
      "train_speed(iter/s)": 1.440583
    },
    {
      "epoch": 1.3797609356925582,
      "grad_norm": 4.578446865081787,
      "learning_rate": 8.235852398207613e-05,
      "loss": 2.2881650924682617,
      "memory(GiB)": 77.56,
      "step": 32205,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.440618
    },
    {
      "epoch": 1.379975151021807,
      "grad_norm": 4.787532329559326,
      "learning_rate": 8.235339327807588e-05,
      "loss": 2.7093544006347656,
      "memory(GiB)": 77.56,
      "step": 32210,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440651
    },
    {
      "epoch": 1.380189366351056,
      "grad_norm": 10.675076484680176,
      "learning_rate": 8.234826198796357e-05,
      "loss": 2.125292205810547,
      "memory(GiB)": 77.56,
      "step": 32215,
      "token_acc": 0.5423076923076923,
      "train_speed(iter/s)": 1.440669
    },
    {
      "epoch": 1.380403581680305,
      "grad_norm": 5.266031265258789,
      "learning_rate": 8.234313011183215e-05,
      "loss": 2.485264778137207,
      "memory(GiB)": 77.56,
      "step": 32220,
      "token_acc": 0.46254071661237783,
      "train_speed(iter/s)": 1.440711
    },
    {
      "epoch": 1.380617797009554,
      "grad_norm": 6.192359447479248,
      "learning_rate": 8.233799764977457e-05,
      "loss": 2.22994327545166,
      "memory(GiB)": 77.56,
      "step": 32225,
      "token_acc": 0.5636363636363636,
      "train_speed(iter/s)": 1.440684
    },
    {
      "epoch": 1.380832012338803,
      "grad_norm": 5.870208740234375,
      "learning_rate": 8.233286460188385e-05,
      "loss": 2.5577566146850588,
      "memory(GiB)": 77.56,
      "step": 32230,
      "token_acc": 0.4197080291970803,
      "train_speed(iter/s)": 1.440687
    },
    {
      "epoch": 1.381046227668052,
      "grad_norm": 4.9664530754089355,
      "learning_rate": 8.232773096825295e-05,
      "loss": 2.8343246459960936,
      "memory(GiB)": 77.56,
      "step": 32235,
      "token_acc": 0.3935483870967742,
      "train_speed(iter/s)": 1.440698
    },
    {
      "epoch": 1.3812604429973008,
      "grad_norm": 4.283202648162842,
      "learning_rate": 8.232259674897486e-05,
      "loss": 2.66262264251709,
      "memory(GiB)": 77.56,
      "step": 32240,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.440743
    },
    {
      "epoch": 1.3814746583265498,
      "grad_norm": 4.331939220428467,
      "learning_rate": 8.231746194414262e-05,
      "loss": 2.4501779556274412,
      "memory(GiB)": 77.56,
      "step": 32245,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.440704
    },
    {
      "epoch": 1.3816888736557988,
      "grad_norm": 4.636662006378174,
      "learning_rate": 8.231232655384924e-05,
      "loss": 2.686531639099121,
      "memory(GiB)": 77.56,
      "step": 32250,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.440671
    },
    {
      "epoch": 1.3819030889850477,
      "grad_norm": 5.0237650871276855,
      "learning_rate": 8.230719057818776e-05,
      "loss": 2.477691078186035,
      "memory(GiB)": 77.56,
      "step": 32255,
      "token_acc": 0.45918367346938777,
      "train_speed(iter/s)": 1.440689
    },
    {
      "epoch": 1.3821173043142967,
      "grad_norm": 5.002540588378906,
      "learning_rate": 8.230205401725122e-05,
      "loss": 2.4150623321533202,
      "memory(GiB)": 77.56,
      "step": 32260,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.440739
    },
    {
      "epoch": 1.3823315196435457,
      "grad_norm": 5.962159633636475,
      "learning_rate": 8.229691687113266e-05,
      "loss": 2.570017623901367,
      "memory(GiB)": 77.56,
      "step": 32265,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.440804
    },
    {
      "epoch": 1.3825457349727945,
      "grad_norm": 5.570696830749512,
      "learning_rate": 8.229177913992515e-05,
      "loss": 2.691242980957031,
      "memory(GiB)": 77.56,
      "step": 32270,
      "token_acc": 0.4732824427480916,
      "train_speed(iter/s)": 1.440803
    },
    {
      "epoch": 1.3827599503020436,
      "grad_norm": 4.509525775909424,
      "learning_rate": 8.228664082372177e-05,
      "loss": 2.468734359741211,
      "memory(GiB)": 77.56,
      "step": 32275,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.440742
    },
    {
      "epoch": 1.3829741656312926,
      "grad_norm": 5.3111491203308105,
      "learning_rate": 8.22815019226156e-05,
      "loss": 2.464916801452637,
      "memory(GiB)": 77.56,
      "step": 32280,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.44077
    },
    {
      "epoch": 1.3831883809605414,
      "grad_norm": 6.811872482299805,
      "learning_rate": 8.227636243669976e-05,
      "loss": 2.6973987579345704,
      "memory(GiB)": 77.56,
      "step": 32285,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.440776
    },
    {
      "epoch": 1.3834025962897905,
      "grad_norm": 4.771921634674072,
      "learning_rate": 8.227122236606733e-05,
      "loss": 2.415981101989746,
      "memory(GiB)": 77.56,
      "step": 32290,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.440796
    },
    {
      "epoch": 1.3836168116190395,
      "grad_norm": 4.9913105964660645,
      "learning_rate": 8.226608171081143e-05,
      "loss": 2.625484657287598,
      "memory(GiB)": 77.56,
      "step": 32295,
      "token_acc": 0.47157190635451507,
      "train_speed(iter/s)": 1.440781
    },
    {
      "epoch": 1.3838310269482883,
      "grad_norm": 4.733423709869385,
      "learning_rate": 8.22609404710252e-05,
      "loss": 2.5317523956298826,
      "memory(GiB)": 77.56,
      "step": 32300,
      "token_acc": 0.4641638225255973,
      "train_speed(iter/s)": 1.440845
    },
    {
      "epoch": 1.3840452422775373,
      "grad_norm": 4.963057041168213,
      "learning_rate": 8.225579864680175e-05,
      "loss": 2.359029769897461,
      "memory(GiB)": 77.56,
      "step": 32305,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.440899
    },
    {
      "epoch": 1.3842594576067864,
      "grad_norm": 4.482953071594238,
      "learning_rate": 8.225065623823427e-05,
      "loss": 2.215733528137207,
      "memory(GiB)": 77.56,
      "step": 32310,
      "token_acc": 0.5486111111111112,
      "train_speed(iter/s)": 1.440909
    },
    {
      "epoch": 1.3844736729360352,
      "grad_norm": 5.477952480316162,
      "learning_rate": 8.224551324541588e-05,
      "loss": 2.6172725677490236,
      "memory(GiB)": 77.56,
      "step": 32315,
      "token_acc": 0.4316546762589928,
      "train_speed(iter/s)": 1.440946
    },
    {
      "epoch": 1.3846878882652842,
      "grad_norm": 5.942564010620117,
      "learning_rate": 8.224036966843978e-05,
      "loss": 2.522656059265137,
      "memory(GiB)": 77.56,
      "step": 32320,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.440926
    },
    {
      "epoch": 1.3849021035945333,
      "grad_norm": 4.379029750823975,
      "learning_rate": 8.223522550739913e-05,
      "loss": 2.588387298583984,
      "memory(GiB)": 77.56,
      "step": 32325,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.440923
    },
    {
      "epoch": 1.385116318923782,
      "grad_norm": 4.905298709869385,
      "learning_rate": 8.223008076238714e-05,
      "loss": 2.4599098205566405,
      "memory(GiB)": 77.56,
      "step": 32330,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.440909
    },
    {
      "epoch": 1.385330534253031,
      "grad_norm": 5.045111656188965,
      "learning_rate": 8.2224935433497e-05,
      "loss": 2.6597064971923827,
      "memory(GiB)": 77.56,
      "step": 32335,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.440963
    },
    {
      "epoch": 1.3855447495822801,
      "grad_norm": 5.351622581481934,
      "learning_rate": 8.221978952082192e-05,
      "loss": 2.9174943923950196,
      "memory(GiB)": 77.56,
      "step": 32340,
      "token_acc": 0.3825503355704698,
      "train_speed(iter/s)": 1.440917
    },
    {
      "epoch": 1.385758964911529,
      "grad_norm": 4.083662986755371,
      "learning_rate": 8.221464302445514e-05,
      "loss": 2.51528263092041,
      "memory(GiB)": 77.56,
      "step": 32345,
      "token_acc": 0.511864406779661,
      "train_speed(iter/s)": 1.440834
    },
    {
      "epoch": 1.385973180240778,
      "grad_norm": 3.813265323638916,
      "learning_rate": 8.220949594448987e-05,
      "loss": 2.3660558700561523,
      "memory(GiB)": 77.56,
      "step": 32350,
      "token_acc": 0.49271137026239065,
      "train_speed(iter/s)": 1.440826
    },
    {
      "epoch": 1.386187395570027,
      "grad_norm": 4.7663774490356445,
      "learning_rate": 8.220434828101937e-05,
      "loss": 2.4540462493896484,
      "memory(GiB)": 77.56,
      "step": 32355,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.440817
    },
    {
      "epoch": 1.3864016108992758,
      "grad_norm": 5.493171691894531,
      "learning_rate": 8.219920003413691e-05,
      "loss": 2.6703540802001955,
      "memory(GiB)": 77.56,
      "step": 32360,
      "token_acc": 0.47280334728033474,
      "train_speed(iter/s)": 1.440797
    },
    {
      "epoch": 1.3866158262285249,
      "grad_norm": 4.496729850769043,
      "learning_rate": 8.219405120393571e-05,
      "loss": 2.7621109008789064,
      "memory(GiB)": 77.56,
      "step": 32365,
      "token_acc": 0.43213296398891965,
      "train_speed(iter/s)": 1.44081
    },
    {
      "epoch": 1.386830041557774,
      "grad_norm": 4.299685478210449,
      "learning_rate": 8.218890179050908e-05,
      "loss": 2.7615472793579103,
      "memory(GiB)": 77.56,
      "step": 32370,
      "token_acc": 0.4471830985915493,
      "train_speed(iter/s)": 1.440865
    },
    {
      "epoch": 1.3870442568870227,
      "grad_norm": 5.835668087005615,
      "learning_rate": 8.21837517939503e-05,
      "loss": 2.6046798706054686,
      "memory(GiB)": 77.56,
      "step": 32375,
      "token_acc": 0.4470588235294118,
      "train_speed(iter/s)": 1.440828
    },
    {
      "epoch": 1.3872584722162717,
      "grad_norm": 4.56839656829834,
      "learning_rate": 8.217860121435267e-05,
      "loss": 2.5579978942871096,
      "memory(GiB)": 77.56,
      "step": 32380,
      "token_acc": 0.42662116040955633,
      "train_speed(iter/s)": 1.440853
    },
    {
      "epoch": 1.3874726875455208,
      "grad_norm": 4.46902322769165,
      "learning_rate": 8.217345005180949e-05,
      "loss": 2.3457687377929686,
      "memory(GiB)": 77.56,
      "step": 32385,
      "token_acc": 0.47157190635451507,
      "train_speed(iter/s)": 1.44087
    },
    {
      "epoch": 1.3876869028747696,
      "grad_norm": 5.586753845214844,
      "learning_rate": 8.216829830641408e-05,
      "loss": 2.686672592163086,
      "memory(GiB)": 77.56,
      "step": 32390,
      "token_acc": 0.42700729927007297,
      "train_speed(iter/s)": 1.440869
    },
    {
      "epoch": 1.3879011182040186,
      "grad_norm": 3.91251802444458,
      "learning_rate": 8.216314597825976e-05,
      "loss": 2.4910045623779298,
      "memory(GiB)": 77.56,
      "step": 32395,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.440875
    },
    {
      "epoch": 1.3881153335332677,
      "grad_norm": 4.722597122192383,
      "learning_rate": 8.215799306743989e-05,
      "loss": 2.690944862365723,
      "memory(GiB)": 77.56,
      "step": 32400,
      "token_acc": 0.4419889502762431,
      "train_speed(iter/s)": 1.440881
    },
    {
      "epoch": 1.3883295488625165,
      "grad_norm": 3.9358503818511963,
      "learning_rate": 8.21528395740478e-05,
      "loss": 2.3778587341308595,
      "memory(GiB)": 77.56,
      "step": 32405,
      "token_acc": 0.5346534653465347,
      "train_speed(iter/s)": 1.440894
    },
    {
      "epoch": 1.3885437641917655,
      "grad_norm": 5.521212577819824,
      "learning_rate": 8.214768549817687e-05,
      "loss": 2.614592933654785,
      "memory(GiB)": 77.56,
      "step": 32410,
      "token_acc": 0.43543543543543545,
      "train_speed(iter/s)": 1.440936
    },
    {
      "epoch": 1.3887579795210145,
      "grad_norm": 4.584962368011475,
      "learning_rate": 8.214253083992046e-05,
      "loss": 2.730294418334961,
      "memory(GiB)": 77.56,
      "step": 32415,
      "token_acc": 0.47477744807121663,
      "train_speed(iter/s)": 1.440973
    },
    {
      "epoch": 1.3889721948502634,
      "grad_norm": 5.215190410614014,
      "learning_rate": 8.213737559937195e-05,
      "loss": 2.5338584899902346,
      "memory(GiB)": 77.56,
      "step": 32420,
      "token_acc": 0.49640287769784175,
      "train_speed(iter/s)": 1.441031
    },
    {
      "epoch": 1.3891864101795124,
      "grad_norm": 4.041851997375488,
      "learning_rate": 8.213221977662473e-05,
      "loss": 2.630046272277832,
      "memory(GiB)": 77.56,
      "step": 32425,
      "token_acc": 0.45481927710843373,
      "train_speed(iter/s)": 1.441107
    },
    {
      "epoch": 1.3894006255087614,
      "grad_norm": 4.636109352111816,
      "learning_rate": 8.212706337177221e-05,
      "loss": 2.447910690307617,
      "memory(GiB)": 77.56,
      "step": 32430,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.441163
    },
    {
      "epoch": 1.3896148408380102,
      "grad_norm": 4.599270343780518,
      "learning_rate": 8.21219063849078e-05,
      "loss": 2.242801856994629,
      "memory(GiB)": 77.56,
      "step": 32435,
      "token_acc": 0.5222672064777328,
      "train_speed(iter/s)": 1.441111
    },
    {
      "epoch": 1.3898290561672593,
      "grad_norm": 3.935060739517212,
      "learning_rate": 8.211674881612492e-05,
      "loss": 2.7042808532714844,
      "memory(GiB)": 77.56,
      "step": 32440,
      "token_acc": 0.45686900958466453,
      "train_speed(iter/s)": 1.441185
    },
    {
      "epoch": 1.3900432714965083,
      "grad_norm": 4.77193546295166,
      "learning_rate": 8.211159066551701e-05,
      "loss": 2.424533653259277,
      "memory(GiB)": 77.56,
      "step": 32445,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.441161
    },
    {
      "epoch": 1.3902574868257571,
      "grad_norm": 7.982471466064453,
      "learning_rate": 8.210643193317751e-05,
      "loss": 2.8776296615600585,
      "memory(GiB)": 77.56,
      "step": 32450,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.44118
    },
    {
      "epoch": 1.3904717021550062,
      "grad_norm": 4.024241924285889,
      "learning_rate": 8.210127261919987e-05,
      "loss": 2.800585174560547,
      "memory(GiB)": 77.56,
      "step": 32455,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.441163
    },
    {
      "epoch": 1.3906859174842552,
      "grad_norm": 6.202906131744385,
      "learning_rate": 8.209611272367758e-05,
      "loss": 2.5529708862304688,
      "memory(GiB)": 77.56,
      "step": 32460,
      "token_acc": 0.4844290657439446,
      "train_speed(iter/s)": 1.441145
    },
    {
      "epoch": 1.390900132813504,
      "grad_norm": 6.042719841003418,
      "learning_rate": 8.209095224670409e-05,
      "loss": 2.1243961334228514,
      "memory(GiB)": 77.56,
      "step": 32465,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.441191
    },
    {
      "epoch": 1.391114348142753,
      "grad_norm": 4.469020843505859,
      "learning_rate": 8.208579118837289e-05,
      "loss": 2.5956594467163088,
      "memory(GiB)": 77.56,
      "step": 32470,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441226
    },
    {
      "epoch": 1.391328563472002,
      "grad_norm": 5.173557281494141,
      "learning_rate": 8.208062954877749e-05,
      "loss": 2.434360694885254,
      "memory(GiB)": 77.56,
      "step": 32475,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.44121
    },
    {
      "epoch": 1.3915427788012509,
      "grad_norm": 5.418885707855225,
      "learning_rate": 8.207546732801139e-05,
      "loss": 2.5377338409423826,
      "memory(GiB)": 77.56,
      "step": 32480,
      "token_acc": 0.4653846153846154,
      "train_speed(iter/s)": 1.441251
    },
    {
      "epoch": 1.3917569941305,
      "grad_norm": 5.52482795715332,
      "learning_rate": 8.207030452616811e-05,
      "loss": 2.545522689819336,
      "memory(GiB)": 77.56,
      "step": 32485,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.441324
    },
    {
      "epoch": 1.391971209459749,
      "grad_norm": 5.534440994262695,
      "learning_rate": 8.206514114334119e-05,
      "loss": 2.809461975097656,
      "memory(GiB)": 77.56,
      "step": 32490,
      "token_acc": 0.41580756013745707,
      "train_speed(iter/s)": 1.441262
    },
    {
      "epoch": 1.392185424788998,
      "grad_norm": 4.601513385772705,
      "learning_rate": 8.205997717962415e-05,
      "loss": 2.5163551330566407,
      "memory(GiB)": 77.56,
      "step": 32495,
      "token_acc": 0.5210355987055016,
      "train_speed(iter/s)": 1.441283
    },
    {
      "epoch": 1.3923996401182468,
      "grad_norm": 3.5171921253204346,
      "learning_rate": 8.205481263511054e-05,
      "loss": 2.668793487548828,
      "memory(GiB)": 77.56,
      "step": 32500,
      "token_acc": 0.463768115942029,
      "train_speed(iter/s)": 1.441309
    },
    {
      "epoch": 1.3923996401182468,
      "eval_loss": 2.177598714828491,
      "eval_runtime": 14.4616,
      "eval_samples_per_second": 6.915,
      "eval_steps_per_second": 6.915,
      "eval_token_acc": 0.4858681022880215,
      "step": 32500
    },
    {
      "epoch": 1.3926138554474958,
      "grad_norm": 4.515955924987793,
      "learning_rate": 8.204964750989393e-05,
      "loss": 2.348715400695801,
      "memory(GiB)": 77.56,
      "step": 32505,
      "token_acc": 0.4870067372473532,
      "train_speed(iter/s)": 1.440388
    },
    {
      "epoch": 1.3928280707767449,
      "grad_norm": 4.06764554977417,
      "learning_rate": 8.204448180406789e-05,
      "loss": 2.540558624267578,
      "memory(GiB)": 77.56,
      "step": 32510,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.440426
    },
    {
      "epoch": 1.3930422861059937,
      "grad_norm": 6.363366603851318,
      "learning_rate": 8.203931551772602e-05,
      "loss": 2.4556230545043944,
      "memory(GiB)": 77.56,
      "step": 32515,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.440418
    },
    {
      "epoch": 1.3932565014352427,
      "grad_norm": 4.377366065979004,
      "learning_rate": 8.203414865096188e-05,
      "loss": 2.5806402206420898,
      "memory(GiB)": 77.56,
      "step": 32520,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.440447
    },
    {
      "epoch": 1.3934707167644917,
      "grad_norm": 6.2758708000183105,
      "learning_rate": 8.20289812038691e-05,
      "loss": 2.291286087036133,
      "memory(GiB)": 77.56,
      "step": 32525,
      "token_acc": 0.5061224489795918,
      "train_speed(iter/s)": 1.440475
    },
    {
      "epoch": 1.3936849320937406,
      "grad_norm": 5.970769882202148,
      "learning_rate": 8.202381317654126e-05,
      "loss": 2.215692138671875,
      "memory(GiB)": 77.56,
      "step": 32530,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.440463
    },
    {
      "epoch": 1.3938991474229896,
      "grad_norm": 4.598732948303223,
      "learning_rate": 8.201864456907203e-05,
      "loss": 2.9922607421875,
      "memory(GiB)": 77.56,
      "step": 32535,
      "token_acc": 0.43440233236151604,
      "train_speed(iter/s)": 1.44049
    },
    {
      "epoch": 1.3941133627522386,
      "grad_norm": 6.370267391204834,
      "learning_rate": 8.201347538155499e-05,
      "loss": 2.4286373138427733,
      "memory(GiB)": 77.56,
      "step": 32540,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.3943275780814874,
      "grad_norm": 5.638474464416504,
      "learning_rate": 8.200830561408382e-05,
      "loss": 2.421356773376465,
      "memory(GiB)": 77.56,
      "step": 32545,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.440468
    },
    {
      "epoch": 1.3945417934107365,
      "grad_norm": 4.560729026794434,
      "learning_rate": 8.200313526675218e-05,
      "loss": 2.4186595916748046,
      "memory(GiB)": 77.56,
      "step": 32550,
      "token_acc": 0.4831804281345566,
      "train_speed(iter/s)": 1.440461
    },
    {
      "epoch": 1.3947560087399855,
      "grad_norm": 6.50902795791626,
      "learning_rate": 8.199796433965373e-05,
      "loss": 2.122319221496582,
      "memory(GiB)": 77.56,
      "step": 32555,
      "token_acc": 0.5096153846153846,
      "train_speed(iter/s)": 1.440477
    },
    {
      "epoch": 1.3949702240692343,
      "grad_norm": 14.841451644897461,
      "learning_rate": 8.19927928328821e-05,
      "loss": 2.5114162445068358,
      "memory(GiB)": 77.56,
      "step": 32560,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.440447
    },
    {
      "epoch": 1.3951844393984834,
      "grad_norm": 6.572962760925293,
      "learning_rate": 8.198762074653104e-05,
      "loss": 2.4489582061767576,
      "memory(GiB)": 77.56,
      "step": 32565,
      "token_acc": 0.4879032258064516,
      "train_speed(iter/s)": 1.440457
    },
    {
      "epoch": 1.3953986547277324,
      "grad_norm": 4.776227951049805,
      "learning_rate": 8.198244808069424e-05,
      "loss": 2.2974660873413084,
      "memory(GiB)": 77.56,
      "step": 32570,
      "token_acc": 0.5382059800664452,
      "train_speed(iter/s)": 1.440439
    },
    {
      "epoch": 1.3956128700569812,
      "grad_norm": 4.21104097366333,
      "learning_rate": 8.197727483546539e-05,
      "loss": 2.791429328918457,
      "memory(GiB)": 77.56,
      "step": 32575,
      "token_acc": 0.4373259052924791,
      "train_speed(iter/s)": 1.440494
    },
    {
      "epoch": 1.3958270853862302,
      "grad_norm": 5.432312965393066,
      "learning_rate": 8.197210101093817e-05,
      "loss": 2.6320913314819334,
      "memory(GiB)": 77.56,
      "step": 32580,
      "token_acc": 0.446875,
      "train_speed(iter/s)": 1.440505
    },
    {
      "epoch": 1.3960413007154793,
      "grad_norm": 4.993788719177246,
      "learning_rate": 8.196692660720638e-05,
      "loss": 2.4033786773681642,
      "memory(GiB)": 77.56,
      "step": 32585,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.440547
    },
    {
      "epoch": 1.396255516044728,
      "grad_norm": 4.767704486846924,
      "learning_rate": 8.196175162436371e-05,
      "loss": 2.764363098144531,
      "memory(GiB)": 77.56,
      "step": 32590,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.440512
    },
    {
      "epoch": 1.3964697313739771,
      "grad_norm": 5.701756477355957,
      "learning_rate": 8.195657606250393e-05,
      "loss": 2.627780532836914,
      "memory(GiB)": 77.56,
      "step": 32595,
      "token_acc": 0.4817073170731707,
      "train_speed(iter/s)": 1.440545
    },
    {
      "epoch": 1.3966839467032262,
      "grad_norm": 4.9805097579956055,
      "learning_rate": 8.195139992172081e-05,
      "loss": 2.1044809341430666,
      "memory(GiB)": 77.56,
      "step": 32600,
      "token_acc": 0.5258620689655172,
      "train_speed(iter/s)": 1.440588
    },
    {
      "epoch": 1.396898162032475,
      "grad_norm": 4.280782699584961,
      "learning_rate": 8.19462232021081e-05,
      "loss": 2.193564605712891,
      "memory(GiB)": 77.56,
      "step": 32605,
      "token_acc": 0.5415282392026578,
      "train_speed(iter/s)": 1.440626
    },
    {
      "epoch": 1.397112377361724,
      "grad_norm": 4.711809158325195,
      "learning_rate": 8.19410459037596e-05,
      "loss": 2.361937141418457,
      "memory(GiB)": 77.56,
      "step": 32610,
      "token_acc": 0.47876447876447875,
      "train_speed(iter/s)": 1.440659
    },
    {
      "epoch": 1.397326592690973,
      "grad_norm": 4.320030212402344,
      "learning_rate": 8.193586802676907e-05,
      "loss": 2.561391830444336,
      "memory(GiB)": 77.56,
      "step": 32615,
      "token_acc": 0.4636363636363636,
      "train_speed(iter/s)": 1.44065
    },
    {
      "epoch": 1.3975408080202218,
      "grad_norm": 4.283494472503662,
      "learning_rate": 8.193068957123034e-05,
      "loss": 2.7902252197265627,
      "memory(GiB)": 77.56,
      "step": 32620,
      "token_acc": 0.4479166666666667,
      "train_speed(iter/s)": 1.440666
    },
    {
      "epoch": 1.3977550233494709,
      "grad_norm": 5.214541435241699,
      "learning_rate": 8.192551053723721e-05,
      "loss": 2.545706558227539,
      "memory(GiB)": 77.56,
      "step": 32625,
      "token_acc": 0.4452296819787986,
      "train_speed(iter/s)": 1.440653
    },
    {
      "epoch": 1.39796923867872,
      "grad_norm": 5.056772232055664,
      "learning_rate": 8.192033092488351e-05,
      "loss": 2.578232192993164,
      "memory(GiB)": 77.56,
      "step": 32630,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.440686
    },
    {
      "epoch": 1.3981834540079687,
      "grad_norm": 5.8353800773620605,
      "learning_rate": 8.191515073426309e-05,
      "loss": 2.396602249145508,
      "memory(GiB)": 77.56,
      "step": 32635,
      "token_acc": 0.4195804195804196,
      "train_speed(iter/s)": 1.440712
    },
    {
      "epoch": 1.3983976693372178,
      "grad_norm": 5.054687023162842,
      "learning_rate": 8.190996996546975e-05,
      "loss": 2.968783378601074,
      "memory(GiB)": 77.56,
      "step": 32640,
      "token_acc": 0.4141791044776119,
      "train_speed(iter/s)": 1.440704
    },
    {
      "epoch": 1.3986118846664668,
      "grad_norm": 5.241652488708496,
      "learning_rate": 8.19047886185974e-05,
      "loss": 2.658024024963379,
      "memory(GiB)": 77.56,
      "step": 32645,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.440668
    },
    {
      "epoch": 1.3988260999957156,
      "grad_norm": 4.599843502044678,
      "learning_rate": 8.189960669373987e-05,
      "loss": 2.4044450759887694,
      "memory(GiB)": 77.56,
      "step": 32650,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.440713
    },
    {
      "epoch": 1.3990403153249646,
      "grad_norm": 4.849252700805664,
      "learning_rate": 8.189442419099101e-05,
      "loss": 2.7404220581054686,
      "memory(GiB)": 77.56,
      "step": 32655,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440777
    },
    {
      "epoch": 1.3992545306542137,
      "grad_norm": 6.469464302062988,
      "learning_rate": 8.188924111044476e-05,
      "loss": 2.2504724502563476,
      "memory(GiB)": 77.56,
      "step": 32660,
      "token_acc": 0.5982532751091703,
      "train_speed(iter/s)": 1.440807
    },
    {
      "epoch": 1.3994687459834627,
      "grad_norm": 4.850805282592773,
      "learning_rate": 8.188405745219498e-05,
      "loss": 2.551934814453125,
      "memory(GiB)": 77.56,
      "step": 32665,
      "token_acc": 0.4977578475336323,
      "train_speed(iter/s)": 1.440771
    },
    {
      "epoch": 1.3996829613127115,
      "grad_norm": 4.660606384277344,
      "learning_rate": 8.18788732163356e-05,
      "loss": 2.1702070236206055,
      "memory(GiB)": 77.56,
      "step": 32670,
      "token_acc": 0.5355805243445693,
      "train_speed(iter/s)": 1.440814
    },
    {
      "epoch": 1.3998971766419606,
      "grad_norm": 4.29949426651001,
      "learning_rate": 8.187368840296052e-05,
      "loss": 2.586586570739746,
      "memory(GiB)": 77.56,
      "step": 32675,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.44084
    },
    {
      "epoch": 1.4001113919712096,
      "grad_norm": 5.140963554382324,
      "learning_rate": 8.186850301216368e-05,
      "loss": 2.3898927688598635,
      "memory(GiB)": 77.56,
      "step": 32680,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.440847
    },
    {
      "epoch": 1.4003256073004584,
      "grad_norm": 4.6134033203125,
      "learning_rate": 8.186331704403902e-05,
      "loss": 2.4320209503173826,
      "memory(GiB)": 77.56,
      "step": 32685,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.440869
    },
    {
      "epoch": 1.4005398226297074,
      "grad_norm": 4.765467166900635,
      "learning_rate": 8.185813049868045e-05,
      "loss": 2.3984859466552733,
      "memory(GiB)": 77.56,
      "step": 32690,
      "token_acc": 0.4404432132963989,
      "train_speed(iter/s)": 1.440899
    },
    {
      "epoch": 1.4007540379589565,
      "grad_norm": 4.149876117706299,
      "learning_rate": 8.185294337618198e-05,
      "loss": 2.3167816162109376,
      "memory(GiB)": 77.56,
      "step": 32695,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.44092
    },
    {
      "epoch": 1.4009682532882053,
      "grad_norm": 4.961634159088135,
      "learning_rate": 8.184775567663759e-05,
      "loss": 2.56131649017334,
      "memory(GiB)": 77.56,
      "step": 32700,
      "token_acc": 0.4847560975609756,
      "train_speed(iter/s)": 1.440923
    },
    {
      "epoch": 1.4011824686174543,
      "grad_norm": 4.568178653717041,
      "learning_rate": 8.18425674001412e-05,
      "loss": 2.2430482864379884,
      "memory(GiB)": 77.56,
      "step": 32705,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.440958
    },
    {
      "epoch": 1.4013966839467034,
      "grad_norm": 4.932185649871826,
      "learning_rate": 8.183737854678684e-05,
      "loss": 2.4853187561035157,
      "memory(GiB)": 77.56,
      "step": 32710,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.44099
    },
    {
      "epoch": 1.4016108992759522,
      "grad_norm": 4.6252217292785645,
      "learning_rate": 8.183218911666849e-05,
      "loss": 2.572653579711914,
      "memory(GiB)": 77.56,
      "step": 32715,
      "token_acc": 0.452755905511811,
      "train_speed(iter/s)": 1.440979
    },
    {
      "epoch": 1.4018251146052012,
      "grad_norm": 5.366158485412598,
      "learning_rate": 8.182699910988018e-05,
      "loss": 2.667098045349121,
      "memory(GiB)": 77.56,
      "step": 32720,
      "token_acc": 0.4454828660436137,
      "train_speed(iter/s)": 1.441022
    },
    {
      "epoch": 1.4020393299344502,
      "grad_norm": 4.5241007804870605,
      "learning_rate": 8.182180852651592e-05,
      "loss": 2.3706123352050783,
      "memory(GiB)": 77.56,
      "step": 32725,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.441031
    },
    {
      "epoch": 1.402253545263699,
      "grad_norm": 4.363215923309326,
      "learning_rate": 8.181661736666974e-05,
      "loss": 2.3623613357543944,
      "memory(GiB)": 77.56,
      "step": 32730,
      "token_acc": 0.5265151515151515,
      "train_speed(iter/s)": 1.441057
    },
    {
      "epoch": 1.402467760592948,
      "grad_norm": 5.448582172393799,
      "learning_rate": 8.181142563043572e-05,
      "loss": 2.347551727294922,
      "memory(GiB)": 77.56,
      "step": 32735,
      "token_acc": 0.5354609929078015,
      "train_speed(iter/s)": 1.441071
    },
    {
      "epoch": 1.4026819759221971,
      "grad_norm": 5.197177410125732,
      "learning_rate": 8.180623331790785e-05,
      "loss": 2.276912498474121,
      "memory(GiB)": 77.56,
      "step": 32740,
      "token_acc": 0.5154185022026432,
      "train_speed(iter/s)": 1.441112
    },
    {
      "epoch": 1.402896191251446,
      "grad_norm": 5.24609375,
      "learning_rate": 8.180104042918025e-05,
      "loss": 2.6191478729248048,
      "memory(GiB)": 77.56,
      "step": 32745,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.441086
    },
    {
      "epoch": 1.403110406580695,
      "grad_norm": 6.560999393463135,
      "learning_rate": 8.179584696434696e-05,
      "loss": 2.5139934539794924,
      "memory(GiB)": 77.56,
      "step": 32750,
      "token_acc": 0.4701492537313433,
      "train_speed(iter/s)": 1.441
    },
    {
      "epoch": 1.403324621909944,
      "grad_norm": 4.0000152587890625,
      "learning_rate": 8.179065292350208e-05,
      "loss": 2.457881736755371,
      "memory(GiB)": 77.56,
      "step": 32755,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.441007
    },
    {
      "epoch": 1.4035388372391928,
      "grad_norm": 6.2242889404296875,
      "learning_rate": 8.178545830673969e-05,
      "loss": 2.547452926635742,
      "memory(GiB)": 77.56,
      "step": 32760,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.44097
    },
    {
      "epoch": 1.4037530525684419,
      "grad_norm": 4.187480926513672,
      "learning_rate": 8.178026311415392e-05,
      "loss": 2.6834732055664063,
      "memory(GiB)": 77.56,
      "step": 32765,
      "token_acc": 0.4447852760736196,
      "train_speed(iter/s)": 1.440991
    },
    {
      "epoch": 1.4039672678976909,
      "grad_norm": 5.03083610534668,
      "learning_rate": 8.177506734583886e-05,
      "loss": 2.808266448974609,
      "memory(GiB)": 77.56,
      "step": 32770,
      "token_acc": 0.4764705882352941,
      "train_speed(iter/s)": 1.441039
    },
    {
      "epoch": 1.4041814832269397,
      "grad_norm": 5.288457870483398,
      "learning_rate": 8.176987100188865e-05,
      "loss": 2.6233535766601563,
      "memory(GiB)": 77.56,
      "step": 32775,
      "token_acc": 0.4312267657992565,
      "train_speed(iter/s)": 1.441043
    },
    {
      "epoch": 1.4043956985561887,
      "grad_norm": 5.455574989318848,
      "learning_rate": 8.176467408239743e-05,
      "loss": 2.259764862060547,
      "memory(GiB)": 77.56,
      "step": 32780,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.441033
    },
    {
      "epoch": 1.4046099138854378,
      "grad_norm": 5.3393778800964355,
      "learning_rate": 8.175947658745934e-05,
      "loss": 2.104327583312988,
      "memory(GiB)": 77.56,
      "step": 32785,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.440987
    },
    {
      "epoch": 1.4048241292146866,
      "grad_norm": 4.289956569671631,
      "learning_rate": 8.175427851716855e-05,
      "loss": 2.731656074523926,
      "memory(GiB)": 77.56,
      "step": 32790,
      "token_acc": 0.4573170731707317,
      "train_speed(iter/s)": 1.441037
    },
    {
      "epoch": 1.4050383445439356,
      "grad_norm": 4.7855329513549805,
      "learning_rate": 8.17490798716192e-05,
      "loss": 2.6186155319213866,
      "memory(GiB)": 77.56,
      "step": 32795,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.44104
    },
    {
      "epoch": 1.4052525598731846,
      "grad_norm": 7.545797348022461,
      "learning_rate": 8.174388065090548e-05,
      "loss": 2.2450260162353515,
      "memory(GiB)": 77.56,
      "step": 32800,
      "token_acc": 0.5473251028806584,
      "train_speed(iter/s)": 1.441069
    },
    {
      "epoch": 1.4054667752024335,
      "grad_norm": 4.962194442749023,
      "learning_rate": 8.17386808551216e-05,
      "loss": 2.2483697891235352,
      "memory(GiB)": 77.56,
      "step": 32805,
      "token_acc": 0.48493975903614456,
      "train_speed(iter/s)": 1.441069
    },
    {
      "epoch": 1.4056809905316825,
      "grad_norm": 4.5178937911987305,
      "learning_rate": 8.173348048436174e-05,
      "loss": 2.231859016418457,
      "memory(GiB)": 77.56,
      "step": 32810,
      "token_acc": 0.5324675324675324,
      "train_speed(iter/s)": 1.441049
    },
    {
      "epoch": 1.4058952058609315,
      "grad_norm": 4.872297763824463,
      "learning_rate": 8.17282795387201e-05,
      "loss": 2.5846189498901366,
      "memory(GiB)": 77.56,
      "step": 32815,
      "token_acc": 0.45394736842105265,
      "train_speed(iter/s)": 1.441021
    },
    {
      "epoch": 1.4061094211901803,
      "grad_norm": 5.8877129554748535,
      "learning_rate": 8.172307801829093e-05,
      "loss": 2.6296873092651367,
      "memory(GiB)": 77.56,
      "step": 32820,
      "token_acc": 0.4628099173553719,
      "train_speed(iter/s)": 1.441037
    },
    {
      "epoch": 1.4063236365194294,
      "grad_norm": 5.066899299621582,
      "learning_rate": 8.171787592316842e-05,
      "loss": 2.498503494262695,
      "memory(GiB)": 77.56,
      "step": 32825,
      "token_acc": 0.49390243902439024,
      "train_speed(iter/s)": 1.440998
    },
    {
      "epoch": 1.4065378518486784,
      "grad_norm": 7.721127033233643,
      "learning_rate": 8.171267325344685e-05,
      "loss": 2.479060745239258,
      "memory(GiB)": 77.56,
      "step": 32830,
      "token_acc": 0.5301724137931034,
      "train_speed(iter/s)": 1.441068
    },
    {
      "epoch": 1.4067520671779272,
      "grad_norm": 5.018352031707764,
      "learning_rate": 8.170747000922045e-05,
      "loss": 2.431307792663574,
      "memory(GiB)": 77.56,
      "step": 32835,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.440997
    },
    {
      "epoch": 1.4069662825071763,
      "grad_norm": 6.421723365783691,
      "learning_rate": 8.17022661905835e-05,
      "loss": 3.0423044204711913,
      "memory(GiB)": 77.56,
      "step": 32840,
      "token_acc": 0.37987012987012986,
      "train_speed(iter/s)": 1.441042
    },
    {
      "epoch": 1.4071804978364253,
      "grad_norm": 4.114084243774414,
      "learning_rate": 8.169706179763023e-05,
      "loss": 2.1239709854125977,
      "memory(GiB)": 77.56,
      "step": 32845,
      "token_acc": 0.5403508771929825,
      "train_speed(iter/s)": 1.44102
    },
    {
      "epoch": 1.407394713165674,
      "grad_norm": 4.6386494636535645,
      "learning_rate": 8.169185683045498e-05,
      "loss": 2.3217145919799806,
      "memory(GiB)": 77.56,
      "step": 32850,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.441017
    },
    {
      "epoch": 1.4076089284949231,
      "grad_norm": 4.626767635345459,
      "learning_rate": 8.168665128915201e-05,
      "loss": 2.830036163330078,
      "memory(GiB)": 77.56,
      "step": 32855,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.441036
    },
    {
      "epoch": 1.4078231438241722,
      "grad_norm": 5.563905715942383,
      "learning_rate": 8.168144517381562e-05,
      "loss": 2.4487443923950196,
      "memory(GiB)": 77.56,
      "step": 32860,
      "token_acc": 0.4735202492211838,
      "train_speed(iter/s)": 1.441063
    },
    {
      "epoch": 1.408037359153421,
      "grad_norm": 5.357914447784424,
      "learning_rate": 8.167623848454014e-05,
      "loss": 2.568429183959961,
      "memory(GiB)": 77.56,
      "step": 32865,
      "token_acc": 0.46785714285714286,
      "train_speed(iter/s)": 1.441117
    },
    {
      "epoch": 1.40825157448267,
      "grad_norm": 4.455787181854248,
      "learning_rate": 8.167103122141988e-05,
      "loss": 2.4503332138061524,
      "memory(GiB)": 77.56,
      "step": 32870,
      "token_acc": 0.46060606060606063,
      "train_speed(iter/s)": 1.44118
    },
    {
      "epoch": 1.408465789811919,
      "grad_norm": 5.0956573486328125,
      "learning_rate": 8.166582338454917e-05,
      "loss": 2.6250228881835938,
      "memory(GiB)": 77.56,
      "step": 32875,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.441161
    },
    {
      "epoch": 1.4086800051411679,
      "grad_norm": 4.343917369842529,
      "learning_rate": 8.166061497402236e-05,
      "loss": 2.610627365112305,
      "memory(GiB)": 77.56,
      "step": 32880,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.441152
    },
    {
      "epoch": 1.408894220470417,
      "grad_norm": 4.916111946105957,
      "learning_rate": 8.165540598993383e-05,
      "loss": 2.6820501327514648,
      "memory(GiB)": 77.56,
      "step": 32885,
      "token_acc": 0.44931506849315067,
      "train_speed(iter/s)": 1.441135
    },
    {
      "epoch": 1.409108435799666,
      "grad_norm": 9.075296401977539,
      "learning_rate": 8.165019643237792e-05,
      "loss": 2.5452728271484375,
      "memory(GiB)": 77.56,
      "step": 32890,
      "token_acc": 0.4959677419354839,
      "train_speed(iter/s)": 1.441151
    },
    {
      "epoch": 1.4093226511289147,
      "grad_norm": 4.417697429656982,
      "learning_rate": 8.164498630144901e-05,
      "loss": 2.4722240447998045,
      "memory(GiB)": 77.56,
      "step": 32895,
      "token_acc": 0.47246376811594204,
      "train_speed(iter/s)": 1.441154
    },
    {
      "epoch": 1.4095368664581638,
      "grad_norm": 5.6175947189331055,
      "learning_rate": 8.163977559724147e-05,
      "loss": 2.930836486816406,
      "memory(GiB)": 77.56,
      "step": 32900,
      "token_acc": 0.44272445820433437,
      "train_speed(iter/s)": 1.441204
    },
    {
      "epoch": 1.4097510817874128,
      "grad_norm": 7.489553451538086,
      "learning_rate": 8.163456431984975e-05,
      "loss": 2.776253890991211,
      "memory(GiB)": 77.56,
      "step": 32905,
      "token_acc": 0.4303030303030303,
      "train_speed(iter/s)": 1.441115
    },
    {
      "epoch": 1.4099652971166616,
      "grad_norm": 4.295735836029053,
      "learning_rate": 8.162935246936819e-05,
      "loss": 2.4325775146484374,
      "memory(GiB)": 77.56,
      "step": 32910,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.44115
    },
    {
      "epoch": 1.4101795124459107,
      "grad_norm": 3.759232997894287,
      "learning_rate": 8.162414004589126e-05,
      "loss": 2.215176010131836,
      "memory(GiB)": 77.56,
      "step": 32915,
      "token_acc": 0.5390334572490706,
      "train_speed(iter/s)": 1.441196
    },
    {
      "epoch": 1.4103937277751597,
      "grad_norm": 5.306053638458252,
      "learning_rate": 8.161892704951334e-05,
      "loss": 2.668818473815918,
      "memory(GiB)": 77.56,
      "step": 32920,
      "token_acc": 0.4576271186440678,
      "train_speed(iter/s)": 1.441198
    },
    {
      "epoch": 1.4106079431044085,
      "grad_norm": 5.074545383453369,
      "learning_rate": 8.161371348032893e-05,
      "loss": 2.4916919708251952,
      "memory(GiB)": 77.56,
      "step": 32925,
      "token_acc": 0.4456140350877193,
      "train_speed(iter/s)": 1.441155
    },
    {
      "epoch": 1.4108221584336575,
      "grad_norm": 5.683573246002197,
      "learning_rate": 8.160849933843241e-05,
      "loss": 2.3398935317993166,
      "memory(GiB)": 77.56,
      "step": 32930,
      "token_acc": 0.5175097276264592,
      "train_speed(iter/s)": 1.441137
    },
    {
      "epoch": 1.4110363737629066,
      "grad_norm": 4.769052505493164,
      "learning_rate": 8.16032846239183e-05,
      "loss": 2.4615779876708985,
      "memory(GiB)": 77.56,
      "step": 32935,
      "token_acc": 0.4789272030651341,
      "train_speed(iter/s)": 1.441115
    },
    {
      "epoch": 1.4112505890921554,
      "grad_norm": 4.571069240570068,
      "learning_rate": 8.159806933688105e-05,
      "loss": 2.540538024902344,
      "memory(GiB)": 77.56,
      "step": 32940,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.441116
    },
    {
      "epoch": 1.4114648044214044,
      "grad_norm": 5.50850248336792,
      "learning_rate": 8.159285347741513e-05,
      "loss": 2.632329559326172,
      "memory(GiB)": 77.56,
      "step": 32945,
      "token_acc": 0.4703703703703704,
      "train_speed(iter/s)": 1.441137
    },
    {
      "epoch": 1.4116790197506535,
      "grad_norm": 5.724573135375977,
      "learning_rate": 8.1587637045615e-05,
      "loss": 2.593389320373535,
      "memory(GiB)": 77.56,
      "step": 32950,
      "token_acc": 0.46546546546546547,
      "train_speed(iter/s)": 1.441098
    },
    {
      "epoch": 1.4118932350799023,
      "grad_norm": 5.0343523025512695,
      "learning_rate": 8.158242004157522e-05,
      "loss": 2.5726451873779297,
      "memory(GiB)": 77.56,
      "step": 32955,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.441094
    },
    {
      "epoch": 1.4121074504091513,
      "grad_norm": 4.84677791595459,
      "learning_rate": 8.157720246539026e-05,
      "loss": 2.5698471069335938,
      "memory(GiB)": 77.56,
      "step": 32960,
      "token_acc": 0.4552238805970149,
      "train_speed(iter/s)": 1.441055
    },
    {
      "epoch": 1.4123216657384003,
      "grad_norm": 4.837327003479004,
      "learning_rate": 8.157198431715466e-05,
      "loss": 2.5950782775878904,
      "memory(GiB)": 77.56,
      "step": 32965,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.44109
    },
    {
      "epoch": 1.4125358810676492,
      "grad_norm": 3.9810128211975098,
      "learning_rate": 8.156676559696294e-05,
      "loss": 2.3519641876220705,
      "memory(GiB)": 77.56,
      "step": 32970,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.441078
    },
    {
      "epoch": 1.4127500963968982,
      "grad_norm": 4.564142227172852,
      "learning_rate": 8.156154630490968e-05,
      "loss": 2.6326671600341798,
      "memory(GiB)": 77.56,
      "step": 32975,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.441076
    },
    {
      "epoch": 1.4129643117261472,
      "grad_norm": 3.8186042308807373,
      "learning_rate": 8.155632644108937e-05,
      "loss": 2.6206520080566404,
      "memory(GiB)": 77.56,
      "step": 32980,
      "token_acc": 0.49693251533742333,
      "train_speed(iter/s)": 1.441023
    },
    {
      "epoch": 1.413178527055396,
      "grad_norm": 4.897378921508789,
      "learning_rate": 8.155110600559661e-05,
      "loss": 2.395633316040039,
      "memory(GiB)": 77.56,
      "step": 32985,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.441093
    },
    {
      "epoch": 1.413392742384645,
      "grad_norm": 4.5664753913879395,
      "learning_rate": 8.154588499852598e-05,
      "loss": 2.5076459884643554,
      "memory(GiB)": 77.56,
      "step": 32990,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.441117
    },
    {
      "epoch": 1.413606957713894,
      "grad_norm": 4.726290702819824,
      "learning_rate": 8.154066341997204e-05,
      "loss": 2.6111370086669923,
      "memory(GiB)": 77.56,
      "step": 32995,
      "token_acc": 0.44787644787644787,
      "train_speed(iter/s)": 1.441114
    },
    {
      "epoch": 1.413821173043143,
      "grad_norm": 5.217080116271973,
      "learning_rate": 8.15354412700294e-05,
      "loss": 2.572787857055664,
      "memory(GiB)": 77.56,
      "step": 33000,
      "token_acc": 0.5170940170940171,
      "train_speed(iter/s)": 1.441136
    },
    {
      "epoch": 1.413821173043143,
      "eval_loss": 2.155951499938965,
      "eval_runtime": 14.7972,
      "eval_samples_per_second": 6.758,
      "eval_steps_per_second": 6.758,
      "eval_token_acc": 0.4605263157894737,
      "step": 33000
    },
    {
      "epoch": 1.414035388372392,
      "grad_norm": 4.319518566131592,
      "learning_rate": 8.153021854879266e-05,
      "loss": 2.4434165954589844,
      "memory(GiB)": 77.56,
      "step": 33005,
      "token_acc": 0.4681647940074906,
      "train_speed(iter/s)": 1.440084
    },
    {
      "epoch": 1.414249603701641,
      "grad_norm": 5.58959436416626,
      "learning_rate": 8.152499525635643e-05,
      "loss": 2.5024646759033202,
      "memory(GiB)": 77.56,
      "step": 33010,
      "token_acc": 0.42105263157894735,
      "train_speed(iter/s)": 1.440127
    },
    {
      "epoch": 1.4144638190308898,
      "grad_norm": 7.674812316894531,
      "learning_rate": 8.151977139281534e-05,
      "loss": 2.3606868743896485,
      "memory(GiB)": 77.56,
      "step": 33015,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.440153
    },
    {
      "epoch": 1.4146780343601388,
      "grad_norm": 4.988476753234863,
      "learning_rate": 8.151454695826402e-05,
      "loss": 2.613323974609375,
      "memory(GiB)": 77.56,
      "step": 33020,
      "token_acc": 0.4542124542124542,
      "train_speed(iter/s)": 1.440207
    },
    {
      "epoch": 1.4148922496893879,
      "grad_norm": 4.285540580749512,
      "learning_rate": 8.150932195279711e-05,
      "loss": 2.3553529739379884,
      "memory(GiB)": 77.56,
      "step": 33025,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.440218
    },
    {
      "epoch": 1.4151064650186367,
      "grad_norm": 5.295781135559082,
      "learning_rate": 8.150409637650928e-05,
      "loss": 2.2589305877685546,
      "memory(GiB)": 77.56,
      "step": 33030,
      "token_acc": 0.5281385281385281,
      "train_speed(iter/s)": 1.440193
    },
    {
      "epoch": 1.4153206803478857,
      "grad_norm": 4.3361921310424805,
      "learning_rate": 8.14988702294952e-05,
      "loss": 2.5230279922485352,
      "memory(GiB)": 77.56,
      "step": 33035,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.440169
    },
    {
      "epoch": 1.4155348956771348,
      "grad_norm": 4.904618740081787,
      "learning_rate": 8.149364351184953e-05,
      "loss": 2.4007289886474608,
      "memory(GiB)": 77.56,
      "step": 33040,
      "token_acc": 0.494949494949495,
      "train_speed(iter/s)": 1.440135
    },
    {
      "epoch": 1.4157491110063836,
      "grad_norm": 4.181363105773926,
      "learning_rate": 8.148841622366699e-05,
      "loss": 2.517375946044922,
      "memory(GiB)": 77.56,
      "step": 33045,
      "token_acc": 0.4579710144927536,
      "train_speed(iter/s)": 1.440093
    },
    {
      "epoch": 1.4159633263356326,
      "grad_norm": 5.231570243835449,
      "learning_rate": 8.148318836504221e-05,
      "loss": 2.3261022567749023,
      "memory(GiB)": 77.56,
      "step": 33050,
      "token_acc": 0.5050505050505051,
      "train_speed(iter/s)": 1.440134
    },
    {
      "epoch": 1.4161775416648816,
      "grad_norm": 5.6807379722595215,
      "learning_rate": 8.147795993606997e-05,
      "loss": 2.480241394042969,
      "memory(GiB)": 77.56,
      "step": 33055,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.440165
    },
    {
      "epoch": 1.4163917569941304,
      "grad_norm": 4.6538004875183105,
      "learning_rate": 8.147273093684493e-05,
      "loss": 2.358638954162598,
      "memory(GiB)": 77.56,
      "step": 33060,
      "token_acc": 0.5291970802919708,
      "train_speed(iter/s)": 1.440204
    },
    {
      "epoch": 1.4166059723233795,
      "grad_norm": 4.609269142150879,
      "learning_rate": 8.146750136746187e-05,
      "loss": 2.7514999389648436,
      "memory(GiB)": 77.56,
      "step": 33065,
      "token_acc": 0.4192546583850932,
      "train_speed(iter/s)": 1.440227
    },
    {
      "epoch": 1.4168201876526285,
      "grad_norm": 5.511987686157227,
      "learning_rate": 8.14622712280155e-05,
      "loss": 2.4519065856933593,
      "memory(GiB)": 77.56,
      "step": 33070,
      "token_acc": 0.4730290456431535,
      "train_speed(iter/s)": 1.440255
    },
    {
      "epoch": 1.4170344029818773,
      "grad_norm": 4.408252716064453,
      "learning_rate": 8.145704051860056e-05,
      "loss": 2.5872501373291015,
      "memory(GiB)": 77.56,
      "step": 33075,
      "token_acc": 0.4855072463768116,
      "train_speed(iter/s)": 1.440206
    },
    {
      "epoch": 1.4172486183111264,
      "grad_norm": 4.394386291503906,
      "learning_rate": 8.145180923931184e-05,
      "loss": 2.4042551040649416,
      "memory(GiB)": 77.56,
      "step": 33080,
      "token_acc": 0.46075085324232085,
      "train_speed(iter/s)": 1.440248
    },
    {
      "epoch": 1.4174628336403754,
      "grad_norm": 5.2622389793396,
      "learning_rate": 8.144657739024408e-05,
      "loss": 2.613916778564453,
      "memory(GiB)": 77.56,
      "step": 33085,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.440314
    },
    {
      "epoch": 1.4176770489696242,
      "grad_norm": 3.2025272846221924,
      "learning_rate": 8.144134497149207e-05,
      "loss": 2.5696949005126952,
      "memory(GiB)": 77.56,
      "step": 33090,
      "token_acc": 0.4879518072289157,
      "train_speed(iter/s)": 1.440379
    },
    {
      "epoch": 1.4178912642988732,
      "grad_norm": 5.146648406982422,
      "learning_rate": 8.14361119831506e-05,
      "loss": 2.5399837493896484,
      "memory(GiB)": 77.56,
      "step": 33095,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.440416
    },
    {
      "epoch": 1.4181054796281223,
      "grad_norm": 4.936645984649658,
      "learning_rate": 8.143087842531447e-05,
      "loss": 2.317360687255859,
      "memory(GiB)": 77.56,
      "step": 33100,
      "token_acc": 0.5159010600706714,
      "train_speed(iter/s)": 1.44048
    },
    {
      "epoch": 1.418319694957371,
      "grad_norm": 4.946645259857178,
      "learning_rate": 8.142564429807851e-05,
      "loss": 2.5198184967041017,
      "memory(GiB)": 77.56,
      "step": 33105,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.440442
    },
    {
      "epoch": 1.4185339102866201,
      "grad_norm": 8.604920387268066,
      "learning_rate": 8.142040960153749e-05,
      "loss": 2.4088836669921876,
      "memory(GiB)": 77.56,
      "step": 33110,
      "token_acc": 0.47345132743362833,
      "train_speed(iter/s)": 1.440396
    },
    {
      "epoch": 1.4187481256158692,
      "grad_norm": 3.7186927795410156,
      "learning_rate": 8.141517433578632e-05,
      "loss": 2.599583053588867,
      "memory(GiB)": 77.56,
      "step": 33115,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.440343
    },
    {
      "epoch": 1.418962340945118,
      "grad_norm": 6.1346845626831055,
      "learning_rate": 8.140993850091977e-05,
      "loss": 2.4547145843505858,
      "memory(GiB)": 77.56,
      "step": 33120,
      "token_acc": 0.453416149068323,
      "train_speed(iter/s)": 1.440381
    },
    {
      "epoch": 1.419176556274367,
      "grad_norm": 3.691970109939575,
      "learning_rate": 8.140470209703273e-05,
      "loss": 2.4248403549194335,
      "memory(GiB)": 77.56,
      "step": 33125,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.440413
    },
    {
      "epoch": 1.419390771603616,
      "grad_norm": 4.270996570587158,
      "learning_rate": 8.139946512422003e-05,
      "loss": 2.259629249572754,
      "memory(GiB)": 77.56,
      "step": 33130,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.44043
    },
    {
      "epoch": 1.4196049869328649,
      "grad_norm": 4.1327033042907715,
      "learning_rate": 8.139422758257658e-05,
      "loss": 2.228342819213867,
      "memory(GiB)": 77.56,
      "step": 33135,
      "token_acc": 0.5047923322683706,
      "train_speed(iter/s)": 1.440342
    },
    {
      "epoch": 1.4198192022621139,
      "grad_norm": 5.204843521118164,
      "learning_rate": 8.138898947219724e-05,
      "loss": 2.8300018310546875,
      "memory(GiB)": 77.56,
      "step": 33140,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.440369
    },
    {
      "epoch": 1.420033417591363,
      "grad_norm": 6.835001468658447,
      "learning_rate": 8.138375079317693e-05,
      "loss": 2.3787372589111326,
      "memory(GiB)": 77.56,
      "step": 33145,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.440396
    },
    {
      "epoch": 1.4202476329206117,
      "grad_norm": 6.716424465179443,
      "learning_rate": 8.137851154561051e-05,
      "loss": 2.496069145202637,
      "memory(GiB)": 77.56,
      "step": 33150,
      "token_acc": 0.4536082474226804,
      "train_speed(iter/s)": 1.440445
    },
    {
      "epoch": 1.4204618482498608,
      "grad_norm": 4.852542877197266,
      "learning_rate": 8.137327172959294e-05,
      "loss": 2.508481979370117,
      "memory(GiB)": 77.56,
      "step": 33155,
      "token_acc": 0.4758364312267658,
      "train_speed(iter/s)": 1.440472
    },
    {
      "epoch": 1.4206760635791098,
      "grad_norm": 4.746988773345947,
      "learning_rate": 8.136803134521912e-05,
      "loss": 2.7087955474853516,
      "memory(GiB)": 77.56,
      "step": 33160,
      "token_acc": 0.44904458598726116,
      "train_speed(iter/s)": 1.440505
    },
    {
      "epoch": 1.4208902789083586,
      "grad_norm": 4.003751277923584,
      "learning_rate": 8.136279039258397e-05,
      "loss": 2.5319950103759767,
      "memory(GiB)": 77.56,
      "step": 33165,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.440556
    },
    {
      "epoch": 1.4211044942376077,
      "grad_norm": 5.177649974822998,
      "learning_rate": 8.135754887178246e-05,
      "loss": 2.4351833343505858,
      "memory(GiB)": 77.56,
      "step": 33170,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.440596
    },
    {
      "epoch": 1.4213187095668567,
      "grad_norm": 4.527562141418457,
      "learning_rate": 8.135230678290953e-05,
      "loss": 2.467445945739746,
      "memory(GiB)": 77.56,
      "step": 33175,
      "token_acc": 0.4709480122324159,
      "train_speed(iter/s)": 1.440629
    },
    {
      "epoch": 1.4215329248961055,
      "grad_norm": 4.752240180969238,
      "learning_rate": 8.134706412606014e-05,
      "loss": 2.393367385864258,
      "memory(GiB)": 77.56,
      "step": 33180,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.440594
    },
    {
      "epoch": 1.4217471402253545,
      "grad_norm": 6.1346540451049805,
      "learning_rate": 8.134182090132931e-05,
      "loss": 2.874431037902832,
      "memory(GiB)": 77.56,
      "step": 33185,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.440578
    },
    {
      "epoch": 1.4219613555546036,
      "grad_norm": 4.7812652587890625,
      "learning_rate": 8.133657710881197e-05,
      "loss": 2.514013671875,
      "memory(GiB)": 77.56,
      "step": 33190,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.440607
    },
    {
      "epoch": 1.4221755708838524,
      "grad_norm": 5.245537757873535,
      "learning_rate": 8.133133274860316e-05,
      "loss": 2.1866266250610353,
      "memory(GiB)": 77.56,
      "step": 33195,
      "token_acc": 0.524,
      "train_speed(iter/s)": 1.440653
    },
    {
      "epoch": 1.4223897862131014,
      "grad_norm": 5.887220859527588,
      "learning_rate": 8.132608782079785e-05,
      "loss": 2.630213737487793,
      "memory(GiB)": 77.56,
      "step": 33200,
      "token_acc": 0.4420731707317073,
      "train_speed(iter/s)": 1.440677
    },
    {
      "epoch": 1.4226040015423504,
      "grad_norm": 4.989991664886475,
      "learning_rate": 8.132084232549107e-05,
      "loss": 2.6953182220458984,
      "memory(GiB)": 77.56,
      "step": 33205,
      "token_acc": 0.4844961240310077,
      "train_speed(iter/s)": 1.440721
    },
    {
      "epoch": 1.4228182168715993,
      "grad_norm": 4.984920501708984,
      "learning_rate": 8.131559626277784e-05,
      "loss": 2.6882429122924805,
      "memory(GiB)": 77.56,
      "step": 33210,
      "token_acc": 0.46075085324232085,
      "train_speed(iter/s)": 1.440663
    },
    {
      "epoch": 1.4230324322008483,
      "grad_norm": 4.190038681030273,
      "learning_rate": 8.131034963275324e-05,
      "loss": 2.522857093811035,
      "memory(GiB)": 77.56,
      "step": 33215,
      "token_acc": 0.47413793103448276,
      "train_speed(iter/s)": 1.440659
    },
    {
      "epoch": 1.4232466475300973,
      "grad_norm": 5.334013938903809,
      "learning_rate": 8.130510243551227e-05,
      "loss": 2.3469905853271484,
      "memory(GiB)": 77.56,
      "step": 33220,
      "token_acc": 0.4623287671232877,
      "train_speed(iter/s)": 1.44067
    },
    {
      "epoch": 1.4234608628593461,
      "grad_norm": 4.802310466766357,
      "learning_rate": 8.129985467115e-05,
      "loss": 2.711520195007324,
      "memory(GiB)": 77.56,
      "step": 33225,
      "token_acc": 0.4635036496350365,
      "train_speed(iter/s)": 1.440702
    },
    {
      "epoch": 1.4236750781885952,
      "grad_norm": 6.179851531982422,
      "learning_rate": 8.12946063397615e-05,
      "loss": 2.578750419616699,
      "memory(GiB)": 77.56,
      "step": 33230,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.4407
    },
    {
      "epoch": 1.4238892935178442,
      "grad_norm": 4.649274826049805,
      "learning_rate": 8.128935744144186e-05,
      "loss": 2.459977722167969,
      "memory(GiB)": 77.56,
      "step": 33235,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.440694
    },
    {
      "epoch": 1.424103508847093,
      "grad_norm": 7.516111373901367,
      "learning_rate": 8.128410797628615e-05,
      "loss": 2.461636543273926,
      "memory(GiB)": 77.56,
      "step": 33240,
      "token_acc": 0.4908424908424908,
      "train_speed(iter/s)": 1.440705
    },
    {
      "epoch": 1.424317724176342,
      "grad_norm": 3.97896409034729,
      "learning_rate": 8.127885794438947e-05,
      "loss": 2.478633499145508,
      "memory(GiB)": 77.56,
      "step": 33245,
      "token_acc": 0.48881789137380194,
      "train_speed(iter/s)": 1.440735
    },
    {
      "epoch": 1.424531939505591,
      "grad_norm": 4.637205600738525,
      "learning_rate": 8.127360734584695e-05,
      "loss": 2.559110641479492,
      "memory(GiB)": 77.56,
      "step": 33250,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.440746
    },
    {
      "epoch": 1.42474615483484,
      "grad_norm": 3.8903262615203857,
      "learning_rate": 8.126835618075368e-05,
      "loss": 2.4842920303344727,
      "memory(GiB)": 77.56,
      "step": 33255,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.440734
    },
    {
      "epoch": 1.424960370164089,
      "grad_norm": 4.428865432739258,
      "learning_rate": 8.126310444920482e-05,
      "loss": 2.4514419555664064,
      "memory(GiB)": 77.56,
      "step": 33260,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.440746
    },
    {
      "epoch": 1.425174585493338,
      "grad_norm": 5.462167739868164,
      "learning_rate": 8.12578521512955e-05,
      "loss": 2.2549291610717774,
      "memory(GiB)": 77.56,
      "step": 33265,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.440782
    },
    {
      "epoch": 1.4253888008225868,
      "grad_norm": 6.441885471343994,
      "learning_rate": 8.125259928712085e-05,
      "loss": 2.4167686462402345,
      "memory(GiB)": 77.56,
      "step": 33270,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.440776
    },
    {
      "epoch": 1.4256030161518358,
      "grad_norm": 5.584488391876221,
      "learning_rate": 8.124734585677606e-05,
      "loss": 2.8530174255371095,
      "memory(GiB)": 77.56,
      "step": 33275,
      "token_acc": 0.40625,
      "train_speed(iter/s)": 1.440778
    },
    {
      "epoch": 1.4258172314810849,
      "grad_norm": 3.919590473175049,
      "learning_rate": 8.124209186035627e-05,
      "loss": 2.5603870391845702,
      "memory(GiB)": 77.56,
      "step": 33280,
      "token_acc": 0.5082508250825083,
      "train_speed(iter/s)": 1.440757
    },
    {
      "epoch": 1.4260314468103337,
      "grad_norm": 6.487667560577393,
      "learning_rate": 8.12368372979567e-05,
      "loss": 2.7708879470825196,
      "memory(GiB)": 77.56,
      "step": 33285,
      "token_acc": 0.4471830985915493,
      "train_speed(iter/s)": 1.440758
    },
    {
      "epoch": 1.4262456621395827,
      "grad_norm": 4.309908390045166,
      "learning_rate": 8.12315821696725e-05,
      "loss": 2.498444747924805,
      "memory(GiB)": 77.56,
      "step": 33290,
      "token_acc": 0.4337748344370861,
      "train_speed(iter/s)": 1.440776
    },
    {
      "epoch": 1.4264598774688317,
      "grad_norm": 4.262087821960449,
      "learning_rate": 8.122632647559891e-05,
      "loss": 2.708254814147949,
      "memory(GiB)": 77.56,
      "step": 33295,
      "token_acc": 0.45425867507886436,
      "train_speed(iter/s)": 1.4408
    },
    {
      "epoch": 1.4266740927980806,
      "grad_norm": 4.266010284423828,
      "learning_rate": 8.122107021583112e-05,
      "loss": 2.502963066101074,
      "memory(GiB)": 77.56,
      "step": 33300,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.440785
    },
    {
      "epoch": 1.4268883081273296,
      "grad_norm": 3.634632110595703,
      "learning_rate": 8.121581339046433e-05,
      "loss": 2.5300493240356445,
      "memory(GiB)": 77.56,
      "step": 33305,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.440799
    },
    {
      "epoch": 1.4271025234565786,
      "grad_norm": 4.688873291015625,
      "learning_rate": 8.121055599959382e-05,
      "loss": 2.2768619537353514,
      "memory(GiB)": 77.56,
      "step": 33310,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.440785
    },
    {
      "epoch": 1.4273167387858274,
      "grad_norm": 3.5685460567474365,
      "learning_rate": 8.120529804331482e-05,
      "loss": 2.3236207962036133,
      "memory(GiB)": 77.56,
      "step": 33315,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.440739
    },
    {
      "epoch": 1.4275309541150765,
      "grad_norm": 5.878200054168701,
      "learning_rate": 8.120003952172255e-05,
      "loss": 2.494319534301758,
      "memory(GiB)": 77.56,
      "step": 33320,
      "token_acc": 0.4607142857142857,
      "train_speed(iter/s)": 1.440738
    },
    {
      "epoch": 1.4277451694443255,
      "grad_norm": 5.4714508056640625,
      "learning_rate": 8.119478043491232e-05,
      "loss": 2.4719051361083983,
      "memory(GiB)": 77.56,
      "step": 33325,
      "token_acc": 0.5308641975308642,
      "train_speed(iter/s)": 1.440722
    },
    {
      "epoch": 1.4279593847735743,
      "grad_norm": 5.3441243171691895,
      "learning_rate": 8.118952078297936e-05,
      "loss": 2.7524444580078127,
      "memory(GiB)": 77.56,
      "step": 33330,
      "token_acc": 0.4388059701492537,
      "train_speed(iter/s)": 1.440764
    },
    {
      "epoch": 1.4281736001028233,
      "grad_norm": 6.843552589416504,
      "learning_rate": 8.118426056601897e-05,
      "loss": 2.4353818893432617,
      "memory(GiB)": 77.56,
      "step": 33335,
      "token_acc": 0.502092050209205,
      "train_speed(iter/s)": 1.440708
    },
    {
      "epoch": 1.4283878154320724,
      "grad_norm": 4.470619201660156,
      "learning_rate": 8.117899978412646e-05,
      "loss": 2.3507705688476563,
      "memory(GiB)": 77.56,
      "step": 33340,
      "token_acc": 0.48014440433212996,
      "train_speed(iter/s)": 1.440726
    },
    {
      "epoch": 1.4286020307613212,
      "grad_norm": 5.78504753112793,
      "learning_rate": 8.11737384373971e-05,
      "loss": 2.5856754302978517,
      "memory(GiB)": 77.56,
      "step": 33345,
      "token_acc": 0.4708029197080292,
      "train_speed(iter/s)": 1.440715
    },
    {
      "epoch": 1.4288162460905702,
      "grad_norm": 5.120090007781982,
      "learning_rate": 8.116847652592626e-05,
      "loss": 2.3425918579101563,
      "memory(GiB)": 77.56,
      "step": 33350,
      "token_acc": 0.5191489361702127,
      "train_speed(iter/s)": 1.440728
    },
    {
      "epoch": 1.4290304614198193,
      "grad_norm": 4.703099727630615,
      "learning_rate": 8.11632140498092e-05,
      "loss": 2.572285842895508,
      "memory(GiB)": 77.56,
      "step": 33355,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.440767
    },
    {
      "epoch": 1.429244676749068,
      "grad_norm": 5.022765636444092,
      "learning_rate": 8.115795100914129e-05,
      "loss": 2.5505302429199217,
      "memory(GiB)": 77.56,
      "step": 33360,
      "token_acc": 0.46586345381526106,
      "train_speed(iter/s)": 1.440705
    },
    {
      "epoch": 1.4294588920783171,
      "grad_norm": 6.761358261108398,
      "learning_rate": 8.115268740401787e-05,
      "loss": 2.746389389038086,
      "memory(GiB)": 77.56,
      "step": 33365,
      "token_acc": 0.5104602510460251,
      "train_speed(iter/s)": 1.440666
    },
    {
      "epoch": 1.4296731074075661,
      "grad_norm": 4.111597537994385,
      "learning_rate": 8.114742323453431e-05,
      "loss": 2.6209781646728514,
      "memory(GiB)": 77.56,
      "step": 33370,
      "token_acc": 0.45791245791245794,
      "train_speed(iter/s)": 1.440675
    },
    {
      "epoch": 1.429887322736815,
      "grad_norm": 5.42866325378418,
      "learning_rate": 8.114215850078595e-05,
      "loss": 2.26412353515625,
      "memory(GiB)": 77.56,
      "step": 33375,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.440706
    },
    {
      "epoch": 1.430101538066064,
      "grad_norm": 4.153022289276123,
      "learning_rate": 8.113689320286817e-05,
      "loss": 2.988064193725586,
      "memory(GiB)": 77.56,
      "step": 33380,
      "token_acc": 0.446875,
      "train_speed(iter/s)": 1.440737
    },
    {
      "epoch": 1.430315753395313,
      "grad_norm": 4.300074100494385,
      "learning_rate": 8.113162734087636e-05,
      "loss": 2.5837913513183595,
      "memory(GiB)": 77.56,
      "step": 33385,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.440804
    },
    {
      "epoch": 1.4305299687245618,
      "grad_norm": 5.344752788543701,
      "learning_rate": 8.112636091490591e-05,
      "loss": 2.658455657958984,
      "memory(GiB)": 77.56,
      "step": 33390,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.440818
    },
    {
      "epoch": 1.4307441840538109,
      "grad_norm": 6.4714765548706055,
      "learning_rate": 8.112109392505225e-05,
      "loss": 2.6288711547851564,
      "memory(GiB)": 77.56,
      "step": 33395,
      "token_acc": 0.4868913857677903,
      "train_speed(iter/s)": 1.440838
    },
    {
      "epoch": 1.43095839938306,
      "grad_norm": 5.565735816955566,
      "learning_rate": 8.111582637141078e-05,
      "loss": 2.7491548538208006,
      "memory(GiB)": 77.56,
      "step": 33400,
      "token_acc": 0.44921875,
      "train_speed(iter/s)": 1.440848
    },
    {
      "epoch": 1.4311726147123087,
      "grad_norm": 5.314371109008789,
      "learning_rate": 8.11105582540769e-05,
      "loss": 2.509299468994141,
      "memory(GiB)": 77.56,
      "step": 33405,
      "token_acc": 0.5361842105263158,
      "train_speed(iter/s)": 1.440834
    },
    {
      "epoch": 1.4313868300415578,
      "grad_norm": 5.368152618408203,
      "learning_rate": 8.11052895731461e-05,
      "loss": 2.621249961853027,
      "memory(GiB)": 77.56,
      "step": 33410,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.440815
    },
    {
      "epoch": 1.4316010453708068,
      "grad_norm": 5.408423900604248,
      "learning_rate": 8.11000203287138e-05,
      "loss": 2.6001571655273437,
      "memory(GiB)": 77.56,
      "step": 33415,
      "token_acc": 0.4479495268138801,
      "train_speed(iter/s)": 1.440862
    },
    {
      "epoch": 1.4318152607000556,
      "grad_norm": 4.377597332000732,
      "learning_rate": 8.109475052087543e-05,
      "loss": 2.71426944732666,
      "memory(GiB)": 77.56,
      "step": 33420,
      "token_acc": 0.4377358490566038,
      "train_speed(iter/s)": 1.440901
    },
    {
      "epoch": 1.4320294760293046,
      "grad_norm": 4.818597316741943,
      "learning_rate": 8.108948014972652e-05,
      "loss": 2.6134973526000977,
      "memory(GiB)": 77.56,
      "step": 33425,
      "token_acc": 0.4291044776119403,
      "train_speed(iter/s)": 1.440959
    },
    {
      "epoch": 1.4322436913585537,
      "grad_norm": 3.9246184825897217,
      "learning_rate": 8.108420921536248e-05,
      "loss": 2.654636573791504,
      "memory(GiB)": 77.56,
      "step": 33430,
      "token_acc": 0.47896440129449835,
      "train_speed(iter/s)": 1.440968
    },
    {
      "epoch": 1.4324579066878025,
      "grad_norm": 4.747319221496582,
      "learning_rate": 8.107893771787885e-05,
      "loss": 2.6029773712158204,
      "memory(GiB)": 77.56,
      "step": 33435,
      "token_acc": 0.43234323432343236,
      "train_speed(iter/s)": 1.440959
    },
    {
      "epoch": 1.4326721220170515,
      "grad_norm": 5.151585578918457,
      "learning_rate": 8.107366565737112e-05,
      "loss": 2.3758398056030274,
      "memory(GiB)": 77.56,
      "step": 33440,
      "token_acc": 0.5251798561151079,
      "train_speed(iter/s)": 1.440962
    },
    {
      "epoch": 1.4328863373463006,
      "grad_norm": 5.947368144989014,
      "learning_rate": 8.106839303393476e-05,
      "loss": 2.5448598861694336,
      "memory(GiB)": 77.56,
      "step": 33445,
      "token_acc": 0.4495114006514658,
      "train_speed(iter/s)": 1.440992
    },
    {
      "epoch": 1.4331005526755494,
      "grad_norm": 5.890367031097412,
      "learning_rate": 8.106311984766535e-05,
      "loss": 2.8148525238037108,
      "memory(GiB)": 77.56,
      "step": 33450,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.441044
    },
    {
      "epoch": 1.4333147680047984,
      "grad_norm": 5.796880722045898,
      "learning_rate": 8.105784609865835e-05,
      "loss": 2.650727081298828,
      "memory(GiB)": 77.56,
      "step": 33455,
      "token_acc": 0.4901185770750988,
      "train_speed(iter/s)": 1.441091
    },
    {
      "epoch": 1.4335289833340474,
      "grad_norm": 4.480291843414307,
      "learning_rate": 8.105257178700935e-05,
      "loss": 2.385071563720703,
      "memory(GiB)": 77.56,
      "step": 33460,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.44112
    },
    {
      "epoch": 1.4337431986632962,
      "grad_norm": 4.908984661102295,
      "learning_rate": 8.104729691281387e-05,
      "loss": 3.141388702392578,
      "memory(GiB)": 77.56,
      "step": 33465,
      "token_acc": 0.38509316770186336,
      "train_speed(iter/s)": 1.441105
    },
    {
      "epoch": 1.4339574139925453,
      "grad_norm": 6.158671855926514,
      "learning_rate": 8.10420214761675e-05,
      "loss": 2.772718048095703,
      "memory(GiB)": 77.56,
      "step": 33470,
      "token_acc": 0.4222873900293255,
      "train_speed(iter/s)": 1.44111
    },
    {
      "epoch": 1.4341716293217943,
      "grad_norm": 4.42659330368042,
      "learning_rate": 8.103674547716577e-05,
      "loss": 2.409808349609375,
      "memory(GiB)": 77.56,
      "step": 33475,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.441106
    },
    {
      "epoch": 1.4343858446510431,
      "grad_norm": 8.199884414672852,
      "learning_rate": 8.10314689159043e-05,
      "loss": 2.42923641204834,
      "memory(GiB)": 77.56,
      "step": 33480,
      "token_acc": 0.5139442231075697,
      "train_speed(iter/s)": 1.441059
    },
    {
      "epoch": 1.4346000599802922,
      "grad_norm": 5.221634387969971,
      "learning_rate": 8.102619179247866e-05,
      "loss": 2.596410369873047,
      "memory(GiB)": 77.56,
      "step": 33485,
      "token_acc": 0.45808383233532934,
      "train_speed(iter/s)": 1.441096
    },
    {
      "epoch": 1.4348142753095412,
      "grad_norm": 5.489154815673828,
      "learning_rate": 8.102091410698445e-05,
      "loss": 2.6875179290771483,
      "memory(GiB)": 77.56,
      "step": 33490,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.441123
    },
    {
      "epoch": 1.43502849063879,
      "grad_norm": 3.84725284576416,
      "learning_rate": 8.101563585951728e-05,
      "loss": 2.7381092071533204,
      "memory(GiB)": 77.56,
      "step": 33495,
      "token_acc": 0.4570552147239264,
      "train_speed(iter/s)": 1.441149
    },
    {
      "epoch": 1.435242705968039,
      "grad_norm": 4.275827407836914,
      "learning_rate": 8.101035705017277e-05,
      "loss": 2.7338294982910156,
      "memory(GiB)": 77.56,
      "step": 33500,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.441155
    },
    {
      "epoch": 1.435242705968039,
      "eval_loss": 2.3013715744018555,
      "eval_runtime": 14.4887,
      "eval_samples_per_second": 6.902,
      "eval_steps_per_second": 6.902,
      "eval_token_acc": 0.4812760055478502,
      "step": 33500
    },
    {
      "epoch": 1.435456921297288,
      "grad_norm": 4.167738914489746,
      "learning_rate": 8.100507767904653e-05,
      "loss": 2.466640853881836,
      "memory(GiB)": 77.56,
      "step": 33505,
      "token_acc": 0.4852240228789323,
      "train_speed(iter/s)": 1.440202
    },
    {
      "epoch": 1.435671136626537,
      "grad_norm": 5.381674766540527,
      "learning_rate": 8.099979774623425e-05,
      "loss": 2.276072311401367,
      "memory(GiB)": 77.56,
      "step": 33510,
      "token_acc": 0.5246636771300448,
      "train_speed(iter/s)": 1.440171
    },
    {
      "epoch": 1.435885351955786,
      "grad_norm": 6.185641765594482,
      "learning_rate": 8.099451725183154e-05,
      "loss": 2.246568298339844,
      "memory(GiB)": 77.56,
      "step": 33515,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.440168
    },
    {
      "epoch": 1.436099567285035,
      "grad_norm": 5.230942249298096,
      "learning_rate": 8.098923619593409e-05,
      "loss": 2.6886354446411134,
      "memory(GiB)": 77.56,
      "step": 33520,
      "token_acc": 0.5145631067961165,
      "train_speed(iter/s)": 1.440159
    },
    {
      "epoch": 1.4363137826142838,
      "grad_norm": 4.901467323303223,
      "learning_rate": 8.098395457863755e-05,
      "loss": 2.153408241271973,
      "memory(GiB)": 77.56,
      "step": 33525,
      "token_acc": 0.5289256198347108,
      "train_speed(iter/s)": 1.440173
    },
    {
      "epoch": 1.4365279979435328,
      "grad_norm": 4.4969682693481445,
      "learning_rate": 8.097867240003761e-05,
      "loss": 2.5498191833496096,
      "memory(GiB)": 77.56,
      "step": 33530,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.440214
    },
    {
      "epoch": 1.4367422132727818,
      "grad_norm": 4.1007561683654785,
      "learning_rate": 8.097338966022993e-05,
      "loss": 2.7983366012573243,
      "memory(GiB)": 77.56,
      "step": 33535,
      "token_acc": 0.4366576819407008,
      "train_speed(iter/s)": 1.44024
    },
    {
      "epoch": 1.4369564286020307,
      "grad_norm": 6.27601957321167,
      "learning_rate": 8.096810635931026e-05,
      "loss": 2.673727035522461,
      "memory(GiB)": 77.56,
      "step": 33540,
      "token_acc": 0.4358974358974359,
      "train_speed(iter/s)": 1.440259
    },
    {
      "epoch": 1.4371706439312797,
      "grad_norm": 4.566864013671875,
      "learning_rate": 8.09628224973743e-05,
      "loss": 2.391421318054199,
      "memory(GiB)": 77.56,
      "step": 33545,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.440217
    },
    {
      "epoch": 1.4373848592605287,
      "grad_norm": 4.772444248199463,
      "learning_rate": 8.095753807451777e-05,
      "loss": 2.640765380859375,
      "memory(GiB)": 77.56,
      "step": 33550,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.440207
    },
    {
      "epoch": 1.4375990745897775,
      "grad_norm": 4.421515464782715,
      "learning_rate": 8.095225309083638e-05,
      "loss": 2.428546333312988,
      "memory(GiB)": 77.56,
      "step": 33555,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.440212
    },
    {
      "epoch": 1.4378132899190266,
      "grad_norm": 4.157040596008301,
      "learning_rate": 8.094696754642591e-05,
      "loss": 2.5101404190063477,
      "memory(GiB)": 77.56,
      "step": 33560,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.44025
    },
    {
      "epoch": 1.4380275052482756,
      "grad_norm": 5.607004642486572,
      "learning_rate": 8.094168144138208e-05,
      "loss": 2.6087614059448243,
      "memory(GiB)": 77.56,
      "step": 33565,
      "token_acc": 0.5097276264591439,
      "train_speed(iter/s)": 1.440266
    },
    {
      "epoch": 1.4382417205775244,
      "grad_norm": 6.247413635253906,
      "learning_rate": 8.093639477580066e-05,
      "loss": 2.7758199691772463,
      "memory(GiB)": 77.56,
      "step": 33570,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.440255
    },
    {
      "epoch": 1.4384559359067735,
      "grad_norm": 4.286452293395996,
      "learning_rate": 8.093110754977744e-05,
      "loss": 2.545228958129883,
      "memory(GiB)": 77.56,
      "step": 33575,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.440269
    },
    {
      "epoch": 1.4386701512360225,
      "grad_norm": 4.8186564445495605,
      "learning_rate": 8.092581976340819e-05,
      "loss": 2.361837387084961,
      "memory(GiB)": 77.56,
      "step": 33580,
      "token_acc": 0.5110294117647058,
      "train_speed(iter/s)": 1.440284
    },
    {
      "epoch": 1.4388843665652713,
      "grad_norm": 4.426604747772217,
      "learning_rate": 8.09205314167887e-05,
      "loss": 2.558700752258301,
      "memory(GiB)": 77.56,
      "step": 33585,
      "token_acc": 0.45787545787545786,
      "train_speed(iter/s)": 1.440293
    },
    {
      "epoch": 1.4390985818945203,
      "grad_norm": 6.376096248626709,
      "learning_rate": 8.091524251001476e-05,
      "loss": 2.339290428161621,
      "memory(GiB)": 77.56,
      "step": 33590,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.440313
    },
    {
      "epoch": 1.4393127972237694,
      "grad_norm": 10.38262939453125,
      "learning_rate": 8.090995304318224e-05,
      "loss": 3.107723426818848,
      "memory(GiB)": 77.56,
      "step": 33595,
      "token_acc": 0.4064748201438849,
      "train_speed(iter/s)": 1.440254
    },
    {
      "epoch": 1.4395270125530182,
      "grad_norm": 5.614473342895508,
      "learning_rate": 8.090466301638688e-05,
      "loss": 2.6898096084594725,
      "memory(GiB)": 77.56,
      "step": 33600,
      "token_acc": 0.4127906976744186,
      "train_speed(iter/s)": 1.440248
    },
    {
      "epoch": 1.4397412278822672,
      "grad_norm": 6.232783794403076,
      "learning_rate": 8.089937242972459e-05,
      "loss": 2.3672855377197264,
      "memory(GiB)": 77.56,
      "step": 33605,
      "token_acc": 0.5203252032520326,
      "train_speed(iter/s)": 1.440292
    },
    {
      "epoch": 1.4399554432115163,
      "grad_norm": 6.109330177307129,
      "learning_rate": 8.089408128329118e-05,
      "loss": 2.4628347396850585,
      "memory(GiB)": 77.56,
      "step": 33610,
      "token_acc": 0.4793650793650794,
      "train_speed(iter/s)": 1.440275
    },
    {
      "epoch": 1.440169658540765,
      "grad_norm": 3.8369626998901367,
      "learning_rate": 8.088878957718249e-05,
      "loss": 2.786264419555664,
      "memory(GiB)": 77.56,
      "step": 33615,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.440301
    },
    {
      "epoch": 1.440383873870014,
      "grad_norm": 5.966667175292969,
      "learning_rate": 8.088349731149441e-05,
      "loss": 2.650204658508301,
      "memory(GiB)": 77.56,
      "step": 33620,
      "token_acc": 0.44015444015444016,
      "train_speed(iter/s)": 1.440333
    },
    {
      "epoch": 1.4405980891992631,
      "grad_norm": 4.497265815734863,
      "learning_rate": 8.087820448632282e-05,
      "loss": 2.65279483795166,
      "memory(GiB)": 77.56,
      "step": 33625,
      "token_acc": 0.4804804804804805,
      "train_speed(iter/s)": 1.440357
    },
    {
      "epoch": 1.440812304528512,
      "grad_norm": 4.815916538238525,
      "learning_rate": 8.087291110176355e-05,
      "loss": 2.3260820388793944,
      "memory(GiB)": 77.56,
      "step": 33630,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.440377
    },
    {
      "epoch": 1.441026519857761,
      "grad_norm": 4.615348815917969,
      "learning_rate": 8.086761715791258e-05,
      "loss": 2.050550079345703,
      "memory(GiB)": 77.56,
      "step": 33635,
      "token_acc": 0.5330578512396694,
      "train_speed(iter/s)": 1.440399
    },
    {
      "epoch": 1.44124073518701,
      "grad_norm": 6.801609516143799,
      "learning_rate": 8.086232265486576e-05,
      "loss": 2.616620635986328,
      "memory(GiB)": 77.56,
      "step": 33640,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.440388
    },
    {
      "epoch": 1.4414549505162588,
      "grad_norm": 5.370550155639648,
      "learning_rate": 8.085702759271899e-05,
      "loss": 2.580626678466797,
      "memory(GiB)": 77.56,
      "step": 33645,
      "token_acc": 0.4515235457063712,
      "train_speed(iter/s)": 1.440395
    },
    {
      "epoch": 1.4416691658455079,
      "grad_norm": 4.841226577758789,
      "learning_rate": 8.085173197156824e-05,
      "loss": 2.6208080291748046,
      "memory(GiB)": 77.56,
      "step": 33650,
      "token_acc": 0.42474916387959866,
      "train_speed(iter/s)": 1.440408
    },
    {
      "epoch": 1.441883381174757,
      "grad_norm": 4.60963773727417,
      "learning_rate": 8.084643579150941e-05,
      "loss": 2.538687896728516,
      "memory(GiB)": 77.56,
      "step": 33655,
      "token_acc": 0.4828897338403042,
      "train_speed(iter/s)": 1.440344
    },
    {
      "epoch": 1.4420975965040057,
      "grad_norm": 5.9481306076049805,
      "learning_rate": 8.084113905263846e-05,
      "loss": 2.4066253662109376,
      "memory(GiB)": 77.56,
      "step": 33660,
      "token_acc": 0.5037878787878788,
      "train_speed(iter/s)": 1.440364
    },
    {
      "epoch": 1.4423118118332547,
      "grad_norm": 6.224748134613037,
      "learning_rate": 8.083584175505134e-05,
      "loss": 2.3255426406860353,
      "memory(GiB)": 77.56,
      "step": 33665,
      "token_acc": 0.49586776859504134,
      "train_speed(iter/s)": 1.440435
    },
    {
      "epoch": 1.4425260271625038,
      "grad_norm": 5.452847003936768,
      "learning_rate": 8.083054389884405e-05,
      "loss": 2.5483604431152345,
      "memory(GiB)": 77.56,
      "step": 33670,
      "token_acc": 0.472,
      "train_speed(iter/s)": 1.4404
    },
    {
      "epoch": 1.4427402424917526,
      "grad_norm": 4.419780731201172,
      "learning_rate": 8.082524548411252e-05,
      "loss": 2.5582042694091798,
      "memory(GiB)": 77.56,
      "step": 33675,
      "token_acc": 0.46546546546546547,
      "train_speed(iter/s)": 1.440401
    },
    {
      "epoch": 1.4429544578210016,
      "grad_norm": 6.002676486968994,
      "learning_rate": 8.081994651095273e-05,
      "loss": 2.4643436431884767,
      "memory(GiB)": 77.56,
      "step": 33680,
      "token_acc": 0.46686746987951805,
      "train_speed(iter/s)": 1.44037
    },
    {
      "epoch": 1.4431686731502507,
      "grad_norm": 4.702930450439453,
      "learning_rate": 8.081464697946072e-05,
      "loss": 2.578280448913574,
      "memory(GiB)": 77.56,
      "step": 33685,
      "token_acc": 0.46742209631728043,
      "train_speed(iter/s)": 1.440331
    },
    {
      "epoch": 1.4433828884794995,
      "grad_norm": 5.671988487243652,
      "learning_rate": 8.080934688973248e-05,
      "loss": 2.604653549194336,
      "memory(GiB)": 77.56,
      "step": 33690,
      "token_acc": 0.4588607594936709,
      "train_speed(iter/s)": 1.440353
    },
    {
      "epoch": 1.4435971038087485,
      "grad_norm": 4.508692741394043,
      "learning_rate": 8.0804046241864e-05,
      "loss": 2.448164939880371,
      "memory(GiB)": 77.56,
      "step": 33695,
      "token_acc": 0.4881889763779528,
      "train_speed(iter/s)": 1.440331
    },
    {
      "epoch": 1.4438113191379975,
      "grad_norm": 5.394710063934326,
      "learning_rate": 8.079874503595133e-05,
      "loss": 2.5911964416503905,
      "memory(GiB)": 77.56,
      "step": 33700,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.440293
    },
    {
      "epoch": 1.4440255344672464,
      "grad_norm": 5.142947673797607,
      "learning_rate": 8.079344327209051e-05,
      "loss": 2.55928955078125,
      "memory(GiB)": 77.56,
      "step": 33705,
      "token_acc": 0.4515235457063712,
      "train_speed(iter/s)": 1.440301
    },
    {
      "epoch": 1.4442397497964954,
      "grad_norm": 4.448504447937012,
      "learning_rate": 8.078814095037758e-05,
      "loss": 2.5307140350341797,
      "memory(GiB)": 77.56,
      "step": 33710,
      "token_acc": 0.4859154929577465,
      "train_speed(iter/s)": 1.44031
    },
    {
      "epoch": 1.4444539651257444,
      "grad_norm": 5.810717582702637,
      "learning_rate": 8.078283807090858e-05,
      "loss": 2.7283184051513674,
      "memory(GiB)": 77.56,
      "step": 33715,
      "token_acc": 0.4457831325301205,
      "train_speed(iter/s)": 1.440318
    },
    {
      "epoch": 1.4446681804549932,
      "grad_norm": 4.456352710723877,
      "learning_rate": 8.077753463377962e-05,
      "loss": 2.76055965423584,
      "memory(GiB)": 77.56,
      "step": 33720,
      "token_acc": 0.444794952681388,
      "train_speed(iter/s)": 1.4404
    },
    {
      "epoch": 1.4448823957842423,
      "grad_norm": 5.7959184646606445,
      "learning_rate": 8.07722306390867e-05,
      "loss": 2.332366371154785,
      "memory(GiB)": 77.56,
      "step": 33725,
      "token_acc": 0.5054945054945055,
      "train_speed(iter/s)": 1.440414
    },
    {
      "epoch": 1.4450966111134913,
      "grad_norm": 6.644979476928711,
      "learning_rate": 8.0766926086926e-05,
      "loss": 2.443693161010742,
      "memory(GiB)": 77.56,
      "step": 33730,
      "token_acc": 0.4854014598540146,
      "train_speed(iter/s)": 1.440367
    },
    {
      "epoch": 1.4453108264427401,
      "grad_norm": 4.913311958312988,
      "learning_rate": 8.076162097739356e-05,
      "loss": 2.470724105834961,
      "memory(GiB)": 77.56,
      "step": 33735,
      "token_acc": 0.4783950617283951,
      "train_speed(iter/s)": 1.440351
    },
    {
      "epoch": 1.4455250417719891,
      "grad_norm": 5.47266149520874,
      "learning_rate": 8.07563153105855e-05,
      "loss": 2.40041561126709,
      "memory(GiB)": 77.56,
      "step": 33740,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.44034
    },
    {
      "epoch": 1.4457392571012382,
      "grad_norm": 5.7080078125,
      "learning_rate": 8.075100908659793e-05,
      "loss": 3.122731399536133,
      "memory(GiB)": 77.56,
      "step": 33745,
      "token_acc": 0.4129032258064516,
      "train_speed(iter/s)": 1.440365
    },
    {
      "epoch": 1.445953472430487,
      "grad_norm": 4.789261817932129,
      "learning_rate": 8.074570230552698e-05,
      "loss": 2.2112518310546876,
      "memory(GiB)": 77.56,
      "step": 33750,
      "token_acc": 0.5272727272727272,
      "train_speed(iter/s)": 1.44038
    },
    {
      "epoch": 1.446167687759736,
      "grad_norm": 5.958988666534424,
      "learning_rate": 8.074039496746881e-05,
      "loss": 2.8660274505615235,
      "memory(GiB)": 77.56,
      "step": 33755,
      "token_acc": 0.4250871080139373,
      "train_speed(iter/s)": 1.440428
    },
    {
      "epoch": 1.446381903088985,
      "grad_norm": 4.641246795654297,
      "learning_rate": 8.073508707251953e-05,
      "loss": 2.675687789916992,
      "memory(GiB)": 77.56,
      "step": 33760,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.440476
    },
    {
      "epoch": 1.4465961184182339,
      "grad_norm": 4.817697525024414,
      "learning_rate": 8.072977862077532e-05,
      "loss": 2.4350305557250977,
      "memory(GiB)": 77.56,
      "step": 33765,
      "token_acc": 0.47878787878787876,
      "train_speed(iter/s)": 1.440417
    },
    {
      "epoch": 1.446810333747483,
      "grad_norm": 4.625699996948242,
      "learning_rate": 8.072446961233236e-05,
      "loss": 2.404711151123047,
      "memory(GiB)": 77.56,
      "step": 33770,
      "token_acc": 0.4602510460251046,
      "train_speed(iter/s)": 1.440367
    },
    {
      "epoch": 1.447024549076732,
      "grad_norm": 4.81039571762085,
      "learning_rate": 8.07191600472868e-05,
      "loss": 2.301344108581543,
      "memory(GiB)": 77.56,
      "step": 33775,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.440397
    },
    {
      "epoch": 1.4472387644059808,
      "grad_norm": 4.775697231292725,
      "learning_rate": 8.071384992573482e-05,
      "loss": 2.5201833724975584,
      "memory(GiB)": 77.56,
      "step": 33780,
      "token_acc": 0.47041420118343197,
      "train_speed(iter/s)": 1.440404
    },
    {
      "epoch": 1.4474529797352298,
      "grad_norm": 6.714394569396973,
      "learning_rate": 8.070853924777266e-05,
      "loss": 2.4296825408935545,
      "memory(GiB)": 77.56,
      "step": 33785,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.440403
    },
    {
      "epoch": 1.4476671950644788,
      "grad_norm": 4.200636386871338,
      "learning_rate": 8.070322801349649e-05,
      "loss": 2.6939130783081056,
      "memory(GiB)": 77.56,
      "step": 33790,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.440429
    },
    {
      "epoch": 1.4478814103937276,
      "grad_norm": 6.712575435638428,
      "learning_rate": 8.069791622300255e-05,
      "loss": 2.3527122497558595,
      "memory(GiB)": 77.56,
      "step": 33795,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.440389
    },
    {
      "epoch": 1.4480956257229767,
      "grad_norm": 8.108104705810547,
      "learning_rate": 8.069260387638705e-05,
      "loss": 2.6927162170410157,
      "memory(GiB)": 77.56,
      "step": 33800,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.440373
    },
    {
      "epoch": 1.4483098410522257,
      "grad_norm": 4.842827796936035,
      "learning_rate": 8.068729097374626e-05,
      "loss": 2.4400306701660157,
      "memory(GiB)": 77.56,
      "step": 33805,
      "token_acc": 0.470404984423676,
      "train_speed(iter/s)": 1.440392
    },
    {
      "epoch": 1.4485240563814745,
      "grad_norm": 4.508891582489014,
      "learning_rate": 8.068197751517638e-05,
      "loss": 2.4819536209106445,
      "memory(GiB)": 77.56,
      "step": 33810,
      "token_acc": 0.432,
      "train_speed(iter/s)": 1.440415
    },
    {
      "epoch": 1.4487382717107236,
      "grad_norm": 5.302507400512695,
      "learning_rate": 8.067666350077372e-05,
      "loss": 2.615432548522949,
      "memory(GiB)": 77.56,
      "step": 33815,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.440389
    },
    {
      "epoch": 1.4489524870399726,
      "grad_norm": 4.059510231018066,
      "learning_rate": 8.067134893063452e-05,
      "loss": 2.618773651123047,
      "memory(GiB)": 77.56,
      "step": 33820,
      "token_acc": 0.44591029023746703,
      "train_speed(iter/s)": 1.440401
    },
    {
      "epoch": 1.4491667023692214,
      "grad_norm": 4.536221027374268,
      "learning_rate": 8.066603380485505e-05,
      "loss": 2.615183639526367,
      "memory(GiB)": 77.56,
      "step": 33825,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.440448
    },
    {
      "epoch": 1.4493809176984704,
      "grad_norm": 4.616865158081055,
      "learning_rate": 8.066071812353162e-05,
      "loss": 2.4374961853027344,
      "memory(GiB)": 77.56,
      "step": 33830,
      "token_acc": 0.47470817120622566,
      "train_speed(iter/s)": 1.440433
    },
    {
      "epoch": 1.4495951330277195,
      "grad_norm": 4.955784320831299,
      "learning_rate": 8.06554018867605e-05,
      "loss": 2.2806678771972657,
      "memory(GiB)": 77.56,
      "step": 33835,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.44038
    },
    {
      "epoch": 1.4498093483569683,
      "grad_norm": 4.6067609786987305,
      "learning_rate": 8.065008509463805e-05,
      "loss": 2.401715850830078,
      "memory(GiB)": 77.56,
      "step": 33840,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.440372
    },
    {
      "epoch": 1.4500235636862173,
      "grad_norm": 5.793132781982422,
      "learning_rate": 8.064476774726054e-05,
      "loss": 2.485592269897461,
      "memory(GiB)": 77.56,
      "step": 33845,
      "token_acc": 0.4748427672955975,
      "train_speed(iter/s)": 1.440347
    },
    {
      "epoch": 1.4502377790154664,
      "grad_norm": 4.4255523681640625,
      "learning_rate": 8.063944984472431e-05,
      "loss": 2.5242336273193358,
      "memory(GiB)": 77.56,
      "step": 33850,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.440307
    },
    {
      "epoch": 1.4504519943447154,
      "grad_norm": 4.728428840637207,
      "learning_rate": 8.063413138712572e-05,
      "loss": 2.763438415527344,
      "memory(GiB)": 77.56,
      "step": 33855,
      "token_acc": 0.4637223974763407,
      "train_speed(iter/s)": 1.440372
    },
    {
      "epoch": 1.4506662096739642,
      "grad_norm": 4.184111595153809,
      "learning_rate": 8.062881237456108e-05,
      "loss": 2.104880714416504,
      "memory(GiB)": 77.56,
      "step": 33860,
      "token_acc": 0.49387755102040815,
      "train_speed(iter/s)": 1.440404
    },
    {
      "epoch": 1.4508804250032132,
      "grad_norm": 4.85262393951416,
      "learning_rate": 8.06234928071268e-05,
      "loss": 2.3321254730224608,
      "memory(GiB)": 77.56,
      "step": 33865,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.440425
    },
    {
      "epoch": 1.4510946403324623,
      "grad_norm": 5.855126857757568,
      "learning_rate": 8.061817268491919e-05,
      "loss": 2.693760108947754,
      "memory(GiB)": 77.56,
      "step": 33870,
      "token_acc": 0.4612903225806452,
      "train_speed(iter/s)": 1.440413
    },
    {
      "epoch": 1.451308855661711,
      "grad_norm": 4.04247522354126,
      "learning_rate": 8.061285200803467e-05,
      "loss": 2.3854419708251955,
      "memory(GiB)": 77.56,
      "step": 33875,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440419
    },
    {
      "epoch": 1.4515230709909601,
      "grad_norm": 5.262057304382324,
      "learning_rate": 8.060753077656964e-05,
      "loss": 2.469320297241211,
      "memory(GiB)": 77.56,
      "step": 33880,
      "token_acc": 0.4230769230769231,
      "train_speed(iter/s)": 1.440386
    },
    {
      "epoch": 1.4517372863202092,
      "grad_norm": 5.64385986328125,
      "learning_rate": 8.060220899062045e-05,
      "loss": 2.6018207550048826,
      "memory(GiB)": 77.56,
      "step": 33885,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.440383
    },
    {
      "epoch": 1.451951501649458,
      "grad_norm": 5.172167778015137,
      "learning_rate": 8.059688665028355e-05,
      "loss": 2.4049049377441407,
      "memory(GiB)": 77.56,
      "step": 33890,
      "token_acc": 0.49174917491749176,
      "train_speed(iter/s)": 1.44039
    },
    {
      "epoch": 1.452165716978707,
      "grad_norm": 4.447646141052246,
      "learning_rate": 8.059156375565535e-05,
      "loss": 2.5212480545043947,
      "memory(GiB)": 77.56,
      "step": 33895,
      "token_acc": 0.4548736462093863,
      "train_speed(iter/s)": 1.440356
    },
    {
      "epoch": 1.452379932307956,
      "grad_norm": 4.732133388519287,
      "learning_rate": 8.058624030683226e-05,
      "loss": 2.435433578491211,
      "memory(GiB)": 77.56,
      "step": 33900,
      "token_acc": 0.49707602339181284,
      "train_speed(iter/s)": 1.440351
    },
    {
      "epoch": 1.4525941476372048,
      "grad_norm": 5.007026195526123,
      "learning_rate": 8.058091630391076e-05,
      "loss": 2.590679931640625,
      "memory(GiB)": 77.56,
      "step": 33905,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.44036
    },
    {
      "epoch": 1.4528083629664539,
      "grad_norm": 4.960092067718506,
      "learning_rate": 8.057559174698725e-05,
      "loss": 2.620238494873047,
      "memory(GiB)": 77.56,
      "step": 33910,
      "token_acc": 0.4267241379310345,
      "train_speed(iter/s)": 1.440333
    },
    {
      "epoch": 1.453022578295703,
      "grad_norm": 5.670502185821533,
      "learning_rate": 8.05702666361582e-05,
      "loss": 2.4999961853027344,
      "memory(GiB)": 77.56,
      "step": 33915,
      "token_acc": 0.445578231292517,
      "train_speed(iter/s)": 1.440369
    },
    {
      "epoch": 1.4532367936249517,
      "grad_norm": 5.052982330322266,
      "learning_rate": 8.056494097152013e-05,
      "loss": 2.5328283309936523,
      "memory(GiB)": 77.56,
      "step": 33920,
      "token_acc": 0.4299363057324841,
      "train_speed(iter/s)": 1.440408
    },
    {
      "epoch": 1.4534510089542008,
      "grad_norm": 4.490011692047119,
      "learning_rate": 8.055961475316947e-05,
      "loss": 2.695473289489746,
      "memory(GiB)": 77.56,
      "step": 33925,
      "token_acc": 0.4858156028368794,
      "train_speed(iter/s)": 1.440443
    },
    {
      "epoch": 1.4536652242834498,
      "grad_norm": 4.542452335357666,
      "learning_rate": 8.055428798120272e-05,
      "loss": 2.7107307434082033,
      "memory(GiB)": 77.56,
      "step": 33930,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.440468
    },
    {
      "epoch": 1.4538794396126986,
      "grad_norm": 4.045337200164795,
      "learning_rate": 8.054896065571638e-05,
      "loss": 2.320671272277832,
      "memory(GiB)": 77.56,
      "step": 33935,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.44046
    },
    {
      "epoch": 1.4540936549419476,
      "grad_norm": 4.828177452087402,
      "learning_rate": 8.054363277680695e-05,
      "loss": 2.5693618774414064,
      "memory(GiB)": 77.56,
      "step": 33940,
      "token_acc": 0.4768211920529801,
      "train_speed(iter/s)": 1.440465
    },
    {
      "epoch": 1.4543078702711967,
      "grad_norm": 4.6465301513671875,
      "learning_rate": 8.053830434457097e-05,
      "loss": 2.666273498535156,
      "memory(GiB)": 77.56,
      "step": 33945,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.440514
    },
    {
      "epoch": 1.4545220856004455,
      "grad_norm": 5.002664089202881,
      "learning_rate": 8.053297535910496e-05,
      "loss": 3.001188850402832,
      "memory(GiB)": 77.56,
      "step": 33950,
      "token_acc": 0.43034055727554177,
      "train_speed(iter/s)": 1.440543
    },
    {
      "epoch": 1.4547363009296945,
      "grad_norm": 3.7473056316375732,
      "learning_rate": 8.052764582050544e-05,
      "loss": 2.404936408996582,
      "memory(GiB)": 77.56,
      "step": 33955,
      "token_acc": 0.49415204678362573,
      "train_speed(iter/s)": 1.440532
    },
    {
      "epoch": 1.4549505162589436,
      "grad_norm": 5.8957366943359375,
      "learning_rate": 8.052231572886902e-05,
      "loss": 2.5693374633789063,
      "memory(GiB)": 77.56,
      "step": 33960,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.440585
    },
    {
      "epoch": 1.4551647315881924,
      "grad_norm": 4.121309280395508,
      "learning_rate": 8.051698508429219e-05,
      "loss": 2.52559871673584,
      "memory(GiB)": 77.56,
      "step": 33965,
      "token_acc": 0.504950495049505,
      "train_speed(iter/s)": 1.440589
    },
    {
      "epoch": 1.4553789469174414,
      "grad_norm": 4.470658302307129,
      "learning_rate": 8.051165388687154e-05,
      "loss": 2.7387414932250977,
      "memory(GiB)": 77.56,
      "step": 33970,
      "token_acc": 0.4523076923076923,
      "train_speed(iter/s)": 1.440595
    },
    {
      "epoch": 1.4555931622466904,
      "grad_norm": 5.512742042541504,
      "learning_rate": 8.050632213670368e-05,
      "loss": 2.5898693084716795,
      "memory(GiB)": 77.56,
      "step": 33975,
      "token_acc": 0.46332046332046334,
      "train_speed(iter/s)": 1.440625
    },
    {
      "epoch": 1.4558073775759393,
      "grad_norm": 7.265410900115967,
      "learning_rate": 8.050098983388516e-05,
      "loss": 2.705134391784668,
      "memory(GiB)": 77.56,
      "step": 33980,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.440646
    },
    {
      "epoch": 1.4560215929051883,
      "grad_norm": 5.042665958404541,
      "learning_rate": 8.04956569785126e-05,
      "loss": 2.427484321594238,
      "memory(GiB)": 77.56,
      "step": 33985,
      "token_acc": 0.5184049079754601,
      "train_speed(iter/s)": 1.440612
    },
    {
      "epoch": 1.4562358082344373,
      "grad_norm": 5.753913402557373,
      "learning_rate": 8.049032357068263e-05,
      "loss": 2.8000616073608398,
      "memory(GiB)": 77.56,
      "step": 33990,
      "token_acc": 0.44525547445255476,
      "train_speed(iter/s)": 1.440589
    },
    {
      "epoch": 1.4564500235636861,
      "grad_norm": 4.5002827644348145,
      "learning_rate": 8.048498961049182e-05,
      "loss": 2.869167137145996,
      "memory(GiB)": 77.56,
      "step": 33995,
      "token_acc": 0.4269005847953216,
      "train_speed(iter/s)": 1.440571
    },
    {
      "epoch": 1.4566642388929352,
      "grad_norm": 4.313014507293701,
      "learning_rate": 8.047965509803684e-05,
      "loss": 2.503142547607422,
      "memory(GiB)": 77.56,
      "step": 34000,
      "token_acc": 0.4725274725274725,
      "train_speed(iter/s)": 1.440535
    },
    {
      "epoch": 1.4566642388929352,
      "eval_loss": 2.207979440689087,
      "eval_runtime": 13.8409,
      "eval_samples_per_second": 7.225,
      "eval_steps_per_second": 7.225,
      "eval_token_acc": 0.46065808297567956,
      "step": 34000
    },
    {
      "epoch": 1.4568784542221842,
      "grad_norm": 3.5770678520202637,
      "learning_rate": 8.04743200334143e-05,
      "loss": 2.4147817611694338,
      "memory(GiB)": 77.56,
      "step": 34005,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.439678
    },
    {
      "epoch": 1.457092669551433,
      "grad_norm": 4.975116729736328,
      "learning_rate": 8.046898441672087e-05,
      "loss": 2.618977165222168,
      "memory(GiB)": 77.56,
      "step": 34010,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.439692
    },
    {
      "epoch": 1.457306884880682,
      "grad_norm": 4.206860542297363,
      "learning_rate": 8.046364824805321e-05,
      "loss": 2.3614410400390624,
      "memory(GiB)": 77.56,
      "step": 34015,
      "token_acc": 0.49050632911392406,
      "train_speed(iter/s)": 1.439701
    },
    {
      "epoch": 1.457521100209931,
      "grad_norm": 4.989882946014404,
      "learning_rate": 8.045831152750798e-05,
      "loss": 2.7668922424316404,
      "memory(GiB)": 77.56,
      "step": 34020,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.439718
    },
    {
      "epoch": 1.4577353155391801,
      "grad_norm": 5.06628942489624,
      "learning_rate": 8.045297425518188e-05,
      "loss": 2.5968347549438477,
      "memory(GiB)": 77.56,
      "step": 34025,
      "token_acc": 0.4962121212121212,
      "train_speed(iter/s)": 1.439756
    },
    {
      "epoch": 1.457949530868429,
      "grad_norm": 8.473918914794922,
      "learning_rate": 8.044763643117157e-05,
      "loss": 2.5514623641967775,
      "memory(GiB)": 77.56,
      "step": 34030,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.439795
    },
    {
      "epoch": 1.458163746197678,
      "grad_norm": 4.81611442565918,
      "learning_rate": 8.044229805557377e-05,
      "loss": 2.508599853515625,
      "memory(GiB)": 77.56,
      "step": 34035,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.439836
    },
    {
      "epoch": 1.458377961526927,
      "grad_norm": 5.073397159576416,
      "learning_rate": 8.043695912848517e-05,
      "loss": 2.504368209838867,
      "memory(GiB)": 77.56,
      "step": 34040,
      "token_acc": 0.493006993006993,
      "train_speed(iter/s)": 1.439839
    },
    {
      "epoch": 1.4585921768561758,
      "grad_norm": 7.687406063079834,
      "learning_rate": 8.043161965000252e-05,
      "loss": 2.450765037536621,
      "memory(GiB)": 77.56,
      "step": 34045,
      "token_acc": 0.45703125,
      "train_speed(iter/s)": 1.439883
    },
    {
      "epoch": 1.4588063921854248,
      "grad_norm": 4.375649452209473,
      "learning_rate": 8.042627962022252e-05,
      "loss": 3.0150527954101562,
      "memory(GiB)": 77.56,
      "step": 34050,
      "token_acc": 0.4061433447098976,
      "train_speed(iter/s)": 1.439857
    },
    {
      "epoch": 1.4590206075146739,
      "grad_norm": 3.875495433807373,
      "learning_rate": 8.042093903924193e-05,
      "loss": 2.516194152832031,
      "memory(GiB)": 77.56,
      "step": 34055,
      "token_acc": 0.49606299212598426,
      "train_speed(iter/s)": 1.439857
    },
    {
      "epoch": 1.4592348228439227,
      "grad_norm": 5.264041900634766,
      "learning_rate": 8.04155979071575e-05,
      "loss": 2.3918903350830076,
      "memory(GiB)": 77.56,
      "step": 34060,
      "token_acc": 0.47410358565737054,
      "train_speed(iter/s)": 1.439868
    },
    {
      "epoch": 1.4594490381731717,
      "grad_norm": 4.1805291175842285,
      "learning_rate": 8.041025622406596e-05,
      "loss": 2.6414152145385743,
      "memory(GiB)": 77.56,
      "step": 34065,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.439872
    },
    {
      "epoch": 1.4596632535024208,
      "grad_norm": 4.5763840675354,
      "learning_rate": 8.040491399006412e-05,
      "loss": 2.211083984375,
      "memory(GiB)": 77.56,
      "step": 34070,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.439862
    },
    {
      "epoch": 1.4598774688316696,
      "grad_norm": 5.411908149719238,
      "learning_rate": 8.039957120524872e-05,
      "loss": 2.4264049530029297,
      "memory(GiB)": 77.56,
      "step": 34075,
      "token_acc": 0.5136986301369864,
      "train_speed(iter/s)": 1.439893
    },
    {
      "epoch": 1.4600916841609186,
      "grad_norm": 4.275822639465332,
      "learning_rate": 8.03942278697166e-05,
      "loss": 2.6225685119628905,
      "memory(GiB)": 77.56,
      "step": 34080,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.439885
    },
    {
      "epoch": 1.4603058994901676,
      "grad_norm": 6.225549221038818,
      "learning_rate": 8.03888839835645e-05,
      "loss": 2.6890220642089844,
      "memory(GiB)": 77.56,
      "step": 34085,
      "token_acc": 0.4553072625698324,
      "train_speed(iter/s)": 1.439887
    },
    {
      "epoch": 1.4605201148194165,
      "grad_norm": 4.766172885894775,
      "learning_rate": 8.038353954688928e-05,
      "loss": 2.6291620254516603,
      "memory(GiB)": 77.56,
      "step": 34090,
      "token_acc": 0.4754601226993865,
      "train_speed(iter/s)": 1.439892
    },
    {
      "epoch": 1.4607343301486655,
      "grad_norm": 4.639849662780762,
      "learning_rate": 8.037819455978774e-05,
      "loss": 2.659840965270996,
      "memory(GiB)": 77.56,
      "step": 34095,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.439871
    },
    {
      "epoch": 1.4609485454779145,
      "grad_norm": 5.083648204803467,
      "learning_rate": 8.037284902235669e-05,
      "loss": 2.4212902069091795,
      "memory(GiB)": 77.56,
      "step": 34100,
      "token_acc": 0.4701195219123506,
      "train_speed(iter/s)": 1.439914
    },
    {
      "epoch": 1.4611627608071633,
      "grad_norm": 3.767834424972534,
      "learning_rate": 8.036750293469302e-05,
      "loss": 2.3488174438476563,
      "memory(GiB)": 77.56,
      "step": 34105,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.439948
    },
    {
      "epoch": 1.4613769761364124,
      "grad_norm": 5.209688186645508,
      "learning_rate": 8.036215629689352e-05,
      "loss": 2.3675933837890626,
      "memory(GiB)": 77.56,
      "step": 34110,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.439971
    },
    {
      "epoch": 1.4615911914656614,
      "grad_norm": 3.7660396099090576,
      "learning_rate": 8.035680910905508e-05,
      "loss": 2.0199440002441404,
      "memory(GiB)": 77.56,
      "step": 34115,
      "token_acc": 0.5400696864111498,
      "train_speed(iter/s)": 1.440005
    },
    {
      "epoch": 1.4618054067949102,
      "grad_norm": 4.780405044555664,
      "learning_rate": 8.035146137127458e-05,
      "loss": 2.6689279556274412,
      "memory(GiB)": 77.56,
      "step": 34120,
      "token_acc": 0.49337748344370863,
      "train_speed(iter/s)": 1.439987
    },
    {
      "epoch": 1.4620196221241593,
      "grad_norm": 4.523158073425293,
      "learning_rate": 8.034611308364888e-05,
      "loss": 2.525370788574219,
      "memory(GiB)": 77.56,
      "step": 34125,
      "token_acc": 0.4641638225255973,
      "train_speed(iter/s)": 1.440022
    },
    {
      "epoch": 1.4622338374534083,
      "grad_norm": 4.344518661499023,
      "learning_rate": 8.034076424627486e-05,
      "loss": 2.5201141357421877,
      "memory(GiB)": 77.56,
      "step": 34130,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.440066
    },
    {
      "epoch": 1.462448052782657,
      "grad_norm": 4.241621971130371,
      "learning_rate": 8.033541485924945e-05,
      "loss": 2.5197959899902345,
      "memory(GiB)": 77.56,
      "step": 34135,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.440021
    },
    {
      "epoch": 1.4626622681119061,
      "grad_norm": 4.401252269744873,
      "learning_rate": 8.033006492266952e-05,
      "loss": 2.6376155853271483,
      "memory(GiB)": 77.56,
      "step": 34140,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.439984
    },
    {
      "epoch": 1.4628764834411552,
      "grad_norm": 5.189802646636963,
      "learning_rate": 8.032471443663203e-05,
      "loss": 2.576529312133789,
      "memory(GiB)": 77.56,
      "step": 34145,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.440004
    },
    {
      "epoch": 1.463090698770404,
      "grad_norm": 5.572022914886475,
      "learning_rate": 8.03193634012339e-05,
      "loss": 2.6915861129760743,
      "memory(GiB)": 77.56,
      "step": 34150,
      "token_acc": 0.4479166666666667,
      "train_speed(iter/s)": 1.440023
    },
    {
      "epoch": 1.463304914099653,
      "grad_norm": 4.5874481201171875,
      "learning_rate": 8.031401181657206e-05,
      "loss": 2.855754089355469,
      "memory(GiB)": 77.56,
      "step": 34155,
      "token_acc": 0.45660377358490567,
      "train_speed(iter/s)": 1.440019
    },
    {
      "epoch": 1.463519129428902,
      "grad_norm": 5.819094181060791,
      "learning_rate": 8.030865968274344e-05,
      "loss": 2.8100864410400392,
      "memory(GiB)": 77.56,
      "step": 34160,
      "token_acc": 0.43157894736842106,
      "train_speed(iter/s)": 1.439986
    },
    {
      "epoch": 1.4637333447581509,
      "grad_norm": 7.241195201873779,
      "learning_rate": 8.030330699984503e-05,
      "loss": 2.7724287033081056,
      "memory(GiB)": 77.56,
      "step": 34165,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.439998
    },
    {
      "epoch": 1.4639475600874,
      "grad_norm": 5.7527995109558105,
      "learning_rate": 8.029795376797377e-05,
      "loss": 2.797378158569336,
      "memory(GiB)": 77.56,
      "step": 34170,
      "token_acc": 0.41839762611275966,
      "train_speed(iter/s)": 1.440013
    },
    {
      "epoch": 1.464161775416649,
      "grad_norm": 5.099137306213379,
      "learning_rate": 8.029259998722667e-05,
      "loss": 2.5382545471191404,
      "memory(GiB)": 77.56,
      "step": 34175,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.440041
    },
    {
      "epoch": 1.4643759907458977,
      "grad_norm": 3.8338465690612793,
      "learning_rate": 8.028724565770071e-05,
      "loss": 2.2501338958740233,
      "memory(GiB)": 77.56,
      "step": 34180,
      "token_acc": 0.5286624203821656,
      "train_speed(iter/s)": 1.440113
    },
    {
      "epoch": 1.4645902060751468,
      "grad_norm": 5.26433801651001,
      "learning_rate": 8.028189077949288e-05,
      "loss": 2.733415412902832,
      "memory(GiB)": 77.56,
      "step": 34185,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.440124
    },
    {
      "epoch": 1.4648044214043958,
      "grad_norm": 5.47374963760376,
      "learning_rate": 8.027653535270019e-05,
      "loss": 2.495956039428711,
      "memory(GiB)": 77.56,
      "step": 34190,
      "token_acc": 0.4774193548387097,
      "train_speed(iter/s)": 1.440126
    },
    {
      "epoch": 1.4650186367336446,
      "grad_norm": 5.073783874511719,
      "learning_rate": 8.027117937741966e-05,
      "loss": 2.509799575805664,
      "memory(GiB)": 77.56,
      "step": 34195,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.440103
    },
    {
      "epoch": 1.4652328520628937,
      "grad_norm": 4.862172603607178,
      "learning_rate": 8.026582285374832e-05,
      "loss": 2.429007911682129,
      "memory(GiB)": 77.56,
      "step": 34200,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.440133
    },
    {
      "epoch": 1.4654470673921427,
      "grad_norm": 5.415630340576172,
      "learning_rate": 8.02604657817832e-05,
      "loss": 2.4519561767578124,
      "memory(GiB)": 77.56,
      "step": 34205,
      "token_acc": 0.5170278637770898,
      "train_speed(iter/s)": 1.440148
    },
    {
      "epoch": 1.4656612827213915,
      "grad_norm": 4.675723075866699,
      "learning_rate": 8.025510816162137e-05,
      "loss": 2.3500009536743165,
      "memory(GiB)": 77.56,
      "step": 34210,
      "token_acc": 0.5130434782608696,
      "train_speed(iter/s)": 1.440096
    },
    {
      "epoch": 1.4658754980506405,
      "grad_norm": 6.152647972106934,
      "learning_rate": 8.024974999335985e-05,
      "loss": 2.4933956146240233,
      "memory(GiB)": 77.56,
      "step": 34215,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.440128
    },
    {
      "epoch": 1.4660897133798896,
      "grad_norm": 4.551023960113525,
      "learning_rate": 8.024439127709575e-05,
      "loss": 2.600601387023926,
      "memory(GiB)": 77.56,
      "step": 34220,
      "token_acc": 0.4290322580645161,
      "train_speed(iter/s)": 1.44012
    },
    {
      "epoch": 1.4663039287091384,
      "grad_norm": 5.539081573486328,
      "learning_rate": 8.023903201292613e-05,
      "loss": 2.3947078704833986,
      "memory(GiB)": 77.56,
      "step": 34225,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.440142
    },
    {
      "epoch": 1.4665181440383874,
      "grad_norm": 6.135366439819336,
      "learning_rate": 8.023367220094809e-05,
      "loss": 2.3898338317871093,
      "memory(GiB)": 77.56,
      "step": 34230,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.440204
    },
    {
      "epoch": 1.4667323593676365,
      "grad_norm": 5.819205284118652,
      "learning_rate": 8.02283118412587e-05,
      "loss": 2.4377296447753904,
      "memory(GiB)": 77.56,
      "step": 34235,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.440243
    },
    {
      "epoch": 1.4669465746968853,
      "grad_norm": 4.292762279510498,
      "learning_rate": 8.022295093395509e-05,
      "loss": 2.4075897216796873,
      "memory(GiB)": 77.56,
      "step": 34240,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.440261
    },
    {
      "epoch": 1.4671607900261343,
      "grad_norm": 5.881063461303711,
      "learning_rate": 8.021758947913436e-05,
      "loss": 2.4541038513183593,
      "memory(GiB)": 77.56,
      "step": 34245,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.440211
    },
    {
      "epoch": 1.4673750053553833,
      "grad_norm": 5.276733875274658,
      "learning_rate": 8.021222747689367e-05,
      "loss": 2.558882713317871,
      "memory(GiB)": 77.56,
      "step": 34250,
      "token_acc": 0.4875,
      "train_speed(iter/s)": 1.440236
    },
    {
      "epoch": 1.4675892206846322,
      "grad_norm": 7.2614336013793945,
      "learning_rate": 8.020686492733011e-05,
      "loss": 2.5538002014160157,
      "memory(GiB)": 77.56,
      "step": 34255,
      "token_acc": 0.4432624113475177,
      "train_speed(iter/s)": 1.44026
    },
    {
      "epoch": 1.4678034360138812,
      "grad_norm": 5.199793815612793,
      "learning_rate": 8.02015018305409e-05,
      "loss": 2.4433361053466798,
      "memory(GiB)": 77.56,
      "step": 34260,
      "token_acc": 0.5207547169811321,
      "train_speed(iter/s)": 1.440248
    },
    {
      "epoch": 1.4680176513431302,
      "grad_norm": 5.548585891723633,
      "learning_rate": 8.019613818662312e-05,
      "loss": 2.35800838470459,
      "memory(GiB)": 77.56,
      "step": 34265,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440284
    },
    {
      "epoch": 1.468231866672379,
      "grad_norm": 3.831486701965332,
      "learning_rate": 8.019077399567398e-05,
      "loss": 2.6825544357299806,
      "memory(GiB)": 77.56,
      "step": 34270,
      "token_acc": 0.4637223974763407,
      "train_speed(iter/s)": 1.440324
    },
    {
      "epoch": 1.468446082001628,
      "grad_norm": 4.019141674041748,
      "learning_rate": 8.018540925779064e-05,
      "loss": 2.5327083587646486,
      "memory(GiB)": 77.56,
      "step": 34275,
      "token_acc": 0.4563380281690141,
      "train_speed(iter/s)": 1.440299
    },
    {
      "epoch": 1.468660297330877,
      "grad_norm": 4.661750316619873,
      "learning_rate": 8.018004397307031e-05,
      "loss": 2.8629390716552736,
      "memory(GiB)": 77.56,
      "step": 34280,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.440337
    },
    {
      "epoch": 1.468874512660126,
      "grad_norm": 5.158905506134033,
      "learning_rate": 8.017467814161015e-05,
      "loss": 2.487758445739746,
      "memory(GiB)": 77.56,
      "step": 34285,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.440325
    },
    {
      "epoch": 1.469088727989375,
      "grad_norm": 5.842710018157959,
      "learning_rate": 8.016931176350741e-05,
      "loss": 2.309349250793457,
      "memory(GiB)": 77.56,
      "step": 34290,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.440307
    },
    {
      "epoch": 1.469302943318624,
      "grad_norm": 4.598109245300293,
      "learning_rate": 8.016394483885928e-05,
      "loss": 2.6250984191894533,
      "memory(GiB)": 77.56,
      "step": 34295,
      "token_acc": 0.4984126984126984,
      "train_speed(iter/s)": 1.44036
    },
    {
      "epoch": 1.4695171586478728,
      "grad_norm": 5.496702671051025,
      "learning_rate": 8.015857736776299e-05,
      "loss": 2.464499282836914,
      "memory(GiB)": 77.56,
      "step": 34300,
      "token_acc": 0.4784172661870504,
      "train_speed(iter/s)": 1.440343
    },
    {
      "epoch": 1.4697313739771218,
      "grad_norm": 6.301141262054443,
      "learning_rate": 8.015320935031579e-05,
      "loss": 2.6534072875976564,
      "memory(GiB)": 77.56,
      "step": 34305,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.440374
    },
    {
      "epoch": 1.4699455893063709,
      "grad_norm": 5.152554512023926,
      "learning_rate": 8.014784078661491e-05,
      "loss": 2.890163612365723,
      "memory(GiB)": 77.56,
      "step": 34310,
      "token_acc": 0.42657342657342656,
      "train_speed(iter/s)": 1.440373
    },
    {
      "epoch": 1.4701598046356197,
      "grad_norm": 4.3119072914123535,
      "learning_rate": 8.01424716767576e-05,
      "loss": 2.4026954650878904,
      "memory(GiB)": 77.56,
      "step": 34315,
      "token_acc": 0.47039473684210525,
      "train_speed(iter/s)": 1.440382
    },
    {
      "epoch": 1.4703740199648687,
      "grad_norm": 6.448112964630127,
      "learning_rate": 8.013710202084115e-05,
      "loss": 2.0851724624633787,
      "memory(GiB)": 77.56,
      "step": 34320,
      "token_acc": 0.5364806866952789,
      "train_speed(iter/s)": 1.440353
    },
    {
      "epoch": 1.4705882352941178,
      "grad_norm": 4.141831398010254,
      "learning_rate": 8.013173181896283e-05,
      "loss": 2.7041042327880858,
      "memory(GiB)": 77.56,
      "step": 34325,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.440367
    },
    {
      "epoch": 1.4708024506233666,
      "grad_norm": 5.361547470092773,
      "learning_rate": 8.012636107121992e-05,
      "loss": 2.039295959472656,
      "memory(GiB)": 77.56,
      "step": 34330,
      "token_acc": 0.5165289256198347,
      "train_speed(iter/s)": 1.440322
    },
    {
      "epoch": 1.4710166659526156,
      "grad_norm": 5.071845054626465,
      "learning_rate": 8.012098977770971e-05,
      "loss": 2.9382888793945314,
      "memory(GiB)": 77.56,
      "step": 34335,
      "token_acc": 0.4311594202898551,
      "train_speed(iter/s)": 1.440334
    },
    {
      "epoch": 1.4712308812818646,
      "grad_norm": 6.151006698608398,
      "learning_rate": 8.011561793852953e-05,
      "loss": 2.5644079208374024,
      "memory(GiB)": 77.56,
      "step": 34340,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.440351
    },
    {
      "epoch": 1.4714450966111134,
      "grad_norm": 3.985409736633301,
      "learning_rate": 8.011024555377667e-05,
      "loss": 2.2681888580322265,
      "memory(GiB)": 77.56,
      "step": 34345,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.440325
    },
    {
      "epoch": 1.4716593119403625,
      "grad_norm": 4.982699394226074,
      "learning_rate": 8.010487262354847e-05,
      "loss": 2.3488372802734374,
      "memory(GiB)": 77.56,
      "step": 34350,
      "token_acc": 0.5190839694656488,
      "train_speed(iter/s)": 1.440413
    },
    {
      "epoch": 1.4718735272696115,
      "grad_norm": 4.712373733520508,
      "learning_rate": 8.009949914794226e-05,
      "loss": 2.4499528884887694,
      "memory(GiB)": 77.56,
      "step": 34355,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.440408
    },
    {
      "epoch": 1.4720877425988603,
      "grad_norm": 6.626208782196045,
      "learning_rate": 8.009412512705539e-05,
      "loss": 2.5087371826171876,
      "memory(GiB)": 77.56,
      "step": 34360,
      "token_acc": 0.4370629370629371,
      "train_speed(iter/s)": 1.440386
    },
    {
      "epoch": 1.4723019579281094,
      "grad_norm": 4.252260684967041,
      "learning_rate": 8.00887505609852e-05,
      "loss": 2.5764888763427733,
      "memory(GiB)": 77.56,
      "step": 34365,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.440402
    },
    {
      "epoch": 1.4725161732573584,
      "grad_norm": 5.475830078125,
      "learning_rate": 8.008337544982909e-05,
      "loss": 2.4365913391113283,
      "memory(GiB)": 77.56,
      "step": 34370,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.440406
    },
    {
      "epoch": 1.4727303885866072,
      "grad_norm": 8.2151460647583,
      "learning_rate": 8.00779997936844e-05,
      "loss": 2.3214399337768556,
      "memory(GiB)": 77.56,
      "step": 34375,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.440429
    },
    {
      "epoch": 1.4729446039158562,
      "grad_norm": 5.181869029998779,
      "learning_rate": 8.007262359264852e-05,
      "loss": 2.286582183837891,
      "memory(GiB)": 77.56,
      "step": 34380,
      "token_acc": 0.49337748344370863,
      "train_speed(iter/s)": 1.44041
    },
    {
      "epoch": 1.4731588192451053,
      "grad_norm": 4.501758098602295,
      "learning_rate": 8.006724684681888e-05,
      "loss": 2.3877639770507812,
      "memory(GiB)": 77.56,
      "step": 34385,
      "token_acc": 0.45724907063197023,
      "train_speed(iter/s)": 1.440394
    },
    {
      "epoch": 1.473373034574354,
      "grad_norm": 4.29125452041626,
      "learning_rate": 8.006186955629282e-05,
      "loss": 2.359326171875,
      "memory(GiB)": 77.56,
      "step": 34390,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440405
    },
    {
      "epoch": 1.4735872499036031,
      "grad_norm": 4.701378345489502,
      "learning_rate": 8.005649172116782e-05,
      "loss": 2.644342803955078,
      "memory(GiB)": 77.56,
      "step": 34395,
      "token_acc": 0.43,
      "train_speed(iter/s)": 1.440341
    },
    {
      "epoch": 1.4738014652328522,
      "grad_norm": 5.0990800857543945,
      "learning_rate": 8.005111334154127e-05,
      "loss": 2.750717544555664,
      "memory(GiB)": 77.56,
      "step": 34400,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440329
    },
    {
      "epoch": 1.474015680562101,
      "grad_norm": 4.347292900085449,
      "learning_rate": 8.004573441751062e-05,
      "loss": 2.541716766357422,
      "memory(GiB)": 77.56,
      "step": 34405,
      "token_acc": 0.4813664596273292,
      "train_speed(iter/s)": 1.440306
    },
    {
      "epoch": 1.47422989589135,
      "grad_norm": 6.2339348793029785,
      "learning_rate": 8.00403549491733e-05,
      "loss": 2.711593246459961,
      "memory(GiB)": 77.56,
      "step": 34410,
      "token_acc": 0.44528301886792454,
      "train_speed(iter/s)": 1.440317
    },
    {
      "epoch": 1.474444111220599,
      "grad_norm": 5.4573235511779785,
      "learning_rate": 8.003497493662678e-05,
      "loss": 2.347819137573242,
      "memory(GiB)": 77.56,
      "step": 34415,
      "token_acc": 0.48534201954397393,
      "train_speed(iter/s)": 1.44032
    },
    {
      "epoch": 1.4746583265498479,
      "grad_norm": 6.5493693351745605,
      "learning_rate": 8.002959437996849e-05,
      "loss": 2.5783565521240233,
      "memory(GiB)": 77.56,
      "step": 34420,
      "token_acc": 0.4849624060150376,
      "train_speed(iter/s)": 1.440303
    },
    {
      "epoch": 1.4748725418790969,
      "grad_norm": 5.091011047363281,
      "learning_rate": 8.002421327929592e-05,
      "loss": 2.512987518310547,
      "memory(GiB)": 77.56,
      "step": 34425,
      "token_acc": 0.4533333333333333,
      "train_speed(iter/s)": 1.440272
    },
    {
      "epoch": 1.475086757208346,
      "grad_norm": 5.855741024017334,
      "learning_rate": 8.001883163470661e-05,
      "loss": 2.4178829193115234,
      "memory(GiB)": 77.56,
      "step": 34430,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.440228
    },
    {
      "epoch": 1.4753009725375947,
      "grad_norm": 5.49994421005249,
      "learning_rate": 8.001344944629796e-05,
      "loss": 2.5528759002685546,
      "memory(GiB)": 77.56,
      "step": 34435,
      "token_acc": 0.4940239043824701,
      "train_speed(iter/s)": 1.440271
    },
    {
      "epoch": 1.4755151878668438,
      "grad_norm": 5.74098539352417,
      "learning_rate": 8.000806671416754e-05,
      "loss": 2.475512886047363,
      "memory(GiB)": 77.56,
      "step": 34440,
      "token_acc": 0.49224806201550386,
      "train_speed(iter/s)": 1.440302
    },
    {
      "epoch": 1.4757294031960928,
      "grad_norm": 4.453502178192139,
      "learning_rate": 8.000268343841283e-05,
      "loss": 2.661197471618652,
      "memory(GiB)": 77.56,
      "step": 34445,
      "token_acc": 0.4065040650406504,
      "train_speed(iter/s)": 1.440334
    },
    {
      "epoch": 1.4759436185253416,
      "grad_norm": 4.792104244232178,
      "learning_rate": 7.999729961913139e-05,
      "loss": 2.5993507385253904,
      "memory(GiB)": 77.56,
      "step": 34450,
      "token_acc": 0.48299319727891155,
      "train_speed(iter/s)": 1.440383
    },
    {
      "epoch": 1.4761578338545907,
      "grad_norm": 5.920977592468262,
      "learning_rate": 7.999191525642069e-05,
      "loss": 2.9311634063720704,
      "memory(GiB)": 77.56,
      "step": 34455,
      "token_acc": 0.4148148148148148,
      "train_speed(iter/s)": 1.440431
    },
    {
      "epoch": 1.4763720491838397,
      "grad_norm": 4.11167049407959,
      "learning_rate": 7.998653035037834e-05,
      "loss": 2.637973976135254,
      "memory(GiB)": 77.56,
      "step": 34460,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.44042
    },
    {
      "epoch": 1.4765862645130885,
      "grad_norm": 5.867785453796387,
      "learning_rate": 7.998114490110185e-05,
      "loss": 2.431184196472168,
      "memory(GiB)": 77.56,
      "step": 34465,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.440384
    },
    {
      "epoch": 1.4768004798423375,
      "grad_norm": 4.38999080657959,
      "learning_rate": 7.997575890868879e-05,
      "loss": 2.4927196502685547,
      "memory(GiB)": 77.56,
      "step": 34470,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.440414
    },
    {
      "epoch": 1.4770146951715866,
      "grad_norm": 5.949841022491455,
      "learning_rate": 7.997037237323675e-05,
      "loss": 2.3301158905029298,
      "memory(GiB)": 77.56,
      "step": 34475,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.440382
    },
    {
      "epoch": 1.4772289105008354,
      "grad_norm": 3.831998586654663,
      "learning_rate": 7.99649852948433e-05,
      "loss": 2.8143234252929688,
      "memory(GiB)": 77.56,
      "step": 34480,
      "token_acc": 0.4046997389033943,
      "train_speed(iter/s)": 1.440376
    },
    {
      "epoch": 1.4774431258300844,
      "grad_norm": 4.94537353515625,
      "learning_rate": 7.995959767360604e-05,
      "loss": 2.339773178100586,
      "memory(GiB)": 77.56,
      "step": 34485,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.440435
    },
    {
      "epoch": 1.4776573411593334,
      "grad_norm": 6.495689868927002,
      "learning_rate": 7.995420950962254e-05,
      "loss": 2.6484256744384767,
      "memory(GiB)": 77.56,
      "step": 34490,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440399
    },
    {
      "epoch": 1.4778715564885823,
      "grad_norm": 4.667993068695068,
      "learning_rate": 7.994882080299044e-05,
      "loss": 2.440870666503906,
      "memory(GiB)": 77.56,
      "step": 34495,
      "token_acc": 0.5058823529411764,
      "train_speed(iter/s)": 1.440427
    },
    {
      "epoch": 1.4780857718178313,
      "grad_norm": 6.729407787322998,
      "learning_rate": 7.994343155380737e-05,
      "loss": 2.530316925048828,
      "memory(GiB)": 77.56,
      "step": 34500,
      "token_acc": 0.46558704453441296,
      "train_speed(iter/s)": 1.440394
    },
    {
      "epoch": 1.4780857718178313,
      "eval_loss": 2.260627508163452,
      "eval_runtime": 14.256,
      "eval_samples_per_second": 7.015,
      "eval_steps_per_second": 7.015,
      "eval_token_acc": 0.46853146853146854,
      "step": 34500
    },
    {
      "epoch": 1.4782999871470803,
      "grad_norm": 5.972782611846924,
      "learning_rate": 7.993804176217093e-05,
      "loss": 2.6705720901489256,
      "memory(GiB)": 77.56,
      "step": 34505,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.439537
    },
    {
      "epoch": 1.4785142024763291,
      "grad_norm": 5.555943965911865,
      "learning_rate": 7.993265142817881e-05,
      "loss": 2.364592933654785,
      "memory(GiB)": 77.56,
      "step": 34510,
      "token_acc": 0.5058823529411764,
      "train_speed(iter/s)": 1.43956
    },
    {
      "epoch": 1.4787284178055782,
      "grad_norm": 4.481245994567871,
      "learning_rate": 7.99272605519286e-05,
      "loss": 2.3735366821289063,
      "memory(GiB)": 77.56,
      "step": 34515,
      "token_acc": 0.445578231292517,
      "train_speed(iter/s)": 1.439557
    },
    {
      "epoch": 1.4789426331348272,
      "grad_norm": 5.238328456878662,
      "learning_rate": 7.9921869133518e-05,
      "loss": 2.487208938598633,
      "memory(GiB)": 77.56,
      "step": 34520,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.439509
    },
    {
      "epoch": 1.479156848464076,
      "grad_norm": 5.508119106292725,
      "learning_rate": 7.991647717304467e-05,
      "loss": 2.756937026977539,
      "memory(GiB)": 77.56,
      "step": 34525,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.439488
    },
    {
      "epoch": 1.479371063793325,
      "grad_norm": 4.803337574005127,
      "learning_rate": 7.99110846706063e-05,
      "loss": 2.5147260665893554,
      "memory(GiB)": 77.56,
      "step": 34530,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439493
    },
    {
      "epoch": 1.479585279122574,
      "grad_norm": 5.185335159301758,
      "learning_rate": 7.990569162630057e-05,
      "loss": 2.4062711715698244,
      "memory(GiB)": 77.56,
      "step": 34535,
      "token_acc": 0.47924528301886793,
      "train_speed(iter/s)": 1.439423
    },
    {
      "epoch": 1.479799494451823,
      "grad_norm": 5.346336364746094,
      "learning_rate": 7.990029804022518e-05,
      "loss": 2.4619701385498045,
      "memory(GiB)": 77.56,
      "step": 34540,
      "token_acc": 0.4924924924924925,
      "train_speed(iter/s)": 1.439417
    },
    {
      "epoch": 1.480013709781072,
      "grad_norm": 5.0395965576171875,
      "learning_rate": 7.989490391247784e-05,
      "loss": 2.1731245040893556,
      "memory(GiB)": 77.56,
      "step": 34545,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.439383
    },
    {
      "epoch": 1.480227925110321,
      "grad_norm": 4.642335891723633,
      "learning_rate": 7.988950924315628e-05,
      "loss": 2.566513442993164,
      "memory(GiB)": 77.56,
      "step": 34550,
      "token_acc": 0.5032258064516129,
      "train_speed(iter/s)": 1.439449
    },
    {
      "epoch": 1.4804421404395698,
      "grad_norm": 4.903952121734619,
      "learning_rate": 7.988411403235823e-05,
      "loss": 2.508007049560547,
      "memory(GiB)": 77.56,
      "step": 34555,
      "token_acc": 0.4716417910447761,
      "train_speed(iter/s)": 1.43951
    },
    {
      "epoch": 1.4806563557688188,
      "grad_norm": 3.999094009399414,
      "learning_rate": 7.987871828018141e-05,
      "loss": 2.6111896514892576,
      "memory(GiB)": 77.56,
      "step": 34560,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.439534
    },
    {
      "epoch": 1.4808705710980679,
      "grad_norm": 5.334130764007568,
      "learning_rate": 7.987332198672356e-05,
      "loss": 2.630277633666992,
      "memory(GiB)": 77.56,
      "step": 34565,
      "token_acc": 0.43944636678200694,
      "train_speed(iter/s)": 1.439545
    },
    {
      "epoch": 1.4810847864273167,
      "grad_norm": 5.522465705871582,
      "learning_rate": 7.986792515208248e-05,
      "loss": 2.4380071640014647,
      "memory(GiB)": 77.56,
      "step": 34570,
      "token_acc": 0.483739837398374,
      "train_speed(iter/s)": 1.439587
    },
    {
      "epoch": 1.4812990017565657,
      "grad_norm": 4.230618476867676,
      "learning_rate": 7.986252777635592e-05,
      "loss": 2.4863521575927736,
      "memory(GiB)": 77.56,
      "step": 34575,
      "token_acc": 0.4684931506849315,
      "train_speed(iter/s)": 1.439575
    },
    {
      "epoch": 1.4815132170858147,
      "grad_norm": 4.854522705078125,
      "learning_rate": 7.985712985964164e-05,
      "loss": 2.2017822265625,
      "memory(GiB)": 77.56,
      "step": 34580,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.439599
    },
    {
      "epoch": 1.4817274324150636,
      "grad_norm": 6.0103020668029785,
      "learning_rate": 7.985173140203745e-05,
      "loss": 2.481979179382324,
      "memory(GiB)": 77.56,
      "step": 34585,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.439589
    },
    {
      "epoch": 1.4819416477443126,
      "grad_norm": 4.527801036834717,
      "learning_rate": 7.984633240364116e-05,
      "loss": 2.6746572494506835,
      "memory(GiB)": 77.56,
      "step": 34590,
      "token_acc": 0.46865671641791046,
      "train_speed(iter/s)": 1.439564
    },
    {
      "epoch": 1.4821558630735616,
      "grad_norm": 6.16262674331665,
      "learning_rate": 7.984093286455055e-05,
      "loss": 2.4438610076904297,
      "memory(GiB)": 77.56,
      "step": 34595,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.439606
    },
    {
      "epoch": 1.4823700784028104,
      "grad_norm": 5.266121864318848,
      "learning_rate": 7.983553278486344e-05,
      "loss": 2.8454837799072266,
      "memory(GiB)": 77.56,
      "step": 34600,
      "token_acc": 0.4310850439882698,
      "train_speed(iter/s)": 1.43964
    },
    {
      "epoch": 1.4825842937320595,
      "grad_norm": 5.369907379150391,
      "learning_rate": 7.983013216467768e-05,
      "loss": 2.5199420928955076,
      "memory(GiB)": 77.56,
      "step": 34605,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.439642
    },
    {
      "epoch": 1.4827985090613085,
      "grad_norm": 8.589921951293945,
      "learning_rate": 7.982473100409107e-05,
      "loss": 2.506466102600098,
      "memory(GiB)": 77.56,
      "step": 34610,
      "token_acc": 0.4575645756457565,
      "train_speed(iter/s)": 1.439691
    },
    {
      "epoch": 1.4830127243905573,
      "grad_norm": 5.31126070022583,
      "learning_rate": 7.981932930320149e-05,
      "loss": 2.659066390991211,
      "memory(GiB)": 77.56,
      "step": 34615,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.439699
    },
    {
      "epoch": 1.4832269397198063,
      "grad_norm": 5.200987815856934,
      "learning_rate": 7.98139270621068e-05,
      "loss": 2.4870738983154297,
      "memory(GiB)": 77.56,
      "step": 34620,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.439713
    },
    {
      "epoch": 1.4834411550490554,
      "grad_norm": 6.352973937988281,
      "learning_rate": 7.980852428090484e-05,
      "loss": 2.5357471466064454,
      "memory(GiB)": 77.56,
      "step": 34625,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.439723
    },
    {
      "epoch": 1.4836553703783042,
      "grad_norm": 5.85141134262085,
      "learning_rate": 7.980312095969351e-05,
      "loss": 2.693460464477539,
      "memory(GiB)": 77.56,
      "step": 34630,
      "token_acc": 0.45819397993311034,
      "train_speed(iter/s)": 1.439776
    },
    {
      "epoch": 1.4838695857075532,
      "grad_norm": 4.918947219848633,
      "learning_rate": 7.979771709857066e-05,
      "loss": 2.4680059432983397,
      "memory(GiB)": 77.56,
      "step": 34635,
      "token_acc": 0.45483870967741935,
      "train_speed(iter/s)": 1.439838
    },
    {
      "epoch": 1.4840838010368023,
      "grad_norm": 4.078084945678711,
      "learning_rate": 7.979231269763425e-05,
      "loss": 2.4555105209350585,
      "memory(GiB)": 77.56,
      "step": 34640,
      "token_acc": 0.5117845117845118,
      "train_speed(iter/s)": 1.439827
    },
    {
      "epoch": 1.484298016366051,
      "grad_norm": 4.0401177406311035,
      "learning_rate": 7.978690775698213e-05,
      "loss": 2.245615768432617,
      "memory(GiB)": 77.56,
      "step": 34645,
      "token_acc": 0.5387323943661971,
      "train_speed(iter/s)": 1.439797
    },
    {
      "epoch": 1.4845122316953,
      "grad_norm": 5.768013000488281,
      "learning_rate": 7.978150227671223e-05,
      "loss": 2.546351432800293,
      "memory(GiB)": 77.56,
      "step": 34650,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.439806
    },
    {
      "epoch": 1.4847264470245491,
      "grad_norm": 5.7937235832214355,
      "learning_rate": 7.977609625692248e-05,
      "loss": 2.3059606552124023,
      "memory(GiB)": 77.56,
      "step": 34655,
      "token_acc": 0.515748031496063,
      "train_speed(iter/s)": 1.439813
    },
    {
      "epoch": 1.484940662353798,
      "grad_norm": 4.96259069442749,
      "learning_rate": 7.977068969771083e-05,
      "loss": 2.3303560256958007,
      "memory(GiB)": 77.56,
      "step": 34660,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.439808
    },
    {
      "epoch": 1.485154877683047,
      "grad_norm": 4.985846042633057,
      "learning_rate": 7.97652825991752e-05,
      "loss": 2.5304471969604494,
      "memory(GiB)": 77.56,
      "step": 34665,
      "token_acc": 0.44879518072289154,
      "train_speed(iter/s)": 1.439825
    },
    {
      "epoch": 1.485369093012296,
      "grad_norm": 4.745893955230713,
      "learning_rate": 7.975987496141354e-05,
      "loss": 2.5578933715820313,
      "memory(GiB)": 77.56,
      "step": 34670,
      "token_acc": 0.5130111524163569,
      "train_speed(iter/s)": 1.439821
    },
    {
      "epoch": 1.4855833083415448,
      "grad_norm": 5.999105930328369,
      "learning_rate": 7.975446678452384e-05,
      "loss": 2.607408142089844,
      "memory(GiB)": 77.56,
      "step": 34675,
      "token_acc": 0.45723684210526316,
      "train_speed(iter/s)": 1.439876
    },
    {
      "epoch": 1.4857975236707939,
      "grad_norm": 8.275153160095215,
      "learning_rate": 7.974905806860407e-05,
      "loss": 2.397511672973633,
      "memory(GiB)": 77.56,
      "step": 34680,
      "token_acc": 0.4744525547445255,
      "train_speed(iter/s)": 1.439828
    },
    {
      "epoch": 1.486011739000043,
      "grad_norm": 3.7610960006713867,
      "learning_rate": 7.974364881375218e-05,
      "loss": 2.3691444396972656,
      "memory(GiB)": 77.56,
      "step": 34685,
      "token_acc": 0.5155038759689923,
      "train_speed(iter/s)": 1.439815
    },
    {
      "epoch": 1.4862259543292917,
      "grad_norm": 4.706734657287598,
      "learning_rate": 7.97382390200662e-05,
      "loss": 2.685099983215332,
      "memory(GiB)": 77.56,
      "step": 34690,
      "token_acc": 0.44594594594594594,
      "train_speed(iter/s)": 1.439811
    },
    {
      "epoch": 1.4864401696585408,
      "grad_norm": 4.949123859405518,
      "learning_rate": 7.973282868764413e-05,
      "loss": 2.124616241455078,
      "memory(GiB)": 77.56,
      "step": 34695,
      "token_acc": 0.5020746887966805,
      "train_speed(iter/s)": 1.4398
    },
    {
      "epoch": 1.4866543849877898,
      "grad_norm": 5.066686630249023,
      "learning_rate": 7.972741781658399e-05,
      "loss": 2.717406463623047,
      "memory(GiB)": 77.56,
      "step": 34700,
      "token_acc": 0.44333333333333336,
      "train_speed(iter/s)": 1.439807
    },
    {
      "epoch": 1.4868686003170386,
      "grad_norm": 5.711490631103516,
      "learning_rate": 7.972200640698377e-05,
      "loss": 2.4084260940551756,
      "memory(GiB)": 77.56,
      "step": 34705,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.439799
    },
    {
      "epoch": 1.4870828156462876,
      "grad_norm": 4.470431327819824,
      "learning_rate": 7.971659445894152e-05,
      "loss": 2.8833309173583985,
      "memory(GiB)": 77.56,
      "step": 34710,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.43983
    },
    {
      "epoch": 1.4872970309755367,
      "grad_norm": 4.8883209228515625,
      "learning_rate": 7.97111819725553e-05,
      "loss": 2.4811426162719727,
      "memory(GiB)": 77.56,
      "step": 34715,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.439869
    },
    {
      "epoch": 1.4875112463047855,
      "grad_norm": 4.814483642578125,
      "learning_rate": 7.970576894792314e-05,
      "loss": 2.6018882751464845,
      "memory(GiB)": 77.56,
      "step": 34720,
      "token_acc": 0.4440993788819876,
      "train_speed(iter/s)": 1.439942
    },
    {
      "epoch": 1.4877254616340345,
      "grad_norm": 7.101289749145508,
      "learning_rate": 7.97003553851431e-05,
      "loss": 2.6370811462402344,
      "memory(GiB)": 77.56,
      "step": 34725,
      "token_acc": 0.4558303886925795,
      "train_speed(iter/s)": 1.439979
    },
    {
      "epoch": 1.4879396769632836,
      "grad_norm": 5.749527454376221,
      "learning_rate": 7.969494128431327e-05,
      "loss": 2.659012222290039,
      "memory(GiB)": 77.56,
      "step": 34730,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.439976
    },
    {
      "epoch": 1.4881538922925324,
      "grad_norm": 5.493963718414307,
      "learning_rate": 7.968952664553172e-05,
      "loss": 2.4715549468994142,
      "memory(GiB)": 77.56,
      "step": 34735,
      "token_acc": 0.4866920152091255,
      "train_speed(iter/s)": 1.439972
    },
    {
      "epoch": 1.4883681076217814,
      "grad_norm": 5.084977149963379,
      "learning_rate": 7.968411146889656e-05,
      "loss": 2.2956264495849608,
      "memory(GiB)": 77.56,
      "step": 34740,
      "token_acc": 0.523972602739726,
      "train_speed(iter/s)": 1.439997
    },
    {
      "epoch": 1.4885823229510304,
      "grad_norm": 6.531550407409668,
      "learning_rate": 7.967869575450587e-05,
      "loss": 2.8708534240722656,
      "memory(GiB)": 77.56,
      "step": 34745,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.440054
    },
    {
      "epoch": 1.4887965382802792,
      "grad_norm": 4.791491508483887,
      "learning_rate": 7.967327950245775e-05,
      "loss": 2.3526893615722657,
      "memory(GiB)": 77.56,
      "step": 34750,
      "token_acc": 0.4541984732824427,
      "train_speed(iter/s)": 1.440083
    },
    {
      "epoch": 1.4890107536095283,
      "grad_norm": 4.072079658508301,
      "learning_rate": 7.966786271285034e-05,
      "loss": 2.3518356323242187,
      "memory(GiB)": 77.56,
      "step": 34755,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.44009
    },
    {
      "epoch": 1.4892249689387773,
      "grad_norm": 4.942510604858398,
      "learning_rate": 7.966244538578177e-05,
      "loss": 2.668620300292969,
      "memory(GiB)": 77.56,
      "step": 34760,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.44011
    },
    {
      "epoch": 1.4894391842680261,
      "grad_norm": 6.58004903793335,
      "learning_rate": 7.965702752135018e-05,
      "loss": 2.452613639831543,
      "memory(GiB)": 77.56,
      "step": 34765,
      "token_acc": 0.4959016393442623,
      "train_speed(iter/s)": 1.440126
    },
    {
      "epoch": 1.4896533995972752,
      "grad_norm": 6.290943145751953,
      "learning_rate": 7.965160911965371e-05,
      "loss": 2.2688323974609377,
      "memory(GiB)": 77.56,
      "step": 34770,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440159
    },
    {
      "epoch": 1.4898676149265242,
      "grad_norm": 4.718799591064453,
      "learning_rate": 7.964619018079054e-05,
      "loss": 2.085108757019043,
      "memory(GiB)": 77.56,
      "step": 34775,
      "token_acc": 0.5351170568561873,
      "train_speed(iter/s)": 1.44017
    },
    {
      "epoch": 1.490081830255773,
      "grad_norm": 5.204291820526123,
      "learning_rate": 7.964077070485881e-05,
      "loss": 2.4292119979858398,
      "memory(GiB)": 77.56,
      "step": 34780,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.440205
    },
    {
      "epoch": 1.490296045585022,
      "grad_norm": 6.061275482177734,
      "learning_rate": 7.963535069195671e-05,
      "loss": 2.5785268783569335,
      "memory(GiB)": 77.56,
      "step": 34785,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.440192
    },
    {
      "epoch": 1.490510260914271,
      "grad_norm": 4.104846477508545,
      "learning_rate": 7.962993014218243e-05,
      "loss": 2.636448097229004,
      "memory(GiB)": 77.56,
      "step": 34790,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.440179
    },
    {
      "epoch": 1.49072447624352,
      "grad_norm": 4.877756595611572,
      "learning_rate": 7.962450905563418e-05,
      "loss": 2.6018190383911133,
      "memory(GiB)": 77.56,
      "step": 34795,
      "token_acc": 0.44089456869009586,
      "train_speed(iter/s)": 1.440181
    },
    {
      "epoch": 1.490938691572769,
      "grad_norm": 5.892388820648193,
      "learning_rate": 7.961908743241016e-05,
      "loss": 2.557236671447754,
      "memory(GiB)": 77.56,
      "step": 34800,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.440163
    },
    {
      "epoch": 1.491152906902018,
      "grad_norm": 4.26841926574707,
      "learning_rate": 7.961366527260858e-05,
      "loss": 2.4700206756591796,
      "memory(GiB)": 77.56,
      "step": 34805,
      "token_acc": 0.41522491349480967,
      "train_speed(iter/s)": 1.440198
    },
    {
      "epoch": 1.4913671222312668,
      "grad_norm": 7.467026233673096,
      "learning_rate": 7.960824257632768e-05,
      "loss": 2.7604074478149414,
      "memory(GiB)": 77.56,
      "step": 34810,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.440161
    },
    {
      "epoch": 1.4915813375605158,
      "grad_norm": 4.229203701019287,
      "learning_rate": 7.960281934366568e-05,
      "loss": 2.5058860778808594,
      "memory(GiB)": 77.56,
      "step": 34815,
      "token_acc": 0.49673202614379086,
      "train_speed(iter/s)": 1.440148
    },
    {
      "epoch": 1.4917955528897648,
      "grad_norm": 5.706217288970947,
      "learning_rate": 7.959739557472085e-05,
      "loss": 2.591847038269043,
      "memory(GiB)": 77.56,
      "step": 34820,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.440175
    },
    {
      "epoch": 1.4920097682190137,
      "grad_norm": 4.805039405822754,
      "learning_rate": 7.959197126959142e-05,
      "loss": 2.5985092163085937,
      "memory(GiB)": 77.56,
      "step": 34825,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.440167
    },
    {
      "epoch": 1.4922239835482627,
      "grad_norm": 5.608213424682617,
      "learning_rate": 7.958654642837569e-05,
      "loss": 2.520285415649414,
      "memory(GiB)": 77.56,
      "step": 34830,
      "token_acc": 0.46779661016949153,
      "train_speed(iter/s)": 1.440204
    },
    {
      "epoch": 1.4924381988775117,
      "grad_norm": 4.211819171905518,
      "learning_rate": 7.95811210511719e-05,
      "loss": 2.406662940979004,
      "memory(GiB)": 77.56,
      "step": 34835,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.440226
    },
    {
      "epoch": 1.4926524142067605,
      "grad_norm": 5.518600940704346,
      "learning_rate": 7.957569513807836e-05,
      "loss": 2.479241943359375,
      "memory(GiB)": 77.56,
      "step": 34840,
      "token_acc": 0.4831804281345566,
      "train_speed(iter/s)": 1.440185
    },
    {
      "epoch": 1.4928666295360096,
      "grad_norm": 4.9143805503845215,
      "learning_rate": 7.957026868919334e-05,
      "loss": 2.5975696563720705,
      "memory(GiB)": 77.56,
      "step": 34845,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.440223
    },
    {
      "epoch": 1.4930808448652586,
      "grad_norm": 4.8514180183410645,
      "learning_rate": 7.95648417046152e-05,
      "loss": 2.597644805908203,
      "memory(GiB)": 77.56,
      "step": 34850,
      "token_acc": 0.4564102564102564,
      "train_speed(iter/s)": 1.44019
    },
    {
      "epoch": 1.4932950601945074,
      "grad_norm": 5.553345203399658,
      "learning_rate": 7.955941418444221e-05,
      "loss": 2.5125654220581053,
      "memory(GiB)": 77.56,
      "step": 34855,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.44022
    },
    {
      "epoch": 1.4935092755237565,
      "grad_norm": 5.789570331573486,
      "learning_rate": 7.955398612877269e-05,
      "loss": 2.646173095703125,
      "memory(GiB)": 77.56,
      "step": 34860,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.44023
    },
    {
      "epoch": 1.4937234908530055,
      "grad_norm": 5.425500392913818,
      "learning_rate": 7.954855753770499e-05,
      "loss": 2.269179916381836,
      "memory(GiB)": 77.56,
      "step": 34865,
      "token_acc": 0.496,
      "train_speed(iter/s)": 1.440224
    },
    {
      "epoch": 1.4939377061822543,
      "grad_norm": 4.963842391967773,
      "learning_rate": 7.954312841133744e-05,
      "loss": 2.6351694107055663,
      "memory(GiB)": 77.56,
      "step": 34870,
      "token_acc": 0.4659090909090909,
      "train_speed(iter/s)": 1.440215
    },
    {
      "epoch": 1.4941519215115033,
      "grad_norm": 5.984636306762695,
      "learning_rate": 7.953769874976842e-05,
      "loss": 2.2100454330444337,
      "memory(GiB)": 77.56,
      "step": 34875,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.440262
    },
    {
      "epoch": 1.4943661368407524,
      "grad_norm": 4.549845218658447,
      "learning_rate": 7.953226855309628e-05,
      "loss": 2.672037887573242,
      "memory(GiB)": 77.56,
      "step": 34880,
      "token_acc": 0.5305343511450382,
      "train_speed(iter/s)": 1.440258
    },
    {
      "epoch": 1.4945803521700012,
      "grad_norm": 5.547774791717529,
      "learning_rate": 7.952683782141939e-05,
      "loss": 2.5649818420410155,
      "memory(GiB)": 77.56,
      "step": 34885,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.440247
    },
    {
      "epoch": 1.4947945674992502,
      "grad_norm": 4.829455375671387,
      "learning_rate": 7.952140655483613e-05,
      "loss": 2.3834651947021483,
      "memory(GiB)": 77.56,
      "step": 34890,
      "token_acc": 0.5050847457627119,
      "train_speed(iter/s)": 1.440248
    },
    {
      "epoch": 1.4950087828284992,
      "grad_norm": 5.101635456085205,
      "learning_rate": 7.95159747534449e-05,
      "loss": 2.51883544921875,
      "memory(GiB)": 77.56,
      "step": 34895,
      "token_acc": 0.4506172839506173,
      "train_speed(iter/s)": 1.440311
    },
    {
      "epoch": 1.495222998157748,
      "grad_norm": 5.088320732116699,
      "learning_rate": 7.95105424173441e-05,
      "loss": 2.4319252014160155,
      "memory(GiB)": 77.56,
      "step": 34900,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.440337
    },
    {
      "epoch": 1.495437213486997,
      "grad_norm": 4.86480188369751,
      "learning_rate": 7.950510954663213e-05,
      "loss": 2.3498004913330077,
      "memory(GiB)": 77.56,
      "step": 34905,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.44032
    },
    {
      "epoch": 1.4956514288162461,
      "grad_norm": 5.9228129386901855,
      "learning_rate": 7.949967614140744e-05,
      "loss": 2.537355422973633,
      "memory(GiB)": 77.56,
      "step": 34910,
      "token_acc": 0.493006993006993,
      "train_speed(iter/s)": 1.440327
    },
    {
      "epoch": 1.495865644145495,
      "grad_norm": 4.44220495223999,
      "learning_rate": 7.949424220176843e-05,
      "loss": 2.6062253952026366,
      "memory(GiB)": 77.56,
      "step": 34915,
      "token_acc": 0.4612903225806452,
      "train_speed(iter/s)": 1.440337
    },
    {
      "epoch": 1.496079859474744,
      "grad_norm": 4.177424430847168,
      "learning_rate": 7.948880772781356e-05,
      "loss": 2.566104507446289,
      "memory(GiB)": 77.56,
      "step": 34920,
      "token_acc": 0.4265232974910394,
      "train_speed(iter/s)": 1.440374
    },
    {
      "epoch": 1.496294074803993,
      "grad_norm": 4.657687187194824,
      "learning_rate": 7.948337271964128e-05,
      "loss": 2.241402816772461,
      "memory(GiB)": 77.56,
      "step": 34925,
      "token_acc": 0.5167785234899329,
      "train_speed(iter/s)": 1.440378
    },
    {
      "epoch": 1.4965082901332418,
      "grad_norm": 4.103175163269043,
      "learning_rate": 7.947793717735003e-05,
      "loss": 2.112751579284668,
      "memory(GiB)": 77.56,
      "step": 34930,
      "token_acc": 0.5302325581395348,
      "train_speed(iter/s)": 1.440396
    },
    {
      "epoch": 1.4967225054624909,
      "grad_norm": 4.832241058349609,
      "learning_rate": 7.947250110103832e-05,
      "loss": 2.7344244003295897,
      "memory(GiB)": 77.56,
      "step": 34935,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.440402
    },
    {
      "epoch": 1.49693672079174,
      "grad_norm": 5.227599620819092,
      "learning_rate": 7.946706449080459e-05,
      "loss": 2.242207336425781,
      "memory(GiB)": 77.56,
      "step": 34940,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.440413
    },
    {
      "epoch": 1.4971509361209887,
      "grad_norm": 5.069202423095703,
      "learning_rate": 7.946162734674734e-05,
      "loss": 2.9288259506225587,
      "memory(GiB)": 77.56,
      "step": 34945,
      "token_acc": 0.41823899371069184,
      "train_speed(iter/s)": 1.440475
    },
    {
      "epoch": 1.4973651514502377,
      "grad_norm": 4.369980335235596,
      "learning_rate": 7.945618966896508e-05,
      "loss": 2.5908447265625,
      "memory(GiB)": 77.56,
      "step": 34950,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.440484
    },
    {
      "epoch": 1.4975793667794868,
      "grad_norm": 4.850968837738037,
      "learning_rate": 7.945075145755632e-05,
      "loss": 2.5278106689453126,
      "memory(GiB)": 77.56,
      "step": 34955,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.440506
    },
    {
      "epoch": 1.4977935821087356,
      "grad_norm": 5.642793655395508,
      "learning_rate": 7.944531271261955e-05,
      "loss": 2.642150115966797,
      "memory(GiB)": 77.56,
      "step": 34960,
      "token_acc": 0.42704626334519574,
      "train_speed(iter/s)": 1.440522
    },
    {
      "epoch": 1.4980077974379846,
      "grad_norm": 7.260000228881836,
      "learning_rate": 7.943987343425335e-05,
      "loss": 2.6825935363769533,
      "memory(GiB)": 77.56,
      "step": 34965,
      "token_acc": 0.49809885931558934,
      "train_speed(iter/s)": 1.440541
    },
    {
      "epoch": 1.4982220127672337,
      "grad_norm": 4.408665180206299,
      "learning_rate": 7.94344336225562e-05,
      "loss": 2.4678133010864256,
      "memory(GiB)": 77.56,
      "step": 34970,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.440556
    },
    {
      "epoch": 1.4984362280964825,
      "grad_norm": 4.447878360748291,
      "learning_rate": 7.942899327762668e-05,
      "loss": 2.5756969451904297,
      "memory(GiB)": 77.56,
      "step": 34975,
      "token_acc": 0.45604395604395603,
      "train_speed(iter/s)": 1.440545
    },
    {
      "epoch": 1.4986504434257315,
      "grad_norm": 3.8783018589019775,
      "learning_rate": 7.942355239956332e-05,
      "loss": 2.423862838745117,
      "memory(GiB)": 77.56,
      "step": 34980,
      "token_acc": 0.4983922829581994,
      "train_speed(iter/s)": 1.440563
    },
    {
      "epoch": 1.4988646587549805,
      "grad_norm": 3.93540358543396,
      "learning_rate": 7.941811098846472e-05,
      "loss": 2.3538373947143554,
      "memory(GiB)": 77.56,
      "step": 34985,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.440603
    },
    {
      "epoch": 1.4990788740842294,
      "grad_norm": 6.256972312927246,
      "learning_rate": 7.941266904442945e-05,
      "loss": 2.3601333618164064,
      "memory(GiB)": 77.56,
      "step": 34990,
      "token_acc": 0.49173553719008267,
      "train_speed(iter/s)": 1.440618
    },
    {
      "epoch": 1.4992930894134784,
      "grad_norm": 4.560388565063477,
      "learning_rate": 7.94072265675561e-05,
      "loss": 2.5219305038452147,
      "memory(GiB)": 77.56,
      "step": 34995,
      "token_acc": 0.4924924924924925,
      "train_speed(iter/s)": 1.440577
    },
    {
      "epoch": 1.4995073047427274,
      "grad_norm": 5.435134410858154,
      "learning_rate": 7.940178355794324e-05,
      "loss": 2.223490524291992,
      "memory(GiB)": 77.56,
      "step": 35000,
      "token_acc": 0.5220883534136547,
      "train_speed(iter/s)": 1.440476
    },
    {
      "epoch": 1.4995073047427274,
      "eval_loss": 1.9680900573730469,
      "eval_runtime": 14.4282,
      "eval_samples_per_second": 6.931,
      "eval_steps_per_second": 6.931,
      "eval_token_acc": 0.5168831168831168,
      "step": 35000
    },
    {
      "epoch": 1.4997215200719762,
      "grad_norm": 6.762289047241211,
      "learning_rate": 7.93963400156895e-05,
      "loss": 2.2263851165771484,
      "memory(GiB)": 77.56,
      "step": 35005,
      "token_acc": 0.5137963843958135,
      "train_speed(iter/s)": 1.439615
    },
    {
      "epoch": 1.4999357354012253,
      "grad_norm": 4.437307834625244,
      "learning_rate": 7.939089594089347e-05,
      "loss": 2.2932712554931642,
      "memory(GiB)": 77.56,
      "step": 35010,
      "token_acc": 0.5155807365439093,
      "train_speed(iter/s)": 1.439636
    },
    {
      "epoch": 1.5001499507304743,
      "grad_norm": 5.13985538482666,
      "learning_rate": 7.93854513336538e-05,
      "loss": 2.381512260437012,
      "memory(GiB)": 77.56,
      "step": 35015,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439668
    },
    {
      "epoch": 1.5003641660597231,
      "grad_norm": 5.981245517730713,
      "learning_rate": 7.93800061940691e-05,
      "loss": 2.6373035430908205,
      "memory(GiB)": 77.56,
      "step": 35020,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.439734
    },
    {
      "epoch": 1.5005783813889721,
      "grad_norm": 4.933376789093018,
      "learning_rate": 7.937456052223804e-05,
      "loss": 2.458828353881836,
      "memory(GiB)": 77.56,
      "step": 35025,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.439745
    },
    {
      "epoch": 1.5007925967182212,
      "grad_norm": 5.496727466583252,
      "learning_rate": 7.936911431825926e-05,
      "loss": 2.8117338180541993,
      "memory(GiB)": 77.56,
      "step": 35030,
      "token_acc": 0.4272151898734177,
      "train_speed(iter/s)": 1.439765
    },
    {
      "epoch": 1.50100681204747,
      "grad_norm": 4.272097110748291,
      "learning_rate": 7.936366758223142e-05,
      "loss": 2.379751777648926,
      "memory(GiB)": 77.56,
      "step": 35035,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.439743
    },
    {
      "epoch": 1.501221027376719,
      "grad_norm": 4.517671585083008,
      "learning_rate": 7.935822031425319e-05,
      "loss": 2.4862644195556642,
      "memory(GiB)": 77.56,
      "step": 35040,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.439798
    },
    {
      "epoch": 1.501435242705968,
      "grad_norm": 4.858052730560303,
      "learning_rate": 7.935277251442326e-05,
      "loss": 2.133636474609375,
      "memory(GiB)": 77.56,
      "step": 35045,
      "token_acc": 0.5275862068965518,
      "train_speed(iter/s)": 1.439829
    },
    {
      "epoch": 1.5016494580352169,
      "grad_norm": 4.567692279815674,
      "learning_rate": 7.934732418284035e-05,
      "loss": 2.324208068847656,
      "memory(GiB)": 77.56,
      "step": 35050,
      "token_acc": 0.528,
      "train_speed(iter/s)": 1.439877
    },
    {
      "epoch": 1.501863673364466,
      "grad_norm": 5.690674304962158,
      "learning_rate": 7.93418753196031e-05,
      "loss": 2.8742040634155273,
      "memory(GiB)": 77.56,
      "step": 35055,
      "token_acc": 0.3952702702702703,
      "train_speed(iter/s)": 1.439897
    },
    {
      "epoch": 1.502077888693715,
      "grad_norm": 5.509589195251465,
      "learning_rate": 7.933642592481026e-05,
      "loss": 2.378958892822266,
      "memory(GiB)": 77.56,
      "step": 35060,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.439867
    },
    {
      "epoch": 1.5022921040229638,
      "grad_norm": 4.702286720275879,
      "learning_rate": 7.933097599856054e-05,
      "loss": 2.2223737716674803,
      "memory(GiB)": 77.56,
      "step": 35065,
      "token_acc": 0.516260162601626,
      "train_speed(iter/s)": 1.439901
    },
    {
      "epoch": 1.5025063193522128,
      "grad_norm": 5.72991418838501,
      "learning_rate": 7.932552554095268e-05,
      "loss": 2.482608413696289,
      "memory(GiB)": 77.56,
      "step": 35070,
      "token_acc": 0.47844827586206895,
      "train_speed(iter/s)": 1.439903
    },
    {
      "epoch": 1.5027205346814618,
      "grad_norm": 7.016448497772217,
      "learning_rate": 7.932007455208542e-05,
      "loss": 2.734857749938965,
      "memory(GiB)": 77.56,
      "step": 35075,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.439913
    },
    {
      "epoch": 1.5029347500107106,
      "grad_norm": 6.139796733856201,
      "learning_rate": 7.93146230320575e-05,
      "loss": 2.5799448013305666,
      "memory(GiB)": 77.56,
      "step": 35080,
      "token_acc": 0.4962962962962963,
      "train_speed(iter/s)": 1.43996
    },
    {
      "epoch": 1.5031489653399597,
      "grad_norm": 9.221030235290527,
      "learning_rate": 7.93091709809677e-05,
      "loss": 2.4534591674804687,
      "memory(GiB)": 77.56,
      "step": 35085,
      "token_acc": 0.47572815533980584,
      "train_speed(iter/s)": 1.440016
    },
    {
      "epoch": 1.5033631806692087,
      "grad_norm": 5.759133338928223,
      "learning_rate": 7.930371839891475e-05,
      "loss": 2.545137405395508,
      "memory(GiB)": 77.56,
      "step": 35090,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.440015
    },
    {
      "epoch": 1.5035773959984575,
      "grad_norm": 7.849509239196777,
      "learning_rate": 7.929826528599746e-05,
      "loss": 2.778175926208496,
      "memory(GiB)": 77.56,
      "step": 35095,
      "token_acc": 0.45774647887323944,
      "train_speed(iter/s)": 1.440003
    },
    {
      "epoch": 1.5037916113277066,
      "grad_norm": 5.452881336212158,
      "learning_rate": 7.929281164231461e-05,
      "loss": 2.6373451232910154,
      "memory(GiB)": 77.56,
      "step": 35100,
      "token_acc": 0.44571428571428573,
      "train_speed(iter/s)": 1.440053
    },
    {
      "epoch": 1.5040058266569556,
      "grad_norm": 4.941193580627441,
      "learning_rate": 7.928735746796501e-05,
      "loss": 2.6310327529907225,
      "memory(GiB)": 77.56,
      "step": 35105,
      "token_acc": 0.4217252396166134,
      "train_speed(iter/s)": 1.440096
    },
    {
      "epoch": 1.5042200419862044,
      "grad_norm": 5.028214931488037,
      "learning_rate": 7.928190276304744e-05,
      "loss": 2.471433639526367,
      "memory(GiB)": 77.56,
      "step": 35110,
      "token_acc": 0.4596774193548387,
      "train_speed(iter/s)": 1.440027
    },
    {
      "epoch": 1.5044342573154537,
      "grad_norm": 5.106569290161133,
      "learning_rate": 7.927644752766074e-05,
      "loss": 2.7555688858032226,
      "memory(GiB)": 77.56,
      "step": 35115,
      "token_acc": 0.46757679180887374,
      "train_speed(iter/s)": 1.44003
    },
    {
      "epoch": 1.5046484726447025,
      "grad_norm": 7.169212818145752,
      "learning_rate": 7.927099176190374e-05,
      "loss": 2.8109527587890626,
      "memory(GiB)": 77.56,
      "step": 35120,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.440062
    },
    {
      "epoch": 1.5048626879739513,
      "grad_norm": 5.62183952331543,
      "learning_rate": 7.926553546587525e-05,
      "loss": 2.6057254791259767,
      "memory(GiB)": 77.56,
      "step": 35125,
      "token_acc": 0.4901315789473684,
      "train_speed(iter/s)": 1.440087
    },
    {
      "epoch": 1.5050769033032005,
      "grad_norm": 5.260104179382324,
      "learning_rate": 7.926007863967413e-05,
      "loss": 2.4580989837646485,
      "memory(GiB)": 77.56,
      "step": 35130,
      "token_acc": 0.4763779527559055,
      "train_speed(iter/s)": 1.440145
    },
    {
      "epoch": 1.5052911186324494,
      "grad_norm": 4.021641731262207,
      "learning_rate": 7.925462128339925e-05,
      "loss": 2.5604425430297852,
      "memory(GiB)": 77.56,
      "step": 35135,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.44018
    },
    {
      "epoch": 1.5055053339616982,
      "grad_norm": 6.118619441986084,
      "learning_rate": 7.924916339714945e-05,
      "loss": 2.6336822509765625,
      "memory(GiB)": 77.56,
      "step": 35140,
      "token_acc": 0.4740740740740741,
      "train_speed(iter/s)": 1.440221
    },
    {
      "epoch": 1.5057195492909474,
      "grad_norm": 4.927286624908447,
      "learning_rate": 7.924370498102363e-05,
      "loss": 2.7585710525512694,
      "memory(GiB)": 77.56,
      "step": 35145,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.440217
    },
    {
      "epoch": 1.5059337646201962,
      "grad_norm": 4.069596290588379,
      "learning_rate": 7.923824603512065e-05,
      "loss": 2.2800817489624023,
      "memory(GiB)": 77.56,
      "step": 35150,
      "token_acc": 0.5092936802973977,
      "train_speed(iter/s)": 1.440217
    },
    {
      "epoch": 1.506147979949445,
      "grad_norm": 3.937621593475342,
      "learning_rate": 7.923278655953943e-05,
      "loss": 2.326829528808594,
      "memory(GiB)": 77.56,
      "step": 35155,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.440218
    },
    {
      "epoch": 1.5063621952786943,
      "grad_norm": 7.246119022369385,
      "learning_rate": 7.922732655437884e-05,
      "loss": 2.4844532012939453,
      "memory(GiB)": 77.56,
      "step": 35160,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.440163
    },
    {
      "epoch": 1.5065764106079431,
      "grad_norm": 5.87959098815918,
      "learning_rate": 7.922186601973782e-05,
      "loss": 2.432872772216797,
      "memory(GiB)": 77.56,
      "step": 35165,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.44019
    },
    {
      "epoch": 1.506790625937192,
      "grad_norm": 4.126819610595703,
      "learning_rate": 7.921640495571529e-05,
      "loss": 2.5085987091064452,
      "memory(GiB)": 77.56,
      "step": 35170,
      "token_acc": 0.4537037037037037,
      "train_speed(iter/s)": 1.440222
    },
    {
      "epoch": 1.5070048412664412,
      "grad_norm": 4.535497188568115,
      "learning_rate": 7.921094336241017e-05,
      "loss": 2.5078815460205077,
      "memory(GiB)": 77.56,
      "step": 35175,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440221
    },
    {
      "epoch": 1.50721905659569,
      "grad_norm": 5.656325340270996,
      "learning_rate": 7.920548123992142e-05,
      "loss": 2.6849136352539062,
      "memory(GiB)": 77.56,
      "step": 35180,
      "token_acc": 0.4111111111111111,
      "train_speed(iter/s)": 1.44019
    },
    {
      "epoch": 1.5074332719249388,
      "grad_norm": 4.733944892883301,
      "learning_rate": 7.920001858834796e-05,
      "loss": 2.6267629623413087,
      "memory(GiB)": 77.56,
      "step": 35185,
      "token_acc": 0.4803921568627451,
      "train_speed(iter/s)": 1.440232
    },
    {
      "epoch": 1.507647487254188,
      "grad_norm": 4.708016395568848,
      "learning_rate": 7.91945554077888e-05,
      "loss": 2.660777473449707,
      "memory(GiB)": 77.56,
      "step": 35190,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.44021
    },
    {
      "epoch": 1.5078617025834369,
      "grad_norm": 6.860857963562012,
      "learning_rate": 7.918909169834285e-05,
      "loss": 2.7101951599121095,
      "memory(GiB)": 77.56,
      "step": 35195,
      "token_acc": 0.4359861591695502,
      "train_speed(iter/s)": 1.440166
    },
    {
      "epoch": 1.5080759179126857,
      "grad_norm": 4.5236077308654785,
      "learning_rate": 7.918362746010914e-05,
      "loss": 2.3776504516601564,
      "memory(GiB)": 77.56,
      "step": 35200,
      "token_acc": 0.4273743016759777,
      "train_speed(iter/s)": 1.440204
    },
    {
      "epoch": 1.508290133241935,
      "grad_norm": 4.508399486541748,
      "learning_rate": 7.917816269318665e-05,
      "loss": 2.7005031585693358,
      "memory(GiB)": 77.56,
      "step": 35205,
      "token_acc": 0.4296875,
      "train_speed(iter/s)": 1.440252
    },
    {
      "epoch": 1.5085043485711838,
      "grad_norm": 5.647885322570801,
      "learning_rate": 7.917269739767434e-05,
      "loss": 2.3213151931762694,
      "memory(GiB)": 77.56,
      "step": 35210,
      "token_acc": 0.4452296819787986,
      "train_speed(iter/s)": 1.440213
    },
    {
      "epoch": 1.5087185639004326,
      "grad_norm": 4.938968181610107,
      "learning_rate": 7.916723157367129e-05,
      "loss": 2.5326332092285155,
      "memory(GiB)": 77.56,
      "step": 35215,
      "token_acc": 0.4588235294117647,
      "train_speed(iter/s)": 1.440186
    },
    {
      "epoch": 1.5089327792296818,
      "grad_norm": 6.525058269500732,
      "learning_rate": 7.916176522127645e-05,
      "loss": 2.8699703216552734,
      "memory(GiB)": 77.56,
      "step": 35220,
      "token_acc": 0.4603658536585366,
      "train_speed(iter/s)": 1.440176
    },
    {
      "epoch": 1.5091469945589306,
      "grad_norm": 4.628689289093018,
      "learning_rate": 7.91562983405889e-05,
      "loss": 2.640371322631836,
      "memory(GiB)": 77.56,
      "step": 35225,
      "token_acc": 0.47808764940239046,
      "train_speed(iter/s)": 1.440213
    },
    {
      "epoch": 1.5093612098881795,
      "grad_norm": 5.026602745056152,
      "learning_rate": 7.915083093170764e-05,
      "loss": 2.506536102294922,
      "memory(GiB)": 77.56,
      "step": 35230,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.440211
    },
    {
      "epoch": 1.5095754252174287,
      "grad_norm": 5.8378682136535645,
      "learning_rate": 7.914536299473173e-05,
      "loss": 2.594832420349121,
      "memory(GiB)": 77.56,
      "step": 35235,
      "token_acc": 0.5020242914979757,
      "train_speed(iter/s)": 1.440212
    },
    {
      "epoch": 1.5097896405466775,
      "grad_norm": 6.31693696975708,
      "learning_rate": 7.913989452976023e-05,
      "loss": 2.743829345703125,
      "memory(GiB)": 77.56,
      "step": 35240,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.440226
    },
    {
      "epoch": 1.5100038558759263,
      "grad_norm": 6.887179851531982,
      "learning_rate": 7.913442553689221e-05,
      "loss": 2.4985849380493166,
      "memory(GiB)": 77.56,
      "step": 35245,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.440242
    },
    {
      "epoch": 1.5102180712051756,
      "grad_norm": 5.0936479568481445,
      "learning_rate": 7.912895601622675e-05,
      "loss": 2.3697273254394533,
      "memory(GiB)": 77.56,
      "step": 35250,
      "token_acc": 0.47183098591549294,
      "train_speed(iter/s)": 1.440247
    },
    {
      "epoch": 1.5104322865344244,
      "grad_norm": 4.387964725494385,
      "learning_rate": 7.912348596786291e-05,
      "loss": 2.3688579559326173,
      "memory(GiB)": 77.56,
      "step": 35255,
      "token_acc": 0.4927007299270073,
      "train_speed(iter/s)": 1.440239
    },
    {
      "epoch": 1.5106465018636732,
      "grad_norm": 5.024767875671387,
      "learning_rate": 7.91180153918998e-05,
      "loss": 2.3366846084594726,
      "memory(GiB)": 77.56,
      "step": 35260,
      "token_acc": 0.5060606060606061,
      "train_speed(iter/s)": 1.440245
    },
    {
      "epoch": 1.5108607171929225,
      "grad_norm": 6.093329906463623,
      "learning_rate": 7.911254428843654e-05,
      "loss": 2.6157827377319336,
      "memory(GiB)": 77.56,
      "step": 35265,
      "token_acc": 0.45791245791245794,
      "train_speed(iter/s)": 1.440281
    },
    {
      "epoch": 1.5110749325221713,
      "grad_norm": 5.016539573669434,
      "learning_rate": 7.910707265757222e-05,
      "loss": 2.4375516891479494,
      "memory(GiB)": 77.56,
      "step": 35270,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440228
    },
    {
      "epoch": 1.51128914785142,
      "grad_norm": 4.038543224334717,
      "learning_rate": 7.910160049940598e-05,
      "loss": 2.3681718826293947,
      "memory(GiB)": 77.56,
      "step": 35275,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.440259
    },
    {
      "epoch": 1.5115033631806694,
      "grad_norm": 5.901710033416748,
      "learning_rate": 7.909612781403695e-05,
      "loss": 2.5667343139648438,
      "memory(GiB)": 77.56,
      "step": 35280,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.440284
    },
    {
      "epoch": 1.5117175785099182,
      "grad_norm": 4.946591377258301,
      "learning_rate": 7.909065460156427e-05,
      "loss": 2.403861427307129,
      "memory(GiB)": 77.56,
      "step": 35285,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.440205
    },
    {
      "epoch": 1.511931793839167,
      "grad_norm": 4.404177665710449,
      "learning_rate": 7.90851808620871e-05,
      "loss": 2.2176586151123048,
      "memory(GiB)": 77.56,
      "step": 35290,
      "token_acc": 0.525096525096525,
      "train_speed(iter/s)": 1.440216
    },
    {
      "epoch": 1.5121460091684162,
      "grad_norm": 5.203978061676025,
      "learning_rate": 7.907970659570457e-05,
      "loss": 2.570529556274414,
      "memory(GiB)": 77.56,
      "step": 35295,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.440203
    },
    {
      "epoch": 1.512360224497665,
      "grad_norm": 4.523471832275391,
      "learning_rate": 7.90742318025159e-05,
      "loss": 2.179259490966797,
      "memory(GiB)": 77.56,
      "step": 35300,
      "token_acc": 0.5210727969348659,
      "train_speed(iter/s)": 1.440208
    },
    {
      "epoch": 1.5125744398269139,
      "grad_norm": 5.4604268074035645,
      "learning_rate": 7.906875648262022e-05,
      "loss": 2.560937690734863,
      "memory(GiB)": 77.56,
      "step": 35305,
      "token_acc": 0.4668769716088328,
      "train_speed(iter/s)": 1.440171
    },
    {
      "epoch": 1.5127886551561631,
      "grad_norm": 4.289139270782471,
      "learning_rate": 7.906328063611677e-05,
      "loss": 2.7063177108764647,
      "memory(GiB)": 77.56,
      "step": 35310,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440144
    },
    {
      "epoch": 1.513002870485412,
      "grad_norm": 4.89067268371582,
      "learning_rate": 7.905780426310472e-05,
      "loss": 2.5784698486328126,
      "memory(GiB)": 77.56,
      "step": 35315,
      "token_acc": 0.4555160142348754,
      "train_speed(iter/s)": 1.440125
    },
    {
      "epoch": 1.5132170858146607,
      "grad_norm": 6.2356367111206055,
      "learning_rate": 7.905232736368328e-05,
      "loss": 2.7093389511108397,
      "memory(GiB)": 77.56,
      "step": 35320,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.440081
    },
    {
      "epoch": 1.51343130114391,
      "grad_norm": 6.2283549308776855,
      "learning_rate": 7.90468499379517e-05,
      "loss": 2.101117706298828,
      "memory(GiB)": 77.56,
      "step": 35325,
      "token_acc": 0.52734375,
      "train_speed(iter/s)": 1.440056
    },
    {
      "epoch": 1.5136455164731588,
      "grad_norm": 6.189523220062256,
      "learning_rate": 7.904137198600917e-05,
      "loss": 2.468982124328613,
      "memory(GiB)": 77.56,
      "step": 35330,
      "token_acc": 0.45695364238410596,
      "train_speed(iter/s)": 1.440079
    },
    {
      "epoch": 1.5138597318024076,
      "grad_norm": 4.029015064239502,
      "learning_rate": 7.903589350795495e-05,
      "loss": 2.6178512573242188,
      "memory(GiB)": 77.56,
      "step": 35335,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.440105
    },
    {
      "epoch": 1.5140739471316569,
      "grad_norm": 4.273472309112549,
      "learning_rate": 7.903041450388828e-05,
      "loss": 2.716708946228027,
      "memory(GiB)": 77.56,
      "step": 35340,
      "token_acc": 0.45544554455445546,
      "train_speed(iter/s)": 1.440098
    },
    {
      "epoch": 1.5142881624609057,
      "grad_norm": 4.4855499267578125,
      "learning_rate": 7.902493497390842e-05,
      "loss": 2.6940485000610352,
      "memory(GiB)": 77.56,
      "step": 35345,
      "token_acc": 0.4495677233429395,
      "train_speed(iter/s)": 1.440123
    },
    {
      "epoch": 1.5145023777901545,
      "grad_norm": 7.15872049331665,
      "learning_rate": 7.901945491811462e-05,
      "loss": 2.457971954345703,
      "memory(GiB)": 77.56,
      "step": 35350,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.440157
    },
    {
      "epoch": 1.5147165931194038,
      "grad_norm": 4.4815850257873535,
      "learning_rate": 7.901397433660617e-05,
      "loss": 2.6729719161987306,
      "memory(GiB)": 77.56,
      "step": 35355,
      "token_acc": 0.4357142857142857,
      "train_speed(iter/s)": 1.440152
    },
    {
      "epoch": 1.5149308084486526,
      "grad_norm": 5.4969401359558105,
      "learning_rate": 7.900849322948239e-05,
      "loss": 2.6149808883666994,
      "memory(GiB)": 77.56,
      "step": 35360,
      "token_acc": 0.42574257425742573,
      "train_speed(iter/s)": 1.440162
    },
    {
      "epoch": 1.5151450237779014,
      "grad_norm": 3.847578287124634,
      "learning_rate": 7.900301159684251e-05,
      "loss": 2.53448600769043,
      "memory(GiB)": 77.56,
      "step": 35365,
      "token_acc": 0.4608695652173913,
      "train_speed(iter/s)": 1.440207
    },
    {
      "epoch": 1.5153592391071506,
      "grad_norm": 5.528143882751465,
      "learning_rate": 7.899752943878589e-05,
      "loss": 2.507093048095703,
      "memory(GiB)": 77.56,
      "step": 35370,
      "token_acc": 0.5084175084175084,
      "train_speed(iter/s)": 1.440203
    },
    {
      "epoch": 1.5155734544363995,
      "grad_norm": 5.079574108123779,
      "learning_rate": 7.89920467554118e-05,
      "loss": 2.6101160049438477,
      "memory(GiB)": 77.56,
      "step": 35375,
      "token_acc": 0.4647058823529412,
      "train_speed(iter/s)": 1.440212
    },
    {
      "epoch": 1.5157876697656483,
      "grad_norm": 8.060430526733398,
      "learning_rate": 7.898656354681961e-05,
      "loss": 2.6383169174194334,
      "memory(GiB)": 77.56,
      "step": 35380,
      "token_acc": 0.4646017699115044,
      "train_speed(iter/s)": 1.440223
    },
    {
      "epoch": 1.5160018850948975,
      "grad_norm": 5.734468936920166,
      "learning_rate": 7.89810798131086e-05,
      "loss": 2.4155319213867186,
      "memory(GiB)": 77.56,
      "step": 35385,
      "token_acc": 0.4681647940074906,
      "train_speed(iter/s)": 1.440259
    },
    {
      "epoch": 1.5162161004241463,
      "grad_norm": 4.224002838134766,
      "learning_rate": 7.897559555437817e-05,
      "loss": 2.6890220642089844,
      "memory(GiB)": 77.56,
      "step": 35390,
      "token_acc": 0.46567164179104475,
      "train_speed(iter/s)": 1.440261
    },
    {
      "epoch": 1.5164303157533952,
      "grad_norm": 4.405418872833252,
      "learning_rate": 7.897011077072763e-05,
      "loss": 2.3410049438476563,
      "memory(GiB)": 77.56,
      "step": 35395,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.440256
    },
    {
      "epoch": 1.5166445310826444,
      "grad_norm": 5.095930099487305,
      "learning_rate": 7.896462546225637e-05,
      "loss": 2.3978046417236327,
      "memory(GiB)": 77.56,
      "step": 35400,
      "token_acc": 0.4628099173553719,
      "train_speed(iter/s)": 1.440244
    },
    {
      "epoch": 1.5168587464118932,
      "grad_norm": 3.97068190574646,
      "learning_rate": 7.895913962906374e-05,
      "loss": 2.5652626037597654,
      "memory(GiB)": 77.56,
      "step": 35405,
      "token_acc": 0.43558282208588955,
      "train_speed(iter/s)": 1.440243
    },
    {
      "epoch": 1.517072961741142,
      "grad_norm": 4.979217052459717,
      "learning_rate": 7.895365327124912e-05,
      "loss": 2.497342109680176,
      "memory(GiB)": 77.56,
      "step": 35410,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.440209
    },
    {
      "epoch": 1.5172871770703913,
      "grad_norm": 4.717621803283691,
      "learning_rate": 7.894816638891194e-05,
      "loss": 2.6755285263061523,
      "memory(GiB)": 77.56,
      "step": 35415,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.440165
    },
    {
      "epoch": 1.51750139239964,
      "grad_norm": 5.189795970916748,
      "learning_rate": 7.894267898215155e-05,
      "loss": 2.686911392211914,
      "memory(GiB)": 77.56,
      "step": 35420,
      "token_acc": 0.43304843304843305,
      "train_speed(iter/s)": 1.440204
    },
    {
      "epoch": 1.517715607728889,
      "grad_norm": 5.874298095703125,
      "learning_rate": 7.893719105106739e-05,
      "loss": 2.5686975479125977,
      "memory(GiB)": 77.56,
      "step": 35425,
      "token_acc": 0.4783950617283951,
      "train_speed(iter/s)": 1.440185
    },
    {
      "epoch": 1.5179298230581382,
      "grad_norm": 4.797722339630127,
      "learning_rate": 7.893170259575886e-05,
      "loss": 2.5290069580078125,
      "memory(GiB)": 77.56,
      "step": 35430,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.440168
    },
    {
      "epoch": 1.518144038387387,
      "grad_norm": 6.44780969619751,
      "learning_rate": 7.892621361632539e-05,
      "loss": 2.531484031677246,
      "memory(GiB)": 77.56,
      "step": 35435,
      "token_acc": 0.4590643274853801,
      "train_speed(iter/s)": 1.440167
    },
    {
      "epoch": 1.5183582537166358,
      "grad_norm": 4.254237174987793,
      "learning_rate": 7.892072411286644e-05,
      "loss": 2.2918649673461915,
      "memory(GiB)": 77.56,
      "step": 35440,
      "token_acc": 0.5230125523012552,
      "train_speed(iter/s)": 1.440174
    },
    {
      "epoch": 1.518572469045885,
      "grad_norm": 5.859513759613037,
      "learning_rate": 7.891523408548143e-05,
      "loss": 2.5549551010131837,
      "memory(GiB)": 77.56,
      "step": 35445,
      "token_acc": 0.4555160142348754,
      "train_speed(iter/s)": 1.440191
    },
    {
      "epoch": 1.5187866843751339,
      "grad_norm": 5.328099250793457,
      "learning_rate": 7.890974353426984e-05,
      "loss": 2.576326370239258,
      "memory(GiB)": 77.56,
      "step": 35450,
      "token_acc": 0.4591439688715953,
      "train_speed(iter/s)": 1.44022
    },
    {
      "epoch": 1.519000899704383,
      "grad_norm": 5.957167625427246,
      "learning_rate": 7.890425245933113e-05,
      "loss": 2.236961555480957,
      "memory(GiB)": 77.56,
      "step": 35455,
      "token_acc": 0.5404411764705882,
      "train_speed(iter/s)": 1.440247
    },
    {
      "epoch": 1.519215115033632,
      "grad_norm": 5.930410385131836,
      "learning_rate": 7.889876086076475e-05,
      "loss": 2.690696907043457,
      "memory(GiB)": 77.56,
      "step": 35460,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.440215
    },
    {
      "epoch": 1.5194293303628807,
      "grad_norm": 4.081855297088623,
      "learning_rate": 7.889326873867022e-05,
      "loss": 2.475139617919922,
      "memory(GiB)": 77.56,
      "step": 35465,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.440259
    },
    {
      "epoch": 1.5196435456921298,
      "grad_norm": 5.731164455413818,
      "learning_rate": 7.888777609314704e-05,
      "loss": 2.626582145690918,
      "memory(GiB)": 77.56,
      "step": 35470,
      "token_acc": 0.4504792332268371,
      "train_speed(iter/s)": 1.440251
    },
    {
      "epoch": 1.5198577610213788,
      "grad_norm": 4.737641334533691,
      "learning_rate": 7.888228292429468e-05,
      "loss": 2.4544776916503905,
      "memory(GiB)": 77.56,
      "step": 35475,
      "token_acc": 0.4691358024691358,
      "train_speed(iter/s)": 1.440265
    },
    {
      "epoch": 1.5200719763506276,
      "grad_norm": 5.118046760559082,
      "learning_rate": 7.887678923221267e-05,
      "loss": 2.8044885635375976,
      "memory(GiB)": 77.56,
      "step": 35480,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.440322
    },
    {
      "epoch": 1.5202861916798767,
      "grad_norm": 4.17631721496582,
      "learning_rate": 7.887129501700055e-05,
      "loss": 2.554097557067871,
      "memory(GiB)": 77.56,
      "step": 35485,
      "token_acc": 0.4574780058651026,
      "train_speed(iter/s)": 1.440339
    },
    {
      "epoch": 1.5205004070091257,
      "grad_norm": 4.571279525756836,
      "learning_rate": 7.886580027875782e-05,
      "loss": 2.3724308013916016,
      "memory(GiB)": 77.56,
      "step": 35490,
      "token_acc": 0.5062761506276151,
      "train_speed(iter/s)": 1.440294
    },
    {
      "epoch": 1.5207146223383745,
      "grad_norm": 5.154933929443359,
      "learning_rate": 7.886030501758404e-05,
      "loss": 2.3690290451049805,
      "memory(GiB)": 77.56,
      "step": 35495,
      "token_acc": 0.5052631578947369,
      "train_speed(iter/s)": 1.440298
    },
    {
      "epoch": 1.5209288376676235,
      "grad_norm": 5.929644584655762,
      "learning_rate": 7.885480923357878e-05,
      "loss": 2.553443717956543,
      "memory(GiB)": 77.56,
      "step": 35500,
      "token_acc": 0.46865671641791046,
      "train_speed(iter/s)": 1.440292
    },
    {
      "epoch": 1.5209288376676235,
      "eval_loss": 2.2393977642059326,
      "eval_runtime": 14.1273,
      "eval_samples_per_second": 7.079,
      "eval_steps_per_second": 7.079,
      "eval_token_acc": 0.47717231222385864,
      "step": 35500
    },
    {
      "epoch": 1.5211430529968726,
      "grad_norm": 4.536977291107178,
      "learning_rate": 7.884931292684157e-05,
      "loss": 2.4208919525146486,
      "memory(GiB)": 77.56,
      "step": 35505,
      "token_acc": 0.4687179487179487,
      "train_speed(iter/s)": 1.439368
    },
    {
      "epoch": 1.5213572683261214,
      "grad_norm": 4.821561336517334,
      "learning_rate": 7.884381609747198e-05,
      "loss": 2.2888898849487305,
      "memory(GiB)": 77.56,
      "step": 35510,
      "token_acc": 0.4803921568627451,
      "train_speed(iter/s)": 1.439342
    },
    {
      "epoch": 1.5215714836553704,
      "grad_norm": 7.081754684448242,
      "learning_rate": 7.883831874556962e-05,
      "loss": 2.5779401779174806,
      "memory(GiB)": 77.56,
      "step": 35515,
      "token_acc": 0.44765342960288806,
      "train_speed(iter/s)": 1.439322
    },
    {
      "epoch": 1.5217856989846195,
      "grad_norm": 6.1978044509887695,
      "learning_rate": 7.883282087123407e-05,
      "loss": 2.5102333068847655,
      "memory(GiB)": 77.56,
      "step": 35520,
      "token_acc": 0.5019607843137255,
      "train_speed(iter/s)": 1.43935
    },
    {
      "epoch": 1.5219999143138683,
      "grad_norm": 4.620543003082275,
      "learning_rate": 7.88273224745649e-05,
      "loss": 2.7689218521118164,
      "memory(GiB)": 77.56,
      "step": 35525,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.439354
    },
    {
      "epoch": 1.5222141296431173,
      "grad_norm": 5.6832780838012695,
      "learning_rate": 7.882182355566177e-05,
      "loss": 2.5620773315429686,
      "memory(GiB)": 77.56,
      "step": 35530,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.439371
    },
    {
      "epoch": 1.5224283449723663,
      "grad_norm": 5.037816047668457,
      "learning_rate": 7.881632411462424e-05,
      "loss": 2.644948387145996,
      "memory(GiB)": 77.56,
      "step": 35535,
      "token_acc": 0.4405594405594406,
      "train_speed(iter/s)": 1.43939
    },
    {
      "epoch": 1.5226425603016152,
      "grad_norm": 5.584239482879639,
      "learning_rate": 7.881082415155198e-05,
      "loss": 2.460175132751465,
      "memory(GiB)": 77.56,
      "step": 35540,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.439385
    },
    {
      "epoch": 1.5228567756308642,
      "grad_norm": 4.644760608673096,
      "learning_rate": 7.880532366654462e-05,
      "loss": 2.578045463562012,
      "memory(GiB)": 77.56,
      "step": 35545,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.439428
    },
    {
      "epoch": 1.5230709909601132,
      "grad_norm": 4.966503620147705,
      "learning_rate": 7.879982265970178e-05,
      "loss": 2.2969470977783204,
      "memory(GiB)": 77.56,
      "step": 35550,
      "token_acc": 0.48398576512455516,
      "train_speed(iter/s)": 1.439469
    },
    {
      "epoch": 1.523285206289362,
      "grad_norm": 5.454427719116211,
      "learning_rate": 7.879432113112316e-05,
      "loss": 2.5244773864746093,
      "memory(GiB)": 77.56,
      "step": 35555,
      "token_acc": 0.48120300751879697,
      "train_speed(iter/s)": 1.439459
    },
    {
      "epoch": 1.523499421618611,
      "grad_norm": 5.873898029327393,
      "learning_rate": 7.87888190809084e-05,
      "loss": 2.512242317199707,
      "memory(GiB)": 77.56,
      "step": 35560,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.439463
    },
    {
      "epoch": 1.52371363694786,
      "grad_norm": 4.819453239440918,
      "learning_rate": 7.878331650915716e-05,
      "loss": 2.350246810913086,
      "memory(GiB)": 77.56,
      "step": 35565,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.439451
    },
    {
      "epoch": 1.523927852277109,
      "grad_norm": 4.41420841217041,
      "learning_rate": 7.877781341596915e-05,
      "loss": 2.604685401916504,
      "memory(GiB)": 77.56,
      "step": 35570,
      "token_acc": 0.434375,
      "train_speed(iter/s)": 1.439494
    },
    {
      "epoch": 1.524142067606358,
      "grad_norm": 5.763274669647217,
      "learning_rate": 7.877230980144404e-05,
      "loss": 2.310897636413574,
      "memory(GiB)": 77.56,
      "step": 35575,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.439488
    },
    {
      "epoch": 1.524356282935607,
      "grad_norm": 5.697576522827148,
      "learning_rate": 7.876680566568157e-05,
      "loss": 2.5009170532226563,
      "memory(GiB)": 77.56,
      "step": 35580,
      "token_acc": 0.49201277955271566,
      "train_speed(iter/s)": 1.439464
    },
    {
      "epoch": 1.5245704982648558,
      "grad_norm": 7.941682815551758,
      "learning_rate": 7.876130100878142e-05,
      "loss": 2.696417808532715,
      "memory(GiB)": 77.56,
      "step": 35585,
      "token_acc": 0.43656716417910446,
      "train_speed(iter/s)": 1.439488
    },
    {
      "epoch": 1.5247847135941048,
      "grad_norm": 6.363534927368164,
      "learning_rate": 7.875579583084331e-05,
      "loss": 2.6718107223510743,
      "memory(GiB)": 77.56,
      "step": 35590,
      "token_acc": 0.46204620462046203,
      "train_speed(iter/s)": 1.439528
    },
    {
      "epoch": 1.5249989289233539,
      "grad_norm": 5.607141017913818,
      "learning_rate": 7.8750290131967e-05,
      "loss": 3.0673961639404297,
      "memory(GiB)": 77.56,
      "step": 35595,
      "token_acc": 0.4163934426229508,
      "train_speed(iter/s)": 1.439542
    },
    {
      "epoch": 1.5252131442526027,
      "grad_norm": 6.786413669586182,
      "learning_rate": 7.874478391225221e-05,
      "loss": 2.569788360595703,
      "memory(GiB)": 77.56,
      "step": 35600,
      "token_acc": 0.519434628975265,
      "train_speed(iter/s)": 1.439549
    },
    {
      "epoch": 1.5254273595818517,
      "grad_norm": 4.321645736694336,
      "learning_rate": 7.87392771717987e-05,
      "loss": 2.5054037094116213,
      "memory(GiB)": 77.56,
      "step": 35605,
      "token_acc": 0.4407894736842105,
      "train_speed(iter/s)": 1.439562
    },
    {
      "epoch": 1.5256415749111008,
      "grad_norm": 4.681307792663574,
      "learning_rate": 7.873376991070623e-05,
      "loss": 2.707870674133301,
      "memory(GiB)": 77.56,
      "step": 35610,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.439535
    },
    {
      "epoch": 1.5258557902403496,
      "grad_norm": 4.680245399475098,
      "learning_rate": 7.872826212907453e-05,
      "loss": 2.8780704498291017,
      "memory(GiB)": 77.56,
      "step": 35615,
      "token_acc": 0.4332129963898917,
      "train_speed(iter/s)": 1.439556
    },
    {
      "epoch": 1.5260700055695986,
      "grad_norm": 4.478838920593262,
      "learning_rate": 7.872275382700345e-05,
      "loss": 2.4710350036621094,
      "memory(GiB)": 77.56,
      "step": 35620,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.439574
    },
    {
      "epoch": 1.5262842208988476,
      "grad_norm": 5.685272693634033,
      "learning_rate": 7.871724500459272e-05,
      "loss": 2.4026954650878904,
      "memory(GiB)": 77.56,
      "step": 35625,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.439578
    },
    {
      "epoch": 1.5264984362280964,
      "grad_norm": 4.8700761795043945,
      "learning_rate": 7.871173566194217e-05,
      "loss": 2.677492141723633,
      "memory(GiB)": 77.56,
      "step": 35630,
      "token_acc": 0.4774774774774775,
      "train_speed(iter/s)": 1.439578
    },
    {
      "epoch": 1.5267126515573455,
      "grad_norm": 5.006597518920898,
      "learning_rate": 7.870622579915158e-05,
      "loss": 2.3438543319702148,
      "memory(GiB)": 77.56,
      "step": 35635,
      "token_acc": 0.5131578947368421,
      "train_speed(iter/s)": 1.439587
    },
    {
      "epoch": 1.5269268668865945,
      "grad_norm": 5.039852619171143,
      "learning_rate": 7.870071541632078e-05,
      "loss": 2.71852970123291,
      "memory(GiB)": 77.56,
      "step": 35640,
      "token_acc": 0.46439628482972134,
      "train_speed(iter/s)": 1.439588
    },
    {
      "epoch": 1.5271410822158433,
      "grad_norm": 6.308069705963135,
      "learning_rate": 7.869520451354961e-05,
      "loss": 2.248679351806641,
      "memory(GiB)": 77.56,
      "step": 35645,
      "token_acc": 0.531496062992126,
      "train_speed(iter/s)": 1.439566
    },
    {
      "epoch": 1.5273552975450924,
      "grad_norm": 5.813370227813721,
      "learning_rate": 7.868969309093788e-05,
      "loss": 2.7753889083862306,
      "memory(GiB)": 77.56,
      "step": 35650,
      "token_acc": 0.43944636678200694,
      "train_speed(iter/s)": 1.439597
    },
    {
      "epoch": 1.5275695128743414,
      "grad_norm": 4.956195831298828,
      "learning_rate": 7.868418114858545e-05,
      "loss": 2.3649627685546877,
      "memory(GiB)": 77.56,
      "step": 35655,
      "token_acc": 0.42379182156133827,
      "train_speed(iter/s)": 1.43962
    },
    {
      "epoch": 1.5277837282035902,
      "grad_norm": 4.613227844238281,
      "learning_rate": 7.867866868659218e-05,
      "loss": 2.7416263580322267,
      "memory(GiB)": 77.56,
      "step": 35660,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.439625
    },
    {
      "epoch": 1.5279979435328392,
      "grad_norm": 4.017142295837402,
      "learning_rate": 7.867315570505792e-05,
      "loss": 2.4690214157104493,
      "memory(GiB)": 77.56,
      "step": 35665,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.439674
    },
    {
      "epoch": 1.5282121588620883,
      "grad_norm": 5.85869836807251,
      "learning_rate": 7.866764220408255e-05,
      "loss": 2.5865554809570312,
      "memory(GiB)": 77.56,
      "step": 35670,
      "token_acc": 0.4559748427672956,
      "train_speed(iter/s)": 1.439706
    },
    {
      "epoch": 1.528426374191337,
      "grad_norm": 5.94870138168335,
      "learning_rate": 7.866212818376594e-05,
      "loss": 2.4458690643310548,
      "memory(GiB)": 77.56,
      "step": 35675,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.439699
    },
    {
      "epoch": 1.5286405895205861,
      "grad_norm": 4.3266448974609375,
      "learning_rate": 7.8656613644208e-05,
      "loss": 2.4299875259399415,
      "memory(GiB)": 77.56,
      "step": 35680,
      "token_acc": 0.4828897338403042,
      "train_speed(iter/s)": 1.439735
    },
    {
      "epoch": 1.5288548048498352,
      "grad_norm": 4.098716735839844,
      "learning_rate": 7.865109858550862e-05,
      "loss": 2.627867317199707,
      "memory(GiB)": 77.56,
      "step": 35685,
      "token_acc": 0.4450402144772118,
      "train_speed(iter/s)": 1.439787
    },
    {
      "epoch": 1.529069020179084,
      "grad_norm": 4.418527603149414,
      "learning_rate": 7.86455830077677e-05,
      "loss": 2.79116153717041,
      "memory(GiB)": 77.56,
      "step": 35690,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.439788
    },
    {
      "epoch": 1.529283235508333,
      "grad_norm": 5.227472305297852,
      "learning_rate": 7.864006691108519e-05,
      "loss": 2.8565208435058596,
      "memory(GiB)": 77.56,
      "step": 35695,
      "token_acc": 0.4054878048780488,
      "train_speed(iter/s)": 1.439796
    },
    {
      "epoch": 1.529497450837582,
      "grad_norm": 5.767401695251465,
      "learning_rate": 7.863455029556098e-05,
      "loss": 2.7329170227050783,
      "memory(GiB)": 77.56,
      "step": 35700,
      "token_acc": 0.45674740484429066,
      "train_speed(iter/s)": 1.439845
    },
    {
      "epoch": 1.5297116661668309,
      "grad_norm": 5.61146879196167,
      "learning_rate": 7.862903316129504e-05,
      "loss": 2.4084505081176757,
      "memory(GiB)": 77.56,
      "step": 35705,
      "token_acc": 0.45126353790613716,
      "train_speed(iter/s)": 1.439898
    },
    {
      "epoch": 1.5299258814960799,
      "grad_norm": 4.584047794342041,
      "learning_rate": 7.862351550838732e-05,
      "loss": 2.3794958114624025,
      "memory(GiB)": 77.56,
      "step": 35710,
      "token_acc": 0.4817073170731707,
      "train_speed(iter/s)": 1.439864
    },
    {
      "epoch": 1.530140096825329,
      "grad_norm": 5.686636924743652,
      "learning_rate": 7.861799733693775e-05,
      "loss": 2.7293888092041017,
      "memory(GiB)": 77.56,
      "step": 35715,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.439888
    },
    {
      "epoch": 1.5303543121545777,
      "grad_norm": 4.240833759307861,
      "learning_rate": 7.86124786470463e-05,
      "loss": 2.501619338989258,
      "memory(GiB)": 77.56,
      "step": 35720,
      "token_acc": 0.46449704142011833,
      "train_speed(iter/s)": 1.439918
    },
    {
      "epoch": 1.5305685274838268,
      "grad_norm": 4.446664810180664,
      "learning_rate": 7.8606959438813e-05,
      "loss": 2.2656715393066404,
      "memory(GiB)": 77.56,
      "step": 35725,
      "token_acc": 0.5037037037037037,
      "train_speed(iter/s)": 1.439929
    },
    {
      "epoch": 1.5307827428130758,
      "grad_norm": 4.771551609039307,
      "learning_rate": 7.860143971233776e-05,
      "loss": 2.5883575439453126,
      "memory(GiB)": 77.56,
      "step": 35730,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.439985
    },
    {
      "epoch": 1.5309969581423246,
      "grad_norm": 5.456704139709473,
      "learning_rate": 7.859591946772062e-05,
      "loss": 2.0569992065429688,
      "memory(GiB)": 77.56,
      "step": 35735,
      "token_acc": 0.5803921568627451,
      "train_speed(iter/s)": 1.44002
    },
    {
      "epoch": 1.5312111734715736,
      "grad_norm": 4.773999214172363,
      "learning_rate": 7.859039870506158e-05,
      "loss": 2.556158447265625,
      "memory(GiB)": 77.56,
      "step": 35740,
      "token_acc": 0.4709141274238227,
      "train_speed(iter/s)": 1.440046
    },
    {
      "epoch": 1.5314253888008227,
      "grad_norm": 5.565639019012451,
      "learning_rate": 7.858487742446063e-05,
      "loss": 2.3145166397094727,
      "memory(GiB)": 77.56,
      "step": 35745,
      "token_acc": 0.5298804780876494,
      "train_speed(iter/s)": 1.439999
    },
    {
      "epoch": 1.5316396041300715,
      "grad_norm": 4.366349220275879,
      "learning_rate": 7.857935562601783e-05,
      "loss": 2.548398017883301,
      "memory(GiB)": 77.56,
      "step": 35750,
      "token_acc": 0.4174174174174174,
      "train_speed(iter/s)": 1.43998
    },
    {
      "epoch": 1.5318538194593205,
      "grad_norm": 5.73050594329834,
      "learning_rate": 7.857383330983319e-05,
      "loss": 2.599017333984375,
      "memory(GiB)": 77.56,
      "step": 35755,
      "token_acc": 0.4306784660766962,
      "train_speed(iter/s)": 1.439999
    },
    {
      "epoch": 1.5320680347885696,
      "grad_norm": 6.19191837310791,
      "learning_rate": 7.856831047600674e-05,
      "loss": 2.465973663330078,
      "memory(GiB)": 77.56,
      "step": 35760,
      "token_acc": 0.47410358565737054,
      "train_speed(iter/s)": 1.440026
    },
    {
      "epoch": 1.5322822501178184,
      "grad_norm": 4.285295486450195,
      "learning_rate": 7.856278712463856e-05,
      "loss": 2.5737236022949217,
      "memory(GiB)": 77.56,
      "step": 35765,
      "token_acc": 0.45686900958466453,
      "train_speed(iter/s)": 1.440002
    },
    {
      "epoch": 1.5324964654470674,
      "grad_norm": 6.242623329162598,
      "learning_rate": 7.855726325582869e-05,
      "loss": 2.522172546386719,
      "memory(GiB)": 77.56,
      "step": 35770,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.440018
    },
    {
      "epoch": 1.5327106807763164,
      "grad_norm": 6.120961666107178,
      "learning_rate": 7.855173886967722e-05,
      "loss": 2.373886489868164,
      "memory(GiB)": 77.56,
      "step": 35775,
      "token_acc": 0.4664429530201342,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 1.5329248961055653,
      "grad_norm": 4.496016025543213,
      "learning_rate": 7.854621396628421e-05,
      "loss": 2.491876411437988,
      "memory(GiB)": 77.56,
      "step": 35780,
      "token_acc": 0.4527687296416938,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 1.5331391114348143,
      "grad_norm": 5.415960311889648,
      "learning_rate": 7.854068854574976e-05,
      "loss": 2.5605112075805665,
      "memory(GiB)": 77.56,
      "step": 35785,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.440039
    },
    {
      "epoch": 1.5333533267640633,
      "grad_norm": 4.093324184417725,
      "learning_rate": 7.853516260817395e-05,
      "loss": 2.450044631958008,
      "memory(GiB)": 77.56,
      "step": 35790,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.440036
    },
    {
      "epoch": 1.5335675420933121,
      "grad_norm": 4.74154806137085,
      "learning_rate": 7.85296361536569e-05,
      "loss": 2.6434085845947264,
      "memory(GiB)": 77.56,
      "step": 35795,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.440034
    },
    {
      "epoch": 1.5337817574225612,
      "grad_norm": 4.665964126586914,
      "learning_rate": 7.852410918229874e-05,
      "loss": 2.134696388244629,
      "memory(GiB)": 77.56,
      "step": 35800,
      "token_acc": 0.55859375,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 1.5339959727518102,
      "grad_norm": 4.803917407989502,
      "learning_rate": 7.851858169419959e-05,
      "loss": 2.351963424682617,
      "memory(GiB)": 77.56,
      "step": 35805,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.440058
    },
    {
      "epoch": 1.534210188081059,
      "grad_norm": 5.831637382507324,
      "learning_rate": 7.851305368945956e-05,
      "loss": 2.5339468002319334,
      "memory(GiB)": 77.56,
      "step": 35810,
      "token_acc": 0.4721189591078067,
      "train_speed(iter/s)": 1.439996
    },
    {
      "epoch": 1.534424403410308,
      "grad_norm": 5.5182204246521,
      "learning_rate": 7.850752516817883e-05,
      "loss": 2.585394859313965,
      "memory(GiB)": 77.56,
      "step": 35815,
      "token_acc": 0.48562300319488816,
      "train_speed(iter/s)": 1.440056
    },
    {
      "epoch": 1.534638618739557,
      "grad_norm": 3.996464967727661,
      "learning_rate": 7.850199613045754e-05,
      "loss": 2.4121829986572267,
      "memory(GiB)": 77.56,
      "step": 35820,
      "token_acc": 0.5044776119402985,
      "train_speed(iter/s)": 1.44005
    },
    {
      "epoch": 1.534852834068806,
      "grad_norm": 4.608769416809082,
      "learning_rate": 7.849646657639585e-05,
      "loss": 2.4529977798461915,
      "memory(GiB)": 77.56,
      "step": 35825,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.440036
    },
    {
      "epoch": 1.535067049398055,
      "grad_norm": 5.1034440994262695,
      "learning_rate": 7.849093650609392e-05,
      "loss": 2.3742450714111327,
      "memory(GiB)": 77.56,
      "step": 35830,
      "token_acc": 0.5254901960784314,
      "train_speed(iter/s)": 1.440017
    },
    {
      "epoch": 1.535281264727304,
      "grad_norm": 4.757763862609863,
      "learning_rate": 7.848540591965197e-05,
      "loss": 2.565390396118164,
      "memory(GiB)": 77.56,
      "step": 35835,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.440023
    },
    {
      "epoch": 1.5354954800565528,
      "grad_norm": 5.375235080718994,
      "learning_rate": 7.847987481717017e-05,
      "loss": 2.7975418090820314,
      "memory(GiB)": 77.56,
      "step": 35840,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.440018
    },
    {
      "epoch": 1.5357096953858018,
      "grad_norm": 4.377996921539307,
      "learning_rate": 7.847434319874871e-05,
      "loss": 2.5620803833007812,
      "memory(GiB)": 77.56,
      "step": 35845,
      "token_acc": 0.4891640866873065,
      "train_speed(iter/s)": 1.440081
    },
    {
      "epoch": 1.5359239107150509,
      "grad_norm": 5.474483966827393,
      "learning_rate": 7.846881106448781e-05,
      "loss": 2.30145149230957,
      "memory(GiB)": 77.56,
      "step": 35850,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.440093
    },
    {
      "epoch": 1.5361381260442997,
      "grad_norm": 6.444149494171143,
      "learning_rate": 7.846327841448771e-05,
      "loss": 2.3971038818359376,
      "memory(GiB)": 77.56,
      "step": 35855,
      "token_acc": 0.49146757679180886,
      "train_speed(iter/s)": 1.440106
    },
    {
      "epoch": 1.5363523413735487,
      "grad_norm": 4.7963948249816895,
      "learning_rate": 7.84577452488486e-05,
      "loss": 2.344122123718262,
      "memory(GiB)": 77.56,
      "step": 35860,
      "token_acc": 0.5052631578947369,
      "train_speed(iter/s)": 1.440089
    },
    {
      "epoch": 1.5365665567027977,
      "grad_norm": 4.2550578117370605,
      "learning_rate": 7.845221156767076e-05,
      "loss": 2.6109573364257814,
      "memory(GiB)": 77.56,
      "step": 35865,
      "token_acc": 0.4339622641509434,
      "train_speed(iter/s)": 1.440114
    },
    {
      "epoch": 1.5367807720320465,
      "grad_norm": 6.8269267082214355,
      "learning_rate": 7.84466773710544e-05,
      "loss": 2.2013324737548827,
      "memory(GiB)": 77.56,
      "step": 35870,
      "token_acc": 0.47035573122529645,
      "train_speed(iter/s)": 1.440161
    },
    {
      "epoch": 1.5369949873612956,
      "grad_norm": 6.0858073234558105,
      "learning_rate": 7.844114265909979e-05,
      "loss": 2.5771881103515626,
      "memory(GiB)": 77.56,
      "step": 35875,
      "token_acc": 0.4843205574912892,
      "train_speed(iter/s)": 1.440177
    },
    {
      "epoch": 1.5372092026905446,
      "grad_norm": 4.231525421142578,
      "learning_rate": 7.843560743190721e-05,
      "loss": 2.3409427642822265,
      "memory(GiB)": 77.56,
      "step": 35880,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.440189
    },
    {
      "epoch": 1.5374234180197934,
      "grad_norm": 6.729775905609131,
      "learning_rate": 7.843007168957693e-05,
      "loss": 2.240573692321777,
      "memory(GiB)": 77.56,
      "step": 35885,
      "token_acc": 0.539568345323741,
      "train_speed(iter/s)": 1.440185
    },
    {
      "epoch": 1.5376376333490425,
      "grad_norm": 3.6165997982025146,
      "learning_rate": 7.842453543220924e-05,
      "loss": 2.641560935974121,
      "memory(GiB)": 77.56,
      "step": 35890,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.44024
    },
    {
      "epoch": 1.5378518486782915,
      "grad_norm": 5.708844184875488,
      "learning_rate": 7.841899865990441e-05,
      "loss": 2.79778938293457,
      "memory(GiB)": 77.56,
      "step": 35895,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.440237
    },
    {
      "epoch": 1.5380660640075403,
      "grad_norm": 4.266225814819336,
      "learning_rate": 7.841346137276276e-05,
      "loss": 2.532614517211914,
      "memory(GiB)": 77.56,
      "step": 35900,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.440234
    },
    {
      "epoch": 1.5382802793367893,
      "grad_norm": 5.179828643798828,
      "learning_rate": 7.84079235708846e-05,
      "loss": 2.3186885833740236,
      "memory(GiB)": 77.56,
      "step": 35905,
      "token_acc": 0.5311355311355311,
      "train_speed(iter/s)": 1.440252
    },
    {
      "epoch": 1.5384944946660384,
      "grad_norm": 4.616061687469482,
      "learning_rate": 7.840238525437027e-05,
      "loss": 2.5573942184448244,
      "memory(GiB)": 77.56,
      "step": 35910,
      "token_acc": 0.47988505747126436,
      "train_speed(iter/s)": 1.440279
    },
    {
      "epoch": 1.5387087099952872,
      "grad_norm": 5.172558307647705,
      "learning_rate": 7.839684642332008e-05,
      "loss": 2.4401458740234374,
      "memory(GiB)": 77.56,
      "step": 35915,
      "token_acc": 0.4844961240310077,
      "train_speed(iter/s)": 1.440276
    },
    {
      "epoch": 1.5389229253245362,
      "grad_norm": 4.917933464050293,
      "learning_rate": 7.839130707783438e-05,
      "loss": 2.329453468322754,
      "memory(GiB)": 77.56,
      "step": 35920,
      "token_acc": 0.5080385852090032,
      "train_speed(iter/s)": 1.440267
    },
    {
      "epoch": 1.5391371406537853,
      "grad_norm": 5.331910133361816,
      "learning_rate": 7.838576721801351e-05,
      "loss": 2.6115036010742188,
      "memory(GiB)": 77.56,
      "step": 35925,
      "token_acc": 0.44594594594594594,
      "train_speed(iter/s)": 1.440266
    },
    {
      "epoch": 1.539351355983034,
      "grad_norm": 4.201162815093994,
      "learning_rate": 7.838022684395786e-05,
      "loss": 2.461620330810547,
      "memory(GiB)": 77.56,
      "step": 35930,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.44029
    },
    {
      "epoch": 1.539565571312283,
      "grad_norm": 4.826604843139648,
      "learning_rate": 7.837468595576777e-05,
      "loss": 2.5974220275878905,
      "memory(GiB)": 77.56,
      "step": 35935,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.440323
    },
    {
      "epoch": 1.5397797866415321,
      "grad_norm": 5.460199356079102,
      "learning_rate": 7.836914455354362e-05,
      "loss": 2.337574005126953,
      "memory(GiB)": 77.56,
      "step": 35940,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.440357
    },
    {
      "epoch": 1.539994001970781,
      "grad_norm": 4.470697402954102,
      "learning_rate": 7.83636026373858e-05,
      "loss": 2.454970359802246,
      "memory(GiB)": 77.56,
      "step": 35945,
      "token_acc": 0.4407894736842105,
      "train_speed(iter/s)": 1.440409
    },
    {
      "epoch": 1.54020821730003,
      "grad_norm": 6.884347915649414,
      "learning_rate": 7.835806020739472e-05,
      "loss": 2.413157653808594,
      "memory(GiB)": 77.56,
      "step": 35950,
      "token_acc": 0.4261168384879725,
      "train_speed(iter/s)": 1.440452
    },
    {
      "epoch": 1.540422432629279,
      "grad_norm": 4.889522552490234,
      "learning_rate": 7.835251726367078e-05,
      "loss": 2.2992578506469727,
      "memory(GiB)": 77.56,
      "step": 35955,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.440406
    },
    {
      "epoch": 1.5406366479585278,
      "grad_norm": 6.5905632972717285,
      "learning_rate": 7.834697380631438e-05,
      "loss": 2.235517883300781,
      "memory(GiB)": 77.56,
      "step": 35960,
      "token_acc": 0.5441176470588235,
      "train_speed(iter/s)": 1.440435
    },
    {
      "epoch": 1.5408508632877769,
      "grad_norm": 5.198739051818848,
      "learning_rate": 7.834142983542597e-05,
      "loss": 2.703260040283203,
      "memory(GiB)": 77.56,
      "step": 35965,
      "token_acc": 0.501577287066246,
      "train_speed(iter/s)": 1.440417
    },
    {
      "epoch": 1.541065078617026,
      "grad_norm": 4.178027629852295,
      "learning_rate": 7.833588535110597e-05,
      "loss": 2.3758218765258787,
      "memory(GiB)": 77.56,
      "step": 35970,
      "token_acc": 0.4854014598540146,
      "train_speed(iter/s)": 1.440427
    },
    {
      "epoch": 1.5412792939462747,
      "grad_norm": 5.941470146179199,
      "learning_rate": 7.833034035345483e-05,
      "loss": 2.4637609481811524,
      "memory(GiB)": 77.56,
      "step": 35975,
      "token_acc": 0.44404332129963897,
      "train_speed(iter/s)": 1.440426
    },
    {
      "epoch": 1.5414935092755238,
      "grad_norm": 5.588265419006348,
      "learning_rate": 7.832479484257298e-05,
      "loss": 2.5613609313964845,
      "memory(GiB)": 77.56,
      "step": 35980,
      "token_acc": 0.41580756013745707,
      "train_speed(iter/s)": 1.440441
    },
    {
      "epoch": 1.5417077246047728,
      "grad_norm": 6.106418609619141,
      "learning_rate": 7.831924881856092e-05,
      "loss": 2.515755271911621,
      "memory(GiB)": 77.56,
      "step": 35985,
      "token_acc": 0.4815950920245399,
      "train_speed(iter/s)": 1.440454
    },
    {
      "epoch": 1.5419219399340216,
      "grad_norm": 4.166738986968994,
      "learning_rate": 7.831370228151909e-05,
      "loss": 2.3808338165283205,
      "memory(GiB)": 77.56,
      "step": 35990,
      "token_acc": 0.5152439024390244,
      "train_speed(iter/s)": 1.440458
    },
    {
      "epoch": 1.5421361552632706,
      "grad_norm": 4.873399257659912,
      "learning_rate": 7.830815523154799e-05,
      "loss": 2.7323768615722654,
      "memory(GiB)": 77.56,
      "step": 35995,
      "token_acc": 0.4520123839009288,
      "train_speed(iter/s)": 1.44048
    },
    {
      "epoch": 1.5423503705925197,
      "grad_norm": 4.752975940704346,
      "learning_rate": 7.83026076687481e-05,
      "loss": 2.5931865692138674,
      "memory(GiB)": 77.56,
      "step": 36000,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.440489
    },
    {
      "epoch": 1.5423503705925197,
      "eval_loss": 2.1286087036132812,
      "eval_runtime": 14.9947,
      "eval_samples_per_second": 6.669,
      "eval_steps_per_second": 6.669,
      "eval_token_acc": 0.48885793871866295,
      "step": 36000
    },
    {
      "epoch": 1.5425645859217685,
      "grad_norm": 4.597265243530273,
      "learning_rate": 7.829705959321993e-05,
      "loss": 2.367877388000488,
      "memory(GiB)": 77.56,
      "step": 36005,
      "token_acc": 0.4871060171919771,
      "train_speed(iter/s)": 1.439519
    },
    {
      "epoch": 1.5427788012510175,
      "grad_norm": 5.153852939605713,
      "learning_rate": 7.829151100506396e-05,
      "loss": 2.6542198181152346,
      "memory(GiB)": 77.56,
      "step": 36010,
      "token_acc": 0.46283783783783783,
      "train_speed(iter/s)": 1.4395
    },
    {
      "epoch": 1.5429930165802666,
      "grad_norm": 4.836722373962402,
      "learning_rate": 7.828596190438075e-05,
      "loss": 2.182339096069336,
      "memory(GiB)": 77.56,
      "step": 36015,
      "token_acc": 0.4980237154150198,
      "train_speed(iter/s)": 1.439547
    },
    {
      "epoch": 1.5432072319095154,
      "grad_norm": 4.521052360534668,
      "learning_rate": 7.828041229127079e-05,
      "loss": 2.690706253051758,
      "memory(GiB)": 77.56,
      "step": 36020,
      "token_acc": 0.43434343434343436,
      "train_speed(iter/s)": 1.439584
    },
    {
      "epoch": 1.5434214472387644,
      "grad_norm": 5.721269607543945,
      "learning_rate": 7.827486216583465e-05,
      "loss": 2.8337615966796874,
      "memory(GiB)": 77.56,
      "step": 36025,
      "token_acc": 0.4538653366583541,
      "train_speed(iter/s)": 1.439596
    },
    {
      "epoch": 1.5436356625680134,
      "grad_norm": 6.398462295532227,
      "learning_rate": 7.826931152817283e-05,
      "loss": 2.1992380142211916,
      "memory(GiB)": 77.56,
      "step": 36030,
      "token_acc": 0.56,
      "train_speed(iter/s)": 1.439617
    },
    {
      "epoch": 1.5438498778972622,
      "grad_norm": 5.473598957061768,
      "learning_rate": 7.826376037838594e-05,
      "loss": 2.2730775833129884,
      "memory(GiB)": 77.56,
      "step": 36035,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439618
    },
    {
      "epoch": 1.5440640932265113,
      "grad_norm": 4.58029842376709,
      "learning_rate": 7.82582087165745e-05,
      "loss": 2.492129325866699,
      "memory(GiB)": 77.56,
      "step": 36040,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.439678
    },
    {
      "epoch": 1.5442783085557603,
      "grad_norm": 5.405659198760986,
      "learning_rate": 7.82526565428391e-05,
      "loss": 2.7293840408325196,
      "memory(GiB)": 77.56,
      "step": 36045,
      "token_acc": 0.43010752688172044,
      "train_speed(iter/s)": 1.439638
    },
    {
      "epoch": 1.5444925238850091,
      "grad_norm": 5.566967010498047,
      "learning_rate": 7.824710385728033e-05,
      "loss": 3.030938911437988,
      "memory(GiB)": 77.56,
      "step": 36050,
      "token_acc": 0.4276729559748428,
      "train_speed(iter/s)": 1.439674
    },
    {
      "epoch": 1.5447067392142582,
      "grad_norm": 4.404112339019775,
      "learning_rate": 7.824155065999879e-05,
      "loss": 2.689794158935547,
      "memory(GiB)": 77.56,
      "step": 36055,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.439698
    },
    {
      "epoch": 1.5449209545435072,
      "grad_norm": 4.441066741943359,
      "learning_rate": 7.823599695109504e-05,
      "loss": 2.8341550827026367,
      "memory(GiB)": 77.56,
      "step": 36060,
      "token_acc": 0.41642228739002934,
      "train_speed(iter/s)": 1.439693
    },
    {
      "epoch": 1.545135169872756,
      "grad_norm": 5.663236141204834,
      "learning_rate": 7.823044273066975e-05,
      "loss": 2.3890533447265625,
      "memory(GiB)": 77.56,
      "step": 36065,
      "token_acc": 0.4980392156862745,
      "train_speed(iter/s)": 1.439656
    },
    {
      "epoch": 1.545349385202005,
      "grad_norm": 3.5448763370513916,
      "learning_rate": 7.822488799882348e-05,
      "loss": 2.541372299194336,
      "memory(GiB)": 77.56,
      "step": 36070,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.439685
    },
    {
      "epoch": 1.545563600531254,
      "grad_norm": 4.116077423095703,
      "learning_rate": 7.821933275565691e-05,
      "loss": 2.534458351135254,
      "memory(GiB)": 77.56,
      "step": 36075,
      "token_acc": 0.50390625,
      "train_speed(iter/s)": 1.439678
    },
    {
      "epoch": 1.545777815860503,
      "grad_norm": 7.806304931640625,
      "learning_rate": 7.821377700127063e-05,
      "loss": 2.5961685180664062,
      "memory(GiB)": 77.56,
      "step": 36080,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.43972
    },
    {
      "epoch": 1.545992031189752,
      "grad_norm": 5.884232044219971,
      "learning_rate": 7.820822073576534e-05,
      "loss": 2.6835458755493162,
      "memory(GiB)": 77.56,
      "step": 36085,
      "token_acc": 0.4701492537313433,
      "train_speed(iter/s)": 1.43968
    },
    {
      "epoch": 1.546206246519001,
      "grad_norm": 3.6564464569091797,
      "learning_rate": 7.820266395924164e-05,
      "loss": 2.40456485748291,
      "memory(GiB)": 77.56,
      "step": 36090,
      "token_acc": 0.46932515337423314,
      "train_speed(iter/s)": 1.439647
    },
    {
      "epoch": 1.5464204618482498,
      "grad_norm": 4.301235198974609,
      "learning_rate": 7.819710667180024e-05,
      "loss": 2.34564323425293,
      "memory(GiB)": 77.56,
      "step": 36095,
      "token_acc": 0.5045592705167173,
      "train_speed(iter/s)": 1.439667
    },
    {
      "epoch": 1.5466346771774988,
      "grad_norm": 4.588385105133057,
      "learning_rate": 7.819154887354182e-05,
      "loss": 2.4966814041137697,
      "memory(GiB)": 77.56,
      "step": 36100,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.439681
    },
    {
      "epoch": 1.5468488925067478,
      "grad_norm": 4.888988494873047,
      "learning_rate": 7.818599056456701e-05,
      "loss": 2.5890830993652343,
      "memory(GiB)": 77.56,
      "step": 36105,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.439718
    },
    {
      "epoch": 1.5470631078359967,
      "grad_norm": 6.6023993492126465,
      "learning_rate": 7.818043174497656e-05,
      "loss": 2.503236770629883,
      "memory(GiB)": 77.56,
      "step": 36110,
      "token_acc": 0.4847457627118644,
      "train_speed(iter/s)": 1.439756
    },
    {
      "epoch": 1.5472773231652457,
      "grad_norm": 4.630427837371826,
      "learning_rate": 7.817487241487115e-05,
      "loss": 2.526523208618164,
      "memory(GiB)": 77.56,
      "step": 36115,
      "token_acc": 0.5015974440894568,
      "train_speed(iter/s)": 1.439782
    },
    {
      "epoch": 1.5474915384944947,
      "grad_norm": 5.061960220336914,
      "learning_rate": 7.816931257435151e-05,
      "loss": 2.577986717224121,
      "memory(GiB)": 77.56,
      "step": 36120,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.439797
    },
    {
      "epoch": 1.5477057538237435,
      "grad_norm": 5.558644771575928,
      "learning_rate": 7.816375222351833e-05,
      "loss": 2.602009391784668,
      "memory(GiB)": 77.56,
      "step": 36125,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.43983
    },
    {
      "epoch": 1.5479199691529926,
      "grad_norm": 4.5223565101623535,
      "learning_rate": 7.815819136247238e-05,
      "loss": 2.6625133514404298,
      "memory(GiB)": 77.56,
      "step": 36130,
      "token_acc": 0.45110410094637227,
      "train_speed(iter/s)": 1.439803
    },
    {
      "epoch": 1.5481341844822416,
      "grad_norm": 4.3092851638793945,
      "learning_rate": 7.815262999131436e-05,
      "loss": 2.432361030578613,
      "memory(GiB)": 77.56,
      "step": 36135,
      "token_acc": 0.5202702702702703,
      "train_speed(iter/s)": 1.439807
    },
    {
      "epoch": 1.5483483998114904,
      "grad_norm": 5.107344150543213,
      "learning_rate": 7.814706811014504e-05,
      "loss": 2.4219039916992187,
      "memory(GiB)": 77.56,
      "step": 36140,
      "token_acc": 0.49032258064516127,
      "train_speed(iter/s)": 1.439782
    },
    {
      "epoch": 1.5485626151407395,
      "grad_norm": 4.754781246185303,
      "learning_rate": 7.814150571906517e-05,
      "loss": 2.3867931365966797,
      "memory(GiB)": 77.56,
      "step": 36145,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.43981
    },
    {
      "epoch": 1.5487768304699885,
      "grad_norm": 4.836016654968262,
      "learning_rate": 7.813594281817555e-05,
      "loss": 2.7338991165161133,
      "memory(GiB)": 77.56,
      "step": 36150,
      "token_acc": 0.4489795918367347,
      "train_speed(iter/s)": 1.439849
    },
    {
      "epoch": 1.5489910457992373,
      "grad_norm": 4.735340595245361,
      "learning_rate": 7.813037940757692e-05,
      "loss": 2.7531713485717773,
      "memory(GiB)": 77.56,
      "step": 36155,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.439827
    },
    {
      "epoch": 1.5492052611284863,
      "grad_norm": 5.50546407699585,
      "learning_rate": 7.812481548737007e-05,
      "loss": 2.559846115112305,
      "memory(GiB)": 77.56,
      "step": 36160,
      "token_acc": 0.4844290657439446,
      "train_speed(iter/s)": 1.439863
    },
    {
      "epoch": 1.5494194764577354,
      "grad_norm": 4.673547744750977,
      "learning_rate": 7.811925105765581e-05,
      "loss": 2.452170753479004,
      "memory(GiB)": 77.56,
      "step": 36165,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.439864
    },
    {
      "epoch": 1.5496336917869842,
      "grad_norm": 4.878746509552002,
      "learning_rate": 7.811368611853493e-05,
      "loss": 2.666554832458496,
      "memory(GiB)": 77.56,
      "step": 36170,
      "token_acc": 0.4744525547445255,
      "train_speed(iter/s)": 1.439876
    },
    {
      "epoch": 1.5498479071162332,
      "grad_norm": 4.516483306884766,
      "learning_rate": 7.810812067010827e-05,
      "loss": 2.686572265625,
      "memory(GiB)": 77.56,
      "step": 36175,
      "token_acc": 0.4295774647887324,
      "train_speed(iter/s)": 1.439919
    },
    {
      "epoch": 1.5500621224454822,
      "grad_norm": 4.7003912925720215,
      "learning_rate": 7.810255471247663e-05,
      "loss": 2.6671051025390624,
      "memory(GiB)": 77.56,
      "step": 36180,
      "token_acc": 0.47019867549668876,
      "train_speed(iter/s)": 1.43996
    },
    {
      "epoch": 1.550276337774731,
      "grad_norm": 4.506163120269775,
      "learning_rate": 7.809698824574085e-05,
      "loss": 2.5628196716308596,
      "memory(GiB)": 77.56,
      "step": 36185,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.440004
    },
    {
      "epoch": 1.55049055310398,
      "grad_norm": 4.055752277374268,
      "learning_rate": 7.809142127000176e-05,
      "loss": 2.416562080383301,
      "memory(GiB)": 77.56,
      "step": 36190,
      "token_acc": 0.4662756598240469,
      "train_speed(iter/s)": 1.440031
    },
    {
      "epoch": 1.5507047684332291,
      "grad_norm": 4.566366672515869,
      "learning_rate": 7.808585378536024e-05,
      "loss": 2.5232059478759767,
      "memory(GiB)": 77.56,
      "step": 36195,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.44006
    },
    {
      "epoch": 1.550918983762478,
      "grad_norm": 5.101639747619629,
      "learning_rate": 7.808028579191711e-05,
      "loss": 2.614002227783203,
      "memory(GiB)": 77.56,
      "step": 36200,
      "token_acc": 0.45045045045045046,
      "train_speed(iter/s)": 1.44008
    },
    {
      "epoch": 1.551133199091727,
      "grad_norm": 5.844062805175781,
      "learning_rate": 7.807471728977327e-05,
      "loss": 2.3080062866210938,
      "memory(GiB)": 77.56,
      "step": 36205,
      "token_acc": 0.4866920152091255,
      "train_speed(iter/s)": 1.440079
    },
    {
      "epoch": 1.551347414420976,
      "grad_norm": 3.9902467727661133,
      "learning_rate": 7.80691482790296e-05,
      "loss": 2.4026275634765626,
      "memory(GiB)": 77.56,
      "step": 36210,
      "token_acc": 0.46348314606741575,
      "train_speed(iter/s)": 1.440096
    },
    {
      "epoch": 1.5515616297502248,
      "grad_norm": 5.366555213928223,
      "learning_rate": 7.806357875978698e-05,
      "loss": 2.6843694686889648,
      "memory(GiB)": 77.56,
      "step": 36215,
      "token_acc": 0.42996742671009774,
      "train_speed(iter/s)": 1.440122
    },
    {
      "epoch": 1.5517758450794739,
      "grad_norm": 7.945376396179199,
      "learning_rate": 7.805800873214628e-05,
      "loss": 2.5908931732177733,
      "memory(GiB)": 77.56,
      "step": 36220,
      "token_acc": 0.463768115942029,
      "train_speed(iter/s)": 1.440121
    },
    {
      "epoch": 1.551990060408723,
      "grad_norm": 5.47683048248291,
      "learning_rate": 7.805243819620845e-05,
      "loss": 2.338287925720215,
      "memory(GiB)": 77.56,
      "step": 36225,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.440076
    },
    {
      "epoch": 1.5522042757379717,
      "grad_norm": 4.751999855041504,
      "learning_rate": 7.80468671520744e-05,
      "loss": 2.6259727478027344,
      "memory(GiB)": 77.56,
      "step": 36230,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.440115
    },
    {
      "epoch": 1.5524184910672207,
      "grad_norm": 5.376540660858154,
      "learning_rate": 7.804129559984501e-05,
      "loss": 2.642112159729004,
      "memory(GiB)": 77.56,
      "step": 36235,
      "token_acc": 0.46846846846846846,
      "train_speed(iter/s)": 1.440112
    },
    {
      "epoch": 1.5526327063964698,
      "grad_norm": 5.526229381561279,
      "learning_rate": 7.803572353962126e-05,
      "loss": 2.418209457397461,
      "memory(GiB)": 77.56,
      "step": 36240,
      "token_acc": 0.45182724252491696,
      "train_speed(iter/s)": 1.440135
    },
    {
      "epoch": 1.5528469217257186,
      "grad_norm": 6.461812496185303,
      "learning_rate": 7.80301509715041e-05,
      "loss": 2.478191947937012,
      "memory(GiB)": 77.56,
      "step": 36245,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.440169
    },
    {
      "epoch": 1.5530611370549676,
      "grad_norm": 4.604681015014648,
      "learning_rate": 7.802457789559443e-05,
      "loss": 2.3935449600219725,
      "memory(GiB)": 77.56,
      "step": 36250,
      "token_acc": 0.47038327526132406,
      "train_speed(iter/s)": 1.440211
    },
    {
      "epoch": 1.5532753523842167,
      "grad_norm": 4.748660087585449,
      "learning_rate": 7.801900431199325e-05,
      "loss": 2.4280540466308596,
      "memory(GiB)": 77.56,
      "step": 36255,
      "token_acc": 0.4962121212121212,
      "train_speed(iter/s)": 1.440221
    },
    {
      "epoch": 1.5534895677134655,
      "grad_norm": 4.3995161056518555,
      "learning_rate": 7.801343022080152e-05,
      "loss": 2.305587577819824,
      "memory(GiB)": 77.56,
      "step": 36260,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.440228
    },
    {
      "epoch": 1.5537037830427145,
      "grad_norm": 5.565771579742432,
      "learning_rate": 7.800785562212025e-05,
      "loss": 2.4840831756591797,
      "memory(GiB)": 77.56,
      "step": 36265,
      "token_acc": 0.462406015037594,
      "train_speed(iter/s)": 1.440214
    },
    {
      "epoch": 1.5539179983719635,
      "grad_norm": 3.6793878078460693,
      "learning_rate": 7.800228051605036e-05,
      "loss": 2.393549346923828,
      "memory(GiB)": 77.56,
      "step": 36270,
      "token_acc": 0.48589341692789967,
      "train_speed(iter/s)": 1.440207
    },
    {
      "epoch": 1.5541322137012124,
      "grad_norm": 4.280629634857178,
      "learning_rate": 7.799670490269291e-05,
      "loss": 2.7498884201049805,
      "memory(GiB)": 77.56,
      "step": 36275,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.440273
    },
    {
      "epoch": 1.5543464290304614,
      "grad_norm": 4.478820323944092,
      "learning_rate": 7.799112878214889e-05,
      "loss": 2.5493656158447267,
      "memory(GiB)": 77.56,
      "step": 36280,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.440292
    },
    {
      "epoch": 1.5545606443597104,
      "grad_norm": 5.53695011138916,
      "learning_rate": 7.798555215451932e-05,
      "loss": 2.5075130462646484,
      "memory(GiB)": 77.56,
      "step": 36285,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.44031
    },
    {
      "epoch": 1.5547748596889592,
      "grad_norm": 5.313154697418213,
      "learning_rate": 7.797997501990522e-05,
      "loss": 2.4900819778442385,
      "memory(GiB)": 77.56,
      "step": 36290,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.44034
    },
    {
      "epoch": 1.5549890750182083,
      "grad_norm": 4.965277671813965,
      "learning_rate": 7.797439737840761e-05,
      "loss": 2.4642307281494142,
      "memory(GiB)": 77.56,
      "step": 36295,
      "token_acc": 0.4600760456273764,
      "train_speed(iter/s)": 1.440352
    },
    {
      "epoch": 1.5552032903474573,
      "grad_norm": 5.644978046417236,
      "learning_rate": 7.796881923012755e-05,
      "loss": 2.5070371627807617,
      "memory(GiB)": 77.56,
      "step": 36300,
      "token_acc": 0.46382978723404256,
      "train_speed(iter/s)": 1.44033
    },
    {
      "epoch": 1.5554175056767061,
      "grad_norm": 4.894263744354248,
      "learning_rate": 7.796324057516611e-05,
      "loss": 2.4139293670654296,
      "memory(GiB)": 77.56,
      "step": 36305,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.440339
    },
    {
      "epoch": 1.5556317210059551,
      "grad_norm": 5.888087749481201,
      "learning_rate": 7.795766141362432e-05,
      "loss": 2.6148553848266602,
      "memory(GiB)": 77.56,
      "step": 36310,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.440305
    },
    {
      "epoch": 1.5558459363352042,
      "grad_norm": 7.530528545379639,
      "learning_rate": 7.795208174560326e-05,
      "loss": 2.598414993286133,
      "memory(GiB)": 77.56,
      "step": 36315,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.440313
    },
    {
      "epoch": 1.556060151664453,
      "grad_norm": 4.82913875579834,
      "learning_rate": 7.794650157120405e-05,
      "loss": 2.1892459869384764,
      "memory(GiB)": 77.56,
      "step": 36320,
      "token_acc": 0.5340909090909091,
      "train_speed(iter/s)": 1.440292
    },
    {
      "epoch": 1.556274366993702,
      "grad_norm": 4.564405918121338,
      "learning_rate": 7.794092089052774e-05,
      "loss": 2.514790725708008,
      "memory(GiB)": 77.56,
      "step": 36325,
      "token_acc": 0.5066225165562914,
      "train_speed(iter/s)": 1.440281
    },
    {
      "epoch": 1.556488582322951,
      "grad_norm": 4.016900062561035,
      "learning_rate": 7.793533970367542e-05,
      "loss": 2.4102779388427735,
      "memory(GiB)": 77.56,
      "step": 36330,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.440339
    },
    {
      "epoch": 1.5567027976521999,
      "grad_norm": 5.928051471710205,
      "learning_rate": 7.792975801074822e-05,
      "loss": 2.5120061874389648,
      "memory(GiB)": 77.56,
      "step": 36335,
      "token_acc": 0.45878136200716846,
      "train_speed(iter/s)": 1.440304
    },
    {
      "epoch": 1.556917012981449,
      "grad_norm": 5.336033821105957,
      "learning_rate": 7.792417581184725e-05,
      "loss": 2.4910240173339844,
      "memory(GiB)": 77.56,
      "step": 36340,
      "token_acc": 0.47633136094674555,
      "train_speed(iter/s)": 1.440342
    },
    {
      "epoch": 1.557131228310698,
      "grad_norm": 6.072386741638184,
      "learning_rate": 7.791859310707366e-05,
      "loss": 2.4716182708740235,
      "memory(GiB)": 77.56,
      "step": 36345,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.440371
    },
    {
      "epoch": 1.5573454436399468,
      "grad_norm": 4.771454811096191,
      "learning_rate": 7.791300989652853e-05,
      "loss": 2.791448211669922,
      "memory(GiB)": 77.56,
      "step": 36350,
      "token_acc": 0.45110410094637227,
      "train_speed(iter/s)": 1.440411
    },
    {
      "epoch": 1.5575596589691958,
      "grad_norm": 6.639774322509766,
      "learning_rate": 7.790742618031307e-05,
      "loss": 2.591164779663086,
      "memory(GiB)": 77.56,
      "step": 36355,
      "token_acc": 0.42955326460481097,
      "train_speed(iter/s)": 1.440454
    },
    {
      "epoch": 1.5577738742984448,
      "grad_norm": 4.969061851501465,
      "learning_rate": 7.79018419585284e-05,
      "loss": 2.256228446960449,
      "memory(GiB)": 77.56,
      "step": 36360,
      "token_acc": 0.5423076923076923,
      "train_speed(iter/s)": 1.440444
    },
    {
      "epoch": 1.5579880896276936,
      "grad_norm": 5.595006942749023,
      "learning_rate": 7.78962572312757e-05,
      "loss": 2.6134761810302733,
      "memory(GiB)": 77.56,
      "step": 36365,
      "token_acc": 0.5016611295681063,
      "train_speed(iter/s)": 1.440414
    },
    {
      "epoch": 1.5582023049569427,
      "grad_norm": 4.773839950561523,
      "learning_rate": 7.78906719986561e-05,
      "loss": 2.3932146072387694,
      "memory(GiB)": 77.56,
      "step": 36370,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.440429
    },
    {
      "epoch": 1.5584165202861917,
      "grad_norm": 4.176609516143799,
      "learning_rate": 7.788508626077084e-05,
      "loss": 2.5070213317871093,
      "memory(GiB)": 77.56,
      "step": 36375,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.440411
    },
    {
      "epoch": 1.5586307356154405,
      "grad_norm": 5.443298816680908,
      "learning_rate": 7.787950001772108e-05,
      "loss": 2.4655118942260743,
      "memory(GiB)": 77.56,
      "step": 36380,
      "token_acc": 0.5085910652920962,
      "train_speed(iter/s)": 1.440429
    },
    {
      "epoch": 1.5588449509446896,
      "grad_norm": 4.701830863952637,
      "learning_rate": 7.7873913269608e-05,
      "loss": 2.3925262451171876,
      "memory(GiB)": 77.56,
      "step": 36385,
      "token_acc": 0.5035211267605634,
      "train_speed(iter/s)": 1.440401
    },
    {
      "epoch": 1.5590591662739386,
      "grad_norm": 6.884942054748535,
      "learning_rate": 7.786832601653287e-05,
      "loss": 2.774361419677734,
      "memory(GiB)": 77.56,
      "step": 36390,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.440403
    },
    {
      "epoch": 1.5592733816031874,
      "grad_norm": 5.308371543884277,
      "learning_rate": 7.786273825859684e-05,
      "loss": 2.405905914306641,
      "memory(GiB)": 77.56,
      "step": 36395,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.44045
    },
    {
      "epoch": 1.5594875969324364,
      "grad_norm": 4.704966068267822,
      "learning_rate": 7.78571499959012e-05,
      "loss": 2.5021812438964846,
      "memory(GiB)": 77.56,
      "step": 36400,
      "token_acc": 0.49673202614379086,
      "train_speed(iter/s)": 1.440444
    },
    {
      "epoch": 1.5597018122616855,
      "grad_norm": 4.557310104370117,
      "learning_rate": 7.785156122854713e-05,
      "loss": 2.3835289001464846,
      "memory(GiB)": 77.56,
      "step": 36405,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.5599160275909343,
      "grad_norm": 4.9684553146362305,
      "learning_rate": 7.784597195663593e-05,
      "loss": 2.506352996826172,
      "memory(GiB)": 77.56,
      "step": 36410,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.440518
    },
    {
      "epoch": 1.5601302429201833,
      "grad_norm": 5.420516014099121,
      "learning_rate": 7.78403821802688e-05,
      "loss": 2.4266332626342773,
      "memory(GiB)": 77.56,
      "step": 36415,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.440486
    },
    {
      "epoch": 1.5603444582494324,
      "grad_norm": 4.634389877319336,
      "learning_rate": 7.783479189954704e-05,
      "loss": 2.4806692123413088,
      "memory(GiB)": 77.56,
      "step": 36420,
      "token_acc": 0.48264984227129337,
      "train_speed(iter/s)": 1.4405
    },
    {
      "epoch": 1.5605586735786812,
      "grad_norm": 4.9568071365356445,
      "learning_rate": 7.78292011145719e-05,
      "loss": 2.3620481491088867,
      "memory(GiB)": 77.56,
      "step": 36425,
      "token_acc": 0.4934640522875817,
      "train_speed(iter/s)": 1.440537
    },
    {
      "epoch": 1.5607728889079302,
      "grad_norm": 4.540524005889893,
      "learning_rate": 7.782360982544469e-05,
      "loss": 2.5699893951416017,
      "memory(GiB)": 77.56,
      "step": 36430,
      "token_acc": 0.44554455445544555,
      "train_speed(iter/s)": 1.440565
    },
    {
      "epoch": 1.5609871042371792,
      "grad_norm": 5.17749547958374,
      "learning_rate": 7.781801803226669e-05,
      "loss": 2.657927894592285,
      "memory(GiB)": 77.56,
      "step": 36435,
      "token_acc": 0.4492307692307692,
      "train_speed(iter/s)": 1.440589
    },
    {
      "epoch": 1.561201319566428,
      "grad_norm": 5.044076919555664,
      "learning_rate": 7.781242573513918e-05,
      "loss": 2.407718849182129,
      "memory(GiB)": 77.56,
      "step": 36440,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.440634
    },
    {
      "epoch": 1.561415534895677,
      "grad_norm": 7.126518726348877,
      "learning_rate": 7.780683293416352e-05,
      "loss": 2.568478012084961,
      "memory(GiB)": 77.56,
      "step": 36445,
      "token_acc": 0.5015974440894568,
      "train_speed(iter/s)": 1.440654
    },
    {
      "epoch": 1.5616297502249261,
      "grad_norm": 4.720685958862305,
      "learning_rate": 7.780123962944097e-05,
      "loss": 2.646109771728516,
      "memory(GiB)": 77.56,
      "step": 36450,
      "token_acc": 0.44077134986225897,
      "train_speed(iter/s)": 1.440701
    },
    {
      "epoch": 1.561843965554175,
      "grad_norm": 5.71162748336792,
      "learning_rate": 7.779564582107289e-05,
      "loss": 2.2690031051635744,
      "memory(GiB)": 77.56,
      "step": 36455,
      "token_acc": 0.540268456375839,
      "train_speed(iter/s)": 1.440715
    },
    {
      "epoch": 1.562058180883424,
      "grad_norm": 5.233338832855225,
      "learning_rate": 7.779005150916061e-05,
      "loss": 2.619420051574707,
      "memory(GiB)": 77.56,
      "step": 36460,
      "token_acc": 0.464,
      "train_speed(iter/s)": 1.440758
    },
    {
      "epoch": 1.562272396212673,
      "grad_norm": 4.673311233520508,
      "learning_rate": 7.778445669380548e-05,
      "loss": 2.5446447372436523,
      "memory(GiB)": 77.56,
      "step": 36465,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.44077
    },
    {
      "epoch": 1.5624866115419218,
      "grad_norm": 6.547863006591797,
      "learning_rate": 7.777886137510885e-05,
      "loss": 2.765332794189453,
      "memory(GiB)": 77.56,
      "step": 36470,
      "token_acc": 0.43272727272727274,
      "train_speed(iter/s)": 1.440766
    },
    {
      "epoch": 1.562700826871171,
      "grad_norm": 4.185351371765137,
      "learning_rate": 7.777326555317209e-05,
      "loss": 2.795298957824707,
      "memory(GiB)": 77.56,
      "step": 36475,
      "token_acc": 0.42628205128205127,
      "train_speed(iter/s)": 1.440779
    },
    {
      "epoch": 1.5629150422004199,
      "grad_norm": 10.662088394165039,
      "learning_rate": 7.776766922809658e-05,
      "loss": 2.4162115097045898,
      "memory(GiB)": 77.56,
      "step": 36480,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.440775
    },
    {
      "epoch": 1.5631292575296687,
      "grad_norm": 5.266146659851074,
      "learning_rate": 7.776207239998368e-05,
      "loss": 2.4303817749023438,
      "memory(GiB)": 77.56,
      "step": 36485,
      "token_acc": 0.4981549815498155,
      "train_speed(iter/s)": 1.440822
    },
    {
      "epoch": 1.563343472858918,
      "grad_norm": 6.210820198059082,
      "learning_rate": 7.775647506893481e-05,
      "loss": 2.5056489944458007,
      "memory(GiB)": 77.56,
      "step": 36490,
      "token_acc": 0.46405228758169936,
      "train_speed(iter/s)": 1.440864
    },
    {
      "epoch": 1.5635576881881668,
      "grad_norm": 5.352214813232422,
      "learning_rate": 7.775087723505133e-05,
      "loss": 2.8598777770996096,
      "memory(GiB)": 77.56,
      "step": 36495,
      "token_acc": 0.41597796143250687,
      "train_speed(iter/s)": 1.440884
    },
    {
      "epoch": 1.5637719035174156,
      "grad_norm": 5.598677635192871,
      "learning_rate": 7.774527889843471e-05,
      "loss": 2.8109331130981445,
      "memory(GiB)": 77.56,
      "step": 36500,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.440888
    },
    {
      "epoch": 1.5637719035174156,
      "eval_loss": 2.434030055999756,
      "eval_runtime": 13.3766,
      "eval_samples_per_second": 7.476,
      "eval_steps_per_second": 7.476,
      "eval_token_acc": 0.4772413793103448,
      "step": 36500
    },
    {
      "epoch": 1.5639861188466648,
      "grad_norm": 4.27955436706543,
      "learning_rate": 7.773968005918631e-05,
      "loss": 2.3682605743408205,
      "memory(GiB)": 77.56,
      "step": 36505,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.440061
    },
    {
      "epoch": 1.5642003341759136,
      "grad_norm": 5.812290191650391,
      "learning_rate": 7.77340807174076e-05,
      "loss": 2.694140815734863,
      "memory(GiB)": 77.56,
      "step": 36510,
      "token_acc": 0.4685714285714286,
      "train_speed(iter/s)": 1.440051
    },
    {
      "epoch": 1.5644145495051625,
      "grad_norm": 4.478960990905762,
      "learning_rate": 7.77284808732e-05,
      "loss": 2.4436033248901365,
      "memory(GiB)": 77.56,
      "step": 36515,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.440059
    },
    {
      "epoch": 1.5646287648344117,
      "grad_norm": 6.991723537445068,
      "learning_rate": 7.772288052666494e-05,
      "loss": 2.541650390625,
      "memory(GiB)": 77.56,
      "step": 36520,
      "token_acc": 0.4414715719063545,
      "train_speed(iter/s)": 1.440037
    },
    {
      "epoch": 1.5648429801636605,
      "grad_norm": 5.616585731506348,
      "learning_rate": 7.771727967790393e-05,
      "loss": 2.5818508148193358,
      "memory(GiB)": 77.56,
      "step": 36525,
      "token_acc": 0.40492957746478875,
      "train_speed(iter/s)": 1.440072
    },
    {
      "epoch": 1.5650571954929093,
      "grad_norm": 5.703603267669678,
      "learning_rate": 7.771167832701835e-05,
      "loss": 2.3734350204467773,
      "memory(GiB)": 77.56,
      "step": 36530,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.440074
    },
    {
      "epoch": 1.5652714108221586,
      "grad_norm": 4.5755414962768555,
      "learning_rate": 7.770607647410975e-05,
      "loss": 2.5518375396728517,
      "memory(GiB)": 77.56,
      "step": 36535,
      "token_acc": 0.453416149068323,
      "train_speed(iter/s)": 1.440087
    },
    {
      "epoch": 1.5654856261514074,
      "grad_norm": 7.752872943878174,
      "learning_rate": 7.770047411927958e-05,
      "loss": 2.6195579528808595,
      "memory(GiB)": 77.56,
      "step": 36540,
      "token_acc": 0.47079037800687284,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 1.5656998414806562,
      "grad_norm": 4.654615879058838,
      "learning_rate": 7.769487126262934e-05,
      "loss": 2.446702766418457,
      "memory(GiB)": 77.56,
      "step": 36545,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440033
    },
    {
      "epoch": 1.5659140568099055,
      "grad_norm": 5.9365715980529785,
      "learning_rate": 7.768926790426052e-05,
      "loss": 2.5963687896728516,
      "memory(GiB)": 77.56,
      "step": 36550,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.440033
    },
    {
      "epoch": 1.5661282721391543,
      "grad_norm": 5.438448429107666,
      "learning_rate": 7.768366404427464e-05,
      "loss": 2.5643732070922853,
      "memory(GiB)": 77.56,
      "step": 36555,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.440063
    },
    {
      "epoch": 1.566342487468403,
      "grad_norm": 5.707797527313232,
      "learning_rate": 7.767805968277322e-05,
      "loss": 2.591194725036621,
      "memory(GiB)": 77.56,
      "step": 36560,
      "token_acc": 0.4131054131054131,
      "train_speed(iter/s)": 1.440102
    },
    {
      "epoch": 1.5665567027976524,
      "grad_norm": 5.371236324310303,
      "learning_rate": 7.767245481985777e-05,
      "loss": 2.7137020111083983,
      "memory(GiB)": 77.56,
      "step": 36565,
      "token_acc": 0.4354243542435424,
      "train_speed(iter/s)": 1.440126
    },
    {
      "epoch": 1.5667709181269012,
      "grad_norm": 4.493969440460205,
      "learning_rate": 7.766684945562986e-05,
      "loss": 2.3718088150024412,
      "memory(GiB)": 77.56,
      "step": 36570,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.440162
    },
    {
      "epoch": 1.56698513345615,
      "grad_norm": 4.753102779388428,
      "learning_rate": 7.7661243590191e-05,
      "loss": 2.6996337890625,
      "memory(GiB)": 77.56,
      "step": 36575,
      "token_acc": 0.45396825396825397,
      "train_speed(iter/s)": 1.440123
    },
    {
      "epoch": 1.5671993487853992,
      "grad_norm": 5.907541751861572,
      "learning_rate": 7.765563722364278e-05,
      "loss": 2.4066436767578123,
      "memory(GiB)": 77.56,
      "step": 36580,
      "token_acc": 0.5369649805447471,
      "train_speed(iter/s)": 1.440123
    },
    {
      "epoch": 1.567413564114648,
      "grad_norm": 5.512517929077148,
      "learning_rate": 7.765003035608676e-05,
      "loss": 2.658072280883789,
      "memory(GiB)": 77.56,
      "step": 36585,
      "token_acc": 0.46715328467153283,
      "train_speed(iter/s)": 1.440146
    },
    {
      "epoch": 1.5676277794438969,
      "grad_norm": 4.607879161834717,
      "learning_rate": 7.764442298762448e-05,
      "loss": 2.757480430603027,
      "memory(GiB)": 77.56,
      "step": 36590,
      "token_acc": 0.4185303514376997,
      "train_speed(iter/s)": 1.440148
    },
    {
      "epoch": 1.5678419947731461,
      "grad_norm": 4.5528645515441895,
      "learning_rate": 7.763881511835754e-05,
      "loss": 2.548477554321289,
      "memory(GiB)": 77.56,
      "step": 36595,
      "token_acc": 0.47540983606557374,
      "train_speed(iter/s)": 1.440204
    },
    {
      "epoch": 1.568056210102395,
      "grad_norm": 4.772520542144775,
      "learning_rate": 7.763320674838756e-05,
      "loss": 2.4298946380615236,
      "memory(GiB)": 77.56,
      "step": 36600,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.440134
    },
    {
      "epoch": 1.5682704254316437,
      "grad_norm": 5.726729869842529,
      "learning_rate": 7.76275978778161e-05,
      "loss": 2.635542869567871,
      "memory(GiB)": 77.56,
      "step": 36605,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.440181
    },
    {
      "epoch": 1.568484640760893,
      "grad_norm": 4.152476787567139,
      "learning_rate": 7.762198850674478e-05,
      "loss": 2.5780864715576173,
      "memory(GiB)": 77.56,
      "step": 36610,
      "token_acc": 0.43234323432343236,
      "train_speed(iter/s)": 1.440193
    },
    {
      "epoch": 1.5686988560901418,
      "grad_norm": 5.2029709815979,
      "learning_rate": 7.761637863527524e-05,
      "loss": 2.6104827880859376,
      "memory(GiB)": 77.56,
      "step": 36615,
      "token_acc": 0.44274809160305345,
      "train_speed(iter/s)": 1.440192
    },
    {
      "epoch": 1.5689130714193906,
      "grad_norm": 4.238569259643555,
      "learning_rate": 7.761076826350911e-05,
      "loss": 2.3544130325317383,
      "memory(GiB)": 77.56,
      "step": 36620,
      "token_acc": 0.5209790209790209,
      "train_speed(iter/s)": 1.440194
    },
    {
      "epoch": 1.5691272867486399,
      "grad_norm": 4.985941410064697,
      "learning_rate": 7.760515739154798e-05,
      "loss": 2.8340641021728517,
      "memory(GiB)": 77.56,
      "step": 36625,
      "token_acc": 0.45544554455445546,
      "train_speed(iter/s)": 1.440245
    },
    {
      "epoch": 1.5693415020778887,
      "grad_norm": 4.0189313888549805,
      "learning_rate": 7.759954601949355e-05,
      "loss": 2.4815437316894533,
      "memory(GiB)": 77.56,
      "step": 36630,
      "token_acc": 0.5259938837920489,
      "train_speed(iter/s)": 1.440238
    },
    {
      "epoch": 1.5695557174071375,
      "grad_norm": 4.385519027709961,
      "learning_rate": 7.759393414744747e-05,
      "loss": 2.5908279418945312,
      "memory(GiB)": 77.56,
      "step": 36635,
      "token_acc": 0.46963562753036436,
      "train_speed(iter/s)": 1.440237
    },
    {
      "epoch": 1.5697699327363868,
      "grad_norm": 7.434064865112305,
      "learning_rate": 7.758832177551136e-05,
      "loss": 2.553477478027344,
      "memory(GiB)": 77.56,
      "step": 36640,
      "token_acc": 0.5328185328185329,
      "train_speed(iter/s)": 1.440221
    },
    {
      "epoch": 1.5699841480656356,
      "grad_norm": 6.4770402908325195,
      "learning_rate": 7.758270890378691e-05,
      "loss": 2.514266014099121,
      "memory(GiB)": 77.56,
      "step": 36645,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.440182
    },
    {
      "epoch": 1.5701983633948844,
      "grad_norm": 4.651577949523926,
      "learning_rate": 7.757709553237584e-05,
      "loss": 2.2985984802246096,
      "memory(GiB)": 77.56,
      "step": 36650,
      "token_acc": 0.5381526104417671,
      "train_speed(iter/s)": 1.440192
    },
    {
      "epoch": 1.5704125787241336,
      "grad_norm": 8.120918273925781,
      "learning_rate": 7.757148166137981e-05,
      "loss": 2.3862056732177734,
      "memory(GiB)": 77.56,
      "step": 36655,
      "token_acc": 0.5095785440613027,
      "train_speed(iter/s)": 1.440209
    },
    {
      "epoch": 1.5706267940533825,
      "grad_norm": 5.80086612701416,
      "learning_rate": 7.756586729090052e-05,
      "loss": 2.622327995300293,
      "memory(GiB)": 77.56,
      "step": 36660,
      "token_acc": 0.43260188087774293,
      "train_speed(iter/s)": 1.440226
    },
    {
      "epoch": 1.5708410093826313,
      "grad_norm": 3.806565523147583,
      "learning_rate": 7.756025242103969e-05,
      "loss": 2.404244041442871,
      "memory(GiB)": 77.56,
      "step": 36665,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.440241
    },
    {
      "epoch": 1.5710552247118805,
      "grad_norm": 4.028979778289795,
      "learning_rate": 7.755463705189902e-05,
      "loss": 2.401742935180664,
      "memory(GiB)": 77.56,
      "step": 36670,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440255
    },
    {
      "epoch": 1.5712694400411293,
      "grad_norm": 6.583826065063477,
      "learning_rate": 7.754902118358027e-05,
      "loss": 2.2654064178466795,
      "memory(GiB)": 77.56,
      "step": 36675,
      "token_acc": 0.5056603773584906,
      "train_speed(iter/s)": 1.440269
    },
    {
      "epoch": 1.5714836553703782,
      "grad_norm": 4.303641319274902,
      "learning_rate": 7.754340481618514e-05,
      "loss": 2.709935188293457,
      "memory(GiB)": 77.56,
      "step": 36680,
      "token_acc": 0.4568345323741007,
      "train_speed(iter/s)": 1.440287
    },
    {
      "epoch": 1.5716978706996274,
      "grad_norm": 6.314344882965088,
      "learning_rate": 7.753778794981541e-05,
      "loss": 2.5506317138671877,
      "memory(GiB)": 77.56,
      "step": 36685,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.440324
    },
    {
      "epoch": 1.5719120860288762,
      "grad_norm": 5.979341506958008,
      "learning_rate": 7.753217058457281e-05,
      "loss": 2.784853935241699,
      "memory(GiB)": 77.56,
      "step": 36690,
      "token_acc": 0.41118421052631576,
      "train_speed(iter/s)": 1.44035
    },
    {
      "epoch": 1.572126301358125,
      "grad_norm": 5.207551956176758,
      "learning_rate": 7.75265527205591e-05,
      "loss": 2.4602256774902345,
      "memory(GiB)": 77.56,
      "step": 36695,
      "token_acc": 0.4785276073619632,
      "train_speed(iter/s)": 1.440334
    },
    {
      "epoch": 1.5723405166873743,
      "grad_norm": 4.366931915283203,
      "learning_rate": 7.752093435787611e-05,
      "loss": 2.306169891357422,
      "memory(GiB)": 77.56,
      "step": 36700,
      "token_acc": 0.4820717131474104,
      "train_speed(iter/s)": 1.440357
    },
    {
      "epoch": 1.572554732016623,
      "grad_norm": 5.995160102844238,
      "learning_rate": 7.751531549662553e-05,
      "loss": 2.157869338989258,
      "memory(GiB)": 77.56,
      "step": 36705,
      "token_acc": 0.5418326693227091,
      "train_speed(iter/s)": 1.440393
    },
    {
      "epoch": 1.572768947345872,
      "grad_norm": 5.33212947845459,
      "learning_rate": 7.750969613690923e-05,
      "loss": 2.617536926269531,
      "memory(GiB)": 77.56,
      "step": 36710,
      "token_acc": 0.4584837545126354,
      "train_speed(iter/s)": 1.440398
    },
    {
      "epoch": 1.5729831626751212,
      "grad_norm": 4.435813903808594,
      "learning_rate": 7.750407627882897e-05,
      "loss": 2.7557437896728514,
      "memory(GiB)": 77.56,
      "step": 36715,
      "token_acc": 0.4289855072463768,
      "train_speed(iter/s)": 1.440371
    },
    {
      "epoch": 1.57319737800437,
      "grad_norm": 5.804585933685303,
      "learning_rate": 7.749845592248654e-05,
      "loss": 2.260823440551758,
      "memory(GiB)": 77.56,
      "step": 36720,
      "token_acc": 0.5354330708661418,
      "train_speed(iter/s)": 1.440395
    },
    {
      "epoch": 1.5734115933336188,
      "grad_norm": 4.43022346496582,
      "learning_rate": 7.749283506798382e-05,
      "loss": 2.3514469146728514,
      "memory(GiB)": 77.56,
      "step": 36725,
      "token_acc": 0.5179282868525896,
      "train_speed(iter/s)": 1.440409
    },
    {
      "epoch": 1.573625808662868,
      "grad_norm": 6.361043453216553,
      "learning_rate": 7.748721371542258e-05,
      "loss": 2.709575080871582,
      "memory(GiB)": 77.56,
      "step": 36730,
      "token_acc": 0.40168539325842695,
      "train_speed(iter/s)": 1.440419
    },
    {
      "epoch": 1.5738400239921169,
      "grad_norm": 4.544958591461182,
      "learning_rate": 7.748159186490469e-05,
      "loss": 2.4843658447265624,
      "memory(GiB)": 77.56,
      "step": 36735,
      "token_acc": 0.49691358024691357,
      "train_speed(iter/s)": 1.440473
    },
    {
      "epoch": 1.5740542393213657,
      "grad_norm": 6.759220123291016,
      "learning_rate": 7.747596951653198e-05,
      "loss": 2.5847455978393556,
      "memory(GiB)": 77.56,
      "step": 36740,
      "token_acc": 0.44039735099337746,
      "train_speed(iter/s)": 1.440469
    },
    {
      "epoch": 1.574268454650615,
      "grad_norm": 6.662232398986816,
      "learning_rate": 7.747034667040632e-05,
      "loss": 2.4833652496337892,
      "memory(GiB)": 77.56,
      "step": 36745,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.440438
    },
    {
      "epoch": 1.5744826699798637,
      "grad_norm": 4.657716751098633,
      "learning_rate": 7.746472332662955e-05,
      "loss": 2.6565189361572266,
      "memory(GiB)": 77.56,
      "step": 36750,
      "token_acc": 0.4303030303030303,
      "train_speed(iter/s)": 1.44037
    },
    {
      "epoch": 1.5746968853091126,
      "grad_norm": 5.314342498779297,
      "learning_rate": 7.745909948530355e-05,
      "loss": 2.6523664474487303,
      "memory(GiB)": 77.56,
      "step": 36755,
      "token_acc": 0.4600760456273764,
      "train_speed(iter/s)": 1.440391
    },
    {
      "epoch": 1.5749111006383618,
      "grad_norm": 10.755656242370605,
      "learning_rate": 7.745347514653021e-05,
      "loss": 2.4207244873046876,
      "memory(GiB)": 77.56,
      "step": 36760,
      "token_acc": 0.5261194029850746,
      "train_speed(iter/s)": 1.440426
    },
    {
      "epoch": 1.5751253159676106,
      "grad_norm": 6.14573860168457,
      "learning_rate": 7.74478503104114e-05,
      "loss": 2.6705108642578126,
      "memory(GiB)": 77.56,
      "step": 36765,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.44046
    },
    {
      "epoch": 1.5753395312968594,
      "grad_norm": 5.781991481781006,
      "learning_rate": 7.744222497704904e-05,
      "loss": 2.738389778137207,
      "memory(GiB)": 77.56,
      "step": 36770,
      "token_acc": 0.45357142857142857,
      "train_speed(iter/s)": 1.440452
    },
    {
      "epoch": 1.5755537466261087,
      "grad_norm": 7.469683647155762,
      "learning_rate": 7.743659914654504e-05,
      "loss": 2.799342727661133,
      "memory(GiB)": 77.56,
      "step": 36775,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.440475
    },
    {
      "epoch": 1.5757679619553575,
      "grad_norm": 5.489133358001709,
      "learning_rate": 7.743097281900131e-05,
      "loss": 2.57935791015625,
      "memory(GiB)": 77.56,
      "step": 36780,
      "token_acc": 0.44648318042813456,
      "train_speed(iter/s)": 1.440493
    },
    {
      "epoch": 1.5759821772846063,
      "grad_norm": 5.444052696228027,
      "learning_rate": 7.742534599451978e-05,
      "loss": 2.4359161376953127,
      "memory(GiB)": 77.56,
      "step": 36785,
      "token_acc": 0.46546546546546547,
      "train_speed(iter/s)": 1.440507
    },
    {
      "epoch": 1.5761963926138556,
      "grad_norm": 6.023439407348633,
      "learning_rate": 7.741971867320237e-05,
      "loss": 2.485719108581543,
      "memory(GiB)": 77.56,
      "step": 36790,
      "token_acc": 0.4699248120300752,
      "train_speed(iter/s)": 1.440523
    },
    {
      "epoch": 1.5764106079431044,
      "grad_norm": 4.417028427124023,
      "learning_rate": 7.741409085515103e-05,
      "loss": 2.799840545654297,
      "memory(GiB)": 77.56,
      "step": 36795,
      "token_acc": 0.4344569288389513,
      "train_speed(iter/s)": 1.440569
    },
    {
      "epoch": 1.5766248232723532,
      "grad_norm": 4.624118328094482,
      "learning_rate": 7.740846254046772e-05,
      "loss": 2.6378807067871093,
      "memory(GiB)": 77.56,
      "step": 36800,
      "token_acc": 0.4147727272727273,
      "train_speed(iter/s)": 1.440559
    },
    {
      "epoch": 1.5768390386016025,
      "grad_norm": 4.87163782119751,
      "learning_rate": 7.74028337292544e-05,
      "loss": 2.703288459777832,
      "memory(GiB)": 77.56,
      "step": 36805,
      "token_acc": 0.4440677966101695,
      "train_speed(iter/s)": 1.440602
    },
    {
      "epoch": 1.5770532539308513,
      "grad_norm": 4.161083698272705,
      "learning_rate": 7.739720442161305e-05,
      "loss": 2.446087646484375,
      "memory(GiB)": 77.56,
      "step": 36810,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440603
    },
    {
      "epoch": 1.5772674692601003,
      "grad_norm": 4.400523662567139,
      "learning_rate": 7.739157461764564e-05,
      "loss": 2.5423789978027345,
      "memory(GiB)": 77.56,
      "step": 36815,
      "token_acc": 0.49240121580547114,
      "train_speed(iter/s)": 1.440575
    },
    {
      "epoch": 1.5774816845893493,
      "grad_norm": 4.913524150848389,
      "learning_rate": 7.738594431745415e-05,
      "loss": 2.7994462966918947,
      "memory(GiB)": 77.56,
      "step": 36820,
      "token_acc": 0.4529616724738676,
      "train_speed(iter/s)": 1.440551
    },
    {
      "epoch": 1.5776958999185982,
      "grad_norm": 4.85852575302124,
      "learning_rate": 7.738031352114059e-05,
      "loss": 2.4869850158691404,
      "memory(GiB)": 77.56,
      "step": 36825,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.440579
    },
    {
      "epoch": 1.5779101152478472,
      "grad_norm": 5.648110389709473,
      "learning_rate": 7.737468222880697e-05,
      "loss": 2.9752532958984377,
      "memory(GiB)": 77.56,
      "step": 36830,
      "token_acc": 0.40181268882175225,
      "train_speed(iter/s)": 1.440608
    },
    {
      "epoch": 1.5781243305770962,
      "grad_norm": 5.381869316101074,
      "learning_rate": 7.736905044055531e-05,
      "loss": 2.3726613998413084,
      "memory(GiB)": 77.56,
      "step": 36835,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.440592
    },
    {
      "epoch": 1.578338545906345,
      "grad_norm": 4.67322301864624,
      "learning_rate": 7.736341815648761e-05,
      "loss": 2.458905029296875,
      "memory(GiB)": 77.56,
      "step": 36840,
      "token_acc": 0.5088495575221239,
      "train_speed(iter/s)": 1.4406
    },
    {
      "epoch": 1.578552761235594,
      "grad_norm": 4.657018184661865,
      "learning_rate": 7.735778537670594e-05,
      "loss": 2.5954566955566407,
      "memory(GiB)": 77.56,
      "step": 36845,
      "token_acc": 0.4651898734177215,
      "train_speed(iter/s)": 1.440571
    },
    {
      "epoch": 1.578766976564843,
      "grad_norm": 4.841012954711914,
      "learning_rate": 7.73521521013123e-05,
      "loss": 2.2852815628051757,
      "memory(GiB)": 77.56,
      "step": 36850,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.440566
    },
    {
      "epoch": 1.578981191894092,
      "grad_norm": 5.37290620803833,
      "learning_rate": 7.734651833040879e-05,
      "loss": 2.6469724655151365,
      "memory(GiB)": 77.56,
      "step": 36855,
      "token_acc": 0.4565826330532213,
      "train_speed(iter/s)": 1.44059
    },
    {
      "epoch": 1.579195407223341,
      "grad_norm": 5.4792256355285645,
      "learning_rate": 7.734088406409744e-05,
      "loss": 2.4698020935058596,
      "memory(GiB)": 77.56,
      "step": 36860,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.440585
    },
    {
      "epoch": 1.57940962255259,
      "grad_norm": 7.3147358894348145,
      "learning_rate": 7.733524930248032e-05,
      "loss": 2.678775405883789,
      "memory(GiB)": 77.56,
      "step": 36865,
      "token_acc": 0.44244604316546765,
      "train_speed(iter/s)": 1.440553
    },
    {
      "epoch": 1.5796238378818388,
      "grad_norm": 4.623478412628174,
      "learning_rate": 7.732961404565953e-05,
      "loss": 2.620505523681641,
      "memory(GiB)": 77.56,
      "step": 36870,
      "token_acc": 0.4715447154471545,
      "train_speed(iter/s)": 1.440502
    },
    {
      "epoch": 1.5798380532110878,
      "grad_norm": 6.907497882843018,
      "learning_rate": 7.732397829373713e-05,
      "loss": 2.7558525085449217,
      "memory(GiB)": 77.56,
      "step": 36875,
      "token_acc": 0.4358108108108108,
      "train_speed(iter/s)": 1.440506
    },
    {
      "epoch": 1.5800522685403369,
      "grad_norm": 3.7904212474823,
      "learning_rate": 7.731834204681522e-05,
      "loss": 2.2528480529785155,
      "memory(GiB)": 77.56,
      "step": 36880,
      "token_acc": 0.5232198142414861,
      "train_speed(iter/s)": 1.440499
    },
    {
      "epoch": 1.5802664838695857,
      "grad_norm": 3.7137997150421143,
      "learning_rate": 7.731270530499595e-05,
      "loss": 2.8632713317871095,
      "memory(GiB)": 77.56,
      "step": 36885,
      "token_acc": 0.4157608695652174,
      "train_speed(iter/s)": 1.440531
    },
    {
      "epoch": 1.5804806991988347,
      "grad_norm": 6.536667823791504,
      "learning_rate": 7.73070680683814e-05,
      "loss": 2.6188026428222657,
      "memory(GiB)": 77.56,
      "step": 36890,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.44055
    },
    {
      "epoch": 1.5806949145280837,
      "grad_norm": 4.936042785644531,
      "learning_rate": 7.730143033707367e-05,
      "loss": 2.4435157775878906,
      "memory(GiB)": 77.56,
      "step": 36895,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.440537
    },
    {
      "epoch": 1.5809091298573326,
      "grad_norm": 5.053994178771973,
      "learning_rate": 7.729579211117492e-05,
      "loss": 2.538191223144531,
      "memory(GiB)": 77.56,
      "step": 36900,
      "token_acc": 0.47147147147147145,
      "train_speed(iter/s)": 1.440549
    },
    {
      "epoch": 1.5811233451865816,
      "grad_norm": 4.722520351409912,
      "learning_rate": 7.729015339078731e-05,
      "loss": 2.4507837295532227,
      "memory(GiB)": 77.56,
      "step": 36905,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.440552
    },
    {
      "epoch": 1.5813375605158306,
      "grad_norm": 5.030966281890869,
      "learning_rate": 7.728451417601297e-05,
      "loss": 2.3558704376220705,
      "memory(GiB)": 77.56,
      "step": 36910,
      "token_acc": 0.4900398406374502,
      "train_speed(iter/s)": 1.440558
    },
    {
      "epoch": 1.5815517758450794,
      "grad_norm": 4.461679458618164,
      "learning_rate": 7.727887446695405e-05,
      "loss": 2.3357072830200196,
      "memory(GiB)": 77.56,
      "step": 36915,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.440559
    },
    {
      "epoch": 1.5817659911743285,
      "grad_norm": 4.808989524841309,
      "learning_rate": 7.727323426371272e-05,
      "loss": 2.5729959487915037,
      "memory(GiB)": 77.56,
      "step": 36920,
      "token_acc": 0.4186046511627907,
      "train_speed(iter/s)": 1.440565
    },
    {
      "epoch": 1.5819802065035775,
      "grad_norm": 4.341703414916992,
      "learning_rate": 7.726759356639119e-05,
      "loss": 2.516896438598633,
      "memory(GiB)": 77.56,
      "step": 36925,
      "token_acc": 0.47038327526132406,
      "train_speed(iter/s)": 1.440602
    },
    {
      "epoch": 1.5821944218328263,
      "grad_norm": 4.133007049560547,
      "learning_rate": 7.726195237509162e-05,
      "loss": 2.483944320678711,
      "memory(GiB)": 77.56,
      "step": 36930,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.440615
    },
    {
      "epoch": 1.5824086371620754,
      "grad_norm": 4.523614406585693,
      "learning_rate": 7.725631068991618e-05,
      "loss": 2.527436065673828,
      "memory(GiB)": 77.56,
      "step": 36935,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.440631
    },
    {
      "epoch": 1.5826228524913244,
      "grad_norm": 6.269434928894043,
      "learning_rate": 7.725066851096714e-05,
      "loss": 2.2298418045043946,
      "memory(GiB)": 77.56,
      "step": 36940,
      "token_acc": 0.5309734513274337,
      "train_speed(iter/s)": 1.440639
    },
    {
      "epoch": 1.5828370678205732,
      "grad_norm": 5.9857940673828125,
      "learning_rate": 7.724502583834665e-05,
      "loss": 2.4994083404541017,
      "memory(GiB)": 77.56,
      "step": 36945,
      "token_acc": 0.4404332129963899,
      "train_speed(iter/s)": 1.44069
    },
    {
      "epoch": 1.5830512831498222,
      "grad_norm": 4.439908504486084,
      "learning_rate": 7.723938267215698e-05,
      "loss": 2.365422821044922,
      "memory(GiB)": 77.56,
      "step": 36950,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.440707
    },
    {
      "epoch": 1.5832654984790713,
      "grad_norm": 7.594779014587402,
      "learning_rate": 7.723373901250032e-05,
      "loss": 2.384699058532715,
      "memory(GiB)": 77.56,
      "step": 36955,
      "token_acc": 0.47388059701492535,
      "train_speed(iter/s)": 1.44075
    },
    {
      "epoch": 1.58347971380832,
      "grad_norm": 4.222997188568115,
      "learning_rate": 7.722809485947895e-05,
      "loss": 2.5913795471191405,
      "memory(GiB)": 77.56,
      "step": 36960,
      "token_acc": 0.4696132596685083,
      "train_speed(iter/s)": 1.440814
    },
    {
      "epoch": 1.5836939291375691,
      "grad_norm": 5.102694988250732,
      "learning_rate": 7.72224502131951e-05,
      "loss": 2.9261005401611326,
      "memory(GiB)": 77.56,
      "step": 36965,
      "token_acc": 0.41237113402061853,
      "train_speed(iter/s)": 1.440781
    },
    {
      "epoch": 1.5839081444668182,
      "grad_norm": 5.263223171234131,
      "learning_rate": 7.721680507375102e-05,
      "loss": 2.398607063293457,
      "memory(GiB)": 77.56,
      "step": 36970,
      "token_acc": 0.4702549575070821,
      "train_speed(iter/s)": 1.440761
    },
    {
      "epoch": 1.584122359796067,
      "grad_norm": 5.182840347290039,
      "learning_rate": 7.721115944124897e-05,
      "loss": 2.590830421447754,
      "memory(GiB)": 77.56,
      "step": 36975,
      "token_acc": 0.4351145038167939,
      "train_speed(iter/s)": 1.440786
    },
    {
      "epoch": 1.584336575125316,
      "grad_norm": 5.2019805908203125,
      "learning_rate": 7.720551331579126e-05,
      "loss": 2.2415897369384767,
      "memory(GiB)": 77.56,
      "step": 36980,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.440753
    },
    {
      "epoch": 1.584550790454565,
      "grad_norm": 6.639218330383301,
      "learning_rate": 7.719986669748013e-05,
      "loss": 2.4859216690063475,
      "memory(GiB)": 77.56,
      "step": 36985,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.440779
    },
    {
      "epoch": 1.5847650057838139,
      "grad_norm": 4.705714702606201,
      "learning_rate": 7.719421958641794e-05,
      "loss": 2.3174625396728517,
      "memory(GiB)": 77.56,
      "step": 36990,
      "token_acc": 0.5232558139534884,
      "train_speed(iter/s)": 1.440825
    },
    {
      "epoch": 1.5849792211130629,
      "grad_norm": 4.4780378341674805,
      "learning_rate": 7.718857198270692e-05,
      "loss": 2.438267135620117,
      "memory(GiB)": 77.56,
      "step": 36995,
      "token_acc": 0.4542857142857143,
      "train_speed(iter/s)": 1.440823
    },
    {
      "epoch": 1.585193436442312,
      "grad_norm": 5.398425579071045,
      "learning_rate": 7.718292388644943e-05,
      "loss": 2.4314022064208984,
      "memory(GiB)": 77.56,
      "step": 37000,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.440812
    },
    {
      "epoch": 1.585193436442312,
      "eval_loss": 2.4495582580566406,
      "eval_runtime": 13.9402,
      "eval_samples_per_second": 7.174,
      "eval_steps_per_second": 7.174,
      "eval_token_acc": 0.45521472392638035,
      "step": 37000
    },
    {
      "epoch": 1.5854076517715607,
      "grad_norm": 4.466859340667725,
      "learning_rate": 7.717727529774777e-05,
      "loss": 2.8922637939453124,
      "memory(GiB)": 77.56,
      "step": 37005,
      "token_acc": 0.45254833040421794,
      "train_speed(iter/s)": 1.440029
    },
    {
      "epoch": 1.5856218671008098,
      "grad_norm": 4.917857646942139,
      "learning_rate": 7.717162621670427e-05,
      "loss": 2.544180679321289,
      "memory(GiB)": 77.56,
      "step": 37010,
      "token_acc": 0.4489795918367347,
      "train_speed(iter/s)": 1.440025
    },
    {
      "epoch": 1.5858360824300588,
      "grad_norm": 6.8532819747924805,
      "learning_rate": 7.716597664342127e-05,
      "loss": 2.503630447387695,
      "memory(GiB)": 77.56,
      "step": 37015,
      "token_acc": 0.4770992366412214,
      "train_speed(iter/s)": 1.440038
    },
    {
      "epoch": 1.5860502977593076,
      "grad_norm": 5.015791416168213,
      "learning_rate": 7.716032657800113e-05,
      "loss": 2.4324644088745115,
      "memory(GiB)": 77.56,
      "step": 37020,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 1.5862645130885566,
      "grad_norm": 3.747614860534668,
      "learning_rate": 7.715467602054618e-05,
      "loss": 2.057494354248047,
      "memory(GiB)": 77.56,
      "step": 37025,
      "token_acc": 0.521551724137931,
      "train_speed(iter/s)": 1.440075
    },
    {
      "epoch": 1.5864787284178057,
      "grad_norm": 4.066998481750488,
      "learning_rate": 7.714902497115881e-05,
      "loss": 2.9439250946044924,
      "memory(GiB)": 77.56,
      "step": 37030,
      "token_acc": 0.41454545454545455,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 1.5866929437470545,
      "grad_norm": 7.244368553161621,
      "learning_rate": 7.714337342994139e-05,
      "loss": 2.2538969039916994,
      "memory(GiB)": 77.56,
      "step": 37035,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.440105
    },
    {
      "epoch": 1.5869071590763035,
      "grad_norm": 5.2623701095581055,
      "learning_rate": 7.71377213969963e-05,
      "loss": 2.591645050048828,
      "memory(GiB)": 77.56,
      "step": 37040,
      "token_acc": 0.4584527220630373,
      "train_speed(iter/s)": 1.440115
    },
    {
      "epoch": 1.5871213744055526,
      "grad_norm": 4.656369209289551,
      "learning_rate": 7.713206887242592e-05,
      "loss": 2.5355520248413086,
      "memory(GiB)": 77.56,
      "step": 37045,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.44013
    },
    {
      "epoch": 1.5873355897348014,
      "grad_norm": 5.218441009521484,
      "learning_rate": 7.712641585633265e-05,
      "loss": 2.663298225402832,
      "memory(GiB)": 77.56,
      "step": 37050,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.440172
    },
    {
      "epoch": 1.5875498050640504,
      "grad_norm": 5.425658702850342,
      "learning_rate": 7.712076234881893e-05,
      "loss": 2.7011905670166017,
      "memory(GiB)": 77.56,
      "step": 37055,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.440215
    },
    {
      "epoch": 1.5877640203932994,
      "grad_norm": 3.627558946609497,
      "learning_rate": 7.711510834998714e-05,
      "loss": 2.633744239807129,
      "memory(GiB)": 77.56,
      "step": 37060,
      "token_acc": 0.44011976047904194,
      "train_speed(iter/s)": 1.440237
    },
    {
      "epoch": 1.5879782357225483,
      "grad_norm": 5.2490057945251465,
      "learning_rate": 7.710945385993975e-05,
      "loss": 2.6325763702392577,
      "memory(GiB)": 77.56,
      "step": 37065,
      "token_acc": 0.47540983606557374,
      "train_speed(iter/s)": 1.440258
    },
    {
      "epoch": 1.5881924510517973,
      "grad_norm": 4.766805171966553,
      "learning_rate": 7.710379887877917e-05,
      "loss": 2.7200057983398436,
      "memory(GiB)": 77.56,
      "step": 37070,
      "token_acc": 0.4322766570605187,
      "train_speed(iter/s)": 1.440314
    },
    {
      "epoch": 1.5884066663810463,
      "grad_norm": 5.99770975112915,
      "learning_rate": 7.709814340660784e-05,
      "loss": 2.4199867248535156,
      "memory(GiB)": 77.56,
      "step": 37075,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.440341
    },
    {
      "epoch": 1.5886208817102951,
      "grad_norm": 5.70997428894043,
      "learning_rate": 7.709248744352822e-05,
      "loss": 2.1313186645507813,
      "memory(GiB)": 77.56,
      "step": 37080,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.440287
    },
    {
      "epoch": 1.5888350970395442,
      "grad_norm": 5.822385311126709,
      "learning_rate": 7.708683098964275e-05,
      "loss": 2.676686668395996,
      "memory(GiB)": 77.56,
      "step": 37085,
      "token_acc": 0.470404984423676,
      "train_speed(iter/s)": 1.44033
    },
    {
      "epoch": 1.5890493123687932,
      "grad_norm": 4.615672588348389,
      "learning_rate": 7.708117404505397e-05,
      "loss": 2.7133426666259766,
      "memory(GiB)": 77.56,
      "step": 37090,
      "token_acc": 0.4836795252225519,
      "train_speed(iter/s)": 1.440328
    },
    {
      "epoch": 1.589263527698042,
      "grad_norm": 3.886209011077881,
      "learning_rate": 7.707551660986429e-05,
      "loss": 2.5003583908081053,
      "memory(GiB)": 77.56,
      "step": 37095,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.440338
    },
    {
      "epoch": 1.589477743027291,
      "grad_norm": 4.47598123550415,
      "learning_rate": 7.706985868417624e-05,
      "loss": 2.4446216583251954,
      "memory(GiB)": 77.56,
      "step": 37100,
      "token_acc": 0.48828125,
      "train_speed(iter/s)": 1.440392
    },
    {
      "epoch": 1.58969195835654,
      "grad_norm": 6.72584342956543,
      "learning_rate": 7.706420026809232e-05,
      "loss": 2.746270751953125,
      "memory(GiB)": 77.56,
      "step": 37105,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.440368
    },
    {
      "epoch": 1.589906173685789,
      "grad_norm": 5.002017974853516,
      "learning_rate": 7.7058541361715e-05,
      "loss": 2.6312305450439455,
      "memory(GiB)": 77.56,
      "step": 37110,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.440383
    },
    {
      "epoch": 1.590120389015038,
      "grad_norm": 4.537996768951416,
      "learning_rate": 7.705288196514682e-05,
      "loss": 2.2976896286010744,
      "memory(GiB)": 77.56,
      "step": 37115,
      "token_acc": 0.5206611570247934,
      "train_speed(iter/s)": 1.440404
    },
    {
      "epoch": 1.590334604344287,
      "grad_norm": 4.235039234161377,
      "learning_rate": 7.70472220784903e-05,
      "loss": 2.727490997314453,
      "memory(GiB)": 77.56,
      "step": 37120,
      "token_acc": 0.434375,
      "train_speed(iter/s)": 1.440385
    },
    {
      "epoch": 1.5905488196735358,
      "grad_norm": 6.362739086151123,
      "learning_rate": 7.704156170184801e-05,
      "loss": 2.4436103820800783,
      "memory(GiB)": 77.56,
      "step": 37125,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.440422
    },
    {
      "epoch": 1.5907630350027848,
      "grad_norm": 5.3750081062316895,
      "learning_rate": 7.703590083532244e-05,
      "loss": 2.4182289123535154,
      "memory(GiB)": 77.56,
      "step": 37130,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.440456
    },
    {
      "epoch": 1.5909772503320339,
      "grad_norm": 3.729543447494507,
      "learning_rate": 7.703023947901618e-05,
      "loss": 2.382952117919922,
      "memory(GiB)": 77.56,
      "step": 37135,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.440457
    },
    {
      "epoch": 1.5911914656612827,
      "grad_norm": 4.52062463760376,
      "learning_rate": 7.702457763303177e-05,
      "loss": 2.5480127334594727,
      "memory(GiB)": 77.56,
      "step": 37140,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.5914056809905317,
      "grad_norm": 5.185436725616455,
      "learning_rate": 7.701891529747178e-05,
      "loss": 2.525675964355469,
      "memory(GiB)": 77.56,
      "step": 37145,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.44056
    },
    {
      "epoch": 1.5916198963197807,
      "grad_norm": 5.667872905731201,
      "learning_rate": 7.70132524724388e-05,
      "loss": 2.530457878112793,
      "memory(GiB)": 77.56,
      "step": 37150,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440559
    },
    {
      "epoch": 1.5918341116490295,
      "grad_norm": 3.952483892440796,
      "learning_rate": 7.70075891580354e-05,
      "loss": 2.4947715759277345,
      "memory(GiB)": 77.56,
      "step": 37155,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.440569
    },
    {
      "epoch": 1.5920483269782786,
      "grad_norm": 4.715612888336182,
      "learning_rate": 7.70019253543642e-05,
      "loss": 2.5695564270019533,
      "memory(GiB)": 77.56,
      "step": 37160,
      "token_acc": 0.49859943977591037,
      "train_speed(iter/s)": 1.44055
    },
    {
      "epoch": 1.5922625423075276,
      "grad_norm": 4.617494583129883,
      "learning_rate": 7.699626106152778e-05,
      "loss": 2.283112144470215,
      "memory(GiB)": 77.56,
      "step": 37165,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.440578
    },
    {
      "epoch": 1.5924767576367764,
      "grad_norm": 5.440049171447754,
      "learning_rate": 7.699059627962877e-05,
      "loss": 2.57884521484375,
      "memory(GiB)": 77.56,
      "step": 37170,
      "token_acc": 0.484251968503937,
      "train_speed(iter/s)": 1.44055
    },
    {
      "epoch": 1.5926909729660255,
      "grad_norm": 5.097413063049316,
      "learning_rate": 7.698493100876979e-05,
      "loss": 2.749297332763672,
      "memory(GiB)": 77.56,
      "step": 37175,
      "token_acc": 0.45483870967741935,
      "train_speed(iter/s)": 1.440583
    },
    {
      "epoch": 1.5929051882952745,
      "grad_norm": 4.4065752029418945,
      "learning_rate": 7.697926524905348e-05,
      "loss": 2.29630126953125,
      "memory(GiB)": 77.56,
      "step": 37180,
      "token_acc": 0.5033112582781457,
      "train_speed(iter/s)": 1.440615
    },
    {
      "epoch": 1.5931194036245233,
      "grad_norm": 4.410834312438965,
      "learning_rate": 7.697359900058245e-05,
      "loss": 2.6966730117797852,
      "memory(GiB)": 77.56,
      "step": 37185,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.440611
    },
    {
      "epoch": 1.5933336189537723,
      "grad_norm": 6.140238285064697,
      "learning_rate": 7.696793226345939e-05,
      "loss": 3.168558692932129,
      "memory(GiB)": 77.56,
      "step": 37190,
      "token_acc": 0.4336283185840708,
      "train_speed(iter/s)": 1.440654
    },
    {
      "epoch": 1.5935478342830214,
      "grad_norm": 4.942981243133545,
      "learning_rate": 7.696226503778694e-05,
      "loss": 2.480720329284668,
      "memory(GiB)": 77.56,
      "step": 37195,
      "token_acc": 0.4672364672364672,
      "train_speed(iter/s)": 1.440671
    },
    {
      "epoch": 1.5937620496122702,
      "grad_norm": 3.73089861869812,
      "learning_rate": 7.695659732366774e-05,
      "loss": 2.577046775817871,
      "memory(GiB)": 77.56,
      "step": 37200,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440642
    },
    {
      "epoch": 1.5939762649415192,
      "grad_norm": 6.346792221069336,
      "learning_rate": 7.695092912120452e-05,
      "loss": 2.663411331176758,
      "memory(GiB)": 77.56,
      "step": 37205,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.440648
    },
    {
      "epoch": 1.5941904802707683,
      "grad_norm": 5.228428363800049,
      "learning_rate": 7.694526043049995e-05,
      "loss": 2.688984489440918,
      "memory(GiB)": 77.56,
      "step": 37210,
      "token_acc": 0.4467455621301775,
      "train_speed(iter/s)": 1.440592
    },
    {
      "epoch": 1.594404695600017,
      "grad_norm": 3.9997048377990723,
      "learning_rate": 7.693959125165666e-05,
      "loss": 2.43411922454834,
      "memory(GiB)": 77.56,
      "step": 37215,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.440577
    },
    {
      "epoch": 1.594618910929266,
      "grad_norm": 7.999871730804443,
      "learning_rate": 7.693392158477745e-05,
      "loss": 2.6133049011230467,
      "memory(GiB)": 77.56,
      "step": 37220,
      "token_acc": 0.4688427299703264,
      "train_speed(iter/s)": 1.440552
    },
    {
      "epoch": 1.5948331262585151,
      "grad_norm": 6.031138896942139,
      "learning_rate": 7.692825142996498e-05,
      "loss": 2.4414501190185547,
      "memory(GiB)": 77.56,
      "step": 37225,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.44055
    },
    {
      "epoch": 1.595047341587764,
      "grad_norm": 5.841692924499512,
      "learning_rate": 7.692258078732196e-05,
      "loss": 2.834478759765625,
      "memory(GiB)": 77.56,
      "step": 37230,
      "token_acc": 0.4440789473684211,
      "train_speed(iter/s)": 1.440571
    },
    {
      "epoch": 1.595261556917013,
      "grad_norm": 6.838806629180908,
      "learning_rate": 7.69169096569511e-05,
      "loss": 2.682520294189453,
      "memory(GiB)": 77.56,
      "step": 37235,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.440575
    },
    {
      "epoch": 1.595475772246262,
      "grad_norm": 6.088563919067383,
      "learning_rate": 7.691123803895523e-05,
      "loss": 2.2300233840942383,
      "memory(GiB)": 77.56,
      "step": 37240,
      "token_acc": 0.5335968379446641,
      "train_speed(iter/s)": 1.440518
    },
    {
      "epoch": 1.5956899875755108,
      "grad_norm": 4.609649658203125,
      "learning_rate": 7.6905565933437e-05,
      "loss": 2.4188270568847656,
      "memory(GiB)": 77.56,
      "step": 37245,
      "token_acc": 0.4844290657439446,
      "train_speed(iter/s)": 1.440499
    },
    {
      "epoch": 1.5959042029047599,
      "grad_norm": 3.5053088665008545,
      "learning_rate": 7.689989334049923e-05,
      "loss": 2.451124572753906,
      "memory(GiB)": 77.56,
      "step": 37250,
      "token_acc": 0.5148514851485149,
      "train_speed(iter/s)": 1.440535
    },
    {
      "epoch": 1.596118418234009,
      "grad_norm": 5.890113830566406,
      "learning_rate": 7.689422026024464e-05,
      "loss": 2.8184814453125,
      "memory(GiB)": 77.56,
      "step": 37255,
      "token_acc": 0.4626334519572954,
      "train_speed(iter/s)": 1.440544
    },
    {
      "epoch": 1.5963326335632577,
      "grad_norm": 5.272008419036865,
      "learning_rate": 7.688854669277604e-05,
      "loss": 2.3964380264282226,
      "memory(GiB)": 77.56,
      "step": 37260,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.440588
    },
    {
      "epoch": 1.5965468488925068,
      "grad_norm": 5.0255022048950195,
      "learning_rate": 7.688287263819617e-05,
      "loss": 2.526440238952637,
      "memory(GiB)": 77.56,
      "step": 37265,
      "token_acc": 0.4876325088339223,
      "train_speed(iter/s)": 1.440604
    },
    {
      "epoch": 1.5967610642217558,
      "grad_norm": 5.126718997955322,
      "learning_rate": 7.687719809660785e-05,
      "loss": 2.237118148803711,
      "memory(GiB)": 77.56,
      "step": 37270,
      "token_acc": 0.5136186770428015,
      "train_speed(iter/s)": 1.440564
    },
    {
      "epoch": 1.5969752795510046,
      "grad_norm": 4.850007057189941,
      "learning_rate": 7.687152306811388e-05,
      "loss": 2.7526214599609373,
      "memory(GiB)": 77.56,
      "step": 37275,
      "token_acc": 0.4713375796178344,
      "train_speed(iter/s)": 1.440571
    },
    {
      "epoch": 1.5971894948802536,
      "grad_norm": 4.5587687492370605,
      "learning_rate": 7.686584755281708e-05,
      "loss": 2.5548351287841795,
      "memory(GiB)": 77.56,
      "step": 37280,
      "token_acc": 0.43465045592705165,
      "train_speed(iter/s)": 1.440578
    },
    {
      "epoch": 1.5974037102095027,
      "grad_norm": 4.840272903442383,
      "learning_rate": 7.686017155082021e-05,
      "loss": 2.474538803100586,
      "memory(GiB)": 77.56,
      "step": 37285,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.440607
    },
    {
      "epoch": 1.5976179255387515,
      "grad_norm": 7.790668964385986,
      "learning_rate": 7.685449506222616e-05,
      "loss": 2.541324806213379,
      "memory(GiB)": 77.56,
      "step": 37290,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.440612
    },
    {
      "epoch": 1.5978321408680005,
      "grad_norm": 5.4809346199035645,
      "learning_rate": 7.684881808713774e-05,
      "loss": 2.4288475036621096,
      "memory(GiB)": 77.56,
      "step": 37295,
      "token_acc": 0.44936708860759494,
      "train_speed(iter/s)": 1.440603
    },
    {
      "epoch": 1.5980463561972496,
      "grad_norm": 6.058802127838135,
      "learning_rate": 7.684314062565779e-05,
      "loss": 2.76653995513916,
      "memory(GiB)": 77.56,
      "step": 37300,
      "token_acc": 0.41237113402061853,
      "train_speed(iter/s)": 1.440612
    },
    {
      "epoch": 1.5982605715264984,
      "grad_norm": 4.5335693359375,
      "learning_rate": 7.683746267788916e-05,
      "loss": 2.386663627624512,
      "memory(GiB)": 77.56,
      "step": 37305,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.440648
    },
    {
      "epoch": 1.5984747868557474,
      "grad_norm": 5.054959774017334,
      "learning_rate": 7.683178424393472e-05,
      "loss": 2.4031471252441405,
      "memory(GiB)": 77.56,
      "step": 37310,
      "token_acc": 0.5036764705882353,
      "train_speed(iter/s)": 1.440678
    },
    {
      "epoch": 1.5986890021849964,
      "grad_norm": 6.449455738067627,
      "learning_rate": 7.682610532389734e-05,
      "loss": 2.6527030944824217,
      "memory(GiB)": 77.56,
      "step": 37315,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.440739
    },
    {
      "epoch": 1.5989032175142452,
      "grad_norm": 5.217899799346924,
      "learning_rate": 7.68204259178799e-05,
      "loss": 2.2669368743896485,
      "memory(GiB)": 77.56,
      "step": 37320,
      "token_acc": 0.4777327935222672,
      "train_speed(iter/s)": 1.440771
    },
    {
      "epoch": 1.5991174328434943,
      "grad_norm": 4.740453720092773,
      "learning_rate": 7.681474602598529e-05,
      "loss": 2.6832502365112303,
      "memory(GiB)": 77.56,
      "step": 37325,
      "token_acc": 0.4180327868852459,
      "train_speed(iter/s)": 1.440794
    },
    {
      "epoch": 1.5993316481727433,
      "grad_norm": 4.466330051422119,
      "learning_rate": 7.68090656483164e-05,
      "loss": 2.578377532958984,
      "memory(GiB)": 77.56,
      "step": 37330,
      "token_acc": 0.4340836012861736,
      "train_speed(iter/s)": 1.440785
    },
    {
      "epoch": 1.5995458635019921,
      "grad_norm": 4.0943779945373535,
      "learning_rate": 7.680338478497613e-05,
      "loss": 2.4776931762695313,
      "memory(GiB)": 77.56,
      "step": 37335,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.440776
    },
    {
      "epoch": 1.5997600788312412,
      "grad_norm": 5.398914337158203,
      "learning_rate": 7.679770343606741e-05,
      "loss": 2.6147722244262694,
      "memory(GiB)": 77.56,
      "step": 37340,
      "token_acc": 0.4681647940074906,
      "train_speed(iter/s)": 1.440797
    },
    {
      "epoch": 1.5999742941604902,
      "grad_norm": 5.088338851928711,
      "learning_rate": 7.679202160169314e-05,
      "loss": 2.4712217330932615,
      "memory(GiB)": 77.56,
      "step": 37345,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.440776
    },
    {
      "epoch": 1.600188509489739,
      "grad_norm": 5.90064811706543,
      "learning_rate": 7.678633928195628e-05,
      "loss": 2.511794662475586,
      "memory(GiB)": 77.56,
      "step": 37350,
      "token_acc": 0.4955223880597015,
      "train_speed(iter/s)": 1.440775
    },
    {
      "epoch": 1.600402724818988,
      "grad_norm": 6.789758205413818,
      "learning_rate": 7.678065647695975e-05,
      "loss": 2.4820667266845704,
      "memory(GiB)": 77.56,
      "step": 37355,
      "token_acc": 0.4584837545126354,
      "train_speed(iter/s)": 1.440778
    },
    {
      "epoch": 1.600616940148237,
      "grad_norm": 7.3594160079956055,
      "learning_rate": 7.67749731868065e-05,
      "loss": 2.6745500564575195,
      "memory(GiB)": 77.56,
      "step": 37360,
      "token_acc": 0.4699248120300752,
      "train_speed(iter/s)": 1.440806
    },
    {
      "epoch": 1.600831155477486,
      "grad_norm": 5.078739166259766,
      "learning_rate": 7.676928941159951e-05,
      "loss": 2.4518218994140626,
      "memory(GiB)": 77.56,
      "step": 37365,
      "token_acc": 0.50920245398773,
      "train_speed(iter/s)": 1.440839
    },
    {
      "epoch": 1.601045370806735,
      "grad_norm": 4.971065521240234,
      "learning_rate": 7.676360515144172e-05,
      "loss": 2.2999269485473635,
      "memory(GiB)": 77.56,
      "step": 37370,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.440816
    },
    {
      "epoch": 1.601259586135984,
      "grad_norm": 5.04205322265625,
      "learning_rate": 7.675792040643611e-05,
      "loss": 2.50014705657959,
      "memory(GiB)": 77.56,
      "step": 37375,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.440805
    },
    {
      "epoch": 1.6014738014652328,
      "grad_norm": 4.661972999572754,
      "learning_rate": 7.675223517668569e-05,
      "loss": 2.547068977355957,
      "memory(GiB)": 77.56,
      "step": 37380,
      "token_acc": 0.44366197183098594,
      "train_speed(iter/s)": 1.440857
    },
    {
      "epoch": 1.6016880167944818,
      "grad_norm": 6.270064353942871,
      "learning_rate": 7.67465494622934e-05,
      "loss": 2.709405517578125,
      "memory(GiB)": 77.56,
      "step": 37385,
      "token_acc": 0.439873417721519,
      "train_speed(iter/s)": 1.440878
    },
    {
      "epoch": 1.6019022321237308,
      "grad_norm": 8.083171844482422,
      "learning_rate": 7.67408632633623e-05,
      "loss": 2.40850830078125,
      "memory(GiB)": 77.56,
      "step": 37390,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.440829
    },
    {
      "epoch": 1.6021164474529797,
      "grad_norm": 4.941572189331055,
      "learning_rate": 7.673517657999538e-05,
      "loss": 2.4612800598144533,
      "memory(GiB)": 77.56,
      "step": 37395,
      "token_acc": 0.4781021897810219,
      "train_speed(iter/s)": 1.440854
    },
    {
      "epoch": 1.6023306627822287,
      "grad_norm": 7.851924419403076,
      "learning_rate": 7.672948941229565e-05,
      "loss": 2.303112602233887,
      "memory(GiB)": 77.56,
      "step": 37400,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.440879
    },
    {
      "epoch": 1.6025448781114777,
      "grad_norm": 6.726553440093994,
      "learning_rate": 7.672380176036615e-05,
      "loss": 2.5341365814208983,
      "memory(GiB)": 77.56,
      "step": 37405,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.440932
    },
    {
      "epoch": 1.6027590934407265,
      "grad_norm": 4.743828773498535,
      "learning_rate": 7.671811362430992e-05,
      "loss": 2.3696929931640627,
      "memory(GiB)": 77.56,
      "step": 37410,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.440963
    },
    {
      "epoch": 1.6029733087699756,
      "grad_norm": 3.8743042945861816,
      "learning_rate": 7.671242500422998e-05,
      "loss": 2.38269157409668,
      "memory(GiB)": 77.56,
      "step": 37415,
      "token_acc": 0.4940828402366864,
      "train_speed(iter/s)": 1.441007
    },
    {
      "epoch": 1.6031875240992246,
      "grad_norm": 4.833866119384766,
      "learning_rate": 7.670673590022939e-05,
      "loss": 2.354164886474609,
      "memory(GiB)": 77.56,
      "step": 37420,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.441027
    },
    {
      "epoch": 1.6034017394284734,
      "grad_norm": 4.122615814208984,
      "learning_rate": 7.670104631241126e-05,
      "loss": 2.5714073181152344,
      "memory(GiB)": 77.56,
      "step": 37425,
      "token_acc": 0.48502994011976047,
      "train_speed(iter/s)": 1.441018
    },
    {
      "epoch": 1.6036159547577225,
      "grad_norm": 5.98354959487915,
      "learning_rate": 7.66953562408786e-05,
      "loss": 2.5156038284301756,
      "memory(GiB)": 77.56,
      "step": 37430,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.440981
    },
    {
      "epoch": 1.6038301700869715,
      "grad_norm": 5.067171573638916,
      "learning_rate": 7.668966568573455e-05,
      "loss": 2.483203887939453,
      "memory(GiB)": 77.56,
      "step": 37435,
      "token_acc": 0.50625,
      "train_speed(iter/s)": 1.440999
    },
    {
      "epoch": 1.6040443854162203,
      "grad_norm": 9.094625473022461,
      "learning_rate": 7.668397464708214e-05,
      "loss": 2.5751674652099608,
      "memory(GiB)": 77.56,
      "step": 37440,
      "token_acc": 0.4588607594936709,
      "train_speed(iter/s)": 1.440993
    },
    {
      "epoch": 1.6042586007454693,
      "grad_norm": 4.778010845184326,
      "learning_rate": 7.667828312502452e-05,
      "loss": 2.664686584472656,
      "memory(GiB)": 77.56,
      "step": 37445,
      "token_acc": 0.4402730375426621,
      "train_speed(iter/s)": 1.441024
    },
    {
      "epoch": 1.6044728160747184,
      "grad_norm": 3.4660446643829346,
      "learning_rate": 7.667259111966476e-05,
      "loss": 2.4331796646118162,
      "memory(GiB)": 77.56,
      "step": 37450,
      "token_acc": 0.4935897435897436,
      "train_speed(iter/s)": 1.441056
    },
    {
      "epoch": 1.6046870314039672,
      "grad_norm": 4.855658531188965,
      "learning_rate": 7.6666898631106e-05,
      "loss": 2.3552242279052735,
      "memory(GiB)": 77.56,
      "step": 37455,
      "token_acc": 0.5369649805447471,
      "train_speed(iter/s)": 1.441061
    },
    {
      "epoch": 1.6049012467332162,
      "grad_norm": 4.5516533851623535,
      "learning_rate": 7.666120565945135e-05,
      "loss": 2.7862833023071287,
      "memory(GiB)": 77.56,
      "step": 37460,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.44107
    },
    {
      "epoch": 1.6051154620624652,
      "grad_norm": 4.8243536949157715,
      "learning_rate": 7.665551220480395e-05,
      "loss": 2.5050823211669924,
      "memory(GiB)": 77.56,
      "step": 37465,
      "token_acc": 0.5036764705882353,
      "train_speed(iter/s)": 1.441077
    },
    {
      "epoch": 1.605329677391714,
      "grad_norm": 3.3661677837371826,
      "learning_rate": 7.664981826726695e-05,
      "loss": 2.2978546142578127,
      "memory(GiB)": 77.56,
      "step": 37470,
      "token_acc": 0.48360655737704916,
      "train_speed(iter/s)": 1.441071
    },
    {
      "epoch": 1.605543892720963,
      "grad_norm": 4.803250312805176,
      "learning_rate": 7.664412384694348e-05,
      "loss": 2.3432857513427736,
      "memory(GiB)": 77.56,
      "step": 37475,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.441083
    },
    {
      "epoch": 1.6057581080502121,
      "grad_norm": 4.503689765930176,
      "learning_rate": 7.663842894393672e-05,
      "loss": 2.4688488006591798,
      "memory(GiB)": 77.56,
      "step": 37480,
      "token_acc": 0.49185667752442996,
      "train_speed(iter/s)": 1.441109
    },
    {
      "epoch": 1.605972323379461,
      "grad_norm": 4.941961765289307,
      "learning_rate": 7.663273355834984e-05,
      "loss": 2.3893104553222657,
      "memory(GiB)": 77.56,
      "step": 37485,
      "token_acc": 0.4962962962962963,
      "train_speed(iter/s)": 1.441081
    },
    {
      "epoch": 1.60618653870871,
      "grad_norm": 6.212223052978516,
      "learning_rate": 7.662703769028599e-05,
      "loss": 2.3285255432128906,
      "memory(GiB)": 77.56,
      "step": 37490,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.441111
    },
    {
      "epoch": 1.606400754037959,
      "grad_norm": 5.007437705993652,
      "learning_rate": 7.662134133984838e-05,
      "loss": 2.505349349975586,
      "memory(GiB)": 77.56,
      "step": 37495,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.441141
    },
    {
      "epoch": 1.6066149693672078,
      "grad_norm": 4.935401439666748,
      "learning_rate": 7.66156445071402e-05,
      "loss": 2.5389307022094725,
      "memory(GiB)": 77.56,
      "step": 37500,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.441172
    },
    {
      "epoch": 1.6066149693672078,
      "eval_loss": 2.4067347049713135,
      "eval_runtime": 14.7178,
      "eval_samples_per_second": 6.794,
      "eval_steps_per_second": 6.794,
      "eval_token_acc": 0.47804878048780486,
      "step": 37500
    },
    {
      "epoch": 1.6068291846964569,
      "grad_norm": 4.806564807891846,
      "learning_rate": 7.660994719226464e-05,
      "loss": 2.657512664794922,
      "memory(GiB)": 77.56,
      "step": 37505,
      "token_acc": 0.47214854111405835,
      "train_speed(iter/s)": 1.440324
    },
    {
      "epoch": 1.607043400025706,
      "grad_norm": 5.706747055053711,
      "learning_rate": 7.660424939532494e-05,
      "loss": 2.6950347900390623,
      "memory(GiB)": 77.56,
      "step": 37510,
      "token_acc": 0.4163934426229508,
      "train_speed(iter/s)": 1.440341
    },
    {
      "epoch": 1.6072576153549547,
      "grad_norm": 6.18770170211792,
      "learning_rate": 7.65985511164243e-05,
      "loss": 2.7094188690185548,
      "memory(GiB)": 77.56,
      "step": 37515,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.440335
    },
    {
      "epoch": 1.6074718306842037,
      "grad_norm": 4.365080833435059,
      "learning_rate": 7.659285235566596e-05,
      "loss": 2.657796096801758,
      "memory(GiB)": 77.56,
      "step": 37520,
      "token_acc": 0.4649350649350649,
      "train_speed(iter/s)": 1.440347
    },
    {
      "epoch": 1.6076860460134528,
      "grad_norm": 5.248068809509277,
      "learning_rate": 7.658715311315314e-05,
      "loss": 2.4294240951538084,
      "memory(GiB)": 77.56,
      "step": 37525,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.440401
    },
    {
      "epoch": 1.6079002613427016,
      "grad_norm": 5.749895095825195,
      "learning_rate": 7.658145338898912e-05,
      "loss": 2.715859031677246,
      "memory(GiB)": 77.56,
      "step": 37530,
      "token_acc": 0.4336283185840708,
      "train_speed(iter/s)": 1.440404
    },
    {
      "epoch": 1.6081144766719506,
      "grad_norm": 4.384402275085449,
      "learning_rate": 7.657575318327712e-05,
      "loss": 2.6229156494140624,
      "memory(GiB)": 77.56,
      "step": 37535,
      "token_acc": 0.4440677966101695,
      "train_speed(iter/s)": 1.440413
    },
    {
      "epoch": 1.6083286920011997,
      "grad_norm": 4.644188404083252,
      "learning_rate": 7.657005249612044e-05,
      "loss": 2.7625591278076174,
      "memory(GiB)": 77.56,
      "step": 37540,
      "token_acc": 0.4498567335243553,
      "train_speed(iter/s)": 1.440456
    },
    {
      "epoch": 1.6085429073304485,
      "grad_norm": 4.940188884735107,
      "learning_rate": 7.65643513276223e-05,
      "loss": 2.4074514389038084,
      "memory(GiB)": 77.56,
      "step": 37545,
      "token_acc": 0.43050847457627117,
      "train_speed(iter/s)": 1.44049
    },
    {
      "epoch": 1.6087571226596975,
      "grad_norm": 5.767033100128174,
      "learning_rate": 7.655864967788605e-05,
      "loss": 2.867677116394043,
      "memory(GiB)": 77.56,
      "step": 37550,
      "token_acc": 0.42136498516320475,
      "train_speed(iter/s)": 1.440489
    },
    {
      "epoch": 1.6089713379889465,
      "grad_norm": 5.383480072021484,
      "learning_rate": 7.655294754701494e-05,
      "loss": 2.480497360229492,
      "memory(GiB)": 77.56,
      "step": 37555,
      "token_acc": 0.4923547400611621,
      "train_speed(iter/s)": 1.440502
    },
    {
      "epoch": 1.6091855533181953,
      "grad_norm": 5.815268516540527,
      "learning_rate": 7.654724493511227e-05,
      "loss": 2.8430389404296874,
      "memory(GiB)": 77.56,
      "step": 37560,
      "token_acc": 0.4175084175084175,
      "train_speed(iter/s)": 1.44051
    },
    {
      "epoch": 1.6093997686474444,
      "grad_norm": 5.3116865158081055,
      "learning_rate": 7.654154184228137e-05,
      "loss": 2.5249053955078127,
      "memory(GiB)": 77.56,
      "step": 37565,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.440471
    },
    {
      "epoch": 1.6096139839766934,
      "grad_norm": 5.164255619049072,
      "learning_rate": 7.653583826862552e-05,
      "loss": 2.545930099487305,
      "memory(GiB)": 77.56,
      "step": 37570,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.6098281993059422,
      "grad_norm": 6.061285972595215,
      "learning_rate": 7.653013421424806e-05,
      "loss": 2.430738830566406,
      "memory(GiB)": 77.56,
      "step": 37575,
      "token_acc": 0.5127118644067796,
      "train_speed(iter/s)": 1.44046
    },
    {
      "epoch": 1.6100424146351913,
      "grad_norm": 5.412038326263428,
      "learning_rate": 7.652442967925236e-05,
      "loss": 2.7430561065673826,
      "memory(GiB)": 77.56,
      "step": 37580,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.440422
    },
    {
      "epoch": 1.6102566299644403,
      "grad_norm": 6.383668899536133,
      "learning_rate": 7.651872466374172e-05,
      "loss": 2.31632137298584,
      "memory(GiB)": 77.56,
      "step": 37585,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.44043
    },
    {
      "epoch": 1.6104708452936891,
      "grad_norm": 8.428053855895996,
      "learning_rate": 7.65130191678195e-05,
      "loss": 2.372853469848633,
      "memory(GiB)": 77.56,
      "step": 37590,
      "token_acc": 0.5280898876404494,
      "train_speed(iter/s)": 1.440393
    },
    {
      "epoch": 1.6106850606229381,
      "grad_norm": 4.508504867553711,
      "learning_rate": 7.650731319158908e-05,
      "loss": 2.4179462432861327,
      "memory(GiB)": 77.56,
      "step": 37595,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.440321
    },
    {
      "epoch": 1.6108992759521872,
      "grad_norm": 5.605442047119141,
      "learning_rate": 7.650160673515381e-05,
      "loss": 2.5736928939819337,
      "memory(GiB)": 77.56,
      "step": 37600,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.440341
    },
    {
      "epoch": 1.611113491281436,
      "grad_norm": 5.4860076904296875,
      "learning_rate": 7.649589979861706e-05,
      "loss": 2.276582145690918,
      "memory(GiB)": 77.56,
      "step": 37605,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.440328
    },
    {
      "epoch": 1.611327706610685,
      "grad_norm": 4.896621227264404,
      "learning_rate": 7.649019238208226e-05,
      "loss": 2.5030078887939453,
      "memory(GiB)": 77.56,
      "step": 37610,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.440361
    },
    {
      "epoch": 1.611541921939934,
      "grad_norm": 4.020619869232178,
      "learning_rate": 7.648448448565275e-05,
      "loss": 2.4690114974975588,
      "memory(GiB)": 77.56,
      "step": 37615,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.440395
    },
    {
      "epoch": 1.6117561372691829,
      "grad_norm": 4.35134744644165,
      "learning_rate": 7.647877610943197e-05,
      "loss": 2.580672264099121,
      "memory(GiB)": 77.56,
      "step": 37620,
      "token_acc": 0.4880952380952381,
      "train_speed(iter/s)": 1.440442
    },
    {
      "epoch": 1.611970352598432,
      "grad_norm": 3.719632387161255,
      "learning_rate": 7.647306725352332e-05,
      "loss": 2.273370361328125,
      "memory(GiB)": 77.56,
      "step": 37625,
      "token_acc": 0.5304878048780488,
      "train_speed(iter/s)": 1.440464
    },
    {
      "epoch": 1.612184567927681,
      "grad_norm": 5.32607889175415,
      "learning_rate": 7.646735791803024e-05,
      "loss": 2.459147644042969,
      "memory(GiB)": 77.56,
      "step": 37630,
      "token_acc": 0.47151898734177217,
      "train_speed(iter/s)": 1.440492
    },
    {
      "epoch": 1.6123987832569298,
      "grad_norm": 5.236227512359619,
      "learning_rate": 7.646164810305611e-05,
      "loss": 2.3994359970092773,
      "memory(GiB)": 77.56,
      "step": 37635,
      "token_acc": 0.5039370078740157,
      "train_speed(iter/s)": 1.44049
    },
    {
      "epoch": 1.6126129985861788,
      "grad_norm": 4.731742858886719,
      "learning_rate": 7.645593780870442e-05,
      "loss": 2.6422397613525392,
      "memory(GiB)": 77.56,
      "step": 37640,
      "token_acc": 0.4790996784565916,
      "train_speed(iter/s)": 1.440471
    },
    {
      "epoch": 1.6128272139154278,
      "grad_norm": 4.112747669219971,
      "learning_rate": 7.645022703507858e-05,
      "loss": 2.458250045776367,
      "memory(GiB)": 77.56,
      "step": 37645,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.440508
    },
    {
      "epoch": 1.6130414292446766,
      "grad_norm": 4.678101539611816,
      "learning_rate": 7.64445157822821e-05,
      "loss": 2.2416994094848635,
      "memory(GiB)": 77.56,
      "step": 37650,
      "token_acc": 0.5251798561151079,
      "train_speed(iter/s)": 1.44046
    },
    {
      "epoch": 1.6132556445739257,
      "grad_norm": 4.763542175292969,
      "learning_rate": 7.643880405041838e-05,
      "loss": 2.4603944778442384,
      "memory(GiB)": 77.56,
      "step": 37655,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.440421
    },
    {
      "epoch": 1.6134698599031747,
      "grad_norm": 4.984938621520996,
      "learning_rate": 7.643309183959094e-05,
      "loss": 2.6370798110961915,
      "memory(GiB)": 77.56,
      "step": 37660,
      "token_acc": 0.46407185628742514,
      "train_speed(iter/s)": 1.440453
    },
    {
      "epoch": 1.6136840752324235,
      "grad_norm": 3.9210801124572754,
      "learning_rate": 7.642737914990324e-05,
      "loss": 2.6241277694702148,
      "memory(GiB)": 77.56,
      "step": 37665,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.440412
    },
    {
      "epoch": 1.6138982905616726,
      "grad_norm": 5.629879951477051,
      "learning_rate": 7.642166598145876e-05,
      "loss": 2.545711135864258,
      "memory(GiB)": 77.56,
      "step": 37670,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.440435
    },
    {
      "epoch": 1.6141125058909216,
      "grad_norm": 5.128633499145508,
      "learning_rate": 7.641595233436104e-05,
      "loss": 2.3216808319091795,
      "memory(GiB)": 77.56,
      "step": 37675,
      "token_acc": 0.4701492537313433,
      "train_speed(iter/s)": 1.440439
    },
    {
      "epoch": 1.6143267212201704,
      "grad_norm": 5.640380382537842,
      "learning_rate": 7.641023820871353e-05,
      "loss": 2.8596923828125,
      "memory(GiB)": 77.56,
      "step": 37680,
      "token_acc": 0.44482758620689655,
      "train_speed(iter/s)": 1.440496
    },
    {
      "epoch": 1.6145409365494194,
      "grad_norm": 5.271227836608887,
      "learning_rate": 7.640452360461982e-05,
      "loss": 2.4152158737182616,
      "memory(GiB)": 77.56,
      "step": 37685,
      "token_acc": 0.5157894736842106,
      "train_speed(iter/s)": 1.440495
    },
    {
      "epoch": 1.6147551518786685,
      "grad_norm": 5.2296295166015625,
      "learning_rate": 7.639880852218338e-05,
      "loss": 2.613365364074707,
      "memory(GiB)": 77.56,
      "step": 37690,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.440515
    },
    {
      "epoch": 1.6149693672079173,
      "grad_norm": 4.282989978790283,
      "learning_rate": 7.639309296150775e-05,
      "loss": 2.387594985961914,
      "memory(GiB)": 77.56,
      "step": 37695,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.440548
    },
    {
      "epoch": 1.6151835825371663,
      "grad_norm": 4.745231628417969,
      "learning_rate": 7.638737692269649e-05,
      "loss": 2.5518789291381836,
      "memory(GiB)": 77.56,
      "step": 37700,
      "token_acc": 0.44814814814814813,
      "train_speed(iter/s)": 1.440564
    },
    {
      "epoch": 1.6153977978664154,
      "grad_norm": 5.613280773162842,
      "learning_rate": 7.638166040585314e-05,
      "loss": 2.257324981689453,
      "memory(GiB)": 77.56,
      "step": 37705,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.440525
    },
    {
      "epoch": 1.6156120131956642,
      "grad_norm": 5.070335388183594,
      "learning_rate": 7.637594341108127e-05,
      "loss": 2.4958402633666994,
      "memory(GiB)": 77.56,
      "step": 37710,
      "token_acc": 0.47706422018348627,
      "train_speed(iter/s)": 1.440492
    },
    {
      "epoch": 1.6158262285249132,
      "grad_norm": 4.498345851898193,
      "learning_rate": 7.637022593848444e-05,
      "loss": 2.5327690124511717,
      "memory(GiB)": 77.56,
      "step": 37715,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.440455
    },
    {
      "epoch": 1.6160404438541622,
      "grad_norm": 4.5153422355651855,
      "learning_rate": 7.636450798816624e-05,
      "loss": 2.7252254486083984,
      "memory(GiB)": 77.56,
      "step": 37720,
      "token_acc": 0.4490861618798956,
      "train_speed(iter/s)": 1.44047
    },
    {
      "epoch": 1.616254659183411,
      "grad_norm": 4.830429553985596,
      "learning_rate": 7.635878956023023e-05,
      "loss": 2.3023822784423826,
      "memory(GiB)": 77.56,
      "step": 37725,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.440426
    },
    {
      "epoch": 1.61646887451266,
      "grad_norm": 5.323904037475586,
      "learning_rate": 7.635307065478003e-05,
      "loss": 2.6148488998413084,
      "memory(GiB)": 77.56,
      "step": 37730,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.4404
    },
    {
      "epoch": 1.6166830898419091,
      "grad_norm": 5.403785705566406,
      "learning_rate": 7.634735127191922e-05,
      "loss": 2.47650032043457,
      "memory(GiB)": 77.56,
      "step": 37735,
      "token_acc": 0.45255474452554745,
      "train_speed(iter/s)": 1.440416
    },
    {
      "epoch": 1.616897305171158,
      "grad_norm": 4.429103851318359,
      "learning_rate": 7.634163141175144e-05,
      "loss": 2.313262176513672,
      "memory(GiB)": 77.56,
      "step": 37740,
      "token_acc": 0.4765625,
      "train_speed(iter/s)": 1.440421
    },
    {
      "epoch": 1.617111520500407,
      "grad_norm": 5.550984859466553,
      "learning_rate": 7.633591107438029e-05,
      "loss": 2.4353511810302733,
      "memory(GiB)": 77.56,
      "step": 37745,
      "token_acc": 0.4653846153846154,
      "train_speed(iter/s)": 1.440435
    },
    {
      "epoch": 1.617325735829656,
      "grad_norm": 4.2899932861328125,
      "learning_rate": 7.63301902599094e-05,
      "loss": 2.345765495300293,
      "memory(GiB)": 77.56,
      "step": 37750,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.440467
    },
    {
      "epoch": 1.6175399511589048,
      "grad_norm": 5.171411991119385,
      "learning_rate": 7.632446896844243e-05,
      "loss": 2.5676803588867188,
      "memory(GiB)": 77.56,
      "step": 37755,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.440463
    },
    {
      "epoch": 1.6177541664881538,
      "grad_norm": 7.154988765716553,
      "learning_rate": 7.631874720008301e-05,
      "loss": 2.253861999511719,
      "memory(GiB)": 77.56,
      "step": 37760,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.440488
    },
    {
      "epoch": 1.6179683818174029,
      "grad_norm": 4.189392566680908,
      "learning_rate": 7.631302495493478e-05,
      "loss": 2.5879135131835938,
      "memory(GiB)": 77.56,
      "step": 37765,
      "token_acc": 0.4537313432835821,
      "train_speed(iter/s)": 1.440483
    },
    {
      "epoch": 1.6181825971466517,
      "grad_norm": 4.117321491241455,
      "learning_rate": 7.630730223310143e-05,
      "loss": 2.268657684326172,
      "memory(GiB)": 77.56,
      "step": 37770,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.440484
    },
    {
      "epoch": 1.6183968124759007,
      "grad_norm": 5.277060508728027,
      "learning_rate": 7.630157903468663e-05,
      "loss": 2.2288436889648438,
      "memory(GiB)": 77.56,
      "step": 37775,
      "token_acc": 0.5304347826086957,
      "train_speed(iter/s)": 1.440486
    },
    {
      "epoch": 1.6186110278051498,
      "grad_norm": 4.592144966125488,
      "learning_rate": 7.629585535979402e-05,
      "loss": 2.6815914154052733,
      "memory(GiB)": 77.56,
      "step": 37780,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.440443
    },
    {
      "epoch": 1.6188252431343986,
      "grad_norm": 6.084299087524414,
      "learning_rate": 7.629013120852736e-05,
      "loss": 2.34604549407959,
      "memory(GiB)": 77.56,
      "step": 37785,
      "token_acc": 0.5315985130111525,
      "train_speed(iter/s)": 1.44041
    },
    {
      "epoch": 1.6190394584636476,
      "grad_norm": 5.6197309494018555,
      "learning_rate": 7.628440658099029e-05,
      "loss": 2.512620735168457,
      "memory(GiB)": 77.56,
      "step": 37790,
      "token_acc": 0.4548872180451128,
      "train_speed(iter/s)": 1.440386
    },
    {
      "epoch": 1.6192536737928966,
      "grad_norm": 5.779651641845703,
      "learning_rate": 7.627868147728654e-05,
      "loss": 2.5897064208984375,
      "memory(GiB)": 77.56,
      "step": 37795,
      "token_acc": 0.4460431654676259,
      "train_speed(iter/s)": 1.440375
    },
    {
      "epoch": 1.6194678891221455,
      "grad_norm": 4.179642200469971,
      "learning_rate": 7.627295589751982e-05,
      "loss": 2.5370941162109375,
      "memory(GiB)": 77.56,
      "step": 37800,
      "token_acc": 0.4855072463768116,
      "train_speed(iter/s)": 1.440402
    },
    {
      "epoch": 1.6196821044513945,
      "grad_norm": 7.664151668548584,
      "learning_rate": 7.626722984179387e-05,
      "loss": 2.5881237030029296,
      "memory(GiB)": 77.56,
      "step": 37805,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.440388
    },
    {
      "epoch": 1.6198963197806435,
      "grad_norm": 4.311489105224609,
      "learning_rate": 7.62615033102124e-05,
      "loss": 2.3895198822021486,
      "memory(GiB)": 77.56,
      "step": 37810,
      "token_acc": 0.509493670886076,
      "train_speed(iter/s)": 1.44041
    },
    {
      "epoch": 1.6201105351098923,
      "grad_norm": 4.460899353027344,
      "learning_rate": 7.625577630287918e-05,
      "loss": 2.559486961364746,
      "memory(GiB)": 77.56,
      "step": 37815,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.440409
    },
    {
      "epoch": 1.6203247504391414,
      "grad_norm": 4.621767997741699,
      "learning_rate": 7.625004881989792e-05,
      "loss": 2.77893180847168,
      "memory(GiB)": 77.56,
      "step": 37820,
      "token_acc": 0.4360655737704918,
      "train_speed(iter/s)": 1.440408
    },
    {
      "epoch": 1.6205389657683904,
      "grad_norm": 5.098792552947998,
      "learning_rate": 7.624432086137241e-05,
      "loss": 2.751088333129883,
      "memory(GiB)": 77.56,
      "step": 37825,
      "token_acc": 0.4426751592356688,
      "train_speed(iter/s)": 1.440423
    },
    {
      "epoch": 1.6207531810976392,
      "grad_norm": 4.143022060394287,
      "learning_rate": 7.623859242740642e-05,
      "loss": 2.7381216049194337,
      "memory(GiB)": 77.56,
      "step": 37830,
      "token_acc": 0.4721189591078067,
      "train_speed(iter/s)": 1.440403
    },
    {
      "epoch": 1.6209673964268885,
      "grad_norm": 6.091585636138916,
      "learning_rate": 7.62328635181037e-05,
      "loss": 2.7297189712524412,
      "memory(GiB)": 77.56,
      "step": 37835,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.440406
    },
    {
      "epoch": 1.6211816117561373,
      "grad_norm": 5.44978141784668,
      "learning_rate": 7.622713413356806e-05,
      "loss": 2.7423160552978514,
      "memory(GiB)": 77.56,
      "step": 37840,
      "token_acc": 0.38699690402476783,
      "train_speed(iter/s)": 1.440403
    },
    {
      "epoch": 1.621395827085386,
      "grad_norm": 5.1050944328308105,
      "learning_rate": 7.622140427390327e-05,
      "loss": 2.499595069885254,
      "memory(GiB)": 77.56,
      "step": 37845,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.440431
    },
    {
      "epoch": 1.6216100424146354,
      "grad_norm": 4.863409042358398,
      "learning_rate": 7.621567393921315e-05,
      "loss": 2.842523765563965,
      "memory(GiB)": 77.56,
      "step": 37850,
      "token_acc": 0.4620938628158845,
      "train_speed(iter/s)": 1.440465
    },
    {
      "epoch": 1.6218242577438842,
      "grad_norm": 6.070392608642578,
      "learning_rate": 7.62099431296015e-05,
      "loss": 2.9843233108520506,
      "memory(GiB)": 77.56,
      "step": 37855,
      "token_acc": 0.38869257950530034,
      "train_speed(iter/s)": 1.440519
    },
    {
      "epoch": 1.622038473073133,
      "grad_norm": 4.390357494354248,
      "learning_rate": 7.620421184517216e-05,
      "loss": 2.491388702392578,
      "memory(GiB)": 77.56,
      "step": 37860,
      "token_acc": 0.46518105849582175,
      "train_speed(iter/s)": 1.440545
    },
    {
      "epoch": 1.6222526884023822,
      "grad_norm": 4.784060955047607,
      "learning_rate": 7.61984800860289e-05,
      "loss": 2.348118782043457,
      "memory(GiB)": 77.56,
      "step": 37865,
      "token_acc": 0.4635036496350365,
      "train_speed(iter/s)": 1.440521
    },
    {
      "epoch": 1.622466903731631,
      "grad_norm": 5.561811447143555,
      "learning_rate": 7.619274785227564e-05,
      "loss": 2.6775644302368162,
      "memory(GiB)": 77.56,
      "step": 37870,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.440527
    },
    {
      "epoch": 1.6226811190608799,
      "grad_norm": 5.449887752532959,
      "learning_rate": 7.618701514401618e-05,
      "loss": 2.7351711273193358,
      "memory(GiB)": 77.56,
      "step": 37875,
      "token_acc": 0.4416058394160584,
      "train_speed(iter/s)": 1.440554
    },
    {
      "epoch": 1.6228953343901291,
      "grad_norm": 5.843167304992676,
      "learning_rate": 7.618128196135435e-05,
      "loss": 2.4031232833862304,
      "memory(GiB)": 77.56,
      "step": 37880,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.440484
    },
    {
      "epoch": 1.623109549719378,
      "grad_norm": 5.339305400848389,
      "learning_rate": 7.617554830439406e-05,
      "loss": 2.3922922134399416,
      "memory(GiB)": 77.56,
      "step": 37885,
      "token_acc": 0.5214521452145214,
      "train_speed(iter/s)": 1.440501
    },
    {
      "epoch": 1.6233237650486267,
      "grad_norm": 7.193346977233887,
      "learning_rate": 7.616981417323914e-05,
      "loss": 2.7266597747802734,
      "memory(GiB)": 77.56,
      "step": 37890,
      "token_acc": 0.48046875,
      "train_speed(iter/s)": 1.440524
    },
    {
      "epoch": 1.623537980377876,
      "grad_norm": 6.7550435066223145,
      "learning_rate": 7.61640795679935e-05,
      "loss": 2.408618354797363,
      "memory(GiB)": 77.56,
      "step": 37895,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.440537
    },
    {
      "epoch": 1.6237521957071248,
      "grad_norm": 5.425740718841553,
      "learning_rate": 7.6158344488761e-05,
      "loss": 2.6974472045898437,
      "memory(GiB)": 77.56,
      "step": 37900,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.440556
    },
    {
      "epoch": 1.6239664110363736,
      "grad_norm": 4.923129081726074,
      "learning_rate": 7.615260893564556e-05,
      "loss": 2.2135669708251955,
      "memory(GiB)": 77.56,
      "step": 37905,
      "token_acc": 0.5275080906148867,
      "train_speed(iter/s)": 1.44055
    },
    {
      "epoch": 1.6241806263656229,
      "grad_norm": 5.52205228805542,
      "learning_rate": 7.614687290875107e-05,
      "loss": 2.7369285583496095,
      "memory(GiB)": 77.56,
      "step": 37910,
      "token_acc": 0.4406779661016949,
      "train_speed(iter/s)": 1.440511
    },
    {
      "epoch": 1.6243948416948717,
      "grad_norm": 4.848884105682373,
      "learning_rate": 7.614113640818145e-05,
      "loss": 2.658232498168945,
      "memory(GiB)": 77.56,
      "step": 37915,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.4405
    },
    {
      "epoch": 1.6246090570241205,
      "grad_norm": 4.7700018882751465,
      "learning_rate": 7.613539943404064e-05,
      "loss": 2.562948226928711,
      "memory(GiB)": 77.56,
      "step": 37920,
      "token_acc": 0.44565217391304346,
      "train_speed(iter/s)": 1.440544
    },
    {
      "epoch": 1.6248232723533698,
      "grad_norm": 5.361043930053711,
      "learning_rate": 7.612966198643254e-05,
      "loss": 2.4788240432739257,
      "memory(GiB)": 77.56,
      "step": 37925,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.440516
    },
    {
      "epoch": 1.6250374876826186,
      "grad_norm": 4.615484714508057,
      "learning_rate": 7.612392406546109e-05,
      "loss": 2.5205101013183593,
      "memory(GiB)": 77.56,
      "step": 37930,
      "token_acc": 0.4967532467532468,
      "train_speed(iter/s)": 1.440487
    },
    {
      "epoch": 1.6252517030118674,
      "grad_norm": 5.751092433929443,
      "learning_rate": 7.611818567123025e-05,
      "loss": 2.799050521850586,
      "memory(GiB)": 77.56,
      "step": 37935,
      "token_acc": 0.4452054794520548,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.6254659183411166,
      "grad_norm": 6.117849826812744,
      "learning_rate": 7.6112446803844e-05,
      "loss": 2.5702619552612305,
      "memory(GiB)": 77.56,
      "step": 37940,
      "token_acc": 0.47634069400630913,
      "train_speed(iter/s)": 1.440508
    },
    {
      "epoch": 1.6256801336703655,
      "grad_norm": 5.259683609008789,
      "learning_rate": 7.610670746340626e-05,
      "loss": 2.4230731964111327,
      "memory(GiB)": 77.56,
      "step": 37945,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.440516
    },
    {
      "epoch": 1.6258943489996143,
      "grad_norm": 4.160335063934326,
      "learning_rate": 7.610096765002104e-05,
      "loss": 2.6499282836914064,
      "memory(GiB)": 77.56,
      "step": 37950,
      "token_acc": 0.4507936507936508,
      "train_speed(iter/s)": 1.440505
    },
    {
      "epoch": 1.6261085643288635,
      "grad_norm": 3.890566110610962,
      "learning_rate": 7.609522736379229e-05,
      "loss": 2.342402458190918,
      "memory(GiB)": 77.56,
      "step": 37955,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.440525
    },
    {
      "epoch": 1.6263227796581123,
      "grad_norm": 9.271947860717773,
      "learning_rate": 7.608948660482403e-05,
      "loss": 2.3067720413208006,
      "memory(GiB)": 77.56,
      "step": 37960,
      "token_acc": 0.4796747967479675,
      "train_speed(iter/s)": 1.440469
    },
    {
      "epoch": 1.6265369949873612,
      "grad_norm": 4.786581039428711,
      "learning_rate": 7.608374537322024e-05,
      "loss": 2.410334587097168,
      "memory(GiB)": 77.56,
      "step": 37965,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.440478
    },
    {
      "epoch": 1.6267512103166104,
      "grad_norm": 4.84458589553833,
      "learning_rate": 7.607800366908493e-05,
      "loss": 2.341404914855957,
      "memory(GiB)": 77.56,
      "step": 37970,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.440465
    },
    {
      "epoch": 1.6269654256458592,
      "grad_norm": 5.3732590675354,
      "learning_rate": 7.607226149252212e-05,
      "loss": 2.887238311767578,
      "memory(GiB)": 77.56,
      "step": 37975,
      "token_acc": 0.43109540636042404,
      "train_speed(iter/s)": 1.440482
    },
    {
      "epoch": 1.627179640975108,
      "grad_norm": 5.549344062805176,
      "learning_rate": 7.606651884363585e-05,
      "loss": 2.504510498046875,
      "memory(GiB)": 77.56,
      "step": 37980,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.440495
    },
    {
      "epoch": 1.6273938563043573,
      "grad_norm": 6.415650367736816,
      "learning_rate": 7.606077572253012e-05,
      "loss": 2.535602569580078,
      "memory(GiB)": 77.56,
      "step": 37985,
      "token_acc": 0.4432624113475177,
      "train_speed(iter/s)": 1.440516
    },
    {
      "epoch": 1.627608071633606,
      "grad_norm": 4.586345195770264,
      "learning_rate": 7.6055032129309e-05,
      "loss": 2.4179458618164062,
      "memory(GiB)": 77.56,
      "step": 37990,
      "token_acc": 0.49032258064516127,
      "train_speed(iter/s)": 1.440533
    },
    {
      "epoch": 1.627822286962855,
      "grad_norm": 5.557847499847412,
      "learning_rate": 7.604928806407654e-05,
      "loss": 2.7187625885009767,
      "memory(GiB)": 77.56,
      "step": 37995,
      "token_acc": 0.444015444015444,
      "train_speed(iter/s)": 1.440572
    },
    {
      "epoch": 1.6280365022921042,
      "grad_norm": 5.628841876983643,
      "learning_rate": 7.604354352693677e-05,
      "loss": 2.528227996826172,
      "memory(GiB)": 77.56,
      "step": 38000,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.440593
    },
    {
      "epoch": 1.6280365022921042,
      "eval_loss": 2.230398654937744,
      "eval_runtime": 14.3111,
      "eval_samples_per_second": 6.988,
      "eval_steps_per_second": 6.988,
      "eval_token_acc": 0.4927536231884058,
      "step": 38000
    },
    {
      "epoch": 1.628250717621353,
      "grad_norm": 3.8304436206817627,
      "learning_rate": 7.60377985179938e-05,
      "loss": 2.668830680847168,
      "memory(GiB)": 77.56,
      "step": 38005,
      "token_acc": 0.4680105170902717,
      "train_speed(iter/s)": 1.439731
    },
    {
      "epoch": 1.6284649329506018,
      "grad_norm": 3.7807648181915283,
      "learning_rate": 7.603205303735166e-05,
      "loss": 2.268621826171875,
      "memory(GiB)": 77.56,
      "step": 38010,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.439728
    },
    {
      "epoch": 1.628679148279851,
      "grad_norm": 4.056646823883057,
      "learning_rate": 7.602630708511448e-05,
      "loss": 2.4808622360229493,
      "memory(GiB)": 77.56,
      "step": 38015,
      "token_acc": 0.4835820895522388,
      "train_speed(iter/s)": 1.439775
    },
    {
      "epoch": 1.6288933636090999,
      "grad_norm": 4.447537899017334,
      "learning_rate": 7.602056066138633e-05,
      "loss": 2.279632568359375,
      "memory(GiB)": 77.56,
      "step": 38020,
      "token_acc": 0.4907749077490775,
      "train_speed(iter/s)": 1.439799
    },
    {
      "epoch": 1.6291075789383487,
      "grad_norm": 5.137543678283691,
      "learning_rate": 7.601481376627131e-05,
      "loss": 2.5810993194580076,
      "memory(GiB)": 77.56,
      "step": 38025,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.439769
    },
    {
      "epoch": 1.629321794267598,
      "grad_norm": 4.957739353179932,
      "learning_rate": 7.600906639987352e-05,
      "loss": 2.527983856201172,
      "memory(GiB)": 77.56,
      "step": 38030,
      "token_acc": 0.5073529411764706,
      "train_speed(iter/s)": 1.439712
    },
    {
      "epoch": 1.6295360095968467,
      "grad_norm": 4.1089935302734375,
      "learning_rate": 7.600331856229712e-05,
      "loss": 2.464487838745117,
      "memory(GiB)": 77.56,
      "step": 38035,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.439744
    },
    {
      "epoch": 1.6297502249260956,
      "grad_norm": 6.254332065582275,
      "learning_rate": 7.59975702536462e-05,
      "loss": 2.3078657150268556,
      "memory(GiB)": 77.56,
      "step": 38040,
      "token_acc": 0.5447470817120622,
      "train_speed(iter/s)": 1.439765
    },
    {
      "epoch": 1.6299644402553448,
      "grad_norm": 4.900516033172607,
      "learning_rate": 7.599182147402491e-05,
      "loss": 2.5902278900146483,
      "memory(GiB)": 77.56,
      "step": 38045,
      "token_acc": 0.47634069400630913,
      "train_speed(iter/s)": 1.439725
    },
    {
      "epoch": 1.6301786555845936,
      "grad_norm": 5.00094747543335,
      "learning_rate": 7.598607222353739e-05,
      "loss": 2.278514862060547,
      "memory(GiB)": 77.56,
      "step": 38050,
      "token_acc": 0.5131086142322098,
      "train_speed(iter/s)": 1.439729
    },
    {
      "epoch": 1.6303928709138424,
      "grad_norm": 7.132684230804443,
      "learning_rate": 7.598032250228779e-05,
      "loss": 2.2978532791137694,
      "memory(GiB)": 77.56,
      "step": 38055,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.439763
    },
    {
      "epoch": 1.6306070862430917,
      "grad_norm": 5.430882453918457,
      "learning_rate": 7.597457231038028e-05,
      "loss": 2.6715049743652344,
      "memory(GiB)": 77.56,
      "step": 38060,
      "token_acc": 0.44569288389513106,
      "train_speed(iter/s)": 1.43979
    },
    {
      "epoch": 1.6308213015723405,
      "grad_norm": 6.83610200881958,
      "learning_rate": 7.596882164791903e-05,
      "loss": 2.5615875244140627,
      "memory(GiB)": 77.56,
      "step": 38065,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.439755
    },
    {
      "epoch": 1.6310355169015893,
      "grad_norm": 5.552457332611084,
      "learning_rate": 7.596307051500821e-05,
      "loss": 2.6205591201782226,
      "memory(GiB)": 77.56,
      "step": 38070,
      "token_acc": 0.4410112359550562,
      "train_speed(iter/s)": 1.439781
    },
    {
      "epoch": 1.6312497322308386,
      "grad_norm": 5.539562225341797,
      "learning_rate": 7.595731891175202e-05,
      "loss": 2.1188907623291016,
      "memory(GiB)": 77.56,
      "step": 38075,
      "token_acc": 0.5524193548387096,
      "train_speed(iter/s)": 1.439802
    },
    {
      "epoch": 1.6314639475600874,
      "grad_norm": 4.295923709869385,
      "learning_rate": 7.595156683825463e-05,
      "loss": 2.400032806396484,
      "memory(GiB)": 77.56,
      "step": 38080,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.439766
    },
    {
      "epoch": 1.6316781628893362,
      "grad_norm": 5.152463912963867,
      "learning_rate": 7.594581429462026e-05,
      "loss": 2.635820770263672,
      "memory(GiB)": 77.56,
      "step": 38085,
      "token_acc": 0.48286604361370716,
      "train_speed(iter/s)": 1.439808
    },
    {
      "epoch": 1.6318923782185855,
      "grad_norm": 6.016039848327637,
      "learning_rate": 7.594006128095314e-05,
      "loss": 2.6588924407958983,
      "memory(GiB)": 77.56,
      "step": 38090,
      "token_acc": 0.427536231884058,
      "train_speed(iter/s)": 1.439872
    },
    {
      "epoch": 1.6321065935478343,
      "grad_norm": 5.937152862548828,
      "learning_rate": 7.593430779735749e-05,
      "loss": 2.530279541015625,
      "memory(GiB)": 77.56,
      "step": 38095,
      "token_acc": 0.4921135646687697,
      "train_speed(iter/s)": 1.439901
    },
    {
      "epoch": 1.632320808877083,
      "grad_norm": 4.711450576782227,
      "learning_rate": 7.592855384393752e-05,
      "loss": 2.4118274688720702,
      "memory(GiB)": 77.56,
      "step": 38100,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.439916
    },
    {
      "epoch": 1.6325350242063323,
      "grad_norm": 7.2043609619140625,
      "learning_rate": 7.592279942079746e-05,
      "loss": 2.615461730957031,
      "memory(GiB)": 77.56,
      "step": 38105,
      "token_acc": 0.4858156028368794,
      "train_speed(iter/s)": 1.43993
    },
    {
      "epoch": 1.6327492395355812,
      "grad_norm": 4.363980770111084,
      "learning_rate": 7.591704452804157e-05,
      "loss": 2.2706024169921877,
      "memory(GiB)": 77.56,
      "step": 38110,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.439904
    },
    {
      "epoch": 1.63296345486483,
      "grad_norm": 3.774569511413574,
      "learning_rate": 7.591128916577412e-05,
      "loss": 2.411854553222656,
      "memory(GiB)": 77.56,
      "step": 38115,
      "token_acc": 0.5197568389057751,
      "train_speed(iter/s)": 1.439926
    },
    {
      "epoch": 1.6331776701940792,
      "grad_norm": 5.024695873260498,
      "learning_rate": 7.590553333409934e-05,
      "loss": 2.5799734115600588,
      "memory(GiB)": 77.56,
      "step": 38120,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.439918
    },
    {
      "epoch": 1.633391885523328,
      "grad_norm": 4.98157262802124,
      "learning_rate": 7.589977703312152e-05,
      "loss": 2.3855262756347657,
      "memory(GiB)": 77.56,
      "step": 38125,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.43993
    },
    {
      "epoch": 1.6336061008525768,
      "grad_norm": 6.1562724113464355,
      "learning_rate": 7.589402026294497e-05,
      "loss": 2.4294706344604493,
      "memory(GiB)": 77.56,
      "step": 38130,
      "token_acc": 0.4576271186440678,
      "train_speed(iter/s)": 1.43996
    },
    {
      "epoch": 1.633820316181826,
      "grad_norm": 4.540532112121582,
      "learning_rate": 7.588826302367394e-05,
      "loss": 2.669905090332031,
      "memory(GiB)": 77.56,
      "step": 38135,
      "token_acc": 0.432258064516129,
      "train_speed(iter/s)": 1.439953
    },
    {
      "epoch": 1.634034531511075,
      "grad_norm": 5.373117923736572,
      "learning_rate": 7.588250531541274e-05,
      "loss": 2.767827796936035,
      "memory(GiB)": 77.56,
      "step": 38140,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.439919
    },
    {
      "epoch": 1.6342487468403237,
      "grad_norm": 4.0611677169799805,
      "learning_rate": 7.587674713826567e-05,
      "loss": 2.810939979553223,
      "memory(GiB)": 77.56,
      "step": 38145,
      "token_acc": 0.4331395348837209,
      "train_speed(iter/s)": 1.439934
    },
    {
      "epoch": 1.634462962169573,
      "grad_norm": 6.725802421569824,
      "learning_rate": 7.587098849233707e-05,
      "loss": 2.441546821594238,
      "memory(GiB)": 77.56,
      "step": 38150,
      "token_acc": 0.4830188679245283,
      "train_speed(iter/s)": 1.439945
    },
    {
      "epoch": 1.6346771774988218,
      "grad_norm": 4.110783576965332,
      "learning_rate": 7.586522937773123e-05,
      "loss": 2.278568077087402,
      "memory(GiB)": 77.56,
      "step": 38155,
      "token_acc": 0.5190311418685121,
      "train_speed(iter/s)": 1.439903
    },
    {
      "epoch": 1.6348913928280706,
      "grad_norm": 5.280030250549316,
      "learning_rate": 7.58594697945525e-05,
      "loss": 2.4644393920898438,
      "memory(GiB)": 77.56,
      "step": 38160,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.439901
    },
    {
      "epoch": 1.6351056081573199,
      "grad_norm": 4.459488868713379,
      "learning_rate": 7.585370974290521e-05,
      "loss": 2.884593963623047,
      "memory(GiB)": 77.56,
      "step": 38165,
      "token_acc": 0.44694533762057875,
      "train_speed(iter/s)": 1.439897
    },
    {
      "epoch": 1.6353198234865687,
      "grad_norm": 5.270105838775635,
      "learning_rate": 7.584794922289371e-05,
      "loss": 2.3009801864624024,
      "memory(GiB)": 77.56,
      "step": 38170,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.439888
    },
    {
      "epoch": 1.6355340388158177,
      "grad_norm": 6.147409915924072,
      "learning_rate": 7.584218823462238e-05,
      "loss": 2.55008544921875,
      "memory(GiB)": 77.56,
      "step": 38175,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.439818
    },
    {
      "epoch": 1.6357482541450667,
      "grad_norm": 4.9222846031188965,
      "learning_rate": 7.583642677819557e-05,
      "loss": 2.828347396850586,
      "memory(GiB)": 77.56,
      "step": 38180,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.439814
    },
    {
      "epoch": 1.6359624694743156,
      "grad_norm": 4.484988212585449,
      "learning_rate": 7.583066485371764e-05,
      "loss": 2.1581960678100587,
      "memory(GiB)": 77.56,
      "step": 38185,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.439828
    },
    {
      "epoch": 1.6361766848035646,
      "grad_norm": 4.387702465057373,
      "learning_rate": 7.582490246129299e-05,
      "loss": 2.6517467498779297,
      "memory(GiB)": 77.56,
      "step": 38190,
      "token_acc": 0.4551971326164875,
      "train_speed(iter/s)": 1.439864
    },
    {
      "epoch": 1.6363909001328136,
      "grad_norm": 3.7711524963378906,
      "learning_rate": 7.5819139601026e-05,
      "loss": 2.3045331954956056,
      "memory(GiB)": 77.56,
      "step": 38195,
      "token_acc": 0.5374149659863946,
      "train_speed(iter/s)": 1.439875
    },
    {
      "epoch": 1.6366051154620624,
      "grad_norm": 4.847904682159424,
      "learning_rate": 7.581337627302107e-05,
      "loss": 2.541432571411133,
      "memory(GiB)": 77.56,
      "step": 38200,
      "token_acc": 0.42435424354243545,
      "train_speed(iter/s)": 1.439916
    },
    {
      "epoch": 1.6368193307913115,
      "grad_norm": 4.269672870635986,
      "learning_rate": 7.580761247738264e-05,
      "loss": 2.569334030151367,
      "memory(GiB)": 77.56,
      "step": 38205,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.439947
    },
    {
      "epoch": 1.6370335461205605,
      "grad_norm": 4.63272762298584,
      "learning_rate": 7.580184821421508e-05,
      "loss": 2.4549869537353515,
      "memory(GiB)": 77.56,
      "step": 38210,
      "token_acc": 0.48348348348348347,
      "train_speed(iter/s)": 1.439939
    },
    {
      "epoch": 1.6372477614498093,
      "grad_norm": 5.64998197555542,
      "learning_rate": 7.579608348362284e-05,
      "loss": 2.4259469985961912,
      "memory(GiB)": 77.56,
      "step": 38215,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.439962
    },
    {
      "epoch": 1.6374619767790584,
      "grad_norm": 7.6386847496032715,
      "learning_rate": 7.579031828571035e-05,
      "loss": 2.329135513305664,
      "memory(GiB)": 77.56,
      "step": 38220,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.439932
    },
    {
      "epoch": 1.6376761921083074,
      "grad_norm": 5.141870498657227,
      "learning_rate": 7.578455262058204e-05,
      "loss": 2.6745733261108398,
      "memory(GiB)": 77.56,
      "step": 38225,
      "token_acc": 0.43597560975609756,
      "train_speed(iter/s)": 1.439968
    },
    {
      "epoch": 1.6378904074375562,
      "grad_norm": 4.4243268966674805,
      "learning_rate": 7.577878648834241e-05,
      "loss": 2.4369958877563476,
      "memory(GiB)": 77.56,
      "step": 38230,
      "token_acc": 0.4781021897810219,
      "train_speed(iter/s)": 1.43998
    },
    {
      "epoch": 1.6381046227668052,
      "grad_norm": 4.563777446746826,
      "learning_rate": 7.577301988909583e-05,
      "loss": 2.731121635437012,
      "memory(GiB)": 77.56,
      "step": 38235,
      "token_acc": 0.4709480122324159,
      "train_speed(iter/s)": 1.439983
    },
    {
      "epoch": 1.6383188380960543,
      "grad_norm": 4.9535675048828125,
      "learning_rate": 7.576725282294683e-05,
      "loss": 2.455893707275391,
      "memory(GiB)": 77.56,
      "step": 38240,
      "token_acc": 0.5029761904761905,
      "train_speed(iter/s)": 1.439996
    },
    {
      "epoch": 1.638533053425303,
      "grad_norm": 5.448330879211426,
      "learning_rate": 7.576148528999989e-05,
      "loss": 2.8143096923828126,
      "memory(GiB)": 77.56,
      "step": 38245,
      "token_acc": 0.4412811387900356,
      "train_speed(iter/s)": 1.439993
    },
    {
      "epoch": 1.6387472687545521,
      "grad_norm": 6.893537998199463,
      "learning_rate": 7.575571729035948e-05,
      "loss": 2.2737138748168944,
      "memory(GiB)": 77.56,
      "step": 38250,
      "token_acc": 0.5772357723577236,
      "train_speed(iter/s)": 1.440021
    },
    {
      "epoch": 1.6389614840838012,
      "grad_norm": 4.84488582611084,
      "learning_rate": 7.574994882413007e-05,
      "loss": 2.4641876220703125,
      "memory(GiB)": 77.56,
      "step": 38255,
      "token_acc": 0.4820359281437126,
      "train_speed(iter/s)": 1.44001
    },
    {
      "epoch": 1.63917569941305,
      "grad_norm": 4.760132789611816,
      "learning_rate": 7.57441798914162e-05,
      "loss": 2.5985626220703124,
      "memory(GiB)": 77.56,
      "step": 38260,
      "token_acc": 0.42679127725856697,
      "train_speed(iter/s)": 1.44002
    },
    {
      "epoch": 1.639389914742299,
      "grad_norm": 7.235931396484375,
      "learning_rate": 7.573841049232236e-05,
      "loss": 2.6217987060546877,
      "memory(GiB)": 77.56,
      "step": 38265,
      "token_acc": 0.4227129337539432,
      "train_speed(iter/s)": 1.44009
    },
    {
      "epoch": 1.639604130071548,
      "grad_norm": 5.110458850860596,
      "learning_rate": 7.573264062695304e-05,
      "loss": 2.593521499633789,
      "memory(GiB)": 77.56,
      "step": 38270,
      "token_acc": 0.4727272727272727,
      "train_speed(iter/s)": 1.440108
    },
    {
      "epoch": 1.6398183454007969,
      "grad_norm": 4.615773677825928,
      "learning_rate": 7.572687029541283e-05,
      "loss": 2.462036895751953,
      "memory(GiB)": 77.56,
      "step": 38275,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 1.6400325607300459,
      "grad_norm": 4.727434158325195,
      "learning_rate": 7.572109949780624e-05,
      "loss": 2.449875259399414,
      "memory(GiB)": 77.56,
      "step": 38280,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.44009
    },
    {
      "epoch": 1.640246776059295,
      "grad_norm": 4.6819047927856445,
      "learning_rate": 7.571532823423777e-05,
      "loss": 2.401778221130371,
      "memory(GiB)": 77.56,
      "step": 38285,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.440032
    },
    {
      "epoch": 1.6404609913885437,
      "grad_norm": 4.814586162567139,
      "learning_rate": 7.570955650481202e-05,
      "loss": 2.6406890869140627,
      "memory(GiB)": 77.56,
      "step": 38290,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.44001
    },
    {
      "epoch": 1.6406752067177928,
      "grad_norm": 5.465790748596191,
      "learning_rate": 7.570378430963355e-05,
      "loss": 2.783735466003418,
      "memory(GiB)": 77.56,
      "step": 38295,
      "token_acc": 0.44565217391304346,
      "train_speed(iter/s)": 1.440068
    },
    {
      "epoch": 1.6408894220470418,
      "grad_norm": 4.85542631149292,
      "learning_rate": 7.56980116488069e-05,
      "loss": 2.239752769470215,
      "memory(GiB)": 77.56,
      "step": 38300,
      "token_acc": 0.525,
      "train_speed(iter/s)": 1.440086
    },
    {
      "epoch": 1.6411036373762906,
      "grad_norm": 5.765982151031494,
      "learning_rate": 7.569223852243666e-05,
      "loss": 2.829648971557617,
      "memory(GiB)": 77.56,
      "step": 38305,
      "token_acc": 0.43884892086330934,
      "train_speed(iter/s)": 1.440099
    },
    {
      "epoch": 1.6413178527055396,
      "grad_norm": 6.030619144439697,
      "learning_rate": 7.568646493062742e-05,
      "loss": 2.4924142837524412,
      "memory(GiB)": 77.56,
      "step": 38310,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.440103
    },
    {
      "epoch": 1.6415320680347887,
      "grad_norm": 4.4107561111450195,
      "learning_rate": 7.568069087348377e-05,
      "loss": 2.5773033142089843,
      "memory(GiB)": 77.56,
      "step": 38315,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.440078
    },
    {
      "epoch": 1.6417462833640375,
      "grad_norm": 5.170173645019531,
      "learning_rate": 7.567491635111033e-05,
      "loss": 2.4269287109375,
      "memory(GiB)": 77.56,
      "step": 38320,
      "token_acc": 0.5167785234899329,
      "train_speed(iter/s)": 1.440132
    },
    {
      "epoch": 1.6419604986932865,
      "grad_norm": 6.318161964416504,
      "learning_rate": 7.566914136361168e-05,
      "loss": 2.3796905517578124,
      "memory(GiB)": 77.56,
      "step": 38325,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.44014
    },
    {
      "epoch": 1.6421747140225356,
      "grad_norm": 4.667794227600098,
      "learning_rate": 7.566336591109245e-05,
      "loss": 2.5578704833984376,
      "memory(GiB)": 77.56,
      "step": 38330,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.440155
    },
    {
      "epoch": 1.6423889293517844,
      "grad_norm": 5.906712532043457,
      "learning_rate": 7.565758999365728e-05,
      "loss": 2.472130012512207,
      "memory(GiB)": 77.56,
      "step": 38335,
      "token_acc": 0.4770992366412214,
      "train_speed(iter/s)": 1.440074
    },
    {
      "epoch": 1.6426031446810334,
      "grad_norm": 4.678191661834717,
      "learning_rate": 7.56518136114108e-05,
      "loss": 2.568874740600586,
      "memory(GiB)": 77.56,
      "step": 38340,
      "token_acc": 0.49185667752442996,
      "train_speed(iter/s)": 1.440087
    },
    {
      "epoch": 1.6428173600102824,
      "grad_norm": 4.495857238769531,
      "learning_rate": 7.564603676445765e-05,
      "loss": 2.4461050033569336,
      "memory(GiB)": 77.56,
      "step": 38345,
      "token_acc": 0.4717514124293785,
      "train_speed(iter/s)": 1.440071
    },
    {
      "epoch": 1.6430315753395313,
      "grad_norm": 4.390526294708252,
      "learning_rate": 7.56402594529025e-05,
      "loss": 2.401481819152832,
      "memory(GiB)": 77.56,
      "step": 38350,
      "token_acc": 0.512280701754386,
      "train_speed(iter/s)": 1.440063
    },
    {
      "epoch": 1.6432457906687803,
      "grad_norm": 5.889782905578613,
      "learning_rate": 7.563448167684996e-05,
      "loss": 2.2590831756591796,
      "memory(GiB)": 77.56,
      "step": 38355,
      "token_acc": 0.5622641509433962,
      "train_speed(iter/s)": 1.440017
    },
    {
      "epoch": 1.6434600059980293,
      "grad_norm": 4.997591972351074,
      "learning_rate": 7.562870343640478e-05,
      "loss": 2.0358966827392577,
      "memory(GiB)": 77.56,
      "step": 38360,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.440009
    },
    {
      "epoch": 1.6436742213272781,
      "grad_norm": 6.388982772827148,
      "learning_rate": 7.562292473167158e-05,
      "loss": 2.380128860473633,
      "memory(GiB)": 77.56,
      "step": 38365,
      "token_acc": 0.5642023346303502,
      "train_speed(iter/s)": 1.440038
    },
    {
      "epoch": 1.6438884366565272,
      "grad_norm": 5.350227355957031,
      "learning_rate": 7.561714556275505e-05,
      "loss": 2.584260940551758,
      "memory(GiB)": 77.56,
      "step": 38370,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.439995
    },
    {
      "epoch": 1.6441026519857762,
      "grad_norm": 5.289626121520996,
      "learning_rate": 7.561136592975993e-05,
      "loss": 2.5791215896606445,
      "memory(GiB)": 77.56,
      "step": 38375,
      "token_acc": 0.44654088050314467,
      "train_speed(iter/s)": 1.43994
    },
    {
      "epoch": 1.644316867315025,
      "grad_norm": 5.083733558654785,
      "learning_rate": 7.560558583279085e-05,
      "loss": 2.2939130783081056,
      "memory(GiB)": 77.56,
      "step": 38380,
      "token_acc": 0.5347222222222222,
      "train_speed(iter/s)": 1.439968
    },
    {
      "epoch": 1.644531082644274,
      "grad_norm": 4.437898635864258,
      "learning_rate": 7.559980527195259e-05,
      "loss": 2.107828903198242,
      "memory(GiB)": 77.56,
      "step": 38385,
      "token_acc": 0.5431654676258992,
      "train_speed(iter/s)": 1.43989
    },
    {
      "epoch": 1.644745297973523,
      "grad_norm": 4.620337009429932,
      "learning_rate": 7.559402424734982e-05,
      "loss": 2.696457290649414,
      "memory(GiB)": 77.56,
      "step": 38390,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.439868
    },
    {
      "epoch": 1.644959513302772,
      "grad_norm": 4.596342086791992,
      "learning_rate": 7.558824275908732e-05,
      "loss": 2.35351619720459,
      "memory(GiB)": 77.56,
      "step": 38395,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.439897
    },
    {
      "epoch": 1.645173728632021,
      "grad_norm": 7.431744575500488,
      "learning_rate": 7.558246080726978e-05,
      "loss": 2.629029655456543,
      "memory(GiB)": 77.56,
      "step": 38400,
      "token_acc": 0.5215686274509804,
      "train_speed(iter/s)": 1.439876
    },
    {
      "epoch": 1.64538794396127,
      "grad_norm": 7.818251132965088,
      "learning_rate": 7.557667839200198e-05,
      "loss": 2.2264585494995117,
      "memory(GiB)": 77.56,
      "step": 38405,
      "token_acc": 0.49224806201550386,
      "train_speed(iter/s)": 1.439827
    },
    {
      "epoch": 1.6456021592905188,
      "grad_norm": 4.419662952423096,
      "learning_rate": 7.557089551338865e-05,
      "loss": 2.3536327362060545,
      "memory(GiB)": 77.56,
      "step": 38410,
      "token_acc": 0.4919093851132686,
      "train_speed(iter/s)": 1.439792
    },
    {
      "epoch": 1.6458163746197678,
      "grad_norm": 4.097082138061523,
      "learning_rate": 7.556511217153455e-05,
      "loss": 2.526063919067383,
      "memory(GiB)": 77.56,
      "step": 38415,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.439816
    },
    {
      "epoch": 1.6460305899490169,
      "grad_norm": 4.693516254425049,
      "learning_rate": 7.555932836654447e-05,
      "loss": 2.4798213958740236,
      "memory(GiB)": 77.56,
      "step": 38420,
      "token_acc": 0.4939759036144578,
      "train_speed(iter/s)": 1.439884
    },
    {
      "epoch": 1.6462448052782657,
      "grad_norm": 4.578656196594238,
      "learning_rate": 7.555354409852318e-05,
      "loss": 2.335084915161133,
      "memory(GiB)": 77.56,
      "step": 38425,
      "token_acc": 0.5171339563862928,
      "train_speed(iter/s)": 1.439919
    },
    {
      "epoch": 1.6464590206075147,
      "grad_norm": 8.257400512695312,
      "learning_rate": 7.554775936757545e-05,
      "loss": 2.684035301208496,
      "memory(GiB)": 77.56,
      "step": 38430,
      "token_acc": 0.46504559270516715,
      "train_speed(iter/s)": 1.439949
    },
    {
      "epoch": 1.6466732359367637,
      "grad_norm": 4.201570987701416,
      "learning_rate": 7.554197417380613e-05,
      "loss": 2.5370201110839843,
      "memory(GiB)": 77.56,
      "step": 38435,
      "token_acc": 0.4401294498381877,
      "train_speed(iter/s)": 1.439971
    },
    {
      "epoch": 1.6468874512660125,
      "grad_norm": 3.7691662311553955,
      "learning_rate": 7.553618851731996e-05,
      "loss": 2.388266754150391,
      "memory(GiB)": 77.56,
      "step": 38440,
      "token_acc": 0.4914772727272727,
      "train_speed(iter/s)": 1.439979
    },
    {
      "epoch": 1.6471016665952616,
      "grad_norm": 4.353640556335449,
      "learning_rate": 7.553040239822179e-05,
      "loss": 2.6990001678466795,
      "memory(GiB)": 77.56,
      "step": 38445,
      "token_acc": 0.4659090909090909,
      "train_speed(iter/s)": 1.44001
    },
    {
      "epoch": 1.6473158819245106,
      "grad_norm": 5.954075336456299,
      "learning_rate": 7.552461581661643e-05,
      "loss": 2.7798843383789062,
      "memory(GiB)": 77.56,
      "step": 38450,
      "token_acc": 0.453416149068323,
      "train_speed(iter/s)": 1.440027
    },
    {
      "epoch": 1.6475300972537594,
      "grad_norm": 4.807617664337158,
      "learning_rate": 7.55188287726087e-05,
      "loss": 2.469046211242676,
      "memory(GiB)": 77.56,
      "step": 38455,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.439986
    },
    {
      "epoch": 1.6477443125830085,
      "grad_norm": 5.687557220458984,
      "learning_rate": 7.551304126630345e-05,
      "loss": 2.815443229675293,
      "memory(GiB)": 77.56,
      "step": 38460,
      "token_acc": 0.42105263157894735,
      "train_speed(iter/s)": 1.439979
    },
    {
      "epoch": 1.6479585279122575,
      "grad_norm": 5.092597961425781,
      "learning_rate": 7.550725329780555e-05,
      "loss": 2.835649108886719,
      "memory(GiB)": 77.56,
      "step": 38465,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.440005
    },
    {
      "epoch": 1.6481727432415063,
      "grad_norm": 4.26067590713501,
      "learning_rate": 7.550146486721981e-05,
      "loss": 2.2664522171020507,
      "memory(GiB)": 77.56,
      "step": 38470,
      "token_acc": 0.4862068965517241,
      "train_speed(iter/s)": 1.439995
    },
    {
      "epoch": 1.6483869585707553,
      "grad_norm": 4.944790363311768,
      "learning_rate": 7.54956759746511e-05,
      "loss": 2.7993053436279296,
      "memory(GiB)": 77.56,
      "step": 38475,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.439973
    },
    {
      "epoch": 1.6486011739000044,
      "grad_norm": 3.6839516162872314,
      "learning_rate": 7.548988662020433e-05,
      "loss": 2.6350074768066407,
      "memory(GiB)": 77.56,
      "step": 38480,
      "token_acc": 0.4830287206266319,
      "train_speed(iter/s)": 1.439994
    },
    {
      "epoch": 1.6488153892292532,
      "grad_norm": 4.4718523025512695,
      "learning_rate": 7.548409680398433e-05,
      "loss": 2.329062271118164,
      "memory(GiB)": 77.56,
      "step": 38485,
      "token_acc": 0.4749034749034749,
      "train_speed(iter/s)": 1.440016
    },
    {
      "epoch": 1.6490296045585022,
      "grad_norm": 5.5267333984375,
      "learning_rate": 7.547830652609601e-05,
      "loss": 2.7577997207641602,
      "memory(GiB)": 77.56,
      "step": 38490,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.43998
    },
    {
      "epoch": 1.6492438198877513,
      "grad_norm": 6.347685813903809,
      "learning_rate": 7.547251578664427e-05,
      "loss": 2.4713376998901366,
      "memory(GiB)": 77.56,
      "step": 38495,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439957
    },
    {
      "epoch": 1.649458035217,
      "grad_norm": 4.946099281311035,
      "learning_rate": 7.546672458573402e-05,
      "loss": 2.0630680084228517,
      "memory(GiB)": 77.56,
      "step": 38500,
      "token_acc": 0.5801886792452831,
      "train_speed(iter/s)": 1.43996
    },
    {
      "epoch": 1.649458035217,
      "eval_loss": 2.434729814529419,
      "eval_runtime": 15.0419,
      "eval_samples_per_second": 6.648,
      "eval_steps_per_second": 6.648,
      "eval_token_acc": 0.43757725587144625,
      "step": 38500
    },
    {
      "epoch": 1.649672250546249,
      "grad_norm": 5.806351661682129,
      "learning_rate": 7.546093292347016e-05,
      "loss": 2.335239028930664,
      "memory(GiB)": 77.56,
      "step": 38505,
      "token_acc": 0.445970695970696,
      "train_speed(iter/s)": 1.439066
    },
    {
      "epoch": 1.6498864658754981,
      "grad_norm": 8.821775436401367,
      "learning_rate": 7.545514079995762e-05,
      "loss": 2.767985534667969,
      "memory(GiB)": 77.56,
      "step": 38510,
      "token_acc": 0.484,
      "train_speed(iter/s)": 1.439032
    },
    {
      "epoch": 1.650100681204747,
      "grad_norm": 5.096793174743652,
      "learning_rate": 7.544934821530132e-05,
      "loss": 2.5442153930664064,
      "memory(GiB)": 77.56,
      "step": 38515,
      "token_acc": 0.4784172661870504,
      "train_speed(iter/s)": 1.439059
    },
    {
      "epoch": 1.650314896533996,
      "grad_norm": 4.561770915985107,
      "learning_rate": 7.544355516960621e-05,
      "loss": 2.6692693710327147,
      "memory(GiB)": 77.56,
      "step": 38520,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.439042
    },
    {
      "epoch": 1.650529111863245,
      "grad_norm": 5.6831769943237305,
      "learning_rate": 7.543776166297723e-05,
      "loss": 2.511821746826172,
      "memory(GiB)": 77.56,
      "step": 38525,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.439069
    },
    {
      "epoch": 1.6507433271924938,
      "grad_norm": 7.861547946929932,
      "learning_rate": 7.543196769551931e-05,
      "loss": 2.3939016342163084,
      "memory(GiB)": 77.56,
      "step": 38530,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.439128
    },
    {
      "epoch": 1.6509575425217429,
      "grad_norm": 3.8708057403564453,
      "learning_rate": 7.542617326733747e-05,
      "loss": 2.6797657012939453,
      "memory(GiB)": 77.56,
      "step": 38535,
      "token_acc": 0.46715328467153283,
      "train_speed(iter/s)": 1.439034
    },
    {
      "epoch": 1.651171757850992,
      "grad_norm": 4.398819446563721,
      "learning_rate": 7.542037837853664e-05,
      "loss": 2.5842761993408203,
      "memory(GiB)": 77.56,
      "step": 38540,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.439059
    },
    {
      "epoch": 1.6513859731802407,
      "grad_norm": 4.454665184020996,
      "learning_rate": 7.541458302922179e-05,
      "loss": 2.436616897583008,
      "memory(GiB)": 77.56,
      "step": 38545,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.439032
    },
    {
      "epoch": 1.6516001885094898,
      "grad_norm": 5.831594944000244,
      "learning_rate": 7.540878721949796e-05,
      "loss": 2.6067119598388673,
      "memory(GiB)": 77.56,
      "step": 38550,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.439052
    },
    {
      "epoch": 1.6518144038387388,
      "grad_norm": 6.809323310852051,
      "learning_rate": 7.54029909494701e-05,
      "loss": 2.7003225326538085,
      "memory(GiB)": 77.56,
      "step": 38555,
      "token_acc": 0.39755351681957185,
      "train_speed(iter/s)": 1.439106
    },
    {
      "epoch": 1.6520286191679876,
      "grad_norm": 4.65964937210083,
      "learning_rate": 7.539719421924322e-05,
      "loss": 2.321906852722168,
      "memory(GiB)": 77.56,
      "step": 38560,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.439127
    },
    {
      "epoch": 1.6522428344972366,
      "grad_norm": 4.276027202606201,
      "learning_rate": 7.539139702892235e-05,
      "loss": 2.284992790222168,
      "memory(GiB)": 77.56,
      "step": 38565,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.439097
    },
    {
      "epoch": 1.6524570498264857,
      "grad_norm": 5.731977939605713,
      "learning_rate": 7.538559937861251e-05,
      "loss": 2.26531982421875,
      "memory(GiB)": 77.56,
      "step": 38570,
      "token_acc": 0.5230125523012552,
      "train_speed(iter/s)": 1.4391
    },
    {
      "epoch": 1.6526712651557345,
      "grad_norm": 5.116015434265137,
      "learning_rate": 7.53798012684187e-05,
      "loss": 2.5026100158691404,
      "memory(GiB)": 77.56,
      "step": 38575,
      "token_acc": 0.4881656804733728,
      "train_speed(iter/s)": 1.439141
    },
    {
      "epoch": 1.6528854804849835,
      "grad_norm": 5.174070835113525,
      "learning_rate": 7.537400269844601e-05,
      "loss": 2.5735965728759767,
      "memory(GiB)": 77.56,
      "step": 38580,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.439061
    },
    {
      "epoch": 1.6530996958142326,
      "grad_norm": 4.859527111053467,
      "learning_rate": 7.536820366879946e-05,
      "loss": 2.5687503814697266,
      "memory(GiB)": 77.56,
      "step": 38585,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.438995
    },
    {
      "epoch": 1.6533139111434814,
      "grad_norm": 4.956950664520264,
      "learning_rate": 7.536240417958409e-05,
      "loss": 2.701076316833496,
      "memory(GiB)": 77.56,
      "step": 38590,
      "token_acc": 0.42724458204334365,
      "train_speed(iter/s)": 1.43905
    },
    {
      "epoch": 1.6535281264727304,
      "grad_norm": 6.706456184387207,
      "learning_rate": 7.535660423090498e-05,
      "loss": 2.4777870178222656,
      "memory(GiB)": 77.56,
      "step": 38595,
      "token_acc": 0.5170940170940171,
      "train_speed(iter/s)": 1.439057
    },
    {
      "epoch": 1.6537423418019794,
      "grad_norm": 4.761669635772705,
      "learning_rate": 7.535080382286718e-05,
      "loss": 2.692028045654297,
      "memory(GiB)": 77.56,
      "step": 38600,
      "token_acc": 0.4612903225806452,
      "train_speed(iter/s)": 1.43902
    },
    {
      "epoch": 1.6539565571312282,
      "grad_norm": 6.3677520751953125,
      "learning_rate": 7.534500295557581e-05,
      "loss": 2.6435422897338867,
      "memory(GiB)": 77.56,
      "step": 38605,
      "token_acc": 0.4470588235294118,
      "train_speed(iter/s)": 1.438996
    },
    {
      "epoch": 1.6541707724604773,
      "grad_norm": 6.014742374420166,
      "learning_rate": 7.533920162913592e-05,
      "loss": 2.396424674987793,
      "memory(GiB)": 77.56,
      "step": 38610,
      "token_acc": 0.5228758169934641,
      "train_speed(iter/s)": 1.438997
    },
    {
      "epoch": 1.6543849877897263,
      "grad_norm": 5.6689581871032715,
      "learning_rate": 7.533339984365265e-05,
      "loss": 2.421131706237793,
      "memory(GiB)": 77.56,
      "step": 38615,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.439014
    },
    {
      "epoch": 1.6545992031189751,
      "grad_norm": 4.582613945007324,
      "learning_rate": 7.532759759923105e-05,
      "loss": 2.5647411346435547,
      "memory(GiB)": 77.56,
      "step": 38620,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.439036
    },
    {
      "epoch": 1.6548134184482242,
      "grad_norm": 5.181107044219971,
      "learning_rate": 7.532179489597626e-05,
      "loss": 2.500691223144531,
      "memory(GiB)": 77.56,
      "step": 38625,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439077
    },
    {
      "epoch": 1.6550276337774732,
      "grad_norm": 9.546935081481934,
      "learning_rate": 7.531599173399342e-05,
      "loss": 2.7980955123901365,
      "memory(GiB)": 77.56,
      "step": 38630,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.439114
    },
    {
      "epoch": 1.655241849106722,
      "grad_norm": 3.835843563079834,
      "learning_rate": 7.531018811338764e-05,
      "loss": 2.670639419555664,
      "memory(GiB)": 77.56,
      "step": 38635,
      "token_acc": 0.4862068965517241,
      "train_speed(iter/s)": 1.439121
    },
    {
      "epoch": 1.655456064435971,
      "grad_norm": 5.509315490722656,
      "learning_rate": 7.530438403426403e-05,
      "loss": 2.517391014099121,
      "memory(GiB)": 77.56,
      "step": 38640,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.439153
    },
    {
      "epoch": 1.65567027976522,
      "grad_norm": 5.535069942474365,
      "learning_rate": 7.529857949672778e-05,
      "loss": 2.7095333099365235,
      "memory(GiB)": 77.56,
      "step": 38645,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.439158
    },
    {
      "epoch": 1.6558844950944689,
      "grad_norm": 7.169466495513916,
      "learning_rate": 7.529277450088405e-05,
      "loss": 2.417226028442383,
      "memory(GiB)": 77.56,
      "step": 38650,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.439155
    },
    {
      "epoch": 1.656098710423718,
      "grad_norm": 4.585759162902832,
      "learning_rate": 7.528696904683797e-05,
      "loss": 2.039259910583496,
      "memory(GiB)": 77.56,
      "step": 38655,
      "token_acc": 0.5546218487394958,
      "train_speed(iter/s)": 1.439169
    },
    {
      "epoch": 1.656312925752967,
      "grad_norm": 4.929565906524658,
      "learning_rate": 7.528116313469473e-05,
      "loss": 2.458763313293457,
      "memory(GiB)": 77.56,
      "step": 38660,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.439206
    },
    {
      "epoch": 1.6565271410822158,
      "grad_norm": 4.293598651885986,
      "learning_rate": 7.52753567645595e-05,
      "loss": 2.4261655807495117,
      "memory(GiB)": 77.56,
      "step": 38665,
      "token_acc": 0.4485049833887043,
      "train_speed(iter/s)": 1.439233
    },
    {
      "epoch": 1.6567413564114648,
      "grad_norm": 6.021685600280762,
      "learning_rate": 7.526954993653747e-05,
      "loss": 2.4622310638427733,
      "memory(GiB)": 77.56,
      "step": 38670,
      "token_acc": 0.4657534246575342,
      "train_speed(iter/s)": 1.439278
    },
    {
      "epoch": 1.6569555717407138,
      "grad_norm": 3.9898576736450195,
      "learning_rate": 7.526374265073384e-05,
      "loss": 2.586587905883789,
      "memory(GiB)": 77.56,
      "step": 38675,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.439276
    },
    {
      "epoch": 1.6571697870699627,
      "grad_norm": 4.119765281677246,
      "learning_rate": 7.525793490725381e-05,
      "loss": 2.7099660873413085,
      "memory(GiB)": 77.56,
      "step": 38680,
      "token_acc": 0.46397694524495675,
      "train_speed(iter/s)": 1.439286
    },
    {
      "epoch": 1.6573840023992117,
      "grad_norm": 4.754705429077148,
      "learning_rate": 7.525212670620261e-05,
      "loss": 2.5766674041748048,
      "memory(GiB)": 77.56,
      "step": 38685,
      "token_acc": 0.483739837398374,
      "train_speed(iter/s)": 1.439322
    },
    {
      "epoch": 1.6575982177284607,
      "grad_norm": 4.713914394378662,
      "learning_rate": 7.524631804768543e-05,
      "loss": 2.9598012924194337,
      "memory(GiB)": 77.56,
      "step": 38690,
      "token_acc": 0.4463087248322148,
      "train_speed(iter/s)": 1.439345
    },
    {
      "epoch": 1.6578124330577095,
      "grad_norm": 5.343391418457031,
      "learning_rate": 7.524050893180752e-05,
      "loss": 2.514819526672363,
      "memory(GiB)": 77.56,
      "step": 38695,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.439354
    },
    {
      "epoch": 1.6580266483869586,
      "grad_norm": 5.777933597564697,
      "learning_rate": 7.523469935867411e-05,
      "loss": 2.5260684967041014,
      "memory(GiB)": 77.56,
      "step": 38700,
      "token_acc": 0.42902208201892744,
      "train_speed(iter/s)": 1.439348
    },
    {
      "epoch": 1.6582408637162076,
      "grad_norm": 4.930129051208496,
      "learning_rate": 7.522888932839045e-05,
      "loss": 2.4257545471191406,
      "memory(GiB)": 77.56,
      "step": 38705,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.439353
    },
    {
      "epoch": 1.6584550790454564,
      "grad_norm": 4.957555770874023,
      "learning_rate": 7.52230788410618e-05,
      "loss": 2.6519437789916993,
      "memory(GiB)": 77.56,
      "step": 38710,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.439384
    },
    {
      "epoch": 1.6586692943747054,
      "grad_norm": 5.4925007820129395,
      "learning_rate": 7.52172678967934e-05,
      "loss": 2.5465240478515625,
      "memory(GiB)": 77.56,
      "step": 38715,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.439379
    },
    {
      "epoch": 1.6588835097039545,
      "grad_norm": 5.156786918640137,
      "learning_rate": 7.521145649569054e-05,
      "loss": 2.633748435974121,
      "memory(GiB)": 77.56,
      "step": 38720,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.43942
    },
    {
      "epoch": 1.6590977250332033,
      "grad_norm": 6.931817531585693,
      "learning_rate": 7.520564463785851e-05,
      "loss": 2.6720380783081055,
      "memory(GiB)": 77.56,
      "step": 38725,
      "token_acc": 0.4280701754385965,
      "train_speed(iter/s)": 1.439383
    },
    {
      "epoch": 1.6593119403624523,
      "grad_norm": 5.221216201782227,
      "learning_rate": 7.519983232340258e-05,
      "loss": 2.3263933181762697,
      "memory(GiB)": 77.56,
      "step": 38730,
      "token_acc": 0.49429657794676807,
      "train_speed(iter/s)": 1.439357
    },
    {
      "epoch": 1.6595261556917014,
      "grad_norm": 6.273737907409668,
      "learning_rate": 7.519401955242803e-05,
      "loss": 2.4995725631713865,
      "memory(GiB)": 77.56,
      "step": 38735,
      "token_acc": 0.4978165938864629,
      "train_speed(iter/s)": 1.43939
    },
    {
      "epoch": 1.6597403710209502,
      "grad_norm": 5.082566261291504,
      "learning_rate": 7.518820632504021e-05,
      "loss": 2.552480125427246,
      "memory(GiB)": 77.56,
      "step": 38740,
      "token_acc": 0.4604904632152589,
      "train_speed(iter/s)": 1.439362
    },
    {
      "epoch": 1.6599545863501992,
      "grad_norm": 4.399657726287842,
      "learning_rate": 7.518239264134439e-05,
      "loss": 2.3195480346679687,
      "memory(GiB)": 77.56,
      "step": 38745,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.439407
    },
    {
      "epoch": 1.6601688016794482,
      "grad_norm": 5.266880989074707,
      "learning_rate": 7.51765785014459e-05,
      "loss": 2.695078468322754,
      "memory(GiB)": 77.56,
      "step": 38750,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.439442
    },
    {
      "epoch": 1.660383017008697,
      "grad_norm": 7.205869674682617,
      "learning_rate": 7.517076390545007e-05,
      "loss": 2.627971076965332,
      "memory(GiB)": 77.56,
      "step": 38755,
      "token_acc": 0.47604790419161674,
      "train_speed(iter/s)": 1.439456
    },
    {
      "epoch": 1.660597232337946,
      "grad_norm": 5.870373725891113,
      "learning_rate": 7.516494885346223e-05,
      "loss": 2.6181596755981444,
      "memory(GiB)": 77.56,
      "step": 38760,
      "token_acc": 0.44404332129963897,
      "train_speed(iter/s)": 1.439481
    },
    {
      "epoch": 1.6608114476671951,
      "grad_norm": 5.468871116638184,
      "learning_rate": 7.515913334558778e-05,
      "loss": 2.3515453338623047,
      "memory(GiB)": 77.56,
      "step": 38765,
      "token_acc": 0.49812734082397003,
      "train_speed(iter/s)": 1.43949
    },
    {
      "epoch": 1.661025662996444,
      "grad_norm": 4.76951789855957,
      "learning_rate": 7.515331738193199e-05,
      "loss": 2.6942441940307615,
      "memory(GiB)": 77.56,
      "step": 38770,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.439486
    },
    {
      "epoch": 1.661239878325693,
      "grad_norm": 6.500946521759033,
      "learning_rate": 7.514750096260026e-05,
      "loss": 2.636835479736328,
      "memory(GiB)": 77.56,
      "step": 38775,
      "token_acc": 0.4430379746835443,
      "train_speed(iter/s)": 1.439481
    },
    {
      "epoch": 1.661454093654942,
      "grad_norm": 6.966447830200195,
      "learning_rate": 7.514168408769798e-05,
      "loss": 2.415618896484375,
      "memory(GiB)": 77.56,
      "step": 38780,
      "token_acc": 0.4653179190751445,
      "train_speed(iter/s)": 1.439532
    },
    {
      "epoch": 1.6616683089841908,
      "grad_norm": 4.5493083000183105,
      "learning_rate": 7.513586675733049e-05,
      "loss": 2.31470947265625,
      "memory(GiB)": 77.56,
      "step": 38785,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.439529
    },
    {
      "epoch": 1.6618825243134399,
      "grad_norm": 4.842322826385498,
      "learning_rate": 7.51300489716032e-05,
      "loss": 2.6461009979248047,
      "memory(GiB)": 77.56,
      "step": 38790,
      "token_acc": 0.4358108108108108,
      "train_speed(iter/s)": 1.439518
    },
    {
      "epoch": 1.662096739642689,
      "grad_norm": 5.590895175933838,
      "learning_rate": 7.51242307306215e-05,
      "loss": 2.746798515319824,
      "memory(GiB)": 77.56,
      "step": 38795,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.439568
    },
    {
      "epoch": 1.6623109549719377,
      "grad_norm": 5.710389137268066,
      "learning_rate": 7.511841203449079e-05,
      "loss": 2.606972885131836,
      "memory(GiB)": 77.56,
      "step": 38800,
      "token_acc": 0.46715328467153283,
      "train_speed(iter/s)": 1.439585
    },
    {
      "epoch": 1.6625251703011867,
      "grad_norm": 5.101905345916748,
      "learning_rate": 7.511259288331649e-05,
      "loss": 2.4387020111083983,
      "memory(GiB)": 77.56,
      "step": 38805,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.439534
    },
    {
      "epoch": 1.6627393856304358,
      "grad_norm": 6.93477725982666,
      "learning_rate": 7.510677327720401e-05,
      "loss": 2.5429611206054688,
      "memory(GiB)": 77.56,
      "step": 38810,
      "token_acc": 0.44666666666666666,
      "train_speed(iter/s)": 1.439528
    },
    {
      "epoch": 1.6629536009596846,
      "grad_norm": 4.492660999298096,
      "learning_rate": 7.510095321625878e-05,
      "loss": 2.3240121841430663,
      "memory(GiB)": 77.56,
      "step": 38815,
      "token_acc": 0.4979919678714859,
      "train_speed(iter/s)": 1.439546
    },
    {
      "epoch": 1.6631678162889336,
      "grad_norm": 3.215885639190674,
      "learning_rate": 7.509513270058624e-05,
      "loss": 2.4668886184692385,
      "memory(GiB)": 77.56,
      "step": 38820,
      "token_acc": 0.5018181818181818,
      "train_speed(iter/s)": 1.439583
    },
    {
      "epoch": 1.6633820316181827,
      "grad_norm": 4.944400787353516,
      "learning_rate": 7.50893117302918e-05,
      "loss": 2.5129093170166015,
      "memory(GiB)": 77.56,
      "step": 38825,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.439581
    },
    {
      "epoch": 1.6635962469474315,
      "grad_norm": 6.282962322235107,
      "learning_rate": 7.508349030548099e-05,
      "loss": 2.451171875,
      "memory(GiB)": 77.56,
      "step": 38830,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.439536
    },
    {
      "epoch": 1.6638104622766805,
      "grad_norm": 7.855851650238037,
      "learning_rate": 7.507766842625918e-05,
      "loss": 2.6627079010009767,
      "memory(GiB)": 77.56,
      "step": 38835,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.439547
    },
    {
      "epoch": 1.6640246776059295,
      "grad_norm": 5.464412689208984,
      "learning_rate": 7.50718460927319e-05,
      "loss": 2.3441877365112305,
      "memory(GiB)": 77.56,
      "step": 38840,
      "token_acc": 0.4919093851132686,
      "train_speed(iter/s)": 1.439553
    },
    {
      "epoch": 1.6642388929351783,
      "grad_norm": 4.975652694702148,
      "learning_rate": 7.506602330500462e-05,
      "loss": 2.6074970245361326,
      "memory(GiB)": 77.56,
      "step": 38845,
      "token_acc": 0.44200626959247646,
      "train_speed(iter/s)": 1.439609
    },
    {
      "epoch": 1.6644531082644274,
      "grad_norm": 5.119026184082031,
      "learning_rate": 7.506020006318279e-05,
      "loss": 2.2321935653686524,
      "memory(GiB)": 77.56,
      "step": 38850,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.439597
    },
    {
      "epoch": 1.6646673235936764,
      "grad_norm": 4.505013465881348,
      "learning_rate": 7.505437636737196e-05,
      "loss": 2.5621501922607424,
      "memory(GiB)": 77.56,
      "step": 38855,
      "token_acc": 0.4662576687116564,
      "train_speed(iter/s)": 1.439573
    },
    {
      "epoch": 1.6648815389229252,
      "grad_norm": 6.21051025390625,
      "learning_rate": 7.504855221767757e-05,
      "loss": 2.8910472869873045,
      "memory(GiB)": 77.56,
      "step": 38860,
      "token_acc": 0.41379310344827586,
      "train_speed(iter/s)": 1.439576
    },
    {
      "epoch": 1.6650957542521743,
      "grad_norm": 5.030711650848389,
      "learning_rate": 7.504272761420517e-05,
      "loss": 2.554920768737793,
      "memory(GiB)": 77.56,
      "step": 38865,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.439508
    },
    {
      "epoch": 1.6653099695814233,
      "grad_norm": 3.88623046875,
      "learning_rate": 7.503690255706026e-05,
      "loss": 2.198586082458496,
      "memory(GiB)": 77.56,
      "step": 38870,
      "token_acc": 0.5086505190311419,
      "train_speed(iter/s)": 1.439468
    },
    {
      "epoch": 1.6655241849106721,
      "grad_norm": 4.617269515991211,
      "learning_rate": 7.503107704634838e-05,
      "loss": 2.6439130783081053,
      "memory(GiB)": 77.56,
      "step": 38875,
      "token_acc": 0.4575342465753425,
      "train_speed(iter/s)": 1.439458
    },
    {
      "epoch": 1.6657384002399211,
      "grad_norm": 6.865589618682861,
      "learning_rate": 7.502525108217506e-05,
      "loss": 2.1496143341064453,
      "memory(GiB)": 77.56,
      "step": 38880,
      "token_acc": 0.5587044534412956,
      "train_speed(iter/s)": 1.439486
    },
    {
      "epoch": 1.6659526155691702,
      "grad_norm": 5.5975022315979,
      "learning_rate": 7.501942466464584e-05,
      "loss": 2.6099880218505858,
      "memory(GiB)": 77.56,
      "step": 38885,
      "token_acc": 0.4529616724738676,
      "train_speed(iter/s)": 1.439506
    },
    {
      "epoch": 1.666166830898419,
      "grad_norm": 4.411645412445068,
      "learning_rate": 7.501359779386627e-05,
      "loss": 2.645598602294922,
      "memory(GiB)": 77.56,
      "step": 38890,
      "token_acc": 0.44805194805194803,
      "train_speed(iter/s)": 1.43954
    },
    {
      "epoch": 1.666381046227668,
      "grad_norm": 3.435889482498169,
      "learning_rate": 7.500777046994192e-05,
      "loss": 2.1572608947753906,
      "memory(GiB)": 77.56,
      "step": 38895,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.439553
    },
    {
      "epoch": 1.666595261556917,
      "grad_norm": 5.93528413772583,
      "learning_rate": 7.500194269297833e-05,
      "loss": 2.5333736419677733,
      "memory(GiB)": 77.56,
      "step": 38900,
      "token_acc": 0.494949494949495,
      "train_speed(iter/s)": 1.43958
    },
    {
      "epoch": 1.6668094768861659,
      "grad_norm": 4.217938423156738,
      "learning_rate": 7.49961144630811e-05,
      "loss": 2.4473541259765623,
      "memory(GiB)": 77.56,
      "step": 38905,
      "token_acc": 0.5075075075075075,
      "train_speed(iter/s)": 1.439609
    },
    {
      "epoch": 1.667023692215415,
      "grad_norm": 4.780902862548828,
      "learning_rate": 7.499028578035581e-05,
      "loss": 2.6170782089233398,
      "memory(GiB)": 77.56,
      "step": 38910,
      "token_acc": 0.4169014084507042,
      "train_speed(iter/s)": 1.439605
    },
    {
      "epoch": 1.667237907544664,
      "grad_norm": 6.853539943695068,
      "learning_rate": 7.498445664490807e-05,
      "loss": 2.4274330139160156,
      "memory(GiB)": 77.56,
      "step": 38915,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.439635
    },
    {
      "epoch": 1.6674521228739128,
      "grad_norm": 4.190219879150391,
      "learning_rate": 7.497862705684345e-05,
      "loss": 2.6190475463867187,
      "memory(GiB)": 77.56,
      "step": 38920,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.439677
    },
    {
      "epoch": 1.6676663382031618,
      "grad_norm": 7.996579170227051,
      "learning_rate": 7.497279701626756e-05,
      "loss": 2.4458694458007812,
      "memory(GiB)": 77.56,
      "step": 38925,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.439673
    },
    {
      "epoch": 1.6678805535324108,
      "grad_norm": 4.658965110778809,
      "learning_rate": 7.496696652328603e-05,
      "loss": 2.435190200805664,
      "memory(GiB)": 77.56,
      "step": 38930,
      "token_acc": 0.5083612040133779,
      "train_speed(iter/s)": 1.439674
    },
    {
      "epoch": 1.6680947688616596,
      "grad_norm": 4.744532108306885,
      "learning_rate": 7.496113557800446e-05,
      "loss": 2.380868148803711,
      "memory(GiB)": 77.56,
      "step": 38935,
      "token_acc": 0.5230263157894737,
      "train_speed(iter/s)": 1.439667
    },
    {
      "epoch": 1.6683089841909087,
      "grad_norm": 4.757541179656982,
      "learning_rate": 7.495530418052855e-05,
      "loss": 2.550774574279785,
      "memory(GiB)": 77.56,
      "step": 38940,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.439651
    },
    {
      "epoch": 1.6685231995201577,
      "grad_norm": 6.012129783630371,
      "learning_rate": 7.494947233096385e-05,
      "loss": 2.455175018310547,
      "memory(GiB)": 77.56,
      "step": 38945,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.439648
    },
    {
      "epoch": 1.6687374148494065,
      "grad_norm": 4.727448463439941,
      "learning_rate": 7.494364002941608e-05,
      "loss": 2.6159389495849608,
      "memory(GiB)": 77.56,
      "step": 38950,
      "token_acc": 0.4613003095975232,
      "train_speed(iter/s)": 1.439674
    },
    {
      "epoch": 1.6689516301786556,
      "grad_norm": 4.4705400466918945,
      "learning_rate": 7.493780727599086e-05,
      "loss": 2.592789649963379,
      "memory(GiB)": 77.56,
      "step": 38955,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.439693
    },
    {
      "epoch": 1.6691658455079046,
      "grad_norm": 5.9246673583984375,
      "learning_rate": 7.493197407079389e-05,
      "loss": 2.5767799377441407,
      "memory(GiB)": 77.56,
      "step": 38960,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.439669
    },
    {
      "epoch": 1.6693800608371534,
      "grad_norm": 5.711269378662109,
      "learning_rate": 7.492614041393079e-05,
      "loss": 2.86291561126709,
      "memory(GiB)": 77.56,
      "step": 38965,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.439595
    },
    {
      "epoch": 1.6695942761664024,
      "grad_norm": 8.412878036499023,
      "learning_rate": 7.492030630550728e-05,
      "loss": 2.722490119934082,
      "memory(GiB)": 77.56,
      "step": 38970,
      "token_acc": 0.45980707395498394,
      "train_speed(iter/s)": 1.439604
    },
    {
      "epoch": 1.6698084914956515,
      "grad_norm": 5.96693754196167,
      "learning_rate": 7.491447174562906e-05,
      "loss": 2.6551231384277343,
      "memory(GiB)": 77.56,
      "step": 38975,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.439614
    },
    {
      "epoch": 1.6700227068249003,
      "grad_norm": 4.954509735107422,
      "learning_rate": 7.49086367344018e-05,
      "loss": 2.6989574432373047,
      "memory(GiB)": 77.56,
      "step": 38980,
      "token_acc": 0.4653846153846154,
      "train_speed(iter/s)": 1.439582
    },
    {
      "epoch": 1.6702369221541493,
      "grad_norm": 5.199118614196777,
      "learning_rate": 7.490280127193122e-05,
      "loss": 2.872369575500488,
      "memory(GiB)": 77.56,
      "step": 38985,
      "token_acc": 0.4310850439882698,
      "train_speed(iter/s)": 1.439565
    },
    {
      "epoch": 1.6704511374833984,
      "grad_norm": 3.8215460777282715,
      "learning_rate": 7.489696535832305e-05,
      "loss": 2.401051902770996,
      "memory(GiB)": 77.56,
      "step": 38990,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.439551
    },
    {
      "epoch": 1.6706653528126472,
      "grad_norm": 3.881770372390747,
      "learning_rate": 7.489112899368298e-05,
      "loss": 2.625141906738281,
      "memory(GiB)": 77.56,
      "step": 38995,
      "token_acc": 0.45592705167173253,
      "train_speed(iter/s)": 1.439566
    },
    {
      "epoch": 1.6708795681418962,
      "grad_norm": 3.871721029281616,
      "learning_rate": 7.488529217811676e-05,
      "loss": 2.5486787796020507,
      "memory(GiB)": 77.56,
      "step": 39000,
      "token_acc": 0.44545454545454544,
      "train_speed(iter/s)": 1.439577
    },
    {
      "epoch": 1.6708795681418962,
      "eval_loss": 2.2289645671844482,
      "eval_runtime": 14.8591,
      "eval_samples_per_second": 6.73,
      "eval_steps_per_second": 6.73,
      "eval_token_acc": 0.47736093143596375,
      "step": 39000
    },
    {
      "epoch": 1.6710937834711452,
      "grad_norm": 4.947300910949707,
      "learning_rate": 7.487945491173012e-05,
      "loss": 2.4421831130981446,
      "memory(GiB)": 77.56,
      "step": 39005,
      "token_acc": 0.4818702290076336,
      "train_speed(iter/s)": 1.438725
    },
    {
      "epoch": 1.671307998800394,
      "grad_norm": 4.8534393310546875,
      "learning_rate": 7.487361719462883e-05,
      "loss": 2.932672882080078,
      "memory(GiB)": 77.56,
      "step": 39010,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.43875
    },
    {
      "epoch": 1.671522214129643,
      "grad_norm": 5.711973190307617,
      "learning_rate": 7.486777902691864e-05,
      "loss": 2.5695960998535154,
      "memory(GiB)": 77.56,
      "step": 39015,
      "token_acc": 0.46932515337423314,
      "train_speed(iter/s)": 1.438754
    },
    {
      "epoch": 1.6717364294588921,
      "grad_norm": 4.816310405731201,
      "learning_rate": 7.48619404087053e-05,
      "loss": 2.316717338562012,
      "memory(GiB)": 77.56,
      "step": 39020,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.43876
    },
    {
      "epoch": 1.671950644788141,
      "grad_norm": 4.691830635070801,
      "learning_rate": 7.485610134009458e-05,
      "loss": 2.816037178039551,
      "memory(GiB)": 77.56,
      "step": 39025,
      "token_acc": 0.4620938628158845,
      "train_speed(iter/s)": 1.438812
    },
    {
      "epoch": 1.67216486011739,
      "grad_norm": 5.378674507141113,
      "learning_rate": 7.485026182119225e-05,
      "loss": 2.6296865463256838,
      "memory(GiB)": 77.56,
      "step": 39030,
      "token_acc": 0.45857988165680474,
      "train_speed(iter/s)": 1.438815
    },
    {
      "epoch": 1.672379075446639,
      "grad_norm": 4.636214256286621,
      "learning_rate": 7.484442185210414e-05,
      "loss": 2.6497900009155275,
      "memory(GiB)": 77.56,
      "step": 39035,
      "token_acc": 0.42244224422442245,
      "train_speed(iter/s)": 1.438835
    },
    {
      "epoch": 1.6725932907758878,
      "grad_norm": 4.445847511291504,
      "learning_rate": 7.483858143293602e-05,
      "loss": 2.561076354980469,
      "memory(GiB)": 77.56,
      "step": 39040,
      "token_acc": 0.479108635097493,
      "train_speed(iter/s)": 1.438861
    },
    {
      "epoch": 1.6728075061051368,
      "grad_norm": 5.082719326019287,
      "learning_rate": 7.483274056379368e-05,
      "loss": 2.3797388076782227,
      "memory(GiB)": 77.56,
      "step": 39045,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.438885
    },
    {
      "epoch": 1.6730217214343859,
      "grad_norm": 7.136971473693848,
      "learning_rate": 7.482689924478297e-05,
      "loss": 2.4964004516601563,
      "memory(GiB)": 77.56,
      "step": 39050,
      "token_acc": 0.48014440433212996,
      "train_speed(iter/s)": 1.438901
    },
    {
      "epoch": 1.6732359367636347,
      "grad_norm": 5.671760559082031,
      "learning_rate": 7.482105747600968e-05,
      "loss": 2.2216861724853514,
      "memory(GiB)": 77.56,
      "step": 39055,
      "token_acc": 0.49624060150375937,
      "train_speed(iter/s)": 1.438887
    },
    {
      "epoch": 1.6734501520928837,
      "grad_norm": 5.063310146331787,
      "learning_rate": 7.481521525757966e-05,
      "loss": 2.498670768737793,
      "memory(GiB)": 77.56,
      "step": 39060,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.438897
    },
    {
      "epoch": 1.6736643674221328,
      "grad_norm": 4.345708847045898,
      "learning_rate": 7.480937258959872e-05,
      "loss": 2.5834796905517576,
      "memory(GiB)": 77.56,
      "step": 39065,
      "token_acc": 0.4495677233429395,
      "train_speed(iter/s)": 1.438895
    },
    {
      "epoch": 1.6738785827513816,
      "grad_norm": 4.285645484924316,
      "learning_rate": 7.480352947217274e-05,
      "loss": 2.7373275756835938,
      "memory(GiB)": 77.56,
      "step": 39070,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.438887
    },
    {
      "epoch": 1.6740927980806306,
      "grad_norm": 4.04364538192749,
      "learning_rate": 7.479768590540755e-05,
      "loss": 2.2201259613037108,
      "memory(GiB)": 77.56,
      "step": 39075,
      "token_acc": 0.5423728813559322,
      "train_speed(iter/s)": 1.438871
    },
    {
      "epoch": 1.6743070134098796,
      "grad_norm": 5.461197376251221,
      "learning_rate": 7.479184188940901e-05,
      "loss": 2.398779106140137,
      "memory(GiB)": 77.56,
      "step": 39080,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.438845
    },
    {
      "epoch": 1.6745212287391285,
      "grad_norm": 5.962795734405518,
      "learning_rate": 7.4785997424283e-05,
      "loss": 2.826894760131836,
      "memory(GiB)": 77.56,
      "step": 39085,
      "token_acc": 0.43312101910828027,
      "train_speed(iter/s)": 1.438841
    },
    {
      "epoch": 1.6747354440683775,
      "grad_norm": 5.444334983825684,
      "learning_rate": 7.47801525101354e-05,
      "loss": 2.699898910522461,
      "memory(GiB)": 77.56,
      "step": 39090,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.438875
    },
    {
      "epoch": 1.6749496593976265,
      "grad_norm": 5.573398590087891,
      "learning_rate": 7.477430714707208e-05,
      "loss": 2.547859764099121,
      "memory(GiB)": 77.56,
      "step": 39095,
      "token_acc": 0.4899598393574297,
      "train_speed(iter/s)": 1.43886
    },
    {
      "epoch": 1.6751638747268753,
      "grad_norm": 6.221633434295654,
      "learning_rate": 7.476846133519896e-05,
      "loss": 2.221527671813965,
      "memory(GiB)": 77.56,
      "step": 39100,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.438907
    },
    {
      "epoch": 1.6753780900561244,
      "grad_norm": 5.120943546295166,
      "learning_rate": 7.476261507462194e-05,
      "loss": 2.5731740951538087,
      "memory(GiB)": 77.56,
      "step": 39105,
      "token_acc": 0.4377224199288256,
      "train_speed(iter/s)": 1.438932
    },
    {
      "epoch": 1.6755923053853734,
      "grad_norm": 4.545486927032471,
      "learning_rate": 7.47567683654469e-05,
      "loss": 2.459206390380859,
      "memory(GiB)": 77.56,
      "step": 39110,
      "token_acc": 0.5016949152542373,
      "train_speed(iter/s)": 1.43892
    },
    {
      "epoch": 1.6758065207146222,
      "grad_norm": 4.305181503295898,
      "learning_rate": 7.475092120777978e-05,
      "loss": 2.2530704498291017,
      "memory(GiB)": 77.56,
      "step": 39115,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438917
    },
    {
      "epoch": 1.6760207360438713,
      "grad_norm": 5.2314910888671875,
      "learning_rate": 7.47450736017265e-05,
      "loss": 2.3604949951171874,
      "memory(GiB)": 77.56,
      "step": 39120,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.438889
    },
    {
      "epoch": 1.6762349513731203,
      "grad_norm": 5.725355625152588,
      "learning_rate": 7.4739225547393e-05,
      "loss": 2.1574092864990235,
      "memory(GiB)": 77.56,
      "step": 39125,
      "token_acc": 0.5255474452554745,
      "train_speed(iter/s)": 1.4389
    },
    {
      "epoch": 1.676449166702369,
      "grad_norm": 6.635068416595459,
      "learning_rate": 7.473337704488523e-05,
      "loss": 2.669292449951172,
      "memory(GiB)": 77.56,
      "step": 39130,
      "token_acc": 0.47696476964769646,
      "train_speed(iter/s)": 1.438859
    },
    {
      "epoch": 1.6766633820316181,
      "grad_norm": 5.123763561248779,
      "learning_rate": 7.472752809430913e-05,
      "loss": 2.5801183700561525,
      "memory(GiB)": 77.56,
      "step": 39135,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.438833
    },
    {
      "epoch": 1.6768775973608672,
      "grad_norm": 5.092013835906982,
      "learning_rate": 7.472167869577066e-05,
      "loss": 2.5641399383544923,
      "memory(GiB)": 77.56,
      "step": 39140,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.43881
    },
    {
      "epoch": 1.677091812690116,
      "grad_norm": 5.053287506103516,
      "learning_rate": 7.471582884937579e-05,
      "loss": 2.685615348815918,
      "memory(GiB)": 77.56,
      "step": 39145,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.438857
    },
    {
      "epoch": 1.677306028019365,
      "grad_norm": 4.772220611572266,
      "learning_rate": 7.470997855523049e-05,
      "loss": 2.6306638717651367,
      "memory(GiB)": 77.56,
      "step": 39150,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.438861
    },
    {
      "epoch": 1.677520243348614,
      "grad_norm": 6.864222049713135,
      "learning_rate": 7.470412781344075e-05,
      "loss": 3.0099872589111327,
      "memory(GiB)": 77.56,
      "step": 39155,
      "token_acc": 0.4139072847682119,
      "train_speed(iter/s)": 1.438888
    },
    {
      "epoch": 1.6777344586778629,
      "grad_norm": 4.832400321960449,
      "learning_rate": 7.469827662411257e-05,
      "loss": 2.3876165390014648,
      "memory(GiB)": 77.56,
      "step": 39160,
      "token_acc": 0.483271375464684,
      "train_speed(iter/s)": 1.438897
    },
    {
      "epoch": 1.677948674007112,
      "grad_norm": 4.6121745109558105,
      "learning_rate": 7.469242498735193e-05,
      "loss": 2.6444320678710938,
      "memory(GiB)": 77.56,
      "step": 39165,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.43893
    },
    {
      "epoch": 1.678162889336361,
      "grad_norm": 6.902131080627441,
      "learning_rate": 7.468657290326486e-05,
      "loss": 2.5464151382446287,
      "memory(GiB)": 77.56,
      "step": 39170,
      "token_acc": 0.47115384615384615,
      "train_speed(iter/s)": 1.438877
    },
    {
      "epoch": 1.6783771046656097,
      "grad_norm": 5.684139728546143,
      "learning_rate": 7.468072037195735e-05,
      "loss": 2.4755186080932616,
      "memory(GiB)": 77.56,
      "step": 39175,
      "token_acc": 0.48606811145510836,
      "train_speed(iter/s)": 1.438901
    },
    {
      "epoch": 1.6785913199948588,
      "grad_norm": 6.832084655761719,
      "learning_rate": 7.467486739353545e-05,
      "loss": 2.6017223358154298,
      "memory(GiB)": 77.56,
      "step": 39180,
      "token_acc": 0.4429065743944637,
      "train_speed(iter/s)": 1.438948
    },
    {
      "epoch": 1.6788055353241078,
      "grad_norm": 5.121083736419678,
      "learning_rate": 7.466901396810517e-05,
      "loss": 2.4421270370483397,
      "memory(GiB)": 77.56,
      "step": 39185,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.438935
    },
    {
      "epoch": 1.6790197506533566,
      "grad_norm": 4.9516119956970215,
      "learning_rate": 7.466316009577258e-05,
      "loss": 2.643556594848633,
      "memory(GiB)": 77.56,
      "step": 39190,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.438908
    },
    {
      "epoch": 1.6792339659826059,
      "grad_norm": 7.260333061218262,
      "learning_rate": 7.465730577664368e-05,
      "loss": 2.1878664016723635,
      "memory(GiB)": 77.56,
      "step": 39195,
      "token_acc": 0.5608856088560885,
      "train_speed(iter/s)": 1.438898
    },
    {
      "epoch": 1.6794481813118547,
      "grad_norm": 4.869746685028076,
      "learning_rate": 7.465145101082458e-05,
      "loss": 2.7983097076416015,
      "memory(GiB)": 77.56,
      "step": 39200,
      "token_acc": 0.42955326460481097,
      "train_speed(iter/s)": 1.438926
    },
    {
      "epoch": 1.6796623966411035,
      "grad_norm": 6.569705963134766,
      "learning_rate": 7.464559579842132e-05,
      "loss": 2.4519372940063477,
      "memory(GiB)": 77.56,
      "step": 39205,
      "token_acc": 0.4669260700389105,
      "train_speed(iter/s)": 1.438975
    },
    {
      "epoch": 1.6798766119703528,
      "grad_norm": 4.4712419509887695,
      "learning_rate": 7.463974013953995e-05,
      "loss": 2.294918441772461,
      "memory(GiB)": 77.56,
      "step": 39210,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.438927
    },
    {
      "epoch": 1.6800908272996016,
      "grad_norm": 4.814571380615234,
      "learning_rate": 7.463388403428659e-05,
      "loss": 2.29299259185791,
      "memory(GiB)": 77.56,
      "step": 39215,
      "token_acc": 0.4866920152091255,
      "train_speed(iter/s)": 1.438916
    },
    {
      "epoch": 1.6803050426288504,
      "grad_norm": 4.6567606925964355,
      "learning_rate": 7.46280274827673e-05,
      "loss": 2.447309875488281,
      "memory(GiB)": 77.56,
      "step": 39220,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.438955
    },
    {
      "epoch": 1.6805192579580996,
      "grad_norm": 3.4669249057769775,
      "learning_rate": 7.46221704850882e-05,
      "loss": 2.331422805786133,
      "memory(GiB)": 77.56,
      "step": 39225,
      "token_acc": 0.46439628482972134,
      "train_speed(iter/s)": 1.438946
    },
    {
      "epoch": 1.6807334732873485,
      "grad_norm": 5.413876056671143,
      "learning_rate": 7.461631304135538e-05,
      "loss": 2.4734737396240236,
      "memory(GiB)": 77.56,
      "step": 39230,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.438957
    },
    {
      "epoch": 1.6809476886165973,
      "grad_norm": 4.175632953643799,
      "learning_rate": 7.461045515167497e-05,
      "loss": 2.620469856262207,
      "memory(GiB)": 77.56,
      "step": 39235,
      "token_acc": 0.4508670520231214,
      "train_speed(iter/s)": 1.438989
    },
    {
      "epoch": 1.6811619039458465,
      "grad_norm": 5.209411144256592,
      "learning_rate": 7.460459681615305e-05,
      "loss": 2.537337875366211,
      "memory(GiB)": 77.56,
      "step": 39240,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.438999
    },
    {
      "epoch": 1.6813761192750953,
      "grad_norm": 6.1368327140808105,
      "learning_rate": 7.45987380348958e-05,
      "loss": 2.5748977661132812,
      "memory(GiB)": 77.56,
      "step": 39245,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.438999
    },
    {
      "epoch": 1.6815903346043442,
      "grad_norm": 5.22243070602417,
      "learning_rate": 7.459287880800933e-05,
      "loss": 2.320934295654297,
      "memory(GiB)": 77.56,
      "step": 39250,
      "token_acc": 0.4855072463768116,
      "train_speed(iter/s)": 1.438998
    },
    {
      "epoch": 1.6818045499335934,
      "grad_norm": 6.507482051849365,
      "learning_rate": 7.458701913559978e-05,
      "loss": 2.7442821502685546,
      "memory(GiB)": 77.56,
      "step": 39255,
      "token_acc": 0.4498567335243553,
      "train_speed(iter/s)": 1.439028
    },
    {
      "epoch": 1.6820187652628422,
      "grad_norm": 5.577680587768555,
      "learning_rate": 7.458115901777334e-05,
      "loss": 2.246876907348633,
      "memory(GiB)": 77.56,
      "step": 39260,
      "token_acc": 0.5259259259259259,
      "train_speed(iter/s)": 1.43905
    },
    {
      "epoch": 1.682232980592091,
      "grad_norm": 5.186855792999268,
      "learning_rate": 7.457529845463612e-05,
      "loss": 2.264640045166016,
      "memory(GiB)": 77.56,
      "step": 39265,
      "token_acc": 0.4876325088339223,
      "train_speed(iter/s)": 1.439095
    },
    {
      "epoch": 1.6824471959213403,
      "grad_norm": 8.413525581359863,
      "learning_rate": 7.456943744629433e-05,
      "loss": 2.429824447631836,
      "memory(GiB)": 77.56,
      "step": 39270,
      "token_acc": 0.527972027972028,
      "train_speed(iter/s)": 1.439138
    },
    {
      "epoch": 1.682661411250589,
      "grad_norm": 4.4329142570495605,
      "learning_rate": 7.456357599285413e-05,
      "loss": 2.4337596893310547,
      "memory(GiB)": 77.56,
      "step": 39275,
      "token_acc": 0.515527950310559,
      "train_speed(iter/s)": 1.439193
    },
    {
      "epoch": 1.682875626579838,
      "grad_norm": 5.147960186004639,
      "learning_rate": 7.455771409442171e-05,
      "loss": 2.7660747528076173,
      "memory(GiB)": 77.56,
      "step": 39280,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.439236
    },
    {
      "epoch": 1.6830898419090872,
      "grad_norm": 4.473089218139648,
      "learning_rate": 7.455185175110325e-05,
      "loss": 2.6357538223266603,
      "memory(GiB)": 77.56,
      "step": 39285,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.439216
    },
    {
      "epoch": 1.683304057238336,
      "grad_norm": 4.885560035705566,
      "learning_rate": 7.454598896300498e-05,
      "loss": 2.568756103515625,
      "memory(GiB)": 77.56,
      "step": 39290,
      "token_acc": 0.43613707165109034,
      "train_speed(iter/s)": 1.439266
    },
    {
      "epoch": 1.6835182725675848,
      "grad_norm": 4.351996421813965,
      "learning_rate": 7.454012573023308e-05,
      "loss": 2.9108325958251955,
      "memory(GiB)": 77.56,
      "step": 39295,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.439277
    },
    {
      "epoch": 1.683732487896834,
      "grad_norm": 4.397448539733887,
      "learning_rate": 7.453426205289379e-05,
      "loss": 2.2326358795166015,
      "memory(GiB)": 77.56,
      "step": 39300,
      "token_acc": 0.5326460481099656,
      "train_speed(iter/s)": 1.439266
    },
    {
      "epoch": 1.6839467032260829,
      "grad_norm": 4.59770393371582,
      "learning_rate": 7.452839793109332e-05,
      "loss": 2.5102912902832033,
      "memory(GiB)": 77.56,
      "step": 39305,
      "token_acc": 0.47648902821316613,
      "train_speed(iter/s)": 1.439299
    },
    {
      "epoch": 1.6841609185553317,
      "grad_norm": 5.023507118225098,
      "learning_rate": 7.452253336493791e-05,
      "loss": 2.489552688598633,
      "memory(GiB)": 77.56,
      "step": 39310,
      "token_acc": 0.45738636363636365,
      "train_speed(iter/s)": 1.43933
    },
    {
      "epoch": 1.684375133884581,
      "grad_norm": 4.828036785125732,
      "learning_rate": 7.451666835453382e-05,
      "loss": 2.7237260818481444,
      "memory(GiB)": 77.56,
      "step": 39315,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.439292
    },
    {
      "epoch": 1.6845893492138297,
      "grad_norm": 4.740131378173828,
      "learning_rate": 7.451080289998729e-05,
      "loss": 2.5221656799316405,
      "memory(GiB)": 77.56,
      "step": 39320,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.439307
    },
    {
      "epoch": 1.6848035645430786,
      "grad_norm": 5.324163436889648,
      "learning_rate": 7.450493700140454e-05,
      "loss": 2.758205604553223,
      "memory(GiB)": 77.56,
      "step": 39325,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.439352
    },
    {
      "epoch": 1.6850177798723278,
      "grad_norm": 5.364895343780518,
      "learning_rate": 7.44990706588919e-05,
      "loss": 2.2419387817382814,
      "memory(GiB)": 77.56,
      "step": 39330,
      "token_acc": 0.5043859649122807,
      "train_speed(iter/s)": 1.439385
    },
    {
      "epoch": 1.6852319952015766,
      "grad_norm": 4.1193060874938965,
      "learning_rate": 7.44932038725556e-05,
      "loss": 2.604151153564453,
      "memory(GiB)": 77.56,
      "step": 39335,
      "token_acc": 0.46264367816091956,
      "train_speed(iter/s)": 1.43941
    },
    {
      "epoch": 1.6854462105308254,
      "grad_norm": 5.7761640548706055,
      "learning_rate": 7.448733664250191e-05,
      "loss": 2.223978805541992,
      "memory(GiB)": 77.56,
      "step": 39340,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.439443
    },
    {
      "epoch": 1.6856604258600747,
      "grad_norm": 5.1349287033081055,
      "learning_rate": 7.448146896883719e-05,
      "loss": 2.589311218261719,
      "memory(GiB)": 77.56,
      "step": 39345,
      "token_acc": 0.4401294498381877,
      "train_speed(iter/s)": 1.439478
    },
    {
      "epoch": 1.6858746411893235,
      "grad_norm": 4.3070244789123535,
      "learning_rate": 7.447560085166768e-05,
      "loss": 2.571135711669922,
      "memory(GiB)": 77.56,
      "step": 39350,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439477
    },
    {
      "epoch": 1.6860888565185723,
      "grad_norm": 5.659824848175049,
      "learning_rate": 7.446973229109969e-05,
      "loss": 2.485415458679199,
      "memory(GiB)": 77.56,
      "step": 39355,
      "token_acc": 0.5041322314049587,
      "train_speed(iter/s)": 1.439532
    },
    {
      "epoch": 1.6863030718478216,
      "grad_norm": 5.9419708251953125,
      "learning_rate": 7.446386328723955e-05,
      "loss": 2.428858757019043,
      "memory(GiB)": 77.56,
      "step": 39360,
      "token_acc": 0.48220064724919093,
      "train_speed(iter/s)": 1.439542
    },
    {
      "epoch": 1.6865172871770704,
      "grad_norm": 5.576047897338867,
      "learning_rate": 7.445799384019358e-05,
      "loss": 2.317838668823242,
      "memory(GiB)": 77.56,
      "step": 39365,
      "token_acc": 0.5228070175438596,
      "train_speed(iter/s)": 1.439578
    },
    {
      "epoch": 1.6867315025063192,
      "grad_norm": 6.139689922332764,
      "learning_rate": 7.44521239500681e-05,
      "loss": 2.2850807189941404,
      "memory(GiB)": 77.56,
      "step": 39370,
      "token_acc": 0.5104602510460251,
      "train_speed(iter/s)": 1.439626
    },
    {
      "epoch": 1.6869457178355685,
      "grad_norm": 5.615159511566162,
      "learning_rate": 7.444625361696948e-05,
      "loss": 2.4922197341918944,
      "memory(GiB)": 77.56,
      "step": 39375,
      "token_acc": 0.4858156028368794,
      "train_speed(iter/s)": 1.439616
    },
    {
      "epoch": 1.6871599331648173,
      "grad_norm": 5.553599834442139,
      "learning_rate": 7.444038284100401e-05,
      "loss": 2.4360607147216795,
      "memory(GiB)": 77.56,
      "step": 39380,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.439638
    },
    {
      "epoch": 1.687374148494066,
      "grad_norm": 5.104459762573242,
      "learning_rate": 7.44345116222781e-05,
      "loss": 2.6728410720825195,
      "memory(GiB)": 77.56,
      "step": 39385,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.43966
    },
    {
      "epoch": 1.6875883638233153,
      "grad_norm": 5.416928768157959,
      "learning_rate": 7.442863996089809e-05,
      "loss": 2.7254703521728514,
      "memory(GiB)": 77.56,
      "step": 39390,
      "token_acc": 0.48823529411764705,
      "train_speed(iter/s)": 1.439696
    },
    {
      "epoch": 1.6878025791525642,
      "grad_norm": 6.580688953399658,
      "learning_rate": 7.442276785697035e-05,
      "loss": 2.766031837463379,
      "memory(GiB)": 77.56,
      "step": 39395,
      "token_acc": 0.4520547945205479,
      "train_speed(iter/s)": 1.43972
    },
    {
      "epoch": 1.688016794481813,
      "grad_norm": 4.260105609893799,
      "learning_rate": 7.441689531060125e-05,
      "loss": 2.2400495529174806,
      "memory(GiB)": 77.56,
      "step": 39400,
      "token_acc": 0.4895397489539749,
      "train_speed(iter/s)": 1.439733
    },
    {
      "epoch": 1.6882310098110622,
      "grad_norm": 5.144050121307373,
      "learning_rate": 7.441102232189721e-05,
      "loss": 2.6564083099365234,
      "memory(GiB)": 77.56,
      "step": 39405,
      "token_acc": 0.4309859154929577,
      "train_speed(iter/s)": 1.439759
    },
    {
      "epoch": 1.688445225140311,
      "grad_norm": 4.711074352264404,
      "learning_rate": 7.440514889096457e-05,
      "loss": 2.779418182373047,
      "memory(GiB)": 77.56,
      "step": 39410,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.439794
    },
    {
      "epoch": 1.6886594404695598,
      "grad_norm": 4.433008193969727,
      "learning_rate": 7.439927501790978e-05,
      "loss": 2.5604726791381838,
      "memory(GiB)": 77.56,
      "step": 39415,
      "token_acc": 0.49783549783549785,
      "train_speed(iter/s)": 1.439789
    },
    {
      "epoch": 1.688873655798809,
      "grad_norm": 4.9867095947265625,
      "learning_rate": 7.439340070283923e-05,
      "loss": 2.6040271759033202,
      "memory(GiB)": 77.56,
      "step": 39420,
      "token_acc": 0.4290322580645161,
      "train_speed(iter/s)": 1.439812
    },
    {
      "epoch": 1.689087871128058,
      "grad_norm": 5.791916847229004,
      "learning_rate": 7.438752594585935e-05,
      "loss": 2.4432435989379884,
      "memory(GiB)": 77.56,
      "step": 39425,
      "token_acc": 0.4555984555984556,
      "train_speed(iter/s)": 1.439826
    },
    {
      "epoch": 1.6893020864573067,
      "grad_norm": 4.592531204223633,
      "learning_rate": 7.438165074707655e-05,
      "loss": 2.6076065063476563,
      "memory(GiB)": 77.56,
      "step": 39430,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.439866
    },
    {
      "epoch": 1.689516301786556,
      "grad_norm": 5.351303577423096,
      "learning_rate": 7.437577510659729e-05,
      "loss": 2.603732872009277,
      "memory(GiB)": 77.56,
      "step": 39435,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.439861
    },
    {
      "epoch": 1.6897305171158048,
      "grad_norm": 8.990999221801758,
      "learning_rate": 7.436989902452798e-05,
      "loss": 2.7704092025756837,
      "memory(GiB)": 77.56,
      "step": 39440,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.439894
    },
    {
      "epoch": 1.6899447324450536,
      "grad_norm": 4.653515815734863,
      "learning_rate": 7.43640225009751e-05,
      "loss": 2.4889642715454103,
      "memory(GiB)": 77.56,
      "step": 39445,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.439928
    },
    {
      "epoch": 1.6901589477743029,
      "grad_norm": 6.005542278289795,
      "learning_rate": 7.43581455360451e-05,
      "loss": 2.432036209106445,
      "memory(GiB)": 77.56,
      "step": 39450,
      "token_acc": 0.45980707395498394,
      "train_speed(iter/s)": 1.439932
    },
    {
      "epoch": 1.6903731631035517,
      "grad_norm": 4.6243414878845215,
      "learning_rate": 7.435226812984443e-05,
      "loss": 2.7409540176391602,
      "memory(GiB)": 77.56,
      "step": 39455,
      "token_acc": 0.4311377245508982,
      "train_speed(iter/s)": 1.439952
    },
    {
      "epoch": 1.6905873784328005,
      "grad_norm": 4.375554084777832,
      "learning_rate": 7.434639028247959e-05,
      "loss": 2.603215217590332,
      "memory(GiB)": 77.56,
      "step": 39460,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.439958
    },
    {
      "epoch": 1.6908015937620497,
      "grad_norm": 4.447576522827148,
      "learning_rate": 7.434051199405705e-05,
      "loss": 2.4326690673828124,
      "memory(GiB)": 77.56,
      "step": 39465,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.439984
    },
    {
      "epoch": 1.6910158090912986,
      "grad_norm": 7.526741981506348,
      "learning_rate": 7.43346332646833e-05,
      "loss": 2.82733211517334,
      "memory(GiB)": 77.56,
      "step": 39470,
      "token_acc": 0.44482758620689655,
      "train_speed(iter/s)": 1.440025
    },
    {
      "epoch": 1.6912300244205474,
      "grad_norm": 6.270843982696533,
      "learning_rate": 7.432875409446483e-05,
      "loss": 2.868579864501953,
      "memory(GiB)": 77.56,
      "step": 39475,
      "token_acc": 0.3835125448028674,
      "train_speed(iter/s)": 1.440028
    },
    {
      "epoch": 1.6914442397497966,
      "grad_norm": 4.88505220413208,
      "learning_rate": 7.432287448350819e-05,
      "loss": 2.5317960739135743,
      "memory(GiB)": 77.56,
      "step": 39480,
      "token_acc": 0.46956521739130436,
      "train_speed(iter/s)": 1.440041
    },
    {
      "epoch": 1.6916584550790454,
      "grad_norm": 5.6399617195129395,
      "learning_rate": 7.431699443191982e-05,
      "loss": 2.3505857467651365,
      "memory(GiB)": 77.56,
      "step": 39485,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.440067
    },
    {
      "epoch": 1.6918726704082943,
      "grad_norm": 4.655662536621094,
      "learning_rate": 7.431111393980631e-05,
      "loss": 2.5938844680786133,
      "memory(GiB)": 77.56,
      "step": 39490,
      "token_acc": 0.5020080321285141,
      "train_speed(iter/s)": 1.440058
    },
    {
      "epoch": 1.6920868857375435,
      "grad_norm": 4.296602249145508,
      "learning_rate": 7.430523300727416e-05,
      "loss": 2.563298225402832,
      "memory(GiB)": 77.56,
      "step": 39495,
      "token_acc": 0.4725609756097561,
      "train_speed(iter/s)": 1.440086
    },
    {
      "epoch": 1.6923011010667923,
      "grad_norm": 5.97561502456665,
      "learning_rate": 7.429935163442991e-05,
      "loss": 2.477747344970703,
      "memory(GiB)": 77.56,
      "step": 39500,
      "token_acc": 0.45255474452554745,
      "train_speed(iter/s)": 1.440135
    },
    {
      "epoch": 1.6923011010667923,
      "eval_loss": 2.2470016479492188,
      "eval_runtime": 14.3308,
      "eval_samples_per_second": 6.978,
      "eval_steps_per_second": 6.978,
      "eval_token_acc": 0.46788990825688076,
      "step": 39500
    },
    {
      "epoch": 1.6925153163960411,
      "grad_norm": 5.674051761627197,
      "learning_rate": 7.429346982138013e-05,
      "loss": 2.5348369598388674,
      "memory(GiB)": 77.56,
      "step": 39505,
      "token_acc": 0.47433460076045625,
      "train_speed(iter/s)": 1.439335
    },
    {
      "epoch": 1.6927295317252904,
      "grad_norm": 4.425677299499512,
      "learning_rate": 7.428758756823134e-05,
      "loss": 2.1379732131958007,
      "memory(GiB)": 77.56,
      "step": 39510,
      "token_acc": 0.5115511551155115,
      "train_speed(iter/s)": 1.439319
    },
    {
      "epoch": 1.6929437470545392,
      "grad_norm": 5.7284464836120605,
      "learning_rate": 7.42817048750901e-05,
      "loss": 2.4620323181152344,
      "memory(GiB)": 77.56,
      "step": 39515,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.439248
    },
    {
      "epoch": 1.693157962383788,
      "grad_norm": 4.829771041870117,
      "learning_rate": 7.427582174206303e-05,
      "loss": 2.7509206771850585,
      "memory(GiB)": 77.56,
      "step": 39520,
      "token_acc": 0.44256756756756754,
      "train_speed(iter/s)": 1.439248
    },
    {
      "epoch": 1.6933721777130373,
      "grad_norm": 4.4447503089904785,
      "learning_rate": 7.426993816925665e-05,
      "loss": 2.5167118072509767,
      "memory(GiB)": 77.56,
      "step": 39525,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.439279
    },
    {
      "epoch": 1.693586393042286,
      "grad_norm": 4.798361778259277,
      "learning_rate": 7.426405415677758e-05,
      "loss": 2.266404151916504,
      "memory(GiB)": 77.56,
      "step": 39530,
      "token_acc": 0.5246478873239436,
      "train_speed(iter/s)": 1.439294
    },
    {
      "epoch": 1.6938006083715351,
      "grad_norm": 3.9646053314208984,
      "learning_rate": 7.425816970473241e-05,
      "loss": 2.3794809341430665,
      "memory(GiB)": 77.56,
      "step": 39535,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.439347
    },
    {
      "epoch": 1.6940148237007842,
      "grad_norm": 4.556328773498535,
      "learning_rate": 7.425228481322775e-05,
      "loss": 2.628735160827637,
      "memory(GiB)": 77.56,
      "step": 39540,
      "token_acc": 0.4690265486725664,
      "train_speed(iter/s)": 1.439364
    },
    {
      "epoch": 1.694229039030033,
      "grad_norm": 4.042719841003418,
      "learning_rate": 7.424639948237019e-05,
      "loss": 2.338087463378906,
      "memory(GiB)": 77.56,
      "step": 39545,
      "token_acc": 0.49415204678362573,
      "train_speed(iter/s)": 1.439369
    },
    {
      "epoch": 1.694443254359282,
      "grad_norm": 4.656288146972656,
      "learning_rate": 7.424051371226636e-05,
      "loss": 2.8349714279174805,
      "memory(GiB)": 77.56,
      "step": 39550,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.439385
    },
    {
      "epoch": 1.694657469688531,
      "grad_norm": 4.094668388366699,
      "learning_rate": 7.423462750302289e-05,
      "loss": 2.197116470336914,
      "memory(GiB)": 77.56,
      "step": 39555,
      "token_acc": 0.5265151515151515,
      "train_speed(iter/s)": 1.439429
    },
    {
      "epoch": 1.6948716850177798,
      "grad_norm": 5.453276634216309,
      "learning_rate": 7.42287408547464e-05,
      "loss": 2.4897510528564455,
      "memory(GiB)": 77.56,
      "step": 39560,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.439489
    },
    {
      "epoch": 1.6950859003470289,
      "grad_norm": 4.060357570648193,
      "learning_rate": 7.422285376754354e-05,
      "loss": 2.645904541015625,
      "memory(GiB)": 77.56,
      "step": 39565,
      "token_acc": 0.4555160142348754,
      "train_speed(iter/s)": 1.439545
    },
    {
      "epoch": 1.695300115676278,
      "grad_norm": 3.969604730606079,
      "learning_rate": 7.421696624152098e-05,
      "loss": 2.4718791961669924,
      "memory(GiB)": 77.56,
      "step": 39570,
      "token_acc": 0.4935897435897436,
      "train_speed(iter/s)": 1.439606
    },
    {
      "epoch": 1.6955143310055267,
      "grad_norm": 4.247460842132568,
      "learning_rate": 7.421107827678533e-05,
      "loss": 2.7582586288452147,
      "memory(GiB)": 77.56,
      "step": 39575,
      "token_acc": 0.44072948328267475,
      "train_speed(iter/s)": 1.439621
    },
    {
      "epoch": 1.6957285463347758,
      "grad_norm": 5.384565353393555,
      "learning_rate": 7.420518987344331e-05,
      "loss": 2.465864562988281,
      "memory(GiB)": 77.56,
      "step": 39580,
      "token_acc": 0.47639484978540775,
      "train_speed(iter/s)": 1.439645
    },
    {
      "epoch": 1.6959427616640248,
      "grad_norm": 7.9196367263793945,
      "learning_rate": 7.419930103160155e-05,
      "loss": 2.4264804840087892,
      "memory(GiB)": 77.56,
      "step": 39585,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.439655
    },
    {
      "epoch": 1.6961569769932736,
      "grad_norm": 5.056312084197998,
      "learning_rate": 7.419341175136677e-05,
      "loss": 2.530421829223633,
      "memory(GiB)": 77.56,
      "step": 39590,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.439663
    },
    {
      "epoch": 1.6963711923225226,
      "grad_norm": 5.398013114929199,
      "learning_rate": 7.418752203284563e-05,
      "loss": 2.5332481384277346,
      "memory(GiB)": 77.56,
      "step": 39595,
      "token_acc": 0.5137254901960784,
      "train_speed(iter/s)": 1.439694
    },
    {
      "epoch": 1.6965854076517717,
      "grad_norm": 4.751768112182617,
      "learning_rate": 7.418163187614485e-05,
      "loss": 2.705938148498535,
      "memory(GiB)": 77.56,
      "step": 39600,
      "token_acc": 0.4028776978417266,
      "train_speed(iter/s)": 1.439715
    },
    {
      "epoch": 1.6967996229810205,
      "grad_norm": 3.972363233566284,
      "learning_rate": 7.41757412813711e-05,
      "loss": 2.5165117263793944,
      "memory(GiB)": 77.56,
      "step": 39605,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.439745
    },
    {
      "epoch": 1.6970138383102695,
      "grad_norm": 4.075873851776123,
      "learning_rate": 7.416985024863115e-05,
      "loss": 2.6509849548339846,
      "memory(GiB)": 77.56,
      "step": 39610,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.439744
    },
    {
      "epoch": 1.6972280536395186,
      "grad_norm": 5.2881760597229,
      "learning_rate": 7.416395877803166e-05,
      "loss": 2.3910858154296877,
      "memory(GiB)": 77.56,
      "step": 39615,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.439757
    },
    {
      "epoch": 1.6974422689687674,
      "grad_norm": 4.936835289001465,
      "learning_rate": 7.415806686967941e-05,
      "loss": 2.311636734008789,
      "memory(GiB)": 77.56,
      "step": 39620,
      "token_acc": 0.5343511450381679,
      "train_speed(iter/s)": 1.439776
    },
    {
      "epoch": 1.6976564842980164,
      "grad_norm": 4.937396049499512,
      "learning_rate": 7.41521745236811e-05,
      "loss": 2.7444005966186524,
      "memory(GiB)": 77.56,
      "step": 39625,
      "token_acc": 0.4270557029177719,
      "train_speed(iter/s)": 1.439779
    },
    {
      "epoch": 1.6978706996272654,
      "grad_norm": 6.1698503494262695,
      "learning_rate": 7.414628174014351e-05,
      "loss": 3.0836111068725587,
      "memory(GiB)": 77.56,
      "step": 39630,
      "token_acc": 0.4124629080118694,
      "train_speed(iter/s)": 1.439819
    },
    {
      "epoch": 1.6980849149565143,
      "grad_norm": 5.699939250946045,
      "learning_rate": 7.414038851917335e-05,
      "loss": 2.6164871215820313,
      "memory(GiB)": 77.56,
      "step": 39635,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.439798
    },
    {
      "epoch": 1.6982991302857633,
      "grad_norm": 5.500759124755859,
      "learning_rate": 7.413449486087742e-05,
      "loss": 2.416820526123047,
      "memory(GiB)": 77.56,
      "step": 39640,
      "token_acc": 0.5120967741935484,
      "train_speed(iter/s)": 1.439776
    },
    {
      "epoch": 1.6985133456150123,
      "grad_norm": 5.812740325927734,
      "learning_rate": 7.412860076536247e-05,
      "loss": 2.7059539794921874,
      "memory(GiB)": 77.56,
      "step": 39645,
      "token_acc": 0.4204081632653061,
      "train_speed(iter/s)": 1.43978
    },
    {
      "epoch": 1.6987275609442611,
      "grad_norm": 5.377533912658691,
      "learning_rate": 7.41227062327353e-05,
      "loss": 2.600864601135254,
      "memory(GiB)": 77.56,
      "step": 39650,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.439821
    },
    {
      "epoch": 1.6989417762735102,
      "grad_norm": 5.6966657638549805,
      "learning_rate": 7.411681126310268e-05,
      "loss": 3.090333366394043,
      "memory(GiB)": 77.56,
      "step": 39655,
      "token_acc": 0.41368078175895767,
      "train_speed(iter/s)": 1.43977
    },
    {
      "epoch": 1.6991559916027592,
      "grad_norm": 4.18341588973999,
      "learning_rate": 7.411091585657139e-05,
      "loss": 2.3464168548583983,
      "memory(GiB)": 77.56,
      "step": 39660,
      "token_acc": 0.49173553719008267,
      "train_speed(iter/s)": 1.439775
    },
    {
      "epoch": 1.699370206932008,
      "grad_norm": 5.513996601104736,
      "learning_rate": 7.410502001324824e-05,
      "loss": 2.5262365341186523,
      "memory(GiB)": 77.56,
      "step": 39665,
      "token_acc": 0.48534201954397393,
      "train_speed(iter/s)": 1.439782
    },
    {
      "epoch": 1.699584422261257,
      "grad_norm": 4.893082618713379,
      "learning_rate": 7.409912373324003e-05,
      "loss": 2.4036827087402344,
      "memory(GiB)": 77.56,
      "step": 39670,
      "token_acc": 0.5210843373493976,
      "train_speed(iter/s)": 1.439826
    },
    {
      "epoch": 1.699798637590506,
      "grad_norm": 6.081119060516357,
      "learning_rate": 7.409322701665359e-05,
      "loss": 2.4552143096923826,
      "memory(GiB)": 77.56,
      "step": 39675,
      "token_acc": 0.4898785425101215,
      "train_speed(iter/s)": 1.439838
    },
    {
      "epoch": 1.700012852919755,
      "grad_norm": 5.193746089935303,
      "learning_rate": 7.408732986359576e-05,
      "loss": 2.6363288879394533,
      "memory(GiB)": 77.56,
      "step": 39680,
      "token_acc": 0.4517133956386293,
      "train_speed(iter/s)": 1.439823
    },
    {
      "epoch": 1.700227068249004,
      "grad_norm": 5.900773048400879,
      "learning_rate": 7.408143227417334e-05,
      "loss": 2.8062456130981444,
      "memory(GiB)": 77.56,
      "step": 39685,
      "token_acc": 0.45686900958466453,
      "train_speed(iter/s)": 1.439822
    },
    {
      "epoch": 1.700441283578253,
      "grad_norm": 5.177570819854736,
      "learning_rate": 7.407553424849318e-05,
      "loss": 2.4233646392822266,
      "memory(GiB)": 77.56,
      "step": 39690,
      "token_acc": 0.4849624060150376,
      "train_speed(iter/s)": 1.439781
    },
    {
      "epoch": 1.7006554989075018,
      "grad_norm": 6.550286293029785,
      "learning_rate": 7.406963578666214e-05,
      "loss": 2.6196516036987303,
      "memory(GiB)": 77.56,
      "step": 39695,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.439823
    },
    {
      "epoch": 1.7008697142367508,
      "grad_norm": 4.333806991577148,
      "learning_rate": 7.406373688878706e-05,
      "loss": 2.631243133544922,
      "memory(GiB)": 77.56,
      "step": 39700,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.439828
    },
    {
      "epoch": 1.7010839295659999,
      "grad_norm": 5.391918182373047,
      "learning_rate": 7.405783755497482e-05,
      "loss": 2.8438425064086914,
      "memory(GiB)": 77.56,
      "step": 39705,
      "token_acc": 0.40599455040871935,
      "train_speed(iter/s)": 1.439851
    },
    {
      "epoch": 1.7012981448952487,
      "grad_norm": 5.443723201751709,
      "learning_rate": 7.405193778533229e-05,
      "loss": 2.687221145629883,
      "memory(GiB)": 77.56,
      "step": 39710,
      "token_acc": 0.46332046332046334,
      "train_speed(iter/s)": 1.4399
    },
    {
      "epoch": 1.7015123602244977,
      "grad_norm": 6.031450271606445,
      "learning_rate": 7.404603757996634e-05,
      "loss": 2.714067268371582,
      "memory(GiB)": 77.56,
      "step": 39715,
      "token_acc": 0.47129909365558914,
      "train_speed(iter/s)": 1.439878
    },
    {
      "epoch": 1.7017265755537467,
      "grad_norm": 5.732964992523193,
      "learning_rate": 7.404013693898385e-05,
      "loss": 2.3155542373657227,
      "memory(GiB)": 77.56,
      "step": 39720,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.439912
    },
    {
      "epoch": 1.7019407908829955,
      "grad_norm": 5.318905353546143,
      "learning_rate": 7.403423586249174e-05,
      "loss": 2.4612009048461916,
      "memory(GiB)": 77.56,
      "step": 39725,
      "token_acc": 0.5078740157480315,
      "train_speed(iter/s)": 1.43989
    },
    {
      "epoch": 1.7021550062122446,
      "grad_norm": 7.824594020843506,
      "learning_rate": 7.402833435059691e-05,
      "loss": 2.730356216430664,
      "memory(GiB)": 77.56,
      "step": 39730,
      "token_acc": 0.462882096069869,
      "train_speed(iter/s)": 1.439872
    },
    {
      "epoch": 1.7023692215414936,
      "grad_norm": 5.203568458557129,
      "learning_rate": 7.402243240340624e-05,
      "loss": 2.5023773193359373,
      "memory(GiB)": 77.56,
      "step": 39735,
      "token_acc": 0.45016077170418006,
      "train_speed(iter/s)": 1.43989
    },
    {
      "epoch": 1.7025834368707424,
      "grad_norm": 6.08251428604126,
      "learning_rate": 7.401653002102669e-05,
      "loss": 2.444143295288086,
      "memory(GiB)": 77.56,
      "step": 39740,
      "token_acc": 0.4959677419354839,
      "train_speed(iter/s)": 1.439855
    },
    {
      "epoch": 1.7027976521999915,
      "grad_norm": 4.962358474731445,
      "learning_rate": 7.401062720356516e-05,
      "loss": 2.8301889419555666,
      "memory(GiB)": 77.56,
      "step": 39745,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.439861
    },
    {
      "epoch": 1.7030118675292405,
      "grad_norm": 8.405646324157715,
      "learning_rate": 7.400472395112861e-05,
      "loss": 2.5224157333374024,
      "memory(GiB)": 77.56,
      "step": 39750,
      "token_acc": 0.4820846905537459,
      "train_speed(iter/s)": 1.439844
    },
    {
      "epoch": 1.7032260828584893,
      "grad_norm": 5.012209892272949,
      "learning_rate": 7.399882026382395e-05,
      "loss": 2.705532455444336,
      "memory(GiB)": 77.56,
      "step": 39755,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.4398
    },
    {
      "epoch": 1.7034402981877383,
      "grad_norm": 4.887112617492676,
      "learning_rate": 7.399291614175815e-05,
      "loss": 2.334407424926758,
      "memory(GiB)": 77.56,
      "step": 39760,
      "token_acc": 0.5064935064935064,
      "train_speed(iter/s)": 1.439839
    },
    {
      "epoch": 1.7036545135169874,
      "grad_norm": 4.952052593231201,
      "learning_rate": 7.398701158503818e-05,
      "loss": 2.404689407348633,
      "memory(GiB)": 77.56,
      "step": 39765,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.439852
    },
    {
      "epoch": 1.7038687288462362,
      "grad_norm": 4.783664703369141,
      "learning_rate": 7.398110659377098e-05,
      "loss": 2.403533363342285,
      "memory(GiB)": 77.56,
      "step": 39770,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.439838
    },
    {
      "epoch": 1.7040829441754852,
      "grad_norm": 4.58333158493042,
      "learning_rate": 7.397520116806354e-05,
      "loss": 2.5737075805664062,
      "memory(GiB)": 77.56,
      "step": 39775,
      "token_acc": 0.44285714285714284,
      "train_speed(iter/s)": 1.439844
    },
    {
      "epoch": 1.7042971595047343,
      "grad_norm": 6.006550312042236,
      "learning_rate": 7.396929530802286e-05,
      "loss": 2.465286636352539,
      "memory(GiB)": 77.56,
      "step": 39780,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.439834
    },
    {
      "epoch": 1.704511374833983,
      "grad_norm": 7.101477146148682,
      "learning_rate": 7.396338901375588e-05,
      "loss": 2.5579864501953127,
      "memory(GiB)": 77.56,
      "step": 39785,
      "token_acc": 0.48134328358208955,
      "train_speed(iter/s)": 1.439898
    },
    {
      "epoch": 1.704725590163232,
      "grad_norm": 3.6944642066955566,
      "learning_rate": 7.395748228536964e-05,
      "loss": 2.457639694213867,
      "memory(GiB)": 77.56,
      "step": 39790,
      "token_acc": 0.4742857142857143,
      "train_speed(iter/s)": 1.43987
    },
    {
      "epoch": 1.7049398054924811,
      "grad_norm": 6.27808141708374,
      "learning_rate": 7.395157512297114e-05,
      "loss": 2.722454071044922,
      "memory(GiB)": 77.56,
      "step": 39795,
      "token_acc": 0.4633431085043988,
      "train_speed(iter/s)": 1.439889
    },
    {
      "epoch": 1.70515402082173,
      "grad_norm": 4.822300910949707,
      "learning_rate": 7.394566752666739e-05,
      "loss": 2.625094985961914,
      "memory(GiB)": 77.56,
      "step": 39800,
      "token_acc": 0.4794952681388013,
      "train_speed(iter/s)": 1.439868
    },
    {
      "epoch": 1.705368236150979,
      "grad_norm": 4.021412372589111,
      "learning_rate": 7.39397594965654e-05,
      "loss": 2.633833312988281,
      "memory(GiB)": 77.56,
      "step": 39805,
      "token_acc": 0.47074468085106386,
      "train_speed(iter/s)": 1.439901
    },
    {
      "epoch": 1.705582451480228,
      "grad_norm": 4.1343207359313965,
      "learning_rate": 7.393385103277222e-05,
      "loss": 2.423944854736328,
      "memory(GiB)": 77.56,
      "step": 39810,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.439913
    },
    {
      "epoch": 1.7057966668094768,
      "grad_norm": 5.7742509841918945,
      "learning_rate": 7.392794213539486e-05,
      "loss": 2.581664276123047,
      "memory(GiB)": 77.56,
      "step": 39815,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.439875
    },
    {
      "epoch": 1.7060108821387259,
      "grad_norm": 6.122373104095459,
      "learning_rate": 7.392203280454039e-05,
      "loss": 2.482340431213379,
      "memory(GiB)": 77.56,
      "step": 39820,
      "token_acc": 0.48562300319488816,
      "train_speed(iter/s)": 1.43989
    },
    {
      "epoch": 1.706225097467975,
      "grad_norm": 6.736767292022705,
      "learning_rate": 7.391612304031585e-05,
      "loss": 2.3063446044921876,
      "memory(GiB)": 77.56,
      "step": 39825,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.439861
    },
    {
      "epoch": 1.7064393127972237,
      "grad_norm": 7.398369789123535,
      "learning_rate": 7.391021284282833e-05,
      "loss": 2.6417974472045898,
      "memory(GiB)": 77.56,
      "step": 39830,
      "token_acc": 0.42412451361867703,
      "train_speed(iter/s)": 1.439877
    },
    {
      "epoch": 1.7066535281264728,
      "grad_norm": 4.481542587280273,
      "learning_rate": 7.390430221218485e-05,
      "loss": 2.402405548095703,
      "memory(GiB)": 77.56,
      "step": 39835,
      "token_acc": 0.48031496062992124,
      "train_speed(iter/s)": 1.439926
    },
    {
      "epoch": 1.7068677434557218,
      "grad_norm": 5.377448081970215,
      "learning_rate": 7.389839114849253e-05,
      "loss": 2.545908737182617,
      "memory(GiB)": 77.56,
      "step": 39840,
      "token_acc": 0.450199203187251,
      "train_speed(iter/s)": 1.439923
    },
    {
      "epoch": 1.7070819587849706,
      "grad_norm": 5.170248985290527,
      "learning_rate": 7.389247965185841e-05,
      "loss": 2.706021308898926,
      "memory(GiB)": 77.56,
      "step": 39845,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.439898
    },
    {
      "epoch": 1.7072961741142196,
      "grad_norm": 4.069697380065918,
      "learning_rate": 7.388656772238964e-05,
      "loss": 2.29998893737793,
      "memory(GiB)": 77.56,
      "step": 39850,
      "token_acc": 0.5194805194805194,
      "train_speed(iter/s)": 1.439912
    },
    {
      "epoch": 1.7075103894434687,
      "grad_norm": 4.980340957641602,
      "learning_rate": 7.388065536019327e-05,
      "loss": 2.6758806228637697,
      "memory(GiB)": 77.56,
      "step": 39855,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.439862
    },
    {
      "epoch": 1.7077246047727175,
      "grad_norm": 4.526433944702148,
      "learning_rate": 7.387474256537642e-05,
      "loss": 2.6746925354003905,
      "memory(GiB)": 77.56,
      "step": 39860,
      "token_acc": 0.42138364779874216,
      "train_speed(iter/s)": 1.439892
    },
    {
      "epoch": 1.7079388201019665,
      "grad_norm": 4.018715858459473,
      "learning_rate": 7.386882933804621e-05,
      "loss": 2.4466697692871096,
      "memory(GiB)": 77.56,
      "step": 39865,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.439906
    },
    {
      "epoch": 1.7081530354312155,
      "grad_norm": 7.17388391494751,
      "learning_rate": 7.386291567830978e-05,
      "loss": 2.3986154556274415,
      "memory(GiB)": 77.56,
      "step": 39870,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.439916
    },
    {
      "epoch": 1.7083672507604644,
      "grad_norm": 6.041895389556885,
      "learning_rate": 7.385700158627424e-05,
      "loss": 2.636348342895508,
      "memory(GiB)": 77.56,
      "step": 39875,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.439911
    },
    {
      "epoch": 1.7085814660897134,
      "grad_norm": 4.120418071746826,
      "learning_rate": 7.385108706204675e-05,
      "loss": 2.544120025634766,
      "memory(GiB)": 77.56,
      "step": 39880,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.439883
    },
    {
      "epoch": 1.7087956814189624,
      "grad_norm": 5.060544013977051,
      "learning_rate": 7.384517210573444e-05,
      "loss": 2.36533203125,
      "memory(GiB)": 77.56,
      "step": 39885,
      "token_acc": 0.5083056478405316,
      "train_speed(iter/s)": 1.439903
    },
    {
      "epoch": 1.7090098967482112,
      "grad_norm": 5.357654571533203,
      "learning_rate": 7.383925671744446e-05,
      "loss": 2.6864748001098633,
      "memory(GiB)": 77.56,
      "step": 39890,
      "token_acc": 0.4606741573033708,
      "train_speed(iter/s)": 1.439915
    },
    {
      "epoch": 1.7092241120774603,
      "grad_norm": 4.374958515167236,
      "learning_rate": 7.383334089728398e-05,
      "loss": 2.5595218658447267,
      "memory(GiB)": 77.56,
      "step": 39895,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.439934
    },
    {
      "epoch": 1.7094383274067093,
      "grad_norm": 4.889717102050781,
      "learning_rate": 7.382742464536017e-05,
      "loss": 2.3687376022338866,
      "memory(GiB)": 77.56,
      "step": 39900,
      "token_acc": 0.4536082474226804,
      "train_speed(iter/s)": 1.439981
    },
    {
      "epoch": 1.7096525427359581,
      "grad_norm": 4.771450519561768,
      "learning_rate": 7.382150796178022e-05,
      "loss": 2.6416671752929686,
      "memory(GiB)": 77.56,
      "step": 39905,
      "token_acc": 0.46200607902735563,
      "train_speed(iter/s)": 1.439971
    },
    {
      "epoch": 1.7098667580652072,
      "grad_norm": 4.958680152893066,
      "learning_rate": 7.38155908466513e-05,
      "loss": 2.5969444274902345,
      "memory(GiB)": 77.56,
      "step": 39910,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.439984
    },
    {
      "epoch": 1.7100809733944562,
      "grad_norm": 4.731327533721924,
      "learning_rate": 7.380967330008063e-05,
      "loss": 2.5432605743408203,
      "memory(GiB)": 77.56,
      "step": 39915,
      "token_acc": 0.47674418604651164,
      "train_speed(iter/s)": 1.439984
    },
    {
      "epoch": 1.710295188723705,
      "grad_norm": 4.742739677429199,
      "learning_rate": 7.380375532217537e-05,
      "loss": 2.3297481536865234,
      "memory(GiB)": 77.56,
      "step": 39920,
      "token_acc": 0.4983277591973244,
      "train_speed(iter/s)": 1.439959
    },
    {
      "epoch": 1.710509404052954,
      "grad_norm": 5.272616863250732,
      "learning_rate": 7.379783691304276e-05,
      "loss": 2.6115653991699217,
      "memory(GiB)": 77.56,
      "step": 39925,
      "token_acc": 0.44776119402985076,
      "train_speed(iter/s)": 1.439977
    },
    {
      "epoch": 1.710723619382203,
      "grad_norm": 5.22698450088501,
      "learning_rate": 7.379191807279002e-05,
      "loss": 2.2420799255371096,
      "memory(GiB)": 77.56,
      "step": 39930,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.439987
    },
    {
      "epoch": 1.7109378347114519,
      "grad_norm": 4.384657859802246,
      "learning_rate": 7.378599880152434e-05,
      "loss": 2.440205764770508,
      "memory(GiB)": 77.56,
      "step": 39935,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.440019
    },
    {
      "epoch": 1.711152050040701,
      "grad_norm": 5.922700881958008,
      "learning_rate": 7.3780079099353e-05,
      "loss": 2.2233476638793945,
      "memory(GiB)": 77.56,
      "step": 39940,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.439979
    },
    {
      "epoch": 1.71136626536995,
      "grad_norm": 4.690421104431152,
      "learning_rate": 7.377415896638322e-05,
      "loss": 3.0306575775146483,
      "memory(GiB)": 77.56,
      "step": 39945,
      "token_acc": 0.43729903536977494,
      "train_speed(iter/s)": 1.43998
    },
    {
      "epoch": 1.7115804806991988,
      "grad_norm": 6.880755424499512,
      "learning_rate": 7.376823840272224e-05,
      "loss": 2.5882944107055663,
      "memory(GiB)": 77.56,
      "step": 39950,
      "token_acc": 0.4485981308411215,
      "train_speed(iter/s)": 1.439973
    },
    {
      "epoch": 1.7117946960284478,
      "grad_norm": 4.755744934082031,
      "learning_rate": 7.376231740847734e-05,
      "loss": 2.4759830474853515,
      "memory(GiB)": 77.56,
      "step": 39955,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.439989
    },
    {
      "epoch": 1.7120089113576968,
      "grad_norm": 4.4547438621521,
      "learning_rate": 7.375639598375574e-05,
      "loss": 2.6666980743408204,
      "memory(GiB)": 77.56,
      "step": 39960,
      "token_acc": 0.4318936877076412,
      "train_speed(iter/s)": 1.439978
    },
    {
      "epoch": 1.7122231266869457,
      "grad_norm": 6.396081447601318,
      "learning_rate": 7.375047412866476e-05,
      "loss": 2.7680437088012697,
      "memory(GiB)": 77.56,
      "step": 39965,
      "token_acc": 0.4522058823529412,
      "train_speed(iter/s)": 1.439976
    },
    {
      "epoch": 1.7124373420161947,
      "grad_norm": 4.523647308349609,
      "learning_rate": 7.374455184331167e-05,
      "loss": 2.577466583251953,
      "memory(GiB)": 77.56,
      "step": 39970,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.440005
    },
    {
      "epoch": 1.7126515573454437,
      "grad_norm": 4.375260829925537,
      "learning_rate": 7.373862912780373e-05,
      "loss": 2.586697578430176,
      "memory(GiB)": 77.56,
      "step": 39975,
      "token_acc": 0.4875,
      "train_speed(iter/s)": 1.440041
    },
    {
      "epoch": 1.7128657726746925,
      "grad_norm": 4.126951217651367,
      "learning_rate": 7.373270598224827e-05,
      "loss": 2.713718223571777,
      "memory(GiB)": 77.56,
      "step": 39980,
      "token_acc": 0.44410876132930516,
      "train_speed(iter/s)": 1.44005
    },
    {
      "epoch": 1.7130799880039416,
      "grad_norm": 5.868197917938232,
      "learning_rate": 7.372678240675256e-05,
      "loss": 2.084040641784668,
      "memory(GiB)": 77.56,
      "step": 39985,
      "token_acc": 0.5331010452961672,
      "train_speed(iter/s)": 1.440071
    },
    {
      "epoch": 1.7132942033331906,
      "grad_norm": 4.283258438110352,
      "learning_rate": 7.372085840142394e-05,
      "loss": 2.444026756286621,
      "memory(GiB)": 77.56,
      "step": 39990,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.440107
    },
    {
      "epoch": 1.7135084186624394,
      "grad_norm": 6.181144714355469,
      "learning_rate": 7.371493396636973e-05,
      "loss": 2.659521293640137,
      "memory(GiB)": 77.56,
      "step": 39995,
      "token_acc": 0.4537313432835821,
      "train_speed(iter/s)": 1.440104
    },
    {
      "epoch": 1.7137226339916884,
      "grad_norm": 4.1230926513671875,
      "learning_rate": 7.370900910169723e-05,
      "loss": 2.811714744567871,
      "memory(GiB)": 77.56,
      "step": 40000,
      "token_acc": 0.44610778443113774,
      "train_speed(iter/s)": 1.44013
    },
    {
      "epoch": 1.7137226339916884,
      "eval_loss": 2.1421358585357666,
      "eval_runtime": 13.9435,
      "eval_samples_per_second": 7.172,
      "eval_steps_per_second": 7.172,
      "eval_token_acc": 0.48042168674698793,
      "step": 40000
    },
    {
      "epoch": 1.7139368493209375,
      "grad_norm": 4.4093241691589355,
      "learning_rate": 7.370308380751378e-05,
      "loss": 2.7694704055786135,
      "memory(GiB)": 77.56,
      "step": 40005,
      "token_acc": 0.4678714859437751,
      "train_speed(iter/s)": 1.439385
    },
    {
      "epoch": 1.7141510646501863,
      "grad_norm": 4.222255229949951,
      "learning_rate": 7.369715808392675e-05,
      "loss": 2.4819902420043944,
      "memory(GiB)": 77.56,
      "step": 40010,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.439422
    },
    {
      "epoch": 1.7143652799794353,
      "grad_norm": 5.060547828674316,
      "learning_rate": 7.369123193104347e-05,
      "loss": 2.6443323135375976,
      "memory(GiB)": 77.56,
      "step": 40015,
      "token_acc": 0.45484949832775917,
      "train_speed(iter/s)": 1.439434
    },
    {
      "epoch": 1.7145794953086844,
      "grad_norm": 6.641648292541504,
      "learning_rate": 7.368530534897128e-05,
      "loss": 2.2589042663574217,
      "memory(GiB)": 77.56,
      "step": 40020,
      "token_acc": 0.49812734082397003,
      "train_speed(iter/s)": 1.43945
    },
    {
      "epoch": 1.7147937106379332,
      "grad_norm": 4.530861854553223,
      "learning_rate": 7.36793783378176e-05,
      "loss": 2.8286441802978515,
      "memory(GiB)": 77.56,
      "step": 40025,
      "token_acc": 0.4258064516129032,
      "train_speed(iter/s)": 1.439499
    },
    {
      "epoch": 1.7150079259671822,
      "grad_norm": 9.142610549926758,
      "learning_rate": 7.367345089768976e-05,
      "loss": 2.3285446166992188,
      "memory(GiB)": 77.56,
      "step": 40030,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.439499
    },
    {
      "epoch": 1.7152221412964312,
      "grad_norm": 11.068404197692871,
      "learning_rate": 7.366752302869513e-05,
      "loss": 2.5566738128662108,
      "memory(GiB)": 77.56,
      "step": 40035,
      "token_acc": 0.48534201954397393,
      "train_speed(iter/s)": 1.439493
    },
    {
      "epoch": 1.71543635662568,
      "grad_norm": 6.662464141845703,
      "learning_rate": 7.366159473094112e-05,
      "loss": 2.384919548034668,
      "memory(GiB)": 77.56,
      "step": 40040,
      "token_acc": 0.4962121212121212,
      "train_speed(iter/s)": 1.439531
    },
    {
      "epoch": 1.715650571954929,
      "grad_norm": 4.910244941711426,
      "learning_rate": 7.365566600453513e-05,
      "loss": 2.5101667404174806,
      "memory(GiB)": 77.56,
      "step": 40045,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.439539
    },
    {
      "epoch": 1.7158647872841781,
      "grad_norm": 5.121800422668457,
      "learning_rate": 7.364973684958456e-05,
      "loss": 2.5433162689208983,
      "memory(GiB)": 77.56,
      "step": 40050,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.439487
    },
    {
      "epoch": 1.716079002613427,
      "grad_norm": 4.069670677185059,
      "learning_rate": 7.364380726619681e-05,
      "loss": 2.6185279846191407,
      "memory(GiB)": 77.56,
      "step": 40055,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.439513
    },
    {
      "epoch": 1.716293217942676,
      "grad_norm": 4.56881046295166,
      "learning_rate": 7.363787725447935e-05,
      "loss": 2.3854040145874023,
      "memory(GiB)": 77.56,
      "step": 40060,
      "token_acc": 0.4946524064171123,
      "train_speed(iter/s)": 1.439536
    },
    {
      "epoch": 1.716507433271925,
      "grad_norm": 6.28679895401001,
      "learning_rate": 7.363194681453953e-05,
      "loss": 2.667953109741211,
      "memory(GiB)": 77.56,
      "step": 40065,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.439571
    },
    {
      "epoch": 1.7167216486011738,
      "grad_norm": 5.500514507293701,
      "learning_rate": 7.362601594648484e-05,
      "loss": 2.5200653076171875,
      "memory(GiB)": 77.56,
      "step": 40070,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.439627
    },
    {
      "epoch": 1.7169358639304229,
      "grad_norm": 8.16668701171875,
      "learning_rate": 7.36200846504227e-05,
      "loss": 2.5562793731689455,
      "memory(GiB)": 77.56,
      "step": 40075,
      "token_acc": 0.45396825396825397,
      "train_speed(iter/s)": 1.439633
    },
    {
      "epoch": 1.717150079259672,
      "grad_norm": 5.220269680023193,
      "learning_rate": 7.361415292646057e-05,
      "loss": 2.1251943588256834,
      "memory(GiB)": 77.56,
      "step": 40080,
      "token_acc": 0.5131578947368421,
      "train_speed(iter/s)": 1.439639
    },
    {
      "epoch": 1.7173642945889207,
      "grad_norm": 5.581764221191406,
      "learning_rate": 7.360822077470592e-05,
      "loss": 2.601914978027344,
      "memory(GiB)": 77.56,
      "step": 40085,
      "token_acc": 0.4244604316546763,
      "train_speed(iter/s)": 1.439678
    },
    {
      "epoch": 1.7175785099181697,
      "grad_norm": 4.447869300842285,
      "learning_rate": 7.360228819526621e-05,
      "loss": 2.872735595703125,
      "memory(GiB)": 77.56,
      "step": 40090,
      "token_acc": 0.41297935103244837,
      "train_speed(iter/s)": 1.439709
    },
    {
      "epoch": 1.7177927252474188,
      "grad_norm": 5.055079936981201,
      "learning_rate": 7.35963551882489e-05,
      "loss": 2.4841999053955077,
      "memory(GiB)": 77.56,
      "step": 40095,
      "token_acc": 0.4144486692015209,
      "train_speed(iter/s)": 1.439765
    },
    {
      "epoch": 1.7180069405766676,
      "grad_norm": 4.955320835113525,
      "learning_rate": 7.359042175376148e-05,
      "loss": 2.6695268630981444,
      "memory(GiB)": 77.56,
      "step": 40100,
      "token_acc": 0.4590643274853801,
      "train_speed(iter/s)": 1.439755
    },
    {
      "epoch": 1.7182211559059166,
      "grad_norm": 5.731344223022461,
      "learning_rate": 7.358448789191142e-05,
      "loss": 2.4094270706176757,
      "memory(GiB)": 77.56,
      "step": 40105,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.439749
    },
    {
      "epoch": 1.7184353712351657,
      "grad_norm": 4.8525800704956055,
      "learning_rate": 7.357855360280626e-05,
      "loss": 2.6583824157714844,
      "memory(GiB)": 77.56,
      "step": 40110,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.439766
    },
    {
      "epoch": 1.7186495865644145,
      "grad_norm": 4.381115913391113,
      "learning_rate": 7.357261888655349e-05,
      "loss": 2.791868209838867,
      "memory(GiB)": 77.56,
      "step": 40115,
      "token_acc": 0.45652173913043476,
      "train_speed(iter/s)": 1.439771
    },
    {
      "epoch": 1.7188638018936635,
      "grad_norm": 5.044618129730225,
      "learning_rate": 7.35666837432606e-05,
      "loss": 2.681040954589844,
      "memory(GiB)": 77.56,
      "step": 40120,
      "token_acc": 0.43820224719101125,
      "train_speed(iter/s)": 1.439776
    },
    {
      "epoch": 1.7190780172229125,
      "grad_norm": 5.593702793121338,
      "learning_rate": 7.356074817303513e-05,
      "loss": 2.636488342285156,
      "memory(GiB)": 77.56,
      "step": 40125,
      "token_acc": 0.4434250764525994,
      "train_speed(iter/s)": 1.439741
    },
    {
      "epoch": 1.7192922325521613,
      "grad_norm": 4.197584629058838,
      "learning_rate": 7.355481217598461e-05,
      "loss": 2.3843318939208986,
      "memory(GiB)": 77.56,
      "step": 40130,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.439757
    },
    {
      "epoch": 1.7195064478814104,
      "grad_norm": 6.246951580047607,
      "learning_rate": 7.354887575221656e-05,
      "loss": 2.4831682205200196,
      "memory(GiB)": 77.56,
      "step": 40135,
      "token_acc": 0.5014749262536873,
      "train_speed(iter/s)": 1.439778
    },
    {
      "epoch": 1.7197206632106594,
      "grad_norm": 5.10792875289917,
      "learning_rate": 7.354293890183854e-05,
      "loss": 2.4761425018310548,
      "memory(GiB)": 77.56,
      "step": 40140,
      "token_acc": 0.44375,
      "train_speed(iter/s)": 1.439806
    },
    {
      "epoch": 1.7199348785399082,
      "grad_norm": 5.021482944488525,
      "learning_rate": 7.353700162495811e-05,
      "loss": 2.6779758453369142,
      "memory(GiB)": 77.56,
      "step": 40145,
      "token_acc": 0.45724907063197023,
      "train_speed(iter/s)": 1.439777
    },
    {
      "epoch": 1.7201490938691573,
      "grad_norm": 6.702967166900635,
      "learning_rate": 7.35310639216828e-05,
      "loss": 2.804609680175781,
      "memory(GiB)": 77.56,
      "step": 40150,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.439798
    },
    {
      "epoch": 1.7203633091984063,
      "grad_norm": 4.085918426513672,
      "learning_rate": 7.35251257921202e-05,
      "loss": 2.5090551376342773,
      "memory(GiB)": 77.56,
      "step": 40155,
      "token_acc": 0.5038759689922481,
      "train_speed(iter/s)": 1.439851
    },
    {
      "epoch": 1.7205775245276551,
      "grad_norm": 5.104355335235596,
      "learning_rate": 7.351918723637788e-05,
      "loss": 2.4670711517333985,
      "memory(GiB)": 77.56,
      "step": 40160,
      "token_acc": 0.4671814671814672,
      "train_speed(iter/s)": 1.439836
    },
    {
      "epoch": 1.7207917398569041,
      "grad_norm": 4.603713512420654,
      "learning_rate": 7.35132482545634e-05,
      "loss": 2.6015100479125977,
      "memory(GiB)": 77.56,
      "step": 40165,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.439837
    },
    {
      "epoch": 1.7210059551861532,
      "grad_norm": 5.6582841873168945,
      "learning_rate": 7.350730884678441e-05,
      "loss": 2.083356475830078,
      "memory(GiB)": 77.56,
      "step": 40170,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.439864
    },
    {
      "epoch": 1.721220170515402,
      "grad_norm": 4.597603797912598,
      "learning_rate": 7.350136901314845e-05,
      "loss": 2.605609321594238,
      "memory(GiB)": 77.56,
      "step": 40175,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.439897
    },
    {
      "epoch": 1.721434385844651,
      "grad_norm": 4.485691070556641,
      "learning_rate": 7.349542875376312e-05,
      "loss": 2.3316486358642576,
      "memory(GiB)": 77.56,
      "step": 40180,
      "token_acc": 0.49554896142433236,
      "train_speed(iter/s)": 1.439915
    },
    {
      "epoch": 1.7216486011739,
      "grad_norm": 5.283058166503906,
      "learning_rate": 7.348948806873609e-05,
      "loss": 2.434284782409668,
      "memory(GiB)": 77.56,
      "step": 40185,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.439935
    },
    {
      "epoch": 1.7218628165031489,
      "grad_norm": 3.7487587928771973,
      "learning_rate": 7.348354695817493e-05,
      "loss": 2.2351150512695312,
      "memory(GiB)": 77.56,
      "step": 40190,
      "token_acc": 0.5301587301587302,
      "train_speed(iter/s)": 1.43996
    },
    {
      "epoch": 1.722077031832398,
      "grad_norm": 5.850497722625732,
      "learning_rate": 7.347760542218729e-05,
      "loss": 2.428093147277832,
      "memory(GiB)": 77.56,
      "step": 40195,
      "token_acc": 0.47035573122529645,
      "train_speed(iter/s)": 1.439974
    },
    {
      "epoch": 1.722291247161647,
      "grad_norm": 6.800174236297607,
      "learning_rate": 7.347166346088081e-05,
      "loss": 2.807633972167969,
      "memory(GiB)": 77.56,
      "step": 40200,
      "token_acc": 0.4141791044776119,
      "train_speed(iter/s)": 1.440031
    },
    {
      "epoch": 1.7225054624908958,
      "grad_norm": 5.8759002685546875,
      "learning_rate": 7.346572107436313e-05,
      "loss": 2.7058908462524416,
      "memory(GiB)": 77.56,
      "step": 40205,
      "token_acc": 0.46745562130177515,
      "train_speed(iter/s)": 1.440019
    },
    {
      "epoch": 1.7227196778201448,
      "grad_norm": 5.107700347900391,
      "learning_rate": 7.345977826274189e-05,
      "loss": 2.612692451477051,
      "memory(GiB)": 77.56,
      "step": 40210,
      "token_acc": 0.49008498583569404,
      "train_speed(iter/s)": 1.440046
    },
    {
      "epoch": 1.7229338931493938,
      "grad_norm": 4.888387680053711,
      "learning_rate": 7.345383502612477e-05,
      "loss": 2.2837772369384766,
      "memory(GiB)": 77.56,
      "step": 40215,
      "token_acc": 0.4867924528301887,
      "train_speed(iter/s)": 1.44
    },
    {
      "epoch": 1.7231481084786426,
      "grad_norm": 5.610747337341309,
      "learning_rate": 7.344789136461942e-05,
      "loss": 2.562638854980469,
      "memory(GiB)": 77.56,
      "step": 40220,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440013
    },
    {
      "epoch": 1.7233623238078917,
      "grad_norm": 4.89550256729126,
      "learning_rate": 7.344194727833354e-05,
      "loss": 2.660519027709961,
      "memory(GiB)": 77.56,
      "step": 40225,
      "token_acc": 0.44074074074074077,
      "train_speed(iter/s)": 1.440051
    },
    {
      "epoch": 1.7235765391371407,
      "grad_norm": 5.416114807128906,
      "learning_rate": 7.343600276737477e-05,
      "loss": 2.574527931213379,
      "memory(GiB)": 77.56,
      "step": 40230,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.440067
    },
    {
      "epoch": 1.7237907544663895,
      "grad_norm": 4.884213924407959,
      "learning_rate": 7.343005783185085e-05,
      "loss": 2.445771408081055,
      "memory(GiB)": 77.56,
      "step": 40235,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.440073
    },
    {
      "epoch": 1.7240049697956386,
      "grad_norm": 4.314901351928711,
      "learning_rate": 7.342411247186942e-05,
      "loss": 2.433149719238281,
      "memory(GiB)": 77.56,
      "step": 40240,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.440096
    },
    {
      "epoch": 1.7242191851248876,
      "grad_norm": 4.202539920806885,
      "learning_rate": 7.341816668753825e-05,
      "loss": 2.4923582077026367,
      "memory(GiB)": 77.56,
      "step": 40245,
      "token_acc": 0.49606299212598426,
      "train_speed(iter/s)": 1.440119
    },
    {
      "epoch": 1.7244334004541364,
      "grad_norm": 4.560611724853516,
      "learning_rate": 7.341222047896501e-05,
      "loss": 2.562312126159668,
      "memory(GiB)": 77.56,
      "step": 40250,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.440164
    },
    {
      "epoch": 1.7246476157833854,
      "grad_norm": 5.547539234161377,
      "learning_rate": 7.340627384625743e-05,
      "loss": 2.4742088317871094,
      "memory(GiB)": 77.56,
      "step": 40255,
      "token_acc": 0.4910394265232975,
      "train_speed(iter/s)": 1.440186
    },
    {
      "epoch": 1.7248618311126345,
      "grad_norm": 6.14340353012085,
      "learning_rate": 7.340032678952325e-05,
      "loss": 2.6248313903808596,
      "memory(GiB)": 77.56,
      "step": 40260,
      "token_acc": 0.4820359281437126,
      "train_speed(iter/s)": 1.440201
    },
    {
      "epoch": 1.7250760464418833,
      "grad_norm": 4.854167938232422,
      "learning_rate": 7.339437930887018e-05,
      "loss": 2.7901308059692385,
      "memory(GiB)": 77.56,
      "step": 40265,
      "token_acc": 0.42244224422442245,
      "train_speed(iter/s)": 1.440242
    },
    {
      "epoch": 1.7252902617711323,
      "grad_norm": 4.3600239753723145,
      "learning_rate": 7.338843140440601e-05,
      "loss": 2.5119213104248046,
      "memory(GiB)": 77.56,
      "step": 40270,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.440244
    },
    {
      "epoch": 1.7255044771003814,
      "grad_norm": 4.788198471069336,
      "learning_rate": 7.338248307623844e-05,
      "loss": 2.515062141418457,
      "memory(GiB)": 77.56,
      "step": 40275,
      "token_acc": 0.46785714285714286,
      "train_speed(iter/s)": 1.440237
    },
    {
      "epoch": 1.7257186924296302,
      "grad_norm": 6.201361656188965,
      "learning_rate": 7.337653432447527e-05,
      "loss": 2.6870807647705077,
      "memory(GiB)": 77.56,
      "step": 40280,
      "token_acc": 0.45565749235474007,
      "train_speed(iter/s)": 1.440215
    },
    {
      "epoch": 1.7259329077588792,
      "grad_norm": 8.603616714477539,
      "learning_rate": 7.337058514922425e-05,
      "loss": 2.7269718170166017,
      "memory(GiB)": 77.56,
      "step": 40285,
      "token_acc": 0.41603053435114506,
      "train_speed(iter/s)": 1.440221
    },
    {
      "epoch": 1.7261471230881282,
      "grad_norm": 4.739927768707275,
      "learning_rate": 7.336463555059316e-05,
      "loss": 2.577268600463867,
      "memory(GiB)": 77.56,
      "step": 40290,
      "token_acc": 0.45695364238410596,
      "train_speed(iter/s)": 1.440273
    },
    {
      "epoch": 1.726361338417377,
      "grad_norm": 4.7373576164245605,
      "learning_rate": 7.335868552868979e-05,
      "loss": 2.313774299621582,
      "memory(GiB)": 77.56,
      "step": 40295,
      "token_acc": 0.5140562248995983,
      "train_speed(iter/s)": 1.440318
    },
    {
      "epoch": 1.726575553746626,
      "grad_norm": 5.206615924835205,
      "learning_rate": 7.335273508362189e-05,
      "loss": 2.613463592529297,
      "memory(GiB)": 77.56,
      "step": 40300,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440332
    },
    {
      "epoch": 1.7267897690758751,
      "grad_norm": 4.584021091461182,
      "learning_rate": 7.334678421549731e-05,
      "loss": 2.2635498046875,
      "memory(GiB)": 77.56,
      "step": 40305,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.440335
    },
    {
      "epoch": 1.727003984405124,
      "grad_norm": 6.164045810699463,
      "learning_rate": 7.334083292442382e-05,
      "loss": 2.4893268585205077,
      "memory(GiB)": 77.56,
      "step": 40310,
      "token_acc": 0.5100401606425703,
      "train_speed(iter/s)": 1.440352
    },
    {
      "epoch": 1.727218199734373,
      "grad_norm": 5.9391584396362305,
      "learning_rate": 7.333488121050925e-05,
      "loss": 2.485063362121582,
      "memory(GiB)": 77.56,
      "step": 40315,
      "token_acc": 0.46814404432132967,
      "train_speed(iter/s)": 1.440309
    },
    {
      "epoch": 1.727432415063622,
      "grad_norm": 5.095292091369629,
      "learning_rate": 7.332892907386142e-05,
      "loss": 2.3862890243530273,
      "memory(GiB)": 77.56,
      "step": 40320,
      "token_acc": 0.49480968858131485,
      "train_speed(iter/s)": 1.44027
    },
    {
      "epoch": 1.7276466303928708,
      "grad_norm": 3.871943712234497,
      "learning_rate": 7.332297651458815e-05,
      "loss": 2.4703323364257814,
      "memory(GiB)": 77.56,
      "step": 40325,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.440295
    },
    {
      "epoch": 1.7278608457221198,
      "grad_norm": 4.925197124481201,
      "learning_rate": 7.33170235327973e-05,
      "loss": 2.485773468017578,
      "memory(GiB)": 77.56,
      "step": 40330,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.440274
    },
    {
      "epoch": 1.7280750610513689,
      "grad_norm": 5.479426860809326,
      "learning_rate": 7.331107012859667e-05,
      "loss": 2.6705928802490235,
      "memory(GiB)": 77.56,
      "step": 40335,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440299
    },
    {
      "epoch": 1.7282892763806177,
      "grad_norm": 4.657426834106445,
      "learning_rate": 7.330511630209414e-05,
      "loss": 2.5968000411987306,
      "memory(GiB)": 77.56,
      "step": 40340,
      "token_acc": 0.4686192468619247,
      "train_speed(iter/s)": 1.440307
    },
    {
      "epoch": 1.7285034917098667,
      "grad_norm": 6.10496711730957,
      "learning_rate": 7.329916205339757e-05,
      "loss": 2.439530944824219,
      "memory(GiB)": 77.56,
      "step": 40345,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.440339
    },
    {
      "epoch": 1.7287177070391158,
      "grad_norm": 5.403059959411621,
      "learning_rate": 7.329320738261484e-05,
      "loss": 2.4468931198120116,
      "memory(GiB)": 77.56,
      "step": 40350,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.440343
    },
    {
      "epoch": 1.7289319223683646,
      "grad_norm": 7.664263725280762,
      "learning_rate": 7.32872522898538e-05,
      "loss": 2.4133537292480467,
      "memory(GiB)": 77.56,
      "step": 40355,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.440344
    },
    {
      "epoch": 1.7291461376976136,
      "grad_norm": 4.846518516540527,
      "learning_rate": 7.328129677522234e-05,
      "loss": 2.357892608642578,
      "memory(GiB)": 77.56,
      "step": 40360,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.440393
    },
    {
      "epoch": 1.7293603530268626,
      "grad_norm": 7.284416198730469,
      "learning_rate": 7.327534083882832e-05,
      "loss": 2.574083709716797,
      "memory(GiB)": 77.56,
      "step": 40365,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.440409
    },
    {
      "epoch": 1.7295745683561115,
      "grad_norm": 5.343515396118164,
      "learning_rate": 7.32693844807797e-05,
      "loss": 2.256794738769531,
      "memory(GiB)": 77.56,
      "step": 40370,
      "token_acc": 0.49173553719008267,
      "train_speed(iter/s)": 1.440428
    },
    {
      "epoch": 1.7297887836853605,
      "grad_norm": 4.493984222412109,
      "learning_rate": 7.326342770118434e-05,
      "loss": 2.4861026763916017,
      "memory(GiB)": 77.56,
      "step": 40375,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.440417
    },
    {
      "epoch": 1.7300029990146095,
      "grad_norm": 5.836231708526611,
      "learning_rate": 7.325747050015017e-05,
      "loss": 2.248910903930664,
      "memory(GiB)": 77.56,
      "step": 40380,
      "token_acc": 0.5140562248995983,
      "train_speed(iter/s)": 1.44042
    },
    {
      "epoch": 1.7302172143438583,
      "grad_norm": 5.4983320236206055,
      "learning_rate": 7.325151287778509e-05,
      "loss": 2.727415084838867,
      "memory(GiB)": 77.56,
      "step": 40385,
      "token_acc": 0.43205574912891986,
      "train_speed(iter/s)": 1.440394
    },
    {
      "epoch": 1.7304314296731074,
      "grad_norm": 4.418530464172363,
      "learning_rate": 7.324555483419707e-05,
      "loss": 2.9007171630859374,
      "memory(GiB)": 77.56,
      "step": 40390,
      "token_acc": 0.40804597701149425,
      "train_speed(iter/s)": 1.440392
    },
    {
      "epoch": 1.7306456450023564,
      "grad_norm": 4.905223846435547,
      "learning_rate": 7.323959636949398e-05,
      "loss": 2.598267364501953,
      "memory(GiB)": 77.56,
      "step": 40395,
      "token_acc": 0.45514950166112955,
      "train_speed(iter/s)": 1.440425
    },
    {
      "epoch": 1.7308598603316052,
      "grad_norm": 6.1761016845703125,
      "learning_rate": 7.323363748378384e-05,
      "loss": 2.7709102630615234,
      "memory(GiB)": 77.56,
      "step": 40400,
      "token_acc": 0.43157894736842106,
      "train_speed(iter/s)": 1.440375
    },
    {
      "epoch": 1.7310740756608543,
      "grad_norm": 4.547285079956055,
      "learning_rate": 7.322767817717454e-05,
      "loss": 2.4007984161376954,
      "memory(GiB)": 77.56,
      "step": 40405,
      "token_acc": 0.44594594594594594,
      "train_speed(iter/s)": 1.440407
    },
    {
      "epoch": 1.7312882909901033,
      "grad_norm": 4.873405933380127,
      "learning_rate": 7.322171844977407e-05,
      "loss": 2.385569763183594,
      "memory(GiB)": 77.56,
      "step": 40410,
      "token_acc": 0.5170278637770898,
      "train_speed(iter/s)": 1.440398
    },
    {
      "epoch": 1.731502506319352,
      "grad_norm": 5.374251842498779,
      "learning_rate": 7.321575830169037e-05,
      "loss": 2.63199520111084,
      "memory(GiB)": 77.56,
      "step": 40415,
      "token_acc": 0.4371069182389937,
      "train_speed(iter/s)": 1.440382
    },
    {
      "epoch": 1.7317167216486011,
      "grad_norm": 5.448232173919678,
      "learning_rate": 7.320979773303144e-05,
      "loss": 2.6868612289428713,
      "memory(GiB)": 77.56,
      "step": 40420,
      "token_acc": 0.4573643410852713,
      "train_speed(iter/s)": 1.440409
    },
    {
      "epoch": 1.7319309369778502,
      "grad_norm": 5.896923542022705,
      "learning_rate": 7.320383674390525e-05,
      "loss": 2.3270648956298827,
      "memory(GiB)": 77.56,
      "step": 40425,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.440434
    },
    {
      "epoch": 1.732145152307099,
      "grad_norm": 4.544893741607666,
      "learning_rate": 7.319787533441981e-05,
      "loss": 2.6786264419555663,
      "memory(GiB)": 77.56,
      "step": 40430,
      "token_acc": 0.4495114006514658,
      "train_speed(iter/s)": 1.440451
    },
    {
      "epoch": 1.732359367636348,
      "grad_norm": 4.721642017364502,
      "learning_rate": 7.319191350468308e-05,
      "loss": 2.3792957305908202,
      "memory(GiB)": 77.56,
      "step": 40435,
      "token_acc": 0.49145299145299143,
      "train_speed(iter/s)": 1.440477
    },
    {
      "epoch": 1.732573582965597,
      "grad_norm": 4.844539642333984,
      "learning_rate": 7.318595125480308e-05,
      "loss": 2.4523681640625,
      "memory(GiB)": 77.56,
      "step": 40440,
      "token_acc": 0.4536082474226804,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.7327877982948459,
      "grad_norm": 6.116079330444336,
      "learning_rate": 7.317998858488783e-05,
      "loss": 2.4160694122314452,
      "memory(GiB)": 77.56,
      "step": 40445,
      "token_acc": 0.5065502183406113,
      "train_speed(iter/s)": 1.440436
    },
    {
      "epoch": 1.733002013624095,
      "grad_norm": 4.909404754638672,
      "learning_rate": 7.317402549504533e-05,
      "loss": 2.4019872665405275,
      "memory(GiB)": 77.56,
      "step": 40450,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.440445
    },
    {
      "epoch": 1.733216228953344,
      "grad_norm": 4.96981143951416,
      "learning_rate": 7.316806198538362e-05,
      "loss": 2.250261688232422,
      "memory(GiB)": 77.56,
      "step": 40455,
      "token_acc": 0.5139442231075697,
      "train_speed(iter/s)": 1.440455
    },
    {
      "epoch": 1.7334304442825927,
      "grad_norm": 4.455904960632324,
      "learning_rate": 7.316209805601074e-05,
      "loss": 2.551084518432617,
      "memory(GiB)": 77.56,
      "step": 40460,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.7336446596118418,
      "grad_norm": 5.141857147216797,
      "learning_rate": 7.315613370703473e-05,
      "loss": 2.539170837402344,
      "memory(GiB)": 77.56,
      "step": 40465,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.440488
    },
    {
      "epoch": 1.7338588749410908,
      "grad_norm": 3.741856098175049,
      "learning_rate": 7.315016893856363e-05,
      "loss": 2.7437156677246093,
      "memory(GiB)": 77.56,
      "step": 40470,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.44047
    },
    {
      "epoch": 1.7340730902703396,
      "grad_norm": 5.1214213371276855,
      "learning_rate": 7.31442037507055e-05,
      "loss": 2.3721084594726562,
      "memory(GiB)": 77.56,
      "step": 40475,
      "token_acc": 0.5134328358208955,
      "train_speed(iter/s)": 1.440447
    },
    {
      "epoch": 1.7342873055995887,
      "grad_norm": 6.05742883682251,
      "learning_rate": 7.313823814356841e-05,
      "loss": 2.6238195419311525,
      "memory(GiB)": 77.56,
      "step": 40480,
      "token_acc": 0.4645390070921986,
      "train_speed(iter/s)": 1.44047
    },
    {
      "epoch": 1.7345015209288377,
      "grad_norm": 4.813930988311768,
      "learning_rate": 7.313227211726044e-05,
      "loss": 2.563539505004883,
      "memory(GiB)": 77.56,
      "step": 40485,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.4405
    },
    {
      "epoch": 1.7347157362580865,
      "grad_norm": 4.328474044799805,
      "learning_rate": 7.312630567188965e-05,
      "loss": 2.696954536437988,
      "memory(GiB)": 77.56,
      "step": 40490,
      "token_acc": 0.4397163120567376,
      "train_speed(iter/s)": 1.440528
    },
    {
      "epoch": 1.7349299515873355,
      "grad_norm": 4.531674385070801,
      "learning_rate": 7.312033880756415e-05,
      "loss": 2.4875301361083983,
      "memory(GiB)": 77.56,
      "step": 40495,
      "token_acc": 0.5057034220532319,
      "train_speed(iter/s)": 1.440489
    },
    {
      "epoch": 1.7351441669165846,
      "grad_norm": 4.905977249145508,
      "learning_rate": 7.311437152439201e-05,
      "loss": 2.4641048431396486,
      "memory(GiB)": 77.56,
      "step": 40500,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.440519
    },
    {
      "epoch": 1.7351441669165846,
      "eval_loss": 2.252992630004883,
      "eval_runtime": 14.9239,
      "eval_samples_per_second": 6.701,
      "eval_steps_per_second": 6.701,
      "eval_token_acc": 0.46740638002773927,
      "step": 40500
    },
    {
      "epoch": 1.7353583822458334,
      "grad_norm": 4.80511999130249,
      "learning_rate": 7.310840382248136e-05,
      "loss": 2.5491262435913087,
      "memory(GiB)": 77.56,
      "step": 40505,
      "token_acc": 0.468063872255489,
      "train_speed(iter/s)": 1.439743
    },
    {
      "epoch": 1.7355725975750824,
      "grad_norm": 6.622159004211426,
      "learning_rate": 7.310243570194029e-05,
      "loss": 2.644375228881836,
      "memory(GiB)": 77.56,
      "step": 40510,
      "token_acc": 0.462406015037594,
      "train_speed(iter/s)": 1.43978
    },
    {
      "epoch": 1.7357868129043315,
      "grad_norm": 5.362545013427734,
      "learning_rate": 7.309646716287692e-05,
      "loss": 2.366105079650879,
      "memory(GiB)": 77.56,
      "step": 40515,
      "token_acc": 0.5075987841945289,
      "train_speed(iter/s)": 1.439795
    },
    {
      "epoch": 1.7360010282335803,
      "grad_norm": 5.001947402954102,
      "learning_rate": 7.30904982053994e-05,
      "loss": 2.302295112609863,
      "memory(GiB)": 77.56,
      "step": 40520,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.439772
    },
    {
      "epoch": 1.7362152435628293,
      "grad_norm": 4.6217041015625,
      "learning_rate": 7.308452882961582e-05,
      "loss": 2.5398284912109377,
      "memory(GiB)": 77.56,
      "step": 40525,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.439747
    },
    {
      "epoch": 1.7364294588920783,
      "grad_norm": 4.250296592712402,
      "learning_rate": 7.307855903563435e-05,
      "loss": 2.555429458618164,
      "memory(GiB)": 77.56,
      "step": 40530,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.439796
    },
    {
      "epoch": 1.7366436742213271,
      "grad_norm": 6.470465660095215,
      "learning_rate": 7.307258882356314e-05,
      "loss": 2.4705730438232423,
      "memory(GiB)": 77.56,
      "step": 40535,
      "token_acc": 0.4981132075471698,
      "train_speed(iter/s)": 1.439741
    },
    {
      "epoch": 1.7368578895505762,
      "grad_norm": 4.547605514526367,
      "learning_rate": 7.306661819351033e-05,
      "loss": 2.268496513366699,
      "memory(GiB)": 77.56,
      "step": 40540,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.439733
    },
    {
      "epoch": 1.7370721048798252,
      "grad_norm": 5.024112701416016,
      "learning_rate": 7.306064714558412e-05,
      "loss": 2.406391143798828,
      "memory(GiB)": 77.56,
      "step": 40545,
      "token_acc": 0.48623853211009177,
      "train_speed(iter/s)": 1.439655
    },
    {
      "epoch": 1.737286320209074,
      "grad_norm": 6.351881980895996,
      "learning_rate": 7.305467567989263e-05,
      "loss": 2.446776580810547,
      "memory(GiB)": 77.56,
      "step": 40550,
      "token_acc": 0.5053763440860215,
      "train_speed(iter/s)": 1.439664
    },
    {
      "epoch": 1.7375005355383233,
      "grad_norm": 6.156836032867432,
      "learning_rate": 7.304870379654407e-05,
      "loss": 2.5387594223022463,
      "memory(GiB)": 77.56,
      "step": 40555,
      "token_acc": 0.4696969696969697,
      "train_speed(iter/s)": 1.439714
    },
    {
      "epoch": 1.737714750867572,
      "grad_norm": 4.673503875732422,
      "learning_rate": 7.304273149564662e-05,
      "loss": 2.4603137969970703,
      "memory(GiB)": 77.56,
      "step": 40560,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.43975
    },
    {
      "epoch": 1.737928966196821,
      "grad_norm": 5.059421539306641,
      "learning_rate": 7.303675877730848e-05,
      "loss": 2.2620702743530274,
      "memory(GiB)": 77.56,
      "step": 40565,
      "token_acc": 0.4820359281437126,
      "train_speed(iter/s)": 1.439795
    },
    {
      "epoch": 1.7381431815260702,
      "grad_norm": 4.947644233703613,
      "learning_rate": 7.303078564163783e-05,
      "loss": 2.7470006942749023,
      "memory(GiB)": 77.56,
      "step": 40570,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.439811
    },
    {
      "epoch": 1.738357396855319,
      "grad_norm": 5.241748809814453,
      "learning_rate": 7.30248120887429e-05,
      "loss": 2.7889389038085937,
      "memory(GiB)": 77.56,
      "step": 40575,
      "token_acc": 0.42105263157894735,
      "train_speed(iter/s)": 1.439828
    },
    {
      "epoch": 1.7385716121845678,
      "grad_norm": 4.976478099822998,
      "learning_rate": 7.301883811873191e-05,
      "loss": 2.6292093276977537,
      "memory(GiB)": 77.56,
      "step": 40580,
      "token_acc": 0.47440273037542663,
      "train_speed(iter/s)": 1.43985
    },
    {
      "epoch": 1.738785827513817,
      "grad_norm": 5.1577677726745605,
      "learning_rate": 7.301286373171305e-05,
      "loss": 2.3530048370361327,
      "memory(GiB)": 77.56,
      "step": 40585,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439832
    },
    {
      "epoch": 1.7390000428430659,
      "grad_norm": 6.54850435256958,
      "learning_rate": 7.30068889277946e-05,
      "loss": 2.4994314193725584,
      "memory(GiB)": 77.56,
      "step": 40590,
      "token_acc": 0.47491638795986624,
      "train_speed(iter/s)": 1.439789
    },
    {
      "epoch": 1.7392142581723147,
      "grad_norm": 6.911805629730225,
      "learning_rate": 7.300091370708476e-05,
      "loss": 2.661682891845703,
      "memory(GiB)": 77.56,
      "step": 40595,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.439832
    },
    {
      "epoch": 1.739428473501564,
      "grad_norm": 5.412624835968018,
      "learning_rate": 7.29949380696918e-05,
      "loss": 2.5546459197998046,
      "memory(GiB)": 77.56,
      "step": 40600,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.439827
    },
    {
      "epoch": 1.7396426888308127,
      "grad_norm": 7.305591106414795,
      "learning_rate": 7.298896201572396e-05,
      "loss": 2.5734928131103514,
      "memory(GiB)": 77.56,
      "step": 40605,
      "token_acc": 0.47577092511013214,
      "train_speed(iter/s)": 1.439839
    },
    {
      "epoch": 1.7398569041600616,
      "grad_norm": 3.903895616531372,
      "learning_rate": 7.298298554528952e-05,
      "loss": 2.597462844848633,
      "memory(GiB)": 77.56,
      "step": 40610,
      "token_acc": 0.4817073170731707,
      "train_speed(iter/s)": 1.43983
    },
    {
      "epoch": 1.7400711194893108,
      "grad_norm": 4.521829605102539,
      "learning_rate": 7.297700865849672e-05,
      "loss": 2.549844169616699,
      "memory(GiB)": 77.56,
      "step": 40615,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.439827
    },
    {
      "epoch": 1.7402853348185596,
      "grad_norm": 6.840327739715576,
      "learning_rate": 7.297103135545387e-05,
      "loss": 2.335257720947266,
      "memory(GiB)": 77.56,
      "step": 40620,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.439856
    },
    {
      "epoch": 1.7404995501478084,
      "grad_norm": 8.244111061096191,
      "learning_rate": 7.296505363626922e-05,
      "loss": 2.7429428100585938,
      "memory(GiB)": 77.56,
      "step": 40625,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.439786
    },
    {
      "epoch": 1.7407137654770577,
      "grad_norm": 4.791749954223633,
      "learning_rate": 7.29590755010511e-05,
      "loss": 2.3054740905761717,
      "memory(GiB)": 77.56,
      "step": 40630,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.439785
    },
    {
      "epoch": 1.7409279808063065,
      "grad_norm": 5.039892196655273,
      "learning_rate": 7.295309694990777e-05,
      "loss": 2.410167694091797,
      "memory(GiB)": 77.56,
      "step": 40635,
      "token_acc": 0.5041666666666667,
      "train_speed(iter/s)": 1.439787
    },
    {
      "epoch": 1.7411421961355553,
      "grad_norm": 5.623873710632324,
      "learning_rate": 7.294711798294758e-05,
      "loss": 2.36167049407959,
      "memory(GiB)": 77.56,
      "step": 40640,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.439802
    },
    {
      "epoch": 1.7413564114648046,
      "grad_norm": 7.293215751647949,
      "learning_rate": 7.29411386002788e-05,
      "loss": 2.4238561630249023,
      "memory(GiB)": 77.56,
      "step": 40645,
      "token_acc": 0.5064377682403434,
      "train_speed(iter/s)": 1.439811
    },
    {
      "epoch": 1.7415706267940534,
      "grad_norm": 3.7725753784179688,
      "learning_rate": 7.293515880200978e-05,
      "loss": 2.543780708312988,
      "memory(GiB)": 77.56,
      "step": 40650,
      "token_acc": 0.47941176470588237,
      "train_speed(iter/s)": 1.43979
    },
    {
      "epoch": 1.7417848421233022,
      "grad_norm": 4.8384881019592285,
      "learning_rate": 7.292917858824885e-05,
      "loss": 2.6027462005615236,
      "memory(GiB)": 77.56,
      "step": 40655,
      "token_acc": 0.44518272425249167,
      "train_speed(iter/s)": 1.439787
    },
    {
      "epoch": 1.7419990574525515,
      "grad_norm": 4.457609176635742,
      "learning_rate": 7.292319795910434e-05,
      "loss": 2.5626333236694334,
      "memory(GiB)": 77.56,
      "step": 40660,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.439737
    },
    {
      "epoch": 1.7422132727818003,
      "grad_norm": 6.2772064208984375,
      "learning_rate": 7.29172169146846e-05,
      "loss": 2.378091049194336,
      "memory(GiB)": 77.56,
      "step": 40665,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.439735
    },
    {
      "epoch": 1.742427488111049,
      "grad_norm": 4.154864311218262,
      "learning_rate": 7.291123545509796e-05,
      "loss": 2.60838623046875,
      "memory(GiB)": 77.56,
      "step": 40670,
      "token_acc": 0.4575757575757576,
      "train_speed(iter/s)": 1.439747
    },
    {
      "epoch": 1.7426417034402983,
      "grad_norm": 4.296637058258057,
      "learning_rate": 7.290525358045279e-05,
      "loss": 2.543315887451172,
      "memory(GiB)": 77.56,
      "step": 40675,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.439745
    },
    {
      "epoch": 1.7428559187695472,
      "grad_norm": 4.667581081390381,
      "learning_rate": 7.289927129085749e-05,
      "loss": 2.6319957733154298,
      "memory(GiB)": 77.56,
      "step": 40680,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.439765
    },
    {
      "epoch": 1.743070134098796,
      "grad_norm": 4.641863822937012,
      "learning_rate": 7.28932885864204e-05,
      "loss": 2.4675020217895507,
      "memory(GiB)": 77.56,
      "step": 40685,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.439802
    },
    {
      "epoch": 1.7432843494280452,
      "grad_norm": 6.663127422332764,
      "learning_rate": 7.28873054672499e-05,
      "loss": 2.5931159973144533,
      "memory(GiB)": 77.56,
      "step": 40690,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439792
    },
    {
      "epoch": 1.743498564757294,
      "grad_norm": 6.932421684265137,
      "learning_rate": 7.288132193345443e-05,
      "loss": 2.7125431060791017,
      "memory(GiB)": 77.56,
      "step": 40695,
      "token_acc": 0.4606413994169096,
      "train_speed(iter/s)": 1.439785
    },
    {
      "epoch": 1.7437127800865428,
      "grad_norm": 6.435955047607422,
      "learning_rate": 7.287533798514231e-05,
      "loss": 2.5684553146362306,
      "memory(GiB)": 77.56,
      "step": 40700,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.439777
    },
    {
      "epoch": 1.743926995415792,
      "grad_norm": 4.493941307067871,
      "learning_rate": 7.2869353622422e-05,
      "loss": 2.286090087890625,
      "memory(GiB)": 77.56,
      "step": 40705,
      "token_acc": 0.4978165938864629,
      "train_speed(iter/s)": 1.439828
    },
    {
      "epoch": 1.744141210745041,
      "grad_norm": 6.2130913734436035,
      "learning_rate": 7.286336884540189e-05,
      "loss": 2.1310964584350587,
      "memory(GiB)": 77.56,
      "step": 40710,
      "token_acc": 0.5057034220532319,
      "train_speed(iter/s)": 1.439855
    },
    {
      "epoch": 1.7443554260742897,
      "grad_norm": 4.717353820800781,
      "learning_rate": 7.28573836541904e-05,
      "loss": 2.5709308624267577,
      "memory(GiB)": 77.56,
      "step": 40715,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.439865
    },
    {
      "epoch": 1.744569641403539,
      "grad_norm": 5.047801971435547,
      "learning_rate": 7.285139804889598e-05,
      "loss": 2.5895423889160156,
      "memory(GiB)": 77.56,
      "step": 40720,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.43985
    },
    {
      "epoch": 1.7447838567327878,
      "grad_norm": 6.486758708953857,
      "learning_rate": 7.284541202962704e-05,
      "loss": 2.865602493286133,
      "memory(GiB)": 77.56,
      "step": 40725,
      "token_acc": 0.421259842519685,
      "train_speed(iter/s)": 1.439873
    },
    {
      "epoch": 1.7449980720620366,
      "grad_norm": 5.440725803375244,
      "learning_rate": 7.283942559649202e-05,
      "loss": 2.575438690185547,
      "memory(GiB)": 77.56,
      "step": 40730,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.439857
    },
    {
      "epoch": 1.7452122873912859,
      "grad_norm": 5.528493404388428,
      "learning_rate": 7.283343874959941e-05,
      "loss": 2.5446495056152343,
      "memory(GiB)": 77.56,
      "step": 40735,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.439841
    },
    {
      "epoch": 1.7454265027205347,
      "grad_norm": 3.583773612976074,
      "learning_rate": 7.282745148905759e-05,
      "loss": 2.553518867492676,
      "memory(GiB)": 77.56,
      "step": 40740,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.439836
    },
    {
      "epoch": 1.7456407180497835,
      "grad_norm": 4.390019416809082,
      "learning_rate": 7.28214638149751e-05,
      "loss": 2.5765850067138674,
      "memory(GiB)": 77.56,
      "step": 40745,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.439827
    },
    {
      "epoch": 1.7458549333790327,
      "grad_norm": 4.111855983734131,
      "learning_rate": 7.28154757274604e-05,
      "loss": 2.502342414855957,
      "memory(GiB)": 77.56,
      "step": 40750,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439828
    },
    {
      "epoch": 1.7460691487082816,
      "grad_norm": 4.50808048248291,
      "learning_rate": 7.280948722662194e-05,
      "loss": 2.421772766113281,
      "memory(GiB)": 77.56,
      "step": 40755,
      "token_acc": 0.45871559633027525,
      "train_speed(iter/s)": 1.439864
    },
    {
      "epoch": 1.7462833640375304,
      "grad_norm": 5.665604114532471,
      "learning_rate": 7.280349831256821e-05,
      "loss": 2.6266101837158202,
      "memory(GiB)": 77.56,
      "step": 40760,
      "token_acc": 0.47183098591549294,
      "train_speed(iter/s)": 1.439902
    },
    {
      "epoch": 1.7464975793667796,
      "grad_norm": 3.7740471363067627,
      "learning_rate": 7.279750898540774e-05,
      "loss": 2.191357421875,
      "memory(GiB)": 77.56,
      "step": 40765,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.439876
    },
    {
      "epoch": 1.7467117946960284,
      "grad_norm": 6.237283706665039,
      "learning_rate": 7.279151924524899e-05,
      "loss": 2.4772674560546877,
      "memory(GiB)": 77.56,
      "step": 40770,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.439898
    },
    {
      "epoch": 1.7469260100252773,
      "grad_norm": 4.670551776885986,
      "learning_rate": 7.278552909220049e-05,
      "loss": 2.4850196838378906,
      "memory(GiB)": 77.56,
      "step": 40775,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.439908
    },
    {
      "epoch": 1.7471402253545265,
      "grad_norm": 4.446109294891357,
      "learning_rate": 7.277953852637076e-05,
      "loss": 2.4417701721191407,
      "memory(GiB)": 77.56,
      "step": 40780,
      "token_acc": 0.45151515151515154,
      "train_speed(iter/s)": 1.439901
    },
    {
      "epoch": 1.7473544406837753,
      "grad_norm": 4.435632228851318,
      "learning_rate": 7.277354754786832e-05,
      "loss": 2.644916534423828,
      "memory(GiB)": 77.56,
      "step": 40785,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.439934
    },
    {
      "epoch": 1.7475686560130241,
      "grad_norm": 4.747522354125977,
      "learning_rate": 7.27675561568017e-05,
      "loss": 2.4676158905029295,
      "memory(GiB)": 77.56,
      "step": 40790,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.43992
    },
    {
      "epoch": 1.7477828713422734,
      "grad_norm": 4.9789018630981445,
      "learning_rate": 7.276156435327946e-05,
      "loss": 2.4871990203857424,
      "memory(GiB)": 77.56,
      "step": 40795,
      "token_acc": 0.4867924528301887,
      "train_speed(iter/s)": 1.439914
    },
    {
      "epoch": 1.7479970866715222,
      "grad_norm": 5.806180477142334,
      "learning_rate": 7.27555721374101e-05,
      "loss": 2.245492935180664,
      "memory(GiB)": 77.56,
      "step": 40800,
      "token_acc": 0.5061728395061729,
      "train_speed(iter/s)": 1.439923
    },
    {
      "epoch": 1.748211302000771,
      "grad_norm": 5.388152599334717,
      "learning_rate": 7.274957950930223e-05,
      "loss": 2.5515552520751954,
      "memory(GiB)": 77.56,
      "step": 40805,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.439936
    },
    {
      "epoch": 1.7484255173300203,
      "grad_norm": 5.357909202575684,
      "learning_rate": 7.274358646906438e-05,
      "loss": 2.761534881591797,
      "memory(GiB)": 77.56,
      "step": 40810,
      "token_acc": 0.4553314121037464,
      "train_speed(iter/s)": 1.439951
    },
    {
      "epoch": 1.748639732659269,
      "grad_norm": 3.8138606548309326,
      "learning_rate": 7.273759301680511e-05,
      "loss": 2.443224334716797,
      "memory(GiB)": 77.56,
      "step": 40815,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.439959
    },
    {
      "epoch": 1.748853947988518,
      "grad_norm": 4.7785444259643555,
      "learning_rate": 7.273159915263303e-05,
      "loss": 2.490665245056152,
      "memory(GiB)": 77.56,
      "step": 40820,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.439958
    },
    {
      "epoch": 1.7490681633177672,
      "grad_norm": 5.238621711730957,
      "learning_rate": 7.27256048766567e-05,
      "loss": 2.3670316696166993,
      "memory(GiB)": 77.56,
      "step": 40825,
      "token_acc": 0.5477178423236515,
      "train_speed(iter/s)": 1.439971
    },
    {
      "epoch": 1.749282378647016,
      "grad_norm": 4.694814205169678,
      "learning_rate": 7.271961018898473e-05,
      "loss": 2.3893117904663086,
      "memory(GiB)": 77.56,
      "step": 40830,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.439967
    },
    {
      "epoch": 1.7494965939762648,
      "grad_norm": 7.3936848640441895,
      "learning_rate": 7.27136150897257e-05,
      "loss": 2.376717758178711,
      "memory(GiB)": 77.56,
      "step": 40835,
      "token_acc": 0.5125,
      "train_speed(iter/s)": 1.439947
    },
    {
      "epoch": 1.749710809305514,
      "grad_norm": 5.1302103996276855,
      "learning_rate": 7.270761957898823e-05,
      "loss": 2.715974044799805,
      "memory(GiB)": 77.56,
      "step": 40840,
      "token_acc": 0.41403508771929826,
      "train_speed(iter/s)": 1.439969
    },
    {
      "epoch": 1.7499250246347628,
      "grad_norm": 4.127560615539551,
      "learning_rate": 7.270162365688092e-05,
      "loss": 2.2237516403198243,
      "memory(GiB)": 77.56,
      "step": 40845,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.439949
    },
    {
      "epoch": 1.7501392399640117,
      "grad_norm": 9.205788612365723,
      "learning_rate": 7.269562732351241e-05,
      "loss": 2.6029834747314453,
      "memory(GiB)": 77.56,
      "step": 40850,
      "token_acc": 0.4978165938864629,
      "train_speed(iter/s)": 1.439966
    },
    {
      "epoch": 1.750353455293261,
      "grad_norm": 5.841698169708252,
      "learning_rate": 7.268963057899132e-05,
      "loss": 2.5094099044799805,
      "memory(GiB)": 77.56,
      "step": 40855,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.439963
    },
    {
      "epoch": 1.7505676706225097,
      "grad_norm": 4.756931304931641,
      "learning_rate": 7.268363342342628e-05,
      "loss": 2.6328990936279295,
      "memory(GiB)": 77.56,
      "step": 40860,
      "token_acc": 0.4551083591331269,
      "train_speed(iter/s)": 1.43999
    },
    {
      "epoch": 1.7507818859517585,
      "grad_norm": 4.951889514923096,
      "learning_rate": 7.267763585692595e-05,
      "loss": 2.719615364074707,
      "memory(GiB)": 77.56,
      "step": 40865,
      "token_acc": 0.4646153846153846,
      "train_speed(iter/s)": 1.440017
    },
    {
      "epoch": 1.7509961012810078,
      "grad_norm": 3.907123565673828,
      "learning_rate": 7.267163787959898e-05,
      "loss": 2.297977256774902,
      "memory(GiB)": 77.56,
      "step": 40870,
      "token_acc": 0.5206611570247934,
      "train_speed(iter/s)": 1.440033
    },
    {
      "epoch": 1.7512103166102566,
      "grad_norm": 5.126631736755371,
      "learning_rate": 7.2665639491554e-05,
      "loss": 2.6837745666503907,
      "memory(GiB)": 77.56,
      "step": 40875,
      "token_acc": 0.42058823529411765,
      "train_speed(iter/s)": 1.440065
    },
    {
      "epoch": 1.7514245319395054,
      "grad_norm": 5.0975751876831055,
      "learning_rate": 7.265964069289972e-05,
      "loss": 2.415487861633301,
      "memory(GiB)": 77.56,
      "step": 40880,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.44007
    },
    {
      "epoch": 1.7516387472687547,
      "grad_norm": 7.13967752456665,
      "learning_rate": 7.265364148374478e-05,
      "loss": 2.5960275650024416,
      "memory(GiB)": 77.56,
      "step": 40885,
      "token_acc": 0.45768025078369906,
      "train_speed(iter/s)": 1.440095
    },
    {
      "epoch": 1.7518529625980035,
      "grad_norm": 4.449210166931152,
      "learning_rate": 7.264764186419788e-05,
      "loss": 2.753332328796387,
      "memory(GiB)": 77.56,
      "step": 40890,
      "token_acc": 0.41637010676156583,
      "train_speed(iter/s)": 1.440158
    },
    {
      "epoch": 1.7520671779272525,
      "grad_norm": 6.141135215759277,
      "learning_rate": 7.26416418343677e-05,
      "loss": 2.5775833129882812,
      "memory(GiB)": 77.56,
      "step": 40895,
      "token_acc": 0.43820224719101125,
      "train_speed(iter/s)": 1.440184
    },
    {
      "epoch": 1.7522813932565016,
      "grad_norm": 5.39845609664917,
      "learning_rate": 7.263564139436294e-05,
      "loss": 2.3636911392211912,
      "memory(GiB)": 77.56,
      "step": 40900,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.440214
    },
    {
      "epoch": 1.7524956085857504,
      "grad_norm": 4.237954139709473,
      "learning_rate": 7.26296405442923e-05,
      "loss": 2.444438934326172,
      "memory(GiB)": 77.56,
      "step": 40905,
      "token_acc": 0.5014925373134328,
      "train_speed(iter/s)": 1.440243
    },
    {
      "epoch": 1.7527098239149994,
      "grad_norm": 4.258236408233643,
      "learning_rate": 7.26236392842645e-05,
      "loss": 2.5143136978149414,
      "memory(GiB)": 77.56,
      "step": 40910,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.440267
    },
    {
      "epoch": 1.7529240392442484,
      "grad_norm": 5.368627071380615,
      "learning_rate": 7.261763761438824e-05,
      "loss": 2.27318115234375,
      "memory(GiB)": 77.56,
      "step": 40915,
      "token_acc": 0.51,
      "train_speed(iter/s)": 1.44027
    },
    {
      "epoch": 1.7531382545734973,
      "grad_norm": 6.429764747619629,
      "learning_rate": 7.261163553477226e-05,
      "loss": 2.5919261932373048,
      "memory(GiB)": 77.56,
      "step": 40920,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.440282
    },
    {
      "epoch": 1.7533524699027463,
      "grad_norm": 6.166162014007568,
      "learning_rate": 7.26056330455253e-05,
      "loss": 2.3379310607910155,
      "memory(GiB)": 77.56,
      "step": 40925,
      "token_acc": 0.5148148148148148,
      "train_speed(iter/s)": 1.440318
    },
    {
      "epoch": 1.7535666852319953,
      "grad_norm": 5.194517612457275,
      "learning_rate": 7.259963014675608e-05,
      "loss": 2.538084030151367,
      "memory(GiB)": 77.56,
      "step": 40930,
      "token_acc": 0.4584615384615385,
      "train_speed(iter/s)": 1.440355
    },
    {
      "epoch": 1.7537809005612441,
      "grad_norm": 4.538455963134766,
      "learning_rate": 7.259362683857336e-05,
      "loss": 2.637641716003418,
      "memory(GiB)": 77.56,
      "step": 40935,
      "token_acc": 0.4690909090909091,
      "train_speed(iter/s)": 1.440388
    },
    {
      "epoch": 1.7539951158904932,
      "grad_norm": 4.2257184982299805,
      "learning_rate": 7.258762312108591e-05,
      "loss": 2.419940185546875,
      "memory(GiB)": 77.56,
      "step": 40940,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.440354
    },
    {
      "epoch": 1.7542093312197422,
      "grad_norm": 5.460875988006592,
      "learning_rate": 7.258161899440246e-05,
      "loss": 2.3541849136352537,
      "memory(GiB)": 77.56,
      "step": 40945,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.44034
    },
    {
      "epoch": 1.754423546548991,
      "grad_norm": 6.190771102905273,
      "learning_rate": 7.257561445863182e-05,
      "loss": 2.557106781005859,
      "memory(GiB)": 77.56,
      "step": 40950,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.440363
    },
    {
      "epoch": 1.75463776187824,
      "grad_norm": 5.101850986480713,
      "learning_rate": 7.256960951388274e-05,
      "loss": 2.0622623443603514,
      "memory(GiB)": 77.56,
      "step": 40955,
      "token_acc": 0.5691699604743083,
      "train_speed(iter/s)": 1.440361
    },
    {
      "epoch": 1.754851977207489,
      "grad_norm": 3.9329185485839844,
      "learning_rate": 7.2563604160264e-05,
      "loss": 2.7492170333862305,
      "memory(GiB)": 77.56,
      "step": 40960,
      "token_acc": 0.4983922829581994,
      "train_speed(iter/s)": 1.440382
    },
    {
      "epoch": 1.755066192536738,
      "grad_norm": 4.481163024902344,
      "learning_rate": 7.25575983978844e-05,
      "loss": 2.3399169921875,
      "memory(GiB)": 77.56,
      "step": 40965,
      "token_acc": 0.5016835016835017,
      "train_speed(iter/s)": 1.440381
    },
    {
      "epoch": 1.755280407865987,
      "grad_norm": 5.865595817565918,
      "learning_rate": 7.255159222685277e-05,
      "loss": 2.3756961822509766,
      "memory(GiB)": 77.56,
      "step": 40970,
      "token_acc": 0.43986254295532645,
      "train_speed(iter/s)": 1.440405
    },
    {
      "epoch": 1.755494623195236,
      "grad_norm": 5.64993953704834,
      "learning_rate": 7.254558564727786e-05,
      "loss": 2.5981561660766603,
      "memory(GiB)": 77.56,
      "step": 40975,
      "token_acc": 0.44516129032258067,
      "train_speed(iter/s)": 1.440419
    },
    {
      "epoch": 1.7557088385244848,
      "grad_norm": 4.915028095245361,
      "learning_rate": 7.253957865926854e-05,
      "loss": 2.503937911987305,
      "memory(GiB)": 77.56,
      "step": 40980,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.440413
    },
    {
      "epoch": 1.7559230538537338,
      "grad_norm": 5.909773826599121,
      "learning_rate": 7.253357126293361e-05,
      "loss": 2.189768600463867,
      "memory(GiB)": 77.56,
      "step": 40985,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.44042
    },
    {
      "epoch": 1.7561372691829829,
      "grad_norm": 4.224458694458008,
      "learning_rate": 7.252756345838187e-05,
      "loss": 2.4832731246948243,
      "memory(GiB)": 77.56,
      "step": 40990,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.440475
    },
    {
      "epoch": 1.7563514845122317,
      "grad_norm": 5.250662803649902,
      "learning_rate": 7.252155524572222e-05,
      "loss": 2.818433952331543,
      "memory(GiB)": 77.56,
      "step": 40995,
      "token_acc": 0.445993031358885,
      "train_speed(iter/s)": 1.440457
    },
    {
      "epoch": 1.7565656998414807,
      "grad_norm": 4.951468467712402,
      "learning_rate": 7.251554662506346e-05,
      "loss": 2.559613037109375,
      "memory(GiB)": 77.56,
      "step": 41000,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.440462
    },
    {
      "epoch": 1.7565656998414807,
      "eval_loss": 2.1070873737335205,
      "eval_runtime": 13.8064,
      "eval_samples_per_second": 7.243,
      "eval_steps_per_second": 7.243,
      "eval_token_acc": 0.502283105022831,
      "step": 41000
    },
    {
      "epoch": 1.7567799151707297,
      "grad_norm": 5.524142742156982,
      "learning_rate": 7.250953759651443e-05,
      "loss": 2.285837173461914,
      "memory(GiB)": 77.56,
      "step": 41005,
      "token_acc": 0.49739311783107404,
      "train_speed(iter/s)": 1.439688
    },
    {
      "epoch": 1.7569941304999785,
      "grad_norm": 5.306302070617676,
      "learning_rate": 7.250352816018402e-05,
      "loss": 2.702947425842285,
      "memory(GiB)": 77.56,
      "step": 41010,
      "token_acc": 0.4539877300613497,
      "train_speed(iter/s)": 1.439664
    },
    {
      "epoch": 1.7572083458292276,
      "grad_norm": 4.793148040771484,
      "learning_rate": 7.24975183161811e-05,
      "loss": 2.4987409591674803,
      "memory(GiB)": 77.56,
      "step": 41015,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.439657
    },
    {
      "epoch": 1.7574225611584766,
      "grad_norm": 4.849996566772461,
      "learning_rate": 7.249150806461453e-05,
      "loss": 2.0000219345092773,
      "memory(GiB)": 77.56,
      "step": 41020,
      "token_acc": 0.5860655737704918,
      "train_speed(iter/s)": 1.439639
    },
    {
      "epoch": 1.7576367764877254,
      "grad_norm": 5.711615562438965,
      "learning_rate": 7.248549740559319e-05,
      "loss": 2.657586669921875,
      "memory(GiB)": 77.56,
      "step": 41025,
      "token_acc": 0.48125,
      "train_speed(iter/s)": 1.439641
    },
    {
      "epoch": 1.7578509918169745,
      "grad_norm": 5.536566257476807,
      "learning_rate": 7.247948633922597e-05,
      "loss": 2.314360809326172,
      "memory(GiB)": 77.56,
      "step": 41030,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.439653
    },
    {
      "epoch": 1.7580652071462235,
      "grad_norm": 4.747581958770752,
      "learning_rate": 7.247347486562177e-05,
      "loss": 2.78458251953125,
      "memory(GiB)": 77.56,
      "step": 41035,
      "token_acc": 0.4527687296416938,
      "train_speed(iter/s)": 1.439661
    },
    {
      "epoch": 1.7582794224754723,
      "grad_norm": 5.256016731262207,
      "learning_rate": 7.246746298488949e-05,
      "loss": 2.5093849182128904,
      "memory(GiB)": 77.56,
      "step": 41040,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.439646
    },
    {
      "epoch": 1.7584936378047213,
      "grad_norm": 4.9303131103515625,
      "learning_rate": 7.246145069713804e-05,
      "loss": 2.653642272949219,
      "memory(GiB)": 77.56,
      "step": 41045,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.439685
    },
    {
      "epoch": 1.7587078531339704,
      "grad_norm": 5.3557233810424805,
      "learning_rate": 7.245543800247634e-05,
      "loss": 2.302123260498047,
      "memory(GiB)": 77.56,
      "step": 41050,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.439664
    },
    {
      "epoch": 1.7589220684632192,
      "grad_norm": 5.448339939117432,
      "learning_rate": 7.244942490101332e-05,
      "loss": 2.6585994720458985,
      "memory(GiB)": 77.56,
      "step": 41055,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.439636
    },
    {
      "epoch": 1.7591362837924682,
      "grad_norm": 4.330277442932129,
      "learning_rate": 7.24434113928579e-05,
      "loss": 2.4772113800048827,
      "memory(GiB)": 77.56,
      "step": 41060,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.439667
    },
    {
      "epoch": 1.7593504991217173,
      "grad_norm": 5.223570823669434,
      "learning_rate": 7.243739747811903e-05,
      "loss": 2.4174676895141602,
      "memory(GiB)": 77.56,
      "step": 41065,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.439694
    },
    {
      "epoch": 1.759564714450966,
      "grad_norm": 4.613308429718018,
      "learning_rate": 7.243138315690567e-05,
      "loss": 2.6890644073486327,
      "memory(GiB)": 77.56,
      "step": 41070,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.439736
    },
    {
      "epoch": 1.759778929780215,
      "grad_norm": 4.7630462646484375,
      "learning_rate": 7.242536842932675e-05,
      "loss": 2.692647171020508,
      "memory(GiB)": 77.56,
      "step": 41075,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.439749
    },
    {
      "epoch": 1.7599931451094641,
      "grad_norm": 4.018398284912109,
      "learning_rate": 7.241935329549125e-05,
      "loss": 2.9527984619140626,
      "memory(GiB)": 77.56,
      "step": 41080,
      "token_acc": 0.3961218836565097,
      "train_speed(iter/s)": 1.439776
    },
    {
      "epoch": 1.760207360438713,
      "grad_norm": 4.387391567230225,
      "learning_rate": 7.241333775550813e-05,
      "loss": 2.555002975463867,
      "memory(GiB)": 77.56,
      "step": 41085,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.43978
    },
    {
      "epoch": 1.760421575767962,
      "grad_norm": 4.582672595977783,
      "learning_rate": 7.240732180948637e-05,
      "loss": 2.626940155029297,
      "memory(GiB)": 77.56,
      "step": 41090,
      "token_acc": 0.44886363636363635,
      "train_speed(iter/s)": 1.439762
    },
    {
      "epoch": 1.760635791097211,
      "grad_norm": 5.070387840270996,
      "learning_rate": 7.240130545753496e-05,
      "loss": 2.461312675476074,
      "memory(GiB)": 77.56,
      "step": 41095,
      "token_acc": 0.46540880503144655,
      "train_speed(iter/s)": 1.439801
    },
    {
      "epoch": 1.7608500064264598,
      "grad_norm": 6.758989334106445,
      "learning_rate": 7.239528869976288e-05,
      "loss": 2.7175647735595705,
      "memory(GiB)": 77.56,
      "step": 41100,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.439832
    },
    {
      "epoch": 1.7610642217557089,
      "grad_norm": 5.756537437438965,
      "learning_rate": 7.238927153627914e-05,
      "loss": 2.6489498138427736,
      "memory(GiB)": 77.56,
      "step": 41105,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.43984
    },
    {
      "epoch": 1.761278437084958,
      "grad_norm": 5.375679969787598,
      "learning_rate": 7.238325396719275e-05,
      "loss": 2.4181610107421876,
      "memory(GiB)": 77.56,
      "step": 41110,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439825
    },
    {
      "epoch": 1.7614926524142067,
      "grad_norm": 5.2676873207092285,
      "learning_rate": 7.23772359926127e-05,
      "loss": 2.7358272552490233,
      "memory(GiB)": 77.56,
      "step": 41115,
      "token_acc": 0.45652173913043476,
      "train_speed(iter/s)": 1.439848
    },
    {
      "epoch": 1.7617068677434558,
      "grad_norm": 4.4719767570495605,
      "learning_rate": 7.237121761264805e-05,
      "loss": 2.507393646240234,
      "memory(GiB)": 77.56,
      "step": 41120,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.439867
    },
    {
      "epoch": 1.7619210830727048,
      "grad_norm": 3.822356939315796,
      "learning_rate": 7.23651988274078e-05,
      "loss": 2.652963066101074,
      "memory(GiB)": 77.56,
      "step": 41125,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.439877
    },
    {
      "epoch": 1.7621352984019536,
      "grad_norm": 6.476071357727051,
      "learning_rate": 7.235917963700098e-05,
      "loss": 2.8798969268798826,
      "memory(GiB)": 77.56,
      "step": 41130,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.439907
    },
    {
      "epoch": 1.7623495137312026,
      "grad_norm": 4.881564140319824,
      "learning_rate": 7.235316004153667e-05,
      "loss": 2.4644643783569338,
      "memory(GiB)": 77.56,
      "step": 41135,
      "token_acc": 0.47854785478547857,
      "train_speed(iter/s)": 1.439933
    },
    {
      "epoch": 1.7625637290604517,
      "grad_norm": 4.694281101226807,
      "learning_rate": 7.234714004112388e-05,
      "loss": 2.473439025878906,
      "memory(GiB)": 77.56,
      "step": 41140,
      "token_acc": 0.4387755102040816,
      "train_speed(iter/s)": 1.439956
    },
    {
      "epoch": 1.7627779443897005,
      "grad_norm": 4.466538906097412,
      "learning_rate": 7.234111963587169e-05,
      "loss": 2.450484848022461,
      "memory(GiB)": 77.56,
      "step": 41145,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.439966
    },
    {
      "epoch": 1.7629921597189495,
      "grad_norm": 3.744291305541992,
      "learning_rate": 7.233509882588914e-05,
      "loss": 2.535540771484375,
      "memory(GiB)": 77.56,
      "step": 41150,
      "token_acc": 0.468586387434555,
      "train_speed(iter/s)": 1.439992
    },
    {
      "epoch": 1.7632063750481985,
      "grad_norm": 5.230288982391357,
      "learning_rate": 7.232907761128535e-05,
      "loss": 2.702992248535156,
      "memory(GiB)": 77.56,
      "step": 41155,
      "token_acc": 0.46607669616519176,
      "train_speed(iter/s)": 1.439998
    },
    {
      "epoch": 1.7634205903774474,
      "grad_norm": 5.842164516448975,
      "learning_rate": 7.232305599216938e-05,
      "loss": 2.5083232879638673,
      "memory(GiB)": 77.56,
      "step": 41160,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.439991
    },
    {
      "epoch": 1.7636348057066964,
      "grad_norm": 5.75670862197876,
      "learning_rate": 7.231703396865029e-05,
      "loss": 2.2387208938598633,
      "memory(GiB)": 77.56,
      "step": 41165,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.439951
    },
    {
      "epoch": 1.7638490210359454,
      "grad_norm": 5.6118998527526855,
      "learning_rate": 7.23110115408372e-05,
      "loss": 2.2752613067626952,
      "memory(GiB)": 77.56,
      "step": 41170,
      "token_acc": 0.49814126394052044,
      "train_speed(iter/s)": 1.439923
    },
    {
      "epoch": 1.7640632363651942,
      "grad_norm": 4.402761459350586,
      "learning_rate": 7.23049887088392e-05,
      "loss": 2.512615776062012,
      "memory(GiB)": 77.56,
      "step": 41175,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.439958
    },
    {
      "epoch": 1.7642774516944433,
      "grad_norm": 4.5815653800964355,
      "learning_rate": 7.22989654727654e-05,
      "loss": 2.5429006576538087,
      "memory(GiB)": 77.56,
      "step": 41180,
      "token_acc": 0.5146579804560261,
      "train_speed(iter/s)": 1.43988
    },
    {
      "epoch": 1.7644916670236923,
      "grad_norm": 5.889723777770996,
      "learning_rate": 7.229294183272495e-05,
      "loss": 2.3612640380859373,
      "memory(GiB)": 77.56,
      "step": 41185,
      "token_acc": 0.4847328244274809,
      "train_speed(iter/s)": 1.439876
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 4.5917887687683105,
      "learning_rate": 7.228691778882693e-05,
      "loss": 2.602444076538086,
      "memory(GiB)": 77.56,
      "step": 41190,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.439844
    },
    {
      "epoch": 1.7649200976821902,
      "grad_norm": 5.886419773101807,
      "learning_rate": 7.228089334118047e-05,
      "loss": 2.5652881622314454,
      "memory(GiB)": 77.56,
      "step": 41195,
      "token_acc": 0.44807121661721067,
      "train_speed(iter/s)": 1.439868
    },
    {
      "epoch": 1.7651343130114392,
      "grad_norm": 4.75514554977417,
      "learning_rate": 7.227486848989475e-05,
      "loss": 2.4506084442138674,
      "memory(GiB)": 77.56,
      "step": 41200,
      "token_acc": 0.465625,
      "train_speed(iter/s)": 1.439886
    },
    {
      "epoch": 1.765348528340688,
      "grad_norm": 6.189976215362549,
      "learning_rate": 7.226884323507887e-05,
      "loss": 2.3964977264404297,
      "memory(GiB)": 77.56,
      "step": 41205,
      "token_acc": 0.4866920152091255,
      "train_speed(iter/s)": 1.439918
    },
    {
      "epoch": 1.765562743669937,
      "grad_norm": 4.561886310577393,
      "learning_rate": 7.226281757684204e-05,
      "loss": 2.515126037597656,
      "memory(GiB)": 77.56,
      "step": 41210,
      "token_acc": 0.4427710843373494,
      "train_speed(iter/s)": 1.439932
    },
    {
      "epoch": 1.765776958999186,
      "grad_norm": 4.111982345581055,
      "learning_rate": 7.225679151529337e-05,
      "loss": 2.322934150695801,
      "memory(GiB)": 77.56,
      "step": 41215,
      "token_acc": 0.496875,
      "train_speed(iter/s)": 1.439923
    },
    {
      "epoch": 1.7659911743284349,
      "grad_norm": 6.711462497711182,
      "learning_rate": 7.225076505054202e-05,
      "loss": 2.3757816314697267,
      "memory(GiB)": 77.56,
      "step": 41220,
      "token_acc": 0.48497854077253216,
      "train_speed(iter/s)": 1.439916
    },
    {
      "epoch": 1.766205389657684,
      "grad_norm": 7.27276086807251,
      "learning_rate": 7.224473818269721e-05,
      "loss": 2.670970153808594,
      "memory(GiB)": 77.56,
      "step": 41225,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.439933
    },
    {
      "epoch": 1.766419604986933,
      "grad_norm": 4.742341041564941,
      "learning_rate": 7.22387109118681e-05,
      "loss": 2.542632293701172,
      "memory(GiB)": 77.56,
      "step": 41230,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.439888
    },
    {
      "epoch": 1.7666338203161818,
      "grad_norm": 3.990126848220825,
      "learning_rate": 7.223268323816386e-05,
      "loss": 2.285426139831543,
      "memory(GiB)": 77.56,
      "step": 41235,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.439871
    },
    {
      "epoch": 1.7668480356454308,
      "grad_norm": 4.492398738861084,
      "learning_rate": 7.222665516169373e-05,
      "loss": 2.734975814819336,
      "memory(GiB)": 77.56,
      "step": 41240,
      "token_acc": 0.43302180685358255,
      "train_speed(iter/s)": 1.439904
    },
    {
      "epoch": 1.7670622509746798,
      "grad_norm": 5.796911239624023,
      "learning_rate": 7.22206266825669e-05,
      "loss": 2.61248836517334,
      "memory(GiB)": 77.56,
      "step": 41245,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.439914
    },
    {
      "epoch": 1.7672764663039287,
      "grad_norm": 6.271621227264404,
      "learning_rate": 7.221459780089255e-05,
      "loss": 2.5292232513427733,
      "memory(GiB)": 77.56,
      "step": 41250,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.43988
    },
    {
      "epoch": 1.7674906816331777,
      "grad_norm": 3.8992645740509033,
      "learning_rate": 7.220856851677994e-05,
      "loss": 2.5859981536865235,
      "memory(GiB)": 77.56,
      "step": 41255,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.439914
    },
    {
      "epoch": 1.7677048969624267,
      "grad_norm": 6.706957817077637,
      "learning_rate": 7.220253883033827e-05,
      "loss": 2.9659212112426756,
      "memory(GiB)": 77.56,
      "step": 41260,
      "token_acc": 0.4251968503937008,
      "train_speed(iter/s)": 1.439925
    },
    {
      "epoch": 1.7679191122916755,
      "grad_norm": 5.120011806488037,
      "learning_rate": 7.219650874167675e-05,
      "loss": 2.4343017578125,
      "memory(GiB)": 77.56,
      "step": 41265,
      "token_acc": 0.47278911564625853,
      "train_speed(iter/s)": 1.439921
    },
    {
      "epoch": 1.7681333276209246,
      "grad_norm": 4.392596244812012,
      "learning_rate": 7.21904782509047e-05,
      "loss": 2.519309425354004,
      "memory(GiB)": 77.56,
      "step": 41270,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.43995
    },
    {
      "epoch": 1.7683475429501736,
      "grad_norm": 5.016544818878174,
      "learning_rate": 7.218444735813132e-05,
      "loss": 2.6991399765014648,
      "memory(GiB)": 77.56,
      "step": 41275,
      "token_acc": 0.47843137254901963,
      "train_speed(iter/s)": 1.439997
    },
    {
      "epoch": 1.7685617582794224,
      "grad_norm": 5.663881301879883,
      "learning_rate": 7.217841606346584e-05,
      "loss": 2.2779376983642576,
      "memory(GiB)": 77.56,
      "step": 41280,
      "token_acc": 0.47555555555555556,
      "train_speed(iter/s)": 1.440032
    },
    {
      "epoch": 1.7687759736086714,
      "grad_norm": 5.234511375427246,
      "learning_rate": 7.217238436701756e-05,
      "loss": 2.5336509704589845,
      "memory(GiB)": 77.56,
      "step": 41285,
      "token_acc": 0.4618181818181818,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 1.7689901889379205,
      "grad_norm": 5.388389587402344,
      "learning_rate": 7.216635226889572e-05,
      "loss": 2.098532867431641,
      "memory(GiB)": 77.56,
      "step": 41290,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.440087
    },
    {
      "epoch": 1.7692044042671693,
      "grad_norm": 5.6218037605285645,
      "learning_rate": 7.216031976920963e-05,
      "loss": 2.6895313262939453,
      "memory(GiB)": 77.56,
      "step": 41295,
      "token_acc": 0.4542124542124542,
      "train_speed(iter/s)": 1.440099
    },
    {
      "epoch": 1.7694186195964183,
      "grad_norm": 4.102258682250977,
      "learning_rate": 7.215428686806855e-05,
      "loss": 2.801017189025879,
      "memory(GiB)": 77.56,
      "step": 41300,
      "token_acc": 0.4006024096385542,
      "train_speed(iter/s)": 1.440121
    },
    {
      "epoch": 1.7696328349256674,
      "grad_norm": 8.6804780960083,
      "learning_rate": 7.214825356558181e-05,
      "loss": 2.545354461669922,
      "memory(GiB)": 77.56,
      "step": 41305,
      "token_acc": 0.49603174603174605,
      "train_speed(iter/s)": 1.440125
    },
    {
      "epoch": 1.7698470502549162,
      "grad_norm": 4.438647270202637,
      "learning_rate": 7.214221986185865e-05,
      "loss": 2.752881622314453,
      "memory(GiB)": 77.56,
      "step": 41310,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.440144
    },
    {
      "epoch": 1.7700612655841652,
      "grad_norm": 4.487037181854248,
      "learning_rate": 7.213618575700841e-05,
      "loss": 2.467152214050293,
      "memory(GiB)": 77.56,
      "step": 41315,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.440163
    },
    {
      "epoch": 1.7702754809134142,
      "grad_norm": 4.7058000564575195,
      "learning_rate": 7.21301512511404e-05,
      "loss": 2.692424774169922,
      "memory(GiB)": 77.56,
      "step": 41320,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.440195
    },
    {
      "epoch": 1.770489696242663,
      "grad_norm": 6.961008071899414,
      "learning_rate": 7.212411634436396e-05,
      "loss": 2.5186532974243163,
      "memory(GiB)": 77.56,
      "step": 41325,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.44022
    },
    {
      "epoch": 1.770703911571912,
      "grad_norm": 5.930924892425537,
      "learning_rate": 7.21180810367884e-05,
      "loss": 2.546651268005371,
      "memory(GiB)": 77.56,
      "step": 41330,
      "token_acc": 0.43870967741935485,
      "train_speed(iter/s)": 1.440207
    },
    {
      "epoch": 1.7709181269011611,
      "grad_norm": 5.790921688079834,
      "learning_rate": 7.211204532852302e-05,
      "loss": 2.636322021484375,
      "memory(GiB)": 77.56,
      "step": 41335,
      "token_acc": 0.44272445820433437,
      "train_speed(iter/s)": 1.440262
    },
    {
      "epoch": 1.77113234223041,
      "grad_norm": 5.1165947914123535,
      "learning_rate": 7.21060092196772e-05,
      "loss": 2.778389739990234,
      "memory(GiB)": 77.56,
      "step": 41340,
      "token_acc": 0.43023255813953487,
      "train_speed(iter/s)": 1.440224
    },
    {
      "epoch": 1.771346557559659,
      "grad_norm": 4.746934413909912,
      "learning_rate": 7.209997271036031e-05,
      "loss": 2.438471221923828,
      "memory(GiB)": 77.56,
      "step": 41345,
      "token_acc": 0.5182186234817814,
      "train_speed(iter/s)": 1.440279
    },
    {
      "epoch": 1.771560772888908,
      "grad_norm": 5.822700500488281,
      "learning_rate": 7.209393580068167e-05,
      "loss": 2.619981575012207,
      "memory(GiB)": 77.56,
      "step": 41350,
      "token_acc": 0.4506172839506173,
      "train_speed(iter/s)": 1.440272
    },
    {
      "epoch": 1.7717749882181568,
      "grad_norm": 6.012607574462891,
      "learning_rate": 7.208789849075065e-05,
      "loss": 2.787109375,
      "memory(GiB)": 77.56,
      "step": 41355,
      "token_acc": 0.432258064516129,
      "train_speed(iter/s)": 1.440295
    },
    {
      "epoch": 1.7719892035474059,
      "grad_norm": 7.867917060852051,
      "learning_rate": 7.208186078067665e-05,
      "loss": 2.358205223083496,
      "memory(GiB)": 77.56,
      "step": 41360,
      "token_acc": 0.5064516129032258,
      "train_speed(iter/s)": 1.440278
    },
    {
      "epoch": 1.772203418876655,
      "grad_norm": 3.890453577041626,
      "learning_rate": 7.2075822670569e-05,
      "loss": 2.6740558624267576,
      "memory(GiB)": 77.56,
      "step": 41365,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.44031
    },
    {
      "epoch": 1.7724176342059037,
      "grad_norm": 4.7364020347595215,
      "learning_rate": 7.206978416053714e-05,
      "loss": 2.524542236328125,
      "memory(GiB)": 77.56,
      "step": 41370,
      "token_acc": 0.46825396825396826,
      "train_speed(iter/s)": 1.440328
    },
    {
      "epoch": 1.7726318495351527,
      "grad_norm": 5.364123821258545,
      "learning_rate": 7.206374525069041e-05,
      "loss": 2.6369821548461916,
      "memory(GiB)": 77.56,
      "step": 41375,
      "token_acc": 0.44086021505376344,
      "train_speed(iter/s)": 1.440343
    },
    {
      "epoch": 1.7728460648644018,
      "grad_norm": 4.983763694763184,
      "learning_rate": 7.205770594113826e-05,
      "loss": 2.583917808532715,
      "memory(GiB)": 77.56,
      "step": 41380,
      "token_acc": 0.43853820598006643,
      "train_speed(iter/s)": 1.440356
    },
    {
      "epoch": 1.7730602801936506,
      "grad_norm": 4.361377716064453,
      "learning_rate": 7.205166623199007e-05,
      "loss": 2.505217170715332,
      "memory(GiB)": 77.56,
      "step": 41385,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.44036
    },
    {
      "epoch": 1.7732744955228996,
      "grad_norm": 6.533573150634766,
      "learning_rate": 7.204562612335526e-05,
      "loss": 2.74563045501709,
      "memory(GiB)": 77.56,
      "step": 41390,
      "token_acc": 0.470404984423676,
      "train_speed(iter/s)": 1.440389
    },
    {
      "epoch": 1.7734887108521487,
      "grad_norm": 5.561290264129639,
      "learning_rate": 7.203958561534324e-05,
      "loss": 2.4945323944091795,
      "memory(GiB)": 77.56,
      "step": 41395,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.440428
    },
    {
      "epoch": 1.7737029261813975,
      "grad_norm": 6.526377201080322,
      "learning_rate": 7.203354470806348e-05,
      "loss": 2.537253189086914,
      "memory(GiB)": 77.56,
      "step": 41400,
      "token_acc": 0.5065502183406113,
      "train_speed(iter/s)": 1.440464
    },
    {
      "epoch": 1.7739171415106465,
      "grad_norm": 4.192866802215576,
      "learning_rate": 7.202750340162536e-05,
      "loss": 2.664411926269531,
      "memory(GiB)": 77.56,
      "step": 41405,
      "token_acc": 0.4786096256684492,
      "train_speed(iter/s)": 1.440498
    },
    {
      "epoch": 1.7741313568398955,
      "grad_norm": 9.008928298950195,
      "learning_rate": 7.202146169613835e-05,
      "loss": 2.7312032699584963,
      "memory(GiB)": 77.56,
      "step": 41410,
      "token_acc": 0.44816053511705684,
      "train_speed(iter/s)": 1.440502
    },
    {
      "epoch": 1.7743455721691443,
      "grad_norm": 4.682097911834717,
      "learning_rate": 7.201541959171191e-05,
      "loss": 2.8470258712768555,
      "memory(GiB)": 77.56,
      "step": 41415,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.440545
    },
    {
      "epoch": 1.7745597874983934,
      "grad_norm": 5.219699859619141,
      "learning_rate": 7.200937708845552e-05,
      "loss": 2.5159173965454102,
      "memory(GiB)": 77.56,
      "step": 41420,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.440569
    },
    {
      "epoch": 1.7747740028276424,
      "grad_norm": 5.969017505645752,
      "learning_rate": 7.200333418647859e-05,
      "loss": 2.7637874603271486,
      "memory(GiB)": 77.56,
      "step": 41425,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.440536
    },
    {
      "epoch": 1.7749882181568912,
      "grad_norm": 4.169270038604736,
      "learning_rate": 7.199729088589063e-05,
      "loss": 2.386057662963867,
      "memory(GiB)": 77.56,
      "step": 41430,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.440499
    },
    {
      "epoch": 1.7752024334861403,
      "grad_norm": 6.329133987426758,
      "learning_rate": 7.199124718680111e-05,
      "loss": 2.8182960510253907,
      "memory(GiB)": 77.56,
      "step": 41435,
      "token_acc": 0.4088050314465409,
      "train_speed(iter/s)": 1.440465
    },
    {
      "epoch": 1.7754166488153893,
      "grad_norm": 5.599954128265381,
      "learning_rate": 7.198520308931951e-05,
      "loss": 2.5343799591064453,
      "memory(GiB)": 77.56,
      "step": 41440,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.440493
    },
    {
      "epoch": 1.775630864144638,
      "grad_norm": 4.3669328689575195,
      "learning_rate": 7.197915859355534e-05,
      "loss": 2.615229606628418,
      "memory(GiB)": 77.56,
      "step": 41445,
      "token_acc": 0.44299674267100975,
      "train_speed(iter/s)": 1.440514
    },
    {
      "epoch": 1.7758450794738871,
      "grad_norm": 6.3242998123168945,
      "learning_rate": 7.197311369961809e-05,
      "loss": 2.278392219543457,
      "memory(GiB)": 77.56,
      "step": 41450,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.440549
    },
    {
      "epoch": 1.7760592948031362,
      "grad_norm": 5.017621040344238,
      "learning_rate": 7.196706840761727e-05,
      "loss": 2.62106876373291,
      "memory(GiB)": 77.56,
      "step": 41455,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.440545
    },
    {
      "epoch": 1.776273510132385,
      "grad_norm": 4.113110065460205,
      "learning_rate": 7.196102271766241e-05,
      "loss": 2.9667633056640623,
      "memory(GiB)": 77.56,
      "step": 41460,
      "token_acc": 0.40607734806629836,
      "train_speed(iter/s)": 1.440548
    },
    {
      "epoch": 1.776487725461634,
      "grad_norm": 6.209756374359131,
      "learning_rate": 7.1954976629863e-05,
      "loss": 2.7616878509521485,
      "memory(GiB)": 77.56,
      "step": 41465,
      "token_acc": 0.4497991967871486,
      "train_speed(iter/s)": 1.44056
    },
    {
      "epoch": 1.776701940790883,
      "grad_norm": 6.022635459899902,
      "learning_rate": 7.194893014432862e-05,
      "loss": 2.424945068359375,
      "memory(GiB)": 77.56,
      "step": 41470,
      "token_acc": 0.45244956772334294,
      "train_speed(iter/s)": 1.440586
    },
    {
      "epoch": 1.7769161561201319,
      "grad_norm": 5.682319164276123,
      "learning_rate": 7.194288326116876e-05,
      "loss": 2.392338180541992,
      "memory(GiB)": 77.56,
      "step": 41475,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.440625
    },
    {
      "epoch": 1.777130371449381,
      "grad_norm": 7.37538480758667,
      "learning_rate": 7.1936835980493e-05,
      "loss": 2.602642059326172,
      "memory(GiB)": 77.56,
      "step": 41480,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.440636
    },
    {
      "epoch": 1.77734458677863,
      "grad_norm": 5.151170253753662,
      "learning_rate": 7.193078830241087e-05,
      "loss": 2.957985496520996,
      "memory(GiB)": 77.56,
      "step": 41485,
      "token_acc": 0.4501510574018127,
      "train_speed(iter/s)": 1.440684
    },
    {
      "epoch": 1.7775588021078788,
      "grad_norm": 5.186006546020508,
      "learning_rate": 7.192474022703194e-05,
      "loss": 2.5356969833374023,
      "memory(GiB)": 77.56,
      "step": 41490,
      "token_acc": 0.519298245614035,
      "train_speed(iter/s)": 1.440696
    },
    {
      "epoch": 1.7777730174371278,
      "grad_norm": 4.7131571769714355,
      "learning_rate": 7.191869175446576e-05,
      "loss": 2.312872886657715,
      "memory(GiB)": 77.56,
      "step": 41495,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.440708
    },
    {
      "epoch": 1.7779872327663768,
      "grad_norm": 4.934205532073975,
      "learning_rate": 7.191264288482194e-05,
      "loss": 2.5455230712890624,
      "memory(GiB)": 77.56,
      "step": 41500,
      "token_acc": 0.47474747474747475,
      "train_speed(iter/s)": 1.440694
    },
    {
      "epoch": 1.7779872327663768,
      "eval_loss": 2.125981330871582,
      "eval_runtime": 14.0156,
      "eval_samples_per_second": 7.135,
      "eval_steps_per_second": 7.135,
      "eval_token_acc": 0.48404255319148937,
      "step": 41500
    },
    {
      "epoch": 1.7782014480956256,
      "grad_norm": 6.567460536956787,
      "learning_rate": 7.190659361821003e-05,
      "loss": 2.3286039352416994,
      "memory(GiB)": 77.56,
      "step": 41505,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.439977
    },
    {
      "epoch": 1.7784156634248747,
      "grad_norm": 5.059206962585449,
      "learning_rate": 7.190054395473962e-05,
      "loss": 2.5230005264282225,
      "memory(GiB)": 77.56,
      "step": 41510,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.439934
    },
    {
      "epoch": 1.7786298787541237,
      "grad_norm": 4.963253021240234,
      "learning_rate": 7.189449389452032e-05,
      "loss": 2.5037582397460936,
      "memory(GiB)": 77.56,
      "step": 41515,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.439951
    },
    {
      "epoch": 1.7788440940833725,
      "grad_norm": 3.9160563945770264,
      "learning_rate": 7.188844343766173e-05,
      "loss": 2.3822832107543945,
      "memory(GiB)": 77.56,
      "step": 41520,
      "token_acc": 0.5201612903225806,
      "train_speed(iter/s)": 1.439973
    },
    {
      "epoch": 1.7790583094126216,
      "grad_norm": 4.47765588760376,
      "learning_rate": 7.188239258427343e-05,
      "loss": 2.6160921096801757,
      "memory(GiB)": 77.56,
      "step": 41525,
      "token_acc": 0.43125,
      "train_speed(iter/s)": 1.439957
    },
    {
      "epoch": 1.7792725247418706,
      "grad_norm": 5.558020114898682,
      "learning_rate": 7.187634133446507e-05,
      "loss": 2.605325126647949,
      "memory(GiB)": 77.56,
      "step": 41530,
      "token_acc": 0.4598337950138504,
      "train_speed(iter/s)": 1.440001
    },
    {
      "epoch": 1.7794867400711194,
      "grad_norm": 4.489594459533691,
      "learning_rate": 7.18702896883463e-05,
      "loss": 2.896640396118164,
      "memory(GiB)": 77.56,
      "step": 41535,
      "token_acc": 0.37942122186495175,
      "train_speed(iter/s)": 1.439965
    },
    {
      "epoch": 1.7797009554003684,
      "grad_norm": 6.1399407386779785,
      "learning_rate": 7.186423764602668e-05,
      "loss": 2.248468208312988,
      "memory(GiB)": 77.56,
      "step": 41540,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.439997
    },
    {
      "epoch": 1.7799151707296175,
      "grad_norm": 5.654233455657959,
      "learning_rate": 7.185818520761589e-05,
      "loss": 2.2671958923339846,
      "memory(GiB)": 77.56,
      "step": 41545,
      "token_acc": 0.4979757085020243,
      "train_speed(iter/s)": 1.440014
    },
    {
      "epoch": 1.7801293860588663,
      "grad_norm": 5.686004638671875,
      "learning_rate": 7.185213237322358e-05,
      "loss": 2.4533288955688475,
      "memory(GiB)": 77.56,
      "step": 41550,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.440038
    },
    {
      "epoch": 1.7803436013881153,
      "grad_norm": 4.564158916473389,
      "learning_rate": 7.184607914295937e-05,
      "loss": 2.3960342407226562,
      "memory(GiB)": 77.56,
      "step": 41555,
      "token_acc": 0.45955882352941174,
      "train_speed(iter/s)": 1.440092
    },
    {
      "epoch": 1.7805578167173644,
      "grad_norm": 4.506593227386475,
      "learning_rate": 7.184002551693296e-05,
      "loss": 2.554522895812988,
      "memory(GiB)": 77.56,
      "step": 41560,
      "token_acc": 0.4738955823293173,
      "train_speed(iter/s)": 1.440085
    },
    {
      "epoch": 1.7807720320466132,
      "grad_norm": 4.5854058265686035,
      "learning_rate": 7.183397149525401e-05,
      "loss": 2.3290922164916994,
      "memory(GiB)": 77.56,
      "step": 41565,
      "token_acc": 0.5130111524163569,
      "train_speed(iter/s)": 1.440122
    },
    {
      "epoch": 1.7809862473758622,
      "grad_norm": 5.484707832336426,
      "learning_rate": 7.182791707803216e-05,
      "loss": 2.6286540985107423,
      "memory(GiB)": 77.56,
      "step": 41570,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.440149
    },
    {
      "epoch": 1.7812004627051112,
      "grad_norm": 4.956236839294434,
      "learning_rate": 7.182186226537714e-05,
      "loss": 2.6242177963256834,
      "memory(GiB)": 77.56,
      "step": 41575,
      "token_acc": 0.47924528301886793,
      "train_speed(iter/s)": 1.440163
    },
    {
      "epoch": 1.78141467803436,
      "grad_norm": 5.367851257324219,
      "learning_rate": 7.18158070573986e-05,
      "loss": 2.675528717041016,
      "memory(GiB)": 77.56,
      "step": 41580,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.440154
    },
    {
      "epoch": 1.781628893363609,
      "grad_norm": 5.175403118133545,
      "learning_rate": 7.180975145420625e-05,
      "loss": 2.335260581970215,
      "memory(GiB)": 77.56,
      "step": 41585,
      "token_acc": 0.5040983606557377,
      "train_speed(iter/s)": 1.4402
    },
    {
      "epoch": 1.7818431086928581,
      "grad_norm": 4.378537178039551,
      "learning_rate": 7.180369545590981e-05,
      "loss": 2.374713134765625,
      "memory(GiB)": 77.56,
      "step": 41590,
      "token_acc": 0.5245398773006135,
      "train_speed(iter/s)": 1.440215
    },
    {
      "epoch": 1.782057324022107,
      "grad_norm": 4.559616565704346,
      "learning_rate": 7.179763906261896e-05,
      "loss": 2.3027820587158203,
      "memory(GiB)": 77.56,
      "step": 41595,
      "token_acc": 0.4714828897338403,
      "train_speed(iter/s)": 1.440221
    },
    {
      "epoch": 1.782271539351356,
      "grad_norm": 6.5058088302612305,
      "learning_rate": 7.179158227444343e-05,
      "loss": 2.5365753173828125,
      "memory(GiB)": 77.56,
      "step": 41600,
      "token_acc": 0.46688741721854304,
      "train_speed(iter/s)": 1.440225
    },
    {
      "epoch": 1.782485754680605,
      "grad_norm": 4.407220363616943,
      "learning_rate": 7.178552509149294e-05,
      "loss": 2.428637886047363,
      "memory(GiB)": 77.56,
      "step": 41605,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.440222
    },
    {
      "epoch": 1.7826999700098538,
      "grad_norm": 3.9212772846221924,
      "learning_rate": 7.177946751387723e-05,
      "loss": 2.4099021911621095,
      "memory(GiB)": 77.56,
      "step": 41610,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.440227
    },
    {
      "epoch": 1.7829141853391028,
      "grad_norm": 4.980103015899658,
      "learning_rate": 7.177340954170604e-05,
      "loss": 2.5643218994140624,
      "memory(GiB)": 77.56,
      "step": 41615,
      "token_acc": 0.47843137254901963,
      "train_speed(iter/s)": 1.440228
    },
    {
      "epoch": 1.7831284006683519,
      "grad_norm": 4.563261985778809,
      "learning_rate": 7.176735117508911e-05,
      "loss": 2.676917839050293,
      "memory(GiB)": 77.56,
      "step": 41620,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.440258
    },
    {
      "epoch": 1.7833426159976007,
      "grad_norm": 4.64477014541626,
      "learning_rate": 7.176129241413619e-05,
      "loss": 2.514088439941406,
      "memory(GiB)": 77.56,
      "step": 41625,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.440298
    },
    {
      "epoch": 1.7835568313268497,
      "grad_norm": 7.422196865081787,
      "learning_rate": 7.175523325895705e-05,
      "loss": 2.656286430358887,
      "memory(GiB)": 77.56,
      "step": 41630,
      "token_acc": 0.4172661870503597,
      "train_speed(iter/s)": 1.440277
    },
    {
      "epoch": 1.7837710466560988,
      "grad_norm": 6.116430282592773,
      "learning_rate": 7.174917370966145e-05,
      "loss": 2.299342727661133,
      "memory(GiB)": 77.56,
      "step": 41635,
      "token_acc": 0.5019607843137255,
      "train_speed(iter/s)": 1.440262
    },
    {
      "epoch": 1.7839852619853476,
      "grad_norm": 4.1989216804504395,
      "learning_rate": 7.174311376635916e-05,
      "loss": 2.3616992950439455,
      "memory(GiB)": 77.56,
      "step": 41640,
      "token_acc": 0.5207667731629393,
      "train_speed(iter/s)": 1.44022
    },
    {
      "epoch": 1.7841994773145966,
      "grad_norm": 4.709729194641113,
      "learning_rate": 7.173705342915998e-05,
      "loss": 2.6079713821411135,
      "memory(GiB)": 77.56,
      "step": 41645,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.440177
    },
    {
      "epoch": 1.7844136926438456,
      "grad_norm": 5.738956451416016,
      "learning_rate": 7.173099269817368e-05,
      "loss": 2.477339172363281,
      "memory(GiB)": 77.56,
      "step": 41650,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.440173
    },
    {
      "epoch": 1.7846279079730945,
      "grad_norm": 5.236964225769043,
      "learning_rate": 7.172493157351006e-05,
      "loss": 2.8129898071289063,
      "memory(GiB)": 77.56,
      "step": 41655,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.440219
    },
    {
      "epoch": 1.7848421233023435,
      "grad_norm": 4.651280879974365,
      "learning_rate": 7.171887005527893e-05,
      "loss": 2.2527658462524416,
      "memory(GiB)": 77.56,
      "step": 41660,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.440241
    },
    {
      "epoch": 1.7850563386315925,
      "grad_norm": 10.790109634399414,
      "learning_rate": 7.17128081435901e-05,
      "loss": 2.887628746032715,
      "memory(GiB)": 77.56,
      "step": 41665,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.440296
    },
    {
      "epoch": 1.7852705539608413,
      "grad_norm": 3.976048231124878,
      "learning_rate": 7.170674583855335e-05,
      "loss": 2.5121671676635744,
      "memory(GiB)": 77.56,
      "step": 41670,
      "token_acc": 0.4620253164556962,
      "train_speed(iter/s)": 1.440316
    },
    {
      "epoch": 1.7854847692900904,
      "grad_norm": 4.391224384307861,
      "learning_rate": 7.170068314027855e-05,
      "loss": 2.365525817871094,
      "memory(GiB)": 77.56,
      "step": 41675,
      "token_acc": 0.49390243902439024,
      "train_speed(iter/s)": 1.440309
    },
    {
      "epoch": 1.7856989846193394,
      "grad_norm": 4.24528169631958,
      "learning_rate": 7.169462004887555e-05,
      "loss": 2.632581329345703,
      "memory(GiB)": 77.56,
      "step": 41680,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.440329
    },
    {
      "epoch": 1.7859131999485882,
      "grad_norm": 4.256743907928467,
      "learning_rate": 7.168855656445412e-05,
      "loss": 2.4761331558227537,
      "memory(GiB)": 77.56,
      "step": 41685,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.440352
    },
    {
      "epoch": 1.7861274152778372,
      "grad_norm": 4.573760509490967,
      "learning_rate": 7.168249268712415e-05,
      "loss": 2.4129867553710938,
      "memory(GiB)": 77.56,
      "step": 41690,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.440385
    },
    {
      "epoch": 1.7863416306070863,
      "grad_norm": 4.989323139190674,
      "learning_rate": 7.16764284169955e-05,
      "loss": 2.4528703689575195,
      "memory(GiB)": 77.56,
      "step": 41695,
      "token_acc": 0.470404984423676,
      "train_speed(iter/s)": 1.44042
    },
    {
      "epoch": 1.786555845936335,
      "grad_norm": 5.008612632751465,
      "learning_rate": 7.167036375417801e-05,
      "loss": 2.6893356323242186,
      "memory(GiB)": 77.56,
      "step": 41700,
      "token_acc": 0.44370860927152317,
      "train_speed(iter/s)": 1.440422
    },
    {
      "epoch": 1.7867700612655841,
      "grad_norm": 5.087997913360596,
      "learning_rate": 7.166429869878154e-05,
      "loss": 2.5272586822509764,
      "memory(GiB)": 77.56,
      "step": 41705,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.440457
    },
    {
      "epoch": 1.7869842765948332,
      "grad_norm": 4.831918239593506,
      "learning_rate": 7.165823325091599e-05,
      "loss": 2.5077728271484374,
      "memory(GiB)": 77.56,
      "step": 41710,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.440468
    },
    {
      "epoch": 1.787198491924082,
      "grad_norm": 4.737553119659424,
      "learning_rate": 7.165216741069122e-05,
      "loss": 2.6069902420043944,
      "memory(GiB)": 77.56,
      "step": 41715,
      "token_acc": 0.4602076124567474,
      "train_speed(iter/s)": 1.44048
    },
    {
      "epoch": 1.787412707253331,
      "grad_norm": 6.522765159606934,
      "learning_rate": 7.164610117821713e-05,
      "loss": 2.6605192184448243,
      "memory(GiB)": 77.56,
      "step": 41720,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.440524
    },
    {
      "epoch": 1.78762692258258,
      "grad_norm": 11.286214828491211,
      "learning_rate": 7.16400345536036e-05,
      "loss": 2.541108512878418,
      "memory(GiB)": 77.56,
      "step": 41725,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.440519
    },
    {
      "epoch": 1.7878411379118289,
      "grad_norm": 5.008591651916504,
      "learning_rate": 7.163396753696057e-05,
      "loss": 2.5759857177734373,
      "memory(GiB)": 77.56,
      "step": 41730,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.440478
    },
    {
      "epoch": 1.788055353241078,
      "grad_norm": 5.311080455780029,
      "learning_rate": 7.162790012839791e-05,
      "loss": 2.367471694946289,
      "memory(GiB)": 77.56,
      "step": 41735,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.440469
    },
    {
      "epoch": 1.788269568570327,
      "grad_norm": 4.494590759277344,
      "learning_rate": 7.162183232802554e-05,
      "loss": 2.9353370666503906,
      "memory(GiB)": 77.56,
      "step": 41740,
      "token_acc": 0.38622754491017963,
      "train_speed(iter/s)": 1.440453
    },
    {
      "epoch": 1.7884837838995757,
      "grad_norm": 4.419131278991699,
      "learning_rate": 7.161576413595339e-05,
      "loss": 2.0296464920043946,
      "memory(GiB)": 77.56,
      "step": 41745,
      "token_acc": 0.5714285714285714,
      "train_speed(iter/s)": 1.44043
    },
    {
      "epoch": 1.7886979992288248,
      "grad_norm": 4.305533409118652,
      "learning_rate": 7.160969555229142e-05,
      "loss": 2.7163219451904297,
      "memory(GiB)": 77.56,
      "step": 41750,
      "token_acc": 0.44376899696048633,
      "train_speed(iter/s)": 1.440439
    },
    {
      "epoch": 1.7889122145580738,
      "grad_norm": 7.3529181480407715,
      "learning_rate": 7.160362657714953e-05,
      "loss": 2.3863460540771486,
      "memory(GiB)": 77.56,
      "step": 41755,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.4404
    },
    {
      "epoch": 1.7891264298873226,
      "grad_norm": 5.84220027923584,
      "learning_rate": 7.159755721063768e-05,
      "loss": 2.4685157775878905,
      "memory(GiB)": 77.56,
      "step": 41760,
      "token_acc": 0.48639455782312924,
      "train_speed(iter/s)": 1.440374
    },
    {
      "epoch": 1.7893406452165717,
      "grad_norm": 5.567548751831055,
      "learning_rate": 7.159148745286582e-05,
      "loss": 2.1888689041137694,
      "memory(GiB)": 77.56,
      "step": 41765,
      "token_acc": 0.5458515283842795,
      "train_speed(iter/s)": 1.440381
    },
    {
      "epoch": 1.7895548605458207,
      "grad_norm": 4.820566177368164,
      "learning_rate": 7.158541730394391e-05,
      "loss": 2.381830596923828,
      "memory(GiB)": 77.56,
      "step": 41770,
      "token_acc": 0.46616541353383456,
      "train_speed(iter/s)": 1.440405
    },
    {
      "epoch": 1.7897690758750695,
      "grad_norm": 4.206056594848633,
      "learning_rate": 7.157934676398192e-05,
      "loss": 2.5655515670776365,
      "memory(GiB)": 77.56,
      "step": 41775,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.440404
    },
    {
      "epoch": 1.7899832912043185,
      "grad_norm": 8.237789154052734,
      "learning_rate": 7.157327583308981e-05,
      "loss": 2.410350227355957,
      "memory(GiB)": 77.56,
      "step": 41780,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.44041
    },
    {
      "epoch": 1.7901975065335676,
      "grad_norm": 5.038957118988037,
      "learning_rate": 7.156720451137759e-05,
      "loss": 2.3720062255859373,
      "memory(GiB)": 77.56,
      "step": 41785,
      "token_acc": 0.4854014598540146,
      "train_speed(iter/s)": 1.440394
    },
    {
      "epoch": 1.7904117218628164,
      "grad_norm": 6.162631511688232,
      "learning_rate": 7.156113279895522e-05,
      "loss": 2.8040966033935546,
      "memory(GiB)": 77.56,
      "step": 41790,
      "token_acc": 0.43944636678200694,
      "train_speed(iter/s)": 1.440451
    },
    {
      "epoch": 1.7906259371920654,
      "grad_norm": 4.644979000091553,
      "learning_rate": 7.15550606959327e-05,
      "loss": 2.8054466247558594,
      "memory(GiB)": 77.56,
      "step": 41795,
      "token_acc": 0.43962848297213625,
      "train_speed(iter/s)": 1.440484
    },
    {
      "epoch": 1.7908401525213145,
      "grad_norm": 4.512179851531982,
      "learning_rate": 7.154898820242003e-05,
      "loss": 2.775253486633301,
      "memory(GiB)": 77.56,
      "step": 41800,
      "token_acc": 0.436426116838488,
      "train_speed(iter/s)": 1.44051
    },
    {
      "epoch": 1.7910543678505633,
      "grad_norm": 8.885587692260742,
      "learning_rate": 7.154291531852723e-05,
      "loss": 2.411014938354492,
      "memory(GiB)": 77.56,
      "step": 41805,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.44056
    },
    {
      "epoch": 1.7912685831798123,
      "grad_norm": 5.238372802734375,
      "learning_rate": 7.153684204436433e-05,
      "loss": 2.8068750381469725,
      "memory(GiB)": 77.56,
      "step": 41810,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.440613
    },
    {
      "epoch": 1.7914827985090613,
      "grad_norm": 6.922854423522949,
      "learning_rate": 7.153076838004129e-05,
      "loss": 2.2262115478515625,
      "memory(GiB)": 77.56,
      "step": 41815,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.440606
    },
    {
      "epoch": 1.7916970138383101,
      "grad_norm": 4.267703056335449,
      "learning_rate": 7.152469432566822e-05,
      "loss": 2.6742364883422853,
      "memory(GiB)": 77.56,
      "step": 41820,
      "token_acc": 0.43909348441926344,
      "train_speed(iter/s)": 1.440627
    },
    {
      "epoch": 1.7919112291675592,
      "grad_norm": 5.662759780883789,
      "learning_rate": 7.151861988135511e-05,
      "loss": 2.392784309387207,
      "memory(GiB)": 77.56,
      "step": 41825,
      "token_acc": 0.5107692307692308,
      "train_speed(iter/s)": 1.440655
    },
    {
      "epoch": 1.7921254444968082,
      "grad_norm": 5.835724353790283,
      "learning_rate": 7.151254504721201e-05,
      "loss": 2.859352684020996,
      "memory(GiB)": 77.56,
      "step": 41830,
      "token_acc": 0.40264026402640263,
      "train_speed(iter/s)": 1.440679
    },
    {
      "epoch": 1.792339659826057,
      "grad_norm": 6.028143882751465,
      "learning_rate": 7.150646982334897e-05,
      "loss": 2.3542957305908203,
      "memory(GiB)": 77.56,
      "step": 41835,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.440731
    },
    {
      "epoch": 1.792553875155306,
      "grad_norm": 3.758532762527466,
      "learning_rate": 7.150039420987606e-05,
      "loss": 2.462593650817871,
      "memory(GiB)": 77.56,
      "step": 41840,
      "token_acc": 0.48059701492537316,
      "train_speed(iter/s)": 1.440762
    },
    {
      "epoch": 1.792768090484555,
      "grad_norm": 5.808251857757568,
      "learning_rate": 7.149431820690335e-05,
      "loss": 2.4858678817749023,
      "memory(GiB)": 77.56,
      "step": 41845,
      "token_acc": 0.5134099616858238,
      "train_speed(iter/s)": 1.440738
    },
    {
      "epoch": 1.792982305813804,
      "grad_norm": 4.783618927001953,
      "learning_rate": 7.14882418145409e-05,
      "loss": 2.5948272705078126,
      "memory(GiB)": 77.56,
      "step": 41850,
      "token_acc": 0.47398843930635837,
      "train_speed(iter/s)": 1.440779
    },
    {
      "epoch": 1.793196521143053,
      "grad_norm": 5.319774627685547,
      "learning_rate": 7.148216503289878e-05,
      "loss": 2.621656036376953,
      "memory(GiB)": 77.56,
      "step": 41855,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440815
    },
    {
      "epoch": 1.793410736472302,
      "grad_norm": 5.4388298988342285,
      "learning_rate": 7.147608786208709e-05,
      "loss": 2.6802688598632813,
      "memory(GiB)": 77.56,
      "step": 41860,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.440878
    },
    {
      "epoch": 1.7936249518015508,
      "grad_norm": 4.757205486297607,
      "learning_rate": 7.147001030221594e-05,
      "loss": 2.3458911895751955,
      "memory(GiB)": 77.56,
      "step": 41865,
      "token_acc": 0.47038327526132406,
      "train_speed(iter/s)": 1.440917
    },
    {
      "epoch": 1.7938391671307998,
      "grad_norm": 4.485103130340576,
      "learning_rate": 7.146393235339539e-05,
      "loss": 2.5630115509033202,
      "memory(GiB)": 77.56,
      "step": 41870,
      "token_acc": 0.4391891891891892,
      "train_speed(iter/s)": 1.440913
    },
    {
      "epoch": 1.7940533824600489,
      "grad_norm": 4.268906116485596,
      "learning_rate": 7.145785401573559e-05,
      "loss": 2.498132514953613,
      "memory(GiB)": 77.56,
      "step": 41875,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.440871
    },
    {
      "epoch": 1.7942675977892977,
      "grad_norm": 4.961365699768066,
      "learning_rate": 7.145177528934663e-05,
      "loss": 2.7005792617797852,
      "memory(GiB)": 77.56,
      "step": 41880,
      "token_acc": 0.4246153846153846,
      "train_speed(iter/s)": 1.440893
    },
    {
      "epoch": 1.7944818131185467,
      "grad_norm": 4.200442314147949,
      "learning_rate": 7.144569617433863e-05,
      "loss": 2.4139476776123048,
      "memory(GiB)": 77.56,
      "step": 41885,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.440933
    },
    {
      "epoch": 1.7946960284477957,
      "grad_norm": 4.790828704833984,
      "learning_rate": 7.143961667082173e-05,
      "loss": 2.6316465377807616,
      "memory(GiB)": 77.56,
      "step": 41890,
      "token_acc": 0.46835443037974683,
      "train_speed(iter/s)": 1.440961
    },
    {
      "epoch": 1.7949102437770446,
      "grad_norm": 5.2089128494262695,
      "learning_rate": 7.143353677890607e-05,
      "loss": 2.5275337219238283,
      "memory(GiB)": 77.56,
      "step": 41895,
      "token_acc": 0.4392857142857143,
      "train_speed(iter/s)": 1.440965
    },
    {
      "epoch": 1.7951244591062938,
      "grad_norm": 5.263805389404297,
      "learning_rate": 7.142745649870177e-05,
      "loss": 2.684798240661621,
      "memory(GiB)": 77.56,
      "step": 41900,
      "token_acc": 0.41637010676156583,
      "train_speed(iter/s)": 1.441001
    },
    {
      "epoch": 1.7953386744355426,
      "grad_norm": 5.7521281242370605,
      "learning_rate": 7.142137583031901e-05,
      "loss": 2.4712223052978515,
      "memory(GiB)": 77.56,
      "step": 41905,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.44102
    },
    {
      "epoch": 1.7955528897647914,
      "grad_norm": 3.884425163269043,
      "learning_rate": 7.141529477386792e-05,
      "loss": 2.5200998306274416,
      "memory(GiB)": 77.56,
      "step": 41910,
      "token_acc": 0.4664179104477612,
      "train_speed(iter/s)": 1.44102
    },
    {
      "epoch": 1.7957671050940407,
      "grad_norm": 4.386074542999268,
      "learning_rate": 7.140921332945868e-05,
      "loss": 2.3738903045654296,
      "memory(GiB)": 77.56,
      "step": 41915,
      "token_acc": 0.49193548387096775,
      "train_speed(iter/s)": 1.441015
    },
    {
      "epoch": 1.7959813204232895,
      "grad_norm": 4.672444820404053,
      "learning_rate": 7.140313149720145e-05,
      "loss": 2.617329216003418,
      "memory(GiB)": 77.56,
      "step": 41920,
      "token_acc": 0.4670487106017192,
      "train_speed(iter/s)": 1.441004
    },
    {
      "epoch": 1.7961955357525383,
      "grad_norm": 4.702471733093262,
      "learning_rate": 7.139704927720644e-05,
      "loss": 2.57045841217041,
      "memory(GiB)": 77.56,
      "step": 41925,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.441046
    },
    {
      "epoch": 1.7964097510817876,
      "grad_norm": 4.501942157745361,
      "learning_rate": 7.139096666958378e-05,
      "loss": 2.451796531677246,
      "memory(GiB)": 77.56,
      "step": 41930,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.441074
    },
    {
      "epoch": 1.7966239664110364,
      "grad_norm": 3.9743943214416504,
      "learning_rate": 7.13848836744437e-05,
      "loss": 2.465041160583496,
      "memory(GiB)": 77.56,
      "step": 41935,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.441066
    },
    {
      "epoch": 1.7968381817402852,
      "grad_norm": 6.735315799713135,
      "learning_rate": 7.137880029189641e-05,
      "loss": 2.5690181732177733,
      "memory(GiB)": 77.56,
      "step": 41940,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.441031
    },
    {
      "epoch": 1.7970523970695345,
      "grad_norm": 4.932923793792725,
      "learning_rate": 7.137271652205208e-05,
      "loss": 2.447648811340332,
      "memory(GiB)": 77.56,
      "step": 41945,
      "token_acc": 0.4743083003952569,
      "train_speed(iter/s)": 1.441081
    },
    {
      "epoch": 1.7972666123987833,
      "grad_norm": 6.011246681213379,
      "learning_rate": 7.136663236502095e-05,
      "loss": 2.8310478210449217,
      "memory(GiB)": 77.56,
      "step": 41950,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.441048
    },
    {
      "epoch": 1.797480827728032,
      "grad_norm": 4.117939472198486,
      "learning_rate": 7.136054782091322e-05,
      "loss": 2.1531585693359374,
      "memory(GiB)": 77.56,
      "step": 41955,
      "token_acc": 0.5574468085106383,
      "train_speed(iter/s)": 1.44108
    },
    {
      "epoch": 1.7976950430572813,
      "grad_norm": 4.647158145904541,
      "learning_rate": 7.135446288983914e-05,
      "loss": 2.5094648361206056,
      "memory(GiB)": 77.56,
      "step": 41960,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.441053
    },
    {
      "epoch": 1.7979092583865302,
      "grad_norm": 4.708792686462402,
      "learning_rate": 7.134837757190891e-05,
      "loss": 2.7043376922607423,
      "memory(GiB)": 77.56,
      "step": 41965,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.441059
    },
    {
      "epoch": 1.798123473715779,
      "grad_norm": 4.866854190826416,
      "learning_rate": 7.134229186723282e-05,
      "loss": 2.5221385955810547,
      "memory(GiB)": 77.56,
      "step": 41970,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.44096
    },
    {
      "epoch": 1.7983376890450282,
      "grad_norm": 4.450485706329346,
      "learning_rate": 7.133620577592108e-05,
      "loss": 2.3979970932006838,
      "memory(GiB)": 77.56,
      "step": 41975,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.440947
    },
    {
      "epoch": 1.798551904374277,
      "grad_norm": 5.129952907562256,
      "learning_rate": 7.133011929808398e-05,
      "loss": 2.607725143432617,
      "memory(GiB)": 77.56,
      "step": 41980,
      "token_acc": 0.4939759036144578,
      "train_speed(iter/s)": 1.440961
    },
    {
      "epoch": 1.7987661197035258,
      "grad_norm": 4.724880695343018,
      "learning_rate": 7.132403243383173e-05,
      "loss": 2.680074691772461,
      "memory(GiB)": 77.56,
      "step": 41985,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440967
    },
    {
      "epoch": 1.798980335032775,
      "grad_norm": 5.4733452796936035,
      "learning_rate": 7.131794518327463e-05,
      "loss": 2.4858415603637694,
      "memory(GiB)": 77.56,
      "step": 41990,
      "token_acc": 0.4697986577181208,
      "train_speed(iter/s)": 1.440939
    },
    {
      "epoch": 1.799194550362024,
      "grad_norm": 5.619987487792969,
      "learning_rate": 7.131185754652299e-05,
      "loss": 2.543406677246094,
      "memory(GiB)": 77.56,
      "step": 41995,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440909
    },
    {
      "epoch": 1.7994087656912727,
      "grad_norm": 4.996868133544922,
      "learning_rate": 7.130576952368701e-05,
      "loss": 2.8067474365234375,
      "memory(GiB)": 77.56,
      "step": 42000,
      "token_acc": 0.4160839160839161,
      "train_speed(iter/s)": 1.440903
    },
    {
      "epoch": 1.7994087656912727,
      "eval_loss": 2.3193671703338623,
      "eval_runtime": 14.288,
      "eval_samples_per_second": 6.999,
      "eval_steps_per_second": 6.999,
      "eval_token_acc": 0.45690834473324216,
      "step": 42000
    },
    {
      "epoch": 1.799622981020522,
      "grad_norm": 6.263830661773682,
      "learning_rate": 7.129968111487706e-05,
      "loss": 2.1180793762207033,
      "memory(GiB)": 77.56,
      "step": 42005,
      "token_acc": 0.46576663452266154,
      "train_speed(iter/s)": 1.440158
    },
    {
      "epoch": 1.7998371963497708,
      "grad_norm": 4.130348205566406,
      "learning_rate": 7.12935923202034e-05,
      "loss": 2.449118423461914,
      "memory(GiB)": 77.56,
      "step": 42010,
      "token_acc": 0.5167785234899329,
      "train_speed(iter/s)": 1.440176
    },
    {
      "epoch": 1.8000514116790196,
      "grad_norm": 14.074151039123535,
      "learning_rate": 7.128750313977633e-05,
      "loss": 2.5615009307861327,
      "memory(GiB)": 77.56,
      "step": 42015,
      "token_acc": 0.44805194805194803,
      "train_speed(iter/s)": 1.440206
    },
    {
      "epoch": 1.8002656270082689,
      "grad_norm": 4.594374656677246,
      "learning_rate": 7.128141357370618e-05,
      "loss": 2.6220375061035157,
      "memory(GiB)": 77.56,
      "step": 42020,
      "token_acc": 0.47076023391812866,
      "train_speed(iter/s)": 1.440201
    },
    {
      "epoch": 1.8004798423375177,
      "grad_norm": 4.808896541595459,
      "learning_rate": 7.127532362210327e-05,
      "loss": 2.4455310821533205,
      "memory(GiB)": 77.56,
      "step": 42025,
      "token_acc": 0.48417721518987344,
      "train_speed(iter/s)": 1.440207
    },
    {
      "epoch": 1.8006940576667665,
      "grad_norm": 6.303009510040283,
      "learning_rate": 7.12692332850779e-05,
      "loss": 2.5016260147094727,
      "memory(GiB)": 77.56,
      "step": 42030,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.44025
    },
    {
      "epoch": 1.8009082729960157,
      "grad_norm": 5.437148094177246,
      "learning_rate": 7.126314256274042e-05,
      "loss": 2.4528053283691404,
      "memory(GiB)": 77.56,
      "step": 42035,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.440278
    },
    {
      "epoch": 1.8011224883252646,
      "grad_norm": 4.603950023651123,
      "learning_rate": 7.125705145520117e-05,
      "loss": 2.4632890701293944,
      "memory(GiB)": 77.56,
      "step": 42040,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.440308
    },
    {
      "epoch": 1.8013367036545134,
      "grad_norm": 5.031585693359375,
      "learning_rate": 7.125095996257048e-05,
      "loss": 2.8365667343139647,
      "memory(GiB)": 77.56,
      "step": 42045,
      "token_acc": 0.39568345323741005,
      "train_speed(iter/s)": 1.440293
    },
    {
      "epoch": 1.8015509189837626,
      "grad_norm": 4.871204853057861,
      "learning_rate": 7.124486808495873e-05,
      "loss": 2.459613800048828,
      "memory(GiB)": 77.56,
      "step": 42050,
      "token_acc": 0.5109717868338558,
      "train_speed(iter/s)": 1.440257
    },
    {
      "epoch": 1.8017651343130114,
      "grad_norm": 4.5483012199401855,
      "learning_rate": 7.123877582247626e-05,
      "loss": 2.7137245178222655,
      "memory(GiB)": 77.56,
      "step": 42055,
      "token_acc": 0.4064516129032258,
      "train_speed(iter/s)": 1.44028
    },
    {
      "epoch": 1.8019793496422603,
      "grad_norm": 6.571033477783203,
      "learning_rate": 7.123268317523344e-05,
      "loss": 2.58399658203125,
      "memory(GiB)": 77.56,
      "step": 42060,
      "token_acc": 0.46987951807228917,
      "train_speed(iter/s)": 1.440281
    },
    {
      "epoch": 1.8021935649715095,
      "grad_norm": 5.01151180267334,
      "learning_rate": 7.122659014334065e-05,
      "loss": 2.4359710693359373,
      "memory(GiB)": 77.56,
      "step": 42065,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.440336
    },
    {
      "epoch": 1.8024077803007583,
      "grad_norm": 5.244607448577881,
      "learning_rate": 7.122049672690828e-05,
      "loss": 2.1736881256103517,
      "memory(GiB)": 77.56,
      "step": 42070,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.440384
    },
    {
      "epoch": 1.8026219956300071,
      "grad_norm": 4.934427738189697,
      "learning_rate": 7.121440292604667e-05,
      "loss": 2.603476333618164,
      "memory(GiB)": 77.56,
      "step": 42075,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440426
    },
    {
      "epoch": 1.8028362109592564,
      "grad_norm": 4.66689920425415,
      "learning_rate": 7.120830874086627e-05,
      "loss": 2.590445899963379,
      "memory(GiB)": 77.56,
      "step": 42080,
      "token_acc": 0.47924528301886793,
      "train_speed(iter/s)": 1.440455
    },
    {
      "epoch": 1.8030504262885052,
      "grad_norm": 4.61002779006958,
      "learning_rate": 7.120221417147747e-05,
      "loss": 2.3773166656494142,
      "memory(GiB)": 77.56,
      "step": 42085,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.440472
    },
    {
      "epoch": 1.803264641617754,
      "grad_norm": 4.537388801574707,
      "learning_rate": 7.119611921799065e-05,
      "loss": 2.6261821746826173,
      "memory(GiB)": 77.56,
      "step": 42090,
      "token_acc": 0.4523076923076923,
      "train_speed(iter/s)": 1.44046
    },
    {
      "epoch": 1.8034788569470033,
      "grad_norm": 4.249468803405762,
      "learning_rate": 7.119002388051627e-05,
      "loss": 2.7216285705566405,
      "memory(GiB)": 77.56,
      "step": 42095,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.440517
    },
    {
      "epoch": 1.803693072276252,
      "grad_norm": 7.35615348815918,
      "learning_rate": 7.118392815916472e-05,
      "loss": 2.5203794479370116,
      "memory(GiB)": 77.56,
      "step": 42100,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.440523
    },
    {
      "epoch": 1.803907287605501,
      "grad_norm": 5.2889509201049805,
      "learning_rate": 7.117783205404644e-05,
      "loss": 2.2091312408447266,
      "memory(GiB)": 77.56,
      "step": 42105,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.440566
    },
    {
      "epoch": 1.8041215029347502,
      "grad_norm": 4.824329853057861,
      "learning_rate": 7.117173556527187e-05,
      "loss": 2.512018013000488,
      "memory(GiB)": 77.56,
      "step": 42110,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.440585
    },
    {
      "epoch": 1.804335718263999,
      "grad_norm": 5.606828212738037,
      "learning_rate": 7.116563869295145e-05,
      "loss": 2.3335639953613283,
      "memory(GiB)": 77.56,
      "step": 42115,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440573
    },
    {
      "epoch": 1.8045499335932478,
      "grad_norm": 4.261765003204346,
      "learning_rate": 7.11595414371956e-05,
      "loss": 2.5213661193847656,
      "memory(GiB)": 77.56,
      "step": 42120,
      "token_acc": 0.4879518072289157,
      "train_speed(iter/s)": 1.440586
    },
    {
      "epoch": 1.804764148922497,
      "grad_norm": 3.8963773250579834,
      "learning_rate": 7.115344379811485e-05,
      "loss": 2.9007081985473633,
      "memory(GiB)": 77.56,
      "step": 42125,
      "token_acc": 0.4243243243243243,
      "train_speed(iter/s)": 1.440591
    },
    {
      "epoch": 1.8049783642517458,
      "grad_norm": 7.668780326843262,
      "learning_rate": 7.11473457758196e-05,
      "loss": 2.4790472030639648,
      "memory(GiB)": 77.56,
      "step": 42130,
      "token_acc": 0.4752851711026616,
      "train_speed(iter/s)": 1.440609
    },
    {
      "epoch": 1.8051925795809947,
      "grad_norm": 5.161706924438477,
      "learning_rate": 7.114124737042035e-05,
      "loss": 2.7120758056640626,
      "memory(GiB)": 77.56,
      "step": 42135,
      "token_acc": 0.42,
      "train_speed(iter/s)": 1.44061
    },
    {
      "epoch": 1.805406794910244,
      "grad_norm": 7.514952182769775,
      "learning_rate": 7.113514858202758e-05,
      "loss": 2.463833236694336,
      "memory(GiB)": 77.56,
      "step": 42140,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.440563
    },
    {
      "epoch": 1.8056210102394927,
      "grad_norm": 4.837766647338867,
      "learning_rate": 7.112904941075175e-05,
      "loss": 2.703973579406738,
      "memory(GiB)": 77.56,
      "step": 42145,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.440602
    },
    {
      "epoch": 1.8058352255687415,
      "grad_norm": 6.498230457305908,
      "learning_rate": 7.112294985670337e-05,
      "loss": 2.2992202758789064,
      "memory(GiB)": 77.56,
      "step": 42150,
      "token_acc": 0.522633744855967,
      "train_speed(iter/s)": 1.440602
    },
    {
      "epoch": 1.8060494408979908,
      "grad_norm": 3.9114623069763184,
      "learning_rate": 7.111684991999295e-05,
      "loss": 2.6470157623291017,
      "memory(GiB)": 77.56,
      "step": 42155,
      "token_acc": 0.4696485623003195,
      "train_speed(iter/s)": 1.440622
    },
    {
      "epoch": 1.8062636562272396,
      "grad_norm": 6.802840709686279,
      "learning_rate": 7.111074960073098e-05,
      "loss": 2.750494956970215,
      "memory(GiB)": 77.56,
      "step": 42160,
      "token_acc": 0.48014440433212996,
      "train_speed(iter/s)": 1.440627
    },
    {
      "epoch": 1.8064778715564884,
      "grad_norm": 4.773049831390381,
      "learning_rate": 7.110464889902796e-05,
      "loss": 2.2988460540771483,
      "memory(GiB)": 77.56,
      "step": 42165,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.440666
    },
    {
      "epoch": 1.8066920868857377,
      "grad_norm": 6.8202924728393555,
      "learning_rate": 7.109854781499446e-05,
      "loss": 2.542022132873535,
      "memory(GiB)": 77.56,
      "step": 42170,
      "token_acc": 0.47854785478547857,
      "train_speed(iter/s)": 1.440672
    },
    {
      "epoch": 1.8069063022149865,
      "grad_norm": 4.792782783508301,
      "learning_rate": 7.109244634874096e-05,
      "loss": 2.8129718780517576,
      "memory(GiB)": 77.56,
      "step": 42175,
      "token_acc": 0.4576923076923077,
      "train_speed(iter/s)": 1.440709
    },
    {
      "epoch": 1.8071205175442353,
      "grad_norm": 5.579444885253906,
      "learning_rate": 7.108634450037802e-05,
      "loss": 2.370299530029297,
      "memory(GiB)": 77.56,
      "step": 42180,
      "token_acc": 0.4980392156862745,
      "train_speed(iter/s)": 1.440714
    },
    {
      "epoch": 1.8073347328734846,
      "grad_norm": 4.456747055053711,
      "learning_rate": 7.108024227001615e-05,
      "loss": 2.6092477798461915,
      "memory(GiB)": 77.56,
      "step": 42185,
      "token_acc": 0.43389830508474575,
      "train_speed(iter/s)": 1.440728
    },
    {
      "epoch": 1.8075489482027334,
      "grad_norm": 5.074801921844482,
      "learning_rate": 7.10741396577659e-05,
      "loss": 2.525266647338867,
      "memory(GiB)": 77.56,
      "step": 42190,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.440734
    },
    {
      "epoch": 1.8077631635319822,
      "grad_norm": 5.257549285888672,
      "learning_rate": 7.106803666373787e-05,
      "loss": 2.597431182861328,
      "memory(GiB)": 77.56,
      "step": 42195,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.44073
    },
    {
      "epoch": 1.8079773788612314,
      "grad_norm": 6.119453430175781,
      "learning_rate": 7.10619332880426e-05,
      "loss": 2.4286149978637694,
      "memory(GiB)": 77.56,
      "step": 42200,
      "token_acc": 0.5105633802816901,
      "train_speed(iter/s)": 1.440764
    },
    {
      "epoch": 1.8081915941904803,
      "grad_norm": 5.287391662597656,
      "learning_rate": 7.105582953079063e-05,
      "loss": 2.454452133178711,
      "memory(GiB)": 77.56,
      "step": 42205,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.440738
    },
    {
      "epoch": 1.808405809519729,
      "grad_norm": 4.636159896850586,
      "learning_rate": 7.104972539209256e-05,
      "loss": 2.6267990112304687,
      "memory(GiB)": 77.56,
      "step": 42210,
      "token_acc": 0.4495114006514658,
      "train_speed(iter/s)": 1.440755
    },
    {
      "epoch": 1.8086200248489783,
      "grad_norm": 3.9708337783813477,
      "learning_rate": 7.104362087205898e-05,
      "loss": 2.505854606628418,
      "memory(GiB)": 77.56,
      "step": 42215,
      "token_acc": 0.47041420118343197,
      "train_speed(iter/s)": 1.440711
    },
    {
      "epoch": 1.8088342401782271,
      "grad_norm": 4.682599067687988,
      "learning_rate": 7.103751597080045e-05,
      "loss": 2.7869260787963865,
      "memory(GiB)": 77.56,
      "step": 42220,
      "token_acc": 0.4338235294117647,
      "train_speed(iter/s)": 1.440726
    },
    {
      "epoch": 1.809048455507476,
      "grad_norm": 5.472672462463379,
      "learning_rate": 7.103141068842759e-05,
      "loss": 2.3040191650390627,
      "memory(GiB)": 77.56,
      "step": 42225,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.44069
    },
    {
      "epoch": 1.8092626708367252,
      "grad_norm": 6.776150226593018,
      "learning_rate": 7.1025305025051e-05,
      "loss": 2.4268171310424806,
      "memory(GiB)": 77.56,
      "step": 42230,
      "token_acc": 0.5146443514644351,
      "train_speed(iter/s)": 1.440698
    },
    {
      "epoch": 1.809476886165974,
      "grad_norm": 4.946876525878906,
      "learning_rate": 7.101919898078128e-05,
      "loss": 2.7492752075195312,
      "memory(GiB)": 77.56,
      "step": 42235,
      "token_acc": 0.46686746987951805,
      "train_speed(iter/s)": 1.440678
    },
    {
      "epoch": 1.809691101495223,
      "grad_norm": 4.6621294021606445,
      "learning_rate": 7.101309255572905e-05,
      "loss": 2.518558311462402,
      "memory(GiB)": 77.56,
      "step": 42240,
      "token_acc": 0.4745222929936306,
      "train_speed(iter/s)": 1.440681
    },
    {
      "epoch": 1.809905316824472,
      "grad_norm": 6.335580825805664,
      "learning_rate": 7.100698575000496e-05,
      "loss": 2.7184579849243162,
      "memory(GiB)": 77.56,
      "step": 42245,
      "token_acc": 0.4452054794520548,
      "train_speed(iter/s)": 1.440694
    },
    {
      "epoch": 1.810119532153721,
      "grad_norm": 5.230987071990967,
      "learning_rate": 7.10008785637196e-05,
      "loss": 2.328630065917969,
      "memory(GiB)": 77.56,
      "step": 42250,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.440652
    },
    {
      "epoch": 1.81033374748297,
      "grad_norm": 6.371697425842285,
      "learning_rate": 7.099477099698361e-05,
      "loss": 2.582619285583496,
      "memory(GiB)": 77.56,
      "step": 42255,
      "token_acc": 0.45739910313901344,
      "train_speed(iter/s)": 1.440671
    },
    {
      "epoch": 1.810547962812219,
      "grad_norm": 8.896299362182617,
      "learning_rate": 7.098866304990767e-05,
      "loss": 2.2750127792358397,
      "memory(GiB)": 77.56,
      "step": 42260,
      "token_acc": 0.5152838427947598,
      "train_speed(iter/s)": 1.440688
    },
    {
      "epoch": 1.8107621781414678,
      "grad_norm": 4.576548099517822,
      "learning_rate": 7.09825547226024e-05,
      "loss": 2.2459184646606447,
      "memory(GiB)": 77.56,
      "step": 42265,
      "token_acc": 0.4715447154471545,
      "train_speed(iter/s)": 1.440688
    },
    {
      "epoch": 1.8109763934707168,
      "grad_norm": 5.128814220428467,
      "learning_rate": 7.097644601517848e-05,
      "loss": 2.376708984375,
      "memory(GiB)": 77.56,
      "step": 42270,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.440689
    },
    {
      "epoch": 1.8111906087999659,
      "grad_norm": 4.97381591796875,
      "learning_rate": 7.097033692774653e-05,
      "loss": 2.457427978515625,
      "memory(GiB)": 77.56,
      "step": 42275,
      "token_acc": 0.46774193548387094,
      "train_speed(iter/s)": 1.440707
    },
    {
      "epoch": 1.8114048241292147,
      "grad_norm": 4.62233829498291,
      "learning_rate": 7.09642274604173e-05,
      "loss": 2.306803512573242,
      "memory(GiB)": 77.56,
      "step": 42280,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440719
    },
    {
      "epoch": 1.8116190394584637,
      "grad_norm": 5.273478031158447,
      "learning_rate": 7.09581176133014e-05,
      "loss": 2.4258193969726562,
      "memory(GiB)": 77.56,
      "step": 42285,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.440749
    },
    {
      "epoch": 1.8118332547877127,
      "grad_norm": 6.072991371154785,
      "learning_rate": 7.095200738650953e-05,
      "loss": 2.5939872741699217,
      "memory(GiB)": 77.56,
      "step": 42290,
      "token_acc": 0.4342105263157895,
      "train_speed(iter/s)": 1.440716
    },
    {
      "epoch": 1.8120474701169615,
      "grad_norm": 5.440391540527344,
      "learning_rate": 7.094589678015242e-05,
      "loss": 2.393227195739746,
      "memory(GiB)": 77.56,
      "step": 42295,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.440729
    },
    {
      "epoch": 1.8122616854462106,
      "grad_norm": 5.05177640914917,
      "learning_rate": 7.093978579434072e-05,
      "loss": 2.6129196166992186,
      "memory(GiB)": 77.56,
      "step": 42300,
      "token_acc": 0.47634069400630913,
      "train_speed(iter/s)": 1.440736
    },
    {
      "epoch": 1.8124759007754596,
      "grad_norm": 4.825653553009033,
      "learning_rate": 7.093367442918515e-05,
      "loss": 2.6763320922851563,
      "memory(GiB)": 77.56,
      "step": 42305,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.440759
    },
    {
      "epoch": 1.8126901161047084,
      "grad_norm": 4.48107385635376,
      "learning_rate": 7.092756268479646e-05,
      "loss": 2.4609107971191406,
      "memory(GiB)": 77.56,
      "step": 42310,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.440765
    },
    {
      "epoch": 1.8129043314339575,
      "grad_norm": 4.72641134262085,
      "learning_rate": 7.092145056128532e-05,
      "loss": 2.511740493774414,
      "memory(GiB)": 77.56,
      "step": 42315,
      "token_acc": 0.4341317365269461,
      "train_speed(iter/s)": 1.440751
    },
    {
      "epoch": 1.8131185467632065,
      "grad_norm": 4.278848648071289,
      "learning_rate": 7.091533805876247e-05,
      "loss": 2.2098731994628906,
      "memory(GiB)": 77.56,
      "step": 42320,
      "token_acc": 0.553030303030303,
      "train_speed(iter/s)": 1.440725
    },
    {
      "epoch": 1.8133327620924553,
      "grad_norm": 5.661434650421143,
      "learning_rate": 7.090922517733867e-05,
      "loss": 2.554936981201172,
      "memory(GiB)": 77.56,
      "step": 42325,
      "token_acc": 0.424812030075188,
      "train_speed(iter/s)": 1.440728
    },
    {
      "epoch": 1.8135469774217043,
      "grad_norm": 5.949657440185547,
      "learning_rate": 7.090311191712464e-05,
      "loss": 2.510664367675781,
      "memory(GiB)": 77.56,
      "step": 42330,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.440738
    },
    {
      "epoch": 1.8137611927509534,
      "grad_norm": 6.50044584274292,
      "learning_rate": 7.08969982782311e-05,
      "loss": 2.4037689208984374,
      "memory(GiB)": 77.56,
      "step": 42335,
      "token_acc": 0.5363984674329502,
      "train_speed(iter/s)": 1.44077
    },
    {
      "epoch": 1.8139754080802022,
      "grad_norm": 5.013998031616211,
      "learning_rate": 7.089088426076885e-05,
      "loss": 2.6579734802246096,
      "memory(GiB)": 77.56,
      "step": 42340,
      "token_acc": 0.4725274725274725,
      "train_speed(iter/s)": 1.44077
    },
    {
      "epoch": 1.8141896234094512,
      "grad_norm": 4.483895778656006,
      "learning_rate": 7.088476986484865e-05,
      "loss": 2.281387710571289,
      "memory(GiB)": 77.56,
      "step": 42345,
      "token_acc": 0.5019305019305019,
      "train_speed(iter/s)": 1.440812
    },
    {
      "epoch": 1.8144038387387003,
      "grad_norm": 4.825186252593994,
      "learning_rate": 7.087865509058124e-05,
      "loss": 2.730850601196289,
      "memory(GiB)": 77.56,
      "step": 42350,
      "token_acc": 0.43288590604026844,
      "train_speed(iter/s)": 1.440823
    },
    {
      "epoch": 1.814618054067949,
      "grad_norm": 4.690332889556885,
      "learning_rate": 7.087253993807741e-05,
      "loss": 2.4748687744140625,
      "memory(GiB)": 77.56,
      "step": 42355,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.440795
    },
    {
      "epoch": 1.814832269397198,
      "grad_norm": 5.284695148468018,
      "learning_rate": 7.086642440744794e-05,
      "loss": 2.4146251678466797,
      "memory(GiB)": 77.56,
      "step": 42360,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.440824
    },
    {
      "epoch": 1.8150464847264471,
      "grad_norm": 5.020896911621094,
      "learning_rate": 7.08603084988036e-05,
      "loss": 2.4790555953979494,
      "memory(GiB)": 77.56,
      "step": 42365,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.440807
    },
    {
      "epoch": 1.815260700055696,
      "grad_norm": 5.374149322509766,
      "learning_rate": 7.085419221225522e-05,
      "loss": 2.7598379135131834,
      "memory(GiB)": 77.56,
      "step": 42370,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.440795
    },
    {
      "epoch": 1.815474915384945,
      "grad_norm": 4.4034600257873535,
      "learning_rate": 7.084807554791358e-05,
      "loss": 2.6706916809082033,
      "memory(GiB)": 77.56,
      "step": 42375,
      "token_acc": 0.428125,
      "train_speed(iter/s)": 1.440817
    },
    {
      "epoch": 1.815689130714194,
      "grad_norm": 3.9676151275634766,
      "learning_rate": 7.08419585058895e-05,
      "loss": 2.5571022033691406,
      "memory(GiB)": 77.56,
      "step": 42380,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.440856
    },
    {
      "epoch": 1.8159033460434428,
      "grad_norm": 5.746986389160156,
      "learning_rate": 7.083584108629379e-05,
      "loss": 2.4747262954711915,
      "memory(GiB)": 77.56,
      "step": 42385,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.440847
    },
    {
      "epoch": 1.8161175613726919,
      "grad_norm": 6.064831733703613,
      "learning_rate": 7.082972328923726e-05,
      "loss": 2.4341896057128904,
      "memory(GiB)": 77.56,
      "step": 42390,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.440861
    },
    {
      "epoch": 1.816331776701941,
      "grad_norm": 9.76562213897705,
      "learning_rate": 7.082360511483077e-05,
      "loss": 2.770310974121094,
      "memory(GiB)": 77.56,
      "step": 42395,
      "token_acc": 0.4227129337539432,
      "train_speed(iter/s)": 1.440855
    },
    {
      "epoch": 1.8165459920311897,
      "grad_norm": 5.372458457946777,
      "learning_rate": 7.081748656318514e-05,
      "loss": 2.2854185104370117,
      "memory(GiB)": 77.56,
      "step": 42400,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.440816
    },
    {
      "epoch": 1.8167602073604388,
      "grad_norm": 4.555784702301025,
      "learning_rate": 7.081136763441119e-05,
      "loss": 2.782064437866211,
      "memory(GiB)": 77.56,
      "step": 42405,
      "token_acc": 0.44660194174757284,
      "train_speed(iter/s)": 1.440793
    },
    {
      "epoch": 1.8169744226896878,
      "grad_norm": 4.1589155197143555,
      "learning_rate": 7.080524832861982e-05,
      "loss": 2.4405752182006837,
      "memory(GiB)": 77.56,
      "step": 42410,
      "token_acc": 0.5070422535211268,
      "train_speed(iter/s)": 1.440764
    },
    {
      "epoch": 1.8171886380189366,
      "grad_norm": 4.964533805847168,
      "learning_rate": 7.079912864592185e-05,
      "loss": 2.1579416275024412,
      "memory(GiB)": 77.56,
      "step": 42415,
      "token_acc": 0.5703703703703704,
      "train_speed(iter/s)": 1.440742
    },
    {
      "epoch": 1.8174028533481856,
      "grad_norm": 4.621743202209473,
      "learning_rate": 7.079300858642814e-05,
      "loss": 2.334096145629883,
      "memory(GiB)": 77.56,
      "step": 42420,
      "token_acc": 0.48659003831417624,
      "train_speed(iter/s)": 1.44073
    },
    {
      "epoch": 1.8176170686774347,
      "grad_norm": 4.793374061584473,
      "learning_rate": 7.078688815024959e-05,
      "loss": 2.5243562698364257,
      "memory(GiB)": 77.56,
      "step": 42425,
      "token_acc": 0.4472573839662447,
      "train_speed(iter/s)": 1.440763
    },
    {
      "epoch": 1.8178312840066835,
      "grad_norm": 6.596546173095703,
      "learning_rate": 7.078076733749706e-05,
      "loss": 2.651638412475586,
      "memory(GiB)": 77.56,
      "step": 42430,
      "token_acc": 0.45791245791245794,
      "train_speed(iter/s)": 1.440778
    },
    {
      "epoch": 1.8180454993359325,
      "grad_norm": 5.211883544921875,
      "learning_rate": 7.077464614828142e-05,
      "loss": 2.6267276763916017,
      "memory(GiB)": 77.56,
      "step": 42435,
      "token_acc": 0.46175637393767704,
      "train_speed(iter/s)": 1.440809
    },
    {
      "epoch": 1.8182597146651815,
      "grad_norm": 5.329939842224121,
      "learning_rate": 7.07685245827136e-05,
      "loss": 2.4424423217773437,
      "memory(GiB)": 77.56,
      "step": 42440,
      "token_acc": 0.4740740740740741,
      "train_speed(iter/s)": 1.440836
    },
    {
      "epoch": 1.8184739299944304,
      "grad_norm": 3.719423294067383,
      "learning_rate": 7.076240264090446e-05,
      "loss": 2.7285940170288088,
      "memory(GiB)": 77.56,
      "step": 42445,
      "token_acc": 0.4461538461538462,
      "train_speed(iter/s)": 1.440835
    },
    {
      "epoch": 1.8186881453236794,
      "grad_norm": 4.613279819488525,
      "learning_rate": 7.075628032296491e-05,
      "loss": 2.375505828857422,
      "memory(GiB)": 77.56,
      "step": 42450,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.440848
    },
    {
      "epoch": 1.8189023606529284,
      "grad_norm": 4.4954938888549805,
      "learning_rate": 7.07501576290059e-05,
      "loss": 2.356053924560547,
      "memory(GiB)": 77.56,
      "step": 42455,
      "token_acc": 0.5182186234817814,
      "train_speed(iter/s)": 1.440787
    },
    {
      "epoch": 1.8191165759821772,
      "grad_norm": 4.431429386138916,
      "learning_rate": 7.074403455913829e-05,
      "loss": 2.754829216003418,
      "memory(GiB)": 77.56,
      "step": 42460,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.440799
    },
    {
      "epoch": 1.8193307913114263,
      "grad_norm": 4.924394130706787,
      "learning_rate": 7.073791111347305e-05,
      "loss": 2.286001205444336,
      "memory(GiB)": 77.56,
      "step": 42465,
      "token_acc": 0.5287356321839081,
      "train_speed(iter/s)": 1.440832
    },
    {
      "epoch": 1.8195450066406753,
      "grad_norm": 4.5436859130859375,
      "learning_rate": 7.07317872921211e-05,
      "loss": 2.4529966354370116,
      "memory(GiB)": 77.56,
      "step": 42470,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.440838
    },
    {
      "epoch": 1.8197592219699241,
      "grad_norm": 5.282332420349121,
      "learning_rate": 7.072566309519338e-05,
      "loss": 2.5372215270996095,
      "memory(GiB)": 77.56,
      "step": 42475,
      "token_acc": 0.47346938775510206,
      "train_speed(iter/s)": 1.440849
    },
    {
      "epoch": 1.8199734372991732,
      "grad_norm": 5.669687271118164,
      "learning_rate": 7.071953852280081e-05,
      "loss": 2.5251893997192383,
      "memory(GiB)": 77.56,
      "step": 42480,
      "token_acc": 0.4625,
      "train_speed(iter/s)": 1.440871
    },
    {
      "epoch": 1.8201876526284222,
      "grad_norm": 4.748483180999756,
      "learning_rate": 7.071341357505437e-05,
      "loss": 2.6501842498779298,
      "memory(GiB)": 77.56,
      "step": 42485,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.440895
    },
    {
      "epoch": 1.820401867957671,
      "grad_norm": 4.6338114738464355,
      "learning_rate": 7.070728825206501e-05,
      "loss": 2.8831497192382813,
      "memory(GiB)": 77.56,
      "step": 42490,
      "token_acc": 0.445859872611465,
      "train_speed(iter/s)": 1.440922
    },
    {
      "epoch": 1.82061608328692,
      "grad_norm": 6.064342498779297,
      "learning_rate": 7.070116255394371e-05,
      "loss": 2.340175819396973,
      "memory(GiB)": 77.56,
      "step": 42495,
      "token_acc": 0.5032894736842105,
      "train_speed(iter/s)": 1.440962
    },
    {
      "epoch": 1.820830298616169,
      "grad_norm": 5.834266662597656,
      "learning_rate": 7.069503648080143e-05,
      "loss": 2.5555051803588866,
      "memory(GiB)": 77.56,
      "step": 42500,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.440978
    },
    {
      "epoch": 1.820830298616169,
      "eval_loss": 2.2241339683532715,
      "eval_runtime": 14.1675,
      "eval_samples_per_second": 7.058,
      "eval_steps_per_second": 7.058,
      "eval_token_acc": 0.47468354430379744,
      "step": 42500
    },
    {
      "epoch": 1.8210445139454179,
      "grad_norm": 6.313887596130371,
      "learning_rate": 7.068891003274915e-05,
      "loss": 2.353474807739258,
      "memory(GiB)": 77.56,
      "step": 42505,
      "token_acc": 0.4801812004530011,
      "train_speed(iter/s)": 1.440276
    },
    {
      "epoch": 1.821258729274667,
      "grad_norm": 5.908014297485352,
      "learning_rate": 7.068278320989785e-05,
      "loss": 2.248538017272949,
      "memory(GiB)": 77.56,
      "step": 42510,
      "token_acc": 0.5165289256198347,
      "train_speed(iter/s)": 1.440285
    },
    {
      "epoch": 1.821472944603916,
      "grad_norm": 4.993363857269287,
      "learning_rate": 7.067665601235853e-05,
      "loss": 2.569161033630371,
      "memory(GiB)": 77.56,
      "step": 42515,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.440284
    },
    {
      "epoch": 1.8216871599331648,
      "grad_norm": 5.7084221839904785,
      "learning_rate": 7.067052844024219e-05,
      "loss": 2.5540475845336914,
      "memory(GiB)": 77.56,
      "step": 42520,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.440283
    },
    {
      "epoch": 1.8219013752624138,
      "grad_norm": 4.787469863891602,
      "learning_rate": 7.066440049365983e-05,
      "loss": 2.291056823730469,
      "memory(GiB)": 77.56,
      "step": 42525,
      "token_acc": 0.5285714285714286,
      "train_speed(iter/s)": 1.44027
    },
    {
      "epoch": 1.8221155905916628,
      "grad_norm": 4.9950642585754395,
      "learning_rate": 7.06582721727225e-05,
      "loss": 2.5157814025878906,
      "memory(GiB)": 77.56,
      "step": 42530,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.440273
    },
    {
      "epoch": 1.8223298059209116,
      "grad_norm": 5.4237542152404785,
      "learning_rate": 7.065214347754115e-05,
      "loss": 2.4063838958740233,
      "memory(GiB)": 77.56,
      "step": 42535,
      "token_acc": 0.49429657794676807,
      "train_speed(iter/s)": 1.440227
    },
    {
      "epoch": 1.8225440212501607,
      "grad_norm": 6.043896198272705,
      "learning_rate": 7.064601440822688e-05,
      "loss": 2.2234670639038088,
      "memory(GiB)": 77.56,
      "step": 42540,
      "token_acc": 0.5152542372881356,
      "train_speed(iter/s)": 1.440256
    },
    {
      "epoch": 1.8227582365794097,
      "grad_norm": 5.785392761230469,
      "learning_rate": 7.063988496489067e-05,
      "loss": 2.357148551940918,
      "memory(GiB)": 77.56,
      "step": 42545,
      "token_acc": 0.5191082802547771,
      "train_speed(iter/s)": 1.440286
    },
    {
      "epoch": 1.8229724519086585,
      "grad_norm": 4.614426136016846,
      "learning_rate": 7.06337551476436e-05,
      "loss": 2.3619728088378906,
      "memory(GiB)": 77.56,
      "step": 42550,
      "token_acc": 0.5354609929078015,
      "train_speed(iter/s)": 1.440311
    },
    {
      "epoch": 1.8231866672379076,
      "grad_norm": 5.35098934173584,
      "learning_rate": 7.062762495659669e-05,
      "loss": 2.5314332962036135,
      "memory(GiB)": 77.56,
      "step": 42555,
      "token_acc": 0.46258503401360546,
      "train_speed(iter/s)": 1.440301
    },
    {
      "epoch": 1.8234008825671566,
      "grad_norm": 8.009761810302734,
      "learning_rate": 7.062149439186101e-05,
      "loss": 2.5030033111572267,
      "memory(GiB)": 77.56,
      "step": 42560,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.440293
    },
    {
      "epoch": 1.8236150978964054,
      "grad_norm": 4.992061614990234,
      "learning_rate": 7.06153634535476e-05,
      "loss": 2.441164016723633,
      "memory(GiB)": 77.56,
      "step": 42565,
      "token_acc": 0.4603658536585366,
      "train_speed(iter/s)": 1.440297
    },
    {
      "epoch": 1.8238293132256544,
      "grad_norm": 5.6058502197265625,
      "learning_rate": 7.060923214176756e-05,
      "loss": 2.9603830337524415,
      "memory(GiB)": 77.56,
      "step": 42570,
      "token_acc": 0.40764331210191085,
      "train_speed(iter/s)": 1.440257
    },
    {
      "epoch": 1.8240435285549035,
      "grad_norm": 4.6703643798828125,
      "learning_rate": 7.060310045663193e-05,
      "loss": 2.729035568237305,
      "memory(GiB)": 77.56,
      "step": 42575,
      "token_acc": 0.4329268292682927,
      "train_speed(iter/s)": 1.440283
    },
    {
      "epoch": 1.8242577438841523,
      "grad_norm": 4.325722694396973,
      "learning_rate": 7.059696839825182e-05,
      "loss": 2.428070068359375,
      "memory(GiB)": 77.56,
      "step": 42580,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.440309
    },
    {
      "epoch": 1.8244719592134013,
      "grad_norm": 4.568417549133301,
      "learning_rate": 7.05908359667383e-05,
      "loss": 2.5685184478759764,
      "memory(GiB)": 77.56,
      "step": 42585,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.440317
    },
    {
      "epoch": 1.8246861745426504,
      "grad_norm": 5.907657623291016,
      "learning_rate": 7.058470316220248e-05,
      "loss": 2.3318790435791015,
      "memory(GiB)": 77.56,
      "step": 42590,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.440301
    },
    {
      "epoch": 1.8249003898718992,
      "grad_norm": 6.304762840270996,
      "learning_rate": 7.057856998475543e-05,
      "loss": 2.5876407623291016,
      "memory(GiB)": 77.56,
      "step": 42595,
      "token_acc": 0.5059288537549407,
      "train_speed(iter/s)": 1.440326
    },
    {
      "epoch": 1.8251146052011482,
      "grad_norm": 5.732647895812988,
      "learning_rate": 7.05724364345083e-05,
      "loss": 2.6107913970947267,
      "memory(GiB)": 77.56,
      "step": 42600,
      "token_acc": 0.45016077170418006,
      "train_speed(iter/s)": 1.440323
    },
    {
      "epoch": 1.8253288205303972,
      "grad_norm": 5.199191093444824,
      "learning_rate": 7.056630251157219e-05,
      "loss": 2.3107242584228516,
      "memory(GiB)": 77.56,
      "step": 42605,
      "token_acc": 0.506578947368421,
      "train_speed(iter/s)": 1.44035
    },
    {
      "epoch": 1.825543035859646,
      "grad_norm": 4.308810710906982,
      "learning_rate": 7.05601682160582e-05,
      "loss": 2.695077133178711,
      "memory(GiB)": 77.56,
      "step": 42610,
      "token_acc": 0.4565826330532213,
      "train_speed(iter/s)": 1.440358
    },
    {
      "epoch": 1.825757251188895,
      "grad_norm": 6.556538105010986,
      "learning_rate": 7.055403354807749e-05,
      "loss": 2.414827346801758,
      "memory(GiB)": 77.56,
      "step": 42615,
      "token_acc": 0.501577287066246,
      "train_speed(iter/s)": 1.440343
    },
    {
      "epoch": 1.8259714665181441,
      "grad_norm": 6.080617427825928,
      "learning_rate": 7.054789850774118e-05,
      "loss": 2.4803916931152346,
      "memory(GiB)": 77.56,
      "step": 42620,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.440339
    },
    {
      "epoch": 1.826185681847393,
      "grad_norm": 6.105247974395752,
      "learning_rate": 7.054176309516041e-05,
      "loss": 2.6499397277832033,
      "memory(GiB)": 77.56,
      "step": 42625,
      "token_acc": 0.4808259587020649,
      "train_speed(iter/s)": 1.440349
    },
    {
      "epoch": 1.826399897176642,
      "grad_norm": 6.8638014793396,
      "learning_rate": 7.053562731044632e-05,
      "loss": 2.483260726928711,
      "memory(GiB)": 77.56,
      "step": 42630,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.440385
    },
    {
      "epoch": 1.826614112505891,
      "grad_norm": 4.559618949890137,
      "learning_rate": 7.05294911537101e-05,
      "loss": 2.4516197204589845,
      "memory(GiB)": 77.56,
      "step": 42635,
      "token_acc": 0.4517133956386293,
      "train_speed(iter/s)": 1.440405
    },
    {
      "epoch": 1.8268283278351398,
      "grad_norm": 6.369504928588867,
      "learning_rate": 7.052335462506285e-05,
      "loss": 2.7407154083251952,
      "memory(GiB)": 77.56,
      "step": 42640,
      "token_acc": 0.44047619047619047,
      "train_speed(iter/s)": 1.440392
    },
    {
      "epoch": 1.8270425431643889,
      "grad_norm": 5.443140029907227,
      "learning_rate": 7.051721772461582e-05,
      "loss": 2.403109550476074,
      "memory(GiB)": 77.56,
      "step": 42645,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.440407
    },
    {
      "epoch": 1.827256758493638,
      "grad_norm": 5.081888675689697,
      "learning_rate": 7.051108045248014e-05,
      "loss": 2.6135513305664064,
      "memory(GiB)": 77.56,
      "step": 42650,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.440404
    },
    {
      "epoch": 1.8274709738228867,
      "grad_norm": 6.575129985809326,
      "learning_rate": 7.050494280876697e-05,
      "loss": 2.4567901611328127,
      "memory(GiB)": 77.56,
      "step": 42655,
      "token_acc": 0.5063829787234042,
      "train_speed(iter/s)": 1.44042
    },
    {
      "epoch": 1.8276851891521357,
      "grad_norm": 4.810214042663574,
      "learning_rate": 7.049880479358754e-05,
      "loss": 2.452205276489258,
      "memory(GiB)": 77.56,
      "step": 42660,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.440411
    },
    {
      "epoch": 1.8278994044813848,
      "grad_norm": 5.769571781158447,
      "learning_rate": 7.049266640705304e-05,
      "loss": 2.678585433959961,
      "memory(GiB)": 77.56,
      "step": 42665,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440399
    },
    {
      "epoch": 1.8281136198106336,
      "grad_norm": 4.388237476348877,
      "learning_rate": 7.048652764927466e-05,
      "loss": 2.7901378631591798,
      "memory(GiB)": 77.56,
      "step": 42670,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.440421
    },
    {
      "epoch": 1.8283278351398826,
      "grad_norm": 4.146549701690674,
      "learning_rate": 7.04803885203636e-05,
      "loss": 2.524386978149414,
      "memory(GiB)": 77.56,
      "step": 42675,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.440461
    },
    {
      "epoch": 1.8285420504691317,
      "grad_norm": 3.608217239379883,
      "learning_rate": 7.047424902043111e-05,
      "loss": 2.617009162902832,
      "memory(GiB)": 77.56,
      "step": 42680,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.440473
    },
    {
      "epoch": 1.8287562657983805,
      "grad_norm": 5.381430625915527,
      "learning_rate": 7.046810914958839e-05,
      "loss": 2.31790828704834,
      "memory(GiB)": 77.56,
      "step": 42685,
      "token_acc": 0.4701195219123506,
      "train_speed(iter/s)": 1.440489
    },
    {
      "epoch": 1.8289704811276295,
      "grad_norm": 5.573160648345947,
      "learning_rate": 7.046196890794666e-05,
      "loss": 2.4803380966186523,
      "memory(GiB)": 77.56,
      "step": 42690,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.8291846964568785,
      "grad_norm": 6.720072269439697,
      "learning_rate": 7.045582829561718e-05,
      "loss": 2.4793569564819338,
      "memory(GiB)": 77.56,
      "step": 42695,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.440516
    },
    {
      "epoch": 1.8293989117861273,
      "grad_norm": 5.089569568634033,
      "learning_rate": 7.044968731271118e-05,
      "loss": 2.31717529296875,
      "memory(GiB)": 77.56,
      "step": 42700,
      "token_acc": 0.4981549815498155,
      "train_speed(iter/s)": 1.440509
    },
    {
      "epoch": 1.8296131271153764,
      "grad_norm": 5.226766109466553,
      "learning_rate": 7.04435459593399e-05,
      "loss": 2.63138484954834,
      "memory(GiB)": 77.56,
      "step": 42705,
      "token_acc": 0.45878136200716846,
      "train_speed(iter/s)": 1.440499
    },
    {
      "epoch": 1.8298273424446254,
      "grad_norm": 5.256031513214111,
      "learning_rate": 7.04374042356146e-05,
      "loss": 2.8003908157348634,
      "memory(GiB)": 77.56,
      "step": 42710,
      "token_acc": 0.4387755102040816,
      "train_speed(iter/s)": 1.440458
    },
    {
      "epoch": 1.8300415577738742,
      "grad_norm": 4.400367259979248,
      "learning_rate": 7.043126214164657e-05,
      "loss": 2.4033090591430666,
      "memory(GiB)": 77.56,
      "step": 42715,
      "token_acc": 0.5140845070422535,
      "train_speed(iter/s)": 1.440485
    },
    {
      "epoch": 1.8302557731031233,
      "grad_norm": 6.759885311126709,
      "learning_rate": 7.042511967754707e-05,
      "loss": 2.7107378005981446,
      "memory(GiB)": 77.56,
      "step": 42720,
      "token_acc": 0.45569620253164556,
      "train_speed(iter/s)": 1.440511
    },
    {
      "epoch": 1.8304699884323723,
      "grad_norm": 6.6062116622924805,
      "learning_rate": 7.041897684342734e-05,
      "loss": 2.503848648071289,
      "memory(GiB)": 77.56,
      "step": 42725,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.440518
    },
    {
      "epoch": 1.830684203761621,
      "grad_norm": 7.270866870880127,
      "learning_rate": 7.04128336393987e-05,
      "loss": 2.848384666442871,
      "memory(GiB)": 77.56,
      "step": 42730,
      "token_acc": 0.4232081911262799,
      "train_speed(iter/s)": 1.440544
    },
    {
      "epoch": 1.8308984190908701,
      "grad_norm": 4.867170810699463,
      "learning_rate": 7.040669006557245e-05,
      "loss": 2.3126014709472655,
      "memory(GiB)": 77.56,
      "step": 42735,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.440524
    },
    {
      "epoch": 1.8311126344201192,
      "grad_norm": 5.4053635597229,
      "learning_rate": 7.040054612205982e-05,
      "loss": 2.4727672576904296,
      "memory(GiB)": 77.56,
      "step": 42740,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.440515
    },
    {
      "epoch": 1.831326849749368,
      "grad_norm": 5.161674976348877,
      "learning_rate": 7.03944018089722e-05,
      "loss": 2.333534812927246,
      "memory(GiB)": 77.56,
      "step": 42745,
      "token_acc": 0.47019867549668876,
      "train_speed(iter/s)": 1.44052
    },
    {
      "epoch": 1.831541065078617,
      "grad_norm": 4.686703205108643,
      "learning_rate": 7.038825712642085e-05,
      "loss": 2.735204315185547,
      "memory(GiB)": 77.56,
      "step": 42750,
      "token_acc": 0.4078014184397163,
      "train_speed(iter/s)": 1.440562
    },
    {
      "epoch": 1.831755280407866,
      "grad_norm": 4.866014003753662,
      "learning_rate": 7.03821120745171e-05,
      "loss": 2.665049362182617,
      "memory(GiB)": 77.56,
      "step": 42755,
      "token_acc": 0.4533333333333333,
      "train_speed(iter/s)": 1.440578
    },
    {
      "epoch": 1.8319694957371149,
      "grad_norm": 4.805840969085693,
      "learning_rate": 7.037596665337226e-05,
      "loss": 2.5504104614257814,
      "memory(GiB)": 77.56,
      "step": 42760,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.440597
    },
    {
      "epoch": 1.832183711066364,
      "grad_norm": 5.937466621398926,
      "learning_rate": 7.036982086309769e-05,
      "loss": 2.3927824020385744,
      "memory(GiB)": 77.56,
      "step": 42765,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.440654
    },
    {
      "epoch": 1.832397926395613,
      "grad_norm": 6.452016830444336,
      "learning_rate": 7.036367470380467e-05,
      "loss": 2.840751838684082,
      "memory(GiB)": 77.56,
      "step": 42770,
      "token_acc": 0.44876325088339225,
      "train_speed(iter/s)": 1.440628
    },
    {
      "epoch": 1.8326121417248618,
      "grad_norm": 7.551222324371338,
      "learning_rate": 7.03575281756046e-05,
      "loss": 2.674056816101074,
      "memory(GiB)": 77.56,
      "step": 42775,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.440636
    },
    {
      "epoch": 1.8328263570541108,
      "grad_norm": 7.136316776275635,
      "learning_rate": 7.03513812786088e-05,
      "loss": 2.59435977935791,
      "memory(GiB)": 77.56,
      "step": 42780,
      "token_acc": 0.4742268041237113,
      "train_speed(iter/s)": 1.44062
    },
    {
      "epoch": 1.8330405723833598,
      "grad_norm": 4.6389899253845215,
      "learning_rate": 7.034523401292866e-05,
      "loss": 2.7101425170898437,
      "memory(GiB)": 77.56,
      "step": 42785,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.440627
    },
    {
      "epoch": 1.8332547877126086,
      "grad_norm": 3.842972993850708,
      "learning_rate": 7.03390863786755e-05,
      "loss": 2.4858264923095703,
      "memory(GiB)": 77.56,
      "step": 42790,
      "token_acc": 0.49002849002849,
      "train_speed(iter/s)": 1.440653
    },
    {
      "epoch": 1.8334690030418577,
      "grad_norm": 5.817306041717529,
      "learning_rate": 7.033293837596073e-05,
      "loss": 2.576206588745117,
      "memory(GiB)": 77.56,
      "step": 42795,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.440662
    },
    {
      "epoch": 1.8336832183711067,
      "grad_norm": 6.170565605163574,
      "learning_rate": 7.032679000489568e-05,
      "loss": 2.592821502685547,
      "memory(GiB)": 77.56,
      "step": 42800,
      "token_acc": 0.45144356955380577,
      "train_speed(iter/s)": 1.440668
    },
    {
      "epoch": 1.8338974337003555,
      "grad_norm": 5.738097190856934,
      "learning_rate": 7.032064126559179e-05,
      "loss": 2.663206100463867,
      "memory(GiB)": 77.56,
      "step": 42805,
      "token_acc": 0.44868035190615835,
      "train_speed(iter/s)": 1.440678
    },
    {
      "epoch": 1.8341116490296046,
      "grad_norm": 5.772311687469482,
      "learning_rate": 7.031449215816041e-05,
      "loss": 2.6962120056152346,
      "memory(GiB)": 77.56,
      "step": 42810,
      "token_acc": 0.4280936454849498,
      "train_speed(iter/s)": 1.440685
    },
    {
      "epoch": 1.8343258643588536,
      "grad_norm": 5.5379509925842285,
      "learning_rate": 7.030834268271294e-05,
      "loss": 2.4956586837768553,
      "memory(GiB)": 77.56,
      "step": 42815,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.440683
    },
    {
      "epoch": 1.8345400796881024,
      "grad_norm": 4.719763278961182,
      "learning_rate": 7.030219283936079e-05,
      "loss": 2.643565368652344,
      "memory(GiB)": 77.56,
      "step": 42820,
      "token_acc": 0.4485294117647059,
      "train_speed(iter/s)": 1.440694
    },
    {
      "epoch": 1.8347542950173514,
      "grad_norm": 4.409863471984863,
      "learning_rate": 7.029604262821539e-05,
      "loss": 2.5345787048339843,
      "memory(GiB)": 77.56,
      "step": 42825,
      "token_acc": 0.4810126582278481,
      "train_speed(iter/s)": 1.440697
    },
    {
      "epoch": 1.8349685103466005,
      "grad_norm": 5.3105268478393555,
      "learning_rate": 7.028989204938812e-05,
      "loss": 2.3162155151367188,
      "memory(GiB)": 77.56,
      "step": 42830,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.440737
    },
    {
      "epoch": 1.8351827256758493,
      "grad_norm": 5.830991268157959,
      "learning_rate": 7.028374110299044e-05,
      "loss": 2.385612678527832,
      "memory(GiB)": 77.56,
      "step": 42835,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.440776
    },
    {
      "epoch": 1.8353969410050983,
      "grad_norm": 6.850379467010498,
      "learning_rate": 7.027758978913374e-05,
      "loss": 2.410422134399414,
      "memory(GiB)": 77.56,
      "step": 42840,
      "token_acc": 0.4858156028368794,
      "train_speed(iter/s)": 1.44078
    },
    {
      "epoch": 1.8356111563343473,
      "grad_norm": 4.204717636108398,
      "learning_rate": 7.02714381079295e-05,
      "loss": 2.3584640502929686,
      "memory(GiB)": 77.56,
      "step": 42845,
      "token_acc": 0.5093167701863354,
      "train_speed(iter/s)": 1.440822
    },
    {
      "epoch": 1.8358253716635962,
      "grad_norm": 3.996864080429077,
      "learning_rate": 7.026528605948913e-05,
      "loss": 2.572443962097168,
      "memory(GiB)": 77.56,
      "step": 42850,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.440809
    },
    {
      "epoch": 1.8360395869928452,
      "grad_norm": 4.448183536529541,
      "learning_rate": 7.025913364392409e-05,
      "loss": 2.6070602416992186,
      "memory(GiB)": 77.56,
      "step": 42855,
      "token_acc": 0.44510385756676557,
      "train_speed(iter/s)": 1.440774
    },
    {
      "epoch": 1.8362538023220942,
      "grad_norm": 6.315032482147217,
      "learning_rate": 7.025298086134586e-05,
      "loss": 2.8378742218017576,
      "memory(GiB)": 77.56,
      "step": 42860,
      "token_acc": 0.425,
      "train_speed(iter/s)": 1.44075
    },
    {
      "epoch": 1.836468017651343,
      "grad_norm": 5.0336761474609375,
      "learning_rate": 7.024682771186586e-05,
      "loss": 2.6184921264648438,
      "memory(GiB)": 77.56,
      "step": 42865,
      "token_acc": 0.484251968503937,
      "train_speed(iter/s)": 1.440768
    },
    {
      "epoch": 1.836682232980592,
      "grad_norm": 5.010344982147217,
      "learning_rate": 7.02406741955956e-05,
      "loss": 2.6325130462646484,
      "memory(GiB)": 77.56,
      "step": 42870,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.440763
    },
    {
      "epoch": 1.8368964483098411,
      "grad_norm": 5.454922199249268,
      "learning_rate": 7.023452031264653e-05,
      "loss": 2.5520700454711913,
      "memory(GiB)": 77.56,
      "step": 42875,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.440767
    },
    {
      "epoch": 1.83711066363909,
      "grad_norm": 5.264733791351318,
      "learning_rate": 7.022836606313015e-05,
      "loss": 2.335470199584961,
      "memory(GiB)": 77.56,
      "step": 42880,
      "token_acc": 0.535593220338983,
      "train_speed(iter/s)": 1.440748
    },
    {
      "epoch": 1.837324878968339,
      "grad_norm": 4.933237552642822,
      "learning_rate": 7.022221144715793e-05,
      "loss": 2.393748474121094,
      "memory(GiB)": 77.56,
      "step": 42885,
      "token_acc": 0.4470588235294118,
      "train_speed(iter/s)": 1.440753
    },
    {
      "epoch": 1.837539094297588,
      "grad_norm": 6.662442207336426,
      "learning_rate": 7.021605646484137e-05,
      "loss": 2.7910985946655273,
      "memory(GiB)": 77.56,
      "step": 42890,
      "token_acc": 0.4368932038834951,
      "train_speed(iter/s)": 1.440682
    },
    {
      "epoch": 1.8377533096268368,
      "grad_norm": 5.211727142333984,
      "learning_rate": 7.020990111629202e-05,
      "loss": 2.9527698516845704,
      "memory(GiB)": 77.56,
      "step": 42895,
      "token_acc": 0.4421768707482993,
      "train_speed(iter/s)": 1.440671
    },
    {
      "epoch": 1.8379675249560858,
      "grad_norm": 5.574032783508301,
      "learning_rate": 7.020374540162132e-05,
      "loss": 2.2710084915161133,
      "memory(GiB)": 77.56,
      "step": 42900,
      "token_acc": 0.49407114624505927,
      "train_speed(iter/s)": 1.440718
    },
    {
      "epoch": 1.8381817402853349,
      "grad_norm": 5.917013168334961,
      "learning_rate": 7.019758932094083e-05,
      "loss": 2.2579360961914063,
      "memory(GiB)": 77.56,
      "step": 42905,
      "token_acc": 0.5194805194805194,
      "train_speed(iter/s)": 1.440726
    },
    {
      "epoch": 1.8383959556145837,
      "grad_norm": 4.261195659637451,
      "learning_rate": 7.019143287436207e-05,
      "loss": 2.517148017883301,
      "memory(GiB)": 77.56,
      "step": 42910,
      "token_acc": 0.5078369905956113,
      "train_speed(iter/s)": 1.440739
    },
    {
      "epoch": 1.8386101709438327,
      "grad_norm": 5.687638282775879,
      "learning_rate": 7.018527606199657e-05,
      "loss": 2.4009639739990236,
      "memory(GiB)": 77.56,
      "step": 42915,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.44071
    },
    {
      "epoch": 1.8388243862730818,
      "grad_norm": 5.801491737365723,
      "learning_rate": 7.017911888395584e-05,
      "loss": 2.4040607452392577,
      "memory(GiB)": 77.56,
      "step": 42920,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.440677
    },
    {
      "epoch": 1.8390386016023306,
      "grad_norm": 5.284247398376465,
      "learning_rate": 7.017296134035145e-05,
      "loss": 2.453633117675781,
      "memory(GiB)": 77.56,
      "step": 42925,
      "token_acc": 0.5205479452054794,
      "train_speed(iter/s)": 1.440694
    },
    {
      "epoch": 1.8392528169315796,
      "grad_norm": 6.3654465675354,
      "learning_rate": 7.016680343129493e-05,
      "loss": 2.527718734741211,
      "memory(GiB)": 77.56,
      "step": 42930,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.440744
    },
    {
      "epoch": 1.8394670322608286,
      "grad_norm": 4.958274841308594,
      "learning_rate": 7.016064515689787e-05,
      "loss": 2.4775012969970702,
      "memory(GiB)": 77.56,
      "step": 42935,
      "token_acc": 0.4900662251655629,
      "train_speed(iter/s)": 1.440736
    },
    {
      "epoch": 1.8396812475900775,
      "grad_norm": 5.551571846008301,
      "learning_rate": 7.015448651727179e-05,
      "loss": 2.655436325073242,
      "memory(GiB)": 77.56,
      "step": 42940,
      "token_acc": 0.4430769230769231,
      "train_speed(iter/s)": 1.44069
    },
    {
      "epoch": 1.8398954629193265,
      "grad_norm": 4.376008033752441,
      "learning_rate": 7.01483275125283e-05,
      "loss": 2.537699890136719,
      "memory(GiB)": 77.56,
      "step": 42945,
      "token_acc": 0.49174917491749176,
      "train_speed(iter/s)": 1.440726
    },
    {
      "epoch": 1.8401096782485755,
      "grad_norm": 5.936407089233398,
      "learning_rate": 7.014216814277894e-05,
      "loss": 2.7400318145751954,
      "memory(GiB)": 77.56,
      "step": 42950,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.440742
    },
    {
      "epoch": 1.8403238935778243,
      "grad_norm": 5.693716049194336,
      "learning_rate": 7.01360084081353e-05,
      "loss": 2.706703948974609,
      "memory(GiB)": 77.56,
      "step": 42955,
      "token_acc": 0.4244604316546763,
      "train_speed(iter/s)": 1.440753
    },
    {
      "epoch": 1.8405381089070734,
      "grad_norm": 6.113903999328613,
      "learning_rate": 7.0129848308709e-05,
      "loss": 2.658347320556641,
      "memory(GiB)": 77.56,
      "step": 42960,
      "token_acc": 0.45058139534883723,
      "train_speed(iter/s)": 1.440785
    },
    {
      "epoch": 1.8407523242363224,
      "grad_norm": 5.083117961883545,
      "learning_rate": 7.012368784461161e-05,
      "loss": 2.833583641052246,
      "memory(GiB)": 77.56,
      "step": 42965,
      "token_acc": 0.4233128834355828,
      "train_speed(iter/s)": 1.440824
    },
    {
      "epoch": 1.8409665395655712,
      "grad_norm": 4.786596775054932,
      "learning_rate": 7.011752701595471e-05,
      "loss": 2.790046501159668,
      "memory(GiB)": 77.56,
      "step": 42970,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.440831
    },
    {
      "epoch": 1.8411807548948202,
      "grad_norm": 5.824346542358398,
      "learning_rate": 7.011136582284998e-05,
      "loss": 2.5863201141357424,
      "memory(GiB)": 77.56,
      "step": 42975,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.44084
    },
    {
      "epoch": 1.8413949702240693,
      "grad_norm": 4.7220845222473145,
      "learning_rate": 7.010520426540897e-05,
      "loss": 2.336712646484375,
      "memory(GiB)": 77.56,
      "step": 42980,
      "token_acc": 0.5179856115107914,
      "train_speed(iter/s)": 1.440843
    },
    {
      "epoch": 1.841609185553318,
      "grad_norm": 5.535294532775879,
      "learning_rate": 7.009904234374332e-05,
      "loss": 2.5778785705566407,
      "memory(GiB)": 77.56,
      "step": 42985,
      "token_acc": 0.4674329501915709,
      "train_speed(iter/s)": 1.440867
    },
    {
      "epoch": 1.8418234008825671,
      "grad_norm": 6.33097505569458,
      "learning_rate": 7.009288005796469e-05,
      "loss": 2.712126541137695,
      "memory(GiB)": 77.56,
      "step": 42990,
      "token_acc": 0.41132075471698115,
      "train_speed(iter/s)": 1.440898
    },
    {
      "epoch": 1.8420376162118162,
      "grad_norm": 4.346415996551514,
      "learning_rate": 7.008671740818466e-05,
      "loss": 2.378527069091797,
      "memory(GiB)": 77.56,
      "step": 42995,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.440894
    },
    {
      "epoch": 1.842251831541065,
      "grad_norm": 4.51521110534668,
      "learning_rate": 7.008055439451491e-05,
      "loss": 2.4386518478393553,
      "memory(GiB)": 77.56,
      "step": 43000,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.440871
    },
    {
      "epoch": 1.842251831541065,
      "eval_loss": 2.1833853721618652,
      "eval_runtime": 14.0243,
      "eval_samples_per_second": 7.131,
      "eval_steps_per_second": 7.131,
      "eval_token_acc": 0.4574468085106383,
      "step": 43000
    },
    {
      "epoch": 1.842466046870314,
      "grad_norm": 4.630211353302002,
      "learning_rate": 7.007439101706708e-05,
      "loss": 2.5586383819580076,
      "memory(GiB)": 77.56,
      "step": 43005,
      "token_acc": 0.46735751295336786,
      "train_speed(iter/s)": 1.440152
    },
    {
      "epoch": 1.842680262199563,
      "grad_norm": 4.667478084564209,
      "learning_rate": 7.006822727595284e-05,
      "loss": 2.679239273071289,
      "memory(GiB)": 77.56,
      "step": 43010,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.440158
    },
    {
      "epoch": 1.8428944775288119,
      "grad_norm": 4.515958309173584,
      "learning_rate": 7.00620631712838e-05,
      "loss": 2.890361785888672,
      "memory(GiB)": 77.56,
      "step": 43015,
      "token_acc": 0.38613861386138615,
      "train_speed(iter/s)": 1.44016
    },
    {
      "epoch": 1.843108692858061,
      "grad_norm": 5.206225872039795,
      "learning_rate": 7.00558987031717e-05,
      "loss": 2.6854297637939455,
      "memory(GiB)": 77.56,
      "step": 43020,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.440175
    },
    {
      "epoch": 1.84332290818731,
      "grad_norm": 4.903097152709961,
      "learning_rate": 7.004973387172818e-05,
      "loss": 2.7004928588867188,
      "memory(GiB)": 77.56,
      "step": 43025,
      "token_acc": 0.44405594405594406,
      "train_speed(iter/s)": 1.440198
    },
    {
      "epoch": 1.8435371235165587,
      "grad_norm": 4.881291389465332,
      "learning_rate": 7.00435686770649e-05,
      "loss": 2.544137954711914,
      "memory(GiB)": 77.56,
      "step": 43030,
      "token_acc": 0.44931506849315067,
      "train_speed(iter/s)": 1.440199
    },
    {
      "epoch": 1.8437513388458078,
      "grad_norm": 5.065155982971191,
      "learning_rate": 7.003740311929358e-05,
      "loss": 2.2918020248413087,
      "memory(GiB)": 77.56,
      "step": 43035,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.440155
    },
    {
      "epoch": 1.8439655541750568,
      "grad_norm": 3.8211097717285156,
      "learning_rate": 7.003123719852592e-05,
      "loss": 2.2812437057495116,
      "memory(GiB)": 77.56,
      "step": 43040,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440169
    },
    {
      "epoch": 1.8441797695043056,
      "grad_norm": 5.353897571563721,
      "learning_rate": 7.00250709148736e-05,
      "loss": 2.7209804534912108,
      "memory(GiB)": 77.56,
      "step": 43045,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.440079
    },
    {
      "epoch": 1.8443939848335547,
      "grad_norm": 4.515683174133301,
      "learning_rate": 7.001890426844833e-05,
      "loss": 2.576288604736328,
      "memory(GiB)": 77.56,
      "step": 43050,
      "token_acc": 0.4391891891891892,
      "train_speed(iter/s)": 1.440098
    },
    {
      "epoch": 1.8446082001628037,
      "grad_norm": 4.109330654144287,
      "learning_rate": 7.001273725936184e-05,
      "loss": 2.634905242919922,
      "memory(GiB)": 77.56,
      "step": 43055,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.440117
    },
    {
      "epoch": 1.8448224154920525,
      "grad_norm": 4.0266265869140625,
      "learning_rate": 7.000656988772583e-05,
      "loss": 2.897775077819824,
      "memory(GiB)": 77.56,
      "step": 43060,
      "token_acc": 0.43641618497109824,
      "train_speed(iter/s)": 1.440127
    },
    {
      "epoch": 1.8450366308213015,
      "grad_norm": 4.472598552703857,
      "learning_rate": 7.000040215365205e-05,
      "loss": 2.56237735748291,
      "memory(GiB)": 77.56,
      "step": 43065,
      "token_acc": 0.49034749034749037,
      "train_speed(iter/s)": 1.440151
    },
    {
      "epoch": 1.8452508461505506,
      "grad_norm": 4.262383460998535,
      "learning_rate": 6.999423405725221e-05,
      "loss": 2.7779729843139647,
      "memory(GiB)": 77.56,
      "step": 43070,
      "token_acc": 0.40425531914893614,
      "train_speed(iter/s)": 1.44018
    },
    {
      "epoch": 1.8454650614797994,
      "grad_norm": 6.024571418762207,
      "learning_rate": 6.998806559863806e-05,
      "loss": 2.6618925094604493,
      "memory(GiB)": 77.56,
      "step": 43075,
      "token_acc": 0.44565217391304346,
      "train_speed(iter/s)": 1.440183
    },
    {
      "epoch": 1.8456792768090484,
      "grad_norm": 5.0662455558776855,
      "learning_rate": 6.998189677792138e-05,
      "loss": 2.6739423751831053,
      "memory(GiB)": 77.56,
      "step": 43080,
      "token_acc": 0.4888268156424581,
      "train_speed(iter/s)": 1.440187
    },
    {
      "epoch": 1.8458934921382975,
      "grad_norm": 6.7062530517578125,
      "learning_rate": 6.997572759521386e-05,
      "loss": 3.038888931274414,
      "memory(GiB)": 77.56,
      "step": 43085,
      "token_acc": 0.3944020356234097,
      "train_speed(iter/s)": 1.440234
    },
    {
      "epoch": 1.8461077074675463,
      "grad_norm": 6.608982086181641,
      "learning_rate": 6.996955805062732e-05,
      "loss": 2.4786960601806642,
      "memory(GiB)": 77.56,
      "step": 43090,
      "token_acc": 0.47924528301886793,
      "train_speed(iter/s)": 1.440245
    },
    {
      "epoch": 1.8463219227967953,
      "grad_norm": 3.9080896377563477,
      "learning_rate": 6.99633881442735e-05,
      "loss": 2.4747425079345704,
      "memory(GiB)": 77.56,
      "step": 43095,
      "token_acc": 0.467966573816156,
      "train_speed(iter/s)": 1.440181
    },
    {
      "epoch": 1.8465361381260443,
      "grad_norm": 4.1277031898498535,
      "learning_rate": 6.995721787626417e-05,
      "loss": 2.425519561767578,
      "memory(GiB)": 77.56,
      "step": 43100,
      "token_acc": 0.5506072874493927,
      "train_speed(iter/s)": 1.440183
    },
    {
      "epoch": 1.8467503534552931,
      "grad_norm": 4.736281871795654,
      "learning_rate": 6.995104724671112e-05,
      "loss": 2.764559745788574,
      "memory(GiB)": 77.56,
      "step": 43105,
      "token_acc": 0.4297994269340974,
      "train_speed(iter/s)": 1.440191
    },
    {
      "epoch": 1.8469645687845422,
      "grad_norm": 4.871133327484131,
      "learning_rate": 6.994487625572613e-05,
      "loss": 2.4187595367431642,
      "memory(GiB)": 77.56,
      "step": 43110,
      "token_acc": 0.46956521739130436,
      "train_speed(iter/s)": 1.440197
    },
    {
      "epoch": 1.8471787841137912,
      "grad_norm": 4.730271816253662,
      "learning_rate": 6.993870490342099e-05,
      "loss": 2.6471309661865234,
      "memory(GiB)": 77.56,
      "step": 43115,
      "token_acc": 0.446875,
      "train_speed(iter/s)": 1.440177
    },
    {
      "epoch": 1.84739299944304,
      "grad_norm": 4.323908805847168,
      "learning_rate": 6.993253318990753e-05,
      "loss": 2.4310916900634765,
      "memory(GiB)": 77.56,
      "step": 43120,
      "token_acc": 0.46835443037974683,
      "train_speed(iter/s)": 1.440178
    },
    {
      "epoch": 1.847607214772289,
      "grad_norm": 5.497196674346924,
      "learning_rate": 6.992636111529752e-05,
      "loss": 2.4413120269775392,
      "memory(GiB)": 77.56,
      "step": 43125,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.440191
    },
    {
      "epoch": 1.847821430101538,
      "grad_norm": 5.502140045166016,
      "learning_rate": 6.992018867970278e-05,
      "loss": 2.975293731689453,
      "memory(GiB)": 77.56,
      "step": 43130,
      "token_acc": 0.4146341463414634,
      "train_speed(iter/s)": 1.440179
    },
    {
      "epoch": 1.848035645430787,
      "grad_norm": 4.811458587646484,
      "learning_rate": 6.991401588323514e-05,
      "loss": 2.208480453491211,
      "memory(GiB)": 77.56,
      "step": 43135,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.440182
    },
    {
      "epoch": 1.848249860760036,
      "grad_norm": 4.557877540588379,
      "learning_rate": 6.990784272600643e-05,
      "loss": 2.1829463958740236,
      "memory(GiB)": 77.56,
      "step": 43140,
      "token_acc": 0.5149253731343284,
      "train_speed(iter/s)": 1.440151
    },
    {
      "epoch": 1.848464076089285,
      "grad_norm": 4.922790050506592,
      "learning_rate": 6.990166920812847e-05,
      "loss": 2.4981847763061524,
      "memory(GiB)": 77.56,
      "step": 43145,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.440178
    },
    {
      "epoch": 1.8486782914185338,
      "grad_norm": 4.187565803527832,
      "learning_rate": 6.989549532971309e-05,
      "loss": 3.0007017135620115,
      "memory(GiB)": 77.56,
      "step": 43150,
      "token_acc": 0.43,
      "train_speed(iter/s)": 1.440214
    },
    {
      "epoch": 1.8488925067477828,
      "grad_norm": 7.6599555015563965,
      "learning_rate": 6.988932109087216e-05,
      "loss": 2.5249298095703123,
      "memory(GiB)": 77.56,
      "step": 43155,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.440219
    },
    {
      "epoch": 1.8491067220770319,
      "grad_norm": 8.369895935058594,
      "learning_rate": 6.988314649171751e-05,
      "loss": 2.2094568252563476,
      "memory(GiB)": 77.56,
      "step": 43160,
      "token_acc": 0.491869918699187,
      "train_speed(iter/s)": 1.440265
    },
    {
      "epoch": 1.8493209374062807,
      "grad_norm": 3.8235225677490234,
      "learning_rate": 6.987697153236102e-05,
      "loss": 2.48919677734375,
      "memory(GiB)": 77.56,
      "step": 43165,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440288
    },
    {
      "epoch": 1.8495351527355297,
      "grad_norm": 3.86279559135437,
      "learning_rate": 6.987079621291455e-05,
      "loss": 2.2730331420898438,
      "memory(GiB)": 77.56,
      "step": 43170,
      "token_acc": 0.5106382978723404,
      "train_speed(iter/s)": 1.44032
    },
    {
      "epoch": 1.8497493680647787,
      "grad_norm": 5.171674728393555,
      "learning_rate": 6.986462053348996e-05,
      "loss": 2.55029296875,
      "memory(GiB)": 77.56,
      "step": 43175,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.440328
    },
    {
      "epoch": 1.8499635833940276,
      "grad_norm": 4.2251811027526855,
      "learning_rate": 6.985844449419913e-05,
      "loss": 2.502684211730957,
      "memory(GiB)": 77.56,
      "step": 43180,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.440346
    },
    {
      "epoch": 1.8501777987232766,
      "grad_norm": 4.920688152313232,
      "learning_rate": 6.985226809515395e-05,
      "loss": 2.8009572982788087,
      "memory(GiB)": 77.56,
      "step": 43185,
      "token_acc": 0.40836012861736337,
      "train_speed(iter/s)": 1.440366
    },
    {
      "epoch": 1.8503920140525256,
      "grad_norm": 5.436327934265137,
      "learning_rate": 6.98460913364663e-05,
      "loss": 2.579128456115723,
      "memory(GiB)": 77.56,
      "step": 43190,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440367
    },
    {
      "epoch": 1.8506062293817744,
      "grad_norm": 5.891781330108643,
      "learning_rate": 6.983991421824811e-05,
      "loss": 2.523370933532715,
      "memory(GiB)": 77.56,
      "step": 43195,
      "token_acc": 0.510548523206751,
      "train_speed(iter/s)": 1.440379
    },
    {
      "epoch": 1.8508204447110235,
      "grad_norm": 4.855978012084961,
      "learning_rate": 6.983373674061126e-05,
      "loss": 2.462619972229004,
      "memory(GiB)": 77.56,
      "step": 43200,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440401
    },
    {
      "epoch": 1.8510346600402725,
      "grad_norm": 4.836047649383545,
      "learning_rate": 6.982755890366766e-05,
      "loss": 2.6963569641113283,
      "memory(GiB)": 77.56,
      "step": 43205,
      "token_acc": 0.4552238805970149,
      "train_speed(iter/s)": 1.440406
    },
    {
      "epoch": 1.8512488753695213,
      "grad_norm": 4.0195231437683105,
      "learning_rate": 6.982138070752923e-05,
      "loss": 2.649702453613281,
      "memory(GiB)": 77.56,
      "step": 43210,
      "token_acc": 0.4905149051490515,
      "train_speed(iter/s)": 1.440429
    },
    {
      "epoch": 1.8514630906987704,
      "grad_norm": 5.107478141784668,
      "learning_rate": 6.981520215230788e-05,
      "loss": 2.583074188232422,
      "memory(GiB)": 77.56,
      "step": 43215,
      "token_acc": 0.4962121212121212,
      "train_speed(iter/s)": 1.44047
    },
    {
      "epoch": 1.8516773060280194,
      "grad_norm": 4.943504333496094,
      "learning_rate": 6.980902323811557e-05,
      "loss": 2.4854690551757814,
      "memory(GiB)": 77.56,
      "step": 43220,
      "token_acc": 0.47470817120622566,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.8518915213572682,
      "grad_norm": 3.9420807361602783,
      "learning_rate": 6.980284396506421e-05,
      "loss": 2.6313228607177734,
      "memory(GiB)": 77.56,
      "step": 43225,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.440502
    },
    {
      "epoch": 1.8521057366865172,
      "grad_norm": 6.036388874053955,
      "learning_rate": 6.979666433326577e-05,
      "loss": 2.7429330825805662,
      "memory(GiB)": 77.56,
      "step": 43230,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.440509
    },
    {
      "epoch": 1.8523199520157663,
      "grad_norm": 6.380568504333496,
      "learning_rate": 6.979048434283218e-05,
      "loss": 2.595009994506836,
      "memory(GiB)": 77.56,
      "step": 43235,
      "token_acc": 0.4208754208754209,
      "train_speed(iter/s)": 1.440563
    },
    {
      "epoch": 1.852534167345015,
      "grad_norm": 4.630065441131592,
      "learning_rate": 6.978430399387541e-05,
      "loss": 2.4058111190795897,
      "memory(GiB)": 77.56,
      "step": 43240,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.44053
    },
    {
      "epoch": 1.8527483826742641,
      "grad_norm": 4.022320747375488,
      "learning_rate": 6.97781232865074e-05,
      "loss": 2.441282272338867,
      "memory(GiB)": 77.56,
      "step": 43245,
      "token_acc": 0.49166666666666664,
      "train_speed(iter/s)": 1.440477
    },
    {
      "epoch": 1.8529625980035132,
      "grad_norm": 4.732589244842529,
      "learning_rate": 6.977194222084013e-05,
      "loss": 2.520916748046875,
      "memory(GiB)": 77.56,
      "step": 43250,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.440478
    },
    {
      "epoch": 1.853176813332762,
      "grad_norm": 5.01125431060791,
      "learning_rate": 6.97657607969856e-05,
      "loss": 2.6507028579711913,
      "memory(GiB)": 77.56,
      "step": 43255,
      "token_acc": 0.4459016393442623,
      "train_speed(iter/s)": 1.440489
    },
    {
      "epoch": 1.8533910286620112,
      "grad_norm": 4.769619464874268,
      "learning_rate": 6.975957901505574e-05,
      "loss": 2.265201377868652,
      "memory(GiB)": 77.56,
      "step": 43260,
      "token_acc": 0.5409836065573771,
      "train_speed(iter/s)": 1.44052
    },
    {
      "epoch": 1.85360524399126,
      "grad_norm": 4.287899971008301,
      "learning_rate": 6.97533968751626e-05,
      "loss": 2.485826110839844,
      "memory(GiB)": 77.56,
      "step": 43265,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.440519
    },
    {
      "epoch": 1.8538194593205088,
      "grad_norm": 5.929051876068115,
      "learning_rate": 6.974721437741813e-05,
      "loss": 2.472885322570801,
      "memory(GiB)": 77.56,
      "step": 43270,
      "token_acc": 0.5176470588235295,
      "train_speed(iter/s)": 1.440488
    },
    {
      "epoch": 1.854033674649758,
      "grad_norm": 4.998896598815918,
      "learning_rate": 6.974103152193434e-05,
      "loss": 2.297836685180664,
      "memory(GiB)": 77.56,
      "step": 43275,
      "token_acc": 0.5390625,
      "train_speed(iter/s)": 1.440517
    },
    {
      "epoch": 1.854247889979007,
      "grad_norm": 5.720040798187256,
      "learning_rate": 6.973484830882326e-05,
      "loss": 2.609451675415039,
      "memory(GiB)": 77.56,
      "step": 43280,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.440492
    },
    {
      "epoch": 1.8544621053082557,
      "grad_norm": 4.926914691925049,
      "learning_rate": 6.97286647381969e-05,
      "loss": 2.4503416061401366,
      "memory(GiB)": 77.56,
      "step": 43285,
      "token_acc": 0.472,
      "train_speed(iter/s)": 1.440517
    },
    {
      "epoch": 1.854676320637505,
      "grad_norm": 6.048356533050537,
      "learning_rate": 6.972248081016724e-05,
      "loss": 2.396846580505371,
      "memory(GiB)": 77.56,
      "step": 43290,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.440521
    },
    {
      "epoch": 1.8548905359667538,
      "grad_norm": 4.139334201812744,
      "learning_rate": 6.971629652484635e-05,
      "loss": 2.3883827209472654,
      "memory(GiB)": 77.56,
      "step": 43295,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440521
    },
    {
      "epoch": 1.8551047512960026,
      "grad_norm": 4.844492435455322,
      "learning_rate": 6.971011188234628e-05,
      "loss": 2.719435119628906,
      "memory(GiB)": 77.56,
      "step": 43300,
      "token_acc": 0.4676470588235294,
      "train_speed(iter/s)": 1.440539
    },
    {
      "epoch": 1.8553189666252519,
      "grad_norm": 5.669625282287598,
      "learning_rate": 6.9703926882779e-05,
      "loss": 2.4841148376464846,
      "memory(GiB)": 77.56,
      "step": 43305,
      "token_acc": 0.4612903225806452,
      "train_speed(iter/s)": 1.440533
    },
    {
      "epoch": 1.8555331819545007,
      "grad_norm": 6.216773986816406,
      "learning_rate": 6.969774152625664e-05,
      "loss": 2.0978271484375,
      "memory(GiB)": 77.56,
      "step": 43310,
      "token_acc": 0.5103734439834025,
      "train_speed(iter/s)": 1.440557
    },
    {
      "epoch": 1.8557473972837495,
      "grad_norm": 5.27077579498291,
      "learning_rate": 6.969155581289119e-05,
      "loss": 2.484029006958008,
      "memory(GiB)": 77.56,
      "step": 43315,
      "token_acc": 0.49794238683127573,
      "train_speed(iter/s)": 1.440545
    },
    {
      "epoch": 1.8559616126129987,
      "grad_norm": 5.679560661315918,
      "learning_rate": 6.968536974279475e-05,
      "loss": 2.672632026672363,
      "memory(GiB)": 77.56,
      "step": 43320,
      "token_acc": 0.4177215189873418,
      "train_speed(iter/s)": 1.440582
    },
    {
      "epoch": 1.8561758279422476,
      "grad_norm": 3.9813849925994873,
      "learning_rate": 6.967918331607937e-05,
      "loss": 2.484807014465332,
      "memory(GiB)": 77.56,
      "step": 43325,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.440592
    },
    {
      "epoch": 1.8563900432714964,
      "grad_norm": 5.795016288757324,
      "learning_rate": 6.967299653285711e-05,
      "loss": 2.193868637084961,
      "memory(GiB)": 77.56,
      "step": 43330,
      "token_acc": 0.453125,
      "train_speed(iter/s)": 1.440593
    },
    {
      "epoch": 1.8566042586007456,
      "grad_norm": 4.828737258911133,
      "learning_rate": 6.966680939324006e-05,
      "loss": 2.525767707824707,
      "memory(GiB)": 77.56,
      "step": 43335,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.440615
    },
    {
      "epoch": 1.8568184739299944,
      "grad_norm": 5.104550838470459,
      "learning_rate": 6.966062189734033e-05,
      "loss": 2.5883762359619142,
      "memory(GiB)": 77.56,
      "step": 43340,
      "token_acc": 0.42671009771986973,
      "train_speed(iter/s)": 1.44063
    },
    {
      "epoch": 1.8570326892592433,
      "grad_norm": 4.25006628036499,
      "learning_rate": 6.965443404526998e-05,
      "loss": 2.370433044433594,
      "memory(GiB)": 77.56,
      "step": 43345,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.440611
    },
    {
      "epoch": 1.8572469045884925,
      "grad_norm": 4.21995735168457,
      "learning_rate": 6.964824583714111e-05,
      "loss": 2.446165084838867,
      "memory(GiB)": 77.56,
      "step": 43350,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.440623
    },
    {
      "epoch": 1.8574611199177413,
      "grad_norm": 4.266446113586426,
      "learning_rate": 6.964205727306586e-05,
      "loss": 2.655420112609863,
      "memory(GiB)": 77.56,
      "step": 43355,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.440629
    },
    {
      "epoch": 1.8576753352469901,
      "grad_norm": 4.508521556854248,
      "learning_rate": 6.963586835315629e-05,
      "loss": 2.5085685729980467,
      "memory(GiB)": 77.56,
      "step": 43360,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.440635
    },
    {
      "epoch": 1.8578895505762394,
      "grad_norm": 4.173177719116211,
      "learning_rate": 6.962967907752455e-05,
      "loss": 2.577441596984863,
      "memory(GiB)": 77.56,
      "step": 43365,
      "token_acc": 0.4668769716088328,
      "train_speed(iter/s)": 1.440649
    },
    {
      "epoch": 1.8581037659054882,
      "grad_norm": 4.742574214935303,
      "learning_rate": 6.962348944628276e-05,
      "loss": 2.2691776275634767,
      "memory(GiB)": 77.56,
      "step": 43370,
      "token_acc": 0.48563218390804597,
      "train_speed(iter/s)": 1.44067
    },
    {
      "epoch": 1.858317981234737,
      "grad_norm": 6.976690292358398,
      "learning_rate": 6.961729945954307e-05,
      "loss": 2.68206787109375,
      "memory(GiB)": 77.56,
      "step": 43375,
      "token_acc": 0.4610169491525424,
      "train_speed(iter/s)": 1.440676
    },
    {
      "epoch": 1.8585321965639863,
      "grad_norm": 5.838136196136475,
      "learning_rate": 6.961110911741757e-05,
      "loss": 2.4876346588134766,
      "memory(GiB)": 77.56,
      "step": 43380,
      "token_acc": 0.4570552147239264,
      "train_speed(iter/s)": 1.440655
    },
    {
      "epoch": 1.858746411893235,
      "grad_norm": 4.771737575531006,
      "learning_rate": 6.960491842001846e-05,
      "loss": 2.4716320037841797,
      "memory(GiB)": 77.56,
      "step": 43385,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.440704
    },
    {
      "epoch": 1.858960627222484,
      "grad_norm": 6.208278656005859,
      "learning_rate": 6.959872736745784e-05,
      "loss": 2.4613088607788085,
      "memory(GiB)": 77.56,
      "step": 43390,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.4407
    },
    {
      "epoch": 1.8591748425517332,
      "grad_norm": 4.553323268890381,
      "learning_rate": 6.95925359598479e-05,
      "loss": 2.575738525390625,
      "memory(GiB)": 77.56,
      "step": 43395,
      "token_acc": 0.4682080924855491,
      "train_speed(iter/s)": 1.44075
    },
    {
      "epoch": 1.859389057880982,
      "grad_norm": 4.627508640289307,
      "learning_rate": 6.95863441973008e-05,
      "loss": 2.4983247756958007,
      "memory(GiB)": 77.56,
      "step": 43400,
      "token_acc": 0.46062992125984253,
      "train_speed(iter/s)": 1.440773
    },
    {
      "epoch": 1.8596032732102308,
      "grad_norm": 6.814741611480713,
      "learning_rate": 6.958015207992867e-05,
      "loss": 2.373073387145996,
      "memory(GiB)": 77.56,
      "step": 43405,
      "token_acc": 0.4509090909090909,
      "train_speed(iter/s)": 1.440789
    },
    {
      "epoch": 1.85981748853948,
      "grad_norm": 5.269682884216309,
      "learning_rate": 6.957395960784374e-05,
      "loss": 2.832663154602051,
      "memory(GiB)": 77.56,
      "step": 43410,
      "token_acc": 0.43656716417910446,
      "train_speed(iter/s)": 1.440806
    },
    {
      "epoch": 1.8600317038687288,
      "grad_norm": 5.185050964355469,
      "learning_rate": 6.956776678115817e-05,
      "loss": 2.564066696166992,
      "memory(GiB)": 77.56,
      "step": 43415,
      "token_acc": 0.4360655737704918,
      "train_speed(iter/s)": 1.440859
    },
    {
      "epoch": 1.8602459191979777,
      "grad_norm": 4.240634918212891,
      "learning_rate": 6.956157359998414e-05,
      "loss": 2.3420948028564452,
      "memory(GiB)": 77.56,
      "step": 43420,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.44086
    },
    {
      "epoch": 1.860460134527227,
      "grad_norm": 6.516208648681641,
      "learning_rate": 6.955538006443386e-05,
      "loss": 2.530893898010254,
      "memory(GiB)": 77.56,
      "step": 43425,
      "token_acc": 0.4775510204081633,
      "train_speed(iter/s)": 1.440844
    },
    {
      "epoch": 1.8606743498564757,
      "grad_norm": 4.886407375335693,
      "learning_rate": 6.954918617461952e-05,
      "loss": 2.5489376068115233,
      "memory(GiB)": 77.56,
      "step": 43430,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.440846
    },
    {
      "epoch": 1.8608885651857245,
      "grad_norm": 5.054930686950684,
      "learning_rate": 6.954299193065334e-05,
      "loss": 2.8205469131469725,
      "memory(GiB)": 77.56,
      "step": 43435,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.44083
    },
    {
      "epoch": 1.8611027805149738,
      "grad_norm": 4.837130069732666,
      "learning_rate": 6.953679733264753e-05,
      "loss": 2.5725114822387694,
      "memory(GiB)": 77.56,
      "step": 43440,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.440827
    },
    {
      "epoch": 1.8613169958442226,
      "grad_norm": 5.3324503898620605,
      "learning_rate": 6.95306023807143e-05,
      "loss": 2.5039031982421873,
      "memory(GiB)": 77.56,
      "step": 43445,
      "token_acc": 0.4897119341563786,
      "train_speed(iter/s)": 1.440839
    },
    {
      "epoch": 1.8615312111734714,
      "grad_norm": 4.094564437866211,
      "learning_rate": 6.952440707496589e-05,
      "loss": 2.612009620666504,
      "memory(GiB)": 77.56,
      "step": 43450,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.440869
    },
    {
      "epoch": 1.8617454265027207,
      "grad_norm": 4.587902069091797,
      "learning_rate": 6.951821141551455e-05,
      "loss": 2.530088424682617,
      "memory(GiB)": 77.56,
      "step": 43455,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.44089
    },
    {
      "epoch": 1.8619596418319695,
      "grad_norm": 5.007899284362793,
      "learning_rate": 6.951201540247249e-05,
      "loss": 2.371768760681152,
      "memory(GiB)": 77.56,
      "step": 43460,
      "token_acc": 0.5064102564102564,
      "train_speed(iter/s)": 1.440896
    },
    {
      "epoch": 1.8621738571612183,
      "grad_norm": 4.972559452056885,
      "learning_rate": 6.950581903595194e-05,
      "loss": 2.7083065032958986,
      "memory(GiB)": 77.56,
      "step": 43465,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.440931
    },
    {
      "epoch": 1.8623880724904676,
      "grad_norm": 4.13785457611084,
      "learning_rate": 6.949962231606522e-05,
      "loss": 2.3697992324829102,
      "memory(GiB)": 77.56,
      "step": 43470,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.440895
    },
    {
      "epoch": 1.8626022878197164,
      "grad_norm": 4.663508892059326,
      "learning_rate": 6.949342524292453e-05,
      "loss": 2.1297985076904298,
      "memory(GiB)": 77.56,
      "step": 43475,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.4409
    },
    {
      "epoch": 1.8628165031489652,
      "grad_norm": 4.668050765991211,
      "learning_rate": 6.948722781664215e-05,
      "loss": 2.5469255447387695,
      "memory(GiB)": 77.56,
      "step": 43480,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.440853
    },
    {
      "epoch": 1.8630307184782144,
      "grad_norm": 4.261608123779297,
      "learning_rate": 6.948103003733036e-05,
      "loss": 2.676869773864746,
      "memory(GiB)": 77.56,
      "step": 43485,
      "token_acc": 0.414985590778098,
      "train_speed(iter/s)": 1.440859
    },
    {
      "epoch": 1.8632449338074633,
      "grad_norm": 5.871016025543213,
      "learning_rate": 6.947483190510144e-05,
      "loss": 2.5678970336914064,
      "memory(GiB)": 77.56,
      "step": 43490,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.440849
    },
    {
      "epoch": 1.863459149136712,
      "grad_norm": 5.957498550415039,
      "learning_rate": 6.946863342006768e-05,
      "loss": 2.7108821868896484,
      "memory(GiB)": 77.56,
      "step": 43495,
      "token_acc": 0.471976401179941,
      "train_speed(iter/s)": 1.440827
    },
    {
      "epoch": 1.8636733644659613,
      "grad_norm": 8.545435905456543,
      "learning_rate": 6.946243458234135e-05,
      "loss": 2.5306770324707033,
      "memory(GiB)": 77.56,
      "step": 43500,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.44086
    },
    {
      "epoch": 1.8636733644659613,
      "eval_loss": 2.236259698867798,
      "eval_runtime": 13.7364,
      "eval_samples_per_second": 7.28,
      "eval_steps_per_second": 7.28,
      "eval_token_acc": 0.4917541229385307,
      "step": 43500
    },
    {
      "epoch": 1.8638875797952101,
      "grad_norm": 3.7839760780334473,
      "learning_rate": 6.945623539203477e-05,
      "loss": 2.7957149505615235,
      "memory(GiB)": 77.56,
      "step": 43505,
      "token_acc": 0.477319587628866,
      "train_speed(iter/s)": 1.440141
    },
    {
      "epoch": 1.864101795124459,
      "grad_norm": 11.7059965133667,
      "learning_rate": 6.945003584926021e-05,
      "loss": 2.688629150390625,
      "memory(GiB)": 77.56,
      "step": 43510,
      "token_acc": 0.4261168384879725,
      "train_speed(iter/s)": 1.440121
    },
    {
      "epoch": 1.8643160104537082,
      "grad_norm": 5.366820812225342,
      "learning_rate": 6.944383595413003e-05,
      "loss": 2.7319902420043944,
      "memory(GiB)": 77.56,
      "step": 43515,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.440143
    },
    {
      "epoch": 1.864530225782957,
      "grad_norm": 4.514744281768799,
      "learning_rate": 6.943763570675652e-05,
      "loss": 2.6621368408203123,
      "memory(GiB)": 77.56,
      "step": 43520,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.440183
    },
    {
      "epoch": 1.8647444411122058,
      "grad_norm": 5.274441242218018,
      "learning_rate": 6.943143510725199e-05,
      "loss": 2.4516849517822266,
      "memory(GiB)": 77.56,
      "step": 43525,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.440136
    },
    {
      "epoch": 1.864958656441455,
      "grad_norm": 5.445152759552002,
      "learning_rate": 6.942523415572882e-05,
      "loss": 2.5138612747192384,
      "memory(GiB)": 77.56,
      "step": 43530,
      "token_acc": 0.5032258064516129,
      "train_speed(iter/s)": 1.440151
    },
    {
      "epoch": 1.865172871770704,
      "grad_norm": 7.207469463348389,
      "learning_rate": 6.941903285229927e-05,
      "loss": 2.7002941131591798,
      "memory(GiB)": 77.56,
      "step": 43535,
      "token_acc": 0.44776119402985076,
      "train_speed(iter/s)": 1.440097
    },
    {
      "epoch": 1.8653870870999527,
      "grad_norm": 3.9434337615966797,
      "learning_rate": 6.941283119707575e-05,
      "loss": 2.586966705322266,
      "memory(GiB)": 77.56,
      "step": 43540,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.440082
    },
    {
      "epoch": 1.865601302429202,
      "grad_norm": 5.480777263641357,
      "learning_rate": 6.940662919017057e-05,
      "loss": 2.4758493423461916,
      "memory(GiB)": 77.56,
      "step": 43545,
      "token_acc": 0.5053763440860215,
      "train_speed(iter/s)": 1.440086
    },
    {
      "epoch": 1.8658155177584508,
      "grad_norm": 4.904507637023926,
      "learning_rate": 6.940042683169609e-05,
      "loss": 2.491636848449707,
      "memory(GiB)": 77.56,
      "step": 43550,
      "token_acc": 0.5080385852090032,
      "train_speed(iter/s)": 1.440107
    },
    {
      "epoch": 1.8660297330876996,
      "grad_norm": 5.072583198547363,
      "learning_rate": 6.939422412176468e-05,
      "loss": 2.304940032958984,
      "memory(GiB)": 77.56,
      "step": 43555,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.440113
    },
    {
      "epoch": 1.8662439484169489,
      "grad_norm": 7.3964643478393555,
      "learning_rate": 6.938802106048873e-05,
      "loss": 2.4718551635742188,
      "memory(GiB)": 77.56,
      "step": 43560,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.440127
    },
    {
      "epoch": 1.8664581637461977,
      "grad_norm": 4.174238681793213,
      "learning_rate": 6.938181764798058e-05,
      "loss": 2.856204605102539,
      "memory(GiB)": 77.56,
      "step": 43565,
      "token_acc": 0.49221183800623053,
      "train_speed(iter/s)": 1.440179
    },
    {
      "epoch": 1.8666723790754465,
      "grad_norm": 5.73566198348999,
      "learning_rate": 6.93756138843526e-05,
      "loss": 2.6958240509033202,
      "memory(GiB)": 77.56,
      "step": 43570,
      "token_acc": 0.4117647058823529,
      "train_speed(iter/s)": 1.440113
    },
    {
      "epoch": 1.8668865944046957,
      "grad_norm": 5.127898216247559,
      "learning_rate": 6.936940976971724e-05,
      "loss": 2.5017662048339844,
      "memory(GiB)": 77.56,
      "step": 43575,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.440146
    },
    {
      "epoch": 1.8671008097339445,
      "grad_norm": 4.297633171081543,
      "learning_rate": 6.936320530418681e-05,
      "loss": 2.6987648010253906,
      "memory(GiB)": 77.56,
      "step": 43580,
      "token_acc": 0.46439628482972134,
      "train_speed(iter/s)": 1.440151
    },
    {
      "epoch": 1.8673150250631934,
      "grad_norm": 5.929309844970703,
      "learning_rate": 6.935700048787379e-05,
      "loss": 2.6566654205322267,
      "memory(GiB)": 77.56,
      "step": 43585,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.44016
    },
    {
      "epoch": 1.8675292403924426,
      "grad_norm": 5.273189067840576,
      "learning_rate": 6.935079532089052e-05,
      "loss": 2.542314147949219,
      "memory(GiB)": 77.56,
      "step": 43590,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.440165
    },
    {
      "epoch": 1.8677434557216914,
      "grad_norm": 6.269961833953857,
      "learning_rate": 6.934458980334944e-05,
      "loss": 2.670563507080078,
      "memory(GiB)": 77.56,
      "step": 43595,
      "token_acc": 0.44106463878326996,
      "train_speed(iter/s)": 1.440175
    },
    {
      "epoch": 1.8679576710509405,
      "grad_norm": 4.557129859924316,
      "learning_rate": 6.933838393536299e-05,
      "loss": 2.5225624084472655,
      "memory(GiB)": 77.56,
      "step": 43600,
      "token_acc": 0.48011363636363635,
      "train_speed(iter/s)": 1.440165
    },
    {
      "epoch": 1.8681718863801895,
      "grad_norm": 4.752987861633301,
      "learning_rate": 6.933217771704356e-05,
      "loss": 2.380077362060547,
      "memory(GiB)": 77.56,
      "step": 43605,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.44017
    },
    {
      "epoch": 1.8683861017094383,
      "grad_norm": 4.288316249847412,
      "learning_rate": 6.932597114850359e-05,
      "loss": 2.9540658950805665,
      "memory(GiB)": 77.56,
      "step": 43610,
      "token_acc": 0.41317365269461076,
      "train_speed(iter/s)": 1.440146
    },
    {
      "epoch": 1.8686003170386873,
      "grad_norm": 5.97092342376709,
      "learning_rate": 6.931976422985554e-05,
      "loss": 2.494019317626953,
      "memory(GiB)": 77.56,
      "step": 43615,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.440171
    },
    {
      "epoch": 1.8688145323679364,
      "grad_norm": 5.262835502624512,
      "learning_rate": 6.931355696121183e-05,
      "loss": 2.2457048416137697,
      "memory(GiB)": 77.56,
      "step": 43620,
      "token_acc": 0.5259259259259259,
      "train_speed(iter/s)": 1.440156
    },
    {
      "epoch": 1.8690287476971852,
      "grad_norm": 5.038191318511963,
      "learning_rate": 6.930734934268491e-05,
      "loss": 2.470929718017578,
      "memory(GiB)": 77.56,
      "step": 43625,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.440107
    },
    {
      "epoch": 1.8692429630264342,
      "grad_norm": 6.955289363861084,
      "learning_rate": 6.930114137438725e-05,
      "loss": 2.6117916107177734,
      "memory(GiB)": 77.56,
      "step": 43630,
      "token_acc": 0.43686006825938567,
      "train_speed(iter/s)": 1.440112
    },
    {
      "epoch": 1.8694571783556833,
      "grad_norm": 5.306889057159424,
      "learning_rate": 6.92949330564313e-05,
      "loss": 2.4621322631835936,
      "memory(GiB)": 77.56,
      "step": 43635,
      "token_acc": 0.47265625,
      "train_speed(iter/s)": 1.440119
    },
    {
      "epoch": 1.869671393684932,
      "grad_norm": 5.303523063659668,
      "learning_rate": 6.928872438892956e-05,
      "loss": 2.3270092010498047,
      "memory(GiB)": 77.56,
      "step": 43640,
      "token_acc": 0.4983277591973244,
      "train_speed(iter/s)": 1.440118
    },
    {
      "epoch": 1.869885609014181,
      "grad_norm": 3.8588778972625732,
      "learning_rate": 6.928251537199446e-05,
      "loss": 2.6262319564819334,
      "memory(GiB)": 77.56,
      "step": 43645,
      "token_acc": 0.4386422976501306,
      "train_speed(iter/s)": 1.440142
    },
    {
      "epoch": 1.8700998243434301,
      "grad_norm": 4.734883785247803,
      "learning_rate": 6.927630600573851e-05,
      "loss": 2.570682716369629,
      "memory(GiB)": 77.56,
      "step": 43650,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.440157
    },
    {
      "epoch": 1.870314039672679,
      "grad_norm": 4.64441442489624,
      "learning_rate": 6.927009629027421e-05,
      "loss": 2.217654800415039,
      "memory(GiB)": 77.56,
      "step": 43655,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.440162
    },
    {
      "epoch": 1.870528255001928,
      "grad_norm": 6.077085494995117,
      "learning_rate": 6.926388622571403e-05,
      "loss": 2.426427459716797,
      "memory(GiB)": 77.56,
      "step": 43660,
      "token_acc": 0.4962121212121212,
      "train_speed(iter/s)": 1.440191
    },
    {
      "epoch": 1.870742470331177,
      "grad_norm": 4.232110977172852,
      "learning_rate": 6.925767581217046e-05,
      "loss": 2.2968572616577148,
      "memory(GiB)": 77.56,
      "step": 43665,
      "token_acc": 0.516728624535316,
      "train_speed(iter/s)": 1.440222
    },
    {
      "epoch": 1.8709566856604258,
      "grad_norm": 3.608761787414551,
      "learning_rate": 6.925146504975606e-05,
      "loss": 2.242470932006836,
      "memory(GiB)": 77.56,
      "step": 43670,
      "token_acc": 0.5113636363636364,
      "train_speed(iter/s)": 1.440212
    },
    {
      "epoch": 1.8711709009896749,
      "grad_norm": 5.739598751068115,
      "learning_rate": 6.92452539385833e-05,
      "loss": 2.247125434875488,
      "memory(GiB)": 77.56,
      "step": 43675,
      "token_acc": 0.5053763440860215,
      "train_speed(iter/s)": 1.440188
    },
    {
      "epoch": 1.871385116318924,
      "grad_norm": 7.395709037780762,
      "learning_rate": 6.92390424787647e-05,
      "loss": 2.343984603881836,
      "memory(GiB)": 77.56,
      "step": 43680,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.440166
    },
    {
      "epoch": 1.8715993316481727,
      "grad_norm": 4.652615547180176,
      "learning_rate": 6.923283067041282e-05,
      "loss": 2.400879669189453,
      "memory(GiB)": 77.56,
      "step": 43685,
      "token_acc": 0.5021097046413502,
      "train_speed(iter/s)": 1.440169
    },
    {
      "epoch": 1.8718135469774217,
      "grad_norm": 7.7594146728515625,
      "learning_rate": 6.922661851364016e-05,
      "loss": 2.393063545227051,
      "memory(GiB)": 77.56,
      "step": 43690,
      "token_acc": 0.4197080291970803,
      "train_speed(iter/s)": 1.440181
    },
    {
      "epoch": 1.8720277623066708,
      "grad_norm": 5.022670269012451,
      "learning_rate": 6.922040600855926e-05,
      "loss": 2.5054611206054687,
      "memory(GiB)": 77.56,
      "step": 43695,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.440162
    },
    {
      "epoch": 1.8722419776359196,
      "grad_norm": 5.376389503479004,
      "learning_rate": 6.921419315528268e-05,
      "loss": 2.703806686401367,
      "memory(GiB)": 77.56,
      "step": 43700,
      "token_acc": 0.44569288389513106,
      "train_speed(iter/s)": 1.440207
    },
    {
      "epoch": 1.8724561929651686,
      "grad_norm": 6.1688761711120605,
      "learning_rate": 6.920797995392297e-05,
      "loss": 2.550992965698242,
      "memory(GiB)": 77.56,
      "step": 43705,
      "token_acc": 0.41638225255972694,
      "train_speed(iter/s)": 1.440213
    },
    {
      "epoch": 1.8726704082944177,
      "grad_norm": 4.517484188079834,
      "learning_rate": 6.920176640459268e-05,
      "loss": 2.6653520584106447,
      "memory(GiB)": 77.56,
      "step": 43710,
      "token_acc": 0.4613259668508287,
      "train_speed(iter/s)": 1.44022
    },
    {
      "epoch": 1.8728846236236665,
      "grad_norm": 4.672995567321777,
      "learning_rate": 6.919555250740439e-05,
      "loss": 2.6150575637817384,
      "memory(GiB)": 77.56,
      "step": 43715,
      "token_acc": 0.4440993788819876,
      "train_speed(iter/s)": 1.440217
    },
    {
      "epoch": 1.8730988389529155,
      "grad_norm": 8.136906623840332,
      "learning_rate": 6.918933826247065e-05,
      "loss": 2.590870666503906,
      "memory(GiB)": 77.56,
      "step": 43720,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.440225
    },
    {
      "epoch": 1.8733130542821645,
      "grad_norm": 6.360536575317383,
      "learning_rate": 6.918312366990405e-05,
      "loss": 2.6951112747192383,
      "memory(GiB)": 77.56,
      "step": 43725,
      "token_acc": 0.43234323432343236,
      "train_speed(iter/s)": 1.440246
    },
    {
      "epoch": 1.8735272696114134,
      "grad_norm": 4.464033126831055,
      "learning_rate": 6.917690872981717e-05,
      "loss": 2.6590789794921874,
      "memory(GiB)": 77.56,
      "step": 43730,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.44024
    },
    {
      "epoch": 1.8737414849406624,
      "grad_norm": 4.957088947296143,
      "learning_rate": 6.917069344232258e-05,
      "loss": 2.2770597457885744,
      "memory(GiB)": 77.56,
      "step": 43735,
      "token_acc": 0.5210727969348659,
      "train_speed(iter/s)": 1.440277
    },
    {
      "epoch": 1.8739557002699114,
      "grad_norm": 6.839737892150879,
      "learning_rate": 6.916447780753291e-05,
      "loss": 2.5905824661254884,
      "memory(GiB)": 77.56,
      "step": 43740,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.440221
    },
    {
      "epoch": 1.8741699155991602,
      "grad_norm": 4.120622158050537,
      "learning_rate": 6.915826182556075e-05,
      "loss": 2.385251998901367,
      "memory(GiB)": 77.56,
      "step": 43745,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.440217
    },
    {
      "epoch": 1.8743841309284093,
      "grad_norm": 8.104998588562012,
      "learning_rate": 6.915204549651871e-05,
      "loss": 2.60362548828125,
      "memory(GiB)": 77.56,
      "step": 43750,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.440196
    },
    {
      "epoch": 1.8745983462576583,
      "grad_norm": 4.279297828674316,
      "learning_rate": 6.914582882051938e-05,
      "loss": 2.6403614044189454,
      "memory(GiB)": 77.56,
      "step": 43755,
      "token_acc": 0.44299674267100975,
      "train_speed(iter/s)": 1.440199
    },
    {
      "epoch": 1.8748125615869071,
      "grad_norm": 6.763187885284424,
      "learning_rate": 6.913961179767543e-05,
      "loss": 2.311290740966797,
      "memory(GiB)": 77.56,
      "step": 43760,
      "token_acc": 0.53515625,
      "train_speed(iter/s)": 1.44022
    },
    {
      "epoch": 1.8750267769161562,
      "grad_norm": 6.343144416809082,
      "learning_rate": 6.913339442809942e-05,
      "loss": 2.5140384674072265,
      "memory(GiB)": 77.56,
      "step": 43765,
      "token_acc": 0.4980237154150198,
      "train_speed(iter/s)": 1.440263
    },
    {
      "epoch": 1.8752409922454052,
      "grad_norm": 4.878091335296631,
      "learning_rate": 6.912717671190407e-05,
      "loss": 2.728389358520508,
      "memory(GiB)": 77.56,
      "step": 43770,
      "token_acc": 0.4970414201183432,
      "train_speed(iter/s)": 1.440311
    },
    {
      "epoch": 1.875455207574654,
      "grad_norm": 5.390368461608887,
      "learning_rate": 6.912095864920193e-05,
      "loss": 2.3072967529296875,
      "memory(GiB)": 77.56,
      "step": 43775,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.440308
    },
    {
      "epoch": 1.875669422903903,
      "grad_norm": 4.705453395843506,
      "learning_rate": 6.911474024010569e-05,
      "loss": 2.5538101196289062,
      "memory(GiB)": 77.56,
      "step": 43780,
      "token_acc": 0.4815950920245399,
      "train_speed(iter/s)": 1.440295
    },
    {
      "epoch": 1.875883638233152,
      "grad_norm": 5.005934715270996,
      "learning_rate": 6.910852148472802e-05,
      "loss": 2.5337921142578126,
      "memory(GiB)": 77.56,
      "step": 43785,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.440297
    },
    {
      "epoch": 1.8760978535624009,
      "grad_norm": 4.590370178222656,
      "learning_rate": 6.910230238318157e-05,
      "loss": 2.1834234237670898,
      "memory(GiB)": 77.56,
      "step": 43790,
      "token_acc": 0.5409252669039146,
      "train_speed(iter/s)": 1.440322
    },
    {
      "epoch": 1.87631206889165,
      "grad_norm": 4.759270191192627,
      "learning_rate": 6.909608293557895e-05,
      "loss": 2.4583187103271484,
      "memory(GiB)": 77.56,
      "step": 43795,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.440292
    },
    {
      "epoch": 1.876526284220899,
      "grad_norm": 4.1492085456848145,
      "learning_rate": 6.908986314203289e-05,
      "loss": 2.724493980407715,
      "memory(GiB)": 77.56,
      "step": 43800,
      "token_acc": 0.42507645259938837,
      "train_speed(iter/s)": 1.440286
    },
    {
      "epoch": 1.8767404995501478,
      "grad_norm": 4.204627513885498,
      "learning_rate": 6.908364300265607e-05,
      "loss": 2.418387603759766,
      "memory(GiB)": 77.56,
      "step": 43805,
      "token_acc": 0.46099290780141844,
      "train_speed(iter/s)": 1.440293
    },
    {
      "epoch": 1.8769547148793968,
      "grad_norm": 5.65638542175293,
      "learning_rate": 6.907742251756114e-05,
      "loss": 2.375998306274414,
      "memory(GiB)": 77.56,
      "step": 43810,
      "token_acc": 0.5444839857651246,
      "train_speed(iter/s)": 1.44027
    },
    {
      "epoch": 1.8771689302086458,
      "grad_norm": 3.6777758598327637,
      "learning_rate": 6.907120168686079e-05,
      "loss": 2.1566015243530274,
      "memory(GiB)": 77.56,
      "step": 43815,
      "token_acc": 0.5214007782101168,
      "train_speed(iter/s)": 1.440262
    },
    {
      "epoch": 1.8773831455378946,
      "grad_norm": 5.5765790939331055,
      "learning_rate": 6.906498051066776e-05,
      "loss": 2.375243377685547,
      "memory(GiB)": 77.56,
      "step": 43820,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.440283
    },
    {
      "epoch": 1.8775973608671437,
      "grad_norm": 4.720075607299805,
      "learning_rate": 6.90587589890947e-05,
      "loss": 2.635455322265625,
      "memory(GiB)": 77.56,
      "step": 43825,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.440304
    },
    {
      "epoch": 1.8778115761963927,
      "grad_norm": 5.195075035095215,
      "learning_rate": 6.905253712225436e-05,
      "loss": 2.779643249511719,
      "memory(GiB)": 77.56,
      "step": 43830,
      "token_acc": 0.4351145038167939,
      "train_speed(iter/s)": 1.440226
    },
    {
      "epoch": 1.8780257915256415,
      "grad_norm": 3.975156545639038,
      "learning_rate": 6.904631491025945e-05,
      "loss": 2.816888427734375,
      "memory(GiB)": 77.56,
      "step": 43835,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.440253
    },
    {
      "epoch": 1.8782400068548906,
      "grad_norm": 4.649935245513916,
      "learning_rate": 6.904009235322265e-05,
      "loss": 2.6185672760009764,
      "memory(GiB)": 77.56,
      "step": 43840,
      "token_acc": 0.43670886075949367,
      "train_speed(iter/s)": 1.440252
    },
    {
      "epoch": 1.8784542221841396,
      "grad_norm": 6.371175289154053,
      "learning_rate": 6.903386945125673e-05,
      "loss": 2.7402807235717774,
      "memory(GiB)": 77.56,
      "step": 43845,
      "token_acc": 0.45121951219512196,
      "train_speed(iter/s)": 1.440244
    },
    {
      "epoch": 1.8786684375133884,
      "grad_norm": 5.791367530822754,
      "learning_rate": 6.90276462044744e-05,
      "loss": 2.267729949951172,
      "memory(GiB)": 77.56,
      "step": 43850,
      "token_acc": 0.5147540983606558,
      "train_speed(iter/s)": 1.440224
    },
    {
      "epoch": 1.8788826528426374,
      "grad_norm": 6.075738906860352,
      "learning_rate": 6.90214226129884e-05,
      "loss": 2.1494869232177733,
      "memory(GiB)": 77.56,
      "step": 43855,
      "token_acc": 0.5021645021645021,
      "train_speed(iter/s)": 1.440254
    },
    {
      "epoch": 1.8790968681718865,
      "grad_norm": 5.984035968780518,
      "learning_rate": 6.901519867691151e-05,
      "loss": 2.7059022903442385,
      "memory(GiB)": 77.56,
      "step": 43860,
      "token_acc": 0.44981412639405205,
      "train_speed(iter/s)": 1.440264
    },
    {
      "epoch": 1.8793110835011353,
      "grad_norm": 4.501453399658203,
      "learning_rate": 6.900897439635646e-05,
      "loss": 2.630645751953125,
      "memory(GiB)": 77.56,
      "step": 43865,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.440267
    },
    {
      "epoch": 1.8795252988303843,
      "grad_norm": 5.558224678039551,
      "learning_rate": 6.900274977143599e-05,
      "loss": 2.4991966247558595,
      "memory(GiB)": 77.56,
      "step": 43870,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.440291
    },
    {
      "epoch": 1.8797395141596334,
      "grad_norm": 4.4733757972717285,
      "learning_rate": 6.89965248022629e-05,
      "loss": 2.2572797775268554,
      "memory(GiB)": 77.56,
      "step": 43875,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.44025
    },
    {
      "epoch": 1.8799537294888822,
      "grad_norm": 4.89325475692749,
      "learning_rate": 6.899029948894993e-05,
      "loss": 2.7734336853027344,
      "memory(GiB)": 77.56,
      "step": 43880,
      "token_acc": 0.43884892086330934,
      "train_speed(iter/s)": 1.440247
    },
    {
      "epoch": 1.8801679448181312,
      "grad_norm": 7.597643852233887,
      "learning_rate": 6.898407383160985e-05,
      "loss": 2.5331729888916015,
      "memory(GiB)": 77.56,
      "step": 43885,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.440274
    },
    {
      "epoch": 1.8803821601473802,
      "grad_norm": 4.127545356750488,
      "learning_rate": 6.89778478303555e-05,
      "loss": 2.31573486328125,
      "memory(GiB)": 77.56,
      "step": 43890,
      "token_acc": 0.49201277955271566,
      "train_speed(iter/s)": 1.440292
    },
    {
      "epoch": 1.880596375476629,
      "grad_norm": 7.149901390075684,
      "learning_rate": 6.897162148529963e-05,
      "loss": 2.661090278625488,
      "memory(GiB)": 77.56,
      "step": 43895,
      "token_acc": 0.44370860927152317,
      "train_speed(iter/s)": 1.440327
    },
    {
      "epoch": 1.880810590805878,
      "grad_norm": 6.092045783996582,
      "learning_rate": 6.896539479655501e-05,
      "loss": 2.7160852432250975,
      "memory(GiB)": 77.56,
      "step": 43900,
      "token_acc": 0.426056338028169,
      "train_speed(iter/s)": 1.44033
    },
    {
      "epoch": 1.8810248061351271,
      "grad_norm": 5.126265048980713,
      "learning_rate": 6.89591677642345e-05,
      "loss": 2.7058479309082033,
      "memory(GiB)": 77.56,
      "step": 43905,
      "token_acc": 0.4766355140186916,
      "train_speed(iter/s)": 1.440329
    },
    {
      "epoch": 1.881239021464376,
      "grad_norm": 5.003473281860352,
      "learning_rate": 6.895294038845087e-05,
      "loss": 2.278493118286133,
      "memory(GiB)": 77.56,
      "step": 43910,
      "token_acc": 0.49850746268656715,
      "train_speed(iter/s)": 1.440315
    },
    {
      "epoch": 1.881453236793625,
      "grad_norm": 4.984987735748291,
      "learning_rate": 6.894671266931697e-05,
      "loss": 2.4715904235839843,
      "memory(GiB)": 77.56,
      "step": 43915,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.440358
    },
    {
      "epoch": 1.881667452122874,
      "grad_norm": 4.243618488311768,
      "learning_rate": 6.894048460694557e-05,
      "loss": 2.678019332885742,
      "memory(GiB)": 77.56,
      "step": 43920,
      "token_acc": 0.42990654205607476,
      "train_speed(iter/s)": 1.440369
    },
    {
      "epoch": 1.8818816674521228,
      "grad_norm": 4.783170223236084,
      "learning_rate": 6.893425620144952e-05,
      "loss": 2.2035579681396484,
      "memory(GiB)": 77.56,
      "step": 43925,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.44037
    },
    {
      "epoch": 1.8820958827813719,
      "grad_norm": 4.1528754234313965,
      "learning_rate": 6.892802745294165e-05,
      "loss": 2.6755613327026366,
      "memory(GiB)": 77.56,
      "step": 43930,
      "token_acc": 0.4131944444444444,
      "train_speed(iter/s)": 1.440378
    },
    {
      "epoch": 1.8823100981106209,
      "grad_norm": 9.812792778015137,
      "learning_rate": 6.892179836153483e-05,
      "loss": 2.6304904937744142,
      "memory(GiB)": 77.56,
      "step": 43935,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.440382
    },
    {
      "epoch": 1.8825243134398697,
      "grad_norm": 5.402609825134277,
      "learning_rate": 6.891556892734188e-05,
      "loss": 2.4052761077880858,
      "memory(GiB)": 77.56,
      "step": 43940,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.440394
    },
    {
      "epoch": 1.8827385287691187,
      "grad_norm": 5.9622979164123535,
      "learning_rate": 6.890933915047565e-05,
      "loss": 2.1981311798095704,
      "memory(GiB)": 77.56,
      "step": 43945,
      "token_acc": 0.5236363636363637,
      "train_speed(iter/s)": 1.440424
    },
    {
      "epoch": 1.8829527440983678,
      "grad_norm": 5.090926647186279,
      "learning_rate": 6.8903109031049e-05,
      "loss": 2.7230417251586916,
      "memory(GiB)": 77.56,
      "step": 43950,
      "token_acc": 0.4307692307692308,
      "train_speed(iter/s)": 1.440456
    },
    {
      "epoch": 1.8831669594276166,
      "grad_norm": 5.329853534698486,
      "learning_rate": 6.88968785691748e-05,
      "loss": 2.1352123260498046,
      "memory(GiB)": 77.56,
      "step": 43955,
      "token_acc": 0.5245901639344263,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.8833811747568656,
      "grad_norm": 5.130246162414551,
      "learning_rate": 6.88906477649659e-05,
      "loss": 2.2296253204345704,
      "memory(GiB)": 77.56,
      "step": 43960,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.440496
    },
    {
      "epoch": 1.8835953900861147,
      "grad_norm": 4.083864212036133,
      "learning_rate": 6.888441661853523e-05,
      "loss": 2.340218734741211,
      "memory(GiB)": 77.56,
      "step": 43965,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.440489
    },
    {
      "epoch": 1.8838096054153635,
      "grad_norm": 5.540615558624268,
      "learning_rate": 6.887818512999562e-05,
      "loss": 2.4362293243408204,
      "memory(GiB)": 77.56,
      "step": 43970,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.440506
    },
    {
      "epoch": 1.8840238207446125,
      "grad_norm": 4.939748764038086,
      "learning_rate": 6.887195329945997e-05,
      "loss": 2.4583242416381834,
      "memory(GiB)": 77.56,
      "step": 43975,
      "token_acc": 0.4778761061946903,
      "train_speed(iter/s)": 1.440511
    },
    {
      "epoch": 1.8842380360738615,
      "grad_norm": 6.531278133392334,
      "learning_rate": 6.88657211270412e-05,
      "loss": 2.2615076065063477,
      "memory(GiB)": 77.56,
      "step": 43980,
      "token_acc": 0.5150375939849624,
      "train_speed(iter/s)": 1.44051
    },
    {
      "epoch": 1.8844522514031103,
      "grad_norm": 4.485387325286865,
      "learning_rate": 6.885948861285219e-05,
      "loss": 2.597138595581055,
      "memory(GiB)": 77.56,
      "step": 43985,
      "token_acc": 0.4702194357366771,
      "train_speed(iter/s)": 1.440501
    },
    {
      "epoch": 1.8846664667323594,
      "grad_norm": 5.031183242797852,
      "learning_rate": 6.885325575700584e-05,
      "loss": 2.4078433990478514,
      "memory(GiB)": 77.56,
      "step": 43990,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.44049
    },
    {
      "epoch": 1.8848806820616084,
      "grad_norm": 5.190521717071533,
      "learning_rate": 6.884702255961508e-05,
      "loss": 2.76110954284668,
      "memory(GiB)": 77.56,
      "step": 43995,
      "token_acc": 0.43465045592705165,
      "train_speed(iter/s)": 1.440474
    },
    {
      "epoch": 1.8850948973908572,
      "grad_norm": 5.4947590827941895,
      "learning_rate": 6.884078902079283e-05,
      "loss": 2.3637725830078127,
      "memory(GiB)": 77.56,
      "step": 44000,
      "token_acc": 0.5148148148148148,
      "train_speed(iter/s)": 1.440485
    },
    {
      "epoch": 1.8850948973908572,
      "eval_loss": 2.4237220287323,
      "eval_runtime": 14.7483,
      "eval_samples_per_second": 6.78,
      "eval_steps_per_second": 6.78,
      "eval_token_acc": 0.44549763033175355,
      "step": 44000
    },
    {
      "epoch": 1.8853091127201063,
      "grad_norm": 6.161308765411377,
      "learning_rate": 6.8834555140652e-05,
      "loss": 2.1780561447143554,
      "memory(GiB)": 77.56,
      "step": 44005,
      "token_acc": 0.45576407506702415,
      "train_speed(iter/s)": 1.439736
    },
    {
      "epoch": 1.8855233280493553,
      "grad_norm": 4.318418502807617,
      "learning_rate": 6.882832091930556e-05,
      "loss": 2.5019474029541016,
      "memory(GiB)": 77.56,
      "step": 44010,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.439723
    },
    {
      "epoch": 1.885737543378604,
      "grad_norm": 4.363212585449219,
      "learning_rate": 6.882208635686641e-05,
      "loss": 2.362237739562988,
      "memory(GiB)": 77.56,
      "step": 44015,
      "token_acc": 0.4896265560165975,
      "train_speed(iter/s)": 1.439713
    },
    {
      "epoch": 1.8859517587078531,
      "grad_norm": 5.084024429321289,
      "learning_rate": 6.881585145344751e-05,
      "loss": 2.341076469421387,
      "memory(GiB)": 77.56,
      "step": 44020,
      "token_acc": 0.5137931034482759,
      "train_speed(iter/s)": 1.439738
    },
    {
      "epoch": 1.8861659740371022,
      "grad_norm": 3.8032565116882324,
      "learning_rate": 6.88096162091618e-05,
      "loss": 2.510670471191406,
      "memory(GiB)": 77.56,
      "step": 44025,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.439706
    },
    {
      "epoch": 1.886380189366351,
      "grad_norm": 7.438320159912109,
      "learning_rate": 6.880338062412228e-05,
      "loss": 2.864375114440918,
      "memory(GiB)": 77.56,
      "step": 44030,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.439719
    },
    {
      "epoch": 1.8865944046956,
      "grad_norm": 5.129100322723389,
      "learning_rate": 6.879714469844185e-05,
      "loss": 2.4701171875,
      "memory(GiB)": 77.56,
      "step": 44035,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.439747
    },
    {
      "epoch": 1.886808620024849,
      "grad_norm": 4.694314002990723,
      "learning_rate": 6.879090843223353e-05,
      "loss": 2.3149692535400392,
      "memory(GiB)": 77.56,
      "step": 44040,
      "token_acc": 0.5305343511450382,
      "train_speed(iter/s)": 1.439737
    },
    {
      "epoch": 1.8870228353540979,
      "grad_norm": 4.69293212890625,
      "learning_rate": 6.878467182561026e-05,
      "loss": 2.3349449157714846,
      "memory(GiB)": 77.56,
      "step": 44045,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.439731
    },
    {
      "epoch": 1.887237050683347,
      "grad_norm": 5.171730041503906,
      "learning_rate": 6.877843487868505e-05,
      "loss": 2.3221935272216796,
      "memory(GiB)": 77.56,
      "step": 44050,
      "token_acc": 0.5259259259259259,
      "train_speed(iter/s)": 1.439745
    },
    {
      "epoch": 1.887451266012596,
      "grad_norm": 5.316656589508057,
      "learning_rate": 6.877219759157087e-05,
      "loss": 2.308821678161621,
      "memory(GiB)": 77.56,
      "step": 44055,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.439757
    },
    {
      "epoch": 1.8876654813418448,
      "grad_norm": 5.144650936126709,
      "learning_rate": 6.876595996438072e-05,
      "loss": 2.5875526428222657,
      "memory(GiB)": 77.56,
      "step": 44060,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.439792
    },
    {
      "epoch": 1.8878796966710938,
      "grad_norm": 3.9537158012390137,
      "learning_rate": 6.875972199722761e-05,
      "loss": 2.427333450317383,
      "memory(GiB)": 77.56,
      "step": 44065,
      "token_acc": 0.5020242914979757,
      "train_speed(iter/s)": 1.439811
    },
    {
      "epoch": 1.8880939120003428,
      "grad_norm": 4.662796974182129,
      "learning_rate": 6.875348369022452e-05,
      "loss": 2.361859130859375,
      "memory(GiB)": 77.56,
      "step": 44070,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43982
    },
    {
      "epoch": 1.8883081273295916,
      "grad_norm": 5.868584632873535,
      "learning_rate": 6.87472450434845e-05,
      "loss": 2.578470802307129,
      "memory(GiB)": 77.56,
      "step": 44075,
      "token_acc": 0.49615384615384617,
      "train_speed(iter/s)": 1.439828
    },
    {
      "epoch": 1.8885223426588407,
      "grad_norm": 4.635788917541504,
      "learning_rate": 6.874100605712054e-05,
      "loss": 2.407032012939453,
      "memory(GiB)": 77.56,
      "step": 44080,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.439819
    },
    {
      "epoch": 1.8887365579880897,
      "grad_norm": 7.193100929260254,
      "learning_rate": 6.873476673124567e-05,
      "loss": 2.6361345291137694,
      "memory(GiB)": 77.56,
      "step": 44085,
      "token_acc": 0.4378698224852071,
      "train_speed(iter/s)": 1.439769
    },
    {
      "epoch": 1.8889507733173385,
      "grad_norm": 5.188387870788574,
      "learning_rate": 6.872852706597295e-05,
      "loss": 2.548508071899414,
      "memory(GiB)": 77.56,
      "step": 44090,
      "token_acc": 0.4430769230769231,
      "train_speed(iter/s)": 1.439805
    },
    {
      "epoch": 1.8891649886465876,
      "grad_norm": 5.298627853393555,
      "learning_rate": 6.872228706141537e-05,
      "loss": 2.3667545318603516,
      "memory(GiB)": 77.56,
      "step": 44095,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.439811
    },
    {
      "epoch": 1.8893792039758366,
      "grad_norm": 6.749616622924805,
      "learning_rate": 6.8716046717686e-05,
      "loss": 2.5558544158935548,
      "memory(GiB)": 77.56,
      "step": 44100,
      "token_acc": 0.5131578947368421,
      "train_speed(iter/s)": 1.439784
    },
    {
      "epoch": 1.8895934193050854,
      "grad_norm": 4.484576225280762,
      "learning_rate": 6.87098060348979e-05,
      "loss": 2.1556907653808595,
      "memory(GiB)": 77.56,
      "step": 44105,
      "token_acc": 0.5073529411764706,
      "train_speed(iter/s)": 1.439789
    },
    {
      "epoch": 1.8898076346343344,
      "grad_norm": 4.849059104919434,
      "learning_rate": 6.87035650131641e-05,
      "loss": 2.708654022216797,
      "memory(GiB)": 77.56,
      "step": 44110,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.439797
    },
    {
      "epoch": 1.8900218499635835,
      "grad_norm": 4.918495178222656,
      "learning_rate": 6.869732365259767e-05,
      "loss": 2.398085021972656,
      "memory(GiB)": 77.56,
      "step": 44115,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.439855
    },
    {
      "epoch": 1.8902360652928323,
      "grad_norm": 4.631740093231201,
      "learning_rate": 6.869108195331169e-05,
      "loss": 2.5333696365356446,
      "memory(GiB)": 77.56,
      "step": 44120,
      "token_acc": 0.4703703703703704,
      "train_speed(iter/s)": 1.439861
    },
    {
      "epoch": 1.8904502806220813,
      "grad_norm": 4.902470588684082,
      "learning_rate": 6.868483991541923e-05,
      "loss": 2.482940673828125,
      "memory(GiB)": 77.56,
      "step": 44125,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.439893
    },
    {
      "epoch": 1.8906644959513303,
      "grad_norm": 4.570365905761719,
      "learning_rate": 6.867859753903336e-05,
      "loss": 2.6711803436279298,
      "memory(GiB)": 77.56,
      "step": 44130,
      "token_acc": 0.445859872611465,
      "train_speed(iter/s)": 1.439893
    },
    {
      "epoch": 1.8908787112805792,
      "grad_norm": 4.237510681152344,
      "learning_rate": 6.867235482426719e-05,
      "loss": 2.4270206451416017,
      "memory(GiB)": 77.56,
      "step": 44135,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.4399
    },
    {
      "epoch": 1.8910929266098282,
      "grad_norm": 5.2100300788879395,
      "learning_rate": 6.866611177123377e-05,
      "loss": 2.335957336425781,
      "memory(GiB)": 77.56,
      "step": 44140,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.439898
    },
    {
      "epoch": 1.8913071419390772,
      "grad_norm": 5.797184467315674,
      "learning_rate": 6.865986838004624e-05,
      "loss": 2.575522613525391,
      "memory(GiB)": 77.56,
      "step": 44145,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.439912
    },
    {
      "epoch": 1.891521357268326,
      "grad_norm": 7.417253017425537,
      "learning_rate": 6.865362465081768e-05,
      "loss": 2.2161609649658205,
      "memory(GiB)": 77.56,
      "step": 44150,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.439931
    },
    {
      "epoch": 1.891735572597575,
      "grad_norm": 5.7866950035095215,
      "learning_rate": 6.864738058366122e-05,
      "loss": 2.4511606216430666,
      "memory(GiB)": 77.56,
      "step": 44155,
      "token_acc": 0.4823943661971831,
      "train_speed(iter/s)": 1.439964
    },
    {
      "epoch": 1.8919497879268241,
      "grad_norm": 3.83992600440979,
      "learning_rate": 6.864113617868997e-05,
      "loss": 2.60604190826416,
      "memory(GiB)": 77.56,
      "step": 44160,
      "token_acc": 0.49271137026239065,
      "train_speed(iter/s)": 1.44
    },
    {
      "epoch": 1.892164003256073,
      "grad_norm": 5.51364278793335,
      "learning_rate": 6.863489143601705e-05,
      "loss": 2.4766021728515626,
      "memory(GiB)": 77.56,
      "step": 44165,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.440022
    },
    {
      "epoch": 1.892378218585322,
      "grad_norm": 8.00137996673584,
      "learning_rate": 6.862864635575559e-05,
      "loss": 2.514451026916504,
      "memory(GiB)": 77.56,
      "step": 44170,
      "token_acc": 0.44921875,
      "train_speed(iter/s)": 1.440023
    },
    {
      "epoch": 1.892592433914571,
      "grad_norm": 7.786540985107422,
      "learning_rate": 6.862240093801873e-05,
      "loss": 2.3569799423217774,
      "memory(GiB)": 77.56,
      "step": 44175,
      "token_acc": 0.4645390070921986,
      "train_speed(iter/s)": 1.440023
    },
    {
      "epoch": 1.8928066492438198,
      "grad_norm": 4.392272472381592,
      "learning_rate": 6.861615518291962e-05,
      "loss": 2.4845392227172853,
      "memory(GiB)": 77.56,
      "step": 44180,
      "token_acc": 0.4680232558139535,
      "train_speed(iter/s)": 1.440018
    },
    {
      "epoch": 1.8930208645730688,
      "grad_norm": 4.896578788757324,
      "learning_rate": 6.860990909057137e-05,
      "loss": 2.821159553527832,
      "memory(GiB)": 77.56,
      "step": 44185,
      "token_acc": 0.4416058394160584,
      "train_speed(iter/s)": 1.440043
    },
    {
      "epoch": 1.8932350799023179,
      "grad_norm": 5.9553680419921875,
      "learning_rate": 6.860366266108716e-05,
      "loss": 2.5175575256347655,
      "memory(GiB)": 77.56,
      "step": 44190,
      "token_acc": 0.46946564885496184,
      "train_speed(iter/s)": 1.440054
    },
    {
      "epoch": 1.8934492952315667,
      "grad_norm": 6.024968147277832,
      "learning_rate": 6.859741589458015e-05,
      "loss": 2.520068550109863,
      "memory(GiB)": 77.56,
      "step": 44195,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.440073
    },
    {
      "epoch": 1.8936635105608157,
      "grad_norm": 5.330214977264404,
      "learning_rate": 6.859116879116352e-05,
      "loss": 2.6640974044799806,
      "memory(GiB)": 77.56,
      "step": 44200,
      "token_acc": 0.4901315789473684,
      "train_speed(iter/s)": 1.440076
    },
    {
      "epoch": 1.8938777258900648,
      "grad_norm": 7.271003723144531,
      "learning_rate": 6.858492135095043e-05,
      "loss": 2.7508787155151366,
      "memory(GiB)": 77.56,
      "step": 44205,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.440047
    },
    {
      "epoch": 1.8940919412193136,
      "grad_norm": 4.811175346374512,
      "learning_rate": 6.857867357405404e-05,
      "loss": 2.1135034561157227,
      "memory(GiB)": 77.56,
      "step": 44210,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.440073
    },
    {
      "epoch": 1.8943061565485626,
      "grad_norm": 4.746237754821777,
      "learning_rate": 6.857242546058756e-05,
      "loss": 2.4407325744628907,
      "memory(GiB)": 77.56,
      "step": 44215,
      "token_acc": 0.5047318611987381,
      "train_speed(iter/s)": 1.440103
    },
    {
      "epoch": 1.8945203718778116,
      "grad_norm": 5.387736797332764,
      "learning_rate": 6.856617701066416e-05,
      "loss": 2.5141532897949217,
      "memory(GiB)": 77.56,
      "step": 44220,
      "token_acc": 0.48026315789473684,
      "train_speed(iter/s)": 1.440124
    },
    {
      "epoch": 1.8947345872070604,
      "grad_norm": 4.8732523918151855,
      "learning_rate": 6.855992822439705e-05,
      "loss": 2.4990543365478515,
      "memory(GiB)": 77.56,
      "step": 44225,
      "token_acc": 0.45874587458745875,
      "train_speed(iter/s)": 1.440127
    },
    {
      "epoch": 1.8949488025363095,
      "grad_norm": 5.696702003479004,
      "learning_rate": 6.855367910189942e-05,
      "loss": 2.3242656707763674,
      "memory(GiB)": 77.56,
      "step": 44230,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.440165
    },
    {
      "epoch": 1.8951630178655585,
      "grad_norm": 5.067339897155762,
      "learning_rate": 6.854742964328451e-05,
      "loss": 2.559010696411133,
      "memory(GiB)": 77.56,
      "step": 44235,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.440151
    },
    {
      "epoch": 1.8953772331948073,
      "grad_norm": 4.362419605255127,
      "learning_rate": 6.854117984866552e-05,
      "loss": 2.544628715515137,
      "memory(GiB)": 77.56,
      "step": 44240,
      "token_acc": 0.4938650306748466,
      "train_speed(iter/s)": 1.44011
    },
    {
      "epoch": 1.8955914485240564,
      "grad_norm": 5.868187427520752,
      "learning_rate": 6.853492971815564e-05,
      "loss": 2.6103952407836912,
      "memory(GiB)": 77.56,
      "step": 44245,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.440115
    },
    {
      "epoch": 1.8958056638533054,
      "grad_norm": 4.793186187744141,
      "learning_rate": 6.852867925186814e-05,
      "loss": 2.580684280395508,
      "memory(GiB)": 77.56,
      "step": 44250,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.440097
    },
    {
      "epoch": 1.8960198791825542,
      "grad_norm": 4.180619716644287,
      "learning_rate": 6.85224284499162e-05,
      "loss": 2.662502479553223,
      "memory(GiB)": 77.56,
      "step": 44255,
      "token_acc": 0.44192634560906513,
      "train_speed(iter/s)": 1.440116
    },
    {
      "epoch": 1.8962340945118032,
      "grad_norm": 4.831403732299805,
      "learning_rate": 6.851617731241312e-05,
      "loss": 2.5747108459472656,
      "memory(GiB)": 77.56,
      "step": 44260,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.440132
    },
    {
      "epoch": 1.8964483098410523,
      "grad_norm": 4.796936988830566,
      "learning_rate": 6.850992583947212e-05,
      "loss": 2.830059814453125,
      "memory(GiB)": 77.56,
      "step": 44265,
      "token_acc": 0.44368600682593856,
      "train_speed(iter/s)": 1.440171
    },
    {
      "epoch": 1.896662525170301,
      "grad_norm": 7.206109523773193,
      "learning_rate": 6.850367403120645e-05,
      "loss": 2.4320091247558593,
      "memory(GiB)": 77.56,
      "step": 44270,
      "token_acc": 0.47157190635451507,
      "train_speed(iter/s)": 1.440134
    },
    {
      "epoch": 1.8968767404995501,
      "grad_norm": 5.738765239715576,
      "learning_rate": 6.849742188772936e-05,
      "loss": 2.4546533584594727,
      "memory(GiB)": 77.56,
      "step": 44275,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.440147
    },
    {
      "epoch": 1.8970909558287992,
      "grad_norm": 4.917630672454834,
      "learning_rate": 6.849116940915412e-05,
      "loss": 2.0927761077880858,
      "memory(GiB)": 77.56,
      "step": 44280,
      "token_acc": 0.5222222222222223,
      "train_speed(iter/s)": 1.440198
    },
    {
      "epoch": 1.897305171158048,
      "grad_norm": 5.305604457855225,
      "learning_rate": 6.8484916595594e-05,
      "loss": 2.3930181503295898,
      "memory(GiB)": 77.56,
      "step": 44285,
      "token_acc": 0.5206896551724138,
      "train_speed(iter/s)": 1.440187
    },
    {
      "epoch": 1.897519386487297,
      "grad_norm": 4.421180725097656,
      "learning_rate": 6.847866344716228e-05,
      "loss": 2.3185192108154298,
      "memory(GiB)": 77.56,
      "step": 44290,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.440168
    },
    {
      "epoch": 1.897733601816546,
      "grad_norm": 5.534777641296387,
      "learning_rate": 6.847240996397224e-05,
      "loss": 2.5180152893066405,
      "memory(GiB)": 77.56,
      "step": 44295,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.440204
    },
    {
      "epoch": 1.8979478171457949,
      "grad_norm": 4.705270767211914,
      "learning_rate": 6.846615614613716e-05,
      "loss": 2.5731258392333984,
      "memory(GiB)": 77.56,
      "step": 44300,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.440219
    },
    {
      "epoch": 1.898162032475044,
      "grad_norm": 5.615041255950928,
      "learning_rate": 6.845990199377033e-05,
      "loss": 2.8154468536376953,
      "memory(GiB)": 77.56,
      "step": 44305,
      "token_acc": 0.4648318042813456,
      "train_speed(iter/s)": 1.440226
    },
    {
      "epoch": 1.898376247804293,
      "grad_norm": 4.8558807373046875,
      "learning_rate": 6.845364750698507e-05,
      "loss": 2.760580062866211,
      "memory(GiB)": 77.56,
      "step": 44310,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.440278
    },
    {
      "epoch": 1.8985904631335417,
      "grad_norm": 5.47043514251709,
      "learning_rate": 6.844739268589467e-05,
      "loss": 2.5955120086669923,
      "memory(GiB)": 77.56,
      "step": 44315,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.440226
    },
    {
      "epoch": 1.8988046784627908,
      "grad_norm": 5.321701526641846,
      "learning_rate": 6.844113753061246e-05,
      "loss": 2.2472354888916017,
      "memory(GiB)": 77.56,
      "step": 44320,
      "token_acc": 0.5201793721973094,
      "train_speed(iter/s)": 1.440235
    },
    {
      "epoch": 1.8990188937920398,
      "grad_norm": 4.491479396820068,
      "learning_rate": 6.843488204125174e-05,
      "loss": 2.733559989929199,
      "memory(GiB)": 77.56,
      "step": 44325,
      "token_acc": 0.42990654205607476,
      "train_speed(iter/s)": 1.440212
    },
    {
      "epoch": 1.8992331091212886,
      "grad_norm": 6.029450416564941,
      "learning_rate": 6.842862621792581e-05,
      "loss": 2.443553161621094,
      "memory(GiB)": 77.56,
      "step": 44330,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.440222
    },
    {
      "epoch": 1.8994473244505377,
      "grad_norm": 5.856253623962402,
      "learning_rate": 6.842237006074805e-05,
      "loss": 2.851237106323242,
      "memory(GiB)": 77.56,
      "step": 44335,
      "token_acc": 0.43465045592705165,
      "train_speed(iter/s)": 1.440278
    },
    {
      "epoch": 1.8996615397797867,
      "grad_norm": 4.865426063537598,
      "learning_rate": 6.841611356983179e-05,
      "loss": 2.3552343368530275,
      "memory(GiB)": 77.56,
      "step": 44340,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.440245
    },
    {
      "epoch": 1.8998757551090355,
      "grad_norm": 4.678223133087158,
      "learning_rate": 6.840985674529033e-05,
      "loss": 2.636866569519043,
      "memory(GiB)": 77.56,
      "step": 44345,
      "token_acc": 0.5054151624548736,
      "train_speed(iter/s)": 1.440282
    },
    {
      "epoch": 1.9000899704382845,
      "grad_norm": 5.230689525604248,
      "learning_rate": 6.840359958723705e-05,
      "loss": 2.7309587478637694,
      "memory(GiB)": 77.56,
      "step": 44350,
      "token_acc": 0.41324921135646686,
      "train_speed(iter/s)": 1.440322
    },
    {
      "epoch": 1.9003041857675336,
      "grad_norm": 6.695852756500244,
      "learning_rate": 6.839734209578532e-05,
      "loss": 2.299959182739258,
      "memory(GiB)": 77.56,
      "step": 44355,
      "token_acc": 0.45110410094637227,
      "train_speed(iter/s)": 1.440365
    },
    {
      "epoch": 1.9005184010967824,
      "grad_norm": 4.074918270111084,
      "learning_rate": 6.839108427104846e-05,
      "loss": 2.2946331024169924,
      "memory(GiB)": 77.56,
      "step": 44360,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.440405
    },
    {
      "epoch": 1.9007326164260314,
      "grad_norm": 6.127739906311035,
      "learning_rate": 6.838482611313985e-05,
      "loss": 2.290567970275879,
      "memory(GiB)": 77.56,
      "step": 44365,
      "token_acc": 0.544,
      "train_speed(iter/s)": 1.440436
    },
    {
      "epoch": 1.9009468317552805,
      "grad_norm": 5.002657413482666,
      "learning_rate": 6.83785676221729e-05,
      "loss": 2.466518211364746,
      "memory(GiB)": 77.56,
      "step": 44370,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.440453
    },
    {
      "epoch": 1.9011610470845293,
      "grad_norm": 5.262711048126221,
      "learning_rate": 6.837230879826093e-05,
      "loss": 2.222349166870117,
      "memory(GiB)": 77.56,
      "step": 44375,
      "token_acc": 0.43389830508474575,
      "train_speed(iter/s)": 1.440419
    },
    {
      "epoch": 1.9013752624137783,
      "grad_norm": 6.165261745452881,
      "learning_rate": 6.836604964151737e-05,
      "loss": 2.5008384704589846,
      "memory(GiB)": 77.56,
      "step": 44380,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.440446
    },
    {
      "epoch": 1.9015894777430273,
      "grad_norm": 5.602413654327393,
      "learning_rate": 6.835979015205558e-05,
      "loss": 2.4265506744384764,
      "memory(GiB)": 77.56,
      "step": 44385,
      "token_acc": 0.5375,
      "train_speed(iter/s)": 1.440423
    },
    {
      "epoch": 1.9018036930722761,
      "grad_norm": 5.043893814086914,
      "learning_rate": 6.835353032998896e-05,
      "loss": 2.2805496215820313,
      "memory(GiB)": 77.56,
      "step": 44390,
      "token_acc": 0.528957528957529,
      "train_speed(iter/s)": 1.44041
    },
    {
      "epoch": 1.9020179084015252,
      "grad_norm": 4.740059852600098,
      "learning_rate": 6.834727017543094e-05,
      "loss": 2.356195068359375,
      "memory(GiB)": 77.56,
      "step": 44395,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440415
    },
    {
      "epoch": 1.9022321237307742,
      "grad_norm": 5.851634979248047,
      "learning_rate": 6.83410096884949e-05,
      "loss": 2.5868406295776367,
      "memory(GiB)": 77.56,
      "step": 44400,
      "token_acc": 0.4901185770750988,
      "train_speed(iter/s)": 1.440429
    },
    {
      "epoch": 1.902446339060023,
      "grad_norm": 5.274365425109863,
      "learning_rate": 6.833474886929425e-05,
      "loss": 2.489689254760742,
      "memory(GiB)": 77.56,
      "step": 44405,
      "token_acc": 0.4759036144578313,
      "train_speed(iter/s)": 1.440437
    },
    {
      "epoch": 1.902660554389272,
      "grad_norm": 4.019134044647217,
      "learning_rate": 6.832848771794245e-05,
      "loss": 2.27807502746582,
      "memory(GiB)": 77.56,
      "step": 44410,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.440452
    },
    {
      "epoch": 1.902874769718521,
      "grad_norm": 4.189376354217529,
      "learning_rate": 6.83222262345529e-05,
      "loss": 2.325949859619141,
      "memory(GiB)": 77.56,
      "step": 44415,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.440407
    },
    {
      "epoch": 1.90308898504777,
      "grad_norm": 5.913931369781494,
      "learning_rate": 6.831596441923902e-05,
      "loss": 2.5688007354736326,
      "memory(GiB)": 77.56,
      "step": 44420,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.440388
    },
    {
      "epoch": 1.903303200377019,
      "grad_norm": 4.257114410400391,
      "learning_rate": 6.830970227211427e-05,
      "loss": 2.7230743408203124,
      "memory(GiB)": 77.56,
      "step": 44425,
      "token_acc": 0.4735202492211838,
      "train_speed(iter/s)": 1.440421
    },
    {
      "epoch": 1.903517415706268,
      "grad_norm": 5.527200698852539,
      "learning_rate": 6.830343979329208e-05,
      "loss": 2.335757827758789,
      "memory(GiB)": 77.56,
      "step": 44430,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.440458
    },
    {
      "epoch": 1.9037316310355168,
      "grad_norm": 4.908282279968262,
      "learning_rate": 6.829717698288591e-05,
      "loss": 2.3787445068359374,
      "memory(GiB)": 77.56,
      "step": 44435,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.440469
    },
    {
      "epoch": 1.9039458463647658,
      "grad_norm": 4.689885139465332,
      "learning_rate": 6.829091384100922e-05,
      "loss": 2.2938655853271483,
      "memory(GiB)": 77.56,
      "step": 44440,
      "token_acc": 0.43866171003717475,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.9041600616940149,
      "grad_norm": 4.703327655792236,
      "learning_rate": 6.828465036777548e-05,
      "loss": 2.5733800888061524,
      "memory(GiB)": 77.56,
      "step": 44445,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.440459
    },
    {
      "epoch": 1.9043742770232637,
      "grad_norm": 6.948575973510742,
      "learning_rate": 6.827838656329812e-05,
      "loss": 2.6196332931518556,
      "memory(GiB)": 77.56,
      "step": 44450,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.440433
    },
    {
      "epoch": 1.9045884923525127,
      "grad_norm": 6.885674476623535,
      "learning_rate": 6.827212242769065e-05,
      "loss": 2.4800430297851563,
      "memory(GiB)": 77.56,
      "step": 44455,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.440456
    },
    {
      "epoch": 1.9048027076817617,
      "grad_norm": 5.4360737800598145,
      "learning_rate": 6.826585796106654e-05,
      "loss": 2.589427947998047,
      "memory(GiB)": 77.56,
      "step": 44460,
      "token_acc": 0.4659090909090909,
      "train_speed(iter/s)": 1.440458
    },
    {
      "epoch": 1.9050169230110106,
      "grad_norm": 4.815432071685791,
      "learning_rate": 6.82595931635393e-05,
      "loss": 2.551029586791992,
      "memory(GiB)": 77.56,
      "step": 44465,
      "token_acc": 0.5163398692810458,
      "train_speed(iter/s)": 1.440482
    },
    {
      "epoch": 1.9052311383402596,
      "grad_norm": 7.217994689941406,
      "learning_rate": 6.825332803522238e-05,
      "loss": 2.3544147491455076,
      "memory(GiB)": 77.56,
      "step": 44470,
      "token_acc": 0.5271966527196653,
      "train_speed(iter/s)": 1.440491
    },
    {
      "epoch": 1.9054453536695086,
      "grad_norm": 4.317183494567871,
      "learning_rate": 6.82470625762293e-05,
      "loss": 2.2244842529296873,
      "memory(GiB)": 77.56,
      "step": 44475,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.440452
    },
    {
      "epoch": 1.9056595689987574,
      "grad_norm": 5.077231407165527,
      "learning_rate": 6.824079678667357e-05,
      "loss": 2.70748291015625,
      "memory(GiB)": 77.56,
      "step": 44480,
      "token_acc": 0.45112781954887216,
      "train_speed(iter/s)": 1.440423
    },
    {
      "epoch": 1.9058737843280065,
      "grad_norm": 5.023132801055908,
      "learning_rate": 6.82345306666687e-05,
      "loss": 2.3879493713378905,
      "memory(GiB)": 77.56,
      "step": 44485,
      "token_acc": 0.47564469914040114,
      "train_speed(iter/s)": 1.440471
    },
    {
      "epoch": 1.9060879996572555,
      "grad_norm": 5.956859588623047,
      "learning_rate": 6.822826421632819e-05,
      "loss": 2.4765167236328125,
      "memory(GiB)": 77.56,
      "step": 44490,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.440481
    },
    {
      "epoch": 1.9063022149865043,
      "grad_norm": 4.102790355682373,
      "learning_rate": 6.822199743576558e-05,
      "loss": 2.5086177825927733,
      "memory(GiB)": 77.56,
      "step": 44495,
      "token_acc": 0.49079754601226994,
      "train_speed(iter/s)": 1.440487
    },
    {
      "epoch": 1.9065164303157534,
      "grad_norm": 4.397863388061523,
      "learning_rate": 6.82157303250944e-05,
      "loss": 2.4902156829833983,
      "memory(GiB)": 77.56,
      "step": 44500,
      "token_acc": 0.4708029197080292,
      "train_speed(iter/s)": 1.440514
    },
    {
      "epoch": 1.9065164303157534,
      "eval_loss": 2.4332528114318848,
      "eval_runtime": 14.922,
      "eval_samples_per_second": 6.702,
      "eval_steps_per_second": 6.702,
      "eval_token_acc": 0.46124523506988563,
      "step": 44500
    },
    {
      "epoch": 1.9067306456450024,
      "grad_norm": 4.878571033477783,
      "learning_rate": 6.820946288442816e-05,
      "loss": 2.7795650482177736,
      "memory(GiB)": 77.56,
      "step": 44505,
      "token_acc": 0.45056867891513563,
      "train_speed(iter/s)": 1.439757
    },
    {
      "epoch": 1.9069448609742512,
      "grad_norm": 5.015744686126709,
      "learning_rate": 6.820319511388043e-05,
      "loss": 2.7151039123535154,
      "memory(GiB)": 77.56,
      "step": 44510,
      "token_acc": 0.449685534591195,
      "train_speed(iter/s)": 1.439787
    },
    {
      "epoch": 1.9071590763035002,
      "grad_norm": 4.074020862579346,
      "learning_rate": 6.819692701356474e-05,
      "loss": 2.9155553817749023,
      "memory(GiB)": 77.56,
      "step": 44515,
      "token_acc": 0.40119760479041916,
      "train_speed(iter/s)": 1.439799
    },
    {
      "epoch": 1.9073732916327493,
      "grad_norm": 3.7463245391845703,
      "learning_rate": 6.819065858359464e-05,
      "loss": 2.5465557098388674,
      "memory(GiB)": 77.56,
      "step": 44520,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.439793
    },
    {
      "epoch": 1.907587506961998,
      "grad_norm": 7.250163555145264,
      "learning_rate": 6.818438982408371e-05,
      "loss": 2.690298080444336,
      "memory(GiB)": 77.56,
      "step": 44525,
      "token_acc": 0.4542124542124542,
      "train_speed(iter/s)": 1.439795
    },
    {
      "epoch": 1.9078017222912471,
      "grad_norm": 5.25191593170166,
      "learning_rate": 6.81781207351455e-05,
      "loss": 2.4210437774658202,
      "memory(GiB)": 77.56,
      "step": 44530,
      "token_acc": 0.49640287769784175,
      "train_speed(iter/s)": 1.439841
    },
    {
      "epoch": 1.9080159376204961,
      "grad_norm": 4.25998592376709,
      "learning_rate": 6.817185131689356e-05,
      "loss": 2.5990951538085936,
      "memory(GiB)": 77.56,
      "step": 44535,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.439863
    },
    {
      "epoch": 1.908230152949745,
      "grad_norm": 4.700165748596191,
      "learning_rate": 6.816558156944151e-05,
      "loss": 2.2443944931030275,
      "memory(GiB)": 77.56,
      "step": 44540,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.43989
    },
    {
      "epoch": 1.908444368278994,
      "grad_norm": 5.219594955444336,
      "learning_rate": 6.81593114929029e-05,
      "loss": 2.5784843444824217,
      "memory(GiB)": 77.56,
      "step": 44545,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.439917
    },
    {
      "epoch": 1.908658583608243,
      "grad_norm": 4.59681510925293,
      "learning_rate": 6.815304108739133e-05,
      "loss": 2.6447582244873047,
      "memory(GiB)": 77.56,
      "step": 44550,
      "token_acc": 0.43636363636363634,
      "train_speed(iter/s)": 1.439908
    },
    {
      "epoch": 1.9088727989374918,
      "grad_norm": 4.393628120422363,
      "learning_rate": 6.814677035302038e-05,
      "loss": 2.506500244140625,
      "memory(GiB)": 77.56,
      "step": 44555,
      "token_acc": 0.4459016393442623,
      "train_speed(iter/s)": 1.439903
    },
    {
      "epoch": 1.9090870142667409,
      "grad_norm": 5.056570529937744,
      "learning_rate": 6.814049928990369e-05,
      "loss": 2.478056526184082,
      "memory(GiB)": 77.56,
      "step": 44560,
      "token_acc": 0.4511784511784512,
      "train_speed(iter/s)": 1.439933
    },
    {
      "epoch": 1.90930122959599,
      "grad_norm": 5.622750282287598,
      "learning_rate": 6.81342278981548e-05,
      "loss": 2.6171329498291014,
      "memory(GiB)": 77.56,
      "step": 44565,
      "token_acc": 0.41810344827586204,
      "train_speed(iter/s)": 1.439965
    },
    {
      "epoch": 1.9095154449252387,
      "grad_norm": 5.701539993286133,
      "learning_rate": 6.812795617788739e-05,
      "loss": 2.4851621627807616,
      "memory(GiB)": 77.56,
      "step": 44570,
      "token_acc": 0.46048109965635736,
      "train_speed(iter/s)": 1.439923
    },
    {
      "epoch": 1.9097296602544878,
      "grad_norm": 5.992129802703857,
      "learning_rate": 6.812168412921504e-05,
      "loss": 2.4025535583496094,
      "memory(GiB)": 77.56,
      "step": 44575,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.439913
    },
    {
      "epoch": 1.9099438755837368,
      "grad_norm": 3.5778889656066895,
      "learning_rate": 6.811541175225138e-05,
      "loss": 2.5711830139160154,
      "memory(GiB)": 77.56,
      "step": 44580,
      "token_acc": 0.4564459930313589,
      "train_speed(iter/s)": 1.439928
    },
    {
      "epoch": 1.9101580909129856,
      "grad_norm": 4.192580223083496,
      "learning_rate": 6.810913904711004e-05,
      "loss": 2.684924507141113,
      "memory(GiB)": 77.56,
      "step": 44585,
      "token_acc": 0.4169381107491857,
      "train_speed(iter/s)": 1.439939
    },
    {
      "epoch": 1.9103723062422346,
      "grad_norm": 4.966233730316162,
      "learning_rate": 6.810286601390466e-05,
      "loss": 2.288302421569824,
      "memory(GiB)": 77.56,
      "step": 44590,
      "token_acc": 0.4880952380952381,
      "train_speed(iter/s)": 1.439929
    },
    {
      "epoch": 1.9105865215714837,
      "grad_norm": 5.189723968505859,
      "learning_rate": 6.809659265274888e-05,
      "loss": 2.3654727935791016,
      "memory(GiB)": 77.56,
      "step": 44595,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.439974
    },
    {
      "epoch": 1.9108007369007325,
      "grad_norm": 4.194835186004639,
      "learning_rate": 6.809031896375636e-05,
      "loss": 2.6107873916625977,
      "memory(GiB)": 77.56,
      "step": 44600,
      "token_acc": 0.4314868804664723,
      "train_speed(iter/s)": 1.439951
    },
    {
      "epoch": 1.9110149522299815,
      "grad_norm": 5.426981449127197,
      "learning_rate": 6.808404494704073e-05,
      "loss": 2.34908447265625,
      "memory(GiB)": 77.56,
      "step": 44605,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.439938
    },
    {
      "epoch": 1.9112291675592306,
      "grad_norm": 5.184592247009277,
      "learning_rate": 6.807777060271566e-05,
      "loss": 2.452161407470703,
      "memory(GiB)": 77.56,
      "step": 44610,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.439973
    },
    {
      "epoch": 1.9114433828884794,
      "grad_norm": 4.717799186706543,
      "learning_rate": 6.807149593089484e-05,
      "loss": 2.276875686645508,
      "memory(GiB)": 77.56,
      "step": 44615,
      "token_acc": 0.5087719298245614,
      "train_speed(iter/s)": 1.439971
    },
    {
      "epoch": 1.9116575982177286,
      "grad_norm": 4.707108974456787,
      "learning_rate": 6.806522093169189e-05,
      "loss": 2.6007253646850588,
      "memory(GiB)": 77.56,
      "step": 44620,
      "token_acc": 0.5021645021645021,
      "train_speed(iter/s)": 1.440014
    },
    {
      "epoch": 1.9118718135469774,
      "grad_norm": 4.245858669281006,
      "learning_rate": 6.805894560522051e-05,
      "loss": 2.5515480041503906,
      "memory(GiB)": 77.56,
      "step": 44625,
      "token_acc": 0.4700854700854701,
      "train_speed(iter/s)": 1.440041
    },
    {
      "epoch": 1.9120860288762263,
      "grad_norm": 5.026432514190674,
      "learning_rate": 6.805266995159442e-05,
      "loss": 2.5758060455322265,
      "memory(GiB)": 77.56,
      "step": 44630,
      "token_acc": 0.49244712990936557,
      "train_speed(iter/s)": 1.440058
    },
    {
      "epoch": 1.9123002442054755,
      "grad_norm": 4.200397491455078,
      "learning_rate": 6.804639397092726e-05,
      "loss": 2.635577583312988,
      "memory(GiB)": 77.56,
      "step": 44635,
      "token_acc": 0.48909657320872274,
      "train_speed(iter/s)": 1.440074
    },
    {
      "epoch": 1.9125144595347243,
      "grad_norm": 3.5709409713745117,
      "learning_rate": 6.804011766333276e-05,
      "loss": 2.190346908569336,
      "memory(GiB)": 77.56,
      "step": 44640,
      "token_acc": 0.5206896551724138,
      "train_speed(iter/s)": 1.440056
    },
    {
      "epoch": 1.9127286748639731,
      "grad_norm": 5.625542640686035,
      "learning_rate": 6.80338410289246e-05,
      "loss": 2.4494842529296874,
      "memory(GiB)": 77.56,
      "step": 44645,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.440024
    },
    {
      "epoch": 1.9129428901932224,
      "grad_norm": 6.4972381591796875,
      "learning_rate": 6.802756406781648e-05,
      "loss": 2.2170576095581054,
      "memory(GiB)": 77.56,
      "step": 44650,
      "token_acc": 0.5328467153284672,
      "train_speed(iter/s)": 1.440061
    },
    {
      "epoch": 1.9131571055224712,
      "grad_norm": 5.853265762329102,
      "learning_rate": 6.802128678012214e-05,
      "loss": 2.648603630065918,
      "memory(GiB)": 77.56,
      "step": 44655,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.440045
    },
    {
      "epoch": 1.91337132085172,
      "grad_norm": 4.965744972229004,
      "learning_rate": 6.801500916595527e-05,
      "loss": 2.413798522949219,
      "memory(GiB)": 77.56,
      "step": 44660,
      "token_acc": 0.4984423676012461,
      "train_speed(iter/s)": 1.440072
    },
    {
      "epoch": 1.9135855361809693,
      "grad_norm": 6.027616024017334,
      "learning_rate": 6.800873122542962e-05,
      "loss": 2.6154855728149413,
      "memory(GiB)": 77.56,
      "step": 44665,
      "token_acc": 0.46325878594249204,
      "train_speed(iter/s)": 1.440053
    },
    {
      "epoch": 1.913799751510218,
      "grad_norm": 5.317164897918701,
      "learning_rate": 6.80024529586589e-05,
      "loss": 2.3016637802124023,
      "memory(GiB)": 77.56,
      "step": 44670,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.440071
    },
    {
      "epoch": 1.914013966839467,
      "grad_norm": 4.941979885101318,
      "learning_rate": 6.799617436575688e-05,
      "loss": 2.164947509765625,
      "memory(GiB)": 77.56,
      "step": 44675,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.440117
    },
    {
      "epoch": 1.9142281821687162,
      "grad_norm": 5.081755638122559,
      "learning_rate": 6.798989544683725e-05,
      "loss": 2.5052539825439455,
      "memory(GiB)": 77.56,
      "step": 44680,
      "token_acc": 0.5052631578947369,
      "train_speed(iter/s)": 1.440124
    },
    {
      "epoch": 1.914442397497965,
      "grad_norm": 5.270116806030273,
      "learning_rate": 6.798361620201382e-05,
      "loss": 2.524852752685547,
      "memory(GiB)": 77.56,
      "step": 44685,
      "token_acc": 0.5174603174603175,
      "train_speed(iter/s)": 1.440104
    },
    {
      "epoch": 1.9146566128272138,
      "grad_norm": 4.910960674285889,
      "learning_rate": 6.797733663140028e-05,
      "loss": 2.371072769165039,
      "memory(GiB)": 77.56,
      "step": 44690,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.440074
    },
    {
      "epoch": 1.914870828156463,
      "grad_norm": 4.892165184020996,
      "learning_rate": 6.797105673511044e-05,
      "loss": 2.616036796569824,
      "memory(GiB)": 77.56,
      "step": 44695,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.440134
    },
    {
      "epoch": 1.9150850434857118,
      "grad_norm": 5.687769412994385,
      "learning_rate": 6.796477651325805e-05,
      "loss": 2.6887134552001952,
      "memory(GiB)": 77.56,
      "step": 44700,
      "token_acc": 0.4954128440366973,
      "train_speed(iter/s)": 1.44012
    },
    {
      "epoch": 1.9152992588149607,
      "grad_norm": 5.398184776306152,
      "learning_rate": 6.795849596595686e-05,
      "loss": 2.6722476959228514,
      "memory(GiB)": 77.56,
      "step": 44705,
      "token_acc": 0.4426751592356688,
      "train_speed(iter/s)": 1.440126
    },
    {
      "epoch": 1.91551347414421,
      "grad_norm": 4.448758125305176,
      "learning_rate": 6.79522150933207e-05,
      "loss": 2.4870203018188475,
      "memory(GiB)": 77.56,
      "step": 44710,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.440139
    },
    {
      "epoch": 1.9157276894734587,
      "grad_norm": 4.954132556915283,
      "learning_rate": 6.794593389546331e-05,
      "loss": 2.6410064697265625,
      "memory(GiB)": 77.56,
      "step": 44715,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.440164
    },
    {
      "epoch": 1.9159419048027075,
      "grad_norm": 7.54559326171875,
      "learning_rate": 6.793965237249848e-05,
      "loss": 2.5560977935791014,
      "memory(GiB)": 77.56,
      "step": 44720,
      "token_acc": 0.46886446886446886,
      "train_speed(iter/s)": 1.440165
    },
    {
      "epoch": 1.9161561201319568,
      "grad_norm": 4.980578899383545,
      "learning_rate": 6.793337052454005e-05,
      "loss": 2.514068603515625,
      "memory(GiB)": 77.56,
      "step": 44725,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.440173
    },
    {
      "epoch": 1.9163703354612056,
      "grad_norm": 5.284402370452881,
      "learning_rate": 6.792708835170177e-05,
      "loss": 2.4049118041992186,
      "memory(GiB)": 77.56,
      "step": 44730,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.440161
    },
    {
      "epoch": 1.9165845507904544,
      "grad_norm": 4.475897789001465,
      "learning_rate": 6.792080585409746e-05,
      "loss": 2.47657470703125,
      "memory(GiB)": 77.56,
      "step": 44735,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440192
    },
    {
      "epoch": 1.9167987661197037,
      "grad_norm": 3.9399573802948,
      "learning_rate": 6.791452303184094e-05,
      "loss": 2.5199440002441404,
      "memory(GiB)": 77.56,
      "step": 44740,
      "token_acc": 0.486404833836858,
      "train_speed(iter/s)": 1.440182
    },
    {
      "epoch": 1.9170129814489525,
      "grad_norm": 4.361889362335205,
      "learning_rate": 6.790823988504606e-05,
      "loss": 2.7094736099243164,
      "memory(GiB)": 77.56,
      "step": 44745,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.440194
    },
    {
      "epoch": 1.9172271967782013,
      "grad_norm": 5.997969627380371,
      "learning_rate": 6.790195641382658e-05,
      "loss": 2.7188392639160157,
      "memory(GiB)": 77.56,
      "step": 44750,
      "token_acc": 0.43567251461988304,
      "train_speed(iter/s)": 1.440185
    },
    {
      "epoch": 1.9174414121074506,
      "grad_norm": 5.397094249725342,
      "learning_rate": 6.789567261829639e-05,
      "loss": 2.5017984390258787,
      "memory(GiB)": 77.56,
      "step": 44755,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.44011
    },
    {
      "epoch": 1.9176556274366994,
      "grad_norm": 6.921408176422119,
      "learning_rate": 6.788938849856929e-05,
      "loss": 2.719912528991699,
      "memory(GiB)": 77.56,
      "step": 44760,
      "token_acc": 0.46204620462046203,
      "train_speed(iter/s)": 1.440135
    },
    {
      "epoch": 1.9178698427659482,
      "grad_norm": 4.085766315460205,
      "learning_rate": 6.788310405475915e-05,
      "loss": 2.15233154296875,
      "memory(GiB)": 77.56,
      "step": 44765,
      "token_acc": 0.5215686274509804,
      "train_speed(iter/s)": 1.440155
    },
    {
      "epoch": 1.9180840580951974,
      "grad_norm": 5.353226184844971,
      "learning_rate": 6.78768192869798e-05,
      "loss": 2.401826858520508,
      "memory(GiB)": 77.56,
      "step": 44770,
      "token_acc": 0.5287769784172662,
      "train_speed(iter/s)": 1.440182
    },
    {
      "epoch": 1.9182982734244463,
      "grad_norm": 4.965895175933838,
      "learning_rate": 6.787053419534508e-05,
      "loss": 2.5608749389648438,
      "memory(GiB)": 77.56,
      "step": 44775,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440166
    },
    {
      "epoch": 1.918512488753695,
      "grad_norm": 4.790609359741211,
      "learning_rate": 6.786424877996887e-05,
      "loss": 2.4662988662719725,
      "memory(GiB)": 77.56,
      "step": 44780,
      "token_acc": 0.47289156626506024,
      "train_speed(iter/s)": 1.440193
    },
    {
      "epoch": 1.9187267040829443,
      "grad_norm": 5.190618515014648,
      "learning_rate": 6.785796304096506e-05,
      "loss": 2.4805007934570313,
      "memory(GiB)": 77.56,
      "step": 44785,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.440178
    },
    {
      "epoch": 1.9189409194121931,
      "grad_norm": 4.605311393737793,
      "learning_rate": 6.78516769784475e-05,
      "loss": 2.8391439437866213,
      "memory(GiB)": 77.56,
      "step": 44790,
      "token_acc": 0.4578313253012048,
      "train_speed(iter/s)": 1.440224
    },
    {
      "epoch": 1.919155134741442,
      "grad_norm": 3.971388578414917,
      "learning_rate": 6.784539059253004e-05,
      "loss": 2.6653919219970703,
      "memory(GiB)": 77.56,
      "step": 44795,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.440247
    },
    {
      "epoch": 1.9193693500706912,
      "grad_norm": 3.805058479309082,
      "learning_rate": 6.783910388332661e-05,
      "loss": 2.362849807739258,
      "memory(GiB)": 77.56,
      "step": 44800,
      "token_acc": 0.45723684210526316,
      "train_speed(iter/s)": 1.440273
    },
    {
      "epoch": 1.91958356539994,
      "grad_norm": 4.532949924468994,
      "learning_rate": 6.783281685095105e-05,
      "loss": 2.414392280578613,
      "memory(GiB)": 77.56,
      "step": 44805,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.440274
    },
    {
      "epoch": 1.9197977807291888,
      "grad_norm": 4.745820999145508,
      "learning_rate": 6.78265294955173e-05,
      "loss": 2.460069274902344,
      "memory(GiB)": 77.56,
      "step": 44810,
      "token_acc": 0.5134099616858238,
      "train_speed(iter/s)": 1.44026
    },
    {
      "epoch": 1.920011996058438,
      "grad_norm": 4.696557998657227,
      "learning_rate": 6.782024181713925e-05,
      "loss": 2.4688270568847654,
      "memory(GiB)": 77.56,
      "step": 44815,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.440289
    },
    {
      "epoch": 1.920226211387687,
      "grad_norm": 4.4076247215271,
      "learning_rate": 6.781395381593082e-05,
      "loss": 2.496187210083008,
      "memory(GiB)": 77.56,
      "step": 44820,
      "token_acc": 0.4367469879518072,
      "train_speed(iter/s)": 1.440298
    },
    {
      "epoch": 1.9204404267169357,
      "grad_norm": 5.914099216461182,
      "learning_rate": 6.780766549200587e-05,
      "loss": 2.5455934524536135,
      "memory(GiB)": 77.56,
      "step": 44825,
      "token_acc": 0.4948805460750853,
      "train_speed(iter/s)": 1.440291
    },
    {
      "epoch": 1.920654642046185,
      "grad_norm": 4.998441219329834,
      "learning_rate": 6.780137684547838e-05,
      "loss": 2.6923410415649416,
      "memory(GiB)": 77.56,
      "step": 44830,
      "token_acc": 0.4166666666666667,
      "train_speed(iter/s)": 1.440326
    },
    {
      "epoch": 1.9208688573754338,
      "grad_norm": 4.390291690826416,
      "learning_rate": 6.779508787646222e-05,
      "loss": 2.4865606307983397,
      "memory(GiB)": 77.56,
      "step": 44835,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.440353
    },
    {
      "epoch": 1.9210830727046826,
      "grad_norm": 7.698446273803711,
      "learning_rate": 6.778879858507138e-05,
      "loss": 2.5965694427490233,
      "memory(GiB)": 77.56,
      "step": 44840,
      "token_acc": 0.45018450184501846,
      "train_speed(iter/s)": 1.440373
    },
    {
      "epoch": 1.9212972880339318,
      "grad_norm": 5.172274112701416,
      "learning_rate": 6.778250897141976e-05,
      "loss": 2.5993576049804688,
      "memory(GiB)": 77.56,
      "step": 44845,
      "token_acc": 0.4429065743944637,
      "train_speed(iter/s)": 1.440407
    },
    {
      "epoch": 1.9215115033631807,
      "grad_norm": 4.587586402893066,
      "learning_rate": 6.777621903562132e-05,
      "loss": 2.297383499145508,
      "memory(GiB)": 77.56,
      "step": 44850,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.440424
    },
    {
      "epoch": 1.9217257186924295,
      "grad_norm": 3.7478761672973633,
      "learning_rate": 6.776992877778997e-05,
      "loss": 2.829428482055664,
      "memory(GiB)": 77.56,
      "step": 44855,
      "token_acc": 0.42810457516339867,
      "train_speed(iter/s)": 1.440488
    },
    {
      "epoch": 1.9219399340216787,
      "grad_norm": 5.2397685050964355,
      "learning_rate": 6.77636381980397e-05,
      "loss": 2.568630409240723,
      "memory(GiB)": 77.56,
      "step": 44860,
      "token_acc": 0.4746376811594203,
      "train_speed(iter/s)": 1.440509
    },
    {
      "epoch": 1.9221541493509275,
      "grad_norm": 9.00521469116211,
      "learning_rate": 6.775734729648447e-05,
      "loss": 2.752686309814453,
      "memory(GiB)": 77.56,
      "step": 44865,
      "token_acc": 0.47035573122529645,
      "train_speed(iter/s)": 1.440503
    },
    {
      "epoch": 1.9223683646801764,
      "grad_norm": 5.078679084777832,
      "learning_rate": 6.775105607323822e-05,
      "loss": 2.4538951873779298,
      "memory(GiB)": 77.56,
      "step": 44870,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.440493
    },
    {
      "epoch": 1.9225825800094256,
      "grad_norm": 3.5641767978668213,
      "learning_rate": 6.774476452841496e-05,
      "loss": 2.509532165527344,
      "memory(GiB)": 77.56,
      "step": 44875,
      "token_acc": 0.4592833876221498,
      "train_speed(iter/s)": 1.440483
    },
    {
      "epoch": 1.9227967953386744,
      "grad_norm": 4.824163913726807,
      "learning_rate": 6.773847266212863e-05,
      "loss": 2.3499183654785156,
      "memory(GiB)": 77.56,
      "step": 44880,
      "token_acc": 0.4942084942084942,
      "train_speed(iter/s)": 1.440487
    },
    {
      "epoch": 1.9230110106679232,
      "grad_norm": 4.4063286781311035,
      "learning_rate": 6.773218047449323e-05,
      "loss": 2.404673767089844,
      "memory(GiB)": 77.56,
      "step": 44885,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.440493
    },
    {
      "epoch": 1.9232252259971725,
      "grad_norm": 4.887564182281494,
      "learning_rate": 6.772588796562274e-05,
      "loss": 2.6430288314819337,
      "memory(GiB)": 77.56,
      "step": 44890,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.440511
    },
    {
      "epoch": 1.9234394413264213,
      "grad_norm": 4.752852439880371,
      "learning_rate": 6.771959513563116e-05,
      "loss": 2.498763847351074,
      "memory(GiB)": 77.56,
      "step": 44895,
      "token_acc": 0.4459016393442623,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.9236536566556701,
      "grad_norm": 4.58977746963501,
      "learning_rate": 6.77133019846325e-05,
      "loss": 2.3855897903442385,
      "memory(GiB)": 77.56,
      "step": 44900,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.440514
    },
    {
      "epoch": 1.9238678719849194,
      "grad_norm": 6.020086288452148,
      "learning_rate": 6.770700851274077e-05,
      "loss": 2.8141883850097655,
      "memory(GiB)": 77.56,
      "step": 44905,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.440546
    },
    {
      "epoch": 1.9240820873141682,
      "grad_norm": 3.6862447261810303,
      "learning_rate": 6.770071472006994e-05,
      "loss": 2.6929437637329103,
      "memory(GiB)": 77.56,
      "step": 44910,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.440548
    },
    {
      "epoch": 1.924296302643417,
      "grad_norm": 4.549350738525391,
      "learning_rate": 6.769442060673409e-05,
      "loss": 2.507354736328125,
      "memory(GiB)": 77.56,
      "step": 44915,
      "token_acc": 0.4513888888888889,
      "train_speed(iter/s)": 1.440546
    },
    {
      "epoch": 1.9245105179726663,
      "grad_norm": 6.182398796081543,
      "learning_rate": 6.768812617284719e-05,
      "loss": 2.4904422760009766,
      "memory(GiB)": 77.56,
      "step": 44920,
      "token_acc": 0.48338368580060426,
      "train_speed(iter/s)": 1.440578
    },
    {
      "epoch": 1.924724733301915,
      "grad_norm": 5.885412693023682,
      "learning_rate": 6.76818314185233e-05,
      "loss": 2.3250518798828126,
      "memory(GiB)": 77.56,
      "step": 44925,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.440568
    },
    {
      "epoch": 1.9249389486311639,
      "grad_norm": 4.711713790893555,
      "learning_rate": 6.767553634387645e-05,
      "loss": 2.893124580383301,
      "memory(GiB)": 77.56,
      "step": 44930,
      "token_acc": 0.4228395061728395,
      "train_speed(iter/s)": 1.440578
    },
    {
      "epoch": 1.9251531639604131,
      "grad_norm": 3.9892210960388184,
      "learning_rate": 6.766924094902067e-05,
      "loss": 2.800054168701172,
      "memory(GiB)": 77.56,
      "step": 44935,
      "token_acc": 0.44648318042813456,
      "train_speed(iter/s)": 1.440585
    },
    {
      "epoch": 1.925367379289662,
      "grad_norm": 5.91239595413208,
      "learning_rate": 6.766294523407002e-05,
      "loss": 2.2573455810546874,
      "memory(GiB)": 77.56,
      "step": 44940,
      "token_acc": 0.5039370078740157,
      "train_speed(iter/s)": 1.440607
    },
    {
      "epoch": 1.9255815946189108,
      "grad_norm": 5.328665733337402,
      "learning_rate": 6.765664919913856e-05,
      "loss": 2.2851661682128905,
      "memory(GiB)": 77.56,
      "step": 44945,
      "token_acc": 0.491869918699187,
      "train_speed(iter/s)": 1.440603
    },
    {
      "epoch": 1.92579580994816,
      "grad_norm": 6.843301296234131,
      "learning_rate": 6.765035284434031e-05,
      "loss": 2.48330078125,
      "memory(GiB)": 77.56,
      "step": 44950,
      "token_acc": 0.4491017964071856,
      "train_speed(iter/s)": 1.440594
    },
    {
      "epoch": 1.9260100252774088,
      "grad_norm": 5.048739910125732,
      "learning_rate": 6.764405616978938e-05,
      "loss": 2.105299377441406,
      "memory(GiB)": 77.56,
      "step": 44955,
      "token_acc": 0.509090909090909,
      "train_speed(iter/s)": 1.440618
    },
    {
      "epoch": 1.9262242406066579,
      "grad_norm": 4.746590614318848,
      "learning_rate": 6.763775917559982e-05,
      "loss": 2.259245681762695,
      "memory(GiB)": 77.56,
      "step": 44960,
      "token_acc": 0.5278810408921933,
      "train_speed(iter/s)": 1.440628
    },
    {
      "epoch": 1.926438455935907,
      "grad_norm": 5.139850616455078,
      "learning_rate": 6.76314618618857e-05,
      "loss": 2.58447265625,
      "memory(GiB)": 77.56,
      "step": 44965,
      "token_acc": 0.4389438943894389,
      "train_speed(iter/s)": 1.440682
    },
    {
      "epoch": 1.9266526712651557,
      "grad_norm": 6.878082275390625,
      "learning_rate": 6.76251642287611e-05,
      "loss": 2.3195016860961912,
      "memory(GiB)": 77.56,
      "step": 44970,
      "token_acc": 0.5077881619937694,
      "train_speed(iter/s)": 1.440702
    },
    {
      "epoch": 1.9268668865944047,
      "grad_norm": 5.152690410614014,
      "learning_rate": 6.761886627634013e-05,
      "loss": 2.4266490936279297,
      "memory(GiB)": 77.56,
      "step": 44975,
      "token_acc": 0.4898785425101215,
      "train_speed(iter/s)": 1.440721
    },
    {
      "epoch": 1.9270811019236538,
      "grad_norm": 6.772485733032227,
      "learning_rate": 6.761256800473686e-05,
      "loss": 2.4867437362670897,
      "memory(GiB)": 77.56,
      "step": 44980,
      "token_acc": 0.4420289855072464,
      "train_speed(iter/s)": 1.440716
    },
    {
      "epoch": 1.9272953172529026,
      "grad_norm": 6.258435249328613,
      "learning_rate": 6.760626941406541e-05,
      "loss": 2.5491941452026365,
      "memory(GiB)": 77.56,
      "step": 44985,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.440747
    },
    {
      "epoch": 1.9275095325821516,
      "grad_norm": 7.090802192687988,
      "learning_rate": 6.759997050443987e-05,
      "loss": 2.7601268768310545,
      "memory(GiB)": 77.56,
      "step": 44990,
      "token_acc": 0.47151898734177217,
      "train_speed(iter/s)": 1.440739
    },
    {
      "epoch": 1.9277237479114007,
      "grad_norm": 4.356286525726318,
      "learning_rate": 6.759367127597436e-05,
      "loss": 2.4777576446533205,
      "memory(GiB)": 77.56,
      "step": 44995,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.440747
    },
    {
      "epoch": 1.9279379632406495,
      "grad_norm": 4.196362495422363,
      "learning_rate": 6.758737172878298e-05,
      "loss": 2.5145463943481445,
      "memory(GiB)": 77.56,
      "step": 45000,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.440746
    },
    {
      "epoch": 1.9279379632406495,
      "eval_loss": 2.272648572921753,
      "eval_runtime": 13.8851,
      "eval_samples_per_second": 7.202,
      "eval_steps_per_second": 7.202,
      "eval_token_acc": 0.4802547770700637,
      "step": 45000
    },
    {
      "epoch": 1.9281521785698985,
      "grad_norm": 4.798551559448242,
      "learning_rate": 6.758107186297987e-05,
      "loss": 2.6651744842529297,
      "memory(GiB)": 77.56,
      "step": 45005,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.440106
    },
    {
      "epoch": 1.9283663938991475,
      "grad_norm": 6.122073173522949,
      "learning_rate": 6.757477167867914e-05,
      "loss": 2.641393280029297,
      "memory(GiB)": 77.56,
      "step": 45010,
      "token_acc": 0.4526627218934911,
      "train_speed(iter/s)": 1.440144
    },
    {
      "epoch": 1.9285806092283964,
      "grad_norm": 7.3333563804626465,
      "learning_rate": 6.756847117599495e-05,
      "loss": 2.451889991760254,
      "memory(GiB)": 77.56,
      "step": 45015,
      "token_acc": 0.5186440677966102,
      "train_speed(iter/s)": 1.440179
    },
    {
      "epoch": 1.9287948245576454,
      "grad_norm": 4.334657669067383,
      "learning_rate": 6.756217035504142e-05,
      "loss": 2.4974742889404298,
      "memory(GiB)": 77.56,
      "step": 45020,
      "token_acc": 0.4887640449438202,
      "train_speed(iter/s)": 1.440185
    },
    {
      "epoch": 1.9290090398868944,
      "grad_norm": 5.305824279785156,
      "learning_rate": 6.755586921593268e-05,
      "loss": 2.381949615478516,
      "memory(GiB)": 77.56,
      "step": 45025,
      "token_acc": 0.49262536873156343,
      "train_speed(iter/s)": 1.440152
    },
    {
      "epoch": 1.9292232552161432,
      "grad_norm": 5.9110212326049805,
      "learning_rate": 6.754956775878294e-05,
      "loss": 2.5106342315673826,
      "memory(GiB)": 77.56,
      "step": 45030,
      "token_acc": 0.44879518072289154,
      "train_speed(iter/s)": 1.440186
    },
    {
      "epoch": 1.9294374705453923,
      "grad_norm": 5.255375385284424,
      "learning_rate": 6.754326598370628e-05,
      "loss": 2.6439294815063477,
      "memory(GiB)": 77.56,
      "step": 45035,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.440199
    },
    {
      "epoch": 1.9296516858746413,
      "grad_norm": 6.076241970062256,
      "learning_rate": 6.75369638908169e-05,
      "loss": 2.4158645629882813,
      "memory(GiB)": 77.56,
      "step": 45040,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.440223
    },
    {
      "epoch": 1.9298659012038901,
      "grad_norm": 4.477628707885742,
      "learning_rate": 6.7530661480229e-05,
      "loss": 2.7268402099609377,
      "memory(GiB)": 77.56,
      "step": 45045,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.440238
    },
    {
      "epoch": 1.9300801165331392,
      "grad_norm": 5.206740856170654,
      "learning_rate": 6.75243587520567e-05,
      "loss": 2.4599994659423827,
      "memory(GiB)": 77.56,
      "step": 45050,
      "token_acc": 0.4717741935483871,
      "train_speed(iter/s)": 1.440253
    },
    {
      "epoch": 1.9302943318623882,
      "grad_norm": 4.140765190124512,
      "learning_rate": 6.751805570641421e-05,
      "loss": 2.6211193084716795,
      "memory(GiB)": 77.56,
      "step": 45055,
      "token_acc": 0.4435483870967742,
      "train_speed(iter/s)": 1.44029
    },
    {
      "epoch": 1.930508547191637,
      "grad_norm": 4.071769714355469,
      "learning_rate": 6.75117523434157e-05,
      "loss": 2.774671936035156,
      "memory(GiB)": 77.56,
      "step": 45060,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.440309
    },
    {
      "epoch": 1.930722762520886,
      "grad_norm": 8.741419792175293,
      "learning_rate": 6.750544866317536e-05,
      "loss": 2.596161460876465,
      "memory(GiB)": 77.56,
      "step": 45065,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.440266
    },
    {
      "epoch": 1.930936977850135,
      "grad_norm": 4.923036098480225,
      "learning_rate": 6.749914466580741e-05,
      "loss": 2.7943634033203124,
      "memory(GiB)": 77.56,
      "step": 45070,
      "token_acc": 0.43304843304843305,
      "train_speed(iter/s)": 1.44027
    },
    {
      "epoch": 1.9311511931793839,
      "grad_norm": 4.8703293800354,
      "learning_rate": 6.749284035142604e-05,
      "loss": 2.632495880126953,
      "memory(GiB)": 77.56,
      "step": 45075,
      "token_acc": 0.4867924528301887,
      "train_speed(iter/s)": 1.440285
    },
    {
      "epoch": 1.931365408508633,
      "grad_norm": 5.802798271179199,
      "learning_rate": 6.748653572014546e-05,
      "loss": 2.656983757019043,
      "memory(GiB)": 77.56,
      "step": 45080,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.44025
    },
    {
      "epoch": 1.931579623837882,
      "grad_norm": 3.893988847732544,
      "learning_rate": 6.748023077207988e-05,
      "loss": 2.6880258560180663,
      "memory(GiB)": 77.56,
      "step": 45085,
      "token_acc": 0.4592833876221498,
      "train_speed(iter/s)": 1.440248
    },
    {
      "epoch": 1.9317938391671308,
      "grad_norm": 3.6687610149383545,
      "learning_rate": 6.747392550734352e-05,
      "loss": 2.22630615234375,
      "memory(GiB)": 77.56,
      "step": 45090,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.440234
    },
    {
      "epoch": 1.9320080544963798,
      "grad_norm": 7.48464298248291,
      "learning_rate": 6.74676199260506e-05,
      "loss": 2.7770549774169924,
      "memory(GiB)": 77.56,
      "step": 45095,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.440266
    },
    {
      "epoch": 1.9322222698256288,
      "grad_norm": 5.711358547210693,
      "learning_rate": 6.746131402831537e-05,
      "loss": 2.421975517272949,
      "memory(GiB)": 77.56,
      "step": 45100,
      "token_acc": 0.49795918367346936,
      "train_speed(iter/s)": 1.440316
    },
    {
      "epoch": 1.9324364851548776,
      "grad_norm": 5.555703639984131,
      "learning_rate": 6.745500781425205e-05,
      "loss": 2.902573013305664,
      "memory(GiB)": 77.56,
      "step": 45105,
      "token_acc": 0.45141065830721006,
      "train_speed(iter/s)": 1.440321
    },
    {
      "epoch": 1.9326507004841267,
      "grad_norm": 4.670094966888428,
      "learning_rate": 6.74487012839749e-05,
      "loss": 2.3642473220825195,
      "memory(GiB)": 77.56,
      "step": 45110,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.440276
    },
    {
      "epoch": 1.9328649158133757,
      "grad_norm": 5.587475776672363,
      "learning_rate": 6.744239443759813e-05,
      "loss": 2.6430273056030273,
      "memory(GiB)": 77.56,
      "step": 45115,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440217
    },
    {
      "epoch": 1.9330791311426245,
      "grad_norm": 5.119340896606445,
      "learning_rate": 6.743608727523605e-05,
      "loss": 2.14676456451416,
      "memory(GiB)": 77.56,
      "step": 45120,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.440203
    },
    {
      "epoch": 1.9332933464718736,
      "grad_norm": 5.657715797424316,
      "learning_rate": 6.742977979700287e-05,
      "loss": 2.621632957458496,
      "memory(GiB)": 77.56,
      "step": 45125,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440213
    },
    {
      "epoch": 1.9335075618011226,
      "grad_norm": 5.945922374725342,
      "learning_rate": 6.74234720030129e-05,
      "loss": 2.449910354614258,
      "memory(GiB)": 77.56,
      "step": 45130,
      "token_acc": 0.5205047318611987,
      "train_speed(iter/s)": 1.440203
    },
    {
      "epoch": 1.9337217771303714,
      "grad_norm": 8.12584114074707,
      "learning_rate": 6.741716389338039e-05,
      "loss": 2.1450578689575197,
      "memory(GiB)": 77.56,
      "step": 45135,
      "token_acc": 0.5283842794759825,
      "train_speed(iter/s)": 1.440231
    },
    {
      "epoch": 1.9339359924596204,
      "grad_norm": 4.752293586730957,
      "learning_rate": 6.74108554682196e-05,
      "loss": 2.821733093261719,
      "memory(GiB)": 77.56,
      "step": 45140,
      "token_acc": 0.4148148148148148,
      "train_speed(iter/s)": 1.440244
    },
    {
      "epoch": 1.9341502077888695,
      "grad_norm": 7.279443740844727,
      "learning_rate": 6.740454672764484e-05,
      "loss": 2.858756256103516,
      "memory(GiB)": 77.56,
      "step": 45145,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.44027
    },
    {
      "epoch": 1.9343644231181183,
      "grad_norm": 6.130493640899658,
      "learning_rate": 6.739823767177039e-05,
      "loss": 2.5429866790771483,
      "memory(GiB)": 77.56,
      "step": 45150,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.440241
    },
    {
      "epoch": 1.9345786384473673,
      "grad_norm": 8.559672355651855,
      "learning_rate": 6.739192830071054e-05,
      "loss": 2.274237632751465,
      "memory(GiB)": 77.56,
      "step": 45155,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.440258
    },
    {
      "epoch": 1.9347928537766164,
      "grad_norm": 5.867745399475098,
      "learning_rate": 6.73856186145796e-05,
      "loss": 2.4978317260742187,
      "memory(GiB)": 77.56,
      "step": 45160,
      "token_acc": 0.5127272727272727,
      "train_speed(iter/s)": 1.440277
    },
    {
      "epoch": 1.9350070691058652,
      "grad_norm": 4.435779094696045,
      "learning_rate": 6.737930861349186e-05,
      "loss": 2.5070775985717773,
      "memory(GiB)": 77.56,
      "step": 45165,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.440217
    },
    {
      "epoch": 1.9352212844351142,
      "grad_norm": 4.044161319732666,
      "learning_rate": 6.737299829756165e-05,
      "loss": 2.232552909851074,
      "memory(GiB)": 77.56,
      "step": 45170,
      "token_acc": 0.4901315789473684,
      "train_speed(iter/s)": 1.440213
    },
    {
      "epoch": 1.9354354997643632,
      "grad_norm": 4.500838756561279,
      "learning_rate": 6.736668766690328e-05,
      "loss": 2.3945945739746093,
      "memory(GiB)": 77.56,
      "step": 45175,
      "token_acc": 0.44574780058651026,
      "train_speed(iter/s)": 1.44025
    },
    {
      "epoch": 1.935649715093612,
      "grad_norm": 4.921265602111816,
      "learning_rate": 6.736037672163107e-05,
      "loss": 2.016653060913086,
      "memory(GiB)": 77.56,
      "step": 45180,
      "token_acc": 0.5225563909774437,
      "train_speed(iter/s)": 1.44026
    },
    {
      "epoch": 1.935863930422861,
      "grad_norm": 5.894145488739014,
      "learning_rate": 6.735406546185934e-05,
      "loss": 2.7764585494995115,
      "memory(GiB)": 77.56,
      "step": 45185,
      "token_acc": 0.46484375,
      "train_speed(iter/s)": 1.440254
    },
    {
      "epoch": 1.9360781457521101,
      "grad_norm": 6.673031330108643,
      "learning_rate": 6.734775388770245e-05,
      "loss": 2.493510437011719,
      "memory(GiB)": 77.56,
      "step": 45190,
      "token_acc": 0.4483870967741935,
      "train_speed(iter/s)": 1.440279
    },
    {
      "epoch": 1.936292361081359,
      "grad_norm": 6.107112884521484,
      "learning_rate": 6.734144199927472e-05,
      "loss": 2.3875770568847656,
      "memory(GiB)": 77.56,
      "step": 45195,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.440264
    },
    {
      "epoch": 1.936506576410608,
      "grad_norm": 5.024647235870361,
      "learning_rate": 6.73351297966905e-05,
      "loss": 2.5815534591674805,
      "memory(GiB)": 77.56,
      "step": 45200,
      "token_acc": 0.47039473684210525,
      "train_speed(iter/s)": 1.440236
    },
    {
      "epoch": 1.936720791739857,
      "grad_norm": 3.971710443496704,
      "learning_rate": 6.732881728006414e-05,
      "loss": 2.6231557846069338,
      "memory(GiB)": 77.56,
      "step": 45205,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.440277
    },
    {
      "epoch": 1.9369350070691058,
      "grad_norm": 4.264815807342529,
      "learning_rate": 6.732250444951e-05,
      "loss": 2.3905750274658204,
      "memory(GiB)": 77.56,
      "step": 45210,
      "token_acc": 0.506896551724138,
      "train_speed(iter/s)": 1.440307
    },
    {
      "epoch": 1.9371492223983549,
      "grad_norm": 4.220180988311768,
      "learning_rate": 6.731619130514243e-05,
      "loss": 2.6343719482421877,
      "memory(GiB)": 77.56,
      "step": 45215,
      "token_acc": 0.4486301369863014,
      "train_speed(iter/s)": 1.440281
    },
    {
      "epoch": 1.9373634377276039,
      "grad_norm": 6.077444553375244,
      "learning_rate": 6.730987784707584e-05,
      "loss": 2.263299560546875,
      "memory(GiB)": 77.56,
      "step": 45220,
      "token_acc": 0.5278969957081545,
      "train_speed(iter/s)": 1.440303
    },
    {
      "epoch": 1.9375776530568527,
      "grad_norm": 5.1404314041137695,
      "learning_rate": 6.730356407542456e-05,
      "loss": 2.5684535980224608,
      "memory(GiB)": 77.56,
      "step": 45225,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.440346
    },
    {
      "epoch": 1.9377918683861017,
      "grad_norm": 6.451222896575928,
      "learning_rate": 6.729724999030297e-05,
      "loss": 2.6400772094726563,
      "memory(GiB)": 77.56,
      "step": 45230,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.44032
    },
    {
      "epoch": 1.9380060837153508,
      "grad_norm": 7.853918552398682,
      "learning_rate": 6.729093559182549e-05,
      "loss": 2.437779998779297,
      "memory(GiB)": 77.56,
      "step": 45235,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.44028
    },
    {
      "epoch": 1.9382202990445996,
      "grad_norm": 4.741629123687744,
      "learning_rate": 6.728462088010647e-05,
      "loss": 2.8410581588745116,
      "memory(GiB)": 77.56,
      "step": 45240,
      "token_acc": 0.44542772861356933,
      "train_speed(iter/s)": 1.440296
    },
    {
      "epoch": 1.9384345143738486,
      "grad_norm": 6.1778178215026855,
      "learning_rate": 6.727830585526035e-05,
      "loss": 2.702621841430664,
      "memory(GiB)": 77.56,
      "step": 45245,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.440314
    },
    {
      "epoch": 1.9386487297030977,
      "grad_norm": 4.489249229431152,
      "learning_rate": 6.727199051740151e-05,
      "loss": 2.316697692871094,
      "memory(GiB)": 77.56,
      "step": 45250,
      "token_acc": 0.5265306122448979,
      "train_speed(iter/s)": 1.440355
    },
    {
      "epoch": 1.9388629450323465,
      "grad_norm": 4.341643333435059,
      "learning_rate": 6.726567486664434e-05,
      "loss": 2.388902473449707,
      "memory(GiB)": 77.56,
      "step": 45255,
      "token_acc": 0.4716417910447761,
      "train_speed(iter/s)": 1.440368
    },
    {
      "epoch": 1.9390771603615955,
      "grad_norm": 4.554438591003418,
      "learning_rate": 6.725935890310328e-05,
      "loss": 2.7564958572387694,
      "memory(GiB)": 77.56,
      "step": 45260,
      "token_acc": 0.419672131147541,
      "train_speed(iter/s)": 1.440417
    },
    {
      "epoch": 1.9392913756908445,
      "grad_norm": 5.8008198738098145,
      "learning_rate": 6.725304262689275e-05,
      "loss": 2.707223129272461,
      "memory(GiB)": 77.56,
      "step": 45265,
      "token_acc": 0.42452830188679247,
      "train_speed(iter/s)": 1.440425
    },
    {
      "epoch": 1.9395055910200933,
      "grad_norm": 7.180793762207031,
      "learning_rate": 6.724672603812717e-05,
      "loss": 2.512354278564453,
      "memory(GiB)": 77.56,
      "step": 45270,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.44045
    },
    {
      "epoch": 1.9397198063493424,
      "grad_norm": 4.4878973960876465,
      "learning_rate": 6.724040913692098e-05,
      "loss": 2.7086597442626954,
      "memory(GiB)": 77.56,
      "step": 45275,
      "token_acc": 0.46099290780141844,
      "train_speed(iter/s)": 1.440477
    },
    {
      "epoch": 1.9399340216785914,
      "grad_norm": 5.306338310241699,
      "learning_rate": 6.723409192338859e-05,
      "loss": 2.6740139007568358,
      "memory(GiB)": 77.56,
      "step": 45280,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.440466
    },
    {
      "epoch": 1.9401482370078402,
      "grad_norm": 6.707126140594482,
      "learning_rate": 6.722777439764446e-05,
      "loss": 2.289545249938965,
      "memory(GiB)": 77.56,
      "step": 45285,
      "token_acc": 0.5280898876404494,
      "train_speed(iter/s)": 1.440469
    },
    {
      "epoch": 1.9403624523370893,
      "grad_norm": 7.800013542175293,
      "learning_rate": 6.722145655980304e-05,
      "loss": 2.498686408996582,
      "memory(GiB)": 77.56,
      "step": 45290,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.44045
    },
    {
      "epoch": 1.9405766676663383,
      "grad_norm": 5.396202564239502,
      "learning_rate": 6.721513840997878e-05,
      "loss": 2.5861148834228516,
      "memory(GiB)": 77.56,
      "step": 45295,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.440417
    },
    {
      "epoch": 1.940790882995587,
      "grad_norm": 5.635911464691162,
      "learning_rate": 6.720881994828612e-05,
      "loss": 2.375554656982422,
      "memory(GiB)": 77.56,
      "step": 45300,
      "token_acc": 0.5098814229249012,
      "train_speed(iter/s)": 1.440463
    },
    {
      "epoch": 1.9410050983248361,
      "grad_norm": 4.748110771179199,
      "learning_rate": 6.720250117483956e-05,
      "loss": 2.6421573638916014,
      "memory(GiB)": 77.56,
      "step": 45305,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.440456
    },
    {
      "epoch": 1.9412193136540852,
      "grad_norm": 3.5759501457214355,
      "learning_rate": 6.719618208975355e-05,
      "loss": 2.7423301696777345,
      "memory(GiB)": 77.56,
      "step": 45310,
      "token_acc": 0.4565826330532213,
      "train_speed(iter/s)": 1.440464
    },
    {
      "epoch": 1.941433528983334,
      "grad_norm": 5.056948661804199,
      "learning_rate": 6.718986269314257e-05,
      "loss": 2.43087158203125,
      "memory(GiB)": 77.56,
      "step": 45315,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.440469
    },
    {
      "epoch": 1.941647744312583,
      "grad_norm": 3.962507724761963,
      "learning_rate": 6.718354298512111e-05,
      "loss": 2.5097015380859373,
      "memory(GiB)": 77.56,
      "step": 45320,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.440485
    },
    {
      "epoch": 1.941861959641832,
      "grad_norm": 4.559586524963379,
      "learning_rate": 6.717722296580364e-05,
      "loss": 2.702666091918945,
      "memory(GiB)": 77.56,
      "step": 45325,
      "token_acc": 0.476038338658147,
      "train_speed(iter/s)": 1.440475
    },
    {
      "epoch": 1.9420761749710809,
      "grad_norm": 5.075676441192627,
      "learning_rate": 6.717090263530466e-05,
      "loss": 2.6583606719970705,
      "memory(GiB)": 77.56,
      "step": 45330,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.440498
    },
    {
      "epoch": 1.94229039030033,
      "grad_norm": 5.0208516120910645,
      "learning_rate": 6.716458199373868e-05,
      "loss": 2.1088844299316407,
      "memory(GiB)": 77.56,
      "step": 45335,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.44052
    },
    {
      "epoch": 1.942504605629579,
      "grad_norm": 5.868277549743652,
      "learning_rate": 6.71582610412202e-05,
      "loss": 2.520858955383301,
      "memory(GiB)": 77.56,
      "step": 45340,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440534
    },
    {
      "epoch": 1.9427188209588278,
      "grad_norm": 4.251816749572754,
      "learning_rate": 6.71519397778637e-05,
      "loss": 2.189457130432129,
      "memory(GiB)": 77.56,
      "step": 45345,
      "token_acc": 0.5198412698412699,
      "train_speed(iter/s)": 1.4405
    },
    {
      "epoch": 1.9429330362880768,
      "grad_norm": 10.378301620483398,
      "learning_rate": 6.714561820378375e-05,
      "loss": 2.5810359954833983,
      "memory(GiB)": 77.56,
      "step": 45350,
      "token_acc": 0.45723684210526316,
      "train_speed(iter/s)": 1.440525
    },
    {
      "epoch": 1.9431472516173258,
      "grad_norm": 5.028270721435547,
      "learning_rate": 6.713929631909483e-05,
      "loss": 2.4318796157836915,
      "memory(GiB)": 77.56,
      "step": 45355,
      "token_acc": 0.5062240663900415,
      "train_speed(iter/s)": 1.440524
    },
    {
      "epoch": 1.9433614669465746,
      "grad_norm": 5.1753249168396,
      "learning_rate": 6.713297412391148e-05,
      "loss": 2.6329109191894533,
      "memory(GiB)": 77.56,
      "step": 45360,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.440544
    },
    {
      "epoch": 1.9435756822758237,
      "grad_norm": 4.587521076202393,
      "learning_rate": 6.712665161834822e-05,
      "loss": 2.2890296936035157,
      "memory(GiB)": 77.56,
      "step": 45365,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.440558
    },
    {
      "epoch": 1.9437898976050727,
      "grad_norm": 3.887960433959961,
      "learning_rate": 6.712032880251962e-05,
      "loss": 2.903134346008301,
      "memory(GiB)": 77.56,
      "step": 45370,
      "token_acc": 0.43389830508474575,
      "train_speed(iter/s)": 1.44056
    },
    {
      "epoch": 1.9440041129343215,
      "grad_norm": 4.430891990661621,
      "learning_rate": 6.711400567654019e-05,
      "loss": 2.468527984619141,
      "memory(GiB)": 77.56,
      "step": 45375,
      "token_acc": 0.4809384164222874,
      "train_speed(iter/s)": 1.440573
    },
    {
      "epoch": 1.9442183282635705,
      "grad_norm": 4.241114139556885,
      "learning_rate": 6.710768224052451e-05,
      "loss": 2.6923124313354494,
      "memory(GiB)": 77.56,
      "step": 45380,
      "token_acc": 0.42136498516320475,
      "train_speed(iter/s)": 1.440584
    },
    {
      "epoch": 1.9444325435928196,
      "grad_norm": 3.6667704582214355,
      "learning_rate": 6.710135849458709e-05,
      "loss": 2.2617155075073243,
      "memory(GiB)": 77.56,
      "step": 45385,
      "token_acc": 0.5131964809384164,
      "train_speed(iter/s)": 1.440599
    },
    {
      "epoch": 1.9446467589220684,
      "grad_norm": 7.395628929138184,
      "learning_rate": 6.709503443884255e-05,
      "loss": 2.6145151138305662,
      "memory(GiB)": 77.56,
      "step": 45390,
      "token_acc": 0.5038461538461538,
      "train_speed(iter/s)": 1.440579
    },
    {
      "epoch": 1.9448609742513174,
      "grad_norm": 4.459097862243652,
      "learning_rate": 6.708871007340541e-05,
      "loss": 2.515476417541504,
      "memory(GiB)": 77.56,
      "step": 45395,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.440598
    },
    {
      "epoch": 1.9450751895805665,
      "grad_norm": 4.505022048950195,
      "learning_rate": 6.708238539839025e-05,
      "loss": 2.3406009674072266,
      "memory(GiB)": 77.56,
      "step": 45400,
      "token_acc": 0.5015105740181269,
      "train_speed(iter/s)": 1.440618
    },
    {
      "epoch": 1.9452894049098153,
      "grad_norm": 6.577566623687744,
      "learning_rate": 6.707606041391165e-05,
      "loss": 2.7817224502563476,
      "memory(GiB)": 77.56,
      "step": 45405,
      "token_acc": 0.41198501872659177,
      "train_speed(iter/s)": 1.440643
    },
    {
      "epoch": 1.9455036202390643,
      "grad_norm": 4.262021064758301,
      "learning_rate": 6.706973512008421e-05,
      "loss": 2.657246398925781,
      "memory(GiB)": 77.56,
      "step": 45410,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.440658
    },
    {
      "epoch": 1.9457178355683133,
      "grad_norm": 4.189906597137451,
      "learning_rate": 6.70634095170225e-05,
      "loss": 2.511100959777832,
      "memory(GiB)": 77.56,
      "step": 45415,
      "token_acc": 0.4541984732824427,
      "train_speed(iter/s)": 1.440674
    },
    {
      "epoch": 1.9459320508975622,
      "grad_norm": 4.1230149269104,
      "learning_rate": 6.705708360484111e-05,
      "loss": 2.6832096099853517,
      "memory(GiB)": 77.56,
      "step": 45420,
      "token_acc": 0.47440273037542663,
      "train_speed(iter/s)": 1.440676
    },
    {
      "epoch": 1.9461462662268112,
      "grad_norm": 5.072758197784424,
      "learning_rate": 6.705075738365468e-05,
      "loss": 2.613764762878418,
      "memory(GiB)": 77.56,
      "step": 45425,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.440597
    },
    {
      "epoch": 1.9463604815560602,
      "grad_norm": 4.666149139404297,
      "learning_rate": 6.704443085357777e-05,
      "loss": 2.439645004272461,
      "memory(GiB)": 77.56,
      "step": 45430,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.440635
    },
    {
      "epoch": 1.946574696885309,
      "grad_norm": 12.834710121154785,
      "learning_rate": 6.7038104014725e-05,
      "loss": 2.3617557525634765,
      "memory(GiB)": 77.56,
      "step": 45435,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.440632
    },
    {
      "epoch": 1.946788912214558,
      "grad_norm": 4.6653008460998535,
      "learning_rate": 6.703177686721098e-05,
      "loss": 2.675905227661133,
      "memory(GiB)": 77.56,
      "step": 45440,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440643
    },
    {
      "epoch": 1.947003127543807,
      "grad_norm": 7.687556743621826,
      "learning_rate": 6.702544941115036e-05,
      "loss": 2.4108144760131838,
      "memory(GiB)": 77.56,
      "step": 45445,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.440635
    },
    {
      "epoch": 1.947217342873056,
      "grad_norm": 4.9494123458862305,
      "learning_rate": 6.701912164665776e-05,
      "loss": 2.415040969848633,
      "memory(GiB)": 77.56,
      "step": 45450,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.440602
    },
    {
      "epoch": 1.947431558202305,
      "grad_norm": 6.328458309173584,
      "learning_rate": 6.70127935738478e-05,
      "loss": 2.236003303527832,
      "memory(GiB)": 77.56,
      "step": 45455,
      "token_acc": 0.5427509293680297,
      "train_speed(iter/s)": 1.440557
    },
    {
      "epoch": 1.947645773531554,
      "grad_norm": 4.472757339477539,
      "learning_rate": 6.700646519283513e-05,
      "loss": 2.6073421478271483,
      "memory(GiB)": 77.56,
      "step": 45460,
      "token_acc": 0.4907749077490775,
      "train_speed(iter/s)": 1.44059
    },
    {
      "epoch": 1.9478599888608028,
      "grad_norm": 6.233323097229004,
      "learning_rate": 6.700013650373438e-05,
      "loss": 2.3979076385498046,
      "memory(GiB)": 77.56,
      "step": 45465,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.440606
    },
    {
      "epoch": 1.9480742041900518,
      "grad_norm": 4.534661769866943,
      "learning_rate": 6.699380750666023e-05,
      "loss": 2.769176483154297,
      "memory(GiB)": 77.56,
      "step": 45470,
      "token_acc": 0.46200607902735563,
      "train_speed(iter/s)": 1.440627
    },
    {
      "epoch": 1.9482884195193009,
      "grad_norm": 4.951756000518799,
      "learning_rate": 6.698747820172728e-05,
      "loss": 2.579067611694336,
      "memory(GiB)": 77.56,
      "step": 45475,
      "token_acc": 0.4161490683229814,
      "train_speed(iter/s)": 1.440667
    },
    {
      "epoch": 1.9485026348485497,
      "grad_norm": 4.349664688110352,
      "learning_rate": 6.698114858905024e-05,
      "loss": 2.4291372299194336,
      "memory(GiB)": 77.56,
      "step": 45480,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.440673
    },
    {
      "epoch": 1.9487168501777987,
      "grad_norm": 4.892009258270264,
      "learning_rate": 6.697481866874377e-05,
      "loss": 2.6150508880615235,
      "memory(GiB)": 77.56,
      "step": 45485,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.440728
    },
    {
      "epoch": 1.9489310655070478,
      "grad_norm": 4.527082920074463,
      "learning_rate": 6.696848844092254e-05,
      "loss": 2.3411970138549805,
      "memory(GiB)": 77.56,
      "step": 45490,
      "token_acc": 0.4440894568690096,
      "train_speed(iter/s)": 1.440684
    },
    {
      "epoch": 1.9491452808362966,
      "grad_norm": 4.689889430999756,
      "learning_rate": 6.696215790570122e-05,
      "loss": 2.333414077758789,
      "memory(GiB)": 77.56,
      "step": 45495,
      "token_acc": 0.4576271186440678,
      "train_speed(iter/s)": 1.440675
    },
    {
      "epoch": 1.9493594961655456,
      "grad_norm": 4.693789958953857,
      "learning_rate": 6.69558270631945e-05,
      "loss": 2.806296539306641,
      "memory(GiB)": 77.56,
      "step": 45500,
      "token_acc": 0.39457831325301207,
      "train_speed(iter/s)": 1.440636
    },
    {
      "epoch": 1.9493594961655456,
      "eval_loss": 2.266148328781128,
      "eval_runtime": 13.4619,
      "eval_samples_per_second": 7.428,
      "eval_steps_per_second": 7.428,
      "eval_token_acc": 0.4514285714285714,
      "step": 45500
    },
    {
      "epoch": 1.9495737114947946,
      "grad_norm": 5.273241996765137,
      "learning_rate": 6.694949591351706e-05,
      "loss": 2.4877269744873045,
      "memory(GiB)": 77.56,
      "step": 45505,
      "token_acc": 0.44410569105691056,
      "train_speed(iter/s)": 1.44
    },
    {
      "epoch": 1.9497879268240434,
      "grad_norm": 5.049350738525391,
      "learning_rate": 6.694316445678363e-05,
      "loss": 2.6786537170410156,
      "memory(GiB)": 77.56,
      "step": 45510,
      "token_acc": 0.44029850746268656,
      "train_speed(iter/s)": 1.44
    },
    {
      "epoch": 1.9500021421532925,
      "grad_norm": 4.017052173614502,
      "learning_rate": 6.693683269310884e-05,
      "loss": 2.529633331298828,
      "memory(GiB)": 77.56,
      "step": 45515,
      "token_acc": 0.5089605734767025,
      "train_speed(iter/s)": 1.439979
    },
    {
      "epoch": 1.9502163574825415,
      "grad_norm": 4.5068159103393555,
      "learning_rate": 6.693050062260747e-05,
      "loss": 2.8497472763061524,
      "memory(GiB)": 77.56,
      "step": 45520,
      "token_acc": 0.44876325088339225,
      "train_speed(iter/s)": 1.439973
    },
    {
      "epoch": 1.9504305728117903,
      "grad_norm": 4.838365077972412,
      "learning_rate": 6.692416824539418e-05,
      "loss": 2.4935396194458006,
      "memory(GiB)": 77.56,
      "step": 45525,
      "token_acc": 0.5033112582781457,
      "train_speed(iter/s)": 1.439976
    },
    {
      "epoch": 1.9506447881410394,
      "grad_norm": 5.676723480224609,
      "learning_rate": 6.69178355615837e-05,
      "loss": 2.3777151107788086,
      "memory(GiB)": 77.56,
      "step": 45530,
      "token_acc": 0.4957627118644068,
      "train_speed(iter/s)": 1.439993
    },
    {
      "epoch": 1.9508590034702884,
      "grad_norm": 6.021706581115723,
      "learning_rate": 6.691150257129077e-05,
      "loss": 2.5612789154052735,
      "memory(GiB)": 77.56,
      "step": 45535,
      "token_acc": 0.4701195219123506,
      "train_speed(iter/s)": 1.440017
    },
    {
      "epoch": 1.9510732187995372,
      "grad_norm": 5.214600563049316,
      "learning_rate": 6.690516927463012e-05,
      "loss": 2.411444664001465,
      "memory(GiB)": 77.56,
      "step": 45540,
      "token_acc": 0.4735202492211838,
      "train_speed(iter/s)": 1.440065
    },
    {
      "epoch": 1.9512874341287862,
      "grad_norm": 4.718339920043945,
      "learning_rate": 6.689883567171646e-05,
      "loss": 2.5888877868652345,
      "memory(GiB)": 77.56,
      "step": 45545,
      "token_acc": 0.4651898734177215,
      "train_speed(iter/s)": 1.440084
    },
    {
      "epoch": 1.9515016494580353,
      "grad_norm": 4.600925922393799,
      "learning_rate": 6.689250176266452e-05,
      "loss": 2.4571081161499024,
      "memory(GiB)": 77.56,
      "step": 45550,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.440073
    },
    {
      "epoch": 1.951715864787284,
      "grad_norm": 4.214907169342041,
      "learning_rate": 6.688616754758908e-05,
      "loss": 2.537423324584961,
      "memory(GiB)": 77.56,
      "step": 45555,
      "token_acc": 0.4727272727272727,
      "train_speed(iter/s)": 1.44006
    },
    {
      "epoch": 1.9519300801165331,
      "grad_norm": 4.54862117767334,
      "learning_rate": 6.687983302660487e-05,
      "loss": 2.7397626876831054,
      "memory(GiB)": 77.56,
      "step": 45560,
      "token_acc": 0.42765273311897106,
      "train_speed(iter/s)": 1.44004
    },
    {
      "epoch": 1.9521442954457822,
      "grad_norm": 4.2088518142700195,
      "learning_rate": 6.687349819982667e-05,
      "loss": 2.4314159393310546,
      "memory(GiB)": 77.56,
      "step": 45565,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.440037
    },
    {
      "epoch": 1.952358510775031,
      "grad_norm": 5.273055076599121,
      "learning_rate": 6.686716306736921e-05,
      "loss": 2.3553470611572265,
      "memory(GiB)": 77.56,
      "step": 45570,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.440011
    },
    {
      "epoch": 1.95257272610428,
      "grad_norm": 6.307897567749023,
      "learning_rate": 6.686082762934725e-05,
      "loss": 2.4227832794189452,
      "memory(GiB)": 77.56,
      "step": 45575,
      "token_acc": 0.48638132295719844,
      "train_speed(iter/s)": 1.440018
    },
    {
      "epoch": 1.952786941433529,
      "grad_norm": 5.8458170890808105,
      "learning_rate": 6.68544918858756e-05,
      "loss": 2.771864318847656,
      "memory(GiB)": 77.56,
      "step": 45580,
      "token_acc": 0.4192546583850932,
      "train_speed(iter/s)": 1.440036
    },
    {
      "epoch": 1.9530011567627779,
      "grad_norm": 6.395783424377441,
      "learning_rate": 6.684815583706902e-05,
      "loss": 2.36529598236084,
      "memory(GiB)": 77.56,
      "step": 45585,
      "token_acc": 0.5182186234817814,
      "train_speed(iter/s)": 1.439999
    },
    {
      "epoch": 1.953215372092027,
      "grad_norm": 5.270660400390625,
      "learning_rate": 6.684181948304228e-05,
      "loss": 2.234212303161621,
      "memory(GiB)": 77.56,
      "step": 45590,
      "token_acc": 0.5035211267605634,
      "train_speed(iter/s)": 1.44003
    },
    {
      "epoch": 1.953429587421276,
      "grad_norm": 4.634381294250488,
      "learning_rate": 6.683548282391019e-05,
      "loss": 2.3774932861328124,
      "memory(GiB)": 77.56,
      "step": 45595,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.440023
    },
    {
      "epoch": 1.9536438027505247,
      "grad_norm": 5.826134204864502,
      "learning_rate": 6.682914585978754e-05,
      "loss": 2.585467529296875,
      "memory(GiB)": 77.56,
      "step": 45600,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.440062
    },
    {
      "epoch": 1.9538580180797738,
      "grad_norm": 3.282994508743286,
      "learning_rate": 6.682280859078911e-05,
      "loss": 2.6191328048706053,
      "memory(GiB)": 77.56,
      "step": 45605,
      "token_acc": 0.5015384615384615,
      "train_speed(iter/s)": 1.440039
    },
    {
      "epoch": 1.9540722334090228,
      "grad_norm": 5.450883865356445,
      "learning_rate": 6.681647101702973e-05,
      "loss": 2.727987861633301,
      "memory(GiB)": 77.56,
      "step": 45610,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.440079
    },
    {
      "epoch": 1.9542864487382716,
      "grad_norm": 4.7945146560668945,
      "learning_rate": 6.681013313862421e-05,
      "loss": 2.798582077026367,
      "memory(GiB)": 77.56,
      "step": 45615,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.44001
    },
    {
      "epoch": 1.9545006640675207,
      "grad_norm": 3.914283275604248,
      "learning_rate": 6.680379495568735e-05,
      "loss": 2.5355690002441404,
      "memory(GiB)": 77.56,
      "step": 45620,
      "token_acc": 0.4542483660130719,
      "train_speed(iter/s)": 1.440041
    },
    {
      "epoch": 1.9547148793967697,
      "grad_norm": 4.840810298919678,
      "learning_rate": 6.679745646833397e-05,
      "loss": 2.332847023010254,
      "memory(GiB)": 77.56,
      "step": 45625,
      "token_acc": 0.49836065573770494,
      "train_speed(iter/s)": 1.440015
    },
    {
      "epoch": 1.9549290947260185,
      "grad_norm": 6.350103378295898,
      "learning_rate": 6.679111767667894e-05,
      "loss": 2.43646240234375,
      "memory(GiB)": 77.56,
      "step": 45630,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.440006
    },
    {
      "epoch": 1.9551433100552675,
      "grad_norm": 5.065173625946045,
      "learning_rate": 6.678477858083703e-05,
      "loss": 2.835140037536621,
      "memory(GiB)": 77.56,
      "step": 45635,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.440022
    },
    {
      "epoch": 1.9553575253845166,
      "grad_norm": 5.095966815948486,
      "learning_rate": 6.677843918092312e-05,
      "loss": 2.516127014160156,
      "memory(GiB)": 77.56,
      "step": 45640,
      "token_acc": 0.486013986013986,
      "train_speed(iter/s)": 1.440008
    },
    {
      "epoch": 1.9555717407137654,
      "grad_norm": 4.866915702819824,
      "learning_rate": 6.677209947705205e-05,
      "loss": 2.562136650085449,
      "memory(GiB)": 77.56,
      "step": 45645,
      "token_acc": 0.43416370106761565,
      "train_speed(iter/s)": 1.440012
    },
    {
      "epoch": 1.9557859560430144,
      "grad_norm": 4.1509175300598145,
      "learning_rate": 6.676575946933867e-05,
      "loss": 2.1394287109375,
      "memory(GiB)": 77.56,
      "step": 45650,
      "token_acc": 0.5096774193548387,
      "train_speed(iter/s)": 1.440026
    },
    {
      "epoch": 1.9560001713722635,
      "grad_norm": 4.327695369720459,
      "learning_rate": 6.675941915789783e-05,
      "loss": 2.4458009719848635,
      "memory(GiB)": 77.56,
      "step": 45655,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.440004
    },
    {
      "epoch": 1.9562143867015123,
      "grad_norm": 6.655233860015869,
      "learning_rate": 6.675307854284438e-05,
      "loss": 2.729585647583008,
      "memory(GiB)": 77.56,
      "step": 45660,
      "token_acc": 0.4067796610169492,
      "train_speed(iter/s)": 1.440007
    },
    {
      "epoch": 1.9564286020307613,
      "grad_norm": 4.3718767166137695,
      "learning_rate": 6.67467376242932e-05,
      "loss": 2.3149494171142577,
      "memory(GiB)": 77.56,
      "step": 45665,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.440002
    },
    {
      "epoch": 1.9566428173600103,
      "grad_norm": 6.582820892333984,
      "learning_rate": 6.674039640235917e-05,
      "loss": 2.5831912994384765,
      "memory(GiB)": 77.56,
      "step": 45670,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.439984
    },
    {
      "epoch": 1.9568570326892591,
      "grad_norm": 4.774419784545898,
      "learning_rate": 6.673405487715713e-05,
      "loss": 2.264492988586426,
      "memory(GiB)": 77.56,
      "step": 45675,
      "token_acc": 0.5210084033613446,
      "train_speed(iter/s)": 1.439973
    },
    {
      "epoch": 1.9570712480185082,
      "grad_norm": 5.1712646484375,
      "learning_rate": 6.6727713048802e-05,
      "loss": 2.4001861572265626,
      "memory(GiB)": 77.56,
      "step": 45680,
      "token_acc": 0.5207547169811321,
      "train_speed(iter/s)": 1.439989
    },
    {
      "epoch": 1.9572854633477572,
      "grad_norm": 6.647982120513916,
      "learning_rate": 6.672137091740867e-05,
      "loss": 2.4587112426757813,
      "memory(GiB)": 77.56,
      "step": 45685,
      "token_acc": 0.45255474452554745,
      "train_speed(iter/s)": 1.44003
    },
    {
      "epoch": 1.957499678677006,
      "grad_norm": 5.97927713394165,
      "learning_rate": 6.671502848309198e-05,
      "loss": 2.41509952545166,
      "memory(GiB)": 77.56,
      "step": 45690,
      "token_acc": 0.4980392156862745,
      "train_speed(iter/s)": 1.440066
    },
    {
      "epoch": 1.957713894006255,
      "grad_norm": 4.979121685028076,
      "learning_rate": 6.670868574596691e-05,
      "loss": 2.54388313293457,
      "memory(GiB)": 77.56,
      "step": 45695,
      "token_acc": 0.44140625,
      "train_speed(iter/s)": 1.440047
    },
    {
      "epoch": 1.957928109335504,
      "grad_norm": 4.9567365646362305,
      "learning_rate": 6.670234270614832e-05,
      "loss": 2.8766857147216798,
      "memory(GiB)": 77.56,
      "step": 45700,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.440045
    },
    {
      "epoch": 1.958142324664753,
      "grad_norm": 4.630826950073242,
      "learning_rate": 6.66959993637511e-05,
      "loss": 2.474800682067871,
      "memory(GiB)": 77.56,
      "step": 45705,
      "token_acc": 0.5288461538461539,
      "train_speed(iter/s)": 1.440044
    },
    {
      "epoch": 1.958356539994002,
      "grad_norm": 4.426127910614014,
      "learning_rate": 6.668965571889021e-05,
      "loss": 2.7333507537841797,
      "memory(GiB)": 77.56,
      "step": 45710,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.440064
    },
    {
      "epoch": 1.958570755323251,
      "grad_norm": 5.418478965759277,
      "learning_rate": 6.668331177168055e-05,
      "loss": 2.374101257324219,
      "memory(GiB)": 77.56,
      "step": 45715,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.44007
    },
    {
      "epoch": 1.9587849706524998,
      "grad_norm": 5.040560245513916,
      "learning_rate": 6.667696752223703e-05,
      "loss": 2.7069873809814453,
      "memory(GiB)": 77.56,
      "step": 45720,
      "token_acc": 0.44981412639405205,
      "train_speed(iter/s)": 1.440092
    },
    {
      "epoch": 1.9589991859817488,
      "grad_norm": 5.640760898590088,
      "learning_rate": 6.667062297067462e-05,
      "loss": 2.40421142578125,
      "memory(GiB)": 77.56,
      "step": 45725,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.440105
    },
    {
      "epoch": 1.9592134013109979,
      "grad_norm": 3.948404312133789,
      "learning_rate": 6.666427811710824e-05,
      "loss": 2.652470016479492,
      "memory(GiB)": 77.56,
      "step": 45730,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.440129
    },
    {
      "epoch": 1.9594276166402467,
      "grad_norm": 6.179126739501953,
      "learning_rate": 6.66579329616528e-05,
      "loss": 2.852280616760254,
      "memory(GiB)": 77.56,
      "step": 45735,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.44018
    },
    {
      "epoch": 1.9596418319694957,
      "grad_norm": 4.214303016662598,
      "learning_rate": 6.66515875044233e-05,
      "loss": 2.3850189208984376,
      "memory(GiB)": 77.56,
      "step": 45740,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.440183
    },
    {
      "epoch": 1.9598560472987447,
      "grad_norm": 4.51772928237915,
      "learning_rate": 6.664524174553467e-05,
      "loss": 2.7407501220703123,
      "memory(GiB)": 77.56,
      "step": 45745,
      "token_acc": 0.4570552147239264,
      "train_speed(iter/s)": 1.44019
    },
    {
      "epoch": 1.9600702626279936,
      "grad_norm": 5.619811058044434,
      "learning_rate": 6.663889568510186e-05,
      "loss": 2.686910057067871,
      "memory(GiB)": 77.56,
      "step": 45750,
      "token_acc": 0.41694915254237286,
      "train_speed(iter/s)": 1.440207
    },
    {
      "epoch": 1.9602844779572426,
      "grad_norm": 3.596792221069336,
      "learning_rate": 6.663254932323986e-05,
      "loss": 2.729270362854004,
      "memory(GiB)": 77.56,
      "step": 45755,
      "token_acc": 0.43217665615141954,
      "train_speed(iter/s)": 1.440221
    },
    {
      "epoch": 1.9604986932864916,
      "grad_norm": 5.189215183258057,
      "learning_rate": 6.662620266006361e-05,
      "loss": 2.6603809356689454,
      "memory(GiB)": 77.56,
      "step": 45760,
      "token_acc": 0.4701086956521739,
      "train_speed(iter/s)": 1.440246
    },
    {
      "epoch": 1.9607129086157404,
      "grad_norm": 5.653921604156494,
      "learning_rate": 6.661985569568811e-05,
      "loss": 2.4463336944580076,
      "memory(GiB)": 77.56,
      "step": 45765,
      "token_acc": 0.5136986301369864,
      "train_speed(iter/s)": 1.440266
    },
    {
      "epoch": 1.9609271239449895,
      "grad_norm": 5.984095096588135,
      "learning_rate": 6.661350843022833e-05,
      "loss": 2.3911781311035156,
      "memory(GiB)": 77.56,
      "step": 45770,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.440303
    },
    {
      "epoch": 1.9611413392742385,
      "grad_norm": 5.861817836761475,
      "learning_rate": 6.660716086379927e-05,
      "loss": 2.728754425048828,
      "memory(GiB)": 77.56,
      "step": 45775,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.440315
    },
    {
      "epoch": 1.9613555546034873,
      "grad_norm": 5.163695335388184,
      "learning_rate": 6.660081299651591e-05,
      "loss": 2.8481945037841796,
      "memory(GiB)": 77.56,
      "step": 45780,
      "token_acc": 0.46273291925465837,
      "train_speed(iter/s)": 1.440364
    },
    {
      "epoch": 1.9615697699327364,
      "grad_norm": 5.630147933959961,
      "learning_rate": 6.659446482849325e-05,
      "loss": 2.5242502212524416,
      "memory(GiB)": 77.56,
      "step": 45785,
      "token_acc": 0.460431654676259,
      "train_speed(iter/s)": 1.440358
    },
    {
      "epoch": 1.9617839852619854,
      "grad_norm": 4.404461860656738,
      "learning_rate": 6.658811635984629e-05,
      "loss": 2.657027816772461,
      "memory(GiB)": 77.56,
      "step": 45790,
      "token_acc": 0.4664179104477612,
      "train_speed(iter/s)": 1.440401
    },
    {
      "epoch": 1.9619982005912342,
      "grad_norm": 4.369134426116943,
      "learning_rate": 6.658176759069004e-05,
      "loss": 2.7699539184570314,
      "memory(GiB)": 77.56,
      "step": 45795,
      "token_acc": 0.48360655737704916,
      "train_speed(iter/s)": 1.44043
    },
    {
      "epoch": 1.9622124159204832,
      "grad_norm": 4.9915924072265625,
      "learning_rate": 6.657541852113953e-05,
      "loss": 2.6560897827148438,
      "memory(GiB)": 77.56,
      "step": 45800,
      "token_acc": 0.423841059602649,
      "train_speed(iter/s)": 1.440389
    },
    {
      "epoch": 1.9624266312497323,
      "grad_norm": 8.534985542297363,
      "learning_rate": 6.656906915130975e-05,
      "loss": 2.514673614501953,
      "memory(GiB)": 77.56,
      "step": 45805,
      "token_acc": 0.46638655462184875,
      "train_speed(iter/s)": 1.440415
    },
    {
      "epoch": 1.962640846578981,
      "grad_norm": 6.863739013671875,
      "learning_rate": 6.656271948131573e-05,
      "loss": 2.553104019165039,
      "memory(GiB)": 77.56,
      "step": 45810,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.440385
    },
    {
      "epoch": 1.9628550619082301,
      "grad_norm": 6.4225897789001465,
      "learning_rate": 6.655636951127254e-05,
      "loss": 2.747235107421875,
      "memory(GiB)": 77.56,
      "step": 45815,
      "token_acc": 0.4342105263157895,
      "train_speed(iter/s)": 1.440399
    },
    {
      "epoch": 1.9630692772374791,
      "grad_norm": 5.554860591888428,
      "learning_rate": 6.655001924129515e-05,
      "loss": 2.8411582946777343,
      "memory(GiB)": 77.56,
      "step": 45820,
      "token_acc": 0.4261744966442953,
      "train_speed(iter/s)": 1.440388
    },
    {
      "epoch": 1.963283492566728,
      "grad_norm": 5.233529567718506,
      "learning_rate": 6.654366867149865e-05,
      "loss": 2.360663414001465,
      "memory(GiB)": 77.56,
      "step": 45825,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.44039
    },
    {
      "epoch": 1.963497707895977,
      "grad_norm": 4.994997024536133,
      "learning_rate": 6.653731780199807e-05,
      "loss": 2.7178909301757814,
      "memory(GiB)": 77.56,
      "step": 45830,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.440354
    },
    {
      "epoch": 1.963711923225226,
      "grad_norm": 4.927798271179199,
      "learning_rate": 6.653096663290847e-05,
      "loss": 2.3548952102661134,
      "memory(GiB)": 77.56,
      "step": 45835,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.440372
    },
    {
      "epoch": 1.9639261385544748,
      "grad_norm": 5.804313659667969,
      "learning_rate": 6.652461516434492e-05,
      "loss": 2.7635629653930662,
      "memory(GiB)": 77.56,
      "step": 45840,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.440402
    },
    {
      "epoch": 1.9641403538837239,
      "grad_norm": 5.223865032196045,
      "learning_rate": 6.651826339642243e-05,
      "loss": 2.681183433532715,
      "memory(GiB)": 77.56,
      "step": 45845,
      "token_acc": 0.4316546762589928,
      "train_speed(iter/s)": 1.440386
    },
    {
      "epoch": 1.964354569212973,
      "grad_norm": 5.4697794914245605,
      "learning_rate": 6.651191132925612e-05,
      "loss": 2.288090133666992,
      "memory(GiB)": 77.56,
      "step": 45850,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.440419
    },
    {
      "epoch": 1.9645687845422217,
      "grad_norm": 4.830199718475342,
      "learning_rate": 6.650555896296104e-05,
      "loss": 2.339397430419922,
      "memory(GiB)": 77.56,
      "step": 45855,
      "token_acc": 0.5042016806722689,
      "train_speed(iter/s)": 1.440422
    },
    {
      "epoch": 1.9647829998714708,
      "grad_norm": 5.951051712036133,
      "learning_rate": 6.649920629765229e-05,
      "loss": 2.4759355545043946,
      "memory(GiB)": 77.56,
      "step": 45860,
      "token_acc": 0.45925925925925926,
      "train_speed(iter/s)": 1.440462
    },
    {
      "epoch": 1.9649972152007198,
      "grad_norm": 5.6724677085876465,
      "learning_rate": 6.649285333344492e-05,
      "loss": 2.5050195693969726,
      "memory(GiB)": 77.56,
      "step": 45865,
      "token_acc": 0.46963562753036436,
      "train_speed(iter/s)": 1.440456
    },
    {
      "epoch": 1.9652114305299686,
      "grad_norm": 5.25377893447876,
      "learning_rate": 6.648650007045407e-05,
      "loss": 2.4837507247924804,
      "memory(GiB)": 77.56,
      "step": 45870,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.440458
    },
    {
      "epoch": 1.9654256458592176,
      "grad_norm": 5.423166751861572,
      "learning_rate": 6.648014650879479e-05,
      "loss": 2.893759346008301,
      "memory(GiB)": 77.56,
      "step": 45875,
      "token_acc": 0.4607142857142857,
      "train_speed(iter/s)": 1.440504
    },
    {
      "epoch": 1.9656398611884667,
      "grad_norm": 4.335208892822266,
      "learning_rate": 6.64737926485822e-05,
      "loss": 2.390877532958984,
      "memory(GiB)": 77.56,
      "step": 45880,
      "token_acc": 0.527972027972028,
      "train_speed(iter/s)": 1.440481
    },
    {
      "epoch": 1.9658540765177155,
      "grad_norm": 4.426405429840088,
      "learning_rate": 6.646743848993139e-05,
      "loss": 2.643740653991699,
      "memory(GiB)": 77.56,
      "step": 45885,
      "token_acc": 0.4417910447761194,
      "train_speed(iter/s)": 1.440457
    },
    {
      "epoch": 1.9660682918469645,
      "grad_norm": 5.99837589263916,
      "learning_rate": 6.64610840329575e-05,
      "loss": 2.7230987548828125,
      "memory(GiB)": 77.56,
      "step": 45890,
      "token_acc": 0.4153846153846154,
      "train_speed(iter/s)": 1.440498
    },
    {
      "epoch": 1.9662825071762136,
      "grad_norm": 6.710351943969727,
      "learning_rate": 6.64547292777756e-05,
      "loss": 2.7149799346923826,
      "memory(GiB)": 77.56,
      "step": 45895,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.440515
    },
    {
      "epoch": 1.9664967225054624,
      "grad_norm": 5.052811145782471,
      "learning_rate": 6.644837422450087e-05,
      "loss": 2.3122949600219727,
      "memory(GiB)": 77.56,
      "step": 45900,
      "token_acc": 0.48120300751879697,
      "train_speed(iter/s)": 1.440496
    },
    {
      "epoch": 1.9667109378347114,
      "grad_norm": 6.387402057647705,
      "learning_rate": 6.644201887324841e-05,
      "loss": 2.6619235992431642,
      "memory(GiB)": 77.56,
      "step": 45905,
      "token_acc": 0.423728813559322,
      "train_speed(iter/s)": 1.440487
    },
    {
      "epoch": 1.9669251531639604,
      "grad_norm": 6.813206672668457,
      "learning_rate": 6.643566322413336e-05,
      "loss": 2.4210391998291017,
      "memory(GiB)": 77.56,
      "step": 45910,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.440483
    },
    {
      "epoch": 1.9671393684932093,
      "grad_norm": 6.530525207519531,
      "learning_rate": 6.642930727727085e-05,
      "loss": 2.380424690246582,
      "memory(GiB)": 77.56,
      "step": 45915,
      "token_acc": 0.4796747967479675,
      "train_speed(iter/s)": 1.440503
    },
    {
      "epoch": 1.9673535838224583,
      "grad_norm": 4.187199592590332,
      "learning_rate": 6.642295103277602e-05,
      "loss": 2.4211254119873047,
      "memory(GiB)": 77.56,
      "step": 45920,
      "token_acc": 0.4774011299435028,
      "train_speed(iter/s)": 1.440496
    },
    {
      "epoch": 1.9675677991517073,
      "grad_norm": 4.4053955078125,
      "learning_rate": 6.641659449076401e-05,
      "loss": 2.7990097045898437,
      "memory(GiB)": 77.56,
      "step": 45925,
      "token_acc": 0.4608433734939759,
      "train_speed(iter/s)": 1.440497
    },
    {
      "epoch": 1.9677820144809561,
      "grad_norm": 4.436799049377441,
      "learning_rate": 6.641023765135001e-05,
      "loss": 2.544133758544922,
      "memory(GiB)": 77.56,
      "step": 45930,
      "token_acc": 0.46607669616519176,
      "train_speed(iter/s)": 1.440473
    },
    {
      "epoch": 1.9679962298102052,
      "grad_norm": 4.896063327789307,
      "learning_rate": 6.640388051464917e-05,
      "loss": 2.408317756652832,
      "memory(GiB)": 77.56,
      "step": 45935,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.440482
    },
    {
      "epoch": 1.9682104451394542,
      "grad_norm": 4.4914469718933105,
      "learning_rate": 6.639752308077662e-05,
      "loss": 2.615721321105957,
      "memory(GiB)": 77.56,
      "step": 45940,
      "token_acc": 0.47560975609756095,
      "train_speed(iter/s)": 1.440503
    },
    {
      "epoch": 1.968424660468703,
      "grad_norm": 6.709591865539551,
      "learning_rate": 6.639116534984758e-05,
      "loss": 2.486857604980469,
      "memory(GiB)": 77.56,
      "step": 45945,
      "token_acc": 0.4632352941176471,
      "train_speed(iter/s)": 1.440523
    },
    {
      "epoch": 1.968638875797952,
      "grad_norm": 4.1377973556518555,
      "learning_rate": 6.63848073219772e-05,
      "loss": 2.4100162506103517,
      "memory(GiB)": 77.56,
      "step": 45950,
      "token_acc": 0.43859649122807015,
      "train_speed(iter/s)": 1.440494
    },
    {
      "epoch": 1.968853091127201,
      "grad_norm": 4.820364952087402,
      "learning_rate": 6.637844899728065e-05,
      "loss": 2.901133728027344,
      "memory(GiB)": 77.56,
      "step": 45955,
      "token_acc": 0.44280442804428044,
      "train_speed(iter/s)": 1.440497
    },
    {
      "epoch": 1.96906730645645,
      "grad_norm": 4.0486907958984375,
      "learning_rate": 6.637209037587315e-05,
      "loss": 2.6329875946044923,
      "memory(GiB)": 77.56,
      "step": 45960,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.440504
    },
    {
      "epoch": 1.969281521785699,
      "grad_norm": 5.785236835479736,
      "learning_rate": 6.636573145786987e-05,
      "loss": 2.518662452697754,
      "memory(GiB)": 77.56,
      "step": 45965,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.440511
    },
    {
      "epoch": 1.969495737114948,
      "grad_norm": 5.312324047088623,
      "learning_rate": 6.635937224338601e-05,
      "loss": 2.66689453125,
      "memory(GiB)": 77.56,
      "step": 45970,
      "token_acc": 0.43234323432343236,
      "train_speed(iter/s)": 1.440531
    },
    {
      "epoch": 1.9697099524441968,
      "grad_norm": 5.460351467132568,
      "learning_rate": 6.635301273253678e-05,
      "loss": 2.398164749145508,
      "memory(GiB)": 77.56,
      "step": 45975,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.969924167773446,
      "grad_norm": 9.177876472473145,
      "learning_rate": 6.634665292543737e-05,
      "loss": 2.5278142929077148,
      "memory(GiB)": 77.56,
      "step": 45980,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.4405
    },
    {
      "epoch": 1.9701383831026948,
      "grad_norm": 4.011397361755371,
      "learning_rate": 6.634029282220303e-05,
      "loss": 2.349286651611328,
      "memory(GiB)": 77.56,
      "step": 45985,
      "token_acc": 0.5300353356890459,
      "train_speed(iter/s)": 1.440511
    },
    {
      "epoch": 1.9703525984319437,
      "grad_norm": 4.624520778656006,
      "learning_rate": 6.633393242294894e-05,
      "loss": 2.2925954818725587,
      "memory(GiB)": 77.56,
      "step": 45990,
      "token_acc": 0.5050505050505051,
      "train_speed(iter/s)": 1.440517
    },
    {
      "epoch": 1.970566813761193,
      "grad_norm": 6.456453800201416,
      "learning_rate": 6.632757172779034e-05,
      "loss": 2.5786991119384766,
      "memory(GiB)": 77.56,
      "step": 45995,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.440515
    },
    {
      "epoch": 1.9707810290904417,
      "grad_norm": 4.135759353637695,
      "learning_rate": 6.632121073684246e-05,
      "loss": 2.1136852264404298,
      "memory(GiB)": 77.56,
      "step": 46000,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.440504
    },
    {
      "epoch": 1.9707810290904417,
      "eval_loss": 2.287435531616211,
      "eval_runtime": 15.0359,
      "eval_samples_per_second": 6.651,
      "eval_steps_per_second": 6.651,
      "eval_token_acc": 0.4641509433962264,
      "step": 46000
    },
    {
      "epoch": 1.9709952444196905,
      "grad_norm": 5.086953639984131,
      "learning_rate": 6.631484945022055e-05,
      "loss": 2.330047035217285,
      "memory(GiB)": 77.56,
      "step": 46005,
      "token_acc": 0.4670433145009416,
      "train_speed(iter/s)": 1.439799
    },
    {
      "epoch": 1.9712094597489398,
      "grad_norm": 5.451361656188965,
      "learning_rate": 6.630848786803983e-05,
      "loss": 2.3972442626953123,
      "memory(GiB)": 77.56,
      "step": 46010,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.439792
    },
    {
      "epoch": 1.9714236750781886,
      "grad_norm": 4.9217987060546875,
      "learning_rate": 6.630212599041556e-05,
      "loss": 2.8109764099121093,
      "memory(GiB)": 77.56,
      "step": 46015,
      "token_acc": 0.4186746987951807,
      "train_speed(iter/s)": 1.439835
    },
    {
      "epoch": 1.9716378904074374,
      "grad_norm": 7.312126636505127,
      "learning_rate": 6.629576381746299e-05,
      "loss": 2.473282051086426,
      "memory(GiB)": 77.56,
      "step": 46020,
      "token_acc": 0.43636363636363634,
      "train_speed(iter/s)": 1.439861
    },
    {
      "epoch": 1.9718521057366867,
      "grad_norm": 4.677431583404541,
      "learning_rate": 6.628940134929736e-05,
      "loss": 2.624419403076172,
      "memory(GiB)": 77.56,
      "step": 46025,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.439862
    },
    {
      "epoch": 1.9720663210659355,
      "grad_norm": 4.504249572753906,
      "learning_rate": 6.628303858603395e-05,
      "loss": 2.371323013305664,
      "memory(GiB)": 77.56,
      "step": 46030,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.439853
    },
    {
      "epoch": 1.9722805363951843,
      "grad_norm": 6.488310813903809,
      "learning_rate": 6.627667552778802e-05,
      "loss": 2.247083282470703,
      "memory(GiB)": 77.56,
      "step": 46035,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.439868
    },
    {
      "epoch": 1.9724947517244336,
      "grad_norm": 5.536812782287598,
      "learning_rate": 6.627031217467484e-05,
      "loss": 2.5022125244140625,
      "memory(GiB)": 77.56,
      "step": 46040,
      "token_acc": 0.4641509433962264,
      "train_speed(iter/s)": 1.439881
    },
    {
      "epoch": 1.9727089670536824,
      "grad_norm": 4.00999116897583,
      "learning_rate": 6.626394852680968e-05,
      "loss": 2.453029441833496,
      "memory(GiB)": 77.56,
      "step": 46045,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.43988
    },
    {
      "epoch": 1.9729231823829312,
      "grad_norm": 6.29418420791626,
      "learning_rate": 6.625758458430785e-05,
      "loss": 2.559286880493164,
      "memory(GiB)": 77.56,
      "step": 46050,
      "token_acc": 0.486013986013986,
      "train_speed(iter/s)": 1.439873
    },
    {
      "epoch": 1.9731373977121804,
      "grad_norm": 5.518522262573242,
      "learning_rate": 6.625122034728463e-05,
      "loss": 2.4911861419677734,
      "memory(GiB)": 77.56,
      "step": 46055,
      "token_acc": 0.4746376811594203,
      "train_speed(iter/s)": 1.439881
    },
    {
      "epoch": 1.9733516130414293,
      "grad_norm": 4.570826530456543,
      "learning_rate": 6.624485581585531e-05,
      "loss": 2.5234804153442383,
      "memory(GiB)": 77.56,
      "step": 46060,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.439864
    },
    {
      "epoch": 1.973565828370678,
      "grad_norm": 4.213891506195068,
      "learning_rate": 6.623849099013517e-05,
      "loss": 2.6506858825683595,
      "memory(GiB)": 77.56,
      "step": 46065,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.439861
    },
    {
      "epoch": 1.9737800436999273,
      "grad_norm": 4.141619682312012,
      "learning_rate": 6.623212587023955e-05,
      "loss": 2.401409149169922,
      "memory(GiB)": 77.56,
      "step": 46070,
      "token_acc": 0.5220588235294118,
      "train_speed(iter/s)": 1.439875
    },
    {
      "epoch": 1.9739942590291761,
      "grad_norm": 4.475203990936279,
      "learning_rate": 6.622576045628374e-05,
      "loss": 2.4911773681640623,
      "memory(GiB)": 77.56,
      "step": 46075,
      "token_acc": 0.44280442804428044,
      "train_speed(iter/s)": 1.439902
    },
    {
      "epoch": 1.974208474358425,
      "grad_norm": 5.4392523765563965,
      "learning_rate": 6.621939474838304e-05,
      "loss": 2.2502946853637695,
      "memory(GiB)": 77.56,
      "step": 46080,
      "token_acc": 0.5181159420289855,
      "train_speed(iter/s)": 1.439919
    },
    {
      "epoch": 1.9744226896876742,
      "grad_norm": 5.93950891494751,
      "learning_rate": 6.621302874665282e-05,
      "loss": 2.4226354598999023,
      "memory(GiB)": 77.56,
      "step": 46085,
      "token_acc": 0.4921135646687697,
      "train_speed(iter/s)": 1.439942
    },
    {
      "epoch": 1.974636905016923,
      "grad_norm": 4.538694858551025,
      "learning_rate": 6.620666245120836e-05,
      "loss": 2.8146881103515624,
      "memory(GiB)": 77.56,
      "step": 46090,
      "token_acc": 0.44360902255639095,
      "train_speed(iter/s)": 1.439967
    },
    {
      "epoch": 1.9748511203461718,
      "grad_norm": 4.988295078277588,
      "learning_rate": 6.620029586216499e-05,
      "loss": 2.610568618774414,
      "memory(GiB)": 77.56,
      "step": 46095,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.439988
    },
    {
      "epoch": 1.975065335675421,
      "grad_norm": 6.159144401550293,
      "learning_rate": 6.61939289796381e-05,
      "loss": 2.6839128494262696,
      "memory(GiB)": 77.56,
      "step": 46100,
      "token_acc": 0.4308176100628931,
      "train_speed(iter/s)": 1.440016
    },
    {
      "epoch": 1.97527955100467,
      "grad_norm": 4.465036392211914,
      "learning_rate": 6.618756180374295e-05,
      "loss": 2.7416873931884767,
      "memory(GiB)": 77.56,
      "step": 46105,
      "token_acc": 0.4560260586319218,
      "train_speed(iter/s)": 1.440026
    },
    {
      "epoch": 1.9754937663339187,
      "grad_norm": 5.041161060333252,
      "learning_rate": 6.618119433459496e-05,
      "loss": 2.0445932388305663,
      "memory(GiB)": 77.56,
      "step": 46110,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.440032
    },
    {
      "epoch": 1.975707981663168,
      "grad_norm": 5.3165507316589355,
      "learning_rate": 6.617482657230945e-05,
      "loss": 2.613577461242676,
      "memory(GiB)": 77.56,
      "step": 46115,
      "token_acc": 0.4276094276094276,
      "train_speed(iter/s)": 1.44003
    },
    {
      "epoch": 1.9759221969924168,
      "grad_norm": 5.273176670074463,
      "learning_rate": 6.61684585170018e-05,
      "loss": 2.156299591064453,
      "memory(GiB)": 77.56,
      "step": 46120,
      "token_acc": 0.5330739299610895,
      "train_speed(iter/s)": 1.440046
    },
    {
      "epoch": 1.9761364123216656,
      "grad_norm": 4.864031791687012,
      "learning_rate": 6.616209016878733e-05,
      "loss": 2.547792434692383,
      "memory(GiB)": 77.56,
      "step": 46125,
      "token_acc": 0.46254071661237783,
      "train_speed(iter/s)": 1.440065
    },
    {
      "epoch": 1.9763506276509148,
      "grad_norm": 5.276404857635498,
      "learning_rate": 6.615572152778147e-05,
      "loss": 2.778274154663086,
      "memory(GiB)": 77.56,
      "step": 46130,
      "token_acc": 0.4494047619047619,
      "train_speed(iter/s)": 1.440042
    },
    {
      "epoch": 1.9765648429801637,
      "grad_norm": 4.446139335632324,
      "learning_rate": 6.614935259409954e-05,
      "loss": 2.7039743423461915,
      "memory(GiB)": 77.56,
      "step": 46135,
      "token_acc": 0.4421052631578947,
      "train_speed(iter/s)": 1.440036
    },
    {
      "epoch": 1.9767790583094125,
      "grad_norm": 4.300412654876709,
      "learning_rate": 6.614298336785692e-05,
      "loss": 2.3138809204101562,
      "memory(GiB)": 77.56,
      "step": 46140,
      "token_acc": 0.5193548387096775,
      "train_speed(iter/s)": 1.440062
    },
    {
      "epoch": 1.9769932736386617,
      "grad_norm": 6.478123664855957,
      "learning_rate": 6.613661384916902e-05,
      "loss": 2.727131652832031,
      "memory(GiB)": 77.56,
      "step": 46145,
      "token_acc": 0.4520547945205479,
      "train_speed(iter/s)": 1.440065
    },
    {
      "epoch": 1.9772074889679105,
      "grad_norm": 5.043158531188965,
      "learning_rate": 6.613024403815125e-05,
      "loss": 2.651078224182129,
      "memory(GiB)": 77.56,
      "step": 46150,
      "token_acc": 0.4276315789473684,
      "train_speed(iter/s)": 1.440081
    },
    {
      "epoch": 1.9774217042971594,
      "grad_norm": 5.35750675201416,
      "learning_rate": 6.612387393491894e-05,
      "loss": 2.6064208984375,
      "memory(GiB)": 77.56,
      "step": 46155,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 1.9776359196264086,
      "grad_norm": 5.122040748596191,
      "learning_rate": 6.611750353958757e-05,
      "loss": 2.5336179733276367,
      "memory(GiB)": 77.56,
      "step": 46160,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.440087
    },
    {
      "epoch": 1.9778501349556574,
      "grad_norm": 4.25455379486084,
      "learning_rate": 6.611113285227247e-05,
      "loss": 2.698091506958008,
      "memory(GiB)": 77.56,
      "step": 46165,
      "token_acc": 0.4523076923076923,
      "train_speed(iter/s)": 1.440104
    },
    {
      "epoch": 1.9780643502849062,
      "grad_norm": 5.582869052886963,
      "learning_rate": 6.610476187308909e-05,
      "loss": 2.8125026702880858,
      "memory(GiB)": 77.56,
      "step": 46170,
      "token_acc": 0.43034055727554177,
      "train_speed(iter/s)": 1.440137
    },
    {
      "epoch": 1.9782785656141555,
      "grad_norm": 4.765628814697266,
      "learning_rate": 6.609839060215287e-05,
      "loss": 2.7479183197021486,
      "memory(GiB)": 77.56,
      "step": 46175,
      "token_acc": 0.4307228915662651,
      "train_speed(iter/s)": 1.440108
    },
    {
      "epoch": 1.9784927809434043,
      "grad_norm": 4.945771217346191,
      "learning_rate": 6.609201903957916e-05,
      "loss": 2.6540557861328127,
      "memory(GiB)": 77.56,
      "step": 46180,
      "token_acc": 0.44412607449856734,
      "train_speed(iter/s)": 1.440084
    },
    {
      "epoch": 1.9787069962726531,
      "grad_norm": 5.673392295837402,
      "learning_rate": 6.608564718548344e-05,
      "loss": 2.469636344909668,
      "memory(GiB)": 77.56,
      "step": 46185,
      "token_acc": 0.4881889763779528,
      "train_speed(iter/s)": 1.440104
    },
    {
      "epoch": 1.9789212116019024,
      "grad_norm": 4.77620792388916,
      "learning_rate": 6.607927503998116e-05,
      "loss": 2.4937023162841796,
      "memory(GiB)": 77.56,
      "step": 46190,
      "token_acc": 0.4980237154150198,
      "train_speed(iter/s)": 1.440085
    },
    {
      "epoch": 1.9791354269311512,
      "grad_norm": 4.5593976974487305,
      "learning_rate": 6.607290260318771e-05,
      "loss": 2.2752307891845702,
      "memory(GiB)": 77.56,
      "step": 46195,
      "token_acc": 0.5369649805447471,
      "train_speed(iter/s)": 1.440107
    },
    {
      "epoch": 1.9793496422604,
      "grad_norm": 5.7640533447265625,
      "learning_rate": 6.606652987521855e-05,
      "loss": 3.032878303527832,
      "memory(GiB)": 77.56,
      "step": 46200,
      "token_acc": 0.42748091603053434,
      "train_speed(iter/s)": 1.44013
    },
    {
      "epoch": 1.9795638575896493,
      "grad_norm": 6.39599609375,
      "learning_rate": 6.606015685618913e-05,
      "loss": 2.6303007125854494,
      "memory(GiB)": 77.56,
      "step": 46205,
      "token_acc": 0.4423076923076923,
      "train_speed(iter/s)": 1.440143
    },
    {
      "epoch": 1.979778072918898,
      "grad_norm": 4.649415493011475,
      "learning_rate": 6.60537835462149e-05,
      "loss": 2.34720401763916,
      "memory(GiB)": 77.56,
      "step": 46210,
      "token_acc": 0.5198863636363636,
      "train_speed(iter/s)": 1.440158
    },
    {
      "epoch": 1.9799922882481469,
      "grad_norm": 10.375418663024902,
      "learning_rate": 6.604740994541133e-05,
      "loss": 2.4568550109863283,
      "memory(GiB)": 77.56,
      "step": 46215,
      "token_acc": 0.5275229357798165,
      "train_speed(iter/s)": 1.440164
    },
    {
      "epoch": 1.9802065035773961,
      "grad_norm": 6.978662014007568,
      "learning_rate": 6.604103605389387e-05,
      "loss": 2.3641246795654296,
      "memory(GiB)": 77.56,
      "step": 46220,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.440145
    },
    {
      "epoch": 1.980420718906645,
      "grad_norm": 4.827588081359863,
      "learning_rate": 6.603466187177799e-05,
      "loss": 2.5591302871704102,
      "memory(GiB)": 77.56,
      "step": 46225,
      "token_acc": 0.5137614678899083,
      "train_speed(iter/s)": 1.440178
    },
    {
      "epoch": 1.9806349342358938,
      "grad_norm": 5.898464679718018,
      "learning_rate": 6.602828739917919e-05,
      "loss": 2.604831314086914,
      "memory(GiB)": 77.56,
      "step": 46230,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.440205
    },
    {
      "epoch": 1.980849149565143,
      "grad_norm": 6.191739082336426,
      "learning_rate": 6.602191263621291e-05,
      "loss": 2.2673355102539063,
      "memory(GiB)": 77.56,
      "step": 46235,
      "token_acc": 0.4939759036144578,
      "train_speed(iter/s)": 1.440233
    },
    {
      "epoch": 1.9810633648943918,
      "grad_norm": 5.841427803039551,
      "learning_rate": 6.601553758299465e-05,
      "loss": 2.1395694732666017,
      "memory(GiB)": 77.56,
      "step": 46240,
      "token_acc": 0.5375,
      "train_speed(iter/s)": 1.440255
    },
    {
      "epoch": 1.9812775802236406,
      "grad_norm": 5.003666400909424,
      "learning_rate": 6.60091622396399e-05,
      "loss": 2.442990303039551,
      "memory(GiB)": 77.56,
      "step": 46245,
      "token_acc": 0.4542483660130719,
      "train_speed(iter/s)": 1.44028
    },
    {
      "epoch": 1.98149179555289,
      "grad_norm": 5.166260719299316,
      "learning_rate": 6.600278660626418e-05,
      "loss": 2.660910224914551,
      "memory(GiB)": 77.56,
      "step": 46250,
      "token_acc": 0.434640522875817,
      "train_speed(iter/s)": 1.440311
    },
    {
      "epoch": 1.9817060108821387,
      "grad_norm": 3.8498947620391846,
      "learning_rate": 6.599641068298295e-05,
      "loss": 2.4890729904174806,
      "memory(GiB)": 77.56,
      "step": 46255,
      "token_acc": 0.5055762081784386,
      "train_speed(iter/s)": 1.440317
    },
    {
      "epoch": 1.9819202262113875,
      "grad_norm": 5.755703926086426,
      "learning_rate": 6.599003446991174e-05,
      "loss": 2.2098081588745115,
      "memory(GiB)": 77.56,
      "step": 46260,
      "token_acc": 0.5633802816901409,
      "train_speed(iter/s)": 1.440349
    },
    {
      "epoch": 1.9821344415406368,
      "grad_norm": 4.347459316253662,
      "learning_rate": 6.598365796716606e-05,
      "loss": 2.7876510620117188,
      "memory(GiB)": 77.56,
      "step": 46265,
      "token_acc": 0.46439628482972134,
      "train_speed(iter/s)": 1.440376
    },
    {
      "epoch": 1.9823486568698856,
      "grad_norm": 4.963770866394043,
      "learning_rate": 6.597728117486141e-05,
      "loss": 2.793642044067383,
      "memory(GiB)": 77.56,
      "step": 46270,
      "token_acc": 0.42452830188679247,
      "train_speed(iter/s)": 1.440392
    },
    {
      "epoch": 1.9825628721991344,
      "grad_norm": 4.950620174407959,
      "learning_rate": 6.597090409311334e-05,
      "loss": 2.7416378021240235,
      "memory(GiB)": 77.56,
      "step": 46275,
      "token_acc": 0.46443514644351463,
      "train_speed(iter/s)": 1.440389
    },
    {
      "epoch": 1.9827770875283837,
      "grad_norm": 4.158836364746094,
      "learning_rate": 6.596452672203736e-05,
      "loss": 2.542458915710449,
      "memory(GiB)": 77.56,
      "step": 46280,
      "token_acc": 0.5375375375375375,
      "train_speed(iter/s)": 1.440383
    },
    {
      "epoch": 1.9829913028576325,
      "grad_norm": 4.593012809753418,
      "learning_rate": 6.595814906174898e-05,
      "loss": 2.7064571380615234,
      "memory(GiB)": 77.56,
      "step": 46285,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.440391
    },
    {
      "epoch": 1.9832055181868813,
      "grad_norm": 5.085115909576416,
      "learning_rate": 6.595177111236377e-05,
      "loss": 2.5841949462890623,
      "memory(GiB)": 77.56,
      "step": 46290,
      "token_acc": 0.49795918367346936,
      "train_speed(iter/s)": 1.440381
    },
    {
      "epoch": 1.9834197335161305,
      "grad_norm": 5.26007604598999,
      "learning_rate": 6.594539287399727e-05,
      "loss": 2.5472856521606446,
      "memory(GiB)": 77.56,
      "step": 46295,
      "token_acc": 0.47904191616766467,
      "train_speed(iter/s)": 1.440422
    },
    {
      "epoch": 1.9836339488453794,
      "grad_norm": 3.1587576866149902,
      "learning_rate": 6.593901434676502e-05,
      "loss": 2.3253494262695313,
      "memory(GiB)": 77.56,
      "step": 46300,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.440427
    },
    {
      "epoch": 1.9838481641746282,
      "grad_norm": 5.255431175231934,
      "learning_rate": 6.593263553078257e-05,
      "loss": 2.7006940841674805,
      "memory(GiB)": 77.56,
      "step": 46305,
      "token_acc": 0.4557377049180328,
      "train_speed(iter/s)": 1.440448
    },
    {
      "epoch": 1.9840623795038774,
      "grad_norm": 4.174441337585449,
      "learning_rate": 6.592625642616546e-05,
      "loss": 2.4668142318725588,
      "memory(GiB)": 77.56,
      "step": 46310,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.44044
    },
    {
      "epoch": 1.9842765948331262,
      "grad_norm": 4.879791736602783,
      "learning_rate": 6.591987703302931e-05,
      "loss": 2.3057186126708986,
      "memory(GiB)": 77.56,
      "step": 46315,
      "token_acc": 0.5371900826446281,
      "train_speed(iter/s)": 1.440457
    },
    {
      "epoch": 1.9844908101623753,
      "grad_norm": 5.802064895629883,
      "learning_rate": 6.591349735148964e-05,
      "loss": 2.4005640029907225,
      "memory(GiB)": 77.56,
      "step": 46320,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.440468
    },
    {
      "epoch": 1.9847050254916243,
      "grad_norm": 4.675039291381836,
      "learning_rate": 6.590711738166204e-05,
      "loss": 2.950305938720703,
      "memory(GiB)": 77.56,
      "step": 46325,
      "token_acc": 0.4539877300613497,
      "train_speed(iter/s)": 1.440488
    },
    {
      "epoch": 1.9849192408208731,
      "grad_norm": 5.377086639404297,
      "learning_rate": 6.590073712366207e-05,
      "loss": 2.6158636093139647,
      "memory(GiB)": 77.56,
      "step": 46330,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.440499
    },
    {
      "epoch": 1.9851334561501222,
      "grad_norm": 5.898637294769287,
      "learning_rate": 6.589435657760535e-05,
      "loss": 2.5784420013427733,
      "memory(GiB)": 77.56,
      "step": 46335,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.440519
    },
    {
      "epoch": 1.9853476714793712,
      "grad_norm": 3.5971996784210205,
      "learning_rate": 6.588797574360744e-05,
      "loss": 2.2648946762084963,
      "memory(GiB)": 77.56,
      "step": 46340,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.440515
    },
    {
      "epoch": 1.98556188680862,
      "grad_norm": 5.021591663360596,
      "learning_rate": 6.588159462178396e-05,
      "loss": 2.5986915588378907,
      "memory(GiB)": 77.56,
      "step": 46345,
      "token_acc": 0.4472843450479233,
      "train_speed(iter/s)": 1.44049
    },
    {
      "epoch": 1.985776102137869,
      "grad_norm": 5.158310890197754,
      "learning_rate": 6.587521321225048e-05,
      "loss": 2.4552642822265627,
      "memory(GiB)": 77.56,
      "step": 46350,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.440495
    },
    {
      "epoch": 1.985990317467118,
      "grad_norm": 3.8151841163635254,
      "learning_rate": 6.586883151512263e-05,
      "loss": 2.422525405883789,
      "memory(GiB)": 77.56,
      "step": 46355,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.440463
    },
    {
      "epoch": 1.9862045327963669,
      "grad_norm": 4.906080722808838,
      "learning_rate": 6.586244953051602e-05,
      "loss": 2.466356086730957,
      "memory(GiB)": 77.56,
      "step": 46360,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 1.986418748125616,
      "grad_norm": 5.610723495483398,
      "learning_rate": 6.585606725854623e-05,
      "loss": 2.7259952545166017,
      "memory(GiB)": 77.56,
      "step": 46365,
      "token_acc": 0.4128113879003559,
      "train_speed(iter/s)": 1.440505
    },
    {
      "epoch": 1.986632963454865,
      "grad_norm": 6.600398540496826,
      "learning_rate": 6.58496846993289e-05,
      "loss": 2.4627485275268555,
      "memory(GiB)": 77.56,
      "step": 46370,
      "token_acc": 0.45110410094637227,
      "train_speed(iter/s)": 1.440499
    },
    {
      "epoch": 1.9868471787841138,
      "grad_norm": 4.842108726501465,
      "learning_rate": 6.58433018529797e-05,
      "loss": 2.661446762084961,
      "memory(GiB)": 77.56,
      "step": 46375,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.4405
    },
    {
      "epoch": 1.9870613941133628,
      "grad_norm": 4.814848899841309,
      "learning_rate": 6.583691871961422e-05,
      "loss": 2.904281997680664,
      "memory(GiB)": 77.56,
      "step": 46380,
      "token_acc": 0.44672131147540983,
      "train_speed(iter/s)": 1.440488
    },
    {
      "epoch": 1.9872756094426118,
      "grad_norm": 4.1047282218933105,
      "learning_rate": 6.583053529934807e-05,
      "loss": 2.713669776916504,
      "memory(GiB)": 77.56,
      "step": 46385,
      "token_acc": 0.45733788395904434,
      "train_speed(iter/s)": 1.440481
    },
    {
      "epoch": 1.9874898247718606,
      "grad_norm": 7.023497104644775,
      "learning_rate": 6.582415159229695e-05,
      "loss": 2.583456039428711,
      "memory(GiB)": 77.56,
      "step": 46390,
      "token_acc": 0.4880239520958084,
      "train_speed(iter/s)": 1.440512
    },
    {
      "epoch": 1.9877040401011097,
      "grad_norm": 5.183415412902832,
      "learning_rate": 6.581776759857647e-05,
      "loss": 2.384836769104004,
      "memory(GiB)": 77.56,
      "step": 46395,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.440503
    },
    {
      "epoch": 1.9879182554303587,
      "grad_norm": 4.711422920227051,
      "learning_rate": 6.58113833183023e-05,
      "loss": 2.1625141143798827,
      "memory(GiB)": 77.56,
      "step": 46400,
      "token_acc": 0.5254777070063694,
      "train_speed(iter/s)": 1.440517
    },
    {
      "epoch": 1.9881324707596075,
      "grad_norm": 4.9224138259887695,
      "learning_rate": 6.580499875159008e-05,
      "loss": 2.6314754486083984,
      "memory(GiB)": 77.56,
      "step": 46405,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.440526
    },
    {
      "epoch": 1.9883466860888566,
      "grad_norm": 4.872269630432129,
      "learning_rate": 6.579861389855548e-05,
      "loss": 2.784394454956055,
      "memory(GiB)": 77.56,
      "step": 46410,
      "token_acc": 0.43209876543209874,
      "train_speed(iter/s)": 1.440507
    },
    {
      "epoch": 1.9885609014181056,
      "grad_norm": 5.029921054840088,
      "learning_rate": 6.579222875931417e-05,
      "loss": 2.4003747940063476,
      "memory(GiB)": 77.56,
      "step": 46415,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.440521
    },
    {
      "epoch": 1.9887751167473544,
      "grad_norm": 5.879994869232178,
      "learning_rate": 6.578584333398184e-05,
      "loss": 2.546307373046875,
      "memory(GiB)": 77.56,
      "step": 46420,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.440541
    },
    {
      "epoch": 1.9889893320766034,
      "grad_norm": 4.8974385261535645,
      "learning_rate": 6.577945762267412e-05,
      "loss": 2.4312274932861326,
      "memory(GiB)": 77.56,
      "step": 46425,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.440506
    },
    {
      "epoch": 1.9892035474058525,
      "grad_norm": 5.272725582122803,
      "learning_rate": 6.577307162550675e-05,
      "loss": 2.1904491424560546,
      "memory(GiB)": 77.56,
      "step": 46430,
      "token_acc": 0.5209790209790209,
      "train_speed(iter/s)": 1.440513
    },
    {
      "epoch": 1.9894177627351013,
      "grad_norm": 5.513241767883301,
      "learning_rate": 6.576668534259536e-05,
      "loss": 2.6153263092041015,
      "memory(GiB)": 77.56,
      "step": 46435,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.44049
    },
    {
      "epoch": 1.9896319780643503,
      "grad_norm": 5.99629020690918,
      "learning_rate": 6.57602987740557e-05,
      "loss": 2.6848064422607423,
      "memory(GiB)": 77.56,
      "step": 46440,
      "token_acc": 0.42805755395683454,
      "train_speed(iter/s)": 1.440497
    },
    {
      "epoch": 1.9898461933935994,
      "grad_norm": 5.065510272979736,
      "learning_rate": 6.575391192000342e-05,
      "loss": 2.973910331726074,
      "memory(GiB)": 77.56,
      "step": 46445,
      "token_acc": 0.4341317365269461,
      "train_speed(iter/s)": 1.440501
    },
    {
      "epoch": 1.9900604087228482,
      "grad_norm": 6.09493350982666,
      "learning_rate": 6.574752478055428e-05,
      "loss": 2.3773401260375975,
      "memory(GiB)": 77.56,
      "step": 46450,
      "token_acc": 0.4528985507246377,
      "train_speed(iter/s)": 1.440502
    },
    {
      "epoch": 1.9902746240520972,
      "grad_norm": 4.089011192321777,
      "learning_rate": 6.574113735582393e-05,
      "loss": 2.4532121658325194,
      "memory(GiB)": 77.56,
      "step": 46455,
      "token_acc": 0.4596774193548387,
      "train_speed(iter/s)": 1.440466
    },
    {
      "epoch": 1.9904888393813462,
      "grad_norm": 4.827455043792725,
      "learning_rate": 6.573474964592813e-05,
      "loss": 2.708393859863281,
      "memory(GiB)": 77.56,
      "step": 46460,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.440492
    },
    {
      "epoch": 1.990703054710595,
      "grad_norm": 4.492869853973389,
      "learning_rate": 6.572836165098258e-05,
      "loss": 2.262235641479492,
      "memory(GiB)": 77.56,
      "step": 46465,
      "token_acc": 0.5585585585585585,
      "train_speed(iter/s)": 1.440494
    },
    {
      "epoch": 1.990917270039844,
      "grad_norm": 5.896934509277344,
      "learning_rate": 6.572197337110299e-05,
      "loss": 2.6777055740356444,
      "memory(GiB)": 77.56,
      "step": 46470,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.440486
    },
    {
      "epoch": 1.9911314853690931,
      "grad_norm": 4.937989234924316,
      "learning_rate": 6.57155848064051e-05,
      "loss": 2.379093360900879,
      "memory(GiB)": 77.56,
      "step": 46475,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.440516
    },
    {
      "epoch": 1.991345700698342,
      "grad_norm": 4.8116536140441895,
      "learning_rate": 6.570919595700467e-05,
      "loss": 2.436779022216797,
      "memory(GiB)": 77.56,
      "step": 46480,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.440512
    },
    {
      "epoch": 1.991559916027591,
      "grad_norm": 4.81565523147583,
      "learning_rate": 6.570280682301739e-05,
      "loss": 2.193369674682617,
      "memory(GiB)": 77.56,
      "step": 46485,
      "token_acc": 0.5375939849624061,
      "train_speed(iter/s)": 1.440544
    },
    {
      "epoch": 1.99177413135684,
      "grad_norm": 5.239287853240967,
      "learning_rate": 6.569641740455905e-05,
      "loss": 2.5726573944091795,
      "memory(GiB)": 77.56,
      "step": 46490,
      "token_acc": 0.4576271186440678,
      "train_speed(iter/s)": 1.440547
    },
    {
      "epoch": 1.9919883466860888,
      "grad_norm": 4.939355850219727,
      "learning_rate": 6.569002770174539e-05,
      "loss": 2.665151596069336,
      "memory(GiB)": 77.56,
      "step": 46495,
      "token_acc": 0.47924528301886793,
      "train_speed(iter/s)": 1.440543
    },
    {
      "epoch": 1.9922025620153379,
      "grad_norm": 5.027363300323486,
      "learning_rate": 6.568363771469214e-05,
      "loss": 2.6751869201660154,
      "memory(GiB)": 77.56,
      "step": 46500,
      "token_acc": 0.45121951219512196,
      "train_speed(iter/s)": 1.440558
    },
    {
      "epoch": 1.9922025620153379,
      "eval_loss": 2.4219417572021484,
      "eval_runtime": 14.8868,
      "eval_samples_per_second": 6.717,
      "eval_steps_per_second": 6.717,
      "eval_token_acc": 0.4422843256379101,
      "step": 46500
    },
    {
      "epoch": 1.9924167773445869,
      "grad_norm": 5.112144470214844,
      "learning_rate": 6.567724744351512e-05,
      "loss": 2.4460798263549806,
      "memory(GiB)": 77.56,
      "step": 46505,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.439871
    },
    {
      "epoch": 1.9926309926738357,
      "grad_norm": 6.163054466247559,
      "learning_rate": 6.567085688833001e-05,
      "loss": 2.5283781051635743,
      "memory(GiB)": 77.56,
      "step": 46510,
      "token_acc": 0.4952681388012618,
      "train_speed(iter/s)": 1.439868
    },
    {
      "epoch": 1.9928452080030847,
      "grad_norm": 5.4156060218811035,
      "learning_rate": 6.566446604925264e-05,
      "loss": 2.5319400787353517,
      "memory(GiB)": 77.56,
      "step": 46515,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.439869
    },
    {
      "epoch": 1.9930594233323338,
      "grad_norm": 5.574679374694824,
      "learning_rate": 6.56580749263988e-05,
      "loss": 2.220624542236328,
      "memory(GiB)": 77.56,
      "step": 46520,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439875
    },
    {
      "epoch": 1.9932736386615826,
      "grad_norm": 3.827199935913086,
      "learning_rate": 6.565168351988422e-05,
      "loss": 2.7762617111206054,
      "memory(GiB)": 77.56,
      "step": 46525,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.439861
    },
    {
      "epoch": 1.9934878539908316,
      "grad_norm": 4.552662372589111,
      "learning_rate": 6.564529182982471e-05,
      "loss": 2.3779298782348635,
      "memory(GiB)": 77.56,
      "step": 46530,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.439852
    },
    {
      "epoch": 1.9937020693200806,
      "grad_norm": 6.504910469055176,
      "learning_rate": 6.563889985633608e-05,
      "loss": 2.4095617294311524,
      "memory(GiB)": 77.56,
      "step": 46535,
      "token_acc": 0.48360655737704916,
      "train_speed(iter/s)": 1.439831
    },
    {
      "epoch": 1.9939162846493295,
      "grad_norm": 6.695801734924316,
      "learning_rate": 6.56325075995341e-05,
      "loss": 2.5215282440185547,
      "memory(GiB)": 77.56,
      "step": 46540,
      "token_acc": 0.460431654676259,
      "train_speed(iter/s)": 1.439846
    },
    {
      "epoch": 1.9941304999785785,
      "grad_norm": 5.770375728607178,
      "learning_rate": 6.562611505953457e-05,
      "loss": 2.6792497634887695,
      "memory(GiB)": 77.56,
      "step": 46545,
      "token_acc": 0.44569288389513106,
      "train_speed(iter/s)": 1.439812
    },
    {
      "epoch": 1.9943447153078275,
      "grad_norm": 7.105743408203125,
      "learning_rate": 6.561972223645331e-05,
      "loss": 2.4988286972045897,
      "memory(GiB)": 77.56,
      "step": 46550,
      "token_acc": 0.5143769968051118,
      "train_speed(iter/s)": 1.439842
    },
    {
      "epoch": 1.9945589306370763,
      "grad_norm": 5.650620460510254,
      "learning_rate": 6.561332913040613e-05,
      "loss": 2.7292051315307617,
      "memory(GiB)": 77.56,
      "step": 46555,
      "token_acc": 0.45692883895131087,
      "train_speed(iter/s)": 1.439828
    },
    {
      "epoch": 1.9947731459663254,
      "grad_norm": 7.187072277069092,
      "learning_rate": 6.560693574150885e-05,
      "loss": 2.366291046142578,
      "memory(GiB)": 77.56,
      "step": 46560,
      "token_acc": 0.47876447876447875,
      "train_speed(iter/s)": 1.439812
    },
    {
      "epoch": 1.9949873612955744,
      "grad_norm": 5.516972541809082,
      "learning_rate": 6.560054206987729e-05,
      "loss": 2.5241676330566407,
      "memory(GiB)": 77.56,
      "step": 46565,
      "token_acc": 0.4506578947368421,
      "train_speed(iter/s)": 1.439795
    },
    {
      "epoch": 1.9952015766248232,
      "grad_norm": 3.9918289184570312,
      "learning_rate": 6.559414811562728e-05,
      "loss": 2.5528985977172853,
      "memory(GiB)": 77.56,
      "step": 46570,
      "token_acc": 0.44047619047619047,
      "train_speed(iter/s)": 1.439792
    },
    {
      "epoch": 1.9954157919540723,
      "grad_norm": 8.8845796585083,
      "learning_rate": 6.558775387887463e-05,
      "loss": 2.7382272720336913,
      "memory(GiB)": 77.56,
      "step": 46575,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.439833
    },
    {
      "epoch": 1.9956300072833213,
      "grad_norm": 5.179216384887695,
      "learning_rate": 6.558135935973521e-05,
      "loss": 2.7244104385375976,
      "memory(GiB)": 77.56,
      "step": 46580,
      "token_acc": 0.44015444015444016,
      "train_speed(iter/s)": 1.439867
    },
    {
      "epoch": 1.99584422261257,
      "grad_norm": 5.242931365966797,
      "learning_rate": 6.557496455832485e-05,
      "loss": 2.298200798034668,
      "memory(GiB)": 77.56,
      "step": 46585,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.439901
    },
    {
      "epoch": 1.9960584379418191,
      "grad_norm": 5.2678542137146,
      "learning_rate": 6.556856947475937e-05,
      "loss": 2.557473373413086,
      "memory(GiB)": 77.56,
      "step": 46590,
      "token_acc": 0.42024539877300615,
      "train_speed(iter/s)": 1.439922
    },
    {
      "epoch": 1.9962726532710682,
      "grad_norm": 5.593785285949707,
      "learning_rate": 6.55621741091547e-05,
      "loss": 2.929859733581543,
      "memory(GiB)": 77.56,
      "step": 46595,
      "token_acc": 0.40522875816993464,
      "train_speed(iter/s)": 1.439924
    },
    {
      "epoch": 1.996486868600317,
      "grad_norm": 5.375649929046631,
      "learning_rate": 6.55557784616266e-05,
      "loss": 2.4373357772827147,
      "memory(GiB)": 77.56,
      "step": 46600,
      "token_acc": 0.48942598187311176,
      "train_speed(iter/s)": 1.439965
    },
    {
      "epoch": 1.996701083929566,
      "grad_norm": 3.787369728088379,
      "learning_rate": 6.554938253229102e-05,
      "loss": 2.5138132095336916,
      "memory(GiB)": 77.56,
      "step": 46605,
      "token_acc": 0.4627831715210356,
      "train_speed(iter/s)": 1.439975
    },
    {
      "epoch": 1.996915299258815,
      "grad_norm": 5.040591239929199,
      "learning_rate": 6.554298632126376e-05,
      "loss": 2.2719184875488283,
      "memory(GiB)": 77.56,
      "step": 46610,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.439991
    },
    {
      "epoch": 1.9971295145880639,
      "grad_norm": 4.640641689300537,
      "learning_rate": 6.553658982866073e-05,
      "loss": 2.4173145294189453,
      "memory(GiB)": 77.56,
      "step": 46615,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.440022
    },
    {
      "epoch": 1.997343729917313,
      "grad_norm": 5.445117950439453,
      "learning_rate": 6.55301930545978e-05,
      "loss": 2.614707183837891,
      "memory(GiB)": 77.56,
      "step": 46620,
      "token_acc": 0.46405228758169936,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 1.997557945246562,
      "grad_norm": 4.248864650726318,
      "learning_rate": 6.552379599919087e-05,
      "loss": 2.593882751464844,
      "memory(GiB)": 77.56,
      "step": 46625,
      "token_acc": 0.46757679180887374,
      "train_speed(iter/s)": 1.440062
    },
    {
      "epoch": 1.9977721605758108,
      "grad_norm": 4.769877910614014,
      "learning_rate": 6.551739866255578e-05,
      "loss": 2.4515695571899414,
      "memory(GiB)": 77.56,
      "step": 46630,
      "token_acc": 0.5101214574898786,
      "train_speed(iter/s)": 1.440088
    },
    {
      "epoch": 1.9979863759050598,
      "grad_norm": 6.324184894561768,
      "learning_rate": 6.551100104480849e-05,
      "loss": 2.6700746536254885,
      "memory(GiB)": 77.56,
      "step": 46635,
      "token_acc": 0.41423948220064727,
      "train_speed(iter/s)": 1.440113
    },
    {
      "epoch": 1.9982005912343088,
      "grad_norm": 4.573021411895752,
      "learning_rate": 6.550460314606486e-05,
      "loss": 2.392468070983887,
      "memory(GiB)": 77.56,
      "step": 46640,
      "token_acc": 0.49185667752442996,
      "train_speed(iter/s)": 1.440117
    },
    {
      "epoch": 1.9984148065635576,
      "grad_norm": 4.98002290725708,
      "learning_rate": 6.549820496644078e-05,
      "loss": 2.700870323181152,
      "memory(GiB)": 77.56,
      "step": 46645,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.440129
    },
    {
      "epoch": 1.9986290218928067,
      "grad_norm": 4.0348405838012695,
      "learning_rate": 6.54918065060522e-05,
      "loss": 2.7478660583496093,
      "memory(GiB)": 77.56,
      "step": 46650,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.440095
    },
    {
      "epoch": 1.9988432372220557,
      "grad_norm": 5.0965704917907715,
      "learning_rate": 6.5485407765015e-05,
      "loss": 2.3058755874633787,
      "memory(GiB)": 77.56,
      "step": 46655,
      "token_acc": 0.5399239543726235,
      "train_speed(iter/s)": 1.440079
    },
    {
      "epoch": 1.9990574525513045,
      "grad_norm": 4.812645435333252,
      "learning_rate": 6.54790087434451e-05,
      "loss": 2.449415588378906,
      "memory(GiB)": 77.56,
      "step": 46660,
      "token_acc": 0.4641638225255973,
      "train_speed(iter/s)": 1.440069
    },
    {
      "epoch": 1.9992716678805535,
      "grad_norm": 6.063690185546875,
      "learning_rate": 6.547260944145845e-05,
      "loss": 2.7162961959838867,
      "memory(GiB)": 77.56,
      "step": 46665,
      "token_acc": 0.4388059701492537,
      "train_speed(iter/s)": 1.440062
    },
    {
      "epoch": 1.9994858832098026,
      "grad_norm": 5.127791404724121,
      "learning_rate": 6.546620985917097e-05,
      "loss": 2.6050086975097657,
      "memory(GiB)": 77.56,
      "step": 46670,
      "token_acc": 0.45896656534954405,
      "train_speed(iter/s)": 1.440071
    },
    {
      "epoch": 1.9997000985390514,
      "grad_norm": 3.860870599746704,
      "learning_rate": 6.545980999669859e-05,
      "loss": 2.809811019897461,
      "memory(GiB)": 77.56,
      "step": 46675,
      "token_acc": 0.4745222929936306,
      "train_speed(iter/s)": 1.440085
    },
    {
      "epoch": 1.9999143138683004,
      "grad_norm": 6.640521049499512,
      "learning_rate": 6.545340985415726e-05,
      "loss": 2.9207460403442385,
      "memory(GiB)": 77.56,
      "step": 46680,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.440069
    },
    {
      "epoch": 2.0001285291975495,
      "grad_norm": 4.379486560821533,
      "learning_rate": 6.54470094316629e-05,
      "loss": 2.554909896850586,
      "memory(GiB)": 77.56,
      "step": 46685,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.440098
    },
    {
      "epoch": 2.0003427445267983,
      "grad_norm": 6.2013163566589355,
      "learning_rate": 6.544060872933146e-05,
      "loss": 2.7635061264038088,
      "memory(GiB)": 77.56,
      "step": 46690,
      "token_acc": 0.44039735099337746,
      "train_speed(iter/s)": 1.440079
    },
    {
      "epoch": 2.000556959856047,
      "grad_norm": 5.286303520202637,
      "learning_rate": 6.543420774727892e-05,
      "loss": 2.4695693969726564,
      "memory(GiB)": 77.56,
      "step": 46695,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.44007
    },
    {
      "epoch": 2.0007711751852963,
      "grad_norm": 4.1759138107299805,
      "learning_rate": 6.542780648562124e-05,
      "loss": 2.6401100158691406,
      "memory(GiB)": 77.56,
      "step": 46700,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.440058
    },
    {
      "epoch": 2.000985390514545,
      "grad_norm": 4.321265697479248,
      "learning_rate": 6.542140494447435e-05,
      "loss": 2.462251091003418,
      "memory(GiB)": 77.56,
      "step": 46705,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.44008
    },
    {
      "epoch": 2.001199605843794,
      "grad_norm": 4.87285852432251,
      "learning_rate": 6.541500312395427e-05,
      "loss": 2.2234607696533204,
      "memory(GiB)": 77.56,
      "step": 46710,
      "token_acc": 0.5137254901960784,
      "train_speed(iter/s)": 1.440097
    },
    {
      "epoch": 2.0014138211730432,
      "grad_norm": 4.688236236572266,
      "learning_rate": 6.540860102417693e-05,
      "loss": 2.3051342010498046,
      "memory(GiB)": 77.56,
      "step": 46715,
      "token_acc": 0.5362903225806451,
      "train_speed(iter/s)": 1.440146
    },
    {
      "epoch": 2.001628036502292,
      "grad_norm": 4.412413120269775,
      "learning_rate": 6.540219864525835e-05,
      "loss": 2.5773540496826173,
      "memory(GiB)": 77.56,
      "step": 46720,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.440168
    },
    {
      "epoch": 2.001842251831541,
      "grad_norm": 6.4903950691223145,
      "learning_rate": 6.539579598731447e-05,
      "loss": 2.61322021484375,
      "memory(GiB)": 77.56,
      "step": 46725,
      "token_acc": 0.4879518072289157,
      "train_speed(iter/s)": 1.440158
    },
    {
      "epoch": 2.00205646716079,
      "grad_norm": 5.339513778686523,
      "learning_rate": 6.53893930504613e-05,
      "loss": 2.4109384536743166,
      "memory(GiB)": 77.56,
      "step": 46730,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.44018
    },
    {
      "epoch": 2.002270682490039,
      "grad_norm": 4.855136394500732,
      "learning_rate": 6.538298983481484e-05,
      "loss": 2.5464282989501954,
      "memory(GiB)": 77.56,
      "step": 46735,
      "token_acc": 0.4307692307692308,
      "train_speed(iter/s)": 1.44017
    },
    {
      "epoch": 2.0024848978192877,
      "grad_norm": 6.4732255935668945,
      "learning_rate": 6.53765863404911e-05,
      "loss": 2.2858535766601564,
      "memory(GiB)": 77.56,
      "step": 46740,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.440203
    },
    {
      "epoch": 2.002699113148537,
      "grad_norm": 5.127806186676025,
      "learning_rate": 6.537018256760606e-05,
      "loss": 2.6056758880615236,
      "memory(GiB)": 77.56,
      "step": 46745,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.440238
    },
    {
      "epoch": 2.002913328477786,
      "grad_norm": 5.070985317230225,
      "learning_rate": 6.536377851627577e-05,
      "loss": 2.4653038024902343,
      "memory(GiB)": 77.56,
      "step": 46750,
      "token_acc": 0.4819277108433735,
      "train_speed(iter/s)": 1.440261
    },
    {
      "epoch": 2.0031275438070346,
      "grad_norm": 5.667211055755615,
      "learning_rate": 6.53573741866162e-05,
      "loss": 2.7156482696533204,
      "memory(GiB)": 77.56,
      "step": 46755,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.440316
    },
    {
      "epoch": 2.003341759136284,
      "grad_norm": 4.165454864501953,
      "learning_rate": 6.53509695787434e-05,
      "loss": 2.5176471710205077,
      "memory(GiB)": 77.56,
      "step": 46760,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.44034
    },
    {
      "epoch": 2.0035559744655327,
      "grad_norm": 5.048638820648193,
      "learning_rate": 6.534456469277337e-05,
      "loss": 2.4855140686035155,
      "memory(GiB)": 77.56,
      "step": 46765,
      "token_acc": 0.4766081871345029,
      "train_speed(iter/s)": 1.440328
    },
    {
      "epoch": 2.0037701897947815,
      "grad_norm": 4.935610771179199,
      "learning_rate": 6.533815952882216e-05,
      "loss": 2.1477176666259767,
      "memory(GiB)": 77.56,
      "step": 46770,
      "token_acc": 0.5325670498084292,
      "train_speed(iter/s)": 1.440306
    },
    {
      "epoch": 2.0039844051240308,
      "grad_norm": 4.180975914001465,
      "learning_rate": 6.53317540870058e-05,
      "loss": 2.447718620300293,
      "memory(GiB)": 77.56,
      "step": 46775,
      "token_acc": 0.5029940119760479,
      "train_speed(iter/s)": 1.440293
    },
    {
      "epoch": 2.0041986204532796,
      "grad_norm": 5.904799938201904,
      "learning_rate": 6.532534836744035e-05,
      "loss": 2.5206777572631838,
      "memory(GiB)": 77.56,
      "step": 46780,
      "token_acc": 0.4713375796178344,
      "train_speed(iter/s)": 1.440279
    },
    {
      "epoch": 2.0044128357825284,
      "grad_norm": 6.350618839263916,
      "learning_rate": 6.531894237024183e-05,
      "loss": 2.4447546005249023,
      "memory(GiB)": 77.56,
      "step": 46785,
      "token_acc": 0.5221843003412969,
      "train_speed(iter/s)": 1.44029
    },
    {
      "epoch": 2.0046270511117776,
      "grad_norm": 4.7659831047058105,
      "learning_rate": 6.53125360955263e-05,
      "loss": 2.087943267822266,
      "memory(GiB)": 77.56,
      "step": 46790,
      "token_acc": 0.5512367491166078,
      "train_speed(iter/s)": 1.44033
    },
    {
      "epoch": 2.0048412664410264,
      "grad_norm": 5.327122211456299,
      "learning_rate": 6.530612954340981e-05,
      "loss": 2.4332603454589843,
      "memory(GiB)": 77.56,
      "step": 46795,
      "token_acc": 0.5141955835962145,
      "train_speed(iter/s)": 1.440345
    },
    {
      "epoch": 2.0050554817702753,
      "grad_norm": 4.535036563873291,
      "learning_rate": 6.529972271400844e-05,
      "loss": 2.4609336853027344,
      "memory(GiB)": 77.56,
      "step": 46800,
      "token_acc": 0.49079754601226994,
      "train_speed(iter/s)": 1.440373
    },
    {
      "epoch": 2.0052696970995245,
      "grad_norm": 5.071470260620117,
      "learning_rate": 6.529331560743821e-05,
      "loss": 2.5745059967041017,
      "memory(GiB)": 77.56,
      "step": 46805,
      "token_acc": 0.46438746438746437,
      "train_speed(iter/s)": 1.44034
    },
    {
      "epoch": 2.0054839124287733,
      "grad_norm": 5.618152141571045,
      "learning_rate": 6.528690822381523e-05,
      "loss": 2.6603822708129883,
      "memory(GiB)": 77.56,
      "step": 46810,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.440361
    },
    {
      "epoch": 2.0056981277580226,
      "grad_norm": 5.464741230010986,
      "learning_rate": 6.528050056325558e-05,
      "loss": 2.4364561080932616,
      "memory(GiB)": 77.56,
      "step": 46815,
      "token_acc": 0.4564459930313589,
      "train_speed(iter/s)": 1.440396
    },
    {
      "epoch": 2.0059123430872714,
      "grad_norm": 5.506270408630371,
      "learning_rate": 6.527409262587533e-05,
      "loss": 2.46496639251709,
      "memory(GiB)": 77.56,
      "step": 46820,
      "token_acc": 0.50625,
      "train_speed(iter/s)": 1.440398
    },
    {
      "epoch": 2.00612655841652,
      "grad_norm": 7.7618560791015625,
      "learning_rate": 6.526768441179056e-05,
      "loss": 2.516340446472168,
      "memory(GiB)": 77.56,
      "step": 46825,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.440392
    },
    {
      "epoch": 2.0063407737457695,
      "grad_norm": 4.279303550720215,
      "learning_rate": 6.526127592111737e-05,
      "loss": 2.6124666213989256,
      "memory(GiB)": 77.56,
      "step": 46830,
      "token_acc": 0.45821325648414984,
      "train_speed(iter/s)": 1.440416
    },
    {
      "epoch": 2.0065549890750183,
      "grad_norm": 6.336496829986572,
      "learning_rate": 6.525486715397183e-05,
      "loss": 2.3455944061279297,
      "memory(GiB)": 77.56,
      "step": 46835,
      "token_acc": 0.5381818181818182,
      "train_speed(iter/s)": 1.440421
    },
    {
      "epoch": 2.006769204404267,
      "grad_norm": 5.114360809326172,
      "learning_rate": 6.524845811047009e-05,
      "loss": 2.3283687591552735,
      "memory(GiB)": 77.56,
      "step": 46840,
      "token_acc": 0.5179282868525896,
      "train_speed(iter/s)": 1.440416
    },
    {
      "epoch": 2.0069834197335163,
      "grad_norm": 6.016843795776367,
      "learning_rate": 6.52420487907282e-05,
      "loss": 2.154891014099121,
      "memory(GiB)": 77.56,
      "step": 46845,
      "token_acc": 0.5276752767527675,
      "train_speed(iter/s)": 1.440417
    },
    {
      "epoch": 2.007197635062765,
      "grad_norm": 4.609485626220703,
      "learning_rate": 6.52356391948623e-05,
      "loss": 2.2795175552368163,
      "memory(GiB)": 77.56,
      "step": 46850,
      "token_acc": 0.5444839857651246,
      "train_speed(iter/s)": 1.44042
    },
    {
      "epoch": 2.007411850392014,
      "grad_norm": 6.186790943145752,
      "learning_rate": 6.522922932298852e-05,
      "loss": 2.7255102157592774,
      "memory(GiB)": 77.56,
      "step": 46855,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.44045
    },
    {
      "epoch": 2.0076260657212632,
      "grad_norm": 4.760570526123047,
      "learning_rate": 6.522281917522295e-05,
      "loss": 2.4536609649658203,
      "memory(GiB)": 77.56,
      "step": 46860,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.440465
    },
    {
      "epoch": 2.007840281050512,
      "grad_norm": 4.162820816040039,
      "learning_rate": 6.521640875168175e-05,
      "loss": 2.3813400268554688,
      "memory(GiB)": 77.56,
      "step": 46865,
      "token_acc": 0.4959785522788204,
      "train_speed(iter/s)": 1.440492
    },
    {
      "epoch": 2.008054496379761,
      "grad_norm": 6.800650596618652,
      "learning_rate": 6.520999805248102e-05,
      "loss": 2.5655967712402346,
      "memory(GiB)": 77.56,
      "step": 46870,
      "token_acc": 0.4810126582278481,
      "train_speed(iter/s)": 1.440518
    },
    {
      "epoch": 2.00826871170901,
      "grad_norm": 6.280663967132568,
      "learning_rate": 6.520358707773691e-05,
      "loss": 2.3437000274658204,
      "memory(GiB)": 77.56,
      "step": 46875,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.440473
    },
    {
      "epoch": 2.008482927038259,
      "grad_norm": 5.3110432624816895,
      "learning_rate": 6.519717582756554e-05,
      "loss": 2.357120323181152,
      "memory(GiB)": 77.56,
      "step": 46880,
      "token_acc": 0.503125,
      "train_speed(iter/s)": 1.440469
    },
    {
      "epoch": 2.0086971423675077,
      "grad_norm": 3.271852731704712,
      "learning_rate": 6.519076430208308e-05,
      "loss": 2.3375322341918947,
      "memory(GiB)": 77.56,
      "step": 46885,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.440502
    },
    {
      "epoch": 2.008911357696757,
      "grad_norm": 5.670036792755127,
      "learning_rate": 6.518435250140569e-05,
      "loss": 2.4166229248046873,
      "memory(GiB)": 77.56,
      "step": 46890,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.440503
    },
    {
      "epoch": 2.009125573026006,
      "grad_norm": 4.191525459289551,
      "learning_rate": 6.517794042564951e-05,
      "loss": 2.3124170303344727,
      "memory(GiB)": 77.56,
      "step": 46895,
      "token_acc": 0.5255972696245734,
      "train_speed(iter/s)": 1.440523
    },
    {
      "epoch": 2.0093397883552546,
      "grad_norm": 5.3819708824157715,
      "learning_rate": 6.51715280749307e-05,
      "loss": 2.7740522384643556,
      "memory(GiB)": 77.56,
      "step": 46900,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.440492
    },
    {
      "epoch": 2.009554003684504,
      "grad_norm": 5.581313610076904,
      "learning_rate": 6.516511544936542e-05,
      "loss": 2.2341598510742187,
      "memory(GiB)": 77.56,
      "step": 46905,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.440451
    },
    {
      "epoch": 2.0097682190137527,
      "grad_norm": 5.02405309677124,
      "learning_rate": 6.515870254906984e-05,
      "loss": 2.3327259063720702,
      "memory(GiB)": 77.56,
      "step": 46910,
      "token_acc": 0.4720496894409938,
      "train_speed(iter/s)": 1.440477
    },
    {
      "epoch": 2.0099824343430015,
      "grad_norm": 4.830881595611572,
      "learning_rate": 6.515228937416015e-05,
      "loss": 2.210921287536621,
      "memory(GiB)": 77.56,
      "step": 46915,
      "token_acc": 0.5255474452554745,
      "train_speed(iter/s)": 1.440443
    },
    {
      "epoch": 2.0101966496722508,
      "grad_norm": 5.542373180389404,
      "learning_rate": 6.514587592475252e-05,
      "loss": 2.493315315246582,
      "memory(GiB)": 77.56,
      "step": 46920,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.440466
    },
    {
      "epoch": 2.0104108650014996,
      "grad_norm": 6.550668239593506,
      "learning_rate": 6.513946220096316e-05,
      "loss": 2.410256195068359,
      "memory(GiB)": 77.56,
      "step": 46925,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.440476
    },
    {
      "epoch": 2.0106250803307484,
      "grad_norm": 5.8648457527160645,
      "learning_rate": 6.513304820290822e-05,
      "loss": 2.301552581787109,
      "memory(GiB)": 77.56,
      "step": 46930,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.440436
    },
    {
      "epoch": 2.0108392956599976,
      "grad_norm": 5.672832012176514,
      "learning_rate": 6.512663393070392e-05,
      "loss": 2.343380355834961,
      "memory(GiB)": 77.56,
      "step": 46935,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.44045
    },
    {
      "epoch": 2.0110535109892465,
      "grad_norm": 5.217822074890137,
      "learning_rate": 6.512021938446646e-05,
      "loss": 2.420007514953613,
      "memory(GiB)": 77.56,
      "step": 46940,
      "token_acc": 0.46747967479674796,
      "train_speed(iter/s)": 1.440478
    },
    {
      "epoch": 2.0112677263184953,
      "grad_norm": 4.892837047576904,
      "learning_rate": 6.511380456431204e-05,
      "loss": 2.4698442459106444,
      "memory(GiB)": 77.56,
      "step": 46945,
      "token_acc": 0.47904191616766467,
      "train_speed(iter/s)": 1.44047
    },
    {
      "epoch": 2.0114819416477445,
      "grad_norm": 4.919934272766113,
      "learning_rate": 6.510738947035687e-05,
      "loss": 2.6515804290771485,
      "memory(GiB)": 77.56,
      "step": 46950,
      "token_acc": 0.4727272727272727,
      "train_speed(iter/s)": 1.440483
    },
    {
      "epoch": 2.0116961569769933,
      "grad_norm": 6.32347297668457,
      "learning_rate": 6.510097410271717e-05,
      "loss": 2.543160820007324,
      "memory(GiB)": 77.56,
      "step": 46955,
      "token_acc": 0.5063291139240507,
      "train_speed(iter/s)": 1.440459
    },
    {
      "epoch": 2.011910372306242,
      "grad_norm": 5.247535705566406,
      "learning_rate": 6.509455846150916e-05,
      "loss": 2.0123329162597656,
      "memory(GiB)": 77.56,
      "step": 46960,
      "token_acc": 0.5529411764705883,
      "train_speed(iter/s)": 1.440479
    },
    {
      "epoch": 2.0121245876354914,
      "grad_norm": 3.964775323867798,
      "learning_rate": 6.508814254684906e-05,
      "loss": 2.4224483489990236,
      "memory(GiB)": 77.56,
      "step": 46965,
      "token_acc": 0.46048109965635736,
      "train_speed(iter/s)": 1.440466
    },
    {
      "epoch": 2.01233880296474,
      "grad_norm": 7.648169994354248,
      "learning_rate": 6.50817263588531e-05,
      "loss": 2.4166791915893553,
      "memory(GiB)": 77.56,
      "step": 46970,
      "token_acc": 0.5158730158730159,
      "train_speed(iter/s)": 1.440471
    },
    {
      "epoch": 2.012553018293989,
      "grad_norm": 4.846028804779053,
      "learning_rate": 6.507530989763754e-05,
      "loss": 2.4553701400756838,
      "memory(GiB)": 77.56,
      "step": 46975,
      "token_acc": 0.42378048780487804,
      "train_speed(iter/s)": 1.440472
    },
    {
      "epoch": 2.0127672336232383,
      "grad_norm": 7.003790855407715,
      "learning_rate": 6.506889316331858e-05,
      "loss": 2.522560882568359,
      "memory(GiB)": 77.56,
      "step": 46980,
      "token_acc": 0.48046875,
      "train_speed(iter/s)": 1.440494
    },
    {
      "epoch": 2.012981448952487,
      "grad_norm": 5.9567060470581055,
      "learning_rate": 6.50624761560125e-05,
      "loss": 2.231857681274414,
      "memory(GiB)": 77.56,
      "step": 46985,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.440499
    },
    {
      "epoch": 2.013195664281736,
      "grad_norm": 4.712182521820068,
      "learning_rate": 6.505605887583552e-05,
      "loss": 2.4869033813476564,
      "memory(GiB)": 77.56,
      "step": 46990,
      "token_acc": 0.46557377049180326,
      "train_speed(iter/s)": 1.440436
    },
    {
      "epoch": 2.013409879610985,
      "grad_norm": 4.36808967590332,
      "learning_rate": 6.50496413229039e-05,
      "loss": 2.0867469787597654,
      "memory(GiB)": 77.56,
      "step": 46995,
      "token_acc": 0.5632183908045977,
      "train_speed(iter/s)": 1.440444
    },
    {
      "epoch": 2.013624094940234,
      "grad_norm": 5.934143543243408,
      "learning_rate": 6.504322349733393e-05,
      "loss": 2.640213203430176,
      "memory(GiB)": 77.56,
      "step": 47000,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.440445
    },
    {
      "epoch": 2.013624094940234,
      "eval_loss": 2.2116146087646484,
      "eval_runtime": 14.3282,
      "eval_samples_per_second": 6.979,
      "eval_steps_per_second": 6.979,
      "eval_token_acc": 0.4889807162534435,
      "step": 47000
    },
    {
      "epoch": 2.013838310269483,
      "grad_norm": 4.928823947906494,
      "learning_rate": 6.503680539924184e-05,
      "loss": 2.2399688720703126,
      "memory(GiB)": 77.56,
      "step": 47005,
      "token_acc": 0.49753208292201384,
      "train_speed(iter/s)": 1.439758
    },
    {
      "epoch": 2.014052525598732,
      "grad_norm": 5.141942977905273,
      "learning_rate": 6.503038702874394e-05,
      "loss": 2.4522708892822265,
      "memory(GiB)": 77.56,
      "step": 47010,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.439704
    },
    {
      "epoch": 2.014266740927981,
      "grad_norm": 5.53627872467041,
      "learning_rate": 6.502396838595646e-05,
      "loss": 2.5175106048583986,
      "memory(GiB)": 77.56,
      "step": 47015,
      "token_acc": 0.4674922600619195,
      "train_speed(iter/s)": 1.439738
    },
    {
      "epoch": 2.0144809562572297,
      "grad_norm": 7.602707862854004,
      "learning_rate": 6.50175494709957e-05,
      "loss": 2.413856315612793,
      "memory(GiB)": 77.56,
      "step": 47020,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.43971
    },
    {
      "epoch": 2.014695171586479,
      "grad_norm": 5.674644470214844,
      "learning_rate": 6.501113028397793e-05,
      "loss": 2.4417697906494142,
      "memory(GiB)": 77.56,
      "step": 47025,
      "token_acc": 0.5259259259259259,
      "train_speed(iter/s)": 1.439694
    },
    {
      "epoch": 2.0149093869157277,
      "grad_norm": 5.88431978225708,
      "learning_rate": 6.500471082501948e-05,
      "loss": 2.6259002685546875,
      "memory(GiB)": 77.56,
      "step": 47030,
      "token_acc": 0.4152823920265781,
      "train_speed(iter/s)": 1.439696
    },
    {
      "epoch": 2.0151236022449766,
      "grad_norm": 7.305669784545898,
      "learning_rate": 6.49982910942366e-05,
      "loss": 2.230219268798828,
      "memory(GiB)": 77.56,
      "step": 47035,
      "token_acc": 0.4891640866873065,
      "train_speed(iter/s)": 1.439662
    },
    {
      "epoch": 2.015337817574226,
      "grad_norm": 4.789058685302734,
      "learning_rate": 6.499187109174561e-05,
      "loss": 2.5457658767700195,
      "memory(GiB)": 77.56,
      "step": 47040,
      "token_acc": 0.496875,
      "train_speed(iter/s)": 1.439626
    },
    {
      "epoch": 2.0155520329034746,
      "grad_norm": 4.733511924743652,
      "learning_rate": 6.498545081766282e-05,
      "loss": 2.266896438598633,
      "memory(GiB)": 77.56,
      "step": 47045,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.439608
    },
    {
      "epoch": 2.0157662482327234,
      "grad_norm": 3.543917417526245,
      "learning_rate": 6.497903027210453e-05,
      "loss": 2.7028133392333986,
      "memory(GiB)": 77.56,
      "step": 47050,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.439643
    },
    {
      "epoch": 2.0159804635619727,
      "grad_norm": 5.878790855407715,
      "learning_rate": 6.497260945518706e-05,
      "loss": 2.4566299438476564,
      "memory(GiB)": 77.56,
      "step": 47055,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.439678
    },
    {
      "epoch": 2.0161946788912215,
      "grad_norm": 4.004664897918701,
      "learning_rate": 6.496618836702672e-05,
      "loss": 2.1462165832519533,
      "memory(GiB)": 77.56,
      "step": 47060,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439684
    },
    {
      "epoch": 2.0164088942204703,
      "grad_norm": 4.003391265869141,
      "learning_rate": 6.49597670077398e-05,
      "loss": 2.233612823486328,
      "memory(GiB)": 77.56,
      "step": 47065,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.439638
    },
    {
      "epoch": 2.0166231095497196,
      "grad_norm": 4.659119606018066,
      "learning_rate": 6.495334537744271e-05,
      "loss": 2.246729850769043,
      "memory(GiB)": 77.56,
      "step": 47070,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.439648
    },
    {
      "epoch": 2.0168373248789684,
      "grad_norm": 5.30164098739624,
      "learning_rate": 6.494692347625174e-05,
      "loss": 2.328721618652344,
      "memory(GiB)": 77.56,
      "step": 47075,
      "token_acc": 0.5606694560669456,
      "train_speed(iter/s)": 1.439615
    },
    {
      "epoch": 2.017051540208217,
      "grad_norm": 6.037069320678711,
      "learning_rate": 6.494050130428323e-05,
      "loss": 2.4070837020874025,
      "memory(GiB)": 77.56,
      "step": 47080,
      "token_acc": 0.477124183006536,
      "train_speed(iter/s)": 1.439617
    },
    {
      "epoch": 2.0172657555374665,
      "grad_norm": 4.738967418670654,
      "learning_rate": 6.493407886165351e-05,
      "loss": 2.786213684082031,
      "memory(GiB)": 77.56,
      "step": 47085,
      "token_acc": 0.4673202614379085,
      "train_speed(iter/s)": 1.439654
    },
    {
      "epoch": 2.0174799708667153,
      "grad_norm": 6.387697219848633,
      "learning_rate": 6.492765614847896e-05,
      "loss": 2.393421173095703,
      "memory(GiB)": 77.56,
      "step": 47090,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.439688
    },
    {
      "epoch": 2.017694186195964,
      "grad_norm": 6.092655658721924,
      "learning_rate": 6.492123316487589e-05,
      "loss": 2.5422796249389648,
      "memory(GiB)": 77.56,
      "step": 47095,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.439684
    },
    {
      "epoch": 2.0179084015252133,
      "grad_norm": 5.014275550842285,
      "learning_rate": 6.49148099109607e-05,
      "loss": 2.3574142456054688,
      "memory(GiB)": 77.56,
      "step": 47100,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.439658
    },
    {
      "epoch": 2.018122616854462,
      "grad_norm": 5.569258689880371,
      "learning_rate": 6.490838638684973e-05,
      "loss": 2.5010326385498045,
      "memory(GiB)": 77.56,
      "step": 47105,
      "token_acc": 0.44565217391304346,
      "train_speed(iter/s)": 1.439626
    },
    {
      "epoch": 2.018336832183711,
      "grad_norm": 4.812034606933594,
      "learning_rate": 6.490196259265934e-05,
      "loss": 2.405223846435547,
      "memory(GiB)": 77.56,
      "step": 47110,
      "token_acc": 0.46439628482972134,
      "train_speed(iter/s)": 1.43962
    },
    {
      "epoch": 2.01855104751296,
      "grad_norm": 5.988837242126465,
      "learning_rate": 6.489553852850596e-05,
      "loss": 2.651731491088867,
      "memory(GiB)": 77.56,
      "step": 47115,
      "token_acc": 0.4536741214057508,
      "train_speed(iter/s)": 1.43963
    },
    {
      "epoch": 2.018765262842209,
      "grad_norm": 4.13277006149292,
      "learning_rate": 6.488911419450591e-05,
      "loss": 2.352150726318359,
      "memory(GiB)": 77.56,
      "step": 47120,
      "token_acc": 0.4927007299270073,
      "train_speed(iter/s)": 1.439639
    },
    {
      "epoch": 2.018979478171458,
      "grad_norm": 5.49845552444458,
      "learning_rate": 6.488268959077557e-05,
      "loss": 2.3887655258178713,
      "memory(GiB)": 77.56,
      "step": 47125,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.439638
    },
    {
      "epoch": 2.019193693500707,
      "grad_norm": 6.266018867492676,
      "learning_rate": 6.487626471743137e-05,
      "loss": 2.502766418457031,
      "memory(GiB)": 77.56,
      "step": 47130,
      "token_acc": 0.4479166666666667,
      "train_speed(iter/s)": 1.439647
    },
    {
      "epoch": 2.019407908829956,
      "grad_norm": 4.618208408355713,
      "learning_rate": 6.486983957458967e-05,
      "loss": 2.315196228027344,
      "memory(GiB)": 77.56,
      "step": 47135,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439657
    },
    {
      "epoch": 2.0196221241592047,
      "grad_norm": 4.724485397338867,
      "learning_rate": 6.486341416236687e-05,
      "loss": 2.527619743347168,
      "memory(GiB)": 77.56,
      "step": 47140,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.439693
    },
    {
      "epoch": 2.019836339488454,
      "grad_norm": 6.78851318359375,
      "learning_rate": 6.485698848087939e-05,
      "loss": 2.1528024673461914,
      "memory(GiB)": 77.56,
      "step": 47145,
      "token_acc": 0.552901023890785,
      "train_speed(iter/s)": 1.439742
    },
    {
      "epoch": 2.020050554817703,
      "grad_norm": 6.4735188484191895,
      "learning_rate": 6.485056253024362e-05,
      "loss": 2.553872299194336,
      "memory(GiB)": 77.56,
      "step": 47150,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.439741
    },
    {
      "epoch": 2.0202647701469516,
      "grad_norm": 4.260812282562256,
      "learning_rate": 6.484413631057599e-05,
      "loss": 2.3631893157958985,
      "memory(GiB)": 77.56,
      "step": 47155,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439731
    },
    {
      "epoch": 2.020478985476201,
      "grad_norm": 4.051983833312988,
      "learning_rate": 6.48377098219929e-05,
      "loss": 2.3516508102416993,
      "memory(GiB)": 77.56,
      "step": 47160,
      "token_acc": 0.5168195718654435,
      "train_speed(iter/s)": 1.439717
    },
    {
      "epoch": 2.0206932008054497,
      "grad_norm": 4.177536964416504,
      "learning_rate": 6.483128306461078e-05,
      "loss": 2.4834449768066404,
      "memory(GiB)": 77.56,
      "step": 47165,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.439707
    },
    {
      "epoch": 2.0209074161346985,
      "grad_norm": 5.729572296142578,
      "learning_rate": 6.482485603854604e-05,
      "loss": 2.5365907669067385,
      "memory(GiB)": 77.56,
      "step": 47170,
      "token_acc": 0.451505016722408,
      "train_speed(iter/s)": 1.439704
    },
    {
      "epoch": 2.0211216314639477,
      "grad_norm": 10.399534225463867,
      "learning_rate": 6.481842874391514e-05,
      "loss": 2.585718536376953,
      "memory(GiB)": 77.56,
      "step": 47175,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.439709
    },
    {
      "epoch": 2.0213358467931966,
      "grad_norm": 6.627134323120117,
      "learning_rate": 6.481200118083449e-05,
      "loss": 2.3237384796142577,
      "memory(GiB)": 77.56,
      "step": 47180,
      "token_acc": 0.48264984227129337,
      "train_speed(iter/s)": 1.439725
    },
    {
      "epoch": 2.0215500621224454,
      "grad_norm": 4.1366167068481445,
      "learning_rate": 6.480557334942055e-05,
      "loss": 2.5460943222045898,
      "memory(GiB)": 77.56,
      "step": 47185,
      "token_acc": 0.4517133956386293,
      "train_speed(iter/s)": 1.439753
    },
    {
      "epoch": 2.0217642774516946,
      "grad_norm": 7.300133228302002,
      "learning_rate": 6.479914524978978e-05,
      "loss": 2.6619863510131836,
      "memory(GiB)": 77.56,
      "step": 47190,
      "token_acc": 0.44224422442244227,
      "train_speed(iter/s)": 1.439769
    },
    {
      "epoch": 2.0219784927809434,
      "grad_norm": 5.019187927246094,
      "learning_rate": 6.479271688205858e-05,
      "loss": 2.5283918380737305,
      "memory(GiB)": 77.56,
      "step": 47195,
      "token_acc": 0.509090909090909,
      "train_speed(iter/s)": 1.439776
    },
    {
      "epoch": 2.0221927081101922,
      "grad_norm": 5.677464485168457,
      "learning_rate": 6.478628824634346e-05,
      "loss": 2.6112157821655275,
      "memory(GiB)": 77.56,
      "step": 47200,
      "token_acc": 0.49382716049382713,
      "train_speed(iter/s)": 1.439766
    },
    {
      "epoch": 2.0224069234394415,
      "grad_norm": 3.9665565490722656,
      "learning_rate": 6.477985934276085e-05,
      "loss": 2.4552179336547852,
      "memory(GiB)": 77.56,
      "step": 47205,
      "token_acc": 0.4612903225806452,
      "train_speed(iter/s)": 1.439774
    },
    {
      "epoch": 2.0226211387686903,
      "grad_norm": 6.540243625640869,
      "learning_rate": 6.477343017142722e-05,
      "loss": 2.491834259033203,
      "memory(GiB)": 77.56,
      "step": 47210,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.439777
    },
    {
      "epoch": 2.022835354097939,
      "grad_norm": 7.2349467277526855,
      "learning_rate": 6.476700073245904e-05,
      "loss": 1.9539417266845702,
      "memory(GiB)": 77.56,
      "step": 47215,
      "token_acc": 0.5700934579439252,
      "train_speed(iter/s)": 1.439774
    },
    {
      "epoch": 2.0230495694271884,
      "grad_norm": 4.291737079620361,
      "learning_rate": 6.47605710259728e-05,
      "loss": 2.1844161987304687,
      "memory(GiB)": 77.56,
      "step": 47220,
      "token_acc": 0.535593220338983,
      "train_speed(iter/s)": 1.439792
    },
    {
      "epoch": 2.023263784756437,
      "grad_norm": 4.990181922912598,
      "learning_rate": 6.475414105208497e-05,
      "loss": 2.6633190155029296,
      "memory(GiB)": 77.56,
      "step": 47225,
      "token_acc": 0.45132743362831856,
      "train_speed(iter/s)": 1.43978
    },
    {
      "epoch": 2.023478000085686,
      "grad_norm": 4.123158931732178,
      "learning_rate": 6.474771081091202e-05,
      "loss": 2.2368452072143556,
      "memory(GiB)": 77.56,
      "step": 47230,
      "token_acc": 0.5031645569620253,
      "train_speed(iter/s)": 1.439784
    },
    {
      "epoch": 2.0236922154149353,
      "grad_norm": 5.408724308013916,
      "learning_rate": 6.474128030257047e-05,
      "loss": 2.443004608154297,
      "memory(GiB)": 77.56,
      "step": 47235,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.439821
    },
    {
      "epoch": 2.023906430744184,
      "grad_norm": 6.242146015167236,
      "learning_rate": 6.473484952717679e-05,
      "loss": 2.7575626373291016,
      "memory(GiB)": 77.56,
      "step": 47240,
      "token_acc": 0.4542483660130719,
      "train_speed(iter/s)": 1.439843
    },
    {
      "epoch": 2.024120646073433,
      "grad_norm": 5.340526103973389,
      "learning_rate": 6.472841848484749e-05,
      "loss": 2.4209598541259765,
      "memory(GiB)": 77.56,
      "step": 47245,
      "token_acc": 0.4778761061946903,
      "train_speed(iter/s)": 1.439856
    },
    {
      "epoch": 2.024334861402682,
      "grad_norm": 5.6754608154296875,
      "learning_rate": 6.472198717569907e-05,
      "loss": 2.383496856689453,
      "memory(GiB)": 77.56,
      "step": 47250,
      "token_acc": 0.48598130841121495,
      "train_speed(iter/s)": 1.439918
    },
    {
      "epoch": 2.024549076731931,
      "grad_norm": 6.373967170715332,
      "learning_rate": 6.471555559984804e-05,
      "loss": 2.695840835571289,
      "memory(GiB)": 77.56,
      "step": 47255,
      "token_acc": 0.4409722222222222,
      "train_speed(iter/s)": 1.439945
    },
    {
      "epoch": 2.0247632920611798,
      "grad_norm": 4.937572479248047,
      "learning_rate": 6.470912375741093e-05,
      "loss": 2.2277603149414062,
      "memory(GiB)": 77.56,
      "step": 47260,
      "token_acc": 0.5059880239520959,
      "train_speed(iter/s)": 1.43994
    },
    {
      "epoch": 2.024977507390429,
      "grad_norm": 3.9886159896850586,
      "learning_rate": 6.470269164850424e-05,
      "loss": 2.7752742767333984,
      "memory(GiB)": 77.56,
      "step": 47265,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.439963
    },
    {
      "epoch": 2.025191722719678,
      "grad_norm": 5.3319220542907715,
      "learning_rate": 6.469625927324448e-05,
      "loss": 2.64019775390625,
      "memory(GiB)": 77.56,
      "step": 47270,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.439975
    },
    {
      "epoch": 2.0254059380489267,
      "grad_norm": 7.283320903778076,
      "learning_rate": 6.468982663174822e-05,
      "loss": 2.2063600540161135,
      "memory(GiB)": 77.56,
      "step": 47275,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.440025
    },
    {
      "epoch": 2.025620153378176,
      "grad_norm": 4.875926971435547,
      "learning_rate": 6.468339372413196e-05,
      "loss": 2.5367393493652344,
      "memory(GiB)": 77.56,
      "step": 47280,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.440015
    },
    {
      "epoch": 2.0258343687074247,
      "grad_norm": 5.891599655151367,
      "learning_rate": 6.467696055051225e-05,
      "loss": 2.403213882446289,
      "memory(GiB)": 77.56,
      "step": 47285,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.440022
    },
    {
      "epoch": 2.0260485840366735,
      "grad_norm": 6.47664213180542,
      "learning_rate": 6.467052711100563e-05,
      "loss": 2.37048225402832,
      "memory(GiB)": 77.56,
      "step": 47290,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.439986
    },
    {
      "epoch": 2.026262799365923,
      "grad_norm": 6.752597808837891,
      "learning_rate": 6.466409340572864e-05,
      "loss": 2.0826969146728516,
      "memory(GiB)": 77.56,
      "step": 47295,
      "token_acc": 0.5498154981549815,
      "train_speed(iter/s)": 1.439992
    },
    {
      "epoch": 2.0264770146951716,
      "grad_norm": 5.42634916305542,
      "learning_rate": 6.465765943479785e-05,
      "loss": 2.737038803100586,
      "memory(GiB)": 77.56,
      "step": 47300,
      "token_acc": 0.4340277777777778,
      "train_speed(iter/s)": 1.440002
    },
    {
      "epoch": 2.0266912300244204,
      "grad_norm": 4.91648006439209,
      "learning_rate": 6.465122519832982e-05,
      "loss": 2.392622375488281,
      "memory(GiB)": 77.56,
      "step": 47305,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.440012
    },
    {
      "epoch": 2.0269054453536697,
      "grad_norm": 5.535400867462158,
      "learning_rate": 6.464479069644109e-05,
      "loss": 2.3067462921142576,
      "memory(GiB)": 77.56,
      "step": 47310,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.439989
    },
    {
      "epoch": 2.0271196606829185,
      "grad_norm": 4.595633506774902,
      "learning_rate": 6.463835592924825e-05,
      "loss": 2.758593940734863,
      "memory(GiB)": 77.56,
      "step": 47315,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.440038
    },
    {
      "epoch": 2.0273338760121673,
      "grad_norm": 6.745107650756836,
      "learning_rate": 6.463192089686786e-05,
      "loss": 2.5533119201660157,
      "memory(GiB)": 77.56,
      "step": 47320,
      "token_acc": 0.4664179104477612,
      "train_speed(iter/s)": 1.440074
    },
    {
      "epoch": 2.0275480913414166,
      "grad_norm": 4.424008369445801,
      "learning_rate": 6.462548559941648e-05,
      "loss": 2.306426429748535,
      "memory(GiB)": 77.56,
      "step": 47325,
      "token_acc": 0.49145299145299143,
      "train_speed(iter/s)": 1.440073
    },
    {
      "epoch": 2.0277623066706654,
      "grad_norm": 5.835791110992432,
      "learning_rate": 6.461905003701072e-05,
      "loss": 2.314540481567383,
      "memory(GiB)": 77.56,
      "step": 47330,
      "token_acc": 0.5423076923076923,
      "train_speed(iter/s)": 1.440085
    },
    {
      "epoch": 2.027976521999914,
      "grad_norm": 6.8580403327941895,
      "learning_rate": 6.461261420976716e-05,
      "loss": 2.4569332122802736,
      "memory(GiB)": 77.56,
      "step": 47335,
      "token_acc": 0.5036231884057971,
      "train_speed(iter/s)": 1.440101
    },
    {
      "epoch": 2.0281907373291634,
      "grad_norm": 6.78116512298584,
      "learning_rate": 6.460617811780238e-05,
      "loss": 2.277758026123047,
      "memory(GiB)": 77.56,
      "step": 47340,
      "token_acc": 0.48909657320872274,
      "train_speed(iter/s)": 1.440116
    },
    {
      "epoch": 2.0284049526584123,
      "grad_norm": 5.802482604980469,
      "learning_rate": 6.459974176123299e-05,
      "loss": 2.4579164505004885,
      "memory(GiB)": 77.56,
      "step": 47345,
      "token_acc": 0.5280898876404494,
      "train_speed(iter/s)": 1.440088
    },
    {
      "epoch": 2.028619167987661,
      "grad_norm": 5.421518325805664,
      "learning_rate": 6.459330514017556e-05,
      "loss": 2.506378746032715,
      "memory(GiB)": 77.56,
      "step": 47350,
      "token_acc": 0.4468864468864469,
      "train_speed(iter/s)": 1.440084
    },
    {
      "epoch": 2.0288333833169103,
      "grad_norm": 4.904119968414307,
      "learning_rate": 6.458686825474674e-05,
      "loss": 2.421441078186035,
      "memory(GiB)": 77.56,
      "step": 47355,
      "token_acc": 0.47384615384615386,
      "train_speed(iter/s)": 1.440041
    },
    {
      "epoch": 2.029047598646159,
      "grad_norm": 6.118224620819092,
      "learning_rate": 6.45804311050631e-05,
      "loss": 2.717330551147461,
      "memory(GiB)": 77.56,
      "step": 47360,
      "token_acc": 0.47214076246334313,
      "train_speed(iter/s)": 1.440061
    },
    {
      "epoch": 2.029261813975408,
      "grad_norm": 3.993043899536133,
      "learning_rate": 6.457399369124129e-05,
      "loss": 2.1258781433105467,
      "memory(GiB)": 77.56,
      "step": 47365,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.440068
    },
    {
      "epoch": 2.029476029304657,
      "grad_norm": 4.975916862487793,
      "learning_rate": 6.456755601339789e-05,
      "loss": 2.434589385986328,
      "memory(GiB)": 77.56,
      "step": 47370,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.440074
    },
    {
      "epoch": 2.029690244633906,
      "grad_norm": 6.480383396148682,
      "learning_rate": 6.456111807164958e-05,
      "loss": 2.5920232772827148,
      "memory(GiB)": 77.56,
      "step": 47375,
      "token_acc": 0.4551282051282051,
      "train_speed(iter/s)": 1.440092
    },
    {
      "epoch": 2.029904459963155,
      "grad_norm": 5.017740726470947,
      "learning_rate": 6.455467986611293e-05,
      "loss": 2.416194534301758,
      "memory(GiB)": 77.56,
      "step": 47380,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.440103
    },
    {
      "epoch": 2.030118675292404,
      "grad_norm": 5.090701103210449,
      "learning_rate": 6.45482413969046e-05,
      "loss": 2.193351936340332,
      "memory(GiB)": 77.56,
      "step": 47385,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.440083
    },
    {
      "epoch": 2.030332890621653,
      "grad_norm": 6.160708427429199,
      "learning_rate": 6.454180266414125e-05,
      "loss": 2.466547966003418,
      "memory(GiB)": 77.56,
      "step": 47390,
      "token_acc": 0.4493927125506073,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 2.0305471059509017,
      "grad_norm": 6.454678535461426,
      "learning_rate": 6.453536366793949e-05,
      "loss": 2.6635141372680664,
      "memory(GiB)": 77.56,
      "step": 47395,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 2.030761321280151,
      "grad_norm": 5.459461688995361,
      "learning_rate": 6.452892440841597e-05,
      "loss": 2.288252067565918,
      "memory(GiB)": 77.56,
      "step": 47400,
      "token_acc": 0.484251968503937,
      "train_speed(iter/s)": 1.44003
    },
    {
      "epoch": 2.0309755366094,
      "grad_norm": 5.283269882202148,
      "learning_rate": 6.452248488568738e-05,
      "loss": 2.5599859237670897,
      "memory(GiB)": 77.56,
      "step": 47405,
      "token_acc": 0.41643835616438357,
      "train_speed(iter/s)": 1.440044
    },
    {
      "epoch": 2.0311897519386486,
      "grad_norm": 5.3509440422058105,
      "learning_rate": 6.451604509987033e-05,
      "loss": 2.2912080764770506,
      "memory(GiB)": 77.56,
      "step": 47410,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.440014
    },
    {
      "epoch": 2.031403967267898,
      "grad_norm": 5.202113151550293,
      "learning_rate": 6.450960505108153e-05,
      "loss": 2.434415817260742,
      "memory(GiB)": 77.56,
      "step": 47415,
      "token_acc": 0.47398843930635837,
      "train_speed(iter/s)": 1.439984
    },
    {
      "epoch": 2.0316181825971467,
      "grad_norm": 4.731122016906738,
      "learning_rate": 6.450316473943763e-05,
      "loss": 2.546631622314453,
      "memory(GiB)": 77.56,
      "step": 47420,
      "token_acc": 0.49523809523809526,
      "train_speed(iter/s)": 1.439986
    },
    {
      "epoch": 2.0318323979263955,
      "grad_norm": 8.232008934020996,
      "learning_rate": 6.449672416505526e-05,
      "loss": 2.2981239318847657,
      "memory(GiB)": 77.56,
      "step": 47425,
      "token_acc": 0.5033112582781457,
      "train_speed(iter/s)": 1.439969
    },
    {
      "epoch": 2.0320466132556447,
      "grad_norm": 6.969511032104492,
      "learning_rate": 6.449028332805117e-05,
      "loss": 2.6346397399902344,
      "memory(GiB)": 77.56,
      "step": 47430,
      "token_acc": 0.44715447154471544,
      "train_speed(iter/s)": 1.440005
    },
    {
      "epoch": 2.0322608285848935,
      "grad_norm": 4.844028472900391,
      "learning_rate": 6.448384222854197e-05,
      "loss": 2.2274232864379884,
      "memory(GiB)": 77.56,
      "step": 47435,
      "token_acc": 0.5153374233128835,
      "train_speed(iter/s)": 1.440007
    },
    {
      "epoch": 2.0324750439141424,
      "grad_norm": 7.408126354217529,
      "learning_rate": 6.447740086664439e-05,
      "loss": 2.150931549072266,
      "memory(GiB)": 77.56,
      "step": 47440,
      "token_acc": 0.5145228215767634,
      "train_speed(iter/s)": 1.440031
    },
    {
      "epoch": 2.0326892592433916,
      "grad_norm": 5.0446295738220215,
      "learning_rate": 6.447095924247512e-05,
      "loss": 2.460962104797363,
      "memory(GiB)": 77.56,
      "step": 47445,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.440036
    },
    {
      "epoch": 2.0329034745726404,
      "grad_norm": 6.160946846008301,
      "learning_rate": 6.446451735615085e-05,
      "loss": 2.326716995239258,
      "memory(GiB)": 77.56,
      "step": 47450,
      "token_acc": 0.5251141552511416,
      "train_speed(iter/s)": 1.440031
    },
    {
      "epoch": 2.0331176899018892,
      "grad_norm": 6.735740661621094,
      "learning_rate": 6.445807520778828e-05,
      "loss": 2.156563568115234,
      "memory(GiB)": 77.56,
      "step": 47455,
      "token_acc": 0.5062240663900415,
      "train_speed(iter/s)": 1.44002
    },
    {
      "epoch": 2.0333319052311385,
      "grad_norm": 5.7068634033203125,
      "learning_rate": 6.445163279750412e-05,
      "loss": 2.3455982208251953,
      "memory(GiB)": 77.56,
      "step": 47460,
      "token_acc": 0.5136986301369864,
      "train_speed(iter/s)": 1.440049
    },
    {
      "epoch": 2.0335461205603873,
      "grad_norm": 5.048141956329346,
      "learning_rate": 6.444519012541507e-05,
      "loss": 2.0552072525024414,
      "memory(GiB)": 77.56,
      "step": 47465,
      "token_acc": 0.5811320754716981,
      "train_speed(iter/s)": 1.440062
    },
    {
      "epoch": 2.033760335889636,
      "grad_norm": 7.297991752624512,
      "learning_rate": 6.443874719163782e-05,
      "loss": 2.100653076171875,
      "memory(GiB)": 77.56,
      "step": 47470,
      "token_acc": 0.5337837837837838,
      "train_speed(iter/s)": 1.440073
    },
    {
      "epoch": 2.0339745512188854,
      "grad_norm": 3.8732073307037354,
      "learning_rate": 6.443230399628916e-05,
      "loss": 2.36053409576416,
      "memory(GiB)": 77.56,
      "step": 47475,
      "token_acc": 0.48360655737704916,
      "train_speed(iter/s)": 1.44005
    },
    {
      "epoch": 2.034188766548134,
      "grad_norm": 5.332747459411621,
      "learning_rate": 6.442586053948576e-05,
      "loss": 2.2015209197998047,
      "memory(GiB)": 77.56,
      "step": 47480,
      "token_acc": 0.5241635687732342,
      "train_speed(iter/s)": 1.440057
    },
    {
      "epoch": 2.034402981877383,
      "grad_norm": 6.69398832321167,
      "learning_rate": 6.441941682134436e-05,
      "loss": 2.2048450469970704,
      "memory(GiB)": 77.56,
      "step": 47485,
      "token_acc": 0.545816733067729,
      "train_speed(iter/s)": 1.440108
    },
    {
      "epoch": 2.0346171972066323,
      "grad_norm": 5.170960903167725,
      "learning_rate": 6.441297284198172e-05,
      "loss": 2.4763031005859375,
      "memory(GiB)": 77.56,
      "step": 47490,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.440151
    },
    {
      "epoch": 2.034831412535881,
      "grad_norm": 5.75846529006958,
      "learning_rate": 6.440652860151452e-05,
      "loss": 2.4016122817993164,
      "memory(GiB)": 77.56,
      "step": 47495,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.440143
    },
    {
      "epoch": 2.03504562786513,
      "grad_norm": 6.008525371551514,
      "learning_rate": 6.440008410005956e-05,
      "loss": 2.5113031387329103,
      "memory(GiB)": 77.56,
      "step": 47500,
      "token_acc": 0.44516129032258067,
      "train_speed(iter/s)": 1.440184
    },
    {
      "epoch": 2.03504562786513,
      "eval_loss": 2.163577079772949,
      "eval_runtime": 14.1428,
      "eval_samples_per_second": 7.071,
      "eval_steps_per_second": 7.071,
      "eval_token_acc": 0.46774193548387094,
      "step": 47500
    },
    {
      "epoch": 2.035259843194379,
      "grad_norm": 4.006160736083984,
      "learning_rate": 6.439363933773358e-05,
      "loss": 2.2517932891845702,
      "memory(GiB)": 77.56,
      "step": 47505,
      "token_acc": 0.474025974025974,
      "train_speed(iter/s)": 1.43954
    },
    {
      "epoch": 2.035474058523628,
      "grad_norm": 5.914662837982178,
      "learning_rate": 6.43871943146533e-05,
      "loss": 2.436318588256836,
      "memory(GiB)": 77.56,
      "step": 47510,
      "token_acc": 0.46647230320699706,
      "train_speed(iter/s)": 1.439531
    },
    {
      "epoch": 2.0356882738528768,
      "grad_norm": 6.909361839294434,
      "learning_rate": 6.43807490309355e-05,
      "loss": 2.337209701538086,
      "memory(GiB)": 77.56,
      "step": 47515,
      "token_acc": 0.525691699604743,
      "train_speed(iter/s)": 1.439505
    },
    {
      "epoch": 2.035902489182126,
      "grad_norm": 4.693491458892822,
      "learning_rate": 6.437430348669695e-05,
      "loss": 2.4208454132080077,
      "memory(GiB)": 77.56,
      "step": 47520,
      "token_acc": 0.4568345323741007,
      "train_speed(iter/s)": 1.439497
    },
    {
      "epoch": 2.036116704511375,
      "grad_norm": 4.709084510803223,
      "learning_rate": 6.436785768205442e-05,
      "loss": 2.5848079681396485,
      "memory(GiB)": 77.56,
      "step": 47525,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.439499
    },
    {
      "epoch": 2.0363309198406236,
      "grad_norm": 5.230836391448975,
      "learning_rate": 6.436141161712465e-05,
      "loss": 2.376507568359375,
      "memory(GiB)": 77.56,
      "step": 47530,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.439485
    },
    {
      "epoch": 2.036545135169873,
      "grad_norm": 4.986629962921143,
      "learning_rate": 6.435496529202446e-05,
      "loss": 2.282478904724121,
      "memory(GiB)": 77.56,
      "step": 47535,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.439509
    },
    {
      "epoch": 2.0367593504991217,
      "grad_norm": 5.57295560836792,
      "learning_rate": 6.434851870687058e-05,
      "loss": 2.5561681747436524,
      "memory(GiB)": 77.56,
      "step": 47540,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.439482
    },
    {
      "epoch": 2.0369735658283705,
      "grad_norm": 4.689891338348389,
      "learning_rate": 6.434207186177982e-05,
      "loss": 2.5525936126708983,
      "memory(GiB)": 77.56,
      "step": 47545,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.439512
    },
    {
      "epoch": 2.03718778115762,
      "grad_norm": 4.075557231903076,
      "learning_rate": 6.4335624756869e-05,
      "loss": 2.3793817520141602,
      "memory(GiB)": 77.56,
      "step": 47550,
      "token_acc": 0.5286195286195287,
      "train_speed(iter/s)": 1.439511
    },
    {
      "epoch": 2.0374019964868686,
      "grad_norm": 4.648125648498535,
      "learning_rate": 6.432917739225486e-05,
      "loss": 2.1042709350585938,
      "memory(GiB)": 77.56,
      "step": 47555,
      "token_acc": 0.5546558704453441,
      "train_speed(iter/s)": 1.439502
    },
    {
      "epoch": 2.0376162118161174,
      "grad_norm": 4.895267009735107,
      "learning_rate": 6.432272976805425e-05,
      "loss": 2.394881820678711,
      "memory(GiB)": 77.56,
      "step": 47560,
      "token_acc": 0.486013986013986,
      "train_speed(iter/s)": 1.439491
    },
    {
      "epoch": 2.0378304271453667,
      "grad_norm": 4.482232093811035,
      "learning_rate": 6.431628188438397e-05,
      "loss": 2.2670919418334963,
      "memory(GiB)": 77.56,
      "step": 47565,
      "token_acc": 0.5311203319502075,
      "train_speed(iter/s)": 1.439503
    },
    {
      "epoch": 2.0380446424746155,
      "grad_norm": 4.007987976074219,
      "learning_rate": 6.430983374136078e-05,
      "loss": 2.51898307800293,
      "memory(GiB)": 77.56,
      "step": 47570,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.439537
    },
    {
      "epoch": 2.0382588578038643,
      "grad_norm": 5.164368152618408,
      "learning_rate": 6.430338533910155e-05,
      "loss": 2.272294044494629,
      "memory(GiB)": 77.56,
      "step": 47575,
      "token_acc": 0.4962962962962963,
      "train_speed(iter/s)": 1.439562
    },
    {
      "epoch": 2.0384730731331135,
      "grad_norm": 4.738786220550537,
      "learning_rate": 6.429693667772308e-05,
      "loss": 2.6083396911621093,
      "memory(GiB)": 77.56,
      "step": 47580,
      "token_acc": 0.4585635359116022,
      "train_speed(iter/s)": 1.439587
    },
    {
      "epoch": 2.0386872884623624,
      "grad_norm": 6.589414596557617,
      "learning_rate": 6.429048775734216e-05,
      "loss": 2.5532367706298826,
      "memory(GiB)": 77.56,
      "step": 47585,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.439585
    },
    {
      "epoch": 2.038901503791611,
      "grad_norm": 5.052786827087402,
      "learning_rate": 6.428403857807568e-05,
      "loss": 2.48738956451416,
      "memory(GiB)": 77.56,
      "step": 47590,
      "token_acc": 0.4646153846153846,
      "train_speed(iter/s)": 1.439579
    },
    {
      "epoch": 2.0391157191208604,
      "grad_norm": 5.194866180419922,
      "learning_rate": 6.427758914004044e-05,
      "loss": 2.323424530029297,
      "memory(GiB)": 77.56,
      "step": 47595,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439577
    },
    {
      "epoch": 2.0393299344501092,
      "grad_norm": 4.9435834884643555,
      "learning_rate": 6.427113944335326e-05,
      "loss": 2.5700027465820314,
      "memory(GiB)": 77.56,
      "step": 47600,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.439584
    },
    {
      "epoch": 2.039544149779358,
      "grad_norm": 4.893200874328613,
      "learning_rate": 6.426468948813102e-05,
      "loss": 2.213448905944824,
      "memory(GiB)": 77.56,
      "step": 47605,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.439617
    },
    {
      "epoch": 2.0397583651086073,
      "grad_norm": 5.314486503601074,
      "learning_rate": 6.425823927449055e-05,
      "loss": 2.1358619689941407,
      "memory(GiB)": 77.56,
      "step": 47610,
      "token_acc": 0.5149501661129569,
      "train_speed(iter/s)": 1.439604
    },
    {
      "epoch": 2.039972580437856,
      "grad_norm": 5.000269412994385,
      "learning_rate": 6.425178880254869e-05,
      "loss": 2.2250171661376954,
      "memory(GiB)": 77.56,
      "step": 47615,
      "token_acc": 0.4871060171919771,
      "train_speed(iter/s)": 1.439622
    },
    {
      "epoch": 2.040186795767105,
      "grad_norm": 4.7385454177856445,
      "learning_rate": 6.424533807242229e-05,
      "loss": 2.5986610412597657,
      "memory(GiB)": 77.56,
      "step": 47620,
      "token_acc": 0.48632218844984804,
      "train_speed(iter/s)": 1.439641
    },
    {
      "epoch": 2.040401011096354,
      "grad_norm": 5.4596266746521,
      "learning_rate": 6.423888708422825e-05,
      "loss": 2.2574934005737304,
      "memory(GiB)": 77.56,
      "step": 47625,
      "token_acc": 0.5451127819548872,
      "train_speed(iter/s)": 1.439624
    },
    {
      "epoch": 2.040615226425603,
      "grad_norm": 4.448336124420166,
      "learning_rate": 6.42324358380834e-05,
      "loss": 2.230917549133301,
      "memory(GiB)": 77.56,
      "step": 47630,
      "token_acc": 0.4847457627118644,
      "train_speed(iter/s)": 1.439644
    },
    {
      "epoch": 2.040829441754852,
      "grad_norm": 7.835684776306152,
      "learning_rate": 6.422598433410465e-05,
      "loss": 2.2535871505737304,
      "memory(GiB)": 77.56,
      "step": 47635,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.439631
    },
    {
      "epoch": 2.041043657084101,
      "grad_norm": 7.610886096954346,
      "learning_rate": 6.421953257240882e-05,
      "loss": 2.476526641845703,
      "memory(GiB)": 77.56,
      "step": 47640,
      "token_acc": 0.47509578544061304,
      "train_speed(iter/s)": 1.439638
    },
    {
      "epoch": 2.04125787241335,
      "grad_norm": 4.8699445724487305,
      "learning_rate": 6.421308055311284e-05,
      "loss": 2.4349857330322267,
      "memory(GiB)": 77.56,
      "step": 47645,
      "token_acc": 0.5050847457627119,
      "train_speed(iter/s)": 1.439652
    },
    {
      "epoch": 2.0414720877425987,
      "grad_norm": 4.174794673919678,
      "learning_rate": 6.420662827633358e-05,
      "loss": 2.4949317932128907,
      "memory(GiB)": 77.56,
      "step": 47650,
      "token_acc": 0.49221183800623053,
      "train_speed(iter/s)": 1.439645
    },
    {
      "epoch": 2.041686303071848,
      "grad_norm": 5.558699131011963,
      "learning_rate": 6.420017574218788e-05,
      "loss": 2.332351875305176,
      "memory(GiB)": 77.56,
      "step": 47655,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.439644
    },
    {
      "epoch": 2.0419005184010968,
      "grad_norm": 8.018543243408203,
      "learning_rate": 6.41937229507927e-05,
      "loss": 2.3778156280517577,
      "memory(GiB)": 77.56,
      "step": 47660,
      "token_acc": 0.509090909090909,
      "train_speed(iter/s)": 1.439623
    },
    {
      "epoch": 2.0421147337303456,
      "grad_norm": 4.871335506439209,
      "learning_rate": 6.418726990226491e-05,
      "loss": 2.40634708404541,
      "memory(GiB)": 77.56,
      "step": 47665,
      "token_acc": 0.5510204081632653,
      "train_speed(iter/s)": 1.439667
    },
    {
      "epoch": 2.042328949059595,
      "grad_norm": 5.5760369300842285,
      "learning_rate": 6.418081659672142e-05,
      "loss": 2.422725868225098,
      "memory(GiB)": 77.56,
      "step": 47670,
      "token_acc": 0.4868913857677903,
      "train_speed(iter/s)": 1.439702
    },
    {
      "epoch": 2.0425431643888436,
      "grad_norm": 6.696816444396973,
      "learning_rate": 6.417436303427914e-05,
      "loss": 2.6715274810791017,
      "memory(GiB)": 77.56,
      "step": 47675,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.439727
    },
    {
      "epoch": 2.0427573797180925,
      "grad_norm": 4.811416149139404,
      "learning_rate": 6.416790921505498e-05,
      "loss": 2.5192203521728516,
      "memory(GiB)": 77.56,
      "step": 47680,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.439735
    },
    {
      "epoch": 2.0429715950473417,
      "grad_norm": 5.809756755828857,
      "learning_rate": 6.416145513916584e-05,
      "loss": 2.5253982543945312,
      "memory(GiB)": 77.56,
      "step": 47685,
      "token_acc": 0.44126984126984126,
      "train_speed(iter/s)": 1.439759
    },
    {
      "epoch": 2.0431858103765905,
      "grad_norm": 6.040470123291016,
      "learning_rate": 6.415500080672866e-05,
      "loss": 2.423827362060547,
      "memory(GiB)": 77.56,
      "step": 47690,
      "token_acc": 0.4582210242587601,
      "train_speed(iter/s)": 1.439735
    },
    {
      "epoch": 2.0434000257058393,
      "grad_norm": 4.674708843231201,
      "learning_rate": 6.414854621786037e-05,
      "loss": 2.4663007736206053,
      "memory(GiB)": 77.56,
      "step": 47695,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.439764
    },
    {
      "epoch": 2.0436142410350886,
      "grad_norm": 5.48889684677124,
      "learning_rate": 6.414209137267788e-05,
      "loss": 2.5493513107299806,
      "memory(GiB)": 77.56,
      "step": 47700,
      "token_acc": 0.4437299035369775,
      "train_speed(iter/s)": 1.439794
    },
    {
      "epoch": 2.0438284563643374,
      "grad_norm": 4.745237350463867,
      "learning_rate": 6.413563627129815e-05,
      "loss": 2.6456165313720703,
      "memory(GiB)": 77.56,
      "step": 47705,
      "token_acc": 0.4469914040114613,
      "train_speed(iter/s)": 1.439783
    },
    {
      "epoch": 2.044042671693586,
      "grad_norm": 4.328041076660156,
      "learning_rate": 6.41291809138381e-05,
      "loss": 2.3519237518310545,
      "memory(GiB)": 77.56,
      "step": 47710,
      "token_acc": 0.5031645569620253,
      "train_speed(iter/s)": 1.439789
    },
    {
      "epoch": 2.0442568870228355,
      "grad_norm": 5.482819557189941,
      "learning_rate": 6.412272530041469e-05,
      "loss": 2.2070241928100587,
      "memory(GiB)": 77.56,
      "step": 47715,
      "token_acc": 0.5038759689922481,
      "train_speed(iter/s)": 1.439748
    },
    {
      "epoch": 2.0444711023520843,
      "grad_norm": 4.37467622756958,
      "learning_rate": 6.411626943114486e-05,
      "loss": 2.417240524291992,
      "memory(GiB)": 77.56,
      "step": 47720,
      "token_acc": 0.47041420118343197,
      "train_speed(iter/s)": 1.439758
    },
    {
      "epoch": 2.044685317681333,
      "grad_norm": 4.715269088745117,
      "learning_rate": 6.410981330614558e-05,
      "loss": 2.2460214614868166,
      "memory(GiB)": 77.56,
      "step": 47725,
      "token_acc": 0.5653846153846154,
      "train_speed(iter/s)": 1.43977
    },
    {
      "epoch": 2.0448995330105824,
      "grad_norm": 4.7994890213012695,
      "learning_rate": 6.410335692553376e-05,
      "loss": 2.4280099868774414,
      "memory(GiB)": 77.56,
      "step": 47730,
      "token_acc": 0.4744525547445255,
      "train_speed(iter/s)": 1.439776
    },
    {
      "epoch": 2.045113748339831,
      "grad_norm": 7.1618170738220215,
      "learning_rate": 6.409690028942643e-05,
      "loss": 2.367707443237305,
      "memory(GiB)": 77.56,
      "step": 47735,
      "token_acc": 0.5282392026578073,
      "train_speed(iter/s)": 1.439705
    },
    {
      "epoch": 2.04532796366908,
      "grad_norm": 4.732382774353027,
      "learning_rate": 6.409044339794052e-05,
      "loss": 2.410608100891113,
      "memory(GiB)": 77.56,
      "step": 47740,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.439682
    },
    {
      "epoch": 2.0455421789983292,
      "grad_norm": 4.81704568862915,
      "learning_rate": 6.408398625119301e-05,
      "loss": 2.5306934356689452,
      "memory(GiB)": 77.56,
      "step": 47745,
      "token_acc": 0.4707792207792208,
      "train_speed(iter/s)": 1.439695
    },
    {
      "epoch": 2.045756394327578,
      "grad_norm": 4.810451984405518,
      "learning_rate": 6.407752884930089e-05,
      "loss": 2.3730548858642577,
      "memory(GiB)": 77.56,
      "step": 47750,
      "token_acc": 0.4984423676012461,
      "train_speed(iter/s)": 1.439709
    },
    {
      "epoch": 2.045970609656827,
      "grad_norm": 4.718359470367432,
      "learning_rate": 6.407107119238111e-05,
      "loss": 2.451678466796875,
      "memory(GiB)": 77.56,
      "step": 47755,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.439742
    },
    {
      "epoch": 2.046184824986076,
      "grad_norm": 3.8796544075012207,
      "learning_rate": 6.40646132805507e-05,
      "loss": 2.5709644317626954,
      "memory(GiB)": 77.56,
      "step": 47760,
      "token_acc": 0.45625,
      "train_speed(iter/s)": 1.439702
    },
    {
      "epoch": 2.046399040315325,
      "grad_norm": 6.005451679229736,
      "learning_rate": 6.405815511392659e-05,
      "loss": 2.2278003692626953,
      "memory(GiB)": 77.56,
      "step": 47765,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.439735
    },
    {
      "epoch": 2.0466132556445737,
      "grad_norm": 5.36748743057251,
      "learning_rate": 6.405169669262583e-05,
      "loss": 2.668032646179199,
      "memory(GiB)": 77.56,
      "step": 47770,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.439726
    },
    {
      "epoch": 2.046827470973823,
      "grad_norm": 5.313390254974365,
      "learning_rate": 6.40452380167654e-05,
      "loss": 2.3970668792724608,
      "memory(GiB)": 77.56,
      "step": 47775,
      "token_acc": 0.5369649805447471,
      "train_speed(iter/s)": 1.439719
    },
    {
      "epoch": 2.047041686303072,
      "grad_norm": 4.933276653289795,
      "learning_rate": 6.403877908646232e-05,
      "loss": 2.3330368041992187,
      "memory(GiB)": 77.56,
      "step": 47780,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.439697
    },
    {
      "epoch": 2.0472559016323206,
      "grad_norm": 4.451749324798584,
      "learning_rate": 6.403231990183358e-05,
      "loss": 2.3309856414794923,
      "memory(GiB)": 77.56,
      "step": 47785,
      "token_acc": 0.494475138121547,
      "train_speed(iter/s)": 1.439704
    },
    {
      "epoch": 2.04747011696157,
      "grad_norm": 4.890246868133545,
      "learning_rate": 6.40258604629962e-05,
      "loss": 2.3902399063110353,
      "memory(GiB)": 77.56,
      "step": 47790,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.439678
    },
    {
      "epoch": 2.0476843322908187,
      "grad_norm": 6.390333652496338,
      "learning_rate": 6.401940077006721e-05,
      "loss": 2.5157808303833007,
      "memory(GiB)": 77.56,
      "step": 47795,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.439674
    },
    {
      "epoch": 2.0478985476200675,
      "grad_norm": 5.395419120788574,
      "learning_rate": 6.401294082316361e-05,
      "loss": 2.6448734283447264,
      "memory(GiB)": 77.56,
      "step": 47800,
      "token_acc": 0.46944444444444444,
      "train_speed(iter/s)": 1.439676
    },
    {
      "epoch": 2.0481127629493168,
      "grad_norm": 6.820577621459961,
      "learning_rate": 6.400648062240243e-05,
      "loss": 2.202439308166504,
      "memory(GiB)": 77.56,
      "step": 47805,
      "token_acc": 0.5442622950819672,
      "train_speed(iter/s)": 1.439688
    },
    {
      "epoch": 2.0483269782785656,
      "grad_norm": 5.206563472747803,
      "learning_rate": 6.400002016790074e-05,
      "loss": 2.2871850967407226,
      "memory(GiB)": 77.56,
      "step": 47810,
      "token_acc": 0.4984709480122324,
      "train_speed(iter/s)": 1.439684
    },
    {
      "epoch": 2.0485411936078144,
      "grad_norm": 5.819280624389648,
      "learning_rate": 6.399355945977554e-05,
      "loss": 2.2564260482788088,
      "memory(GiB)": 77.56,
      "step": 47815,
      "token_acc": 0.53,
      "train_speed(iter/s)": 1.439673
    },
    {
      "epoch": 2.0487554089370636,
      "grad_norm": 5.17987585067749,
      "learning_rate": 6.39870984981439e-05,
      "loss": 2.349131965637207,
      "memory(GiB)": 77.56,
      "step": 47820,
      "token_acc": 0.5311355311355311,
      "train_speed(iter/s)": 1.439713
    },
    {
      "epoch": 2.0489696242663125,
      "grad_norm": 5.169327735900879,
      "learning_rate": 6.398063728312283e-05,
      "loss": 2.422494888305664,
      "memory(GiB)": 77.56,
      "step": 47825,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.439681
    },
    {
      "epoch": 2.0491838395955613,
      "grad_norm": 8.17672348022461,
      "learning_rate": 6.39741758148294e-05,
      "loss": 2.0970438003540037,
      "memory(GiB)": 77.56,
      "step": 47830,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.439668
    },
    {
      "epoch": 2.0493980549248105,
      "grad_norm": 5.379665851593018,
      "learning_rate": 6.396771409338068e-05,
      "loss": 2.691663360595703,
      "memory(GiB)": 77.56,
      "step": 47835,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.439681
    },
    {
      "epoch": 2.0496122702540593,
      "grad_norm": 6.921357154846191,
      "learning_rate": 6.39612521188937e-05,
      "loss": 2.5036855697631837,
      "memory(GiB)": 77.56,
      "step": 47840,
      "token_acc": 0.49337748344370863,
      "train_speed(iter/s)": 1.439695
    },
    {
      "epoch": 2.049826485583308,
      "grad_norm": 7.54922342300415,
      "learning_rate": 6.395478989148556e-05,
      "loss": 2.3679580688476562,
      "memory(GiB)": 77.56,
      "step": 47845,
      "token_acc": 0.47633136094674555,
      "train_speed(iter/s)": 1.439655
    },
    {
      "epoch": 2.0500407009125574,
      "grad_norm": 5.228775501251221,
      "learning_rate": 6.39483274112733e-05,
      "loss": 2.2357929229736326,
      "memory(GiB)": 77.56,
      "step": 47850,
      "token_acc": 0.5232974910394266,
      "train_speed(iter/s)": 1.439684
    },
    {
      "epoch": 2.0502549162418062,
      "grad_norm": 4.800392150878906,
      "learning_rate": 6.3941864678374e-05,
      "loss": 2.427528953552246,
      "memory(GiB)": 77.56,
      "step": 47855,
      "token_acc": 0.46827794561933533,
      "train_speed(iter/s)": 1.439704
    },
    {
      "epoch": 2.050469131571055,
      "grad_norm": 4.886074066162109,
      "learning_rate": 6.393540169290475e-05,
      "loss": 2.254031181335449,
      "memory(GiB)": 77.56,
      "step": 47860,
      "token_acc": 0.5374149659863946,
      "train_speed(iter/s)": 1.439692
    },
    {
      "epoch": 2.0506833469003043,
      "grad_norm": 5.674720287322998,
      "learning_rate": 6.392893845498262e-05,
      "loss": 2.5538101196289062,
      "memory(GiB)": 77.56,
      "step": 47865,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.439687
    },
    {
      "epoch": 2.050897562229553,
      "grad_norm": 6.776648998260498,
      "learning_rate": 6.392247496472472e-05,
      "loss": 2.4056728363037108,
      "memory(GiB)": 77.56,
      "step": 47870,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.439671
    },
    {
      "epoch": 2.051111777558802,
      "grad_norm": 7.172952175140381,
      "learning_rate": 6.391601122224811e-05,
      "loss": 2.341914939880371,
      "memory(GiB)": 77.56,
      "step": 47875,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.439658
    },
    {
      "epoch": 2.051325992888051,
      "grad_norm": 6.251924514770508,
      "learning_rate": 6.39095472276699e-05,
      "loss": 2.646720123291016,
      "memory(GiB)": 77.56,
      "step": 47880,
      "token_acc": 0.4492307692307692,
      "train_speed(iter/s)": 1.439682
    },
    {
      "epoch": 2.0515402082173,
      "grad_norm": 4.300154685974121,
      "learning_rate": 6.39030829811072e-05,
      "loss": 2.4142295837402346,
      "memory(GiB)": 77.56,
      "step": 47885,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.439722
    },
    {
      "epoch": 2.051754423546549,
      "grad_norm": 5.152960300445557,
      "learning_rate": 6.38966184826771e-05,
      "loss": 2.3202417373657225,
      "memory(GiB)": 77.56,
      "step": 47890,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.439708
    },
    {
      "epoch": 2.051968638875798,
      "grad_norm": 7.4465179443359375,
      "learning_rate": 6.389015373249674e-05,
      "loss": 2.7647781372070312,
      "memory(GiB)": 77.56,
      "step": 47895,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.439733
    },
    {
      "epoch": 2.052182854205047,
      "grad_norm": 5.134638786315918,
      "learning_rate": 6.388368873068321e-05,
      "loss": 2.5917724609375,
      "memory(GiB)": 77.56,
      "step": 47900,
      "token_acc": 0.4820846905537459,
      "train_speed(iter/s)": 1.439756
    },
    {
      "epoch": 2.0523970695342957,
      "grad_norm": 4.606905937194824,
      "learning_rate": 6.387722347735362e-05,
      "loss": 2.1808450698852537,
      "memory(GiB)": 77.56,
      "step": 47905,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.439783
    },
    {
      "epoch": 2.052611284863545,
      "grad_norm": 5.044373035430908,
      "learning_rate": 6.387075797262513e-05,
      "loss": 2.676270294189453,
      "memory(GiB)": 77.56,
      "step": 47910,
      "token_acc": 0.44074074074074077,
      "train_speed(iter/s)": 1.439789
    },
    {
      "epoch": 2.0528255001927938,
      "grad_norm": 5.0664591789245605,
      "learning_rate": 6.386429221661483e-05,
      "loss": 2.8006845474243165,
      "memory(GiB)": 77.56,
      "step": 47915,
      "token_acc": 0.4318936877076412,
      "train_speed(iter/s)": 1.439793
    },
    {
      "epoch": 2.0530397155220426,
      "grad_norm": 7.783517360687256,
      "learning_rate": 6.385782620943986e-05,
      "loss": 2.3821617126464845,
      "memory(GiB)": 77.56,
      "step": 47920,
      "token_acc": 0.5106382978723404,
      "train_speed(iter/s)": 1.439799
    },
    {
      "epoch": 2.053253930851292,
      "grad_norm": 4.918902397155762,
      "learning_rate": 6.385135995121739e-05,
      "loss": 2.245688629150391,
      "memory(GiB)": 77.56,
      "step": 47925,
      "token_acc": 0.5222929936305732,
      "train_speed(iter/s)": 1.439806
    },
    {
      "epoch": 2.0534681461805406,
      "grad_norm": 5.0280351638793945,
      "learning_rate": 6.384489344206455e-05,
      "loss": 2.2529088973999025,
      "memory(GiB)": 77.56,
      "step": 47930,
      "token_acc": 0.5164473684210527,
      "train_speed(iter/s)": 1.43981
    },
    {
      "epoch": 2.0536823615097894,
      "grad_norm": 5.787678241729736,
      "learning_rate": 6.383842668209845e-05,
      "loss": 2.4657352447509764,
      "memory(GiB)": 77.56,
      "step": 47935,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.439805
    },
    {
      "epoch": 2.0538965768390387,
      "grad_norm": 5.114063262939453,
      "learning_rate": 6.383195967143629e-05,
      "loss": 2.274463081359863,
      "memory(GiB)": 77.56,
      "step": 47940,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.43982
    },
    {
      "epoch": 2.0541107921682875,
      "grad_norm": 5.533620834350586,
      "learning_rate": 6.38254924101952e-05,
      "loss": 2.320192337036133,
      "memory(GiB)": 77.56,
      "step": 47945,
      "token_acc": 0.5278969957081545,
      "train_speed(iter/s)": 1.439854
    },
    {
      "epoch": 2.0543250074975363,
      "grad_norm": 4.123553276062012,
      "learning_rate": 6.381902489849233e-05,
      "loss": 2.285234832763672,
      "memory(GiB)": 77.56,
      "step": 47950,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.439856
    },
    {
      "epoch": 2.0545392228267856,
      "grad_norm": 8.138646125793457,
      "learning_rate": 6.381255713644488e-05,
      "loss": 2.4138904571533204,
      "memory(GiB)": 77.56,
      "step": 47955,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.439861
    },
    {
      "epoch": 2.0547534381560344,
      "grad_norm": 4.103034973144531,
      "learning_rate": 6.380608912416999e-05,
      "loss": 2.406050682067871,
      "memory(GiB)": 77.56,
      "step": 47960,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.439842
    },
    {
      "epoch": 2.054967653485283,
      "grad_norm": 5.271156311035156,
      "learning_rate": 6.379962086178485e-05,
      "loss": 2.463789176940918,
      "memory(GiB)": 77.56,
      "step": 47965,
      "token_acc": 0.49508196721311476,
      "train_speed(iter/s)": 1.439869
    },
    {
      "epoch": 2.0551818688145325,
      "grad_norm": 5.393589019775391,
      "learning_rate": 6.379315234940664e-05,
      "loss": 2.263259696960449,
      "memory(GiB)": 77.56,
      "step": 47970,
      "token_acc": 0.531986531986532,
      "train_speed(iter/s)": 1.439865
    },
    {
      "epoch": 2.0553960841437813,
      "grad_norm": 7.665388584136963,
      "learning_rate": 6.378668358715252e-05,
      "loss": 2.3992361068725585,
      "memory(GiB)": 77.56,
      "step": 47975,
      "token_acc": 0.4983277591973244,
      "train_speed(iter/s)": 1.439877
    },
    {
      "epoch": 2.05561029947303,
      "grad_norm": 6.8578338623046875,
      "learning_rate": 6.378021457513971e-05,
      "loss": 2.3825326919555665,
      "memory(GiB)": 77.56,
      "step": 47980,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.439876
    },
    {
      "epoch": 2.0558245148022793,
      "grad_norm": 4.834499359130859,
      "learning_rate": 6.377374531348538e-05,
      "loss": 2.1360877990722655,
      "memory(GiB)": 77.56,
      "step": 47985,
      "token_acc": 0.531986531986532,
      "train_speed(iter/s)": 1.439868
    },
    {
      "epoch": 2.056038730131528,
      "grad_norm": 4.213587760925293,
      "learning_rate": 6.376727580230671e-05,
      "loss": 2.2522083282470704,
      "memory(GiB)": 77.56,
      "step": 47990,
      "token_acc": 0.5375375375375375,
      "train_speed(iter/s)": 1.439845
    },
    {
      "epoch": 2.056252945460777,
      "grad_norm": 7.45883846282959,
      "learning_rate": 6.376080604172096e-05,
      "loss": 2.5970041275024416,
      "memory(GiB)": 77.56,
      "step": 47995,
      "token_acc": 0.4280821917808219,
      "train_speed(iter/s)": 1.439829
    },
    {
      "epoch": 2.0564671607900262,
      "grad_norm": 4.569416522979736,
      "learning_rate": 6.375433603184528e-05,
      "loss": 2.4430402755737304,
      "memory(GiB)": 77.56,
      "step": 48000,
      "token_acc": 0.4781420765027322,
      "train_speed(iter/s)": 1.439867
    },
    {
      "epoch": 2.0564671607900262,
      "eval_loss": 2.2916102409362793,
      "eval_runtime": 13.9451,
      "eval_samples_per_second": 7.171,
      "eval_steps_per_second": 7.171,
      "eval_token_acc": 0.45955882352941174,
      "step": 48000
    },
    {
      "epoch": 2.056681376119275,
      "grad_norm": 4.693818092346191,
      "learning_rate": 6.37478657727969e-05,
      "loss": 2.2714290618896484,
      "memory(GiB)": 77.56,
      "step": 48005,
      "token_acc": 0.4605377276669558,
      "train_speed(iter/s)": 1.439278
    },
    {
      "epoch": 2.056895591448524,
      "grad_norm": 5.718568325042725,
      "learning_rate": 6.374139526469304e-05,
      "loss": 2.1184856414794924,
      "memory(GiB)": 77.56,
      "step": 48010,
      "token_acc": 0.5287356321839081,
      "train_speed(iter/s)": 1.439207
    },
    {
      "epoch": 2.057109806777773,
      "grad_norm": 6.2125935554504395,
      "learning_rate": 6.373492450765093e-05,
      "loss": 2.271101951599121,
      "memory(GiB)": 77.56,
      "step": 48015,
      "token_acc": 0.5131195335276968,
      "train_speed(iter/s)": 1.439228
    },
    {
      "epoch": 2.057324022107022,
      "grad_norm": 5.286134719848633,
      "learning_rate": 6.372845350178776e-05,
      "loss": 2.4091039657592774,
      "memory(GiB)": 77.56,
      "step": 48020,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.439274
    },
    {
      "epoch": 2.0575382374362707,
      "grad_norm": 5.216477870941162,
      "learning_rate": 6.372198224722077e-05,
      "loss": 2.2434417724609377,
      "memory(GiB)": 77.56,
      "step": 48025,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.439289
    },
    {
      "epoch": 2.05775245276552,
      "grad_norm": 4.990546226501465,
      "learning_rate": 6.37155107440672e-05,
      "loss": 2.143111801147461,
      "memory(GiB)": 77.56,
      "step": 48030,
      "token_acc": 0.5601659751037344,
      "train_speed(iter/s)": 1.4393
    },
    {
      "epoch": 2.057966668094769,
      "grad_norm": 4.992410182952881,
      "learning_rate": 6.370903899244429e-05,
      "loss": 2.544847869873047,
      "memory(GiB)": 77.56,
      "step": 48035,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.439272
    },
    {
      "epoch": 2.0581808834240176,
      "grad_norm": 5.420811653137207,
      "learning_rate": 6.370256699246929e-05,
      "loss": 2.363251304626465,
      "memory(GiB)": 77.56,
      "step": 48040,
      "token_acc": 0.4789156626506024,
      "train_speed(iter/s)": 1.439275
    },
    {
      "epoch": 2.058395098753267,
      "grad_norm": 4.947816848754883,
      "learning_rate": 6.369609474425944e-05,
      "loss": 2.317617416381836,
      "memory(GiB)": 77.56,
      "step": 48045,
      "token_acc": 0.5133333333333333,
      "train_speed(iter/s)": 1.439291
    },
    {
      "epoch": 2.0586093140825157,
      "grad_norm": 4.4387335777282715,
      "learning_rate": 6.368962224793195e-05,
      "loss": 2.631804275512695,
      "memory(GiB)": 77.56,
      "step": 48050,
      "token_acc": 0.4394904458598726,
      "train_speed(iter/s)": 1.439295
    },
    {
      "epoch": 2.0588235294117645,
      "grad_norm": 3.9364869594573975,
      "learning_rate": 6.368314950360415e-05,
      "loss": 2.2472118377685546,
      "memory(GiB)": 77.56,
      "step": 48055,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.439269
    },
    {
      "epoch": 2.0590377447410138,
      "grad_norm": 6.670133590698242,
      "learning_rate": 6.367667651139324e-05,
      "loss": 2.486860466003418,
      "memory(GiB)": 77.56,
      "step": 48060,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.439273
    },
    {
      "epoch": 2.0592519600702626,
      "grad_norm": 6.60884952545166,
      "learning_rate": 6.367020327141651e-05,
      "loss": 2.564601707458496,
      "memory(GiB)": 77.56,
      "step": 48065,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.439268
    },
    {
      "epoch": 2.0594661753995114,
      "grad_norm": 4.956416130065918,
      "learning_rate": 6.366372978379122e-05,
      "loss": 2.4545724868774412,
      "memory(GiB)": 77.56,
      "step": 48070,
      "token_acc": 0.4985507246376812,
      "train_speed(iter/s)": 1.439277
    },
    {
      "epoch": 2.0596803907287606,
      "grad_norm": 4.4853081703186035,
      "learning_rate": 6.365725604863466e-05,
      "loss": 2.598590850830078,
      "memory(GiB)": 77.56,
      "step": 48075,
      "token_acc": 0.5250836120401338,
      "train_speed(iter/s)": 1.439305
    },
    {
      "epoch": 2.0598946060580094,
      "grad_norm": 4.909699440002441,
      "learning_rate": 6.365078206606408e-05,
      "loss": 2.706365203857422,
      "memory(GiB)": 77.56,
      "step": 48080,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.439309
    },
    {
      "epoch": 2.0601088213872583,
      "grad_norm": 8.361443519592285,
      "learning_rate": 6.364430783619679e-05,
      "loss": 2.3414520263671874,
      "memory(GiB)": 77.56,
      "step": 48085,
      "token_acc": 0.48135593220338985,
      "train_speed(iter/s)": 1.43931
    },
    {
      "epoch": 2.0603230367165075,
      "grad_norm": 6.474850177764893,
      "learning_rate": 6.363783335915005e-05,
      "loss": 2.5392520904541014,
      "memory(GiB)": 77.56,
      "step": 48090,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.439341
    },
    {
      "epoch": 2.0605372520457563,
      "grad_norm": 6.0981011390686035,
      "learning_rate": 6.36313586350412e-05,
      "loss": 2.15423526763916,
      "memory(GiB)": 77.56,
      "step": 48095,
      "token_acc": 0.5335968379446641,
      "train_speed(iter/s)": 1.439349
    },
    {
      "epoch": 2.060751467375005,
      "grad_norm": 5.5894083976745605,
      "learning_rate": 6.362488366398746e-05,
      "loss": 2.670083999633789,
      "memory(GiB)": 77.56,
      "step": 48100,
      "token_acc": 0.4409448818897638,
      "train_speed(iter/s)": 1.439382
    },
    {
      "epoch": 2.0609656827042544,
      "grad_norm": 5.242283344268799,
      "learning_rate": 6.361840844610619e-05,
      "loss": 2.7428956985473634,
      "memory(GiB)": 77.56,
      "step": 48105,
      "token_acc": 0.4441340782122905,
      "train_speed(iter/s)": 1.43941
    },
    {
      "epoch": 2.061179898033503,
      "grad_norm": 6.626741409301758,
      "learning_rate": 6.361193298151466e-05,
      "loss": 2.204768180847168,
      "memory(GiB)": 77.56,
      "step": 48110,
      "token_acc": 0.562992125984252,
      "train_speed(iter/s)": 1.439376
    },
    {
      "epoch": 2.061394113362752,
      "grad_norm": 4.67233419418335,
      "learning_rate": 6.360545727033023e-05,
      "loss": 2.4921072006225584,
      "memory(GiB)": 77.56,
      "step": 48115,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.439373
    },
    {
      "epoch": 2.0616083286920013,
      "grad_norm": 5.033863067626953,
      "learning_rate": 6.359898131267014e-05,
      "loss": 2.411849784851074,
      "memory(GiB)": 77.56,
      "step": 48120,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.439405
    },
    {
      "epoch": 2.06182254402125,
      "grad_norm": 6.651470184326172,
      "learning_rate": 6.359250510865177e-05,
      "loss": 2.49924373626709,
      "memory(GiB)": 77.56,
      "step": 48125,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.439419
    },
    {
      "epoch": 2.062036759350499,
      "grad_norm": 5.903689861297607,
      "learning_rate": 6.358602865839242e-05,
      "loss": 2.346999740600586,
      "memory(GiB)": 77.56,
      "step": 48130,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.4394
    },
    {
      "epoch": 2.062250974679748,
      "grad_norm": 5.883366107940674,
      "learning_rate": 6.357955196200939e-05,
      "loss": 2.416558265686035,
      "memory(GiB)": 77.56,
      "step": 48135,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.439395
    },
    {
      "epoch": 2.062465190008997,
      "grad_norm": 6.490179061889648,
      "learning_rate": 6.357307501962007e-05,
      "loss": 2.1859272003173826,
      "memory(GiB)": 77.56,
      "step": 48140,
      "token_acc": 0.4894894894894895,
      "train_speed(iter/s)": 1.439415
    },
    {
      "epoch": 2.062679405338246,
      "grad_norm": 5.267295837402344,
      "learning_rate": 6.356659783134174e-05,
      "loss": 2.303095817565918,
      "memory(GiB)": 77.56,
      "step": 48145,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.439407
    },
    {
      "epoch": 2.062893620667495,
      "grad_norm": 4.725438594818115,
      "learning_rate": 6.356012039729177e-05,
      "loss": 2.503006362915039,
      "memory(GiB)": 77.56,
      "step": 48150,
      "token_acc": 0.48589341692789967,
      "train_speed(iter/s)": 1.439425
    },
    {
      "epoch": 2.063107835996744,
      "grad_norm": 5.783420085906982,
      "learning_rate": 6.35536427175875e-05,
      "loss": 2.253975486755371,
      "memory(GiB)": 77.56,
      "step": 48155,
      "token_acc": 0.5063291139240507,
      "train_speed(iter/s)": 1.439433
    },
    {
      "epoch": 2.0633220513259927,
      "grad_norm": 4.804093360900879,
      "learning_rate": 6.354716479234629e-05,
      "loss": 2.283300018310547,
      "memory(GiB)": 77.56,
      "step": 48160,
      "token_acc": 0.5410447761194029,
      "train_speed(iter/s)": 1.439453
    },
    {
      "epoch": 2.063536266655242,
      "grad_norm": 5.716576099395752,
      "learning_rate": 6.354068662168546e-05,
      "loss": 2.600575828552246,
      "memory(GiB)": 77.56,
      "step": 48165,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.439441
    },
    {
      "epoch": 2.0637504819844907,
      "grad_norm": 4.64011287689209,
      "learning_rate": 6.353420820572242e-05,
      "loss": 2.1770219802856445,
      "memory(GiB)": 77.56,
      "step": 48170,
      "token_acc": 0.5575539568345323,
      "train_speed(iter/s)": 1.439442
    },
    {
      "epoch": 2.0639646973137395,
      "grad_norm": 5.3839826583862305,
      "learning_rate": 6.352772954457447e-05,
      "loss": 2.416168975830078,
      "memory(GiB)": 77.56,
      "step": 48175,
      "token_acc": 0.49429657794676807,
      "train_speed(iter/s)": 1.439394
    },
    {
      "epoch": 2.064178912642989,
      "grad_norm": 6.550118923187256,
      "learning_rate": 6.3521250638359e-05,
      "loss": 2.33520393371582,
      "memory(GiB)": 77.56,
      "step": 48180,
      "token_acc": 0.5221843003412969,
      "train_speed(iter/s)": 1.439329
    },
    {
      "epoch": 2.0643931279722376,
      "grad_norm": 5.122165679931641,
      "learning_rate": 6.351477148719343e-05,
      "loss": 2.526004600524902,
      "memory(GiB)": 77.56,
      "step": 48185,
      "token_acc": 0.46647230320699706,
      "train_speed(iter/s)": 1.439329
    },
    {
      "epoch": 2.0646073433014864,
      "grad_norm": 5.075963497161865,
      "learning_rate": 6.350829209119509e-05,
      "loss": 2.615765380859375,
      "memory(GiB)": 77.56,
      "step": 48190,
      "token_acc": 0.44402985074626866,
      "train_speed(iter/s)": 1.439318
    },
    {
      "epoch": 2.0648215586307357,
      "grad_norm": 5.452728748321533,
      "learning_rate": 6.350181245048135e-05,
      "loss": 2.231444549560547,
      "memory(GiB)": 77.56,
      "step": 48195,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.43934
    },
    {
      "epoch": 2.0650357739599845,
      "grad_norm": 6.405764102935791,
      "learning_rate": 6.349533256516962e-05,
      "loss": 2.572401428222656,
      "memory(GiB)": 77.56,
      "step": 48200,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.439344
    },
    {
      "epoch": 2.0652499892892333,
      "grad_norm": 4.045449733734131,
      "learning_rate": 6.348885243537728e-05,
      "loss": 2.498097038269043,
      "memory(GiB)": 77.56,
      "step": 48205,
      "token_acc": 0.46258503401360546,
      "train_speed(iter/s)": 1.43935
    },
    {
      "epoch": 2.0654642046184826,
      "grad_norm": 5.261559963226318,
      "learning_rate": 6.348237206122172e-05,
      "loss": 2.444377899169922,
      "memory(GiB)": 77.56,
      "step": 48210,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.439381
    },
    {
      "epoch": 2.0656784199477314,
      "grad_norm": 5.384941577911377,
      "learning_rate": 6.347589144282034e-05,
      "loss": 2.262255096435547,
      "memory(GiB)": 77.56,
      "step": 48215,
      "token_acc": 0.5321428571428571,
      "train_speed(iter/s)": 1.439341
    },
    {
      "epoch": 2.06589263527698,
      "grad_norm": 7.680131435394287,
      "learning_rate": 6.346941058029055e-05,
      "loss": 2.614405059814453,
      "memory(GiB)": 77.56,
      "step": 48220,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439336
    },
    {
      "epoch": 2.0661068506062295,
      "grad_norm": 4.867196559906006,
      "learning_rate": 6.346292947374975e-05,
      "loss": 2.2359460830688476,
      "memory(GiB)": 77.56,
      "step": 48225,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.439385
    },
    {
      "epoch": 2.0663210659354783,
      "grad_norm": 5.427309513092041,
      "learning_rate": 6.345644812331538e-05,
      "loss": 2.4799144744873045,
      "memory(GiB)": 77.56,
      "step": 48230,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.439368
    },
    {
      "epoch": 2.066535281264727,
      "grad_norm": 5.110914707183838,
      "learning_rate": 6.344996652910479e-05,
      "loss": 2.3211389541625977,
      "memory(GiB)": 77.56,
      "step": 48235,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.439359
    },
    {
      "epoch": 2.0667494965939763,
      "grad_norm": 7.088987827301025,
      "learning_rate": 6.344348469123545e-05,
      "loss": 2.7160480499267576,
      "memory(GiB)": 77.56,
      "step": 48240,
      "token_acc": 0.47440273037542663,
      "train_speed(iter/s)": 1.43938
    },
    {
      "epoch": 2.066963711923225,
      "grad_norm": 4.280320644378662,
      "learning_rate": 6.343700260982479e-05,
      "loss": 2.2539031982421873,
      "memory(GiB)": 77.56,
      "step": 48245,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439394
    },
    {
      "epoch": 2.0671779272524744,
      "grad_norm": 4.63068962097168,
      "learning_rate": 6.343052028499019e-05,
      "loss": 2.3382373809814454,
      "memory(GiB)": 77.56,
      "step": 48250,
      "token_acc": 0.5347985347985348,
      "train_speed(iter/s)": 1.439388
    },
    {
      "epoch": 2.067392142581723,
      "grad_norm": 5.030412673950195,
      "learning_rate": 6.342403771684916e-05,
      "loss": 2.321604919433594,
      "memory(GiB)": 77.56,
      "step": 48255,
      "token_acc": 0.5037037037037037,
      "train_speed(iter/s)": 1.439361
    },
    {
      "epoch": 2.067606357910972,
      "grad_norm": 4.675687313079834,
      "learning_rate": 6.341755490551906e-05,
      "loss": 2.3080495834350585,
      "memory(GiB)": 77.56,
      "step": 48260,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.439355
    },
    {
      "epoch": 2.0678205732402213,
      "grad_norm": 5.396175861358643,
      "learning_rate": 6.341107185111737e-05,
      "loss": 2.2108844757080077,
      "memory(GiB)": 77.56,
      "step": 48265,
      "token_acc": 0.5570032573289903,
      "train_speed(iter/s)": 1.439359
    },
    {
      "epoch": 2.06803478856947,
      "grad_norm": 7.695731163024902,
      "learning_rate": 6.340458855376156e-05,
      "loss": 2.3261009216308595,
      "memory(GiB)": 77.56,
      "step": 48270,
      "token_acc": 0.5365853658536586,
      "train_speed(iter/s)": 1.439347
    },
    {
      "epoch": 2.068249003898719,
      "grad_norm": 5.517941474914551,
      "learning_rate": 6.339810501356903e-05,
      "loss": 2.0953414916992186,
      "memory(GiB)": 77.56,
      "step": 48275,
      "token_acc": 0.5743801652892562,
      "train_speed(iter/s)": 1.439317
    },
    {
      "epoch": 2.068463219227968,
      "grad_norm": 4.48099946975708,
      "learning_rate": 6.339162123065727e-05,
      "loss": 2.293553352355957,
      "memory(GiB)": 77.56,
      "step": 48280,
      "token_acc": 0.5328467153284672,
      "train_speed(iter/s)": 1.439318
    },
    {
      "epoch": 2.068677434557217,
      "grad_norm": 6.049262523651123,
      "learning_rate": 6.338513720514371e-05,
      "loss": 2.1654096603393556,
      "memory(GiB)": 77.56,
      "step": 48285,
      "token_acc": 0.5607142857142857,
      "train_speed(iter/s)": 1.439314
    },
    {
      "epoch": 2.068891649886466,
      "grad_norm": 5.270562171936035,
      "learning_rate": 6.337865293714584e-05,
      "loss": 2.644285774230957,
      "memory(GiB)": 77.56,
      "step": 48290,
      "token_acc": 0.45180722891566266,
      "train_speed(iter/s)": 1.439299
    },
    {
      "epoch": 2.069105865215715,
      "grad_norm": 5.95337438583374,
      "learning_rate": 6.337216842678114e-05,
      "loss": 2.506064224243164,
      "memory(GiB)": 77.56,
      "step": 48295,
      "token_acc": 0.4879032258064516,
      "train_speed(iter/s)": 1.439294
    },
    {
      "epoch": 2.069320080544964,
      "grad_norm": 5.2512993812561035,
      "learning_rate": 6.336568367416706e-05,
      "loss": 2.4757450103759764,
      "memory(GiB)": 77.56,
      "step": 48300,
      "token_acc": 0.47232472324723246,
      "train_speed(iter/s)": 1.439302
    },
    {
      "epoch": 2.0695342958742127,
      "grad_norm": 6.013294219970703,
      "learning_rate": 6.335919867942107e-05,
      "loss": 2.5989797592163084,
      "memory(GiB)": 77.56,
      "step": 48305,
      "token_acc": 0.44765342960288806,
      "train_speed(iter/s)": 1.4393
    },
    {
      "epoch": 2.069748511203462,
      "grad_norm": 5.916513919830322,
      "learning_rate": 6.335271344266066e-05,
      "loss": 2.4197723388671877,
      "memory(GiB)": 77.56,
      "step": 48310,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.439324
    },
    {
      "epoch": 2.0699627265327107,
      "grad_norm": 5.032732963562012,
      "learning_rate": 6.334622796400333e-05,
      "loss": 2.382965087890625,
      "memory(GiB)": 77.56,
      "step": 48315,
      "token_acc": 0.484,
      "train_speed(iter/s)": 1.439356
    },
    {
      "epoch": 2.0701769418619596,
      "grad_norm": 5.509790897369385,
      "learning_rate": 6.333974224356656e-05,
      "loss": 2.4743606567382814,
      "memory(GiB)": 77.56,
      "step": 48320,
      "token_acc": 0.4818181818181818,
      "train_speed(iter/s)": 1.439346
    },
    {
      "epoch": 2.070391157191209,
      "grad_norm": 6.152044296264648,
      "learning_rate": 6.333325628146785e-05,
      "loss": 2.601472854614258,
      "memory(GiB)": 77.56,
      "step": 48325,
      "token_acc": 0.4880952380952381,
      "train_speed(iter/s)": 1.439381
    },
    {
      "epoch": 2.0706053725204576,
      "grad_norm": 5.371792316436768,
      "learning_rate": 6.332677007782468e-05,
      "loss": 2.3854007720947266,
      "memory(GiB)": 77.56,
      "step": 48330,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.439348
    },
    {
      "epoch": 2.0708195878497064,
      "grad_norm": 6.3165740966796875,
      "learning_rate": 6.33202836327546e-05,
      "loss": 2.7836769104003904,
      "memory(GiB)": 77.56,
      "step": 48335,
      "token_acc": 0.4605809128630705,
      "train_speed(iter/s)": 1.439319
    },
    {
      "epoch": 2.0710338031789557,
      "grad_norm": 5.348511695861816,
      "learning_rate": 6.331379694637504e-05,
      "loss": 2.050336456298828,
      "memory(GiB)": 77.56,
      "step": 48340,
      "token_acc": 0.5205479452054794,
      "train_speed(iter/s)": 1.439325
    },
    {
      "epoch": 2.0712480185082045,
      "grad_norm": 8.009041786193848,
      "learning_rate": 6.33073100188036e-05,
      "loss": 2.5793895721435547,
      "memory(GiB)": 77.56,
      "step": 48345,
      "token_acc": 0.43342776203966005,
      "train_speed(iter/s)": 1.439348
    },
    {
      "epoch": 2.0714622338374533,
      "grad_norm": 5.634845733642578,
      "learning_rate": 6.330082285015773e-05,
      "loss": 2.254982566833496,
      "memory(GiB)": 77.56,
      "step": 48350,
      "token_acc": 0.5311203319502075,
      "train_speed(iter/s)": 1.439348
    },
    {
      "epoch": 2.0716764491667026,
      "grad_norm": 5.762125015258789,
      "learning_rate": 6.329433544055499e-05,
      "loss": 2.1113458633422852,
      "memory(GiB)": 77.56,
      "step": 48355,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439349
    },
    {
      "epoch": 2.0718906644959514,
      "grad_norm": 5.638726234436035,
      "learning_rate": 6.328784779011288e-05,
      "loss": 2.404056167602539,
      "memory(GiB)": 77.56,
      "step": 48360,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.43931
    },
    {
      "epoch": 2.0721048798252,
      "grad_norm": 4.821144104003906,
      "learning_rate": 6.328135989894896e-05,
      "loss": 2.169768524169922,
      "memory(GiB)": 77.56,
      "step": 48365,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.43929
    },
    {
      "epoch": 2.0723190951544495,
      "grad_norm": 5.7469916343688965,
      "learning_rate": 6.327487176718075e-05,
      "loss": 2.2687667846679687,
      "memory(GiB)": 77.56,
      "step": 48370,
      "token_acc": 0.5363984674329502,
      "train_speed(iter/s)": 1.439267
    },
    {
      "epoch": 2.0725333104836983,
      "grad_norm": 4.575311183929443,
      "learning_rate": 6.326838339492577e-05,
      "loss": 2.321147918701172,
      "memory(GiB)": 77.56,
      "step": 48375,
      "token_acc": 0.5031446540880503,
      "train_speed(iter/s)": 1.439289
    },
    {
      "epoch": 2.072747525812947,
      "grad_norm": 6.307142734527588,
      "learning_rate": 6.326189478230157e-05,
      "loss": 2.359288787841797,
      "memory(GiB)": 77.56,
      "step": 48380,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.439281
    },
    {
      "epoch": 2.0729617411421963,
      "grad_norm": 4.568991661071777,
      "learning_rate": 6.325540592942571e-05,
      "loss": 2.618624687194824,
      "memory(GiB)": 77.56,
      "step": 48385,
      "token_acc": 0.46394984326018807,
      "train_speed(iter/s)": 1.439293
    },
    {
      "epoch": 2.073175956471445,
      "grad_norm": 5.1485724449157715,
      "learning_rate": 6.324891683641576e-05,
      "loss": 2.348969650268555,
      "memory(GiB)": 77.56,
      "step": 48390,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.439279
    },
    {
      "epoch": 2.073390171800694,
      "grad_norm": 6.932342052459717,
      "learning_rate": 6.324242750338922e-05,
      "loss": 2.2965444564819335,
      "memory(GiB)": 77.56,
      "step": 48395,
      "token_acc": 0.5264900662251656,
      "train_speed(iter/s)": 1.439252
    },
    {
      "epoch": 2.073604387129943,
      "grad_norm": 5.213193416595459,
      "learning_rate": 6.323593793046371e-05,
      "loss": 2.507697868347168,
      "memory(GiB)": 77.56,
      "step": 48400,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.439205
    },
    {
      "epoch": 2.073818602459192,
      "grad_norm": 6.838024616241455,
      "learning_rate": 6.322944811775676e-05,
      "loss": 2.2458683013916017,
      "memory(GiB)": 77.56,
      "step": 48405,
      "token_acc": 0.525,
      "train_speed(iter/s)": 1.439204
    },
    {
      "epoch": 2.074032817788441,
      "grad_norm": 6.040023326873779,
      "learning_rate": 6.322295806538594e-05,
      "loss": 2.5360385894775392,
      "memory(GiB)": 77.56,
      "step": 48410,
      "token_acc": 0.5103448275862069,
      "train_speed(iter/s)": 1.439233
    },
    {
      "epoch": 2.07424703311769,
      "grad_norm": 6.14741325378418,
      "learning_rate": 6.321646777346883e-05,
      "loss": 2.425777053833008,
      "memory(GiB)": 77.56,
      "step": 48415,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.439245
    },
    {
      "epoch": 2.074461248446939,
      "grad_norm": 6.541590213775635,
      "learning_rate": 6.320997724212304e-05,
      "loss": 2.2811193466186523,
      "memory(GiB)": 77.56,
      "step": 48420,
      "token_acc": 0.5290322580645161,
      "train_speed(iter/s)": 1.439245
    },
    {
      "epoch": 2.0746754637761877,
      "grad_norm": 5.597967147827148,
      "learning_rate": 6.320348647146608e-05,
      "loss": 2.3456005096435546,
      "memory(GiB)": 77.56,
      "step": 48425,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.439247
    },
    {
      "epoch": 2.074889679105437,
      "grad_norm": 4.073702812194824,
      "learning_rate": 6.319699546161559e-05,
      "loss": 2.480477142333984,
      "memory(GiB)": 77.56,
      "step": 48430,
      "token_acc": 0.4839650145772595,
      "train_speed(iter/s)": 1.439246
    },
    {
      "epoch": 2.075103894434686,
      "grad_norm": 5.629161357879639,
      "learning_rate": 6.319050421268915e-05,
      "loss": 2.342411422729492,
      "memory(GiB)": 77.56,
      "step": 48435,
      "token_acc": 0.5014577259475219,
      "train_speed(iter/s)": 1.439286
    },
    {
      "epoch": 2.0753181097639346,
      "grad_norm": 5.1040568351745605,
      "learning_rate": 6.318401272480435e-05,
      "loss": 2.461376190185547,
      "memory(GiB)": 77.56,
      "step": 48440,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.439307
    },
    {
      "epoch": 2.075532325093184,
      "grad_norm": 4.828313827514648,
      "learning_rate": 6.31775209980788e-05,
      "loss": 2.3649301528930664,
      "memory(GiB)": 77.56,
      "step": 48445,
      "token_acc": 0.48026315789473684,
      "train_speed(iter/s)": 1.439319
    },
    {
      "epoch": 2.0757465404224327,
      "grad_norm": 4.40201473236084,
      "learning_rate": 6.317102903263008e-05,
      "loss": 2.528664207458496,
      "memory(GiB)": 77.56,
      "step": 48450,
      "token_acc": 0.48517520215633425,
      "train_speed(iter/s)": 1.439345
    },
    {
      "epoch": 2.0759607557516815,
      "grad_norm": 5.237856864929199,
      "learning_rate": 6.316453682857582e-05,
      "loss": 2.1672229766845703,
      "memory(GiB)": 77.56,
      "step": 48455,
      "token_acc": 0.5314685314685315,
      "train_speed(iter/s)": 1.439361
    },
    {
      "epoch": 2.0761749710809307,
      "grad_norm": 5.823591709136963,
      "learning_rate": 6.315804438603363e-05,
      "loss": 2.324350357055664,
      "memory(GiB)": 77.56,
      "step": 48460,
      "token_acc": 0.4927007299270073,
      "train_speed(iter/s)": 1.439391
    },
    {
      "epoch": 2.0763891864101796,
      "grad_norm": 5.161983489990234,
      "learning_rate": 6.315155170512111e-05,
      "loss": 2.345977783203125,
      "memory(GiB)": 77.56,
      "step": 48465,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.439408
    },
    {
      "epoch": 2.0766034017394284,
      "grad_norm": 4.3810200691223145,
      "learning_rate": 6.31450587859559e-05,
      "loss": 2.633358192443848,
      "memory(GiB)": 77.56,
      "step": 48470,
      "token_acc": 0.4326923076923077,
      "train_speed(iter/s)": 1.439439
    },
    {
      "epoch": 2.0768176170686776,
      "grad_norm": 4.507627964019775,
      "learning_rate": 6.313856562865564e-05,
      "loss": 2.285697174072266,
      "memory(GiB)": 77.56,
      "step": 48475,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.439398
    },
    {
      "epoch": 2.0770318323979264,
      "grad_norm": 5.572567939758301,
      "learning_rate": 6.313207223333793e-05,
      "loss": 2.541775703430176,
      "memory(GiB)": 77.56,
      "step": 48480,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.439397
    },
    {
      "epoch": 2.0772460477271752,
      "grad_norm": 6.2764997482299805,
      "learning_rate": 6.312557860012039e-05,
      "loss": 2.255630683898926,
      "memory(GiB)": 77.56,
      "step": 48485,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.439452
    },
    {
      "epoch": 2.0774602630564245,
      "grad_norm": 5.716599941253662,
      "learning_rate": 6.31190847291207e-05,
      "loss": 2.53277645111084,
      "memory(GiB)": 77.56,
      "step": 48490,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.43945
    },
    {
      "epoch": 2.0776744783856733,
      "grad_norm": 5.113265514373779,
      "learning_rate": 6.311259062045647e-05,
      "loss": 2.8372575759887697,
      "memory(GiB)": 77.56,
      "step": 48495,
      "token_acc": 0.4117647058823529,
      "train_speed(iter/s)": 1.439471
    },
    {
      "epoch": 2.077888693714922,
      "grad_norm": 4.587961673736572,
      "learning_rate": 6.310609627424537e-05,
      "loss": 2.313138961791992,
      "memory(GiB)": 77.56,
      "step": 48500,
      "token_acc": 0.47540983606557374,
      "train_speed(iter/s)": 1.439487
    },
    {
      "epoch": 2.077888693714922,
      "eval_loss": 2.0567266941070557,
      "eval_runtime": 14.6253,
      "eval_samples_per_second": 6.837,
      "eval_steps_per_second": 6.837,
      "eval_token_acc": 0.4861294583883752,
      "step": 48500
    },
    {
      "epoch": 2.0781029090441714,
      "grad_norm": 5.349452495574951,
      "learning_rate": 6.309960169060504e-05,
      "loss": 2.59619255065918,
      "memory(GiB)": 77.56,
      "step": 48505,
      "token_acc": 0.4685185185185185,
      "train_speed(iter/s)": 1.438787
    },
    {
      "epoch": 2.07831712437342,
      "grad_norm": 3.8574986457824707,
      "learning_rate": 6.309310686965313e-05,
      "loss": 2.4214851379394533,
      "memory(GiB)": 77.56,
      "step": 48510,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.438761
    },
    {
      "epoch": 2.078531339702669,
      "grad_norm": 4.647470474243164,
      "learning_rate": 6.308661181150733e-05,
      "loss": 2.2915735244750977,
      "memory(GiB)": 77.56,
      "step": 48515,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438761
    },
    {
      "epoch": 2.0787455550319183,
      "grad_norm": 5.418434143066406,
      "learning_rate": 6.308011651628525e-05,
      "loss": 2.5039791107177733,
      "memory(GiB)": 77.56,
      "step": 48520,
      "token_acc": 0.47878787878787876,
      "train_speed(iter/s)": 1.438792
    },
    {
      "epoch": 2.078959770361167,
      "grad_norm": 5.092519760131836,
      "learning_rate": 6.30736209841046e-05,
      "loss": 2.503418731689453,
      "memory(GiB)": 77.56,
      "step": 48525,
      "token_acc": 0.4766355140186916,
      "train_speed(iter/s)": 1.438809
    },
    {
      "epoch": 2.079173985690416,
      "grad_norm": 4.155118465423584,
      "learning_rate": 6.306712521508306e-05,
      "loss": 2.1027830123901365,
      "memory(GiB)": 77.56,
      "step": 48530,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.438828
    },
    {
      "epoch": 2.079388201019665,
      "grad_norm": 6.210097789764404,
      "learning_rate": 6.306062920933829e-05,
      "loss": 2.4000572204589843,
      "memory(GiB)": 77.56,
      "step": 48535,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.438836
    },
    {
      "epoch": 2.079602416348914,
      "grad_norm": 5.417062759399414,
      "learning_rate": 6.305413296698795e-05,
      "loss": 2.4895183563232424,
      "memory(GiB)": 77.56,
      "step": 48540,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.438841
    },
    {
      "epoch": 2.0798166316781628,
      "grad_norm": 5.2669196128845215,
      "learning_rate": 6.304763648814974e-05,
      "loss": 2.216512680053711,
      "memory(GiB)": 77.56,
      "step": 48545,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.438851
    },
    {
      "epoch": 2.080030847007412,
      "grad_norm": 5.809610843658447,
      "learning_rate": 6.304113977294137e-05,
      "loss": 2.299217414855957,
      "memory(GiB)": 77.56,
      "step": 48550,
      "token_acc": 0.4423963133640553,
      "train_speed(iter/s)": 1.438865
    },
    {
      "epoch": 2.080245062336661,
      "grad_norm": 8.177706718444824,
      "learning_rate": 6.30346428214805e-05,
      "loss": 2.2650827407836913,
      "memory(GiB)": 77.56,
      "step": 48555,
      "token_acc": 0.5193798449612403,
      "train_speed(iter/s)": 1.438903
    },
    {
      "epoch": 2.0804592776659097,
      "grad_norm": 5.567712783813477,
      "learning_rate": 6.302814563388487e-05,
      "loss": 2.455456352233887,
      "memory(GiB)": 77.56,
      "step": 48560,
      "token_acc": 0.4539877300613497,
      "train_speed(iter/s)": 1.438902
    },
    {
      "epoch": 2.080673492995159,
      "grad_norm": 4.803197860717773,
      "learning_rate": 6.302164821027217e-05,
      "loss": 2.2091701507568358,
      "memory(GiB)": 77.56,
      "step": 48565,
      "token_acc": 0.5341880341880342,
      "train_speed(iter/s)": 1.438889
    },
    {
      "epoch": 2.0808877083244077,
      "grad_norm": 6.3997368812561035,
      "learning_rate": 6.301515055076007e-05,
      "loss": 2.2463455200195312,
      "memory(GiB)": 77.56,
      "step": 48570,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.438931
    },
    {
      "epoch": 2.0811019236536565,
      "grad_norm": 5.555520534515381,
      "learning_rate": 6.300865265546632e-05,
      "loss": 2.7440542221069335,
      "memory(GiB)": 77.56,
      "step": 48575,
      "token_acc": 0.43131868131868134,
      "train_speed(iter/s)": 1.438944
    },
    {
      "epoch": 2.081316138982906,
      "grad_norm": 5.450163841247559,
      "learning_rate": 6.300215452450862e-05,
      "loss": 2.409371566772461,
      "memory(GiB)": 77.56,
      "step": 48580,
      "token_acc": 0.49514563106796117,
      "train_speed(iter/s)": 1.438966
    },
    {
      "epoch": 2.0815303543121546,
      "grad_norm": 5.4445672035217285,
      "learning_rate": 6.29956561580047e-05,
      "loss": 2.444070053100586,
      "memory(GiB)": 77.56,
      "step": 48585,
      "token_acc": 0.48328267477203646,
      "train_speed(iter/s)": 1.438989
    },
    {
      "epoch": 2.0817445696414034,
      "grad_norm": 4.557222843170166,
      "learning_rate": 6.298915755607228e-05,
      "loss": 2.245400238037109,
      "memory(GiB)": 77.56,
      "step": 48590,
      "token_acc": 0.5426621160409556,
      "train_speed(iter/s)": 1.438981
    },
    {
      "epoch": 2.0819587849706527,
      "grad_norm": 4.618313312530518,
      "learning_rate": 6.298265871882908e-05,
      "loss": 2.2743204116821287,
      "memory(GiB)": 77.56,
      "step": 48595,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.438986
    },
    {
      "epoch": 2.0821730002999015,
      "grad_norm": 4.655125617980957,
      "learning_rate": 6.297615964639283e-05,
      "loss": 2.1316608428955077,
      "memory(GiB)": 77.56,
      "step": 48600,
      "token_acc": 0.5641025641025641,
      "train_speed(iter/s)": 1.438966
    },
    {
      "epoch": 2.0823872156291503,
      "grad_norm": 6.335999011993408,
      "learning_rate": 6.29696603388813e-05,
      "loss": 2.473579216003418,
      "memory(GiB)": 77.56,
      "step": 48605,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.438968
    },
    {
      "epoch": 2.0826014309583996,
      "grad_norm": 7.8644514083862305,
      "learning_rate": 6.296316079641218e-05,
      "loss": 2.8764577865600587,
      "memory(GiB)": 77.56,
      "step": 48610,
      "token_acc": 0.4422535211267606,
      "train_speed(iter/s)": 1.438956
    },
    {
      "epoch": 2.0828156462876484,
      "grad_norm": 6.0232462882995605,
      "learning_rate": 6.295666101910325e-05,
      "loss": 2.3139520645141602,
      "memory(GiB)": 77.56,
      "step": 48615,
      "token_acc": 0.48638132295719844,
      "train_speed(iter/s)": 1.438943
    },
    {
      "epoch": 2.083029861616897,
      "grad_norm": 8.493483543395996,
      "learning_rate": 6.295016100707226e-05,
      "loss": 2.5897979736328125,
      "memory(GiB)": 77.56,
      "step": 48620,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.438938
    },
    {
      "epoch": 2.0832440769461464,
      "grad_norm": 5.3446269035339355,
      "learning_rate": 6.294366076043695e-05,
      "loss": 2.4428714752197265,
      "memory(GiB)": 77.56,
      "step": 48625,
      "token_acc": 0.4692982456140351,
      "train_speed(iter/s)": 1.438893
    },
    {
      "epoch": 2.0834582922753953,
      "grad_norm": 6.484088897705078,
      "learning_rate": 6.293716027931507e-05,
      "loss": 2.476865768432617,
      "memory(GiB)": 77.56,
      "step": 48630,
      "token_acc": 0.45936395759717313,
      "train_speed(iter/s)": 1.438919
    },
    {
      "epoch": 2.083672507604644,
      "grad_norm": 5.911398410797119,
      "learning_rate": 6.293065956382442e-05,
      "loss": 2.558656692504883,
      "memory(GiB)": 77.56,
      "step": 48635,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.438903
    },
    {
      "epoch": 2.0838867229338933,
      "grad_norm": 4.09732723236084,
      "learning_rate": 6.292415861408273e-05,
      "loss": 2.475154685974121,
      "memory(GiB)": 77.56,
      "step": 48640,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.43889
    },
    {
      "epoch": 2.084100938263142,
      "grad_norm": 4.679836750030518,
      "learning_rate": 6.291765743020779e-05,
      "loss": 2.631758689880371,
      "memory(GiB)": 77.56,
      "step": 48645,
      "token_acc": 0.48562300319488816,
      "train_speed(iter/s)": 1.438941
    },
    {
      "epoch": 2.084315153592391,
      "grad_norm": 4.731095314025879,
      "learning_rate": 6.291115601231736e-05,
      "loss": 2.4554443359375,
      "memory(GiB)": 77.56,
      "step": 48650,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.438936
    },
    {
      "epoch": 2.08452936892164,
      "grad_norm": 9.38693618774414,
      "learning_rate": 6.290465436052921e-05,
      "loss": 2.5012958526611326,
      "memory(GiB)": 77.56,
      "step": 48655,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.438957
    },
    {
      "epoch": 2.084743584250889,
      "grad_norm": 6.419809818267822,
      "learning_rate": 6.289815247496117e-05,
      "loss": 2.483686828613281,
      "memory(GiB)": 77.56,
      "step": 48660,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438948
    },
    {
      "epoch": 2.084957799580138,
      "grad_norm": 5.986326694488525,
      "learning_rate": 6.289165035573098e-05,
      "loss": 2.166529655456543,
      "memory(GiB)": 77.56,
      "step": 48665,
      "token_acc": 0.5465587044534413,
      "train_speed(iter/s)": 1.438939
    },
    {
      "epoch": 2.085172014909387,
      "grad_norm": 4.805379867553711,
      "learning_rate": 6.288514800295647e-05,
      "loss": 2.393752670288086,
      "memory(GiB)": 77.56,
      "step": 48670,
      "token_acc": 0.47307692307692306,
      "train_speed(iter/s)": 1.438961
    },
    {
      "epoch": 2.085386230238636,
      "grad_norm": 4.750682353973389,
      "learning_rate": 6.287864541675542e-05,
      "loss": 2.4145248413085936,
      "memory(GiB)": 77.56,
      "step": 48675,
      "token_acc": 0.519298245614035,
      "train_speed(iter/s)": 1.439011
    },
    {
      "epoch": 2.0856004455678847,
      "grad_norm": 5.7357707023620605,
      "learning_rate": 6.287214259724559e-05,
      "loss": 2.3581592559814455,
      "memory(GiB)": 77.56,
      "step": 48680,
      "token_acc": 0.5196850393700787,
      "train_speed(iter/s)": 1.439039
    },
    {
      "epoch": 2.085814660897134,
      "grad_norm": 5.493290424346924,
      "learning_rate": 6.286563954454485e-05,
      "loss": 2.458401679992676,
      "memory(GiB)": 77.56,
      "step": 48685,
      "token_acc": 0.4749034749034749,
      "train_speed(iter/s)": 1.439082
    },
    {
      "epoch": 2.0860288762263828,
      "grad_norm": 7.9068732261657715,
      "learning_rate": 6.285913625877097e-05,
      "loss": 2.5358303070068358,
      "memory(GiB)": 77.56,
      "step": 48690,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439099
    },
    {
      "epoch": 2.0862430915556316,
      "grad_norm": 4.675535678863525,
      "learning_rate": 6.285263274004179e-05,
      "loss": 2.4378604888916016,
      "memory(GiB)": 77.56,
      "step": 48695,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.439113
    },
    {
      "epoch": 2.086457306884881,
      "grad_norm": 4.346827983856201,
      "learning_rate": 6.284612898847508e-05,
      "loss": 2.3036895751953126,
      "memory(GiB)": 77.56,
      "step": 48700,
      "token_acc": 0.5076335877862596,
      "train_speed(iter/s)": 1.439104
    },
    {
      "epoch": 2.0866715222141297,
      "grad_norm": 6.296469211578369,
      "learning_rate": 6.283962500418872e-05,
      "loss": 2.703408050537109,
      "memory(GiB)": 77.56,
      "step": 48705,
      "token_acc": 0.47384615384615386,
      "train_speed(iter/s)": 1.439131
    },
    {
      "epoch": 2.0868857375433785,
      "grad_norm": 5.321903228759766,
      "learning_rate": 6.28331207873005e-05,
      "loss": 2.334235954284668,
      "memory(GiB)": 77.56,
      "step": 48710,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.439125
    },
    {
      "epoch": 2.0870999528726277,
      "grad_norm": 5.126958847045898,
      "learning_rate": 6.282661633792826e-05,
      "loss": 2.7237152099609374,
      "memory(GiB)": 77.56,
      "step": 48715,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.439151
    },
    {
      "epoch": 2.0873141682018765,
      "grad_norm": 4.70961332321167,
      "learning_rate": 6.282011165618984e-05,
      "loss": 2.456793785095215,
      "memory(GiB)": 77.56,
      "step": 48720,
      "token_acc": 0.4692737430167598,
      "train_speed(iter/s)": 1.439172
    },
    {
      "epoch": 2.0875283835311254,
      "grad_norm": 5.794946670532227,
      "learning_rate": 6.281360674220305e-05,
      "loss": 2.369644355773926,
      "memory(GiB)": 77.56,
      "step": 48725,
      "token_acc": 0.5015479876160991,
      "train_speed(iter/s)": 1.439184
    },
    {
      "epoch": 2.0877425988603746,
      "grad_norm": 4.590949535369873,
      "learning_rate": 6.280710159608578e-05,
      "loss": 2.6604488372802733,
      "memory(GiB)": 77.56,
      "step": 48730,
      "token_acc": 0.47876447876447875,
      "train_speed(iter/s)": 1.439186
    },
    {
      "epoch": 2.0879568141896234,
      "grad_norm": 4.3121018409729,
      "learning_rate": 6.280059621795583e-05,
      "loss": 2.250263786315918,
      "memory(GiB)": 77.56,
      "step": 48735,
      "token_acc": 0.513595166163142,
      "train_speed(iter/s)": 1.43914
    },
    {
      "epoch": 2.0881710295188722,
      "grad_norm": 5.6718902587890625,
      "learning_rate": 6.279409060793109e-05,
      "loss": 2.4860706329345703,
      "memory(GiB)": 77.56,
      "step": 48740,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.439169
    },
    {
      "epoch": 2.0883852448481215,
      "grad_norm": 4.941648006439209,
      "learning_rate": 6.278758476612937e-05,
      "loss": 2.503778266906738,
      "memory(GiB)": 77.56,
      "step": 48745,
      "token_acc": 0.47157190635451507,
      "train_speed(iter/s)": 1.439161
    },
    {
      "epoch": 2.0885994601773703,
      "grad_norm": 5.030585765838623,
      "learning_rate": 6.278107869266859e-05,
      "loss": 2.29956169128418,
      "memory(GiB)": 77.56,
      "step": 48750,
      "token_acc": 0.5313432835820896,
      "train_speed(iter/s)": 1.439115
    },
    {
      "epoch": 2.088813675506619,
      "grad_norm": 5.252609729766846,
      "learning_rate": 6.277457238766655e-05,
      "loss": 2.123383331298828,
      "memory(GiB)": 77.56,
      "step": 48755,
      "token_acc": 0.5394190871369294,
      "train_speed(iter/s)": 1.439125
    },
    {
      "epoch": 2.0890278908358684,
      "grad_norm": 5.799145221710205,
      "learning_rate": 6.276806585124116e-05,
      "loss": 2.483392906188965,
      "memory(GiB)": 77.56,
      "step": 48760,
      "token_acc": 0.5052631578947369,
      "train_speed(iter/s)": 1.43912
    },
    {
      "epoch": 2.089242106165117,
      "grad_norm": 4.487954139709473,
      "learning_rate": 6.276155908351027e-05,
      "loss": 2.4665964126586912,
      "memory(GiB)": 77.56,
      "step": 48765,
      "token_acc": 0.49174917491749176,
      "train_speed(iter/s)": 1.439105
    },
    {
      "epoch": 2.089456321494366,
      "grad_norm": 5.483849048614502,
      "learning_rate": 6.275505208459178e-05,
      "loss": 2.424612808227539,
      "memory(GiB)": 77.56,
      "step": 48770,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.439124
    },
    {
      "epoch": 2.0896705368236153,
      "grad_norm": 6.1339030265808105,
      "learning_rate": 6.274854485460355e-05,
      "loss": 2.3976306915283203,
      "memory(GiB)": 77.56,
      "step": 48775,
      "token_acc": 0.5112540192926045,
      "train_speed(iter/s)": 1.439125
    },
    {
      "epoch": 2.089884752152864,
      "grad_norm": 4.765876770019531,
      "learning_rate": 6.274203739366347e-05,
      "loss": 2.3025442123413087,
      "memory(GiB)": 77.56,
      "step": 48780,
      "token_acc": 0.5076335877862596,
      "train_speed(iter/s)": 1.439123
    },
    {
      "epoch": 2.090098967482113,
      "grad_norm": 6.810534954071045,
      "learning_rate": 6.273552970188942e-05,
      "loss": 2.316607666015625,
      "memory(GiB)": 77.56,
      "step": 48785,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.439161
    },
    {
      "epoch": 2.090313182811362,
      "grad_norm": 5.791784763336182,
      "learning_rate": 6.272902177939933e-05,
      "loss": 2.345306396484375,
      "memory(GiB)": 77.56,
      "step": 48790,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.439161
    },
    {
      "epoch": 2.090527398140611,
      "grad_norm": 4.690967082977295,
      "learning_rate": 6.272251362631107e-05,
      "loss": 2.3964767456054688,
      "memory(GiB)": 77.56,
      "step": 48795,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439175
    },
    {
      "epoch": 2.0907416134698598,
      "grad_norm": 5.468694686889648,
      "learning_rate": 6.27160052427425e-05,
      "loss": 2.508285140991211,
      "memory(GiB)": 77.56,
      "step": 48800,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.439164
    },
    {
      "epoch": 2.090955828799109,
      "grad_norm": 4.597681045532227,
      "learning_rate": 6.27094966288116e-05,
      "loss": 2.254531478881836,
      "memory(GiB)": 77.56,
      "step": 48805,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.439188
    },
    {
      "epoch": 2.091170044128358,
      "grad_norm": 5.0382537841796875,
      "learning_rate": 6.270298778463624e-05,
      "loss": 2.6489679336547853,
      "memory(GiB)": 77.56,
      "step": 48810,
      "token_acc": 0.5021097046413502,
      "train_speed(iter/s)": 1.439214
    },
    {
      "epoch": 2.0913842594576066,
      "grad_norm": 5.8131489753723145,
      "learning_rate": 6.269647871033432e-05,
      "loss": 2.5048757553100587,
      "memory(GiB)": 77.56,
      "step": 48815,
      "token_acc": 0.4769736842105263,
      "train_speed(iter/s)": 1.439188
    },
    {
      "epoch": 2.091598474786856,
      "grad_norm": 4.653921127319336,
      "learning_rate": 6.26899694060238e-05,
      "loss": 2.340727615356445,
      "memory(GiB)": 77.56,
      "step": 48820,
      "token_acc": 0.47604790419161674,
      "train_speed(iter/s)": 1.439173
    },
    {
      "epoch": 2.0918126901161047,
      "grad_norm": 4.954652309417725,
      "learning_rate": 6.26834598718226e-05,
      "loss": 2.1758090972900392,
      "memory(GiB)": 77.56,
      "step": 48825,
      "token_acc": 0.5229357798165137,
      "train_speed(iter/s)": 1.439192
    },
    {
      "epoch": 2.0920269054453535,
      "grad_norm": 6.207099437713623,
      "learning_rate": 6.267695010784859e-05,
      "loss": 2.5750011444091796,
      "memory(GiB)": 77.56,
      "step": 48830,
      "token_acc": 0.4479166666666667,
      "train_speed(iter/s)": 1.439211
    },
    {
      "epoch": 2.092241120774603,
      "grad_norm": 5.960329055786133,
      "learning_rate": 6.267044011421975e-05,
      "loss": 2.672381782531738,
      "memory(GiB)": 77.56,
      "step": 48835,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.43922
    },
    {
      "epoch": 2.0924553361038516,
      "grad_norm": 4.969362258911133,
      "learning_rate": 6.2663929891054e-05,
      "loss": 2.771834945678711,
      "memory(GiB)": 77.56,
      "step": 48840,
      "token_acc": 0.4461538461538462,
      "train_speed(iter/s)": 1.439231
    },
    {
      "epoch": 2.0926695514331004,
      "grad_norm": 5.441162586212158,
      "learning_rate": 6.265741943846926e-05,
      "loss": 2.423880958557129,
      "memory(GiB)": 77.56,
      "step": 48845,
      "token_acc": 0.5241379310344828,
      "train_speed(iter/s)": 1.439245
    },
    {
      "epoch": 2.0928837667623497,
      "grad_norm": 8.320979118347168,
      "learning_rate": 6.265090875658353e-05,
      "loss": 2.3598764419555662,
      "memory(GiB)": 77.56,
      "step": 48850,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.439272
    },
    {
      "epoch": 2.0930979820915985,
      "grad_norm": 5.704608917236328,
      "learning_rate": 6.264439784551472e-05,
      "loss": 2.1590160369873046,
      "memory(GiB)": 77.56,
      "step": 48855,
      "token_acc": 0.5323741007194245,
      "train_speed(iter/s)": 1.439272
    },
    {
      "epoch": 2.0933121974208473,
      "grad_norm": 5.506229877471924,
      "learning_rate": 6.263788670538075e-05,
      "loss": 2.025997352600098,
      "memory(GiB)": 77.56,
      "step": 48860,
      "token_acc": 0.6032388663967612,
      "train_speed(iter/s)": 1.4393
    },
    {
      "epoch": 2.0935264127500965,
      "grad_norm": 4.564820289611816,
      "learning_rate": 6.263137533629963e-05,
      "loss": 2.3568538665771483,
      "memory(GiB)": 77.56,
      "step": 48865,
      "token_acc": 0.5412541254125413,
      "train_speed(iter/s)": 1.439319
    },
    {
      "epoch": 2.0937406280793454,
      "grad_norm": 5.8116960525512695,
      "learning_rate": 6.262486373838929e-05,
      "loss": 2.299452781677246,
      "memory(GiB)": 77.56,
      "step": 48870,
      "token_acc": 0.5401459854014599,
      "train_speed(iter/s)": 1.439344
    },
    {
      "epoch": 2.093954843408594,
      "grad_norm": 3.955968141555786,
      "learning_rate": 6.261835191176769e-05,
      "loss": 2.393795204162598,
      "memory(GiB)": 77.56,
      "step": 48875,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43937
    },
    {
      "epoch": 2.0941690587378434,
      "grad_norm": 5.866372585296631,
      "learning_rate": 6.261183985655281e-05,
      "loss": 2.3203685760498045,
      "memory(GiB)": 77.56,
      "step": 48880,
      "token_acc": 0.5372549019607843,
      "train_speed(iter/s)": 1.439379
    },
    {
      "epoch": 2.0943832740670922,
      "grad_norm": 4.214962959289551,
      "learning_rate": 6.260532757286264e-05,
      "loss": 2.567597579956055,
      "memory(GiB)": 77.56,
      "step": 48885,
      "token_acc": 0.4525993883792049,
      "train_speed(iter/s)": 1.439383
    },
    {
      "epoch": 2.094597489396341,
      "grad_norm": 4.494652271270752,
      "learning_rate": 6.259881506081512e-05,
      "loss": 2.3942338943481447,
      "memory(GiB)": 77.56,
      "step": 48890,
      "token_acc": 0.46440677966101696,
      "train_speed(iter/s)": 1.439355
    },
    {
      "epoch": 2.0948117047255903,
      "grad_norm": 6.093865871429443,
      "learning_rate": 6.259230232052826e-05,
      "loss": 2.6650035858154295,
      "memory(GiB)": 77.56,
      "step": 48895,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.439344
    },
    {
      "epoch": 2.095025920054839,
      "grad_norm": 5.515293121337891,
      "learning_rate": 6.258578935212e-05,
      "loss": 2.2267967224121095,
      "memory(GiB)": 77.56,
      "step": 48900,
      "token_acc": 0.5414201183431953,
      "train_speed(iter/s)": 1.439379
    },
    {
      "epoch": 2.095240135384088,
      "grad_norm": 5.29227876663208,
      "learning_rate": 6.257927615570839e-05,
      "loss": 2.3304256439208983,
      "memory(GiB)": 77.56,
      "step": 48905,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.43939
    },
    {
      "epoch": 2.095454350713337,
      "grad_norm": 6.785962104797363,
      "learning_rate": 6.257276273141139e-05,
      "loss": 2.422073173522949,
      "memory(GiB)": 77.56,
      "step": 48910,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.439353
    },
    {
      "epoch": 2.095668566042586,
      "grad_norm": 4.055444240570068,
      "learning_rate": 6.256624907934699e-05,
      "loss": 2.2771997451782227,
      "memory(GiB)": 77.56,
      "step": 48915,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.439364
    },
    {
      "epoch": 2.095882781371835,
      "grad_norm": 5.131942272186279,
      "learning_rate": 6.25597351996332e-05,
      "loss": 2.3919715881347656,
      "memory(GiB)": 77.56,
      "step": 48920,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.439383
    },
    {
      "epoch": 2.096096996701084,
      "grad_norm": 7.323632717132568,
      "learning_rate": 6.255322109238803e-05,
      "loss": 2.4990726470947267,
      "memory(GiB)": 77.56,
      "step": 48925,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.439387
    },
    {
      "epoch": 2.096311212030333,
      "grad_norm": 5.591567516326904,
      "learning_rate": 6.254670675772947e-05,
      "loss": 2.7983993530273437,
      "memory(GiB)": 77.56,
      "step": 48930,
      "token_acc": 0.4186746987951807,
      "train_speed(iter/s)": 1.439409
    },
    {
      "epoch": 2.0965254273595817,
      "grad_norm": 4.745422840118408,
      "learning_rate": 6.254019219577556e-05,
      "loss": 2.853195381164551,
      "memory(GiB)": 77.56,
      "step": 48935,
      "token_acc": 0.4680232558139535,
      "train_speed(iter/s)": 1.439413
    },
    {
      "epoch": 2.096739642688831,
      "grad_norm": 4.464240074157715,
      "learning_rate": 6.253367740664431e-05,
      "loss": 2.533761405944824,
      "memory(GiB)": 77.56,
      "step": 48940,
      "token_acc": 0.501628664495114,
      "train_speed(iter/s)": 1.439435
    },
    {
      "epoch": 2.0969538580180798,
      "grad_norm": 7.061866760253906,
      "learning_rate": 6.252716239045372e-05,
      "loss": 2.2278575897216797,
      "memory(GiB)": 77.56,
      "step": 48945,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.439458
    },
    {
      "epoch": 2.0971680733473286,
      "grad_norm": 4.807161331176758,
      "learning_rate": 6.252064714732185e-05,
      "loss": 2.4195178985595702,
      "memory(GiB)": 77.56,
      "step": 48950,
      "token_acc": 0.5031847133757962,
      "train_speed(iter/s)": 1.439442
    },
    {
      "epoch": 2.097382288676578,
      "grad_norm": 7.726494789123535,
      "learning_rate": 6.25141316773667e-05,
      "loss": 2.4144935607910156,
      "memory(GiB)": 77.56,
      "step": 48955,
      "token_acc": 0.486646884272997,
      "train_speed(iter/s)": 1.439454
    },
    {
      "epoch": 2.0975965040058266,
      "grad_norm": 5.9279465675354,
      "learning_rate": 6.250761598070632e-05,
      "loss": 2.538133239746094,
      "memory(GiB)": 77.56,
      "step": 48960,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.439462
    },
    {
      "epoch": 2.0978107193350755,
      "grad_norm": 5.672407627105713,
      "learning_rate": 6.250110005745874e-05,
      "loss": 2.5083023071289063,
      "memory(GiB)": 77.56,
      "step": 48965,
      "token_acc": 0.49612403100775193,
      "train_speed(iter/s)": 1.439445
    },
    {
      "epoch": 2.0980249346643247,
      "grad_norm": 5.007399082183838,
      "learning_rate": 6.249458390774201e-05,
      "loss": 2.5312580108642577,
      "memory(GiB)": 77.56,
      "step": 48970,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.439436
    },
    {
      "epoch": 2.0982391499935735,
      "grad_norm": 8.42153263092041,
      "learning_rate": 6.248806753167417e-05,
      "loss": 2.308869743347168,
      "memory(GiB)": 77.56,
      "step": 48975,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.439417
    },
    {
      "epoch": 2.0984533653228223,
      "grad_norm": 4.837903022766113,
      "learning_rate": 6.248155092937326e-05,
      "loss": 2.3988525390625,
      "memory(GiB)": 77.56,
      "step": 48980,
      "token_acc": 0.4981132075471698,
      "train_speed(iter/s)": 1.439427
    },
    {
      "epoch": 2.0986675806520716,
      "grad_norm": 4.984959602355957,
      "learning_rate": 6.247503410095737e-05,
      "loss": 2.5160776138305665,
      "memory(GiB)": 77.56,
      "step": 48985,
      "token_acc": 0.45652173913043476,
      "train_speed(iter/s)": 1.439413
    },
    {
      "epoch": 2.0988817959813204,
      "grad_norm": 5.135047435760498,
      "learning_rate": 6.246851704654451e-05,
      "loss": 2.214380645751953,
      "memory(GiB)": 77.56,
      "step": 48990,
      "token_acc": 0.5655172413793104,
      "train_speed(iter/s)": 1.439426
    },
    {
      "epoch": 2.099096011310569,
      "grad_norm": 6.444039821624756,
      "learning_rate": 6.246199976625277e-05,
      "loss": 2.469023323059082,
      "memory(GiB)": 77.56,
      "step": 48995,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.439444
    },
    {
      "epoch": 2.0993102266398185,
      "grad_norm": 4.798272132873535,
      "learning_rate": 6.245548226020024e-05,
      "loss": 2.409999465942383,
      "memory(GiB)": 77.56,
      "step": 49000,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.439477
    },
    {
      "epoch": 2.0993102266398185,
      "eval_loss": 2.222402811050415,
      "eval_runtime": 13.8381,
      "eval_samples_per_second": 7.226,
      "eval_steps_per_second": 7.226,
      "eval_token_acc": 0.461212976022567,
      "step": 49000
    },
    {
      "epoch": 2.0995244419690673,
      "grad_norm": 6.333439826965332,
      "learning_rate": 6.244896452850494e-05,
      "loss": 2.6709468841552733,
      "memory(GiB)": 77.56,
      "step": 49005,
      "token_acc": 0.456,
      "train_speed(iter/s)": 1.438812
    },
    {
      "epoch": 2.099738657298316,
      "grad_norm": 4.671489715576172,
      "learning_rate": 6.244244657128498e-05,
      "loss": 2.3175464630126954,
      "memory(GiB)": 77.56,
      "step": 49010,
      "token_acc": 0.5355648535564853,
      "train_speed(iter/s)": 1.438861
    },
    {
      "epoch": 2.0999528726275654,
      "grad_norm": 5.851550579071045,
      "learning_rate": 6.243592838865842e-05,
      "loss": 2.3232490539550783,
      "memory(GiB)": 77.56,
      "step": 49015,
      "token_acc": 0.504,
      "train_speed(iter/s)": 1.438884
    },
    {
      "epoch": 2.100167087956814,
      "grad_norm": 6.119845390319824,
      "learning_rate": 6.242940998074336e-05,
      "loss": 2.5297639846801756,
      "memory(GiB)": 77.56,
      "step": 49020,
      "token_acc": 0.46441947565543074,
      "train_speed(iter/s)": 1.438855
    },
    {
      "epoch": 2.100381303286063,
      "grad_norm": 5.294820785522461,
      "learning_rate": 6.242289134765788e-05,
      "loss": 2.530585289001465,
      "memory(GiB)": 77.56,
      "step": 49025,
      "token_acc": 0.49382716049382713,
      "train_speed(iter/s)": 1.438855
    },
    {
      "epoch": 2.1005955186153122,
      "grad_norm": 5.15852689743042,
      "learning_rate": 6.241637248952006e-05,
      "loss": 2.457562255859375,
      "memory(GiB)": 77.56,
      "step": 49030,
      "token_acc": 0.47865853658536583,
      "train_speed(iter/s)": 1.438875
    },
    {
      "epoch": 2.100809733944561,
      "grad_norm": 6.752788543701172,
      "learning_rate": 6.2409853406448e-05,
      "loss": 2.4850791931152343,
      "memory(GiB)": 77.56,
      "step": 49035,
      "token_acc": 0.45394736842105265,
      "train_speed(iter/s)": 1.438904
    },
    {
      "epoch": 2.10102394927381,
      "grad_norm": 3.656010389328003,
      "learning_rate": 6.240333409855983e-05,
      "loss": 2.648000717163086,
      "memory(GiB)": 77.56,
      "step": 49040,
      "token_acc": 0.4653179190751445,
      "train_speed(iter/s)": 1.438951
    },
    {
      "epoch": 2.101238164603059,
      "grad_norm": 5.753542423248291,
      "learning_rate": 6.239681456597361e-05,
      "loss": 1.9628711700439454,
      "memory(GiB)": 77.56,
      "step": 49045,
      "token_acc": 0.5206611570247934,
      "train_speed(iter/s)": 1.438956
    },
    {
      "epoch": 2.101452379932308,
      "grad_norm": 4.903069019317627,
      "learning_rate": 6.239029480880747e-05,
      "loss": 2.4458799362182617,
      "memory(GiB)": 77.56,
      "step": 49050,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.438933
    },
    {
      "epoch": 2.1016665952615567,
      "grad_norm": 4.62379789352417,
      "learning_rate": 6.238377482717951e-05,
      "loss": 2.2904151916503905,
      "memory(GiB)": 77.56,
      "step": 49055,
      "token_acc": 0.5353159851301115,
      "train_speed(iter/s)": 1.43897
    },
    {
      "epoch": 2.101880810590806,
      "grad_norm": 4.9703049659729,
      "learning_rate": 6.237725462120784e-05,
      "loss": 2.2056434631347654,
      "memory(GiB)": 77.56,
      "step": 49060,
      "token_acc": 0.5387453874538746,
      "train_speed(iter/s)": 1.438942
    },
    {
      "epoch": 2.102095025920055,
      "grad_norm": 5.410557746887207,
      "learning_rate": 6.237073419101061e-05,
      "loss": 2.29724063873291,
      "memory(GiB)": 77.56,
      "step": 49065,
      "token_acc": 0.5407166123778502,
      "train_speed(iter/s)": 1.438857
    },
    {
      "epoch": 2.1023092412493036,
      "grad_norm": 4.390076637268066,
      "learning_rate": 6.236421353670592e-05,
      "loss": 2.665765380859375,
      "memory(GiB)": 77.56,
      "step": 49070,
      "token_acc": 0.46835443037974683,
      "train_speed(iter/s)": 1.43883
    },
    {
      "epoch": 2.102523456578553,
      "grad_norm": 5.287171840667725,
      "learning_rate": 6.235769265841191e-05,
      "loss": 2.5007225036621095,
      "memory(GiB)": 77.56,
      "step": 49075,
      "token_acc": 0.45569620253164556,
      "train_speed(iter/s)": 1.43886
    },
    {
      "epoch": 2.1027376719078017,
      "grad_norm": 6.390717029571533,
      "learning_rate": 6.235117155624671e-05,
      "loss": 2.561604690551758,
      "memory(GiB)": 77.56,
      "step": 49080,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.438871
    },
    {
      "epoch": 2.1029518872370505,
      "grad_norm": 6.165492534637451,
      "learning_rate": 6.234465023032844e-05,
      "loss": 2.133884620666504,
      "memory(GiB)": 77.56,
      "step": 49085,
      "token_acc": 0.5323741007194245,
      "train_speed(iter/s)": 1.43889
    },
    {
      "epoch": 2.1031661025662998,
      "grad_norm": 4.798530578613281,
      "learning_rate": 6.233812868077525e-05,
      "loss": 2.6677425384521483,
      "memory(GiB)": 77.56,
      "step": 49090,
      "token_acc": 0.47381546134663344,
      "train_speed(iter/s)": 1.438923
    },
    {
      "epoch": 2.1033803178955486,
      "grad_norm": 5.187280654907227,
      "learning_rate": 6.233160690770528e-05,
      "loss": 2.5310846328735352,
      "memory(GiB)": 77.56,
      "step": 49095,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.438904
    },
    {
      "epoch": 2.1035945332247974,
      "grad_norm": 5.205384731292725,
      "learning_rate": 6.23250849112367e-05,
      "loss": 2.503949737548828,
      "memory(GiB)": 77.56,
      "step": 49100,
      "token_acc": 0.4670487106017192,
      "train_speed(iter/s)": 1.43889
    },
    {
      "epoch": 2.1038087485540466,
      "grad_norm": 3.6947855949401855,
      "learning_rate": 6.231856269148762e-05,
      "loss": 2.215028762817383,
      "memory(GiB)": 77.56,
      "step": 49105,
      "token_acc": 0.5046153846153846,
      "train_speed(iter/s)": 1.438888
    },
    {
      "epoch": 2.1040229638832955,
      "grad_norm": 5.002501010894775,
      "learning_rate": 6.231204024857624e-05,
      "loss": 2.3441003799438476,
      "memory(GiB)": 77.56,
      "step": 49110,
      "token_acc": 0.5268817204301075,
      "train_speed(iter/s)": 1.438917
    },
    {
      "epoch": 2.1042371792125443,
      "grad_norm": 5.260482311248779,
      "learning_rate": 6.23055175826207e-05,
      "loss": 2.4012269973754883,
      "memory(GiB)": 77.56,
      "step": 49115,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.438922
    },
    {
      "epoch": 2.1044513945417935,
      "grad_norm": 7.0244221687316895,
      "learning_rate": 6.229899469373917e-05,
      "loss": 2.4538755416870117,
      "memory(GiB)": 77.56,
      "step": 49120,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.438931
    },
    {
      "epoch": 2.1046656098710423,
      "grad_norm": 5.103029727935791,
      "learning_rate": 6.229247158204981e-05,
      "loss": 2.3814722061157227,
      "memory(GiB)": 77.56,
      "step": 49125,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.43891
    },
    {
      "epoch": 2.104879825200291,
      "grad_norm": 5.0265703201293945,
      "learning_rate": 6.228594824767078e-05,
      "loss": 2.51201171875,
      "memory(GiB)": 77.56,
      "step": 49130,
      "token_acc": 0.46616541353383456,
      "train_speed(iter/s)": 1.438912
    },
    {
      "epoch": 2.1050940405295404,
      "grad_norm": 6.083210468292236,
      "learning_rate": 6.227942469072027e-05,
      "loss": 2.340061378479004,
      "memory(GiB)": 77.56,
      "step": 49135,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.438914
    },
    {
      "epoch": 2.1053082558587892,
      "grad_norm": 5.102031707763672,
      "learning_rate": 6.227290091131648e-05,
      "loss": 2.9406700134277344,
      "memory(GiB)": 77.56,
      "step": 49140,
      "token_acc": 0.4141791044776119,
      "train_speed(iter/s)": 1.438934
    },
    {
      "epoch": 2.105522471188038,
      "grad_norm": 5.642523765563965,
      "learning_rate": 6.22663769095776e-05,
      "loss": 2.439640426635742,
      "memory(GiB)": 77.56,
      "step": 49145,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.438937
    },
    {
      "epoch": 2.1057366865172873,
      "grad_norm": 5.119693756103516,
      "learning_rate": 6.225985268562175e-05,
      "loss": 2.68221492767334,
      "memory(GiB)": 77.56,
      "step": 49150,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.438952
    },
    {
      "epoch": 2.105950901846536,
      "grad_norm": 6.095335483551025,
      "learning_rate": 6.22533282395672e-05,
      "loss": 2.130111312866211,
      "memory(GiB)": 77.56,
      "step": 49155,
      "token_acc": 0.5241157556270096,
      "train_speed(iter/s)": 1.438974
    },
    {
      "epoch": 2.106165117175785,
      "grad_norm": 6.140216827392578,
      "learning_rate": 6.22468035715321e-05,
      "loss": 2.455864906311035,
      "memory(GiB)": 77.56,
      "step": 49160,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.438923
    },
    {
      "epoch": 2.106379332505034,
      "grad_norm": 5.604785919189453,
      "learning_rate": 6.224027868163467e-05,
      "loss": 2.2982151031494142,
      "memory(GiB)": 77.56,
      "step": 49165,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.438932
    },
    {
      "epoch": 2.106593547834283,
      "grad_norm": 5.83743953704834,
      "learning_rate": 6.223375356999311e-05,
      "loss": 2.18133544921875,
      "memory(GiB)": 77.56,
      "step": 49170,
      "token_acc": 0.5410447761194029,
      "train_speed(iter/s)": 1.438963
    },
    {
      "epoch": 2.106807763163532,
      "grad_norm": 3.9852590560913086,
      "learning_rate": 6.222722823672562e-05,
      "loss": 2.4130273818969727,
      "memory(GiB)": 77.56,
      "step": 49175,
      "token_acc": 0.47161572052401746,
      "train_speed(iter/s)": 1.438952
    },
    {
      "epoch": 2.107021978492781,
      "grad_norm": 5.204185485839844,
      "learning_rate": 6.222070268195041e-05,
      "loss": 2.62127799987793,
      "memory(GiB)": 77.56,
      "step": 49180,
      "token_acc": 0.48559670781893005,
      "train_speed(iter/s)": 1.438993
    },
    {
      "epoch": 2.10723619382203,
      "grad_norm": 5.959209442138672,
      "learning_rate": 6.221417690578574e-05,
      "loss": 2.514321517944336,
      "memory(GiB)": 77.56,
      "step": 49185,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.439038
    },
    {
      "epoch": 2.1074504091512787,
      "grad_norm": 5.792915344238281,
      "learning_rate": 6.220765090834977e-05,
      "loss": 2.6759796142578125,
      "memory(GiB)": 77.56,
      "step": 49190,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.439096
    },
    {
      "epoch": 2.107664624480528,
      "grad_norm": 6.1992573738098145,
      "learning_rate": 6.220112468976076e-05,
      "loss": 2.231016159057617,
      "memory(GiB)": 77.56,
      "step": 49195,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.439102
    },
    {
      "epoch": 2.1078788398097767,
      "grad_norm": 5.198278427124023,
      "learning_rate": 6.219459825013694e-05,
      "loss": 2.3910972595214846,
      "memory(GiB)": 77.56,
      "step": 49200,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.439099
    },
    {
      "epoch": 2.1080930551390256,
      "grad_norm": 6.3471574783325195,
      "learning_rate": 6.218807158959652e-05,
      "loss": 2.1711345672607423,
      "memory(GiB)": 77.56,
      "step": 49205,
      "token_acc": 0.5055762081784386,
      "train_speed(iter/s)": 1.439134
    },
    {
      "epoch": 2.108307270468275,
      "grad_norm": 4.638065338134766,
      "learning_rate": 6.218154470825775e-05,
      "loss": 2.402248764038086,
      "memory(GiB)": 77.56,
      "step": 49210,
      "token_acc": 0.4364820846905538,
      "train_speed(iter/s)": 1.439128
    },
    {
      "epoch": 2.1085214857975236,
      "grad_norm": 5.397317409515381,
      "learning_rate": 6.217501760623889e-05,
      "loss": 2.3646902084350585,
      "memory(GiB)": 77.56,
      "step": 49215,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.439085
    },
    {
      "epoch": 2.1087357011267724,
      "grad_norm": 5.058228492736816,
      "learning_rate": 6.216849028365815e-05,
      "loss": 2.475019836425781,
      "memory(GiB)": 77.56,
      "step": 49220,
      "token_acc": 0.49823321554770317,
      "train_speed(iter/s)": 1.439093
    },
    {
      "epoch": 2.1089499164560217,
      "grad_norm": 5.189786911010742,
      "learning_rate": 6.216196274063379e-05,
      "loss": 2.4395408630371094,
      "memory(GiB)": 77.56,
      "step": 49225,
      "token_acc": 0.4758364312267658,
      "train_speed(iter/s)": 1.439108
    },
    {
      "epoch": 2.1091641317852705,
      "grad_norm": 5.8689680099487305,
      "learning_rate": 6.215543497728407e-05,
      "loss": 2.4147130966186525,
      "memory(GiB)": 77.56,
      "step": 49230,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.439149
    },
    {
      "epoch": 2.1093783471145193,
      "grad_norm": 7.903217792510986,
      "learning_rate": 6.214890699372724e-05,
      "loss": 2.2708404541015623,
      "memory(GiB)": 77.56,
      "step": 49235,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.439149
    },
    {
      "epoch": 2.1095925624437686,
      "grad_norm": 4.991354465484619,
      "learning_rate": 6.214237879008157e-05,
      "loss": 2.434276580810547,
      "memory(GiB)": 77.56,
      "step": 49240,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.439168
    },
    {
      "epoch": 2.1098067777730174,
      "grad_norm": 6.885368824005127,
      "learning_rate": 6.213585036646531e-05,
      "loss": 2.2261051177978515,
      "memory(GiB)": 77.56,
      "step": 49245,
      "token_acc": 0.4984520123839009,
      "train_speed(iter/s)": 1.439165
    },
    {
      "epoch": 2.110020993102266,
      "grad_norm": 8.591289520263672,
      "learning_rate": 6.212932172299674e-05,
      "loss": 2.4186332702636717,
      "memory(GiB)": 77.56,
      "step": 49250,
      "token_acc": 0.49356223175965663,
      "train_speed(iter/s)": 1.439122
    },
    {
      "epoch": 2.1102352084315155,
      "grad_norm": 6.000884056091309,
      "learning_rate": 6.212279285979412e-05,
      "loss": 2.32342643737793,
      "memory(GiB)": 77.56,
      "step": 49255,
      "token_acc": 0.528052805280528,
      "train_speed(iter/s)": 1.439132
    },
    {
      "epoch": 2.1104494237607643,
      "grad_norm": 5.541501998901367,
      "learning_rate": 6.211626377697575e-05,
      "loss": 2.3411909103393556,
      "memory(GiB)": 77.56,
      "step": 49260,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.439131
    },
    {
      "epoch": 2.110663639090013,
      "grad_norm": 4.93800687789917,
      "learning_rate": 6.210973447465988e-05,
      "loss": 2.314703178405762,
      "memory(GiB)": 77.56,
      "step": 49265,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.439127
    },
    {
      "epoch": 2.1108778544192623,
      "grad_norm": 4.915526390075684,
      "learning_rate": 6.210320495296484e-05,
      "loss": 2.8657541275024414,
      "memory(GiB)": 77.56,
      "step": 49270,
      "token_acc": 0.40498442367601245,
      "train_speed(iter/s)": 1.439128
    },
    {
      "epoch": 2.111092069748511,
      "grad_norm": 5.895676136016846,
      "learning_rate": 6.209667521200886e-05,
      "loss": 2.4640504837036135,
      "memory(GiB)": 77.56,
      "step": 49275,
      "token_acc": 0.44363636363636366,
      "train_speed(iter/s)": 1.43912
    },
    {
      "epoch": 2.11130628507776,
      "grad_norm": 5.05529260635376,
      "learning_rate": 6.209014525191025e-05,
      "loss": 2.178257179260254,
      "memory(GiB)": 77.56,
      "step": 49280,
      "token_acc": 0.5193548387096775,
      "train_speed(iter/s)": 1.439096
    },
    {
      "epoch": 2.1115205004070092,
      "grad_norm": 6.255785942077637,
      "learning_rate": 6.208361507278735e-05,
      "loss": 2.2371715545654296,
      "memory(GiB)": 77.56,
      "step": 49285,
      "token_acc": 0.5508771929824562,
      "train_speed(iter/s)": 1.439044
    },
    {
      "epoch": 2.111734715736258,
      "grad_norm": 5.389954566955566,
      "learning_rate": 6.207708467475842e-05,
      "loss": 2.3669509887695312,
      "memory(GiB)": 77.56,
      "step": 49290,
      "token_acc": 0.5227272727272727,
      "train_speed(iter/s)": 1.43905
    },
    {
      "epoch": 2.111948931065507,
      "grad_norm": 4.614659309387207,
      "learning_rate": 6.207055405794176e-05,
      "loss": 2.2462226867675783,
      "memory(GiB)": 77.56,
      "step": 49295,
      "token_acc": 0.48923076923076925,
      "train_speed(iter/s)": 1.439043
    },
    {
      "epoch": 2.112163146394756,
      "grad_norm": 4.930994987487793,
      "learning_rate": 6.20640232224557e-05,
      "loss": 2.4049846649169924,
      "memory(GiB)": 77.56,
      "step": 49300,
      "token_acc": 0.5084175084175084,
      "train_speed(iter/s)": 1.439023
    },
    {
      "epoch": 2.112377361724005,
      "grad_norm": 4.987551689147949,
      "learning_rate": 6.205749216841855e-05,
      "loss": 2.333563232421875,
      "memory(GiB)": 77.56,
      "step": 49305,
      "token_acc": 0.4911660777385159,
      "train_speed(iter/s)": 1.439001
    },
    {
      "epoch": 2.1125915770532537,
      "grad_norm": 5.033623695373535,
      "learning_rate": 6.20509608959486e-05,
      "loss": 2.192428398132324,
      "memory(GiB)": 77.56,
      "step": 49310,
      "token_acc": 0.537117903930131,
      "train_speed(iter/s)": 1.439017
    },
    {
      "epoch": 2.112805792382503,
      "grad_norm": 5.2437663078308105,
      "learning_rate": 6.20444294051642e-05,
      "loss": 2.475846862792969,
      "memory(GiB)": 77.56,
      "step": 49315,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.43904
    },
    {
      "epoch": 2.113020007711752,
      "grad_norm": 8.206925392150879,
      "learning_rate": 6.203789769618365e-05,
      "loss": 2.491926574707031,
      "memory(GiB)": 77.56,
      "step": 49320,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.439009
    },
    {
      "epoch": 2.1132342230410006,
      "grad_norm": 5.43293571472168,
      "learning_rate": 6.203136576912529e-05,
      "loss": 2.3664140701293945,
      "memory(GiB)": 77.56,
      "step": 49325,
      "token_acc": 0.5490196078431373,
      "train_speed(iter/s)": 1.439008
    },
    {
      "epoch": 2.11344843837025,
      "grad_norm": 5.0613908767700195,
      "learning_rate": 6.202483362410748e-05,
      "loss": 2.609480285644531,
      "memory(GiB)": 77.56,
      "step": 49330,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.438993
    },
    {
      "epoch": 2.1136626536994987,
      "grad_norm": 5.422504425048828,
      "learning_rate": 6.20183012612485e-05,
      "loss": 2.3954341888427733,
      "memory(GiB)": 77.56,
      "step": 49335,
      "token_acc": 0.4576923076923077,
      "train_speed(iter/s)": 1.438998
    },
    {
      "epoch": 2.1138768690287475,
      "grad_norm": 4.421570777893066,
      "learning_rate": 6.201176868066674e-05,
      "loss": 2.3289791107177735,
      "memory(GiB)": 77.56,
      "step": 49340,
      "token_acc": 0.486013986013986,
      "train_speed(iter/s)": 1.439025
    },
    {
      "epoch": 2.1140910843579968,
      "grad_norm": 5.543737888336182,
      "learning_rate": 6.20052358824805e-05,
      "loss": 2.5759489059448244,
      "memory(GiB)": 77.56,
      "step": 49345,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.439019
    },
    {
      "epoch": 2.1143052996872456,
      "grad_norm": 6.646541595458984,
      "learning_rate": 6.199870286680817e-05,
      "loss": 2.3740303039550783,
      "memory(GiB)": 77.56,
      "step": 49350,
      "token_acc": 0.4977168949771689,
      "train_speed(iter/s)": 1.439042
    },
    {
      "epoch": 2.1145195150164944,
      "grad_norm": 5.940126419067383,
      "learning_rate": 6.199216963376806e-05,
      "loss": 2.2782978057861327,
      "memory(GiB)": 77.56,
      "step": 49355,
      "token_acc": 0.4970588235294118,
      "train_speed(iter/s)": 1.439068
    },
    {
      "epoch": 2.1147337303457436,
      "grad_norm": 5.637457847595215,
      "learning_rate": 6.198563618347857e-05,
      "loss": 2.5789794921875,
      "memory(GiB)": 77.56,
      "step": 49360,
      "token_acc": 0.45819397993311034,
      "train_speed(iter/s)": 1.43909
    },
    {
      "epoch": 2.1149479456749924,
      "grad_norm": 6.14561653137207,
      "learning_rate": 6.197910251605803e-05,
      "loss": 2.1523054122924803,
      "memory(GiB)": 77.56,
      "step": 49365,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.439064
    },
    {
      "epoch": 2.1151621610042413,
      "grad_norm": 5.5074286460876465,
      "learning_rate": 6.19725686316248e-05,
      "loss": 2.470820426940918,
      "memory(GiB)": 77.56,
      "step": 49370,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.439053
    },
    {
      "epoch": 2.1153763763334905,
      "grad_norm": 6.865657329559326,
      "learning_rate": 6.196603453029728e-05,
      "loss": 2.365422821044922,
      "memory(GiB)": 77.56,
      "step": 49375,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.439026
    },
    {
      "epoch": 2.1155905916627393,
      "grad_norm": 6.0406599044799805,
      "learning_rate": 6.19595002121938e-05,
      "loss": 2.446185493469238,
      "memory(GiB)": 77.56,
      "step": 49380,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.439008
    },
    {
      "epoch": 2.115804806991988,
      "grad_norm": 5.8405680656433105,
      "learning_rate": 6.195296567743277e-05,
      "loss": 2.678752899169922,
      "memory(GiB)": 77.56,
      "step": 49385,
      "token_acc": 0.48221343873517786,
      "train_speed(iter/s)": 1.439013
    },
    {
      "epoch": 2.1160190223212374,
      "grad_norm": 6.350005626678467,
      "learning_rate": 6.194643092613254e-05,
      "loss": 2.4396173477172853,
      "memory(GiB)": 77.56,
      "step": 49390,
      "token_acc": 0.45051194539249145,
      "train_speed(iter/s)": 1.43904
    },
    {
      "epoch": 2.116233237650486,
      "grad_norm": 5.962313175201416,
      "learning_rate": 6.193989595841151e-05,
      "loss": 2.490930366516113,
      "memory(GiB)": 77.56,
      "step": 49395,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.438995
    },
    {
      "epoch": 2.116447452979735,
      "grad_norm": 4.057485580444336,
      "learning_rate": 6.193336077438807e-05,
      "loss": 2.2414833068847657,
      "memory(GiB)": 77.56,
      "step": 49400,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.439013
    },
    {
      "epoch": 2.1166616683089843,
      "grad_norm": 6.552286624908447,
      "learning_rate": 6.192682537418061e-05,
      "loss": 2.437358856201172,
      "memory(GiB)": 77.56,
      "step": 49405,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.438998
    },
    {
      "epoch": 2.116875883638233,
      "grad_norm": 6.112018585205078,
      "learning_rate": 6.19202897579075e-05,
      "loss": 2.4001922607421875,
      "memory(GiB)": 77.56,
      "step": 49410,
      "token_acc": 0.52734375,
      "train_speed(iter/s)": 1.439001
    },
    {
      "epoch": 2.117090098967482,
      "grad_norm": 7.433084011077881,
      "learning_rate": 6.191375392568718e-05,
      "loss": 2.3160715103149414,
      "memory(GiB)": 77.56,
      "step": 49415,
      "token_acc": 0.5037878787878788,
      "train_speed(iter/s)": 1.439015
    },
    {
      "epoch": 2.117304314296731,
      "grad_norm": 5.40130615234375,
      "learning_rate": 6.190721787763801e-05,
      "loss": 2.536113166809082,
      "memory(GiB)": 77.56,
      "step": 49420,
      "token_acc": 0.4709480122324159,
      "train_speed(iter/s)": 1.439017
    },
    {
      "epoch": 2.11751852962598,
      "grad_norm": 5.7429280281066895,
      "learning_rate": 6.190068161387844e-05,
      "loss": 2.4757179260253905,
      "memory(GiB)": 77.56,
      "step": 49425,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.439046
    },
    {
      "epoch": 2.117732744955229,
      "grad_norm": 4.870671272277832,
      "learning_rate": 6.189414513452685e-05,
      "loss": 2.4686864852905273,
      "memory(GiB)": 77.56,
      "step": 49430,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.439083
    },
    {
      "epoch": 2.117946960284478,
      "grad_norm": 5.639690399169922,
      "learning_rate": 6.188760843970166e-05,
      "loss": 2.4599655151367186,
      "memory(GiB)": 77.56,
      "step": 49435,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.439084
    },
    {
      "epoch": 2.118161175613727,
      "grad_norm": 4.582322120666504,
      "learning_rate": 6.188107152952129e-05,
      "loss": 2.2679817199707033,
      "memory(GiB)": 77.56,
      "step": 49440,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.439066
    },
    {
      "epoch": 2.1183753909429757,
      "grad_norm": 5.99381160736084,
      "learning_rate": 6.187453440410418e-05,
      "loss": 2.381431770324707,
      "memory(GiB)": 77.56,
      "step": 49445,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.439078
    },
    {
      "epoch": 2.118589606272225,
      "grad_norm": 3.7383217811584473,
      "learning_rate": 6.186799706356872e-05,
      "loss": 2.377843475341797,
      "memory(GiB)": 77.56,
      "step": 49450,
      "token_acc": 0.4755043227665706,
      "train_speed(iter/s)": 1.439052
    },
    {
      "epoch": 2.1188038216014737,
      "grad_norm": 4.4830098152160645,
      "learning_rate": 6.186145950803337e-05,
      "loss": 2.103716278076172,
      "memory(GiB)": 77.56,
      "step": 49455,
      "token_acc": 0.5266903914590747,
      "train_speed(iter/s)": 1.439036
    },
    {
      "epoch": 2.1190180369307225,
      "grad_norm": 6.509287357330322,
      "learning_rate": 6.185492173761655e-05,
      "loss": 2.693792533874512,
      "memory(GiB)": 77.56,
      "step": 49460,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.439072
    },
    {
      "epoch": 2.119232252259972,
      "grad_norm": 6.000534534454346,
      "learning_rate": 6.184838375243671e-05,
      "loss": 2.4342653274536135,
      "memory(GiB)": 77.56,
      "step": 49465,
      "token_acc": 0.4981549815498155,
      "train_speed(iter/s)": 1.439063
    },
    {
      "epoch": 2.1194464675892206,
      "grad_norm": 4.785678863525391,
      "learning_rate": 6.184184555261227e-05,
      "loss": 2.715536689758301,
      "memory(GiB)": 77.56,
      "step": 49470,
      "token_acc": 0.4413793103448276,
      "train_speed(iter/s)": 1.438981
    },
    {
      "epoch": 2.11966068291847,
      "grad_norm": 6.359776020050049,
      "learning_rate": 6.18353071382617e-05,
      "loss": 2.6815032958984375,
      "memory(GiB)": 77.56,
      "step": 49475,
      "token_acc": 0.40498442367601245,
      "train_speed(iter/s)": 1.438998
    },
    {
      "epoch": 2.1198748982477187,
      "grad_norm": 3.8615801334381104,
      "learning_rate": 6.182876850950344e-05,
      "loss": 2.593334197998047,
      "memory(GiB)": 77.56,
      "step": 49480,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.438969
    },
    {
      "epoch": 2.1200891135769675,
      "grad_norm": 5.856441497802734,
      "learning_rate": 6.182222966645593e-05,
      "loss": 2.129928207397461,
      "memory(GiB)": 77.56,
      "step": 49485,
      "token_acc": 0.5451127819548872,
      "train_speed(iter/s)": 1.438958
    },
    {
      "epoch": 2.1203033289062168,
      "grad_norm": 4.4080424308776855,
      "learning_rate": 6.181569060923765e-05,
      "loss": 2.4936798095703123,
      "memory(GiB)": 77.56,
      "step": 49490,
      "token_acc": 0.5086505190311419,
      "train_speed(iter/s)": 1.438949
    },
    {
      "epoch": 2.1205175442354656,
      "grad_norm": 5.6301164627075195,
      "learning_rate": 6.180915133796705e-05,
      "loss": 2.132694053649902,
      "memory(GiB)": 77.56,
      "step": 49495,
      "token_acc": 0.5445859872611465,
      "train_speed(iter/s)": 1.438956
    },
    {
      "epoch": 2.1207317595647144,
      "grad_norm": 6.104939937591553,
      "learning_rate": 6.180261185276259e-05,
      "loss": 2.311728286743164,
      "memory(GiB)": 77.56,
      "step": 49500,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.438969
    },
    {
      "epoch": 2.1207317595647144,
      "eval_loss": 2.243361473083496,
      "eval_runtime": 14.0098,
      "eval_samples_per_second": 7.138,
      "eval_steps_per_second": 7.138,
      "eval_token_acc": 0.4934876989869754,
      "step": 49500
    },
    {
      "epoch": 2.1209459748939636,
      "grad_norm": 6.549467086791992,
      "learning_rate": 6.179607215374274e-05,
      "loss": 2.3752391815185545,
      "memory(GiB)": 77.56,
      "step": 49505,
      "token_acc": 0.4994652406417112,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 2.1211601902232124,
      "grad_norm": 5.327155590057373,
      "learning_rate": 6.178953224102599e-05,
      "loss": 2.3207958221435545,
      "memory(GiB)": 77.56,
      "step": 49510,
      "token_acc": 0.4623287671232877,
      "train_speed(iter/s)": 1.438342
    },
    {
      "epoch": 2.1213744055524613,
      "grad_norm": 6.941089153289795,
      "learning_rate": 6.178299211473081e-05,
      "loss": 2.448649215698242,
      "memory(GiB)": 77.56,
      "step": 49515,
      "token_acc": 0.5220588235294118,
      "train_speed(iter/s)": 1.438358
    },
    {
      "epoch": 2.1215886208817105,
      "grad_norm": 6.1966376304626465,
      "learning_rate": 6.177645177497566e-05,
      "loss": 2.5064371109008787,
      "memory(GiB)": 77.56,
      "step": 49520,
      "token_acc": 0.5153374233128835,
      "train_speed(iter/s)": 1.438369
    },
    {
      "epoch": 2.1218028362109593,
      "grad_norm": 6.111876010894775,
      "learning_rate": 6.176991122187904e-05,
      "loss": 2.3103778839111326,
      "memory(GiB)": 77.56,
      "step": 49525,
      "token_acc": 0.5398550724637681,
      "train_speed(iter/s)": 1.43832
    },
    {
      "epoch": 2.122017051540208,
      "grad_norm": 5.384545803070068,
      "learning_rate": 6.176337045555944e-05,
      "loss": 2.4785642623901367,
      "memory(GiB)": 77.56,
      "step": 49530,
      "token_acc": 0.5015384615384615,
      "train_speed(iter/s)": 1.438349
    },
    {
      "epoch": 2.1222312668694574,
      "grad_norm": 5.431389808654785,
      "learning_rate": 6.175682947613534e-05,
      "loss": 2.4664621353149414,
      "memory(GiB)": 77.56,
      "step": 49535,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.43837
    },
    {
      "epoch": 2.122445482198706,
      "grad_norm": 5.330932140350342,
      "learning_rate": 6.175028828372527e-05,
      "loss": 2.87606258392334,
      "memory(GiB)": 77.56,
      "step": 49540,
      "token_acc": 0.434640522875817,
      "train_speed(iter/s)": 1.438388
    },
    {
      "epoch": 2.122659697527955,
      "grad_norm": 7.038581371307373,
      "learning_rate": 6.174374687844769e-05,
      "loss": 2.4076940536499025,
      "memory(GiB)": 77.56,
      "step": 49545,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.438408
    },
    {
      "epoch": 2.1228739128572043,
      "grad_norm": 4.5212507247924805,
      "learning_rate": 6.173720526042112e-05,
      "loss": 2.0131912231445312,
      "memory(GiB)": 77.56,
      "step": 49550,
      "token_acc": 0.5261194029850746,
      "train_speed(iter/s)": 1.438396
    },
    {
      "epoch": 2.123088128186453,
      "grad_norm": 4.725074291229248,
      "learning_rate": 6.173066342976405e-05,
      "loss": 2.1782270431518556,
      "memory(GiB)": 77.56,
      "step": 49555,
      "token_acc": 0.5490909090909091,
      "train_speed(iter/s)": 1.438429
    },
    {
      "epoch": 2.123302343515702,
      "grad_norm": 8.704798698425293,
      "learning_rate": 6.172412138659504e-05,
      "loss": 2.2778594970703123,
      "memory(GiB)": 77.56,
      "step": 49560,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.438455
    },
    {
      "epoch": 2.123516558844951,
      "grad_norm": 6.381246089935303,
      "learning_rate": 6.171757913103255e-05,
      "loss": 2.361995887756348,
      "memory(GiB)": 77.56,
      "step": 49565,
      "token_acc": 0.4623287671232877,
      "train_speed(iter/s)": 1.438465
    },
    {
      "epoch": 2.1237307741742,
      "grad_norm": 5.788327217102051,
      "learning_rate": 6.171103666319514e-05,
      "loss": 2.3147878646850586,
      "memory(GiB)": 77.56,
      "step": 49570,
      "token_acc": 0.4808259587020649,
      "train_speed(iter/s)": 1.438503
    },
    {
      "epoch": 2.123944989503449,
      "grad_norm": 4.380969047546387,
      "learning_rate": 6.17044939832013e-05,
      "loss": 2.3820045471191404,
      "memory(GiB)": 77.56,
      "step": 49575,
      "token_acc": 0.5080906148867314,
      "train_speed(iter/s)": 1.438477
    },
    {
      "epoch": 2.124159204832698,
      "grad_norm": 5.585661888122559,
      "learning_rate": 6.169795109116957e-05,
      "loss": 2.5776718139648436,
      "memory(GiB)": 77.56,
      "step": 49580,
      "token_acc": 0.46923076923076923,
      "train_speed(iter/s)": 1.438464
    },
    {
      "epoch": 2.124373420161947,
      "grad_norm": 7.426137924194336,
      "learning_rate": 6.169140798721847e-05,
      "loss": 2.4600198745727537,
      "memory(GiB)": 77.56,
      "step": 49585,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.438492
    },
    {
      "epoch": 2.1245876354911957,
      "grad_norm": 4.669302940368652,
      "learning_rate": 6.168486467146658e-05,
      "loss": 2.5644407272338867,
      "memory(GiB)": 77.56,
      "step": 49590,
      "token_acc": 0.4520547945205479,
      "train_speed(iter/s)": 1.438441
    },
    {
      "epoch": 2.124801850820445,
      "grad_norm": 5.455381393432617,
      "learning_rate": 6.167832114403238e-05,
      "loss": 2.2322168350219727,
      "memory(GiB)": 77.56,
      "step": 49595,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.438465
    },
    {
      "epoch": 2.1250160661496937,
      "grad_norm": 5.240556240081787,
      "learning_rate": 6.167177740503444e-05,
      "loss": 2.333906364440918,
      "memory(GiB)": 77.56,
      "step": 49600,
      "token_acc": 0.519298245614035,
      "train_speed(iter/s)": 1.438493
    },
    {
      "epoch": 2.1252302814789426,
      "grad_norm": 4.892724514007568,
      "learning_rate": 6.166523345459132e-05,
      "loss": 2.451043701171875,
      "memory(GiB)": 77.56,
      "step": 49605,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.438495
    },
    {
      "epoch": 2.125444496808192,
      "grad_norm": 5.030633449554443,
      "learning_rate": 6.165868929282155e-05,
      "loss": 2.3205389022827148,
      "memory(GiB)": 77.56,
      "step": 49610,
      "token_acc": 0.4824561403508772,
      "train_speed(iter/s)": 1.438474
    },
    {
      "epoch": 2.1256587121374406,
      "grad_norm": 4.008750915527344,
      "learning_rate": 6.165214491984367e-05,
      "loss": 2.527475929260254,
      "memory(GiB)": 77.56,
      "step": 49615,
      "token_acc": 0.46756756756756757,
      "train_speed(iter/s)": 1.438483
    },
    {
      "epoch": 2.1258729274666894,
      "grad_norm": 4.7009687423706055,
      "learning_rate": 6.164560033577626e-05,
      "loss": 2.470628547668457,
      "memory(GiB)": 77.56,
      "step": 49620,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.438497
    },
    {
      "epoch": 2.1260871427959387,
      "grad_norm": 4.111307621002197,
      "learning_rate": 6.163905554073787e-05,
      "loss": 2.073130416870117,
      "memory(GiB)": 77.56,
      "step": 49625,
      "token_acc": 0.5418326693227091,
      "train_speed(iter/s)": 1.438493
    },
    {
      "epoch": 2.1263013581251875,
      "grad_norm": 6.109747886657715,
      "learning_rate": 6.16325105348471e-05,
      "loss": 2.5180421829223634,
      "memory(GiB)": 77.56,
      "step": 49630,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.438515
    },
    {
      "epoch": 2.1265155734544363,
      "grad_norm": 5.7932538986206055,
      "learning_rate": 6.162596531822247e-05,
      "loss": 2.5330270767211913,
      "memory(GiB)": 77.56,
      "step": 49635,
      "token_acc": 0.434375,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 2.1267297887836856,
      "grad_norm": 5.666923999786377,
      "learning_rate": 6.161941989098256e-05,
      "loss": 2.4763904571533204,
      "memory(GiB)": 77.56,
      "step": 49640,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.438489
    },
    {
      "epoch": 2.1269440041129344,
      "grad_norm": 4.667091369628906,
      "learning_rate": 6.161287425324597e-05,
      "loss": 2.394595146179199,
      "memory(GiB)": 77.56,
      "step": 49645,
      "token_acc": 0.48589341692789967,
      "train_speed(iter/s)": 1.438525
    },
    {
      "epoch": 2.127158219442183,
      "grad_norm": 5.370915412902832,
      "learning_rate": 6.160632840513127e-05,
      "loss": 2.3501203536987303,
      "memory(GiB)": 77.56,
      "step": 49650,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.438525
    },
    {
      "epoch": 2.1273724347714325,
      "grad_norm": 5.324522018432617,
      "learning_rate": 6.159978234675704e-05,
      "loss": 2.243113136291504,
      "memory(GiB)": 77.56,
      "step": 49655,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438543
    },
    {
      "epoch": 2.1275866501006813,
      "grad_norm": 5.306806564331055,
      "learning_rate": 6.159323607824188e-05,
      "loss": 2.176255798339844,
      "memory(GiB)": 77.56,
      "step": 49660,
      "token_acc": 0.5282258064516129,
      "train_speed(iter/s)": 1.438496
    },
    {
      "epoch": 2.12780086542993,
      "grad_norm": 5.817471027374268,
      "learning_rate": 6.158668959970437e-05,
      "loss": 2.46315975189209,
      "memory(GiB)": 77.56,
      "step": 49665,
      "token_acc": 0.48348348348348347,
      "train_speed(iter/s)": 1.438487
    },
    {
      "epoch": 2.1280150807591793,
      "grad_norm": 4.693881511688232,
      "learning_rate": 6.158014291126311e-05,
      "loss": 2.2356786727905273,
      "memory(GiB)": 77.56,
      "step": 49670,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.438474
    },
    {
      "epoch": 2.128229296088428,
      "grad_norm": 5.200475215911865,
      "learning_rate": 6.15735960130367e-05,
      "loss": 2.420536422729492,
      "memory(GiB)": 77.56,
      "step": 49675,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.438492
    },
    {
      "epoch": 2.128443511417677,
      "grad_norm": 5.993760108947754,
      "learning_rate": 6.156704890514372e-05,
      "loss": 2.2368011474609375,
      "memory(GiB)": 77.56,
      "step": 49680,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.438526
    },
    {
      "epoch": 2.128657726746926,
      "grad_norm": 5.373075008392334,
      "learning_rate": 6.156050158770282e-05,
      "loss": 2.2701204299926756,
      "memory(GiB)": 77.56,
      "step": 49685,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.438545
    },
    {
      "epoch": 2.128871942076175,
      "grad_norm": 3.6746273040771484,
      "learning_rate": 6.155395406083257e-05,
      "loss": 2.7333545684814453,
      "memory(GiB)": 77.56,
      "step": 49690,
      "token_acc": 0.4293785310734463,
      "train_speed(iter/s)": 1.43851
    },
    {
      "epoch": 2.129086157405424,
      "grad_norm": 4.016968250274658,
      "learning_rate": 6.154740632465162e-05,
      "loss": 2.507294464111328,
      "memory(GiB)": 77.56,
      "step": 49695,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.438507
    },
    {
      "epoch": 2.129300372734673,
      "grad_norm": 6.886704444885254,
      "learning_rate": 6.154085837927857e-05,
      "loss": 2.233457565307617,
      "memory(GiB)": 77.56,
      "step": 49700,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.438516
    },
    {
      "epoch": 2.129514588063922,
      "grad_norm": 4.527353286743164,
      "learning_rate": 6.153431022483205e-05,
      "loss": 2.5117422103881837,
      "memory(GiB)": 77.56,
      "step": 49705,
      "token_acc": 0.5053763440860215,
      "train_speed(iter/s)": 1.438547
    },
    {
      "epoch": 2.1297288033931707,
      "grad_norm": 6.1918625831604,
      "learning_rate": 6.152776186143067e-05,
      "loss": 2.406984329223633,
      "memory(GiB)": 77.56,
      "step": 49710,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.438546
    },
    {
      "epoch": 2.12994301872242,
      "grad_norm": 4.611532688140869,
      "learning_rate": 6.152121328919307e-05,
      "loss": 2.220952606201172,
      "memory(GiB)": 77.56,
      "step": 49715,
      "token_acc": 0.5344129554655871,
      "train_speed(iter/s)": 1.438572
    },
    {
      "epoch": 2.130157234051669,
      "grad_norm": 4.31071138381958,
      "learning_rate": 6.15146645082379e-05,
      "loss": 2.1495359420776365,
      "memory(GiB)": 77.56,
      "step": 49720,
      "token_acc": 0.5106382978723404,
      "train_speed(iter/s)": 1.438562
    },
    {
      "epoch": 2.1303714493809176,
      "grad_norm": 6.270028114318848,
      "learning_rate": 6.150811551868377e-05,
      "loss": 2.477192687988281,
      "memory(GiB)": 77.56,
      "step": 49725,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438544
    },
    {
      "epoch": 2.130585664710167,
      "grad_norm": 7.548161506652832,
      "learning_rate": 6.15015663206493e-05,
      "loss": 2.397829055786133,
      "memory(GiB)": 77.56,
      "step": 49730,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.438569
    },
    {
      "epoch": 2.1307998800394157,
      "grad_norm": 5.996013641357422,
      "learning_rate": 6.149501691425321e-05,
      "loss": 2.490370178222656,
      "memory(GiB)": 77.56,
      "step": 49735,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.438567
    },
    {
      "epoch": 2.1310140953686645,
      "grad_norm": 5.669908046722412,
      "learning_rate": 6.148846729961409e-05,
      "loss": 2.2922748565673827,
      "memory(GiB)": 77.56,
      "step": 49740,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.438574
    },
    {
      "epoch": 2.1312283106979137,
      "grad_norm": 5.151236534118652,
      "learning_rate": 6.148191747685061e-05,
      "loss": 2.2669656753540037,
      "memory(GiB)": 77.56,
      "step": 49745,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.438589
    },
    {
      "epoch": 2.1314425260271626,
      "grad_norm": 5.144192695617676,
      "learning_rate": 6.147536744608143e-05,
      "loss": 2.5211219787597656,
      "memory(GiB)": 77.56,
      "step": 49750,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.438557
    },
    {
      "epoch": 2.1316567413564114,
      "grad_norm": 4.636655807495117,
      "learning_rate": 6.146881720742519e-05,
      "loss": 2.3756664276123045,
      "memory(GiB)": 77.56,
      "step": 49755,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.43855
    },
    {
      "epoch": 2.1318709566856606,
      "grad_norm": 5.384429454803467,
      "learning_rate": 6.146226676100058e-05,
      "loss": 2.387874794006348,
      "memory(GiB)": 77.56,
      "step": 49760,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.438536
    },
    {
      "epoch": 2.1320851720149094,
      "grad_norm": 6.880685806274414,
      "learning_rate": 6.145571610692624e-05,
      "loss": 2.2560420989990235,
      "memory(GiB)": 77.56,
      "step": 49765,
      "token_acc": 0.5233333333333333,
      "train_speed(iter/s)": 1.438511
    },
    {
      "epoch": 2.1322993873441582,
      "grad_norm": 7.248161792755127,
      "learning_rate": 6.144916524532086e-05,
      "loss": 2.4489181518554686,
      "memory(GiB)": 77.56,
      "step": 49770,
      "token_acc": 0.47244094488188976,
      "train_speed(iter/s)": 1.438508
    },
    {
      "epoch": 2.1325136026734075,
      "grad_norm": 5.082353591918945,
      "learning_rate": 6.144261417630313e-05,
      "loss": 2.3674482345581054,
      "memory(GiB)": 77.56,
      "step": 49775,
      "token_acc": 0.512987012987013,
      "train_speed(iter/s)": 1.438538
    },
    {
      "epoch": 2.1327278180026563,
      "grad_norm": 5.5833892822265625,
      "learning_rate": 6.143606289999169e-05,
      "loss": 2.5191928863525392,
      "memory(GiB)": 77.56,
      "step": 49780,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.132942033331905,
      "grad_norm": 4.669332504272461,
      "learning_rate": 6.142951141650527e-05,
      "loss": 2.293395233154297,
      "memory(GiB)": 77.56,
      "step": 49785,
      "token_acc": 0.5202702702702703,
      "train_speed(iter/s)": 1.438561
    },
    {
      "epoch": 2.1331562486611544,
      "grad_norm": 5.875611782073975,
      "learning_rate": 6.14229597259625e-05,
      "loss": 2.6193126678466796,
      "memory(GiB)": 77.56,
      "step": 49790,
      "token_acc": 0.4772117962466488,
      "train_speed(iter/s)": 1.438588
    },
    {
      "epoch": 2.133370463990403,
      "grad_norm": 4.320123195648193,
      "learning_rate": 6.141640782848211e-05,
      "loss": 2.146860122680664,
      "memory(GiB)": 77.56,
      "step": 49795,
      "token_acc": 0.5634920634920635,
      "train_speed(iter/s)": 1.438616
    },
    {
      "epoch": 2.133584679319652,
      "grad_norm": 5.646020889282227,
      "learning_rate": 6.140985572418276e-05,
      "loss": 2.3647481918334963,
      "memory(GiB)": 77.56,
      "step": 49800,
      "token_acc": 0.4854014598540146,
      "train_speed(iter/s)": 1.438618
    },
    {
      "epoch": 2.1337988946489013,
      "grad_norm": 6.506009101867676,
      "learning_rate": 6.14033034131832e-05,
      "loss": 2.6079477310180663,
      "memory(GiB)": 77.56,
      "step": 49805,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.438647
    },
    {
      "epoch": 2.13401310997815,
      "grad_norm": 5.5941853523254395,
      "learning_rate": 6.13967508956021e-05,
      "loss": 2.3892213821411135,
      "memory(GiB)": 77.56,
      "step": 49810,
      "token_acc": 0.4837758112094395,
      "train_speed(iter/s)": 1.438648
    },
    {
      "epoch": 2.134227325307399,
      "grad_norm": 4.989762306213379,
      "learning_rate": 6.139019817155815e-05,
      "loss": 2.0783016204833986,
      "memory(GiB)": 77.56,
      "step": 49815,
      "token_acc": 0.5133333333333333,
      "train_speed(iter/s)": 1.438676
    },
    {
      "epoch": 2.134441540636648,
      "grad_norm": 6.979779243469238,
      "learning_rate": 6.13836452411701e-05,
      "loss": 2.5517240524291993,
      "memory(GiB)": 77.56,
      "step": 49820,
      "token_acc": 0.45724907063197023,
      "train_speed(iter/s)": 1.438704
    },
    {
      "epoch": 2.134655755965897,
      "grad_norm": 5.5464582443237305,
      "learning_rate": 6.137709210455661e-05,
      "loss": 2.4006580352783202,
      "memory(GiB)": 77.56,
      "step": 49825,
      "token_acc": 0.5124653739612188,
      "train_speed(iter/s)": 1.438699
    },
    {
      "epoch": 2.1348699712951458,
      "grad_norm": 6.28291130065918,
      "learning_rate": 6.137053876183644e-05,
      "loss": 2.1528804779052733,
      "memory(GiB)": 77.56,
      "step": 49830,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438698
    },
    {
      "epoch": 2.135084186624395,
      "grad_norm": 5.401803493499756,
      "learning_rate": 6.136398521312829e-05,
      "loss": 2.4813947677612305,
      "memory(GiB)": 77.56,
      "step": 49835,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.135298401953644,
      "grad_norm": 6.829617977142334,
      "learning_rate": 6.135743145855088e-05,
      "loss": 2.7040019989013673,
      "memory(GiB)": 77.56,
      "step": 49840,
      "token_acc": 0.46048109965635736,
      "train_speed(iter/s)": 1.438646
    },
    {
      "epoch": 2.1355126172828927,
      "grad_norm": 6.5611138343811035,
      "learning_rate": 6.135087749822296e-05,
      "loss": 2.496912384033203,
      "memory(GiB)": 77.56,
      "step": 49845,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.438662
    },
    {
      "epoch": 2.135726832612142,
      "grad_norm": 4.983579158782959,
      "learning_rate": 6.134432333226324e-05,
      "loss": 2.4228458404541016,
      "memory(GiB)": 77.56,
      "step": 49850,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.438684
    },
    {
      "epoch": 2.1359410479413907,
      "grad_norm": 7.406137943267822,
      "learning_rate": 6.133776896079045e-05,
      "loss": 2.5163949966430663,
      "memory(GiB)": 77.56,
      "step": 49855,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.438704
    },
    {
      "epoch": 2.1361552632706395,
      "grad_norm": 6.890848159790039,
      "learning_rate": 6.133121438392336e-05,
      "loss": 2.358189010620117,
      "memory(GiB)": 77.56,
      "step": 49860,
      "token_acc": 0.4790874524714829,
      "train_speed(iter/s)": 1.438727
    },
    {
      "epoch": 2.136369478599889,
      "grad_norm": 5.278731822967529,
      "learning_rate": 6.132465960178069e-05,
      "loss": 2.1852678298950194,
      "memory(GiB)": 77.56,
      "step": 49865,
      "token_acc": 0.5409252669039146,
      "train_speed(iter/s)": 1.438713
    },
    {
      "epoch": 2.1365836939291376,
      "grad_norm": 6.262353420257568,
      "learning_rate": 6.131810461448118e-05,
      "loss": 2.576774787902832,
      "memory(GiB)": 77.56,
      "step": 49870,
      "token_acc": 0.5036764705882353,
      "train_speed(iter/s)": 1.438726
    },
    {
      "epoch": 2.1367979092583864,
      "grad_norm": 6.995584011077881,
      "learning_rate": 6.131154942214356e-05,
      "loss": 2.5000476837158203,
      "memory(GiB)": 77.56,
      "step": 49875,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.438716
    },
    {
      "epoch": 2.1370121245876357,
      "grad_norm": 5.0994343757629395,
      "learning_rate": 6.130499402488665e-05,
      "loss": 2.340886116027832,
      "memory(GiB)": 77.56,
      "step": 49880,
      "token_acc": 0.519163763066202,
      "train_speed(iter/s)": 1.438685
    },
    {
      "epoch": 2.1372263399168845,
      "grad_norm": 3.9778294563293457,
      "learning_rate": 6.129843842282915e-05,
      "loss": 2.0544652938842773,
      "memory(GiB)": 77.56,
      "step": 49885,
      "token_acc": 0.5673469387755102,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 2.1374405552461333,
      "grad_norm": 4.39173698425293,
      "learning_rate": 6.129188261608985e-05,
      "loss": 2.2172895431518556,
      "memory(GiB)": 77.56,
      "step": 49890,
      "token_acc": 0.5399239543726235,
      "train_speed(iter/s)": 1.438672
    },
    {
      "epoch": 2.1376547705753826,
      "grad_norm": 6.487553119659424,
      "learning_rate": 6.12853266047875e-05,
      "loss": 2.509355354309082,
      "memory(GiB)": 77.56,
      "step": 49895,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 2.1378689859046314,
      "grad_norm": 5.5855488777160645,
      "learning_rate": 6.127877038904087e-05,
      "loss": 2.5908069610595703,
      "memory(GiB)": 77.56,
      "step": 49900,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.438651
    },
    {
      "epoch": 2.13808320123388,
      "grad_norm": 6.047486305236816,
      "learning_rate": 6.127221396896876e-05,
      "loss": 2.8822341918945313,
      "memory(GiB)": 77.56,
      "step": 49905,
      "token_acc": 0.4623955431754875,
      "train_speed(iter/s)": 1.438647
    },
    {
      "epoch": 2.1382974165631294,
      "grad_norm": 5.487802028656006,
      "learning_rate": 6.126565734468987e-05,
      "loss": 2.446123504638672,
      "memory(GiB)": 77.56,
      "step": 49910,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.438655
    },
    {
      "epoch": 2.1385116318923783,
      "grad_norm": 5.983086585998535,
      "learning_rate": 6.125910051632305e-05,
      "loss": 2.622086524963379,
      "memory(GiB)": 77.56,
      "step": 49915,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.138725847221627,
      "grad_norm": 4.839722633361816,
      "learning_rate": 6.125254348398708e-05,
      "loss": 2.53487548828125,
      "memory(GiB)": 77.56,
      "step": 49920,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438665
    },
    {
      "epoch": 2.1389400625508763,
      "grad_norm": 5.114991188049316,
      "learning_rate": 6.124598624780071e-05,
      "loss": 2.6296823501586912,
      "memory(GiB)": 77.56,
      "step": 49925,
      "token_acc": 0.42382271468144045,
      "train_speed(iter/s)": 1.438671
    },
    {
      "epoch": 2.139154277880125,
      "grad_norm": 5.7839837074279785,
      "learning_rate": 6.123942880788276e-05,
      "loss": 2.483086585998535,
      "memory(GiB)": 77.56,
      "step": 49930,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.438701
    },
    {
      "epoch": 2.139368493209374,
      "grad_norm": 4.412889003753662,
      "learning_rate": 6.123287116435201e-05,
      "loss": 2.650364875793457,
      "memory(GiB)": 77.56,
      "step": 49935,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.438735
    },
    {
      "epoch": 2.139582708538623,
      "grad_norm": 6.163586616516113,
      "learning_rate": 6.122631331732726e-05,
      "loss": 2.525053024291992,
      "memory(GiB)": 77.56,
      "step": 49940,
      "token_acc": 0.4525316455696203,
      "train_speed(iter/s)": 1.438765
    },
    {
      "epoch": 2.139796923867872,
      "grad_norm": 6.026120662689209,
      "learning_rate": 6.121975526692731e-05,
      "loss": 2.1024471282958985,
      "memory(GiB)": 77.56,
      "step": 49945,
      "token_acc": 0.5707762557077626,
      "train_speed(iter/s)": 1.438785
    },
    {
      "epoch": 2.140011139197121,
      "grad_norm": 5.671998977661133,
      "learning_rate": 6.121319701327097e-05,
      "loss": 2.443772315979004,
      "memory(GiB)": 77.56,
      "step": 49950,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.438811
    },
    {
      "epoch": 2.14022535452637,
      "grad_norm": 5.431333065032959,
      "learning_rate": 6.120663855647706e-05,
      "loss": 2.2522510528564452,
      "memory(GiB)": 77.56,
      "step": 49955,
      "token_acc": 0.5036764705882353,
      "train_speed(iter/s)": 1.438827
    },
    {
      "epoch": 2.140439569855619,
      "grad_norm": 6.220781326293945,
      "learning_rate": 6.120007989666437e-05,
      "loss": 2.1643314361572266,
      "memory(GiB)": 77.56,
      "step": 49960,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.438809
    },
    {
      "epoch": 2.1406537851848677,
      "grad_norm": 6.45214319229126,
      "learning_rate": 6.119352103395172e-05,
      "loss": 2.625122833251953,
      "memory(GiB)": 77.56,
      "step": 49965,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.438782
    },
    {
      "epoch": 2.140868000514117,
      "grad_norm": 5.414968967437744,
      "learning_rate": 6.118696196845793e-05,
      "loss": 2.4727462768554687,
      "memory(GiB)": 77.56,
      "step": 49970,
      "token_acc": 0.4452296819787986,
      "train_speed(iter/s)": 1.438804
    },
    {
      "epoch": 2.1410822158433658,
      "grad_norm": 4.752712726593018,
      "learning_rate": 6.118040270030185e-05,
      "loss": 2.5828969955444334,
      "memory(GiB)": 77.56,
      "step": 49975,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.438824
    },
    {
      "epoch": 2.1412964311726146,
      "grad_norm": 4.766554355621338,
      "learning_rate": 6.117384322960228e-05,
      "loss": 2.353059005737305,
      "memory(GiB)": 77.56,
      "step": 49980,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.438827
    },
    {
      "epoch": 2.141510646501864,
      "grad_norm": 5.561649322509766,
      "learning_rate": 6.116728355647805e-05,
      "loss": 2.318983459472656,
      "memory(GiB)": 77.56,
      "step": 49985,
      "token_acc": 0.484,
      "train_speed(iter/s)": 1.438846
    },
    {
      "epoch": 2.1417248618311127,
      "grad_norm": 4.975546836853027,
      "learning_rate": 6.1160723681048e-05,
      "loss": 2.4256845474243165,
      "memory(GiB)": 77.56,
      "step": 49990,
      "token_acc": 0.4948805460750853,
      "train_speed(iter/s)": 1.438847
    },
    {
      "epoch": 2.1419390771603615,
      "grad_norm": 4.863134384155273,
      "learning_rate": 6.115416360343099e-05,
      "loss": 2.343097686767578,
      "memory(GiB)": 77.56,
      "step": 49995,
      "token_acc": 0.5350553505535055,
      "train_speed(iter/s)": 1.438854
    },
    {
      "epoch": 2.1421532924896107,
      "grad_norm": 5.984571933746338,
      "learning_rate": 6.114760332374582e-05,
      "loss": 2.2845407485961915,
      "memory(GiB)": 77.56,
      "step": 50000,
      "token_acc": 0.5310077519379846,
      "train_speed(iter/s)": 1.438858
    },
    {
      "epoch": 2.1421532924896107,
      "eval_loss": 2.2455437183380127,
      "eval_runtime": 15.7782,
      "eval_samples_per_second": 6.338,
      "eval_steps_per_second": 6.338,
      "eval_token_acc": 0.4928571428571429,
      "step": 50000
    },
    {
      "epoch": 2.1423675078188595,
      "grad_norm": 5.6391921043396,
      "learning_rate": 6.114104284211139e-05,
      "loss": 2.2938869476318358,
      "memory(GiB)": 77.56,
      "step": 50005,
      "token_acc": 0.49537512846865367,
      "train_speed(iter/s)": 1.438212
    },
    {
      "epoch": 2.1425817231481084,
      "grad_norm": 7.031339168548584,
      "learning_rate": 6.11344821586465e-05,
      "loss": 2.5754207611083983,
      "memory(GiB)": 77.56,
      "step": 50010,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 2.1427959384773576,
      "grad_norm": 5.62471866607666,
      "learning_rate": 6.112792127347001e-05,
      "loss": 2.482496452331543,
      "memory(GiB)": 77.56,
      "step": 50015,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.438206
    },
    {
      "epoch": 2.1430101538066064,
      "grad_norm": 4.942071914672852,
      "learning_rate": 6.112136018670079e-05,
      "loss": 2.132308578491211,
      "memory(GiB)": 77.56,
      "step": 50020,
      "token_acc": 0.5223367697594502,
      "train_speed(iter/s)": 1.438202
    },
    {
      "epoch": 2.1432243691358552,
      "grad_norm": 4.971907138824463,
      "learning_rate": 6.111479889845772e-05,
      "loss": 2.2501907348632812,
      "memory(GiB)": 77.56,
      "step": 50025,
      "token_acc": 0.515748031496063,
      "train_speed(iter/s)": 1.438185
    },
    {
      "epoch": 2.1434385844651045,
      "grad_norm": 11.107024192810059,
      "learning_rate": 6.110823740885962e-05,
      "loss": 2.597299003601074,
      "memory(GiB)": 77.56,
      "step": 50030,
      "token_acc": 0.4984894259818731,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 2.1436527997943533,
      "grad_norm": 5.977148532867432,
      "learning_rate": 6.110167571802538e-05,
      "loss": 2.653176116943359,
      "memory(GiB)": 77.56,
      "step": 50035,
      "token_acc": 0.44108761329305135,
      "train_speed(iter/s)": 1.43817
    },
    {
      "epoch": 2.143867015123602,
      "grad_norm": 6.594445705413818,
      "learning_rate": 6.109511382607388e-05,
      "loss": 2.425271415710449,
      "memory(GiB)": 77.56,
      "step": 50040,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.438165
    },
    {
      "epoch": 2.1440812304528514,
      "grad_norm": 5.630459308624268,
      "learning_rate": 6.108855173312397e-05,
      "loss": 2.4234024047851563,
      "memory(GiB)": 77.56,
      "step": 50045,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.438168
    },
    {
      "epoch": 2.1442954457821,
      "grad_norm": 5.08636474609375,
      "learning_rate": 6.108198943929457e-05,
      "loss": 2.39666633605957,
      "memory(GiB)": 77.56,
      "step": 50050,
      "token_acc": 0.5306859205776173,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 2.144509661111349,
      "grad_norm": 4.75391149520874,
      "learning_rate": 6.107542694470452e-05,
      "loss": 2.2972488403320312,
      "memory(GiB)": 77.56,
      "step": 50055,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.438133
    },
    {
      "epoch": 2.1447238764405983,
      "grad_norm": 5.415741920471191,
      "learning_rate": 6.10688642494727e-05,
      "loss": 2.722870635986328,
      "memory(GiB)": 77.56,
      "step": 50060,
      "token_acc": 0.46458923512747874,
      "train_speed(iter/s)": 1.438167
    },
    {
      "epoch": 2.144938091769847,
      "grad_norm": 5.590670108795166,
      "learning_rate": 6.106230135371804e-05,
      "loss": 2.4435903549194338,
      "memory(GiB)": 77.56,
      "step": 50065,
      "token_acc": 0.4941860465116279,
      "train_speed(iter/s)": 1.438164
    },
    {
      "epoch": 2.145152307099096,
      "grad_norm": 5.37048864364624,
      "learning_rate": 6.105573825755942e-05,
      "loss": 2.6174755096435547,
      "memory(GiB)": 77.56,
      "step": 50070,
      "token_acc": 0.4641975308641975,
      "train_speed(iter/s)": 1.438178
    },
    {
      "epoch": 2.145366522428345,
      "grad_norm": 4.424235820770264,
      "learning_rate": 6.104917496111574e-05,
      "loss": 2.1999488830566407,
      "memory(GiB)": 77.56,
      "step": 50075,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438176
    },
    {
      "epoch": 2.145580737757594,
      "grad_norm": 5.697660446166992,
      "learning_rate": 6.104261146450588e-05,
      "loss": 2.5041751861572266,
      "memory(GiB)": 77.56,
      "step": 50080,
      "token_acc": 0.4738372093023256,
      "train_speed(iter/s)": 1.438163
    },
    {
      "epoch": 2.1457949530868428,
      "grad_norm": 5.865833282470703,
      "learning_rate": 6.103604776784872e-05,
      "loss": 2.5066795349121094,
      "memory(GiB)": 77.56,
      "step": 50085,
      "token_acc": 0.46496815286624205,
      "train_speed(iter/s)": 1.438146
    },
    {
      "epoch": 2.146009168416092,
      "grad_norm": 5.480571269989014,
      "learning_rate": 6.102948387126325e-05,
      "loss": 2.1818000793457033,
      "memory(GiB)": 77.56,
      "step": 50090,
      "token_acc": 0.512280701754386,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 2.146223383745341,
      "grad_norm": 6.711512565612793,
      "learning_rate": 6.10229197748683e-05,
      "loss": 2.623927688598633,
      "memory(GiB)": 77.56,
      "step": 50095,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.43817
    },
    {
      "epoch": 2.1464375990745896,
      "grad_norm": 4.391927719116211,
      "learning_rate": 6.101635547878285e-05,
      "loss": 2.3694786071777343,
      "memory(GiB)": 77.56,
      "step": 50100,
      "token_acc": 0.46112600536193027,
      "train_speed(iter/s)": 1.438212
    },
    {
      "epoch": 2.146651814403839,
      "grad_norm": 5.752460956573486,
      "learning_rate": 6.100979098312576e-05,
      "loss": 2.518646240234375,
      "memory(GiB)": 77.56,
      "step": 50105,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.438243
    },
    {
      "epoch": 2.1468660297330877,
      "grad_norm": 5.81132173538208,
      "learning_rate": 6.100322628801599e-05,
      "loss": 2.3162628173828126,
      "memory(GiB)": 77.56,
      "step": 50110,
      "token_acc": 0.5149253731343284,
      "train_speed(iter/s)": 1.43828
    },
    {
      "epoch": 2.1470802450623365,
      "grad_norm": 5.15714168548584,
      "learning_rate": 6.0996661393572454e-05,
      "loss": 2.4816158294677733,
      "memory(GiB)": 77.56,
      "step": 50115,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 2.147294460391586,
      "grad_norm": 9.179927825927734,
      "learning_rate": 6.099009629991408e-05,
      "loss": 2.149810791015625,
      "memory(GiB)": 77.56,
      "step": 50120,
      "token_acc": 0.5665236051502146,
      "train_speed(iter/s)": 1.438319
    },
    {
      "epoch": 2.1475086757208346,
      "grad_norm": 5.6330695152282715,
      "learning_rate": 6.098353100715981e-05,
      "loss": 2.1747529983520506,
      "memory(GiB)": 77.56,
      "step": 50125,
      "token_acc": 0.5053763440860215,
      "train_speed(iter/s)": 1.438327
    },
    {
      "epoch": 2.1477228910500834,
      "grad_norm": 4.714677333831787,
      "learning_rate": 6.0976965515428554e-05,
      "loss": 2.393399238586426,
      "memory(GiB)": 77.56,
      "step": 50130,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.438341
    },
    {
      "epoch": 2.1479371063793327,
      "grad_norm": 4.813422203063965,
      "learning_rate": 6.097039982483927e-05,
      "loss": 2.2043251037597655,
      "memory(GiB)": 77.56,
      "step": 50135,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.438363
    },
    {
      "epoch": 2.1481513217085815,
      "grad_norm": 5.016312122344971,
      "learning_rate": 6.0963833935510916e-05,
      "loss": 2.3490016937255858,
      "memory(GiB)": 77.56,
      "step": 50140,
      "token_acc": 0.49466192170818507,
      "train_speed(iter/s)": 1.438369
    },
    {
      "epoch": 2.1483655370378303,
      "grad_norm": 4.861772060394287,
      "learning_rate": 6.0957267847562414e-05,
      "loss": 2.557559776306152,
      "memory(GiB)": 77.56,
      "step": 50145,
      "token_acc": 0.4553846153846154,
      "train_speed(iter/s)": 1.43839
    },
    {
      "epoch": 2.1485797523670795,
      "grad_norm": 4.889413356781006,
      "learning_rate": 6.095070156111274e-05,
      "loss": 2.385935401916504,
      "memory(GiB)": 77.56,
      "step": 50150,
      "token_acc": 0.4421052631578947,
      "train_speed(iter/s)": 1.438378
    },
    {
      "epoch": 2.1487939676963284,
      "grad_norm": 5.498384475708008,
      "learning_rate": 6.094413507628084e-05,
      "loss": 2.437391662597656,
      "memory(GiB)": 77.56,
      "step": 50155,
      "token_acc": 0.4733893557422969,
      "train_speed(iter/s)": 1.438388
    },
    {
      "epoch": 2.149008183025577,
      "grad_norm": 5.706067085266113,
      "learning_rate": 6.093756839318565e-05,
      "loss": 2.2755405426025392,
      "memory(GiB)": 77.56,
      "step": 50160,
      "token_acc": 0.5234657039711191,
      "train_speed(iter/s)": 1.438411
    },
    {
      "epoch": 2.1492223983548264,
      "grad_norm": 6.289199352264404,
      "learning_rate": 6.093100151194615e-05,
      "loss": 2.3919721603393556,
      "memory(GiB)": 77.56,
      "step": 50165,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.438425
    },
    {
      "epoch": 2.1494366136840752,
      "grad_norm": 5.646265029907227,
      "learning_rate": 6.09244344326813e-05,
      "loss": 2.469646453857422,
      "memory(GiB)": 77.56,
      "step": 50170,
      "token_acc": 0.49390243902439024,
      "train_speed(iter/s)": 1.438412
    },
    {
      "epoch": 2.149650829013324,
      "grad_norm": 5.634893894195557,
      "learning_rate": 6.091786715551008e-05,
      "loss": 2.32348747253418,
      "memory(GiB)": 77.56,
      "step": 50175,
      "token_acc": 0.5019305019305019,
      "train_speed(iter/s)": 1.438388
    },
    {
      "epoch": 2.1498650443425733,
      "grad_norm": 6.0155463218688965,
      "learning_rate": 6.091129968055146e-05,
      "loss": 2.5108104705810548,
      "memory(GiB)": 77.56,
      "step": 50180,
      "token_acc": 0.5145631067961165,
      "train_speed(iter/s)": 1.438377
    },
    {
      "epoch": 2.150079259671822,
      "grad_norm": 5.3230109214782715,
      "learning_rate": 6.09047320079244e-05,
      "loss": 2.3352073669433593,
      "memory(GiB)": 77.56,
      "step": 50185,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.438349
    },
    {
      "epoch": 2.150293475001071,
      "grad_norm": 5.95028018951416,
      "learning_rate": 6.0898164137747893e-05,
      "loss": 2.077684783935547,
      "memory(GiB)": 77.56,
      "step": 50190,
      "token_acc": 0.5226480836236934,
      "train_speed(iter/s)": 1.438344
    },
    {
      "epoch": 2.15050769033032,
      "grad_norm": 7.436186790466309,
      "learning_rate": 6.089159607014092e-05,
      "loss": 2.4042360305786135,
      "memory(GiB)": 77.56,
      "step": 50195,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.438344
    },
    {
      "epoch": 2.150721905659569,
      "grad_norm": 4.912622451782227,
      "learning_rate": 6.0885027805222484e-05,
      "loss": 2.5977706909179688,
      "memory(GiB)": 77.56,
      "step": 50200,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.438324
    },
    {
      "epoch": 2.150936120988818,
      "grad_norm": 6.073189735412598,
      "learning_rate": 6.0878459343111517e-05,
      "loss": 2.28847599029541,
      "memory(GiB)": 77.56,
      "step": 50205,
      "token_acc": 0.4756554307116105,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 2.151150336318067,
      "grad_norm": 6.330996990203857,
      "learning_rate": 6.087189068392709e-05,
      "loss": 2.3986082077026367,
      "memory(GiB)": 77.56,
      "step": 50210,
      "token_acc": 0.5089605734767025,
      "train_speed(iter/s)": 1.438328
    },
    {
      "epoch": 2.151364551647316,
      "grad_norm": 8.425272941589355,
      "learning_rate": 6.0865321827788154e-05,
      "loss": 2.206868553161621,
      "memory(GiB)": 77.56,
      "step": 50215,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438324
    },
    {
      "epoch": 2.1515787669765647,
      "grad_norm": 5.978322505950928,
      "learning_rate": 6.085875277481372e-05,
      "loss": 2.1524681091308593,
      "memory(GiB)": 77.56,
      "step": 50220,
      "token_acc": 0.5152542372881356,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 2.151792982305814,
      "grad_norm": 4.493298053741455,
      "learning_rate": 6.08521835251228e-05,
      "loss": 2.3736846923828123,
      "memory(GiB)": 77.56,
      "step": 50225,
      "token_acc": 0.5085910652920962,
      "train_speed(iter/s)": 1.438332
    },
    {
      "epoch": 2.1520071976350628,
      "grad_norm": 7.0440521240234375,
      "learning_rate": 6.084561407883438e-05,
      "loss": 2.971976089477539,
      "memory(GiB)": 77.56,
      "step": 50230,
      "token_acc": 0.39841688654353563,
      "train_speed(iter/s)": 1.438345
    },
    {
      "epoch": 2.1522214129643116,
      "grad_norm": 5.06019401550293,
      "learning_rate": 6.083904443606751e-05,
      "loss": 2.262111854553223,
      "memory(GiB)": 77.56,
      "step": 50235,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.438372
    },
    {
      "epoch": 2.152435628293561,
      "grad_norm": 3.6124343872070312,
      "learning_rate": 6.083247459694117e-05,
      "loss": 2.3987369537353516,
      "memory(GiB)": 77.56,
      "step": 50240,
      "token_acc": 0.4968152866242038,
      "train_speed(iter/s)": 1.438404
    },
    {
      "epoch": 2.1526498436228096,
      "grad_norm": 4.730318069458008,
      "learning_rate": 6.0825904561574374e-05,
      "loss": 2.4393558502197266,
      "memory(GiB)": 77.56,
      "step": 50245,
      "token_acc": 0.4954682779456193,
      "train_speed(iter/s)": 1.438423
    },
    {
      "epoch": 2.1528640589520585,
      "grad_norm": 7.20909309387207,
      "learning_rate": 6.081933433008617e-05,
      "loss": 2.157692718505859,
      "memory(GiB)": 77.56,
      "step": 50250,
      "token_acc": 0.5222551928783383,
      "train_speed(iter/s)": 1.438445
    },
    {
      "epoch": 2.1530782742813077,
      "grad_norm": 9.454322814941406,
      "learning_rate": 6.081276390259559e-05,
      "loss": 2.418796920776367,
      "memory(GiB)": 77.56,
      "step": 50255,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.43846
    },
    {
      "epoch": 2.1532924896105565,
      "grad_norm": 5.906976699829102,
      "learning_rate": 6.0806193279221634e-05,
      "loss": 2.2224166870117186,
      "memory(GiB)": 77.56,
      "step": 50260,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.43846
    },
    {
      "epoch": 2.1535067049398053,
      "grad_norm": 7.277247905731201,
      "learning_rate": 6.079962246008336e-05,
      "loss": 2.634247970581055,
      "memory(GiB)": 77.56,
      "step": 50265,
      "token_acc": 0.44983818770226536,
      "train_speed(iter/s)": 1.438491
    },
    {
      "epoch": 2.1537209202690546,
      "grad_norm": 5.129485130310059,
      "learning_rate": 6.07930514452998e-05,
      "loss": 2.341689682006836,
      "memory(GiB)": 77.56,
      "step": 50270,
      "token_acc": 0.5304054054054054,
      "train_speed(iter/s)": 1.438496
    },
    {
      "epoch": 2.1539351355983034,
      "grad_norm": 5.050314426422119,
      "learning_rate": 6.0786480234989976e-05,
      "loss": 2.390590476989746,
      "memory(GiB)": 77.56,
      "step": 50275,
      "token_acc": 0.44398340248962653,
      "train_speed(iter/s)": 1.438507
    },
    {
      "epoch": 2.154149350927552,
      "grad_norm": 5.15317440032959,
      "learning_rate": 6.0779908829272936e-05,
      "loss": 2.0559864044189453,
      "memory(GiB)": 77.56,
      "step": 50280,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.438518
    },
    {
      "epoch": 2.1543635662568015,
      "grad_norm": 5.07619047164917,
      "learning_rate": 6.077333722826775e-05,
      "loss": 2.4736354827880858,
      "memory(GiB)": 77.56,
      "step": 50285,
      "token_acc": 0.4921259842519685,
      "train_speed(iter/s)": 1.438502
    },
    {
      "epoch": 2.1545777815860503,
      "grad_norm": 7.170868396759033,
      "learning_rate": 6.076676543209344e-05,
      "loss": 2.4416608810424805,
      "memory(GiB)": 77.56,
      "step": 50290,
      "token_acc": 0.532051282051282,
      "train_speed(iter/s)": 1.438548
    },
    {
      "epoch": 2.154791996915299,
      "grad_norm": 6.5216851234436035,
      "learning_rate": 6.07601934408691e-05,
      "loss": 2.1830108642578123,
      "memory(GiB)": 77.56,
      "step": 50295,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.438531
    },
    {
      "epoch": 2.1550062122445484,
      "grad_norm": 5.386215686798096,
      "learning_rate": 6.075362125471374e-05,
      "loss": 2.333147430419922,
      "memory(GiB)": 77.56,
      "step": 50300,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.438535
    },
    {
      "epoch": 2.155220427573797,
      "grad_norm": 4.919002532958984,
      "learning_rate": 6.0747048873746446e-05,
      "loss": 2.3592729568481445,
      "memory(GiB)": 77.56,
      "step": 50305,
      "token_acc": 0.5267175572519084,
      "train_speed(iter/s)": 1.438569
    },
    {
      "epoch": 2.155434642903046,
      "grad_norm": 5.851461887359619,
      "learning_rate": 6.074047629808629e-05,
      "loss": 2.5147979736328123,
      "memory(GiB)": 77.56,
      "step": 50310,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.438575
    },
    {
      "epoch": 2.1556488582322952,
      "grad_norm": 5.775365352630615,
      "learning_rate": 6.073390352785232e-05,
      "loss": 2.216945838928223,
      "memory(GiB)": 77.56,
      "step": 50315,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.43859
    },
    {
      "epoch": 2.155863073561544,
      "grad_norm": 5.318002700805664,
      "learning_rate": 6.0727330563163624e-05,
      "loss": 2.3767499923706055,
      "memory(GiB)": 77.56,
      "step": 50320,
      "token_acc": 0.4492753623188406,
      "train_speed(iter/s)": 1.438621
    },
    {
      "epoch": 2.156077288890793,
      "grad_norm": 5.274514198303223,
      "learning_rate": 6.072075740413926e-05,
      "loss": 2.6518791198730467,
      "memory(GiB)": 77.56,
      "step": 50325,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.438629
    },
    {
      "epoch": 2.156291504220042,
      "grad_norm": 5.2085185050964355,
      "learning_rate": 6.071418405089834e-05,
      "loss": 2.455303955078125,
      "memory(GiB)": 77.56,
      "step": 50330,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.438639
    },
    {
      "epoch": 2.156505719549291,
      "grad_norm": 4.752186298370361,
      "learning_rate": 6.070761050355991e-05,
      "loss": 2.59373779296875,
      "memory(GiB)": 77.56,
      "step": 50335,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.438665
    },
    {
      "epoch": 2.1567199348785397,
      "grad_norm": 6.599400520324707,
      "learning_rate": 6.070103676224308e-05,
      "loss": 2.4299339294433593,
      "memory(GiB)": 77.56,
      "step": 50340,
      "token_acc": 0.459214501510574,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 2.156934150207789,
      "grad_norm": 4.464208602905273,
      "learning_rate": 6.069446282706692e-05,
      "loss": 2.295475387573242,
      "memory(GiB)": 77.56,
      "step": 50345,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.43865
    },
    {
      "epoch": 2.157148365537038,
      "grad_norm": 6.727293014526367,
      "learning_rate": 6.068788869815054e-05,
      "loss": 2.5154487609863283,
      "memory(GiB)": 77.56,
      "step": 50350,
      "token_acc": 0.4900662251655629,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 2.1573625808662866,
      "grad_norm": 6.149333477020264,
      "learning_rate": 6.068131437561303e-05,
      "loss": 2.3686569213867186,
      "memory(GiB)": 77.56,
      "step": 50355,
      "token_acc": 0.47244094488188976,
      "train_speed(iter/s)": 1.438691
    },
    {
      "epoch": 2.157576796195536,
      "grad_norm": 4.366853713989258,
      "learning_rate": 6.067473985957349e-05,
      "loss": 2.402776336669922,
      "memory(GiB)": 77.56,
      "step": 50360,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.43871
    },
    {
      "epoch": 2.1577910115247847,
      "grad_norm": 4.948853492736816,
      "learning_rate": 6.066816515015101e-05,
      "loss": 2.26220703125,
      "memory(GiB)": 77.56,
      "step": 50365,
      "token_acc": 0.46863468634686345,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 2.1580052268540335,
      "grad_norm": 6.735204219818115,
      "learning_rate": 6.0661590247464736e-05,
      "loss": 2.371020698547363,
      "memory(GiB)": 77.56,
      "step": 50370,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.438633
    },
    {
      "epoch": 2.1582194421832828,
      "grad_norm": 5.736177444458008,
      "learning_rate": 6.065501515163374e-05,
      "loss": 2.5300086975097655,
      "memory(GiB)": 77.56,
      "step": 50375,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.438652
    },
    {
      "epoch": 2.1584336575125316,
      "grad_norm": 7.273033142089844,
      "learning_rate": 6.064843986277715e-05,
      "loss": 2.9673542022705077,
      "memory(GiB)": 77.56,
      "step": 50380,
      "token_acc": 0.4222972972972973,
      "train_speed(iter/s)": 1.438659
    },
    {
      "epoch": 2.1586478728417804,
      "grad_norm": 4.439969062805176,
      "learning_rate": 6.064186438101409e-05,
      "loss": 2.4679986953735353,
      "memory(GiB)": 77.56,
      "step": 50385,
      "token_acc": 0.46835443037974683,
      "train_speed(iter/s)": 1.438654
    },
    {
      "epoch": 2.1588620881710296,
      "grad_norm": 4.850871562957764,
      "learning_rate": 6.063528870646367e-05,
      "loss": 2.368244171142578,
      "memory(GiB)": 77.56,
      "step": 50390,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.438617
    },
    {
      "epoch": 2.1590763035002785,
      "grad_norm": 5.8203043937683105,
      "learning_rate": 6.0628712839245005e-05,
      "loss": 2.4497161865234376,
      "memory(GiB)": 77.56,
      "step": 50395,
      "token_acc": 0.5052264808362369,
      "train_speed(iter/s)": 1.438639
    },
    {
      "epoch": 2.1592905188295273,
      "grad_norm": 6.4210734367370605,
      "learning_rate": 6.0622136779477254e-05,
      "loss": 2.456562614440918,
      "memory(GiB)": 77.56,
      "step": 50400,
      "token_acc": 0.4953560371517028,
      "train_speed(iter/s)": 1.438671
    },
    {
      "epoch": 2.1595047341587765,
      "grad_norm": 4.872335910797119,
      "learning_rate": 6.0615560527279514e-05,
      "loss": 2.2009693145751954,
      "memory(GiB)": 77.56,
      "step": 50405,
      "token_acc": 0.534375,
      "train_speed(iter/s)": 1.438699
    },
    {
      "epoch": 2.1597189494880253,
      "grad_norm": 4.752109527587891,
      "learning_rate": 6.060898408277096e-05,
      "loss": 2.5521974563598633,
      "memory(GiB)": 77.56,
      "step": 50410,
      "token_acc": 0.47230320699708456,
      "train_speed(iter/s)": 1.43873
    },
    {
      "epoch": 2.159933164817274,
      "grad_norm": 4.424752712249756,
      "learning_rate": 6.06024074460707e-05,
      "loss": 2.467576026916504,
      "memory(GiB)": 77.56,
      "step": 50415,
      "token_acc": 0.4793650793650794,
      "train_speed(iter/s)": 1.438714
    },
    {
      "epoch": 2.1601473801465234,
      "grad_norm": 4.672728538513184,
      "learning_rate": 6.059583061729787e-05,
      "loss": 2.4791168212890624,
      "memory(GiB)": 77.56,
      "step": 50420,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.438702
    },
    {
      "epoch": 2.1603615954757722,
      "grad_norm": 5.194528102874756,
      "learning_rate": 6.058925359657164e-05,
      "loss": 2.430134963989258,
      "memory(GiB)": 77.56,
      "step": 50425,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.438705
    },
    {
      "epoch": 2.160575810805021,
      "grad_norm": 4.794474124908447,
      "learning_rate": 6.058267638401114e-05,
      "loss": 2.731005859375,
      "memory(GiB)": 77.56,
      "step": 50430,
      "token_acc": 0.4522058823529412,
      "train_speed(iter/s)": 1.438734
    },
    {
      "epoch": 2.1607900261342703,
      "grad_norm": 5.209463119506836,
      "learning_rate": 6.057609897973552e-05,
      "loss": 2.6820161819458006,
      "memory(GiB)": 77.56,
      "step": 50435,
      "token_acc": 0.437984496124031,
      "train_speed(iter/s)": 1.438703
    },
    {
      "epoch": 2.161004241463519,
      "grad_norm": 4.501694679260254,
      "learning_rate": 6.056952138386397e-05,
      "loss": 2.462152099609375,
      "memory(GiB)": 77.56,
      "step": 50440,
      "token_acc": 0.4260355029585799,
      "train_speed(iter/s)": 1.438705
    },
    {
      "epoch": 2.161218456792768,
      "grad_norm": 5.322403430938721,
      "learning_rate": 6.056294359651562e-05,
      "loss": 2.1826005935668946,
      "memory(GiB)": 77.56,
      "step": 50445,
      "token_acc": 0.5755395683453237,
      "train_speed(iter/s)": 1.438692
    },
    {
      "epoch": 2.161432672122017,
      "grad_norm": 6.76984977722168,
      "learning_rate": 6.0556365617809615e-05,
      "loss": 2.304361343383789,
      "memory(GiB)": 77.56,
      "step": 50450,
      "token_acc": 0.5163398692810458,
      "train_speed(iter/s)": 1.438684
    },
    {
      "epoch": 2.161646887451266,
      "grad_norm": 4.454043865203857,
      "learning_rate": 6.0549787447865166e-05,
      "loss": 2.467548370361328,
      "memory(GiB)": 77.56,
      "step": 50455,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.438677
    },
    {
      "epoch": 2.161861102780515,
      "grad_norm": 6.561191558837891,
      "learning_rate": 6.0543209086801434e-05,
      "loss": 2.564452362060547,
      "memory(GiB)": 77.56,
      "step": 50460,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.438681
    },
    {
      "epoch": 2.162075318109764,
      "grad_norm": 5.802851676940918,
      "learning_rate": 6.053663053473754e-05,
      "loss": 2.1944671630859376,
      "memory(GiB)": 77.56,
      "step": 50465,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.438719
    },
    {
      "epoch": 2.162289533439013,
      "grad_norm": 5.064087390899658,
      "learning_rate": 6.053005179179273e-05,
      "loss": 2.431778144836426,
      "memory(GiB)": 77.56,
      "step": 50470,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.438746
    },
    {
      "epoch": 2.1625037487682617,
      "grad_norm": 8.673851013183594,
      "learning_rate": 6.052347285808615e-05,
      "loss": 2.200387954711914,
      "memory(GiB)": 77.56,
      "step": 50475,
      "token_acc": 0.5322033898305085,
      "train_speed(iter/s)": 1.438779
    },
    {
      "epoch": 2.162717964097511,
      "grad_norm": 7.30638313293457,
      "learning_rate": 6.051689373373698e-05,
      "loss": 2.772796058654785,
      "memory(GiB)": 77.56,
      "step": 50480,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.438763
    },
    {
      "epoch": 2.1629321794267597,
      "grad_norm": 5.439511775970459,
      "learning_rate": 6.0510314418864413e-05,
      "loss": 2.2394184112548827,
      "memory(GiB)": 77.56,
      "step": 50485,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.438774
    },
    {
      "epoch": 2.1631463947560086,
      "grad_norm": 5.40219259262085,
      "learning_rate": 6.050373491358764e-05,
      "loss": 2.2772876739501955,
      "memory(GiB)": 77.56,
      "step": 50490,
      "token_acc": 0.5364238410596026,
      "train_speed(iter/s)": 1.438742
    },
    {
      "epoch": 2.163360610085258,
      "grad_norm": 5.540431976318359,
      "learning_rate": 6.049715521802587e-05,
      "loss": 2.6008733749389648,
      "memory(GiB)": 77.56,
      "step": 50495,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.438718
    },
    {
      "epoch": 2.1635748254145066,
      "grad_norm": 5.823878765106201,
      "learning_rate": 6.0490575332298274e-05,
      "loss": 2.362906646728516,
      "memory(GiB)": 77.56,
      "step": 50500,
      "token_acc": 0.48598130841121495,
      "train_speed(iter/s)": 1.438739
    },
    {
      "epoch": 2.1635748254145066,
      "eval_loss": 2.3349204063415527,
      "eval_runtime": 15.0511,
      "eval_samples_per_second": 6.644,
      "eval_steps_per_second": 6.644,
      "eval_token_acc": 0.4614427860696517,
      "step": 50500
    },
    {
      "epoch": 2.1637890407437554,
      "grad_norm": 5.63339376449585,
      "learning_rate": 6.048399525652406e-05,
      "loss": 2.63114013671875,
      "memory(GiB)": 77.56,
      "step": 50505,
      "token_acc": 0.46698564593301434,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 2.1640032560730047,
      "grad_norm": 7.447271823883057,
      "learning_rate": 6.0477414990822444e-05,
      "loss": 2.4979364395141603,
      "memory(GiB)": 77.56,
      "step": 50510,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.43811
    },
    {
      "epoch": 2.1642174714022535,
      "grad_norm": 5.807193279266357,
      "learning_rate": 6.0470834535312636e-05,
      "loss": 2.468741226196289,
      "memory(GiB)": 77.56,
      "step": 50515,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.438122
    },
    {
      "epoch": 2.1644316867315023,
      "grad_norm": 5.4508867263793945,
      "learning_rate": 6.046425389011382e-05,
      "loss": 2.6488554000854494,
      "memory(GiB)": 77.56,
      "step": 50520,
      "token_acc": 0.4461538461538462,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 2.1646459020607516,
      "grad_norm": 5.475977897644043,
      "learning_rate": 6.045767305534524e-05,
      "loss": 2.4003440856933596,
      "memory(GiB)": 77.56,
      "step": 50525,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.438108
    },
    {
      "epoch": 2.1648601173900004,
      "grad_norm": 9.473495483398438,
      "learning_rate": 6.045109203112611e-05,
      "loss": 2.3714120864868162,
      "memory(GiB)": 77.56,
      "step": 50530,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.438129
    },
    {
      "epoch": 2.165074332719249,
      "grad_norm": 5.357614994049072,
      "learning_rate": 6.044451081757563e-05,
      "loss": 2.2716957092285157,
      "memory(GiB)": 77.56,
      "step": 50535,
      "token_acc": 0.5347222222222222,
      "train_speed(iter/s)": 1.438134
    },
    {
      "epoch": 2.1652885480484985,
      "grad_norm": 8.74575138092041,
      "learning_rate": 6.043792941481303e-05,
      "loss": 2.482444190979004,
      "memory(GiB)": 77.56,
      "step": 50540,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.438118
    },
    {
      "epoch": 2.1655027633777473,
      "grad_norm": 5.383188247680664,
      "learning_rate": 6.0431347822957574e-05,
      "loss": 2.466129684448242,
      "memory(GiB)": 77.56,
      "step": 50545,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 2.165716978706996,
      "grad_norm": 5.982137203216553,
      "learning_rate": 6.042476604212844e-05,
      "loss": 2.315919876098633,
      "memory(GiB)": 77.56,
      "step": 50550,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.438158
    },
    {
      "epoch": 2.1659311940362453,
      "grad_norm": 5.913150787353516,
      "learning_rate": 6.041818407244492e-05,
      "loss": 2.41861515045166,
      "memory(GiB)": 77.56,
      "step": 50555,
      "token_acc": 0.4847457627118644,
      "train_speed(iter/s)": 1.438199
    },
    {
      "epoch": 2.166145409365494,
      "grad_norm": 4.576161861419678,
      "learning_rate": 6.0411601914026205e-05,
      "loss": 2.2976364135742187,
      "memory(GiB)": 77.56,
      "step": 50560,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.438211
    },
    {
      "epoch": 2.166359624694743,
      "grad_norm": 5.48606538772583,
      "learning_rate": 6.040501956699155e-05,
      "loss": 2.5144866943359374,
      "memory(GiB)": 77.56,
      "step": 50565,
      "token_acc": 0.4603658536585366,
      "train_speed(iter/s)": 1.438231
    },
    {
      "epoch": 2.1665738400239922,
      "grad_norm": 5.112666130065918,
      "learning_rate": 6.039843703146022e-05,
      "loss": 2.504762649536133,
      "memory(GiB)": 77.56,
      "step": 50570,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.438194
    },
    {
      "epoch": 2.166788055353241,
      "grad_norm": 5.325357913970947,
      "learning_rate": 6.039185430755143e-05,
      "loss": 2.5218271255493163,
      "memory(GiB)": 77.56,
      "step": 50575,
      "token_acc": 0.4489051094890511,
      "train_speed(iter/s)": 1.438199
    },
    {
      "epoch": 2.16700227068249,
      "grad_norm": 4.974330425262451,
      "learning_rate": 6.038527139538445e-05,
      "loss": 2.3442657470703123,
      "memory(GiB)": 77.56,
      "step": 50580,
      "token_acc": 0.5076452599388379,
      "train_speed(iter/s)": 1.438234
    },
    {
      "epoch": 2.167216486011739,
      "grad_norm": 5.249184608459473,
      "learning_rate": 6.0378688295078556e-05,
      "loss": 2.2773159027099608,
      "memory(GiB)": 77.56,
      "step": 50585,
      "token_acc": 0.5125,
      "train_speed(iter/s)": 1.43826
    },
    {
      "epoch": 2.167430701340988,
      "grad_norm": 5.24116849899292,
      "learning_rate": 6.037210500675298e-05,
      "loss": 2.8669221878051756,
      "memory(GiB)": 77.56,
      "step": 50590,
      "token_acc": 0.41494845360824745,
      "train_speed(iter/s)": 1.43826
    },
    {
      "epoch": 2.1676449166702367,
      "grad_norm": 5.492655277252197,
      "learning_rate": 6.036552153052698e-05,
      "loss": 2.393923759460449,
      "memory(GiB)": 77.56,
      "step": 50595,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.438275
    },
    {
      "epoch": 2.167859131999486,
      "grad_norm": 5.0687055587768555,
      "learning_rate": 6.035893786651985e-05,
      "loss": 2.402924156188965,
      "memory(GiB)": 77.56,
      "step": 50600,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.438293
    },
    {
      "epoch": 2.168073347328735,
      "grad_norm": 6.7476396560668945,
      "learning_rate": 6.035235401485084e-05,
      "loss": 2.2508779525756837,
      "memory(GiB)": 77.56,
      "step": 50605,
      "token_acc": 0.5320754716981132,
      "train_speed(iter/s)": 1.438331
    },
    {
      "epoch": 2.1682875626579836,
      "grad_norm": 5.155445575714111,
      "learning_rate": 6.034576997563921e-05,
      "loss": 2.6313072204589845,
      "memory(GiB)": 77.56,
      "step": 50610,
      "token_acc": 0.4564459930313589,
      "train_speed(iter/s)": 1.438352
    },
    {
      "epoch": 2.168501777987233,
      "grad_norm": 5.576014995574951,
      "learning_rate": 6.0339185749004265e-05,
      "loss": 2.3745452880859377,
      "memory(GiB)": 77.56,
      "step": 50615,
      "token_acc": 0.5109717868338558,
      "train_speed(iter/s)": 1.438366
    },
    {
      "epoch": 2.1687159933164817,
      "grad_norm": 5.3393049240112305,
      "learning_rate": 6.033260133506528e-05,
      "loss": 2.7261726379394533,
      "memory(GiB)": 77.56,
      "step": 50620,
      "token_acc": 0.438871473354232,
      "train_speed(iter/s)": 1.438346
    },
    {
      "epoch": 2.1689302086457305,
      "grad_norm": 4.502142429351807,
      "learning_rate": 6.03260167339415e-05,
      "loss": 2.405743408203125,
      "memory(GiB)": 77.56,
      "step": 50625,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438366
    },
    {
      "epoch": 2.1691444239749798,
      "grad_norm": 4.860012531280518,
      "learning_rate": 6.031943194575227e-05,
      "loss": 2.5214256286621093,
      "memory(GiB)": 77.56,
      "step": 50630,
      "token_acc": 0.5086505190311419,
      "train_speed(iter/s)": 1.43839
    },
    {
      "epoch": 2.1693586393042286,
      "grad_norm": 7.530333042144775,
      "learning_rate": 6.031284697061683e-05,
      "loss": 2.290510559082031,
      "memory(GiB)": 77.56,
      "step": 50635,
      "token_acc": 0.4911242603550296,
      "train_speed(iter/s)": 1.438414
    },
    {
      "epoch": 2.1695728546334774,
      "grad_norm": 7.081682205200195,
      "learning_rate": 6.030626180865451e-05,
      "loss": 2.4867294311523436,
      "memory(GiB)": 77.56,
      "step": 50640,
      "token_acc": 0.4380664652567976,
      "train_speed(iter/s)": 1.438391
    },
    {
      "epoch": 2.1697870699627266,
      "grad_norm": 5.617022514343262,
      "learning_rate": 6.029967645998459e-05,
      "loss": 2.5478572845458984,
      "memory(GiB)": 77.56,
      "step": 50645,
      "token_acc": 0.5126050420168067,
      "train_speed(iter/s)": 1.438383
    },
    {
      "epoch": 2.1700012852919754,
      "grad_norm": 5.664225101470947,
      "learning_rate": 6.0293090924726346e-05,
      "loss": 2.430221748352051,
      "memory(GiB)": 77.56,
      "step": 50650,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438393
    },
    {
      "epoch": 2.1702155006212243,
      "grad_norm": 7.483113765716553,
      "learning_rate": 6.028650520299912e-05,
      "loss": 2.2885259628295898,
      "memory(GiB)": 77.56,
      "step": 50655,
      "token_acc": 0.5451388888888888,
      "train_speed(iter/s)": 1.438394
    },
    {
      "epoch": 2.1704297159504735,
      "grad_norm": 6.605968475341797,
      "learning_rate": 6.0279919294922206e-05,
      "loss": 2.8503719329833985,
      "memory(GiB)": 77.56,
      "step": 50660,
      "token_acc": 0.4475806451612903,
      "train_speed(iter/s)": 1.438375
    },
    {
      "epoch": 2.1706439312797223,
      "grad_norm": 5.806797981262207,
      "learning_rate": 6.02733332006149e-05,
      "loss": 2.3515445709228517,
      "memory(GiB)": 77.56,
      "step": 50665,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438405
    },
    {
      "epoch": 2.170858146608971,
      "grad_norm": 6.882453918457031,
      "learning_rate": 6.026674692019654e-05,
      "loss": 2.3562469482421875,
      "memory(GiB)": 77.56,
      "step": 50670,
      "token_acc": 0.5056603773584906,
      "train_speed(iter/s)": 1.438394
    },
    {
      "epoch": 2.1710723619382204,
      "grad_norm": 5.481173515319824,
      "learning_rate": 6.0260160453786416e-05,
      "loss": 2.620244598388672,
      "memory(GiB)": 77.56,
      "step": 50675,
      "token_acc": 0.4779874213836478,
      "train_speed(iter/s)": 1.438356
    },
    {
      "epoch": 2.171286577267469,
      "grad_norm": 6.531098365783691,
      "learning_rate": 6.025357380150387e-05,
      "loss": 2.4211912155151367,
      "memory(GiB)": 77.56,
      "step": 50680,
      "token_acc": 0.5272108843537415,
      "train_speed(iter/s)": 1.438332
    },
    {
      "epoch": 2.171500792596718,
      "grad_norm": 6.274392604827881,
      "learning_rate": 6.02469869634682e-05,
      "loss": 2.415005111694336,
      "memory(GiB)": 77.56,
      "step": 50685,
      "token_acc": 0.49279538904899134,
      "train_speed(iter/s)": 1.438361
    },
    {
      "epoch": 2.1717150079259673,
      "grad_norm": 5.343069553375244,
      "learning_rate": 6.0240399939798766e-05,
      "loss": 2.479130744934082,
      "memory(GiB)": 77.56,
      "step": 50690,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.438364
    },
    {
      "epoch": 2.171929223255216,
      "grad_norm": 5.851415634155273,
      "learning_rate": 6.023381273061487e-05,
      "loss": 2.395462417602539,
      "memory(GiB)": 77.56,
      "step": 50695,
      "token_acc": 0.4901185770750988,
      "train_speed(iter/s)": 1.438332
    },
    {
      "epoch": 2.172143438584465,
      "grad_norm": 7.054051876068115,
      "learning_rate": 6.0227225336035866e-05,
      "loss": 2.254651641845703,
      "memory(GiB)": 77.56,
      "step": 50700,
      "token_acc": 0.5340136054421769,
      "train_speed(iter/s)": 1.438325
    },
    {
      "epoch": 2.172357653913714,
      "grad_norm": 5.3499345779418945,
      "learning_rate": 6.022063775618107e-05,
      "loss": 2.645551300048828,
      "memory(GiB)": 77.56,
      "step": 50705,
      "token_acc": 0.4368932038834951,
      "train_speed(iter/s)": 1.438301
    },
    {
      "epoch": 2.172571869242963,
      "grad_norm": 4.720377445220947,
      "learning_rate": 6.0214049991169844e-05,
      "loss": 2.3728002548217773,
      "memory(GiB)": 77.56,
      "step": 50710,
      "token_acc": 0.5029585798816568,
      "train_speed(iter/s)": 1.438307
    },
    {
      "epoch": 2.172786084572212,
      "grad_norm": 5.480470657348633,
      "learning_rate": 6.0207462041121524e-05,
      "loss": 2.592828369140625,
      "memory(GiB)": 77.56,
      "step": 50715,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.438292
    },
    {
      "epoch": 2.173000299901461,
      "grad_norm": 6.5919718742370605,
      "learning_rate": 6.0200873906155455e-05,
      "loss": 2.308619499206543,
      "memory(GiB)": 77.56,
      "step": 50720,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.438327
    },
    {
      "epoch": 2.17321451523071,
      "grad_norm": 5.305961608886719,
      "learning_rate": 6.0194285586390955e-05,
      "loss": 2.450691986083984,
      "memory(GiB)": 77.56,
      "step": 50725,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.438314
    },
    {
      "epoch": 2.1734287305599587,
      "grad_norm": 5.529691696166992,
      "learning_rate": 6.0187697081947434e-05,
      "loss": 2.3037506103515626,
      "memory(GiB)": 77.56,
      "step": 50730,
      "token_acc": 0.5287356321839081,
      "train_speed(iter/s)": 1.438332
    },
    {
      "epoch": 2.173642945889208,
      "grad_norm": 5.136481285095215,
      "learning_rate": 6.0181108392944216e-05,
      "loss": 2.4477033615112305,
      "memory(GiB)": 77.56,
      "step": 50735,
      "token_acc": 0.48502994011976047,
      "train_speed(iter/s)": 1.438358
    },
    {
      "epoch": 2.1738571612184567,
      "grad_norm": 6.894428253173828,
      "learning_rate": 6.017451951950067e-05,
      "loss": 2.6204248428344727,
      "memory(GiB)": 77.56,
      "step": 50740,
      "token_acc": 0.47580645161290325,
      "train_speed(iter/s)": 1.438369
    },
    {
      "epoch": 2.1740713765477055,
      "grad_norm": 6.091804504394531,
      "learning_rate": 6.0167930461736165e-05,
      "loss": 2.5189701080322267,
      "memory(GiB)": 77.56,
      "step": 50745,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.438399
    },
    {
      "epoch": 2.174285591876955,
      "grad_norm": 5.559927940368652,
      "learning_rate": 6.016134121977006e-05,
      "loss": 2.2092323303222656,
      "memory(GiB)": 77.56,
      "step": 50750,
      "token_acc": 0.5642023346303502,
      "train_speed(iter/s)": 1.43841
    },
    {
      "epoch": 2.1744998072062036,
      "grad_norm": 5.553183555603027,
      "learning_rate": 6.015475179372173e-05,
      "loss": 2.5787242889404296,
      "memory(GiB)": 77.56,
      "step": 50755,
      "token_acc": 0.45353159851301117,
      "train_speed(iter/s)": 1.438394
    },
    {
      "epoch": 2.1747140225354524,
      "grad_norm": 5.36630916595459,
      "learning_rate": 6.0148162183710534e-05,
      "loss": 2.3338794708251953,
      "memory(GiB)": 77.56,
      "step": 50760,
      "token_acc": 0.5173745173745173,
      "train_speed(iter/s)": 1.438408
    },
    {
      "epoch": 2.1749282378647017,
      "grad_norm": 5.367036819458008,
      "learning_rate": 6.014157238985587e-05,
      "loss": 2.266572380065918,
      "memory(GiB)": 77.56,
      "step": 50765,
      "token_acc": 0.5291970802919708,
      "train_speed(iter/s)": 1.43838
    },
    {
      "epoch": 2.1751424531939505,
      "grad_norm": 6.967119216918945,
      "learning_rate": 6.0134982412277095e-05,
      "loss": 2.2577402114868166,
      "memory(GiB)": 77.56,
      "step": 50770,
      "token_acc": 0.5068027210884354,
      "train_speed(iter/s)": 1.438372
    },
    {
      "epoch": 2.1753566685231993,
      "grad_norm": 6.080873966217041,
      "learning_rate": 6.0128392251093624e-05,
      "loss": 2.2987483978271483,
      "memory(GiB)": 77.56,
      "step": 50775,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.438372
    },
    {
      "epoch": 2.1755708838524486,
      "grad_norm": 8.971758842468262,
      "learning_rate": 6.012180190642481e-05,
      "loss": 2.63602237701416,
      "memory(GiB)": 77.56,
      "step": 50780,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.438379
    },
    {
      "epoch": 2.1757850991816974,
      "grad_norm": 6.279210567474365,
      "learning_rate": 6.011521137839007e-05,
      "loss": 2.659720230102539,
      "memory(GiB)": 77.56,
      "step": 50785,
      "token_acc": 0.4421768707482993,
      "train_speed(iter/s)": 1.438369
    },
    {
      "epoch": 2.175999314510946,
      "grad_norm": 6.57822322845459,
      "learning_rate": 6.0108620667108794e-05,
      "loss": 2.3965044021606445,
      "memory(GiB)": 77.56,
      "step": 50790,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.438377
    },
    {
      "epoch": 2.1762135298401954,
      "grad_norm": 5.5568647384643555,
      "learning_rate": 6.010202977270035e-05,
      "loss": 2.5202722549438477,
      "memory(GiB)": 77.56,
      "step": 50795,
      "token_acc": 0.4981684981684982,
      "train_speed(iter/s)": 1.438402
    },
    {
      "epoch": 2.1764277451694443,
      "grad_norm": 5.045445919036865,
      "learning_rate": 6.009543869528417e-05,
      "loss": 2.3829681396484377,
      "memory(GiB)": 77.56,
      "step": 50800,
      "token_acc": 0.5175718849840255,
      "train_speed(iter/s)": 1.43842
    },
    {
      "epoch": 2.176641960498693,
      "grad_norm": 4.594754219055176,
      "learning_rate": 6.008884743497966e-05,
      "loss": 2.480429267883301,
      "memory(GiB)": 77.56,
      "step": 50805,
      "token_acc": 0.45977011494252873,
      "train_speed(iter/s)": 1.438422
    },
    {
      "epoch": 2.1768561758279423,
      "grad_norm": 5.978048324584961,
      "learning_rate": 6.00822559919062e-05,
      "loss": 2.185072135925293,
      "memory(GiB)": 77.56,
      "step": 50810,
      "token_acc": 0.5245283018867924,
      "train_speed(iter/s)": 1.438439
    },
    {
      "epoch": 2.177070391157191,
      "grad_norm": 6.25172758102417,
      "learning_rate": 6.007566436618321e-05,
      "loss": 2.718180465698242,
      "memory(GiB)": 77.56,
      "step": 50815,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.438403
    },
    {
      "epoch": 2.17728460648644,
      "grad_norm": 6.709914684295654,
      "learning_rate": 6.006907255793013e-05,
      "loss": 2.6451622009277345,
      "memory(GiB)": 77.56,
      "step": 50820,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.438417
    },
    {
      "epoch": 2.177498821815689,
      "grad_norm": 4.794750690460205,
      "learning_rate": 6.006248056726634e-05,
      "loss": 2.0978515625,
      "memory(GiB)": 77.56,
      "step": 50825,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.438434
    },
    {
      "epoch": 2.177713037144938,
      "grad_norm": 5.17275333404541,
      "learning_rate": 6.005588839431129e-05,
      "loss": 2.4510494232177735,
      "memory(GiB)": 77.56,
      "step": 50830,
      "token_acc": 0.465625,
      "train_speed(iter/s)": 1.438478
    },
    {
      "epoch": 2.177927252474187,
      "grad_norm": 5.74208927154541,
      "learning_rate": 6.0049296039184364e-05,
      "loss": 2.42938346862793,
      "memory(GiB)": 77.56,
      "step": 50835,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438458
    },
    {
      "epoch": 2.178141467803436,
      "grad_norm": 5.702775955200195,
      "learning_rate": 6.0042703502005015e-05,
      "loss": 2.3877586364746093,
      "memory(GiB)": 77.56,
      "step": 50840,
      "token_acc": 0.48742138364779874,
      "train_speed(iter/s)": 1.438488
    },
    {
      "epoch": 2.178355683132685,
      "grad_norm": 6.729453086853027,
      "learning_rate": 6.00361107828927e-05,
      "loss": 2.3723974227905273,
      "memory(GiB)": 77.56,
      "step": 50845,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.438503
    },
    {
      "epoch": 2.1785698984619337,
      "grad_norm": 6.154310703277588,
      "learning_rate": 6.00295178819668e-05,
      "loss": 2.5813852310180665,
      "memory(GiB)": 77.56,
      "step": 50850,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.438527
    },
    {
      "epoch": 2.178784113791183,
      "grad_norm": 7.639402389526367,
      "learning_rate": 6.002292479934678e-05,
      "loss": 2.7543163299560547,
      "memory(GiB)": 77.56,
      "step": 50855,
      "token_acc": 0.4601449275362319,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 2.178998329120432,
      "grad_norm": 5.869032859802246,
      "learning_rate": 6.0016331535152084e-05,
      "loss": 2.389212226867676,
      "memory(GiB)": 77.56,
      "step": 50860,
      "token_acc": 0.44573643410852715,
      "train_speed(iter/s)": 1.438556
    },
    {
      "epoch": 2.1792125444496806,
      "grad_norm": 4.972807884216309,
      "learning_rate": 6.000973808950214e-05,
      "loss": 2.421988105773926,
      "memory(GiB)": 77.56,
      "step": 50865,
      "token_acc": 0.525679758308157,
      "train_speed(iter/s)": 1.43857
    },
    {
      "epoch": 2.17942675977893,
      "grad_norm": 6.6491522789001465,
      "learning_rate": 6.000314446251638e-05,
      "loss": 2.5481613159179686,
      "memory(GiB)": 77.56,
      "step": 50870,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.438578
    },
    {
      "epoch": 2.1796409751081787,
      "grad_norm": 5.59084939956665,
      "learning_rate": 5.99965506543143e-05,
      "loss": 2.310906410217285,
      "memory(GiB)": 77.56,
      "step": 50875,
      "token_acc": 0.49508196721311476,
      "train_speed(iter/s)": 1.438587
    },
    {
      "epoch": 2.1798551904374275,
      "grad_norm": 6.800754547119141,
      "learning_rate": 5.9989956665015324e-05,
      "loss": 2.799560546875,
      "memory(GiB)": 77.56,
      "step": 50880,
      "token_acc": 0.4158415841584158,
      "train_speed(iter/s)": 1.43857
    },
    {
      "epoch": 2.1800694057666767,
      "grad_norm": 4.771638870239258,
      "learning_rate": 5.99833624947389e-05,
      "loss": 2.548948287963867,
      "memory(GiB)": 77.56,
      "step": 50885,
      "token_acc": 0.5095057034220533,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 2.1802836210959256,
      "grad_norm": 6.158051490783691,
      "learning_rate": 5.997676814360451e-05,
      "loss": 2.431191825866699,
      "memory(GiB)": 77.56,
      "step": 50890,
      "token_acc": 0.46886446886446886,
      "train_speed(iter/s)": 1.438612
    },
    {
      "epoch": 2.1804978364251744,
      "grad_norm": 6.7467803955078125,
      "learning_rate": 5.9970173611731616e-05,
      "loss": 2.714176559448242,
      "memory(GiB)": 77.56,
      "step": 50895,
      "token_acc": 0.43109540636042404,
      "train_speed(iter/s)": 1.43861
    },
    {
      "epoch": 2.1807120517544236,
      "grad_norm": 5.072213172912598,
      "learning_rate": 5.996357889923965e-05,
      "loss": 2.329721450805664,
      "memory(GiB)": 77.56,
      "step": 50900,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.438567
    },
    {
      "epoch": 2.1809262670836724,
      "grad_norm": 5.626143932342529,
      "learning_rate": 5.995698400624813e-05,
      "loss": 2.12532901763916,
      "memory(GiB)": 77.56,
      "step": 50905,
      "token_acc": 0.5059760956175299,
      "train_speed(iter/s)": 1.438556
    },
    {
      "epoch": 2.1811404824129212,
      "grad_norm": 5.469083786010742,
      "learning_rate": 5.995038893287648e-05,
      "loss": 2.5412452697753904,
      "memory(GiB)": 77.56,
      "step": 50910,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.438609
    },
    {
      "epoch": 2.1813546977421705,
      "grad_norm": 5.214121341705322,
      "learning_rate": 5.994379367924421e-05,
      "loss": 2.3003984451293946,
      "memory(GiB)": 77.56,
      "step": 50915,
      "token_acc": 0.532,
      "train_speed(iter/s)": 1.438636
    },
    {
      "epoch": 2.1815689130714193,
      "grad_norm": 5.522370338439941,
      "learning_rate": 5.993719824547079e-05,
      "loss": 2.4568899154663084,
      "memory(GiB)": 77.56,
      "step": 50920,
      "token_acc": 0.4483870967741935,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 2.181783128400668,
      "grad_norm": 6.3648858070373535,
      "learning_rate": 5.9930602631675705e-05,
      "loss": 2.533412551879883,
      "memory(GiB)": 77.56,
      "step": 50925,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.438619
    },
    {
      "epoch": 2.1819973437299174,
      "grad_norm": 6.041980266571045,
      "learning_rate": 5.992400683797843e-05,
      "loss": 2.4639968872070312,
      "memory(GiB)": 77.56,
      "step": 50930,
      "token_acc": 0.5177304964539007,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 2.182211559059166,
      "grad_norm": 6.086566925048828,
      "learning_rate": 5.991741086449848e-05,
      "loss": 2.2928543090820312,
      "memory(GiB)": 77.56,
      "step": 50935,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.438578
    },
    {
      "epoch": 2.182425774388415,
      "grad_norm": 5.191893577575684,
      "learning_rate": 5.991081471135531e-05,
      "loss": 2.32578182220459,
      "memory(GiB)": 77.56,
      "step": 50940,
      "token_acc": 0.4959349593495935,
      "train_speed(iter/s)": 1.438599
    },
    {
      "epoch": 2.1826399897176643,
      "grad_norm": 5.302101135253906,
      "learning_rate": 5.990421837866843e-05,
      "loss": 2.479727935791016,
      "memory(GiB)": 77.56,
      "step": 50945,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.438617
    },
    {
      "epoch": 2.182854205046913,
      "grad_norm": 5.075984954833984,
      "learning_rate": 5.989762186655736e-05,
      "loss": 2.6966833114624023,
      "memory(GiB)": 77.56,
      "step": 50950,
      "token_acc": 0.44931506849315067,
      "train_speed(iter/s)": 1.438643
    },
    {
      "epoch": 2.183068420376162,
      "grad_norm": 7.5072340965271,
      "learning_rate": 5.989102517514158e-05,
      "loss": 2.3890151977539062,
      "memory(GiB)": 77.56,
      "step": 50955,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.438639
    },
    {
      "epoch": 2.183282635705411,
      "grad_norm": 4.678971767425537,
      "learning_rate": 5.9884428304540595e-05,
      "loss": 2.442661094665527,
      "memory(GiB)": 77.56,
      "step": 50960,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.438662
    },
    {
      "epoch": 2.18349685103466,
      "grad_norm": 5.9851460456848145,
      "learning_rate": 5.987783125487394e-05,
      "loss": 2.358523368835449,
      "memory(GiB)": 77.56,
      "step": 50965,
      "token_acc": 0.5109170305676856,
      "train_speed(iter/s)": 1.438674
    },
    {
      "epoch": 2.1837110663639088,
      "grad_norm": 4.827874660491943,
      "learning_rate": 5.987123402626108e-05,
      "loss": 2.4888254165649415,
      "memory(GiB)": 77.56,
      "step": 50970,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.438703
    },
    {
      "epoch": 2.183925281693158,
      "grad_norm": 5.580270767211914,
      "learning_rate": 5.986463661882157e-05,
      "loss": 2.5833593368530274,
      "memory(GiB)": 77.56,
      "step": 50975,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.438712
    },
    {
      "epoch": 2.184139497022407,
      "grad_norm": 6.3668718338012695,
      "learning_rate": 5.985803903267491e-05,
      "loss": 2.554352569580078,
      "memory(GiB)": 77.56,
      "step": 50980,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.438718
    },
    {
      "epoch": 2.1843537123516557,
      "grad_norm": 6.56332540512085,
      "learning_rate": 5.985144126794061e-05,
      "loss": 2.382846641540527,
      "memory(GiB)": 77.56,
      "step": 50985,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.438677
    },
    {
      "epoch": 2.184567927680905,
      "grad_norm": 6.39066743850708,
      "learning_rate": 5.984484332473823e-05,
      "loss": 2.6113391876220704,
      "memory(GiB)": 77.56,
      "step": 50990,
      "token_acc": 0.4664179104477612,
      "train_speed(iter/s)": 1.438653
    },
    {
      "epoch": 2.1847821430101537,
      "grad_norm": 5.2281413078308105,
      "learning_rate": 5.983824520318728e-05,
      "loss": 2.4021209716796874,
      "memory(GiB)": 77.56,
      "step": 50995,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.438669
    },
    {
      "epoch": 2.1849963583394025,
      "grad_norm": 4.64334774017334,
      "learning_rate": 5.983164690340727e-05,
      "loss": 2.5635854721069338,
      "memory(GiB)": 77.56,
      "step": 51000,
      "token_acc": 0.44984802431610943,
      "train_speed(iter/s)": 1.438714
    },
    {
      "epoch": 2.1849963583394025,
      "eval_loss": 2.136322021484375,
      "eval_runtime": 14.4488,
      "eval_samples_per_second": 6.921,
      "eval_steps_per_second": 6.921,
      "eval_token_acc": 0.4740853658536585,
      "step": 51000
    },
    {
      "epoch": 2.185210573668652,
      "grad_norm": 5.247880935668945,
      "learning_rate": 5.982504842551777e-05,
      "loss": 2.575246810913086,
      "memory(GiB)": 77.56,
      "step": 51005,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 2.1854247889979006,
      "grad_norm": 5.502328395843506,
      "learning_rate": 5.981844976963831e-05,
      "loss": 2.43762149810791,
      "memory(GiB)": 77.56,
      "step": 51010,
      "token_acc": 0.4668769716088328,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 2.1856390043271494,
      "grad_norm": 5.834925651550293,
      "learning_rate": 5.981185093588839e-05,
      "loss": 2.573842239379883,
      "memory(GiB)": 77.56,
      "step": 51015,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 2.1858532196563987,
      "grad_norm": 5.116877555847168,
      "learning_rate": 5.980525192438761e-05,
      "loss": 2.3351455688476563,
      "memory(GiB)": 77.56,
      "step": 51020,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.438095
    },
    {
      "epoch": 2.1860674349856475,
      "grad_norm": 4.3217902183532715,
      "learning_rate": 5.979865273525549e-05,
      "loss": 2.5266494750976562,
      "memory(GiB)": 77.56,
      "step": 51025,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.438082
    },
    {
      "epoch": 2.1862816503148967,
      "grad_norm": 5.159086227416992,
      "learning_rate": 5.9792053368611565e-05,
      "loss": 2.587189483642578,
      "memory(GiB)": 77.56,
      "step": 51030,
      "token_acc": 0.498567335243553,
      "train_speed(iter/s)": 1.438114
    },
    {
      "epoch": 2.1864958656441456,
      "grad_norm": 6.224012851715088,
      "learning_rate": 5.978545382457543e-05,
      "loss": 2.5629861831665037,
      "memory(GiB)": 77.56,
      "step": 51035,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.43814
    },
    {
      "epoch": 2.1867100809733944,
      "grad_norm": 6.302873611450195,
      "learning_rate": 5.977885410326661e-05,
      "loss": 2.541942596435547,
      "memory(GiB)": 77.56,
      "step": 51040,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.438169
    },
    {
      "epoch": 2.1869242963026436,
      "grad_norm": 8.17940616607666,
      "learning_rate": 5.977225420480468e-05,
      "loss": 2.209524726867676,
      "memory(GiB)": 77.56,
      "step": 51045,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.438195
    },
    {
      "epoch": 2.1871385116318924,
      "grad_norm": 5.27204704284668,
      "learning_rate": 5.97656541293092e-05,
      "loss": 2.1242889404296874,
      "memory(GiB)": 77.56,
      "step": 51050,
      "token_acc": 0.5285171102661597,
      "train_speed(iter/s)": 1.438221
    },
    {
      "epoch": 2.1873527269611412,
      "grad_norm": 6.283603668212891,
      "learning_rate": 5.975905387689973e-05,
      "loss": 2.445973777770996,
      "memory(GiB)": 77.56,
      "step": 51055,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.438217
    },
    {
      "epoch": 2.1875669422903905,
      "grad_norm": 4.831215858459473,
      "learning_rate": 5.9752453447695834e-05,
      "loss": 2.3898681640625,
      "memory(GiB)": 77.56,
      "step": 51060,
      "token_acc": 0.5117845117845118,
      "train_speed(iter/s)": 1.438232
    },
    {
      "epoch": 2.1877811576196393,
      "grad_norm": 5.528214931488037,
      "learning_rate": 5.974585284181712e-05,
      "loss": 2.543265533447266,
      "memory(GiB)": 77.56,
      "step": 51065,
      "token_acc": 0.43613707165109034,
      "train_speed(iter/s)": 1.438243
    },
    {
      "epoch": 2.187995372948888,
      "grad_norm": 4.38115930557251,
      "learning_rate": 5.973925205938311e-05,
      "loss": 2.4410064697265623,
      "memory(GiB)": 77.56,
      "step": 51070,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.438262
    },
    {
      "epoch": 2.1882095882781374,
      "grad_norm": 5.574254989624023,
      "learning_rate": 5.973265110051344e-05,
      "loss": 2.2609447479248046,
      "memory(GiB)": 77.56,
      "step": 51075,
      "token_acc": 0.5346820809248555,
      "train_speed(iter/s)": 1.438231
    },
    {
      "epoch": 2.188423803607386,
      "grad_norm": 5.398826599121094,
      "learning_rate": 5.9726049965327656e-05,
      "loss": 2.464238929748535,
      "memory(GiB)": 77.56,
      "step": 51080,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.43827
    },
    {
      "epoch": 2.188638018936635,
      "grad_norm": 5.23271369934082,
      "learning_rate": 5.9719448653945344e-05,
      "loss": 2.221196746826172,
      "memory(GiB)": 77.56,
      "step": 51085,
      "token_acc": 0.5652173913043478,
      "train_speed(iter/s)": 1.438283
    },
    {
      "epoch": 2.1888522342658843,
      "grad_norm": 5.4251909255981445,
      "learning_rate": 5.9712847166486105e-05,
      "loss": 2.4523088455200197,
      "memory(GiB)": 77.56,
      "step": 51090,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 2.189066449595133,
      "grad_norm": 4.862835884094238,
      "learning_rate": 5.9706245503069534e-05,
      "loss": 2.3679885864257812,
      "memory(GiB)": 77.56,
      "step": 51095,
      "token_acc": 0.5170068027210885,
      "train_speed(iter/s)": 1.438284
    },
    {
      "epoch": 2.189280664924382,
      "grad_norm": 6.4454121589660645,
      "learning_rate": 5.9699643663815205e-05,
      "loss": 2.5079633712768556,
      "memory(GiB)": 77.56,
      "step": 51100,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.438266
    },
    {
      "epoch": 2.189494880253631,
      "grad_norm": 5.596668720245361,
      "learning_rate": 5.969304164884275e-05,
      "loss": 2.5075233459472654,
      "memory(GiB)": 77.56,
      "step": 51105,
      "token_acc": 0.45182724252491696,
      "train_speed(iter/s)": 1.438302
    },
    {
      "epoch": 2.18970909558288,
      "grad_norm": 5.3330230712890625,
      "learning_rate": 5.968643945827176e-05,
      "loss": 2.5485298156738283,
      "memory(GiB)": 77.56,
      "step": 51110,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.438302
    },
    {
      "epoch": 2.1899233109121288,
      "grad_norm": 4.852079391479492,
      "learning_rate": 5.9679837092221815e-05,
      "loss": 2.387131690979004,
      "memory(GiB)": 77.56,
      "step": 51115,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.438276
    },
    {
      "epoch": 2.190137526241378,
      "grad_norm": 4.6014580726623535,
      "learning_rate": 5.967323455081255e-05,
      "loss": 2.265755271911621,
      "memory(GiB)": 77.56,
      "step": 51120,
      "token_acc": 0.5040650406504065,
      "train_speed(iter/s)": 1.438286
    },
    {
      "epoch": 2.190351741570627,
      "grad_norm": 6.057758808135986,
      "learning_rate": 5.966663183416357e-05,
      "loss": 2.3789167404174805,
      "memory(GiB)": 77.56,
      "step": 51125,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438306
    },
    {
      "epoch": 2.1905659568998757,
      "grad_norm": 6.111120700836182,
      "learning_rate": 5.966002894239446e-05,
      "loss": 2.5509174346923826,
      "memory(GiB)": 77.56,
      "step": 51130,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 2.190780172229125,
      "grad_norm": 5.215373516082764,
      "learning_rate": 5.965342587562489e-05,
      "loss": 2.3354923248291017,
      "memory(GiB)": 77.56,
      "step": 51135,
      "token_acc": 0.4956772334293948,
      "train_speed(iter/s)": 1.438385
    },
    {
      "epoch": 2.1909943875583737,
      "grad_norm": 5.302890777587891,
      "learning_rate": 5.9646822633974454e-05,
      "loss": 2.4579946517944338,
      "memory(GiB)": 77.56,
      "step": 51140,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438338
    },
    {
      "epoch": 2.1912086028876225,
      "grad_norm": 7.1614556312561035,
      "learning_rate": 5.964021921756277e-05,
      "loss": 2.5008249282836914,
      "memory(GiB)": 77.56,
      "step": 51145,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 2.191422818216872,
      "grad_norm": 5.565342903137207,
      "learning_rate": 5.963361562650946e-05,
      "loss": 2.4689800262451174,
      "memory(GiB)": 77.56,
      "step": 51150,
      "token_acc": 0.46987951807228917,
      "train_speed(iter/s)": 1.438352
    },
    {
      "epoch": 2.1916370335461206,
      "grad_norm": 4.475517272949219,
      "learning_rate": 5.962701186093419e-05,
      "loss": 2.5771085739135744,
      "memory(GiB)": 77.56,
      "step": 51155,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.438374
    },
    {
      "epoch": 2.1918512488753694,
      "grad_norm": 5.476842880249023,
      "learning_rate": 5.962040792095656e-05,
      "loss": 2.450410079956055,
      "memory(GiB)": 77.56,
      "step": 51160,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.438389
    },
    {
      "epoch": 2.1920654642046187,
      "grad_norm": 6.096722602844238,
      "learning_rate": 5.96138038066962e-05,
      "loss": 2.377358627319336,
      "memory(GiB)": 77.56,
      "step": 51165,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.438406
    },
    {
      "epoch": 2.1922796795338675,
      "grad_norm": 6.062929153442383,
      "learning_rate": 5.960719951827278e-05,
      "loss": 2.460158920288086,
      "memory(GiB)": 77.56,
      "step": 51170,
      "token_acc": 0.4910394265232975,
      "train_speed(iter/s)": 1.438379
    },
    {
      "epoch": 2.1924938948631163,
      "grad_norm": 5.080893039703369,
      "learning_rate": 5.960059505580593e-05,
      "loss": 2.382518196105957,
      "memory(GiB)": 77.56,
      "step": 51175,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.438372
    },
    {
      "epoch": 2.1927081101923656,
      "grad_norm": 6.196957111358643,
      "learning_rate": 5.9593990419415294e-05,
      "loss": 2.1168500900268556,
      "memory(GiB)": 77.56,
      "step": 51180,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.43839
    },
    {
      "epoch": 2.1929223255216144,
      "grad_norm": 7.209712982177734,
      "learning_rate": 5.9587385609220516e-05,
      "loss": 2.4931427001953126,
      "memory(GiB)": 77.56,
      "step": 51185,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.438404
    },
    {
      "epoch": 2.193136540850863,
      "grad_norm": 5.172015190124512,
      "learning_rate": 5.958078062534126e-05,
      "loss": 2.451609802246094,
      "memory(GiB)": 77.56,
      "step": 51190,
      "token_acc": 0.4983277591973244,
      "train_speed(iter/s)": 1.438421
    },
    {
      "epoch": 2.1933507561801124,
      "grad_norm": 7.0667901039123535,
      "learning_rate": 5.957417546789717e-05,
      "loss": 2.3835845947265626,
      "memory(GiB)": 77.56,
      "step": 51195,
      "token_acc": 0.5405405405405406,
      "train_speed(iter/s)": 1.438419
    },
    {
      "epoch": 2.1935649715093612,
      "grad_norm": 6.388824939727783,
      "learning_rate": 5.956757013700791e-05,
      "loss": 2.527954864501953,
      "memory(GiB)": 77.56,
      "step": 51200,
      "token_acc": 0.5150214592274678,
      "train_speed(iter/s)": 1.438443
    },
    {
      "epoch": 2.19377918683861,
      "grad_norm": 6.017939567565918,
      "learning_rate": 5.956096463279314e-05,
      "loss": 2.4327980041503907,
      "memory(GiB)": 77.56,
      "step": 51205,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.43848
    },
    {
      "epoch": 2.1939934021678593,
      "grad_norm": 5.817434310913086,
      "learning_rate": 5.955435895537253e-05,
      "loss": 2.4561895370483398,
      "memory(GiB)": 77.56,
      "step": 51210,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.438491
    },
    {
      "epoch": 2.194207617497108,
      "grad_norm": 4.726731777191162,
      "learning_rate": 5.9547753104865746e-05,
      "loss": 2.4411893844604493,
      "memory(GiB)": 77.56,
      "step": 51215,
      "token_acc": 0.4816326530612245,
      "train_speed(iter/s)": 1.438498
    },
    {
      "epoch": 2.194421832826357,
      "grad_norm": 8.09035587310791,
      "learning_rate": 5.954114708139247e-05,
      "loss": 2.5014892578125,
      "memory(GiB)": 77.56,
      "step": 51220,
      "token_acc": 0.453125,
      "train_speed(iter/s)": 1.438489
    },
    {
      "epoch": 2.194636048155606,
      "grad_norm": 5.522891521453857,
      "learning_rate": 5.953454088507236e-05,
      "loss": 2.330046272277832,
      "memory(GiB)": 77.56,
      "step": 51225,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.438498
    },
    {
      "epoch": 2.194850263484855,
      "grad_norm": 6.340538501739502,
      "learning_rate": 5.952793451602507e-05,
      "loss": 2.178668975830078,
      "memory(GiB)": 77.56,
      "step": 51230,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.438496
    },
    {
      "epoch": 2.195064478814104,
      "grad_norm": 5.124255657196045,
      "learning_rate": 5.9521327974370334e-05,
      "loss": 2.484984588623047,
      "memory(GiB)": 77.56,
      "step": 51235,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.438513
    },
    {
      "epoch": 2.195278694143353,
      "grad_norm": 7.406636714935303,
      "learning_rate": 5.95147212602278e-05,
      "loss": 2.5710737228393556,
      "memory(GiB)": 77.56,
      "step": 51240,
      "token_acc": 0.4476744186046512,
      "train_speed(iter/s)": 1.438484
    },
    {
      "epoch": 2.195492909472602,
      "grad_norm": 5.909469127655029,
      "learning_rate": 5.950811437371716e-05,
      "loss": 2.448769760131836,
      "memory(GiB)": 77.56,
      "step": 51245,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.438464
    },
    {
      "epoch": 2.1957071248018507,
      "grad_norm": 5.257267475128174,
      "learning_rate": 5.950150731495813e-05,
      "loss": 2.4028879165649415,
      "memory(GiB)": 77.56,
      "step": 51250,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.438479
    },
    {
      "epoch": 2.1959213401311,
      "grad_norm": 4.758358478546143,
      "learning_rate": 5.949490008407037e-05,
      "loss": 2.5705322265625,
      "memory(GiB)": 77.56,
      "step": 51255,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.438516
    },
    {
      "epoch": 2.1961355554603488,
      "grad_norm": 5.432770729064941,
      "learning_rate": 5.9488292681173585e-05,
      "loss": 2.188041114807129,
      "memory(GiB)": 77.56,
      "step": 51260,
      "token_acc": 0.5475285171102662,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 2.1963497707895976,
      "grad_norm": 4.859992504119873,
      "learning_rate": 5.948168510638748e-05,
      "loss": 2.528558921813965,
      "memory(GiB)": 77.56,
      "step": 51265,
      "token_acc": 0.4830188679245283,
      "train_speed(iter/s)": 1.438527
    },
    {
      "epoch": 2.196563986118847,
      "grad_norm": 6.28976583480835,
      "learning_rate": 5.9475077359831766e-05,
      "loss": 2.1902523040771484,
      "memory(GiB)": 77.56,
      "step": 51270,
      "token_acc": 0.5451388888888888,
      "train_speed(iter/s)": 1.438523
    },
    {
      "epoch": 2.1967782014480957,
      "grad_norm": 4.740793228149414,
      "learning_rate": 5.9468469441626116e-05,
      "loss": 2.7493324279785156,
      "memory(GiB)": 77.56,
      "step": 51275,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.438523
    },
    {
      "epoch": 2.1969924167773445,
      "grad_norm": 6.288125514984131,
      "learning_rate": 5.946186135189027e-05,
      "loss": 2.7736921310424805,
      "memory(GiB)": 77.56,
      "step": 51280,
      "token_acc": 0.4843205574912892,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 2.1972066321065937,
      "grad_norm": 6.039641857147217,
      "learning_rate": 5.945525309074393e-05,
      "loss": 2.4235759735107423,
      "memory(GiB)": 77.56,
      "step": 51285,
      "token_acc": 0.4948805460750853,
      "train_speed(iter/s)": 1.438526
    },
    {
      "epoch": 2.1974208474358425,
      "grad_norm": 5.337045669555664,
      "learning_rate": 5.944864465830681e-05,
      "loss": 2.7233800888061523,
      "memory(GiB)": 77.56,
      "step": 51290,
      "token_acc": 0.48348348348348347,
      "train_speed(iter/s)": 1.438566
    },
    {
      "epoch": 2.1976350627650914,
      "grad_norm": 6.917486667633057,
      "learning_rate": 5.944203605469863e-05,
      "loss": 2.5187665939331056,
      "memory(GiB)": 77.56,
      "step": 51295,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.43859
    },
    {
      "epoch": 2.1978492780943406,
      "grad_norm": 4.805302143096924,
      "learning_rate": 5.943542728003911e-05,
      "loss": 2.369694137573242,
      "memory(GiB)": 77.56,
      "step": 51300,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 2.1980634934235894,
      "grad_norm": 5.587535858154297,
      "learning_rate": 5.9428818334447976e-05,
      "loss": 2.416248893737793,
      "memory(GiB)": 77.56,
      "step": 51305,
      "token_acc": 0.47232472324723246,
      "train_speed(iter/s)": 1.438618
    },
    {
      "epoch": 2.1982777087528382,
      "grad_norm": 5.735699653625488,
      "learning_rate": 5.9422209218044956e-05,
      "loss": 2.4903430938720703,
      "memory(GiB)": 77.56,
      "step": 51310,
      "token_acc": 0.4868804664723032,
      "train_speed(iter/s)": 1.438651
    },
    {
      "epoch": 2.1984919240820875,
      "grad_norm": 5.903115749359131,
      "learning_rate": 5.941559993094976e-05,
      "loss": 2.28411808013916,
      "memory(GiB)": 77.56,
      "step": 51315,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.1987061394113363,
      "grad_norm": 4.929615497589111,
      "learning_rate": 5.9408990473282145e-05,
      "loss": 2.1097713470458985,
      "memory(GiB)": 77.56,
      "step": 51320,
      "token_acc": 0.5433333333333333,
      "train_speed(iter/s)": 1.43864
    },
    {
      "epoch": 2.198920354740585,
      "grad_norm": 5.301702499389648,
      "learning_rate": 5.940238084516184e-05,
      "loss": 2.680929183959961,
      "memory(GiB)": 77.56,
      "step": 51325,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.438613
    },
    {
      "epoch": 2.1991345700698344,
      "grad_norm": 5.708707809448242,
      "learning_rate": 5.9395771046708594e-05,
      "loss": 2.2984172821044924,
      "memory(GiB)": 77.56,
      "step": 51330,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.438589
    },
    {
      "epoch": 2.199348785399083,
      "grad_norm": 5.186180114746094,
      "learning_rate": 5.9389161078042143e-05,
      "loss": 2.2402292251586915,
      "memory(GiB)": 77.56,
      "step": 51335,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.438577
    },
    {
      "epoch": 2.199563000728332,
      "grad_norm": 4.457876205444336,
      "learning_rate": 5.9382550939282234e-05,
      "loss": 2.462636184692383,
      "memory(GiB)": 77.56,
      "step": 51340,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.1997772160575813,
      "grad_norm": 4.908180236816406,
      "learning_rate": 5.9375940630548597e-05,
      "loss": 2.3931446075439453,
      "memory(GiB)": 77.56,
      "step": 51345,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.43856
    },
    {
      "epoch": 2.19999143138683,
      "grad_norm": 5.65999174118042,
      "learning_rate": 5.9369330151961e-05,
      "loss": 2.395417594909668,
      "memory(GiB)": 77.56,
      "step": 51350,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 2.200205646716079,
      "grad_norm": 5.591470241546631,
      "learning_rate": 5.9362719503639216e-05,
      "loss": 2.4952018737792967,
      "memory(GiB)": 77.56,
      "step": 51355,
      "token_acc": 0.4420731707317073,
      "train_speed(iter/s)": 1.438561
    },
    {
      "epoch": 2.200419862045328,
      "grad_norm": 4.321208953857422,
      "learning_rate": 5.9356108685702974e-05,
      "loss": 2.672606658935547,
      "memory(GiB)": 77.56,
      "step": 51360,
      "token_acc": 0.49375,
      "train_speed(iter/s)": 1.438555
    },
    {
      "epoch": 2.200634077374577,
      "grad_norm": 4.107922554016113,
      "learning_rate": 5.934949769827205e-05,
      "loss": 2.3642757415771483,
      "memory(GiB)": 77.56,
      "step": 51365,
      "token_acc": 0.5015673981191222,
      "train_speed(iter/s)": 1.438568
    },
    {
      "epoch": 2.2008482927038258,
      "grad_norm": 4.427626132965088,
      "learning_rate": 5.9342886541466204e-05,
      "loss": 2.5454174041748048,
      "memory(GiB)": 77.56,
      "step": 51370,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438583
    },
    {
      "epoch": 2.201062508033075,
      "grad_norm": 5.962900638580322,
      "learning_rate": 5.93362752154052e-05,
      "loss": 2.4481891632080077,
      "memory(GiB)": 77.56,
      "step": 51375,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 2.201276723362324,
      "grad_norm": 5.660308361053467,
      "learning_rate": 5.9329663720208826e-05,
      "loss": 2.493613433837891,
      "memory(GiB)": 77.56,
      "step": 51380,
      "token_acc": 0.44569288389513106,
      "train_speed(iter/s)": 1.438574
    },
    {
      "epoch": 2.2014909386915726,
      "grad_norm": 5.489604473114014,
      "learning_rate": 5.932305205599683e-05,
      "loss": 2.7598297119140627,
      "memory(GiB)": 77.56,
      "step": 51385,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 2.201705154020822,
      "grad_norm": 5.936769008636475,
      "learning_rate": 5.931644022288899e-05,
      "loss": 2.371737480163574,
      "memory(GiB)": 77.56,
      "step": 51390,
      "token_acc": 0.45089285714285715,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 2.2019193693500707,
      "grad_norm": 4.960988521575928,
      "learning_rate": 5.9309828221005115e-05,
      "loss": 2.4654918670654298,
      "memory(GiB)": 77.56,
      "step": 51395,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.438543
    },
    {
      "epoch": 2.2021335846793195,
      "grad_norm": 6.509946346282959,
      "learning_rate": 5.930321605046496e-05,
      "loss": 2.7539703369140627,
      "memory(GiB)": 77.56,
      "step": 51400,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.43856
    },
    {
      "epoch": 2.202347800008569,
      "grad_norm": 5.776191234588623,
      "learning_rate": 5.9296603711388324e-05,
      "loss": 2.575841522216797,
      "memory(GiB)": 77.56,
      "step": 51405,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.438594
    },
    {
      "epoch": 2.2025620153378176,
      "grad_norm": 6.656833171844482,
      "learning_rate": 5.928999120389499e-05,
      "loss": 2.5711742401123048,
      "memory(GiB)": 77.56,
      "step": 51410,
      "token_acc": 0.47706422018348627,
      "train_speed(iter/s)": 1.43859
    },
    {
      "epoch": 2.2027762306670664,
      "grad_norm": 4.375270366668701,
      "learning_rate": 5.928337852810475e-05,
      "loss": 2.601535606384277,
      "memory(GiB)": 77.56,
      "step": 51415,
      "token_acc": 0.452991452991453,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 2.2029904459963157,
      "grad_norm": 4.985630989074707,
      "learning_rate": 5.927676568413739e-05,
      "loss": 2.402043342590332,
      "memory(GiB)": 77.56,
      "step": 51420,
      "token_acc": 0.48220064724919093,
      "train_speed(iter/s)": 1.438544
    },
    {
      "epoch": 2.2032046613255645,
      "grad_norm": 5.533224105834961,
      "learning_rate": 5.9270152672112725e-05,
      "loss": 2.276856231689453,
      "memory(GiB)": 77.56,
      "step": 51425,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.43859
    },
    {
      "epoch": 2.2034188766548133,
      "grad_norm": 5.177923202514648,
      "learning_rate": 5.9263539492150557e-05,
      "loss": 2.5078428268432615,
      "memory(GiB)": 77.56,
      "step": 51430,
      "token_acc": 0.4907749077490775,
      "train_speed(iter/s)": 1.438578
    },
    {
      "epoch": 2.2036330919840625,
      "grad_norm": 5.188407897949219,
      "learning_rate": 5.9256926144370663e-05,
      "loss": 2.50731201171875,
      "memory(GiB)": 77.56,
      "step": 51435,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.438539
    },
    {
      "epoch": 2.2038473073133114,
      "grad_norm": 8.031177520751953,
      "learning_rate": 5.9250312628892877e-05,
      "loss": 2.6311120986938477,
      "memory(GiB)": 77.56,
      "step": 51440,
      "token_acc": 0.4784172661870504,
      "train_speed(iter/s)": 1.438546
    },
    {
      "epoch": 2.20406152264256,
      "grad_norm": 4.27857780456543,
      "learning_rate": 5.9243698945837014e-05,
      "loss": 2.434372329711914,
      "memory(GiB)": 77.56,
      "step": 51445,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 2.2042757379718094,
      "grad_norm": 5.051950454711914,
      "learning_rate": 5.923708509532284e-05,
      "loss": 2.433517646789551,
      "memory(GiB)": 77.56,
      "step": 51450,
      "token_acc": 0.46996466431095407,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 2.2044899533010582,
      "grad_norm": 5.938570022583008,
      "learning_rate": 5.923047107747024e-05,
      "loss": 2.32891731262207,
      "memory(GiB)": 77.56,
      "step": 51455,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.438583
    },
    {
      "epoch": 2.204704168630307,
      "grad_norm": 5.0678324699401855,
      "learning_rate": 5.9223856892398975e-05,
      "loss": 2.3945594787597657,
      "memory(GiB)": 77.56,
      "step": 51460,
      "token_acc": 0.5201238390092879,
      "train_speed(iter/s)": 1.438617
    },
    {
      "epoch": 2.2049183839595563,
      "grad_norm": 5.1532158851623535,
      "learning_rate": 5.921724254022889e-05,
      "loss": 2.245484161376953,
      "memory(GiB)": 77.56,
      "step": 51465,
      "token_acc": 0.49557522123893805,
      "train_speed(iter/s)": 1.438589
    },
    {
      "epoch": 2.205132599288805,
      "grad_norm": 6.8095808029174805,
      "learning_rate": 5.921062802107982e-05,
      "loss": 2.64456844329834,
      "memory(GiB)": 77.56,
      "step": 51470,
      "token_acc": 0.48046875,
      "train_speed(iter/s)": 1.43861
    },
    {
      "epoch": 2.205346814618054,
      "grad_norm": 6.624456882476807,
      "learning_rate": 5.920401333507157e-05,
      "loss": 2.413959503173828,
      "memory(GiB)": 77.56,
      "step": 51475,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.438627
    },
    {
      "epoch": 2.205561029947303,
      "grad_norm": 5.202213764190674,
      "learning_rate": 5.9197398482324e-05,
      "loss": 2.625692939758301,
      "memory(GiB)": 77.56,
      "step": 51480,
      "token_acc": 0.4707792207792208,
      "train_speed(iter/s)": 1.43867
    },
    {
      "epoch": 2.205775245276552,
      "grad_norm": 5.417222023010254,
      "learning_rate": 5.919078346295693e-05,
      "loss": 2.4275264739990234,
      "memory(GiB)": 77.56,
      "step": 51485,
      "token_acc": 0.4735202492211838,
      "train_speed(iter/s)": 1.43869
    },
    {
      "epoch": 2.205989460605801,
      "grad_norm": 5.424493789672852,
      "learning_rate": 5.918416827709018e-05,
      "loss": 2.1249998092651365,
      "memory(GiB)": 77.56,
      "step": 51490,
      "token_acc": 0.5241379310344828,
      "train_speed(iter/s)": 1.438659
    },
    {
      "epoch": 2.20620367593505,
      "grad_norm": 5.903954029083252,
      "learning_rate": 5.917755292484361e-05,
      "loss": 2.3350547790527343,
      "memory(GiB)": 77.56,
      "step": 51495,
      "token_acc": 0.5118343195266272,
      "train_speed(iter/s)": 1.438679
    },
    {
      "epoch": 2.206417891264299,
      "grad_norm": 4.104979038238525,
      "learning_rate": 5.917093740633707e-05,
      "loss": 2.266367721557617,
      "memory(GiB)": 77.56,
      "step": 51500,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.438671
    },
    {
      "epoch": 2.206417891264299,
      "eval_loss": 2.3244543075561523,
      "eval_runtime": 14.2381,
      "eval_samples_per_second": 7.023,
      "eval_steps_per_second": 7.023,
      "eval_token_acc": 0.4762532981530343,
      "step": 51500
    },
    {
      "epoch": 2.2066321065935477,
      "grad_norm": 4.286702632904053,
      "learning_rate": 5.916432172169038e-05,
      "loss": 2.5060062408447266,
      "memory(GiB)": 77.56,
      "step": 51505,
      "token_acc": 0.4753623188405797,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 2.206846321922797,
      "grad_norm": 6.096134185791016,
      "learning_rate": 5.9157705871023426e-05,
      "loss": 2.478118133544922,
      "memory(GiB)": 77.56,
      "step": 51510,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.438036
    },
    {
      "epoch": 2.2070605372520458,
      "grad_norm": 4.494735240936279,
      "learning_rate": 5.915108985445603e-05,
      "loss": 2.555459403991699,
      "memory(GiB)": 77.56,
      "step": 51515,
      "token_acc": 0.4759036144578313,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 2.2072747525812946,
      "grad_norm": 6.113365650177002,
      "learning_rate": 5.914447367210805e-05,
      "loss": 2.418491172790527,
      "memory(GiB)": 77.56,
      "step": 51520,
      "token_acc": 0.5477031802120141,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 2.207488967910544,
      "grad_norm": 6.196977138519287,
      "learning_rate": 5.913785732409937e-05,
      "loss": 2.4084068298339845,
      "memory(GiB)": 77.56,
      "step": 51525,
      "token_acc": 0.488,
      "train_speed(iter/s)": 1.438027
    },
    {
      "epoch": 2.2077031832397926,
      "grad_norm": 6.043222427368164,
      "learning_rate": 5.913124081054981e-05,
      "loss": 1.9862640380859375,
      "memory(GiB)": 77.56,
      "step": 51530,
      "token_acc": 0.5401785714285714,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 2.2079173985690415,
      "grad_norm": 4.764535427093506,
      "learning_rate": 5.912462413157926e-05,
      "loss": 2.5969568252563477,
      "memory(GiB)": 77.56,
      "step": 51535,
      "token_acc": 0.43823529411764706,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 2.2081316138982907,
      "grad_norm": 5.156094074249268,
      "learning_rate": 5.91180072873076e-05,
      "loss": 2.5567049026489257,
      "memory(GiB)": 77.56,
      "step": 51540,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438104
    },
    {
      "epoch": 2.2083458292275395,
      "grad_norm": 4.799515247344971,
      "learning_rate": 5.9111390277854675e-05,
      "loss": 2.700315475463867,
      "memory(GiB)": 77.56,
      "step": 51545,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438136
    },
    {
      "epoch": 2.2085600445567883,
      "grad_norm": 4.821591377258301,
      "learning_rate": 5.910477310334036e-05,
      "loss": 2.553404426574707,
      "memory(GiB)": 77.56,
      "step": 51550,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.438164
    },
    {
      "epoch": 2.2087742598860376,
      "grad_norm": 6.490218162536621,
      "learning_rate": 5.9098155763884554e-05,
      "loss": 2.213795852661133,
      "memory(GiB)": 77.56,
      "step": 51555,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.438178
    },
    {
      "epoch": 2.2089884752152864,
      "grad_norm": 6.389596939086914,
      "learning_rate": 5.909153825960711e-05,
      "loss": 2.5539928436279298,
      "memory(GiB)": 77.56,
      "step": 51560,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438155
    },
    {
      "epoch": 2.209202690544535,
      "grad_norm": 6.180983543395996,
      "learning_rate": 5.908492059062794e-05,
      "loss": 2.3557964324951173,
      "memory(GiB)": 77.56,
      "step": 51565,
      "token_acc": 0.5109170305676856,
      "train_speed(iter/s)": 1.438161
    },
    {
      "epoch": 2.2094169058737845,
      "grad_norm": 5.574581623077393,
      "learning_rate": 5.907830275706689e-05,
      "loss": 2.4755584716796877,
      "memory(GiB)": 77.56,
      "step": 51570,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.438156
    },
    {
      "epoch": 2.2096311212030333,
      "grad_norm": 4.493891716003418,
      "learning_rate": 5.907168475904388e-05,
      "loss": 2.510402488708496,
      "memory(GiB)": 77.56,
      "step": 51575,
      "token_acc": 0.5146579804560261,
      "train_speed(iter/s)": 1.438189
    },
    {
      "epoch": 2.209845336532282,
      "grad_norm": 5.137444972991943,
      "learning_rate": 5.906506659667878e-05,
      "loss": 2.4336313247680663,
      "memory(GiB)": 77.56,
      "step": 51580,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.438213
    },
    {
      "epoch": 2.2100595518615314,
      "grad_norm": 6.816127777099609,
      "learning_rate": 5.905844827009151e-05,
      "loss": 1.9409049987792968,
      "memory(GiB)": 77.56,
      "step": 51585,
      "token_acc": 0.5714285714285714,
      "train_speed(iter/s)": 1.438235
    },
    {
      "epoch": 2.21027376719078,
      "grad_norm": 5.381969451904297,
      "learning_rate": 5.905182977940195e-05,
      "loss": 2.7115360260009767,
      "memory(GiB)": 77.56,
      "step": 51590,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.438228
    },
    {
      "epoch": 2.210487982520029,
      "grad_norm": 6.080172061920166,
      "learning_rate": 5.904521112472999e-05,
      "loss": 2.5088022232055662,
      "memory(GiB)": 77.56,
      "step": 51595,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.438211
    },
    {
      "epoch": 2.2107021978492782,
      "grad_norm": 4.577946662902832,
      "learning_rate": 5.903859230619556e-05,
      "loss": 2.7236812591552733,
      "memory(GiB)": 77.56,
      "step": 51600,
      "token_acc": 0.4752475247524752,
      "train_speed(iter/s)": 1.438202
    },
    {
      "epoch": 2.210916413178527,
      "grad_norm": 5.804032802581787,
      "learning_rate": 5.903197332391853e-05,
      "loss": 2.7177486419677734,
      "memory(GiB)": 77.56,
      "step": 51605,
      "token_acc": 0.4200626959247649,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 2.211130628507776,
      "grad_norm": 5.156857967376709,
      "learning_rate": 5.902535417801884e-05,
      "loss": 2.487660598754883,
      "memory(GiB)": 77.56,
      "step": 51610,
      "token_acc": 0.49266862170087977,
      "train_speed(iter/s)": 1.438213
    },
    {
      "epoch": 2.211344843837025,
      "grad_norm": 3.858626365661621,
      "learning_rate": 5.901873486861641e-05,
      "loss": 2.305619812011719,
      "memory(GiB)": 77.56,
      "step": 51615,
      "token_acc": 0.5140562248995983,
      "train_speed(iter/s)": 1.438194
    },
    {
      "epoch": 2.211559059166274,
      "grad_norm": 4.362866401672363,
      "learning_rate": 5.901211539583111e-05,
      "loss": 2.546542167663574,
      "memory(GiB)": 77.56,
      "step": 51620,
      "token_acc": 0.47129909365558914,
      "train_speed(iter/s)": 1.438222
    },
    {
      "epoch": 2.2117732744955227,
      "grad_norm": 5.550348281860352,
      "learning_rate": 5.900549575978291e-05,
      "loss": 2.7103553771972657,
      "memory(GiB)": 77.56,
      "step": 51625,
      "token_acc": 0.41358024691358025,
      "train_speed(iter/s)": 1.438227
    },
    {
      "epoch": 2.211987489824772,
      "grad_norm": 7.457678318023682,
      "learning_rate": 5.899887596059171e-05,
      "loss": 2.773662567138672,
      "memory(GiB)": 77.56,
      "step": 51630,
      "token_acc": 0.4503105590062112,
      "train_speed(iter/s)": 1.438234
    },
    {
      "epoch": 2.212201705154021,
      "grad_norm": 5.990923881530762,
      "learning_rate": 5.899225599837741e-05,
      "loss": 2.7374265670776365,
      "memory(GiB)": 77.56,
      "step": 51635,
      "token_acc": 0.43537414965986393,
      "train_speed(iter/s)": 1.438209
    },
    {
      "epoch": 2.2124159204832696,
      "grad_norm": 5.70811653137207,
      "learning_rate": 5.8985635873259956e-05,
      "loss": 2.471603775024414,
      "memory(GiB)": 77.56,
      "step": 51640,
      "token_acc": 0.46726190476190477,
      "train_speed(iter/s)": 1.438214
    },
    {
      "epoch": 2.212630135812519,
      "grad_norm": 4.653250217437744,
      "learning_rate": 5.8979015585359296e-05,
      "loss": 2.6476161956787108,
      "memory(GiB)": 77.56,
      "step": 51645,
      "token_acc": 0.4369747899159664,
      "train_speed(iter/s)": 1.438204
    },
    {
      "epoch": 2.2128443511417677,
      "grad_norm": 5.632009029388428,
      "learning_rate": 5.897239513479532e-05,
      "loss": 2.2976472854614256,
      "memory(GiB)": 77.56,
      "step": 51650,
      "token_acc": 0.4674329501915709,
      "train_speed(iter/s)": 1.438238
    },
    {
      "epoch": 2.2130585664710165,
      "grad_norm": 5.307003974914551,
      "learning_rate": 5.896577452168801e-05,
      "loss": 2.5293455123901367,
      "memory(GiB)": 77.56,
      "step": 51655,
      "token_acc": 0.48909657320872274,
      "train_speed(iter/s)": 1.438277
    },
    {
      "epoch": 2.2132727818002658,
      "grad_norm": 4.1953125,
      "learning_rate": 5.8959153746157294e-05,
      "loss": 2.3829559326171874,
      "memory(GiB)": 77.56,
      "step": 51660,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.43826
    },
    {
      "epoch": 2.2134869971295146,
      "grad_norm": 5.508793830871582,
      "learning_rate": 5.895253280832308e-05,
      "loss": 2.4228404998779296,
      "memory(GiB)": 77.56,
      "step": 51665,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438262
    },
    {
      "epoch": 2.2137012124587634,
      "grad_norm": 5.665297508239746,
      "learning_rate": 5.894591170830536e-05,
      "loss": 2.4385372161865235,
      "memory(GiB)": 77.56,
      "step": 51670,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438291
    },
    {
      "epoch": 2.2139154277880126,
      "grad_norm": 5.090893268585205,
      "learning_rate": 5.893929044622404e-05,
      "loss": 2.1795680999755858,
      "memory(GiB)": 77.56,
      "step": 51675,
      "token_acc": 0.5359477124183006,
      "train_speed(iter/s)": 1.438304
    },
    {
      "epoch": 2.2141296431172615,
      "grad_norm": 4.7929887771606445,
      "learning_rate": 5.8932669022199095e-05,
      "loss": 2.09849853515625,
      "memory(GiB)": 77.56,
      "step": 51680,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 2.2143438584465103,
      "grad_norm": 6.889700889587402,
      "learning_rate": 5.892604743635045e-05,
      "loss": 2.566457748413086,
      "memory(GiB)": 77.56,
      "step": 51685,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 2.2145580737757595,
      "grad_norm": 5.650369167327881,
      "learning_rate": 5.891942568879811e-05,
      "loss": 2.613294792175293,
      "memory(GiB)": 77.56,
      "step": 51690,
      "token_acc": 0.47333333333333333,
      "train_speed(iter/s)": 1.438343
    },
    {
      "epoch": 2.2147722891050083,
      "grad_norm": 4.616402626037598,
      "learning_rate": 5.8912803779662e-05,
      "loss": 2.2758647918701174,
      "memory(GiB)": 77.56,
      "step": 51695,
      "token_acc": 0.47257383966244726,
      "train_speed(iter/s)": 1.438365
    },
    {
      "epoch": 2.214986504434257,
      "grad_norm": 6.8922810554504395,
      "learning_rate": 5.890618170906208e-05,
      "loss": 2.462801933288574,
      "memory(GiB)": 77.56,
      "step": 51700,
      "token_acc": 0.4769736842105263,
      "train_speed(iter/s)": 1.438396
    },
    {
      "epoch": 2.2152007197635064,
      "grad_norm": 4.460061073303223,
      "learning_rate": 5.889955947711834e-05,
      "loss": 2.406911087036133,
      "memory(GiB)": 77.56,
      "step": 51705,
      "token_acc": 0.5037878787878788,
      "train_speed(iter/s)": 1.438404
    },
    {
      "epoch": 2.215414935092755,
      "grad_norm": 6.122963905334473,
      "learning_rate": 5.8892937083950704e-05,
      "loss": 2.2896242141723633,
      "memory(GiB)": 77.56,
      "step": 51710,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.438421
    },
    {
      "epoch": 2.215629150422004,
      "grad_norm": 4.520376682281494,
      "learning_rate": 5.8886314529679196e-05,
      "loss": 2.5812835693359375,
      "memory(GiB)": 77.56,
      "step": 51715,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.438388
    },
    {
      "epoch": 2.2158433657512533,
      "grad_norm": 4.933167934417725,
      "learning_rate": 5.8879691814423744e-05,
      "loss": 2.4599193572998046,
      "memory(GiB)": 77.56,
      "step": 51720,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.438407
    },
    {
      "epoch": 2.216057581080502,
      "grad_norm": 4.681436538696289,
      "learning_rate": 5.8873068938304355e-05,
      "loss": 2.5470909118652343,
      "memory(GiB)": 77.56,
      "step": 51725,
      "token_acc": 0.4489051094890511,
      "train_speed(iter/s)": 1.438426
    },
    {
      "epoch": 2.216271796409751,
      "grad_norm": 5.931581974029541,
      "learning_rate": 5.8866445901441e-05,
      "loss": 2.547966194152832,
      "memory(GiB)": 77.56,
      "step": 51730,
      "token_acc": 0.4708029197080292,
      "train_speed(iter/s)": 1.43845
    },
    {
      "epoch": 2.216486011739,
      "grad_norm": 6.11769437789917,
      "learning_rate": 5.885982270395366e-05,
      "loss": 2.3104244232177735,
      "memory(GiB)": 77.56,
      "step": 51735,
      "token_acc": 0.5205992509363296,
      "train_speed(iter/s)": 1.43843
    },
    {
      "epoch": 2.216700227068249,
      "grad_norm": 4.562624931335449,
      "learning_rate": 5.885319934596233e-05,
      "loss": 2.6275177001953125,
      "memory(GiB)": 77.56,
      "step": 51740,
      "token_acc": 0.44807121661721067,
      "train_speed(iter/s)": 1.438447
    },
    {
      "epoch": 2.216914442397498,
      "grad_norm": 5.758816242218018,
      "learning_rate": 5.884657582758698e-05,
      "loss": 2.2551509857177736,
      "memory(GiB)": 77.56,
      "step": 51745,
      "token_acc": 0.5230125523012552,
      "train_speed(iter/s)": 1.43847
    },
    {
      "epoch": 2.217128657726747,
      "grad_norm": 6.4284467697143555,
      "learning_rate": 5.8839952148947594e-05,
      "loss": 2.382111740112305,
      "memory(GiB)": 77.56,
      "step": 51750,
      "token_acc": 0.5183946488294314,
      "train_speed(iter/s)": 1.438486
    },
    {
      "epoch": 2.217342873055996,
      "grad_norm": 5.3027143478393555,
      "learning_rate": 5.8833328310164215e-05,
      "loss": 2.307453155517578,
      "memory(GiB)": 77.56,
      "step": 51755,
      "token_acc": 0.5648854961832062,
      "train_speed(iter/s)": 1.438507
    },
    {
      "epoch": 2.2175570883852447,
      "grad_norm": 7.027342796325684,
      "learning_rate": 5.882670431135677e-05,
      "loss": 2.6029586791992188,
      "memory(GiB)": 77.56,
      "step": 51760,
      "token_acc": 0.4329896907216495,
      "train_speed(iter/s)": 1.438504
    },
    {
      "epoch": 2.217771303714494,
      "grad_norm": 4.677776336669922,
      "learning_rate": 5.882008015264532e-05,
      "loss": 2.6335948944091796,
      "memory(GiB)": 77.56,
      "step": 51765,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.438501
    },
    {
      "epoch": 2.2179855190437427,
      "grad_norm": 5.307129383087158,
      "learning_rate": 5.8813455834149846e-05,
      "loss": 2.38995304107666,
      "memory(GiB)": 77.56,
      "step": 51770,
      "token_acc": 0.4703389830508475,
      "train_speed(iter/s)": 1.438518
    },
    {
      "epoch": 2.2181997343729916,
      "grad_norm": 5.592504501342773,
      "learning_rate": 5.880683135599034e-05,
      "loss": 2.632982635498047,
      "memory(GiB)": 77.56,
      "step": 51775,
      "token_acc": 0.4177215189873418,
      "train_speed(iter/s)": 1.438551
    },
    {
      "epoch": 2.218413949702241,
      "grad_norm": 7.6745524406433105,
      "learning_rate": 5.880020671828683e-05,
      "loss": 2.3372930526733398,
      "memory(GiB)": 77.56,
      "step": 51780,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.438552
    },
    {
      "epoch": 2.2186281650314896,
      "grad_norm": 7.014909744262695,
      "learning_rate": 5.879358192115932e-05,
      "loss": 2.380252647399902,
      "memory(GiB)": 77.56,
      "step": 51785,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.438534
    },
    {
      "epoch": 2.2188423803607384,
      "grad_norm": 4.42019510269165,
      "learning_rate": 5.8786956964727834e-05,
      "loss": 2.44866828918457,
      "memory(GiB)": 77.56,
      "step": 51790,
      "token_acc": 0.5100864553314121,
      "train_speed(iter/s)": 1.438503
    },
    {
      "epoch": 2.2190565956899877,
      "grad_norm": 5.494714736938477,
      "learning_rate": 5.878033184911236e-05,
      "loss": 2.292796516418457,
      "memory(GiB)": 77.56,
      "step": 51795,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.438519
    },
    {
      "epoch": 2.2192708110192365,
      "grad_norm": 6.420318603515625,
      "learning_rate": 5.877370657443294e-05,
      "loss": 2.430602264404297,
      "memory(GiB)": 77.56,
      "step": 51800,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 2.2194850263484853,
      "grad_norm": 4.252730846405029,
      "learning_rate": 5.876708114080961e-05,
      "loss": 2.567221832275391,
      "memory(GiB)": 77.56,
      "step": 51805,
      "token_acc": 0.4670487106017192,
      "train_speed(iter/s)": 1.438525
    },
    {
      "epoch": 2.2196992416777346,
      "grad_norm": 4.965419769287109,
      "learning_rate": 5.876045554836237e-05,
      "loss": 2.1858219146728515,
      "memory(GiB)": 77.56,
      "step": 51810,
      "token_acc": 0.556420233463035,
      "train_speed(iter/s)": 1.438544
    },
    {
      "epoch": 2.2199134570069834,
      "grad_norm": 4.175282001495361,
      "learning_rate": 5.875382979721127e-05,
      "loss": 2.6881969451904295,
      "memory(GiB)": 77.56,
      "step": 51815,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.220127672336232,
      "grad_norm": 6.368675708770752,
      "learning_rate": 5.874720388747632e-05,
      "loss": 2.541495132446289,
      "memory(GiB)": 77.56,
      "step": 51820,
      "token_acc": 0.4711111111111111,
      "train_speed(iter/s)": 1.438543
    },
    {
      "epoch": 2.2203418876654815,
      "grad_norm": 5.018083095550537,
      "learning_rate": 5.874057781927756e-05,
      "loss": 2.4667293548583986,
      "memory(GiB)": 77.56,
      "step": 51825,
      "token_acc": 0.440625,
      "train_speed(iter/s)": 1.438538
    },
    {
      "epoch": 2.2205561029947303,
      "grad_norm": 5.82100248336792,
      "learning_rate": 5.8733951592735045e-05,
      "loss": 2.636016273498535,
      "memory(GiB)": 77.56,
      "step": 51830,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.438536
    },
    {
      "epoch": 2.220770318323979,
      "grad_norm": 6.133962631225586,
      "learning_rate": 5.8727325207968806e-05,
      "loss": 2.5272632598876954,
      "memory(GiB)": 77.56,
      "step": 51835,
      "token_acc": 0.4349315068493151,
      "train_speed(iter/s)": 1.438523
    },
    {
      "epoch": 2.2209845336532283,
      "grad_norm": 5.133821964263916,
      "learning_rate": 5.872069866509887e-05,
      "loss": 2.3145822525024413,
      "memory(GiB)": 77.56,
      "step": 51840,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.438529
    },
    {
      "epoch": 2.221198748982477,
      "grad_norm": 10.58356761932373,
      "learning_rate": 5.871407196424532e-05,
      "loss": 2.1660438537597657,
      "memory(GiB)": 77.56,
      "step": 51845,
      "token_acc": 0.4854771784232365,
      "train_speed(iter/s)": 1.438528
    },
    {
      "epoch": 2.221412964311726,
      "grad_norm": 5.412009239196777,
      "learning_rate": 5.870744510552817e-05,
      "loss": 2.1739456176757814,
      "memory(GiB)": 77.56,
      "step": 51850,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 2.2216271796409752,
      "grad_norm": 6.137373447418213,
      "learning_rate": 5.8700818089067474e-05,
      "loss": 2.479261779785156,
      "memory(GiB)": 77.56,
      "step": 51855,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.438569
    },
    {
      "epoch": 2.221841394970224,
      "grad_norm": 6.139931678771973,
      "learning_rate": 5.8694190914983317e-05,
      "loss": 2.496595001220703,
      "memory(GiB)": 77.56,
      "step": 51860,
      "token_acc": 0.46215139442231074,
      "train_speed(iter/s)": 1.438573
    },
    {
      "epoch": 2.222055610299473,
      "grad_norm": 7.457913398742676,
      "learning_rate": 5.868756358339572e-05,
      "loss": 2.512097930908203,
      "memory(GiB)": 77.56,
      "step": 51865,
      "token_acc": 0.49693251533742333,
      "train_speed(iter/s)": 1.438601
    },
    {
      "epoch": 2.222269825628722,
      "grad_norm": 5.847672462463379,
      "learning_rate": 5.8680936094424754e-05,
      "loss": 2.334166717529297,
      "memory(GiB)": 77.56,
      "step": 51870,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.438625
    },
    {
      "epoch": 2.222484040957971,
      "grad_norm": 4.2275071144104,
      "learning_rate": 5.8674308448190506e-05,
      "loss": 2.639978790283203,
      "memory(GiB)": 77.56,
      "step": 51875,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.438606
    },
    {
      "epoch": 2.2226982562872197,
      "grad_norm": 5.724484920501709,
      "learning_rate": 5.8667680644813005e-05,
      "loss": 2.419038009643555,
      "memory(GiB)": 77.56,
      "step": 51880,
      "token_acc": 0.46075085324232085,
      "train_speed(iter/s)": 1.438609
    },
    {
      "epoch": 2.222912471616469,
      "grad_norm": 5.660974979400635,
      "learning_rate": 5.8661052684412354e-05,
      "loss": 2.2676227569580076,
      "memory(GiB)": 77.56,
      "step": 51885,
      "token_acc": 0.4578313253012048,
      "train_speed(iter/s)": 1.438634
    },
    {
      "epoch": 2.223126686945718,
      "grad_norm": 6.126853942871094,
      "learning_rate": 5.86544245671086e-05,
      "loss": 2.6737537384033203,
      "memory(GiB)": 77.56,
      "step": 51890,
      "token_acc": 0.4189723320158103,
      "train_speed(iter/s)": 1.438643
    },
    {
      "epoch": 2.2233409022749666,
      "grad_norm": 6.544710636138916,
      "learning_rate": 5.8647796293021826e-05,
      "loss": 2.8202558517456056,
      "memory(GiB)": 77.56,
      "step": 51895,
      "token_acc": 0.44609665427509293,
      "train_speed(iter/s)": 1.438644
    },
    {
      "epoch": 2.223555117604216,
      "grad_norm": 6.707851409912109,
      "learning_rate": 5.864116786227212e-05,
      "loss": 2.3110851287841796,
      "memory(GiB)": 77.56,
      "step": 51900,
      "token_acc": 0.49843260188087773,
      "train_speed(iter/s)": 1.438658
    },
    {
      "epoch": 2.2237693329334647,
      "grad_norm": 5.187964916229248,
      "learning_rate": 5.863453927497954e-05,
      "loss": 2.4777355194091797,
      "memory(GiB)": 77.56,
      "step": 51905,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.438642
    },
    {
      "epoch": 2.2239835482627135,
      "grad_norm": 4.107360363006592,
      "learning_rate": 5.8627910531264176e-05,
      "loss": 2.625115966796875,
      "memory(GiB)": 77.56,
      "step": 51910,
      "token_acc": 0.4619883040935672,
      "train_speed(iter/s)": 1.438635
    },
    {
      "epoch": 2.2241977635919628,
      "grad_norm": 5.4470720291137695,
      "learning_rate": 5.862128163124613e-05,
      "loss": 2.4041315078735352,
      "memory(GiB)": 77.56,
      "step": 51915,
      "token_acc": 0.5202702702702703,
      "train_speed(iter/s)": 1.438623
    },
    {
      "epoch": 2.2244119789212116,
      "grad_norm": 6.143822193145752,
      "learning_rate": 5.861465257504548e-05,
      "loss": 2.435796356201172,
      "memory(GiB)": 77.56,
      "step": 51920,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.438612
    },
    {
      "epoch": 2.2246261942504604,
      "grad_norm": 10.512497901916504,
      "learning_rate": 5.8608023362782316e-05,
      "loss": 2.5388055801391602,
      "memory(GiB)": 77.56,
      "step": 51925,
      "token_acc": 0.4536082474226804,
      "train_speed(iter/s)": 1.438643
    },
    {
      "epoch": 2.2248404095797096,
      "grad_norm": 7.161112308502197,
      "learning_rate": 5.8601393994576734e-05,
      "loss": 2.444790267944336,
      "memory(GiB)": 77.56,
      "step": 51930,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.438609
    },
    {
      "epoch": 2.2250546249089584,
      "grad_norm": 6.528739929199219,
      "learning_rate": 5.859476447054884e-05,
      "loss": 2.480727767944336,
      "memory(GiB)": 77.56,
      "step": 51935,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.438618
    },
    {
      "epoch": 2.2252688402382073,
      "grad_norm": 4.665379047393799,
      "learning_rate": 5.8588134790818707e-05,
      "loss": 2.0166717529296876,
      "memory(GiB)": 77.56,
      "step": 51940,
      "token_acc": 0.5393258426966292,
      "train_speed(iter/s)": 1.438603
    },
    {
      "epoch": 2.2254830555674565,
      "grad_norm": 7.0894694328308105,
      "learning_rate": 5.858150495550646e-05,
      "loss": 2.2285882949829103,
      "memory(GiB)": 77.56,
      "step": 51945,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.438625
    },
    {
      "epoch": 2.2256972708967053,
      "grad_norm": 5.141116619110107,
      "learning_rate": 5.857487496473221e-05,
      "loss": 2.2541152954101564,
      "memory(GiB)": 77.56,
      "step": 51950,
      "token_acc": 0.5432098765432098,
      "train_speed(iter/s)": 1.438622
    },
    {
      "epoch": 2.225911486225954,
      "grad_norm": 6.123048305511475,
      "learning_rate": 5.856824481861605e-05,
      "loss": 2.383865165710449,
      "memory(GiB)": 77.56,
      "step": 51955,
      "token_acc": 0.4627831715210356,
      "train_speed(iter/s)": 1.438603
    },
    {
      "epoch": 2.2261257015552034,
      "grad_norm": 7.0820441246032715,
      "learning_rate": 5.85616145172781e-05,
      "loss": 2.54730339050293,
      "memory(GiB)": 77.56,
      "step": 51960,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.438598
    },
    {
      "epoch": 2.226339916884452,
      "grad_norm": 6.379467010498047,
      "learning_rate": 5.855498406083847e-05,
      "loss": 2.458445739746094,
      "memory(GiB)": 77.56,
      "step": 51965,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.438636
    },
    {
      "epoch": 2.226554132213701,
      "grad_norm": 6.872741222381592,
      "learning_rate": 5.854835344941727e-05,
      "loss": 2.640496826171875,
      "memory(GiB)": 77.56,
      "step": 51970,
      "token_acc": 0.43416370106761565,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 2.2267683475429503,
      "grad_norm": 5.836672306060791,
      "learning_rate": 5.854172268313465e-05,
      "loss": 2.2867881774902346,
      "memory(GiB)": 77.56,
      "step": 51975,
      "token_acc": 0.5547445255474452,
      "train_speed(iter/s)": 1.438619
    },
    {
      "epoch": 2.226982562872199,
      "grad_norm": 5.967794895172119,
      "learning_rate": 5.8535091762110695e-05,
      "loss": 2.4145483016967773,
      "memory(GiB)": 77.56,
      "step": 51980,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.438599
    },
    {
      "epoch": 2.227196778201448,
      "grad_norm": 6.1902899742126465,
      "learning_rate": 5.852846068646554e-05,
      "loss": 2.2402278900146486,
      "memory(GiB)": 77.56,
      "step": 51985,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438626
    },
    {
      "epoch": 2.227410993530697,
      "grad_norm": 5.117874622344971,
      "learning_rate": 5.8521829456319334e-05,
      "loss": 2.5297452926635744,
      "memory(GiB)": 77.56,
      "step": 51990,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.43861
    },
    {
      "epoch": 2.227625208859946,
      "grad_norm": 6.199989318847656,
      "learning_rate": 5.851519807179219e-05,
      "loss": 2.2609928131103514,
      "memory(GiB)": 77.56,
      "step": 51995,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.438601
    },
    {
      "epoch": 2.227839424189195,
      "grad_norm": 6.211108207702637,
      "learning_rate": 5.850856653300424e-05,
      "loss": 2.674707794189453,
      "memory(GiB)": 77.56,
      "step": 52000,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 2.227839424189195,
      "eval_loss": 2.3591954708099365,
      "eval_runtime": 14.591,
      "eval_samples_per_second": 6.854,
      "eval_steps_per_second": 6.854,
      "eval_token_acc": 0.4573248407643312,
      "step": 52000
    },
    {
      "epoch": 2.228053639518444,
      "grad_norm": 4.987310886383057,
      "learning_rate": 5.850193484007563e-05,
      "loss": 2.3738496780395506,
      "memory(GiB)": 77.56,
      "step": 52005,
      "token_acc": 0.46702317290552586,
      "train_speed(iter/s)": 1.437957
    },
    {
      "epoch": 2.228267854847693,
      "grad_norm": 7.356802940368652,
      "learning_rate": 5.849530299312649e-05,
      "loss": 2.2208412170410154,
      "memory(GiB)": 77.56,
      "step": 52010,
      "token_acc": 0.524822695035461,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 2.2284820701769417,
      "grad_norm": 4.580575942993164,
      "learning_rate": 5.848867099227696e-05,
      "loss": 2.5155324935913086,
      "memory(GiB)": 77.56,
      "step": 52015,
      "token_acc": 0.4887459807073955,
      "train_speed(iter/s)": 1.438001
    },
    {
      "epoch": 2.228696285506191,
      "grad_norm": 5.334826946258545,
      "learning_rate": 5.848203883764721e-05,
      "loss": 2.5269001007080076,
      "memory(GiB)": 77.56,
      "step": 52020,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 2.2289105008354397,
      "grad_norm": 5.351432800292969,
      "learning_rate": 5.8475406529357356e-05,
      "loss": 2.78003044128418,
      "memory(GiB)": 77.56,
      "step": 52025,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 2.2291247161646885,
      "grad_norm": 5.526770114898682,
      "learning_rate": 5.8468774067527575e-05,
      "loss": 2.8215051651000977,
      "memory(GiB)": 77.56,
      "step": 52030,
      "token_acc": 0.3993610223642173,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 2.229338931493938,
      "grad_norm": 5.490999698638916,
      "learning_rate": 5.8462141452277995e-05,
      "loss": 2.418069839477539,
      "memory(GiB)": 77.56,
      "step": 52035,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 2.2295531468231866,
      "grad_norm": 7.099259853363037,
      "learning_rate": 5.845550868372879e-05,
      "loss": 2.4227771759033203,
      "memory(GiB)": 77.56,
      "step": 52040,
      "token_acc": 0.4632352941176471,
      "train_speed(iter/s)": 1.438081
    },
    {
      "epoch": 2.2297673621524354,
      "grad_norm": 6.0998735427856445,
      "learning_rate": 5.844887576200012e-05,
      "loss": 2.3925891876220704,
      "memory(GiB)": 77.56,
      "step": 52045,
      "token_acc": 0.5122699386503068,
      "train_speed(iter/s)": 1.438078
    },
    {
      "epoch": 2.2299815774816847,
      "grad_norm": 4.469804286956787,
      "learning_rate": 5.844224268721214e-05,
      "loss": 2.340729904174805,
      "memory(GiB)": 77.56,
      "step": 52050,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.438053
    },
    {
      "epoch": 2.2301957928109335,
      "grad_norm": 4.652035713195801,
      "learning_rate": 5.843560945948499e-05,
      "loss": 2.371796989440918,
      "memory(GiB)": 77.56,
      "step": 52055,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 2.2304100081401823,
      "grad_norm": 5.386318206787109,
      "learning_rate": 5.8428976078938877e-05,
      "loss": 2.4027633666992188,
      "memory(GiB)": 77.56,
      "step": 52060,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 2.2306242234694316,
      "grad_norm": 5.4634270668029785,
      "learning_rate": 5.842234254569396e-05,
      "loss": 2.408706855773926,
      "memory(GiB)": 77.56,
      "step": 52065,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.438047
    },
    {
      "epoch": 2.2308384387986804,
      "grad_norm": 4.541078090667725,
      "learning_rate": 5.84157088598704e-05,
      "loss": 2.631957244873047,
      "memory(GiB)": 77.56,
      "step": 52070,
      "token_acc": 0.444794952681388,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 2.231052654127929,
      "grad_norm": 6.064865589141846,
      "learning_rate": 5.840907502158839e-05,
      "loss": 2.4243370056152345,
      "memory(GiB)": 77.56,
      "step": 52075,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.438034
    },
    {
      "epoch": 2.2312668694571784,
      "grad_norm": 3.8478260040283203,
      "learning_rate": 5.84024410309681e-05,
      "loss": 2.5084304809570312,
      "memory(GiB)": 77.56,
      "step": 52080,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 2.2314810847864273,
      "grad_norm": 4.559310436248779,
      "learning_rate": 5.839580688812969e-05,
      "loss": 2.4697303771972656,
      "memory(GiB)": 77.56,
      "step": 52085,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 2.231695300115676,
      "grad_norm": 6.388387680053711,
      "learning_rate": 5.8389172593193365e-05,
      "loss": 2.638285827636719,
      "memory(GiB)": 77.56,
      "step": 52090,
      "token_acc": 0.46006389776357826,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 2.2319095154449253,
      "grad_norm": 4.712356090545654,
      "learning_rate": 5.838253814627932e-05,
      "loss": 2.4830198287963867,
      "memory(GiB)": 77.56,
      "step": 52095,
      "token_acc": 0.5244755244755245,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 2.232123730774174,
      "grad_norm": 6.152145862579346,
      "learning_rate": 5.8375903547507724e-05,
      "loss": 2.402608299255371,
      "memory(GiB)": 77.56,
      "step": 52100,
      "token_acc": 0.4785992217898833,
      "train_speed(iter/s)": 1.438046
    },
    {
      "epoch": 2.232337946103423,
      "grad_norm": 5.250881195068359,
      "learning_rate": 5.836926879699879e-05,
      "loss": 2.586806297302246,
      "memory(GiB)": 77.56,
      "step": 52105,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 2.232552161432672,
      "grad_norm": 7.42572021484375,
      "learning_rate": 5.83626338948727e-05,
      "loss": 2.5476520538330076,
      "memory(GiB)": 77.56,
      "step": 52110,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 2.232766376761921,
      "grad_norm": 4.829923152923584,
      "learning_rate": 5.835599884124964e-05,
      "loss": 2.460332489013672,
      "memory(GiB)": 77.56,
      "step": 52115,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.438046
    },
    {
      "epoch": 2.2329805920911703,
      "grad_norm": 5.859281063079834,
      "learning_rate": 5.8349363636249835e-05,
      "loss": 2.6722280502319338,
      "memory(GiB)": 77.56,
      "step": 52120,
      "token_acc": 0.4479166666666667,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 2.233194807420419,
      "grad_norm": 4.803500175476074,
      "learning_rate": 5.834272827999345e-05,
      "loss": 2.4590518951416014,
      "memory(GiB)": 77.56,
      "step": 52125,
      "token_acc": 0.47278911564625853,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 2.233409022749668,
      "grad_norm": 4.993882656097412,
      "learning_rate": 5.833609277260074e-05,
      "loss": 2.3638330459594727,
      "memory(GiB)": 77.56,
      "step": 52130,
      "token_acc": 0.504,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 2.233623238078917,
      "grad_norm": 6.235511302947998,
      "learning_rate": 5.8329457114191886e-05,
      "loss": 2.392010498046875,
      "memory(GiB)": 77.56,
      "step": 52135,
      "token_acc": 0.5047619047619047,
      "train_speed(iter/s)": 1.438097
    },
    {
      "epoch": 2.233837453408166,
      "grad_norm": 6.040309906005859,
      "learning_rate": 5.832282130488711e-05,
      "loss": 2.4152313232421876,
      "memory(GiB)": 77.56,
      "step": 52140,
      "token_acc": 0.5158730158730159,
      "train_speed(iter/s)": 1.438114
    },
    {
      "epoch": 2.234051668737415,
      "grad_norm": 5.303132057189941,
      "learning_rate": 5.8316185344806596e-05,
      "loss": 2.158448600769043,
      "memory(GiB)": 77.56,
      "step": 52145,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.438134
    },
    {
      "epoch": 2.234265884066664,
      "grad_norm": 5.439749240875244,
      "learning_rate": 5.8309549234070605e-05,
      "loss": 2.461848831176758,
      "memory(GiB)": 77.56,
      "step": 52150,
      "token_acc": 0.4911660777385159,
      "train_speed(iter/s)": 1.438136
    },
    {
      "epoch": 2.234480099395913,
      "grad_norm": 4.345194339752197,
      "learning_rate": 5.8302912972799315e-05,
      "loss": 2.5911281585693358,
      "memory(GiB)": 77.56,
      "step": 52155,
      "token_acc": 0.48324022346368717,
      "train_speed(iter/s)": 1.438113
    },
    {
      "epoch": 2.2346943147251617,
      "grad_norm": 5.080658912658691,
      "learning_rate": 5.8296276561112985e-05,
      "loss": 2.3790657043457033,
      "memory(GiB)": 77.56,
      "step": 52160,
      "token_acc": 0.49429657794676807,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 2.234908530054411,
      "grad_norm": 4.814274787902832,
      "learning_rate": 5.828963999913182e-05,
      "loss": 2.2174882888793945,
      "memory(GiB)": 77.56,
      "step": 52165,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.438142
    },
    {
      "epoch": 2.2351227453836597,
      "grad_norm": 4.737116813659668,
      "learning_rate": 5.8283003286976035e-05,
      "loss": 2.3507022857666016,
      "memory(GiB)": 77.56,
      "step": 52170,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.438137
    },
    {
      "epoch": 2.2353369607129085,
      "grad_norm": 5.720367908477783,
      "learning_rate": 5.827636642476589e-05,
      "loss": 2.209674072265625,
      "memory(GiB)": 77.56,
      "step": 52175,
      "token_acc": 0.5483870967741935,
      "train_speed(iter/s)": 1.438184
    },
    {
      "epoch": 2.235551176042158,
      "grad_norm": 5.919825077056885,
      "learning_rate": 5.826972941262161e-05,
      "loss": 2.558554840087891,
      "memory(GiB)": 77.56,
      "step": 52180,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.438194
    },
    {
      "epoch": 2.2357653913714066,
      "grad_norm": 5.511066913604736,
      "learning_rate": 5.826309225066341e-05,
      "loss": 2.4316606521606445,
      "memory(GiB)": 77.56,
      "step": 52185,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.438209
    },
    {
      "epoch": 2.2359796067006554,
      "grad_norm": 4.484046459197998,
      "learning_rate": 5.825645493901155e-05,
      "loss": 2.3739242553710938,
      "memory(GiB)": 77.56,
      "step": 52190,
      "token_acc": 0.4723926380368098,
      "train_speed(iter/s)": 1.438246
    },
    {
      "epoch": 2.2361938220299047,
      "grad_norm": 5.770759105682373,
      "learning_rate": 5.824981747778626e-05,
      "loss": 2.4203216552734377,
      "memory(GiB)": 77.56,
      "step": 52195,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438259
    },
    {
      "epoch": 2.2364080373591535,
      "grad_norm": 4.978071689605713,
      "learning_rate": 5.824317986710778e-05,
      "loss": 2.4877138137817383,
      "memory(GiB)": 77.56,
      "step": 52200,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.438261
    },
    {
      "epoch": 2.2366222526884023,
      "grad_norm": 5.100183486938477,
      "learning_rate": 5.823654210709637e-05,
      "loss": 2.5518028259277346,
      "memory(GiB)": 77.56,
      "step": 52205,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.438285
    },
    {
      "epoch": 2.2368364680176516,
      "grad_norm": 6.134519577026367,
      "learning_rate": 5.8229904197872284e-05,
      "loss": 2.7520109176635743,
      "memory(GiB)": 77.56,
      "step": 52210,
      "token_acc": 0.4272151898734177,
      "train_speed(iter/s)": 1.438244
    },
    {
      "epoch": 2.2370506833469004,
      "grad_norm": 5.845499038696289,
      "learning_rate": 5.822326613955574e-05,
      "loss": 2.6151599884033203,
      "memory(GiB)": 77.56,
      "step": 52215,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.438244
    },
    {
      "epoch": 2.237264898676149,
      "grad_norm": 5.837034702301025,
      "learning_rate": 5.821662793226704e-05,
      "loss": 2.2620176315307616,
      "memory(GiB)": 77.56,
      "step": 52220,
      "token_acc": 0.547244094488189,
      "train_speed(iter/s)": 1.438238
    },
    {
      "epoch": 2.2374791140053985,
      "grad_norm": 5.221806049346924,
      "learning_rate": 5.820998957612641e-05,
      "loss": 2.3668664932250976,
      "memory(GiB)": 77.56,
      "step": 52225,
      "token_acc": 0.47470817120622566,
      "train_speed(iter/s)": 1.438253
    },
    {
      "epoch": 2.2376933293346473,
      "grad_norm": 4.990957260131836,
      "learning_rate": 5.820335107125412e-05,
      "loss": 2.4245464324951174,
      "memory(GiB)": 77.56,
      "step": 52230,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.438241
    },
    {
      "epoch": 2.237907544663896,
      "grad_norm": 4.406238079071045,
      "learning_rate": 5.819671241777043e-05,
      "loss": 2.5115472793579103,
      "memory(GiB)": 77.56,
      "step": 52235,
      "token_acc": 0.44666666666666666,
      "train_speed(iter/s)": 1.438236
    },
    {
      "epoch": 2.2381217599931453,
      "grad_norm": 5.33854341506958,
      "learning_rate": 5.819007361579558e-05,
      "loss": 2.3464244842529296,
      "memory(GiB)": 77.56,
      "step": 52240,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.438265
    },
    {
      "epoch": 2.238335975322394,
      "grad_norm": 4.555159568786621,
      "learning_rate": 5.818343466544989e-05,
      "loss": 2.3547286987304688,
      "memory(GiB)": 77.56,
      "step": 52245,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.438254
    },
    {
      "epoch": 2.238550190651643,
      "grad_norm": 6.115626335144043,
      "learning_rate": 5.8176795566853606e-05,
      "loss": 2.206055450439453,
      "memory(GiB)": 77.56,
      "step": 52250,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.438235
    },
    {
      "epoch": 2.238764405980892,
      "grad_norm": 6.728278160095215,
      "learning_rate": 5.817015632012699e-05,
      "loss": 2.427275848388672,
      "memory(GiB)": 77.56,
      "step": 52255,
      "token_acc": 0.4641509433962264,
      "train_speed(iter/s)": 1.438241
    },
    {
      "epoch": 2.238978621310141,
      "grad_norm": 5.419593811035156,
      "learning_rate": 5.816351692539033e-05,
      "loss": 2.164683723449707,
      "memory(GiB)": 77.56,
      "step": 52260,
      "token_acc": 0.5634920634920635,
      "train_speed(iter/s)": 1.438281
    },
    {
      "epoch": 2.23919283663939,
      "grad_norm": 4.266914367675781,
      "learning_rate": 5.81568773827639e-05,
      "loss": 2.502466583251953,
      "memory(GiB)": 77.56,
      "step": 52265,
      "token_acc": 0.5087719298245614,
      "train_speed(iter/s)": 1.4383
    },
    {
      "epoch": 2.239407051968639,
      "grad_norm": 4.814389228820801,
      "learning_rate": 5.815023769236798e-05,
      "loss": 2.730708122253418,
      "memory(GiB)": 77.56,
      "step": 52270,
      "token_acc": 0.4780058651026393,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 2.239621267297888,
      "grad_norm": 5.215169906616211,
      "learning_rate": 5.814359785432286e-05,
      "loss": 2.356830596923828,
      "memory(GiB)": 77.56,
      "step": 52275,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.438281
    },
    {
      "epoch": 2.2398354826271367,
      "grad_norm": 6.906881809234619,
      "learning_rate": 5.8136957868748844e-05,
      "loss": 2.4186727523803713,
      "memory(GiB)": 77.56,
      "step": 52280,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.438296
    },
    {
      "epoch": 2.240049697956386,
      "grad_norm": 5.0663981437683105,
      "learning_rate": 5.813031773576618e-05,
      "loss": 2.3654603958129883,
      "memory(GiB)": 77.56,
      "step": 52285,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.438264
    },
    {
      "epoch": 2.240263913285635,
      "grad_norm": 6.552148818969727,
      "learning_rate": 5.81236774554952e-05,
      "loss": 2.625449371337891,
      "memory(GiB)": 77.56,
      "step": 52290,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.438269
    },
    {
      "epoch": 2.2404781286148836,
      "grad_norm": 4.879304885864258,
      "learning_rate": 5.811703702805618e-05,
      "loss": 2.2694137573242186,
      "memory(GiB)": 77.56,
      "step": 52295,
      "token_acc": 0.5191082802547771,
      "train_speed(iter/s)": 1.438279
    },
    {
      "epoch": 2.240692343944133,
      "grad_norm": 4.904812335968018,
      "learning_rate": 5.811039645356941e-05,
      "loss": 2.1864694595336913,
      "memory(GiB)": 77.56,
      "step": 52300,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.438285
    },
    {
      "epoch": 2.2409065592733817,
      "grad_norm": 6.095534324645996,
      "learning_rate": 5.810375573215521e-05,
      "loss": 2.602367401123047,
      "memory(GiB)": 77.56,
      "step": 52305,
      "token_acc": 0.4984520123839009,
      "train_speed(iter/s)": 1.438298
    },
    {
      "epoch": 2.2411207746026305,
      "grad_norm": 6.377011299133301,
      "learning_rate": 5.809711486393388e-05,
      "loss": 2.2531185150146484,
      "memory(GiB)": 77.56,
      "step": 52310,
      "token_acc": 0.5103734439834025,
      "train_speed(iter/s)": 1.438313
    },
    {
      "epoch": 2.2413349899318797,
      "grad_norm": 5.126463413238525,
      "learning_rate": 5.8090473849025685e-05,
      "loss": 2.141797637939453,
      "memory(GiB)": 77.56,
      "step": 52315,
      "token_acc": 0.5182186234817814,
      "train_speed(iter/s)": 1.438314
    },
    {
      "epoch": 2.2415492052611286,
      "grad_norm": 8.175921440124512,
      "learning_rate": 5.8083832687551e-05,
      "loss": 2.5346071243286135,
      "memory(GiB)": 77.56,
      "step": 52320,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.438315
    },
    {
      "epoch": 2.2417634205903774,
      "grad_norm": 9.176375389099121,
      "learning_rate": 5.807719137963009e-05,
      "loss": 2.480098915100098,
      "memory(GiB)": 77.56,
      "step": 52325,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.438336
    },
    {
      "epoch": 2.2419776359196266,
      "grad_norm": 5.458570957183838,
      "learning_rate": 5.807054992538328e-05,
      "loss": 2.5901315689086912,
      "memory(GiB)": 77.56,
      "step": 52330,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.43835
    },
    {
      "epoch": 2.2421918512488754,
      "grad_norm": 5.865347385406494,
      "learning_rate": 5.806390832493089e-05,
      "loss": 2.528288650512695,
      "memory(GiB)": 77.56,
      "step": 52335,
      "token_acc": 0.4847328244274809,
      "train_speed(iter/s)": 1.438351
    },
    {
      "epoch": 2.2424060665781242,
      "grad_norm": 5.235307693481445,
      "learning_rate": 5.805726657839324e-05,
      "loss": 2.5595371246337892,
      "memory(GiB)": 77.56,
      "step": 52340,
      "token_acc": 0.4293785310734463,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 2.2426202819073735,
      "grad_norm": 5.518535137176514,
      "learning_rate": 5.805062468589064e-05,
      "loss": 2.1073171615600588,
      "memory(GiB)": 77.56,
      "step": 52345,
      "token_acc": 0.5382059800664452,
      "train_speed(iter/s)": 1.43831
    },
    {
      "epoch": 2.2428344972366223,
      "grad_norm": 5.909546375274658,
      "learning_rate": 5.8043982647543426e-05,
      "loss": 2.144661712646484,
      "memory(GiB)": 77.56,
      "step": 52350,
      "token_acc": 0.5687732342007435,
      "train_speed(iter/s)": 1.438324
    },
    {
      "epoch": 2.243048712565871,
      "grad_norm": 5.499137878417969,
      "learning_rate": 5.803734046347192e-05,
      "loss": 2.4163820266723635,
      "memory(GiB)": 77.56,
      "step": 52355,
      "token_acc": 0.5178571428571429,
      "train_speed(iter/s)": 1.438327
    },
    {
      "epoch": 2.2432629278951204,
      "grad_norm": 5.9609055519104,
      "learning_rate": 5.8030698133796445e-05,
      "loss": 2.4384990692138673,
      "memory(GiB)": 77.56,
      "step": 52360,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.438336
    },
    {
      "epoch": 2.243477143224369,
      "grad_norm": 6.5359930992126465,
      "learning_rate": 5.802405565863735e-05,
      "loss": 2.4144208908081053,
      "memory(GiB)": 77.56,
      "step": 52365,
      "token_acc": 0.4977973568281938,
      "train_speed(iter/s)": 1.438328
    },
    {
      "epoch": 2.243691358553618,
      "grad_norm": 9.94322681427002,
      "learning_rate": 5.8017413038114965e-05,
      "loss": 2.3920829772949217,
      "memory(GiB)": 77.56,
      "step": 52370,
      "token_acc": 0.5267175572519084,
      "train_speed(iter/s)": 1.438345
    },
    {
      "epoch": 2.2439055738828673,
      "grad_norm": 4.635645866394043,
      "learning_rate": 5.8010770272349615e-05,
      "loss": 2.440418243408203,
      "memory(GiB)": 77.56,
      "step": 52375,
      "token_acc": 0.49221183800623053,
      "train_speed(iter/s)": 1.438394
    },
    {
      "epoch": 2.244119789212116,
      "grad_norm": 4.728506565093994,
      "learning_rate": 5.8004127361461644e-05,
      "loss": 2.575151824951172,
      "memory(GiB)": 77.56,
      "step": 52380,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.438367
    },
    {
      "epoch": 2.244334004541365,
      "grad_norm": 5.283487796783447,
      "learning_rate": 5.799748430557139e-05,
      "loss": 2.39417724609375,
      "memory(GiB)": 77.56,
      "step": 52385,
      "token_acc": 0.5313531353135313,
      "train_speed(iter/s)": 1.438366
    },
    {
      "epoch": 2.244548219870614,
      "grad_norm": 5.265110969543457,
      "learning_rate": 5.799084110479921e-05,
      "loss": 2.6639053344726564,
      "memory(GiB)": 77.56,
      "step": 52390,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.438358
    },
    {
      "epoch": 2.244762435199863,
      "grad_norm": 6.173926830291748,
      "learning_rate": 5.798419775926546e-05,
      "loss": 2.472491455078125,
      "memory(GiB)": 77.56,
      "step": 52395,
      "token_acc": 0.4582043343653251,
      "train_speed(iter/s)": 1.438358
    },
    {
      "epoch": 2.2449766505291118,
      "grad_norm": 4.308659076690674,
      "learning_rate": 5.7977554269090475e-05,
      "loss": 2.3466522216796877,
      "memory(GiB)": 77.56,
      "step": 52400,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.438393
    },
    {
      "epoch": 2.245190865858361,
      "grad_norm": 5.5615715980529785,
      "learning_rate": 5.7970910634394594e-05,
      "loss": 2.4914134979248046,
      "memory(GiB)": 77.56,
      "step": 52405,
      "token_acc": 0.4866666666666667,
      "train_speed(iter/s)": 1.438418
    },
    {
      "epoch": 2.24540508118761,
      "grad_norm": 6.581422805786133,
      "learning_rate": 5.796426685529821e-05,
      "loss": 2.561817932128906,
      "memory(GiB)": 77.56,
      "step": 52410,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.438405
    },
    {
      "epoch": 2.2456192965168587,
      "grad_norm": 4.762694835662842,
      "learning_rate": 5.795762293192164e-05,
      "loss": 2.320370674133301,
      "memory(GiB)": 77.56,
      "step": 52415,
      "token_acc": 0.49038461538461536,
      "train_speed(iter/s)": 1.438419
    },
    {
      "epoch": 2.245833511846108,
      "grad_norm": 5.242889404296875,
      "learning_rate": 5.7950978864385286e-05,
      "loss": 2.396922302246094,
      "memory(GiB)": 77.56,
      "step": 52420,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.438438
    },
    {
      "epoch": 2.2460477271753567,
      "grad_norm": 4.6974968910217285,
      "learning_rate": 5.7944334652809485e-05,
      "loss": 1.9638542175292968,
      "memory(GiB)": 77.56,
      "step": 52425,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438455
    },
    {
      "epoch": 2.2462619425046055,
      "grad_norm": 5.275105953216553,
      "learning_rate": 5.7937690297314594e-05,
      "loss": 2.6029062271118164,
      "memory(GiB)": 77.56,
      "step": 52430,
      "token_acc": 0.4507042253521127,
      "train_speed(iter/s)": 1.438447
    },
    {
      "epoch": 2.246476157833855,
      "grad_norm": 8.600444793701172,
      "learning_rate": 5.793104579802102e-05,
      "loss": 2.369175338745117,
      "memory(GiB)": 77.56,
      "step": 52435,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.43843
    },
    {
      "epoch": 2.2466903731631036,
      "grad_norm": 4.77775239944458,
      "learning_rate": 5.79244011550491e-05,
      "loss": 2.4584651947021485,
      "memory(GiB)": 77.56,
      "step": 52440,
      "token_acc": 0.46075085324232085,
      "train_speed(iter/s)": 1.438442
    },
    {
      "epoch": 2.2469045884923524,
      "grad_norm": 6.812375545501709,
      "learning_rate": 5.7917756368519217e-05,
      "loss": 2.665845489501953,
      "memory(GiB)": 77.56,
      "step": 52445,
      "token_acc": 0.45357142857142857,
      "train_speed(iter/s)": 1.438461
    },
    {
      "epoch": 2.2471188038216017,
      "grad_norm": 6.556591033935547,
      "learning_rate": 5.7911111438551754e-05,
      "loss": 2.3507244110107424,
      "memory(GiB)": 77.56,
      "step": 52450,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.438473
    },
    {
      "epoch": 2.2473330191508505,
      "grad_norm": 5.057736873626709,
      "learning_rate": 5.7904466365267097e-05,
      "loss": 2.29601936340332,
      "memory(GiB)": 77.56,
      "step": 52455,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.438488
    },
    {
      "epoch": 2.2475472344800993,
      "grad_norm": 6.139328479766846,
      "learning_rate": 5.789782114878559e-05,
      "loss": 2.318193244934082,
      "memory(GiB)": 77.56,
      "step": 52460,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.43849
    },
    {
      "epoch": 2.2477614498093486,
      "grad_norm": 5.3854289054870605,
      "learning_rate": 5.789117578922767e-05,
      "loss": 2.544055938720703,
      "memory(GiB)": 77.56,
      "step": 52465,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.438501
    },
    {
      "epoch": 2.2479756651385974,
      "grad_norm": 7.402945518493652,
      "learning_rate": 5.7884530286713687e-05,
      "loss": 2.385725212097168,
      "memory(GiB)": 77.56,
      "step": 52470,
      "token_acc": 0.45182724252491696,
      "train_speed(iter/s)": 1.438477
    },
    {
      "epoch": 2.248189880467846,
      "grad_norm": 5.561273097991943,
      "learning_rate": 5.787788464136403e-05,
      "loss": 2.3282379150390624,
      "memory(GiB)": 77.56,
      "step": 52475,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.438488
    },
    {
      "epoch": 2.2484040957970954,
      "grad_norm": 7.599419116973877,
      "learning_rate": 5.787123885329913e-05,
      "loss": 2.2904230117797852,
      "memory(GiB)": 77.56,
      "step": 52480,
      "token_acc": 0.4826254826254826,
      "train_speed(iter/s)": 1.43844
    },
    {
      "epoch": 2.2486183111263442,
      "grad_norm": 5.4341607093811035,
      "learning_rate": 5.786459292263934e-05,
      "loss": 2.273422431945801,
      "memory(GiB)": 77.56,
      "step": 52485,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.438443
    },
    {
      "epoch": 2.248832526455593,
      "grad_norm": 5.609556674957275,
      "learning_rate": 5.785794684950506e-05,
      "loss": 2.547252082824707,
      "memory(GiB)": 77.56,
      "step": 52490,
      "token_acc": 0.4797507788161994,
      "train_speed(iter/s)": 1.438445
    },
    {
      "epoch": 2.2490467417848423,
      "grad_norm": 4.623444557189941,
      "learning_rate": 5.7851300634016724e-05,
      "loss": 2.6107341766357424,
      "memory(GiB)": 77.56,
      "step": 52495,
      "token_acc": 0.438953488372093,
      "train_speed(iter/s)": 1.438411
    },
    {
      "epoch": 2.249260957114091,
      "grad_norm": 7.347949981689453,
      "learning_rate": 5.784465427629469e-05,
      "loss": 2.8286808013916014,
      "memory(GiB)": 77.56,
      "step": 52500,
      "token_acc": 0.43853820598006643,
      "train_speed(iter/s)": 1.438421
    },
    {
      "epoch": 2.249260957114091,
      "eval_loss": 2.180190324783325,
      "eval_runtime": 14.1531,
      "eval_samples_per_second": 7.066,
      "eval_steps_per_second": 7.066,
      "eval_token_acc": 0.4816326530612245,
      "step": 52500
    },
    {
      "epoch": 2.24947517244334,
      "grad_norm": 6.299923896789551,
      "learning_rate": 5.783800777645939e-05,
      "loss": 2.5754859924316404,
      "memory(GiB)": 77.56,
      "step": 52505,
      "token_acc": 0.47147147147147145,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 2.249689387772589,
      "grad_norm": 4.94052791595459,
      "learning_rate": 5.783136113463125e-05,
      "loss": 2.3202953338623047,
      "memory(GiB)": 77.56,
      "step": 52510,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 2.249903603101838,
      "grad_norm": 6.667887210845947,
      "learning_rate": 5.782471435093063e-05,
      "loss": 2.8009078979492186,
      "memory(GiB)": 77.56,
      "step": 52515,
      "token_acc": 0.4199395770392749,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 2.250117818431087,
      "grad_norm": 5.917612552642822,
      "learning_rate": 5.7818067425477976e-05,
      "loss": 2.840686798095703,
      "memory(GiB)": 77.56,
      "step": 52520,
      "token_acc": 0.43790849673202614,
      "train_speed(iter/s)": 1.437917
    },
    {
      "epoch": 2.250332033760336,
      "grad_norm": 6.140458106994629,
      "learning_rate": 5.781142035839371e-05,
      "loss": 2.2187252044677734,
      "memory(GiB)": 77.56,
      "step": 52525,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 2.250546249089585,
      "grad_norm": 4.763411045074463,
      "learning_rate": 5.7804773149798216e-05,
      "loss": 2.2728418350219726,
      "memory(GiB)": 77.56,
      "step": 52530,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.437943
    },
    {
      "epoch": 2.2507604644188337,
      "grad_norm": 5.218469619750977,
      "learning_rate": 5.7798125799811944e-05,
      "loss": 2.6449504852294923,
      "memory(GiB)": 77.56,
      "step": 52535,
      "token_acc": 0.45864661654135336,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 2.250974679748083,
      "grad_norm": 9.126242637634277,
      "learning_rate": 5.77914783085553e-05,
      "loss": 2.4818254470825196,
      "memory(GiB)": 77.56,
      "step": 52540,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 2.2511888950773318,
      "grad_norm": 5.561603546142578,
      "learning_rate": 5.778483067614874e-05,
      "loss": 2.635569953918457,
      "memory(GiB)": 77.56,
      "step": 52545,
      "token_acc": 0.4554140127388535,
      "train_speed(iter/s)": 1.437942
    },
    {
      "epoch": 2.2514031104065806,
      "grad_norm": 4.865551948547363,
      "learning_rate": 5.7778182902712644e-05,
      "loss": 2.3304779052734377,
      "memory(GiB)": 77.56,
      "step": 52550,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.437922
    },
    {
      "epoch": 2.25161732573583,
      "grad_norm": 5.232170104980469,
      "learning_rate": 5.777153498836748e-05,
      "loss": 2.522089385986328,
      "memory(GiB)": 77.56,
      "step": 52555,
      "token_acc": 0.46218487394957986,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 2.2518315410650787,
      "grad_norm": 13.78891372680664,
      "learning_rate": 5.776488693323366e-05,
      "loss": 2.43114013671875,
      "memory(GiB)": 77.56,
      "step": 52560,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 2.2520457563943275,
      "grad_norm": 8.460672378540039,
      "learning_rate": 5.775823873743165e-05,
      "loss": 2.300827217102051,
      "memory(GiB)": 77.56,
      "step": 52565,
      "token_acc": 0.5822222222222222,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 2.2522599717235767,
      "grad_norm": 6.5112690925598145,
      "learning_rate": 5.775159040108185e-05,
      "loss": 2.1495765686035155,
      "memory(GiB)": 77.56,
      "step": 52570,
      "token_acc": 0.5033112582781457,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 2.2524741870528255,
      "grad_norm": 4.030750274658203,
      "learning_rate": 5.7744941924304716e-05,
      "loss": 2.4456748962402344,
      "memory(GiB)": 77.56,
      "step": 52575,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 2.2526884023820744,
      "grad_norm": 5.141251087188721,
      "learning_rate": 5.77382933072207e-05,
      "loss": 2.512584686279297,
      "memory(GiB)": 77.56,
      "step": 52580,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 2.2529026177113236,
      "grad_norm": 4.528842449188232,
      "learning_rate": 5.773164454995026e-05,
      "loss": 2.2243797302246096,
      "memory(GiB)": 77.56,
      "step": 52585,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 2.2531168330405724,
      "grad_norm": 6.190828323364258,
      "learning_rate": 5.77249956526138e-05,
      "loss": 2.2826152801513673,
      "memory(GiB)": 77.56,
      "step": 52590,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.437955
    },
    {
      "epoch": 2.2533310483698212,
      "grad_norm": 4.875882625579834,
      "learning_rate": 5.7718346615331806e-05,
      "loss": 2.585957145690918,
      "memory(GiB)": 77.56,
      "step": 52595,
      "token_acc": 0.5249169435215947,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 2.2535452636990705,
      "grad_norm": 6.09178352355957,
      "learning_rate": 5.771169743822473e-05,
      "loss": 2.415875244140625,
      "memory(GiB)": 77.56,
      "step": 52600,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 2.2537594790283193,
      "grad_norm": 6.055955410003662,
      "learning_rate": 5.770504812141301e-05,
      "loss": 2.4428234100341797,
      "memory(GiB)": 77.56,
      "step": 52605,
      "token_acc": 0.51985559566787,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 2.253973694357568,
      "grad_norm": 4.312917232513428,
      "learning_rate": 5.7698398665017104e-05,
      "loss": 2.5624116897583007,
      "memory(GiB)": 77.56,
      "step": 52610,
      "token_acc": 0.49859943977591037,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 2.2541879096868174,
      "grad_norm": 4.953600883483887,
      "learning_rate": 5.7691749069157505e-05,
      "loss": 2.6711801528930663,
      "memory(GiB)": 77.56,
      "step": 52615,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.437944
    },
    {
      "epoch": 2.254402125016066,
      "grad_norm": 9.856009483337402,
      "learning_rate": 5.768509933395465e-05,
      "loss": 2.1410646438598633,
      "memory(GiB)": 77.56,
      "step": 52620,
      "token_acc": 0.5209125475285171,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 2.254616340345315,
      "grad_norm": 5.957549571990967,
      "learning_rate": 5.7678449459529015e-05,
      "loss": 2.4958564758300783,
      "memory(GiB)": 77.56,
      "step": 52625,
      "token_acc": 0.4254658385093168,
      "train_speed(iter/s)": 1.437948
    },
    {
      "epoch": 2.2548305556745643,
      "grad_norm": 6.903100490570068,
      "learning_rate": 5.7671799446001075e-05,
      "loss": 2.2911500930786133,
      "memory(GiB)": 77.56,
      "step": 52630,
      "token_acc": 0.49823321554770317,
      "train_speed(iter/s)": 1.437946
    },
    {
      "epoch": 2.255044771003813,
      "grad_norm": 5.000712871551514,
      "learning_rate": 5.766514929349126e-05,
      "loss": 2.5998674392700196,
      "memory(GiB)": 77.56,
      "step": 52635,
      "token_acc": 0.4555984555984556,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 2.255258986333062,
      "grad_norm": 4.973024368286133,
      "learning_rate": 5.7658499002120104e-05,
      "loss": 2.53555908203125,
      "memory(GiB)": 77.56,
      "step": 52640,
      "token_acc": 0.49107142857142855,
      "train_speed(iter/s)": 1.437917
    },
    {
      "epoch": 2.255473201662311,
      "grad_norm": 4.722836494445801,
      "learning_rate": 5.765184857200804e-05,
      "loss": 2.5097539901733397,
      "memory(GiB)": 77.56,
      "step": 52645,
      "token_acc": 0.5107033639143731,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 2.25568741699156,
      "grad_norm": 5.660754680633545,
      "learning_rate": 5.764519800327556e-05,
      "loss": 2.1958438873291017,
      "memory(GiB)": 77.56,
      "step": 52650,
      "token_acc": 0.5261437908496732,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 2.2559016323208088,
      "grad_norm": 5.0157623291015625,
      "learning_rate": 5.7638547296043154e-05,
      "loss": 2.4194915771484373,
      "memory(GiB)": 77.56,
      "step": 52655,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.437857
    },
    {
      "epoch": 2.256115847650058,
      "grad_norm": 5.594944477081299,
      "learning_rate": 5.76318964504313e-05,
      "loss": 2.4303096771240233,
      "memory(GiB)": 77.56,
      "step": 52660,
      "token_acc": 0.5049180327868853,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 2.256330062979307,
      "grad_norm": 6.9311418533325195,
      "learning_rate": 5.7625245466560474e-05,
      "loss": 2.4899091720581055,
      "memory(GiB)": 77.56,
      "step": 52665,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 2.2565442783085556,
      "grad_norm": 7.694872856140137,
      "learning_rate": 5.761859434455118e-05,
      "loss": 2.1439022064208983,
      "memory(GiB)": 77.56,
      "step": 52670,
      "token_acc": 0.5387755102040817,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 2.256758493637805,
      "grad_norm": 7.846240043640137,
      "learning_rate": 5.761194308452389e-05,
      "loss": 2.2495460510253906,
      "memory(GiB)": 77.56,
      "step": 52675,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 2.2569727089670537,
      "grad_norm": 4.7755255699157715,
      "learning_rate": 5.760529168659912e-05,
      "loss": 2.3682403564453125,
      "memory(GiB)": 77.56,
      "step": 52680,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 2.2571869242963025,
      "grad_norm": 4.929680347442627,
      "learning_rate": 5.759864015089735e-05,
      "loss": 2.440260887145996,
      "memory(GiB)": 77.56,
      "step": 52685,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 2.257401139625552,
      "grad_norm": 6.361323356628418,
      "learning_rate": 5.7591988477539104e-05,
      "loss": 2.2576662063598634,
      "memory(GiB)": 77.56,
      "step": 52690,
      "token_acc": 0.5150501672240803,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 2.2576153549548006,
      "grad_norm": 6.733607769012451,
      "learning_rate": 5.758533666664485e-05,
      "loss": 2.329274559020996,
      "memory(GiB)": 77.56,
      "step": 52695,
      "token_acc": 0.5078864353312302,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 2.2578295702840494,
      "grad_norm": 4.716760635375977,
      "learning_rate": 5.757868471833512e-05,
      "loss": 2.097896957397461,
      "memory(GiB)": 77.56,
      "step": 52700,
      "token_acc": 0.5415162454873647,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 2.2580437856132987,
      "grad_norm": 5.216626167297363,
      "learning_rate": 5.757203263273039e-05,
      "loss": 2.395073890686035,
      "memory(GiB)": 77.56,
      "step": 52705,
      "token_acc": 0.4462025316455696,
      "train_speed(iter/s)": 1.438092
    },
    {
      "epoch": 2.2582580009425475,
      "grad_norm": 9.615227699279785,
      "learning_rate": 5.756538040995119e-05,
      "loss": 2.5581438064575197,
      "memory(GiB)": 77.56,
      "step": 52710,
      "token_acc": 0.48348348348348347,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 2.2584722162717963,
      "grad_norm": 7.783900737762451,
      "learning_rate": 5.7558728050118036e-05,
      "loss": 2.490049362182617,
      "memory(GiB)": 77.56,
      "step": 52715,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 2.2586864316010455,
      "grad_norm": 4.691930770874023,
      "learning_rate": 5.755207555335142e-05,
      "loss": 2.5530912399291994,
      "memory(GiB)": 77.56,
      "step": 52720,
      "token_acc": 0.5070821529745042,
      "train_speed(iter/s)": 1.438071
    },
    {
      "epoch": 2.2589006469302944,
      "grad_norm": 5.784709453582764,
      "learning_rate": 5.7545422919771874e-05,
      "loss": 2.3213569641113283,
      "memory(GiB)": 77.56,
      "step": 52725,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 2.259114862259543,
      "grad_norm": 4.669852256774902,
      "learning_rate": 5.753877014949992e-05,
      "loss": 2.0753929138183596,
      "memory(GiB)": 77.56,
      "step": 52730,
      "token_acc": 0.5253164556962026,
      "train_speed(iter/s)": 1.4381
    },
    {
      "epoch": 2.2593290775887924,
      "grad_norm": 4.522634029388428,
      "learning_rate": 5.753211724265606e-05,
      "loss": 2.205304718017578,
      "memory(GiB)": 77.56,
      "step": 52735,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.43807
    },
    {
      "epoch": 2.2595432929180412,
      "grad_norm": 5.349833965301514,
      "learning_rate": 5.7525464199360844e-05,
      "loss": 2.4013298034667967,
      "memory(GiB)": 77.56,
      "step": 52740,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.438057
    },
    {
      "epoch": 2.25975750824729,
      "grad_norm": 5.460079669952393,
      "learning_rate": 5.751881101973479e-05,
      "loss": 2.4229297637939453,
      "memory(GiB)": 77.56,
      "step": 52745,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 2.2599717235765393,
      "grad_norm": 5.337129592895508,
      "learning_rate": 5.75121577038984e-05,
      "loss": 2.470130729675293,
      "memory(GiB)": 77.56,
      "step": 52750,
      "token_acc": 0.5037037037037037,
      "train_speed(iter/s)": 1.438047
    },
    {
      "epoch": 2.260185938905788,
      "grad_norm": 6.955590724945068,
      "learning_rate": 5.750550425197224e-05,
      "loss": 2.163749885559082,
      "memory(GiB)": 77.56,
      "step": 52755,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.438049
    },
    {
      "epoch": 2.260400154235037,
      "grad_norm": 4.7066216468811035,
      "learning_rate": 5.749885066407683e-05,
      "loss": 2.409752082824707,
      "memory(GiB)": 77.56,
      "step": 52760,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 2.260614369564286,
      "grad_norm": 5.495039463043213,
      "learning_rate": 5.74921969403327e-05,
      "loss": 2.5739295959472654,
      "memory(GiB)": 77.56,
      "step": 52765,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 2.260828584893535,
      "grad_norm": 5.087043762207031,
      "learning_rate": 5.74855430808604e-05,
      "loss": 2.327122688293457,
      "memory(GiB)": 77.56,
      "step": 52770,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 2.261042800222784,
      "grad_norm": 5.553941249847412,
      "learning_rate": 5.7478889085780476e-05,
      "loss": 2.530124473571777,
      "memory(GiB)": 77.56,
      "step": 52775,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 2.261257015552033,
      "grad_norm": 6.690070629119873,
      "learning_rate": 5.7472234955213435e-05,
      "loss": 2.708115577697754,
      "memory(GiB)": 77.56,
      "step": 52780,
      "token_acc": 0.4368932038834951,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 2.261471230881282,
      "grad_norm": 5.2283220291137695,
      "learning_rate": 5.7465580689279864e-05,
      "loss": 2.3802984237670897,
      "memory(GiB)": 77.56,
      "step": 52785,
      "token_acc": 0.47876447876447875,
      "train_speed(iter/s)": 1.438116
    },
    {
      "epoch": 2.2616854462105307,
      "grad_norm": 5.123905658721924,
      "learning_rate": 5.745892628810029e-05,
      "loss": 2.2786844253540037,
      "memory(GiB)": 77.56,
      "step": 52790,
      "token_acc": 0.5136986301369864,
      "train_speed(iter/s)": 1.438114
    },
    {
      "epoch": 2.26189966153978,
      "grad_norm": 6.019182205200195,
      "learning_rate": 5.745227175179526e-05,
      "loss": 2.4592945098876955,
      "memory(GiB)": 77.56,
      "step": 52795,
      "token_acc": 0.4624277456647399,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 2.2621138768690288,
      "grad_norm": 6.499453544616699,
      "learning_rate": 5.744561708048536e-05,
      "loss": 2.9474178314208985,
      "memory(GiB)": 77.56,
      "step": 52800,
      "token_acc": 0.4134275618374558,
      "train_speed(iter/s)": 1.438063
    },
    {
      "epoch": 2.2623280921982776,
      "grad_norm": 5.409016132354736,
      "learning_rate": 5.743896227429111e-05,
      "loss": 2.3522409439086913,
      "memory(GiB)": 77.56,
      "step": 52805,
      "token_acc": 0.5053475935828877,
      "train_speed(iter/s)": 1.438085
    },
    {
      "epoch": 2.262542307527527,
      "grad_norm": 4.505898475646973,
      "learning_rate": 5.743230733333307e-05,
      "loss": 2.6855234146118163,
      "memory(GiB)": 77.56,
      "step": 52810,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 2.2627565228567756,
      "grad_norm": 5.21828031539917,
      "learning_rate": 5.7425652257731834e-05,
      "loss": 2.4964494705200195,
      "memory(GiB)": 77.56,
      "step": 52815,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.438112
    },
    {
      "epoch": 2.2629707381860245,
      "grad_norm": 6.4684319496154785,
      "learning_rate": 5.741899704760791e-05,
      "loss": 2.5311044692993163,
      "memory(GiB)": 77.56,
      "step": 52820,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.43813
    },
    {
      "epoch": 2.2631849535152737,
      "grad_norm": 4.992574691772461,
      "learning_rate": 5.741234170308193e-05,
      "loss": 2.5661808013916017,
      "memory(GiB)": 77.56,
      "step": 52825,
      "token_acc": 0.48014440433212996,
      "train_speed(iter/s)": 1.438159
    },
    {
      "epoch": 2.2633991688445225,
      "grad_norm": 6.602748870849609,
      "learning_rate": 5.74056862242744e-05,
      "loss": 2.5633052825927733,
      "memory(GiB)": 77.56,
      "step": 52830,
      "token_acc": 0.5097402597402597,
      "train_speed(iter/s)": 1.438152
    },
    {
      "epoch": 2.2636133841737713,
      "grad_norm": 7.113536357879639,
      "learning_rate": 5.7399030611305913e-05,
      "loss": 2.1402908325195313,
      "memory(GiB)": 77.56,
      "step": 52835,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.438172
    },
    {
      "epoch": 2.2638275995030206,
      "grad_norm": 6.32389497756958,
      "learning_rate": 5.739237486429707e-05,
      "loss": 2.4931793212890625,
      "memory(GiB)": 77.56,
      "step": 52840,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.438167
    },
    {
      "epoch": 2.2640418148322694,
      "grad_norm": 5.483723163604736,
      "learning_rate": 5.738571898336842e-05,
      "loss": 2.538656997680664,
      "memory(GiB)": 77.56,
      "step": 52845,
      "token_acc": 0.4744744744744745,
      "train_speed(iter/s)": 1.438162
    },
    {
      "epoch": 2.264256030161518,
      "grad_norm": 5.115888595581055,
      "learning_rate": 5.737906296864053e-05,
      "loss": 2.2663623809814455,
      "memory(GiB)": 77.56,
      "step": 52850,
      "token_acc": 0.5634920634920635,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 2.2644702454907675,
      "grad_norm": 5.4134297370910645,
      "learning_rate": 5.737240682023399e-05,
      "loss": 2.4715688705444334,
      "memory(GiB)": 77.56,
      "step": 52855,
      "token_acc": 0.5103448275862069,
      "train_speed(iter/s)": 1.43816
    },
    {
      "epoch": 2.2646844608200163,
      "grad_norm": 4.665398120880127,
      "learning_rate": 5.73657505382694e-05,
      "loss": 2.578397750854492,
      "memory(GiB)": 77.56,
      "step": 52860,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.438154
    },
    {
      "epoch": 2.264898676149265,
      "grad_norm": 5.527158737182617,
      "learning_rate": 5.735909412286731e-05,
      "loss": 2.3311594009399412,
      "memory(GiB)": 77.56,
      "step": 52865,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.43814
    },
    {
      "epoch": 2.2651128914785144,
      "grad_norm": 4.6213698387146,
      "learning_rate": 5.735243757414833e-05,
      "loss": 2.382043647766113,
      "memory(GiB)": 77.56,
      "step": 52870,
      "token_acc": 0.4391891891891892,
      "train_speed(iter/s)": 1.438129
    },
    {
      "epoch": 2.265327106807763,
      "grad_norm": 4.389588832855225,
      "learning_rate": 5.734578089223306e-05,
      "loss": 2.744268608093262,
      "memory(GiB)": 77.56,
      "step": 52875,
      "token_acc": 0.4262734584450402,
      "train_speed(iter/s)": 1.438129
    },
    {
      "epoch": 2.265541322137012,
      "grad_norm": 5.95206356048584,
      "learning_rate": 5.7339124077242066e-05,
      "loss": 2.5268640518188477,
      "memory(GiB)": 77.56,
      "step": 52880,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.438124
    },
    {
      "epoch": 2.2657555374662612,
      "grad_norm": 5.782310962677002,
      "learning_rate": 5.7332467129295964e-05,
      "loss": 2.2619747161865233,
      "memory(GiB)": 77.56,
      "step": 52885,
      "token_acc": 0.5280898876404494,
      "train_speed(iter/s)": 1.438146
    },
    {
      "epoch": 2.26596975279551,
      "grad_norm": 5.011475086212158,
      "learning_rate": 5.732581004851534e-05,
      "loss": 2.434720993041992,
      "memory(GiB)": 77.56,
      "step": 52890,
      "token_acc": 0.45849802371541504,
      "train_speed(iter/s)": 1.438157
    },
    {
      "epoch": 2.266183968124759,
      "grad_norm": 4.993654251098633,
      "learning_rate": 5.731915283502079e-05,
      "loss": 2.340412139892578,
      "memory(GiB)": 77.56,
      "step": 52895,
      "token_acc": 0.47388059701492535,
      "train_speed(iter/s)": 1.438186
    },
    {
      "epoch": 2.266398183454008,
      "grad_norm": 4.088520050048828,
      "learning_rate": 5.731249548893291e-05,
      "loss": 2.396742057800293,
      "memory(GiB)": 77.56,
      "step": 52900,
      "token_acc": 0.5222929936305732,
      "train_speed(iter/s)": 1.438204
    },
    {
      "epoch": 2.266612398783257,
      "grad_norm": 4.804503917694092,
      "learning_rate": 5.730583801037234e-05,
      "loss": 2.3832279205322267,
      "memory(GiB)": 77.56,
      "step": 52905,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.438205
    },
    {
      "epoch": 2.2668266141125057,
      "grad_norm": 4.973869800567627,
      "learning_rate": 5.729918039945963e-05,
      "loss": 2.6502588272094725,
      "memory(GiB)": 77.56,
      "step": 52910,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.438232
    },
    {
      "epoch": 2.267040829441755,
      "grad_norm": 5.373922824859619,
      "learning_rate": 5.729252265631545e-05,
      "loss": 2.3283412933349608,
      "memory(GiB)": 77.56,
      "step": 52915,
      "token_acc": 0.5017064846416383,
      "train_speed(iter/s)": 1.43823
    },
    {
      "epoch": 2.267255044771004,
      "grad_norm": 5.5015997886657715,
      "learning_rate": 5.728586478106037e-05,
      "loss": 2.685133934020996,
      "memory(GiB)": 77.56,
      "step": 52920,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 2.2674692601002526,
      "grad_norm": 7.28200626373291,
      "learning_rate": 5.727920677381501e-05,
      "loss": 2.42260684967041,
      "memory(GiB)": 77.56,
      "step": 52925,
      "token_acc": 0.5150375939849624,
      "train_speed(iter/s)": 1.438278
    },
    {
      "epoch": 2.267683475429502,
      "grad_norm": 6.621476650238037,
      "learning_rate": 5.72725486347e-05,
      "loss": 2.368482208251953,
      "memory(GiB)": 77.56,
      "step": 52930,
      "token_acc": 0.48606811145510836,
      "train_speed(iter/s)": 1.438279
    },
    {
      "epoch": 2.2678976907587507,
      "grad_norm": 5.441981792449951,
      "learning_rate": 5.726589036383594e-05,
      "loss": 2.7897417068481447,
      "memory(GiB)": 77.56,
      "step": 52935,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.438277
    },
    {
      "epoch": 2.2681119060879995,
      "grad_norm": 6.121260643005371,
      "learning_rate": 5.725923196134345e-05,
      "loss": 2.4538955688476562,
      "memory(GiB)": 77.56,
      "step": 52940,
      "token_acc": 0.46757679180887374,
      "train_speed(iter/s)": 1.438291
    },
    {
      "epoch": 2.2683261214172488,
      "grad_norm": 5.303750991821289,
      "learning_rate": 5.725257342734318e-05,
      "loss": 2.3754655838012697,
      "memory(GiB)": 77.56,
      "step": 52945,
      "token_acc": 0.4725274725274725,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 2.2685403367464976,
      "grad_norm": 5.178130626678467,
      "learning_rate": 5.7245914761955744e-05,
      "loss": 2.293476867675781,
      "memory(GiB)": 77.56,
      "step": 52950,
      "token_acc": 0.5098684210526315,
      "train_speed(iter/s)": 1.43832
    },
    {
      "epoch": 2.2687545520757464,
      "grad_norm": 4.849772930145264,
      "learning_rate": 5.7239255965301755e-05,
      "loss": 2.196510124206543,
      "memory(GiB)": 77.56,
      "step": 52955,
      "token_acc": 0.5622317596566524,
      "train_speed(iter/s)": 1.438353
    },
    {
      "epoch": 2.2689687674049956,
      "grad_norm": 7.176445960998535,
      "learning_rate": 5.723259703750186e-05,
      "loss": 2.5891637802124023,
      "memory(GiB)": 77.56,
      "step": 52960,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.438364
    },
    {
      "epoch": 2.2691829827342445,
      "grad_norm": 5.645754814147949,
      "learning_rate": 5.722593797867667e-05,
      "loss": 2.2313116073608397,
      "memory(GiB)": 77.56,
      "step": 52965,
      "token_acc": 0.5724381625441696,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 2.2693971980634933,
      "grad_norm": 4.929851531982422,
      "learning_rate": 5.721927878894685e-05,
      "loss": 2.5327665328979494,
      "memory(GiB)": 77.56,
      "step": 52970,
      "token_acc": 0.4820846905537459,
      "train_speed(iter/s)": 1.438423
    },
    {
      "epoch": 2.2696114133927425,
      "grad_norm": 4.901386737823486,
      "learning_rate": 5.721261946843302e-05,
      "loss": 2.5152469635009767,
      "memory(GiB)": 77.56,
      "step": 52975,
      "token_acc": 0.45426829268292684,
      "train_speed(iter/s)": 1.438445
    },
    {
      "epoch": 2.2698256287219913,
      "grad_norm": 8.600111961364746,
      "learning_rate": 5.72059600172558e-05,
      "loss": 2.6597412109375,
      "memory(GiB)": 77.56,
      "step": 52980,
      "token_acc": 0.5074183976261127,
      "train_speed(iter/s)": 1.438463
    },
    {
      "epoch": 2.27003984405124,
      "grad_norm": 4.8522419929504395,
      "learning_rate": 5.7199300435535884e-05,
      "loss": 2.4320371627807615,
      "memory(GiB)": 77.56,
      "step": 52985,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.438494
    },
    {
      "epoch": 2.2702540593804894,
      "grad_norm": 5.510733127593994,
      "learning_rate": 5.7192640723393874e-05,
      "loss": 2.4039257049560545,
      "memory(GiB)": 77.56,
      "step": 52990,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.438528
    },
    {
      "epoch": 2.270468274709738,
      "grad_norm": 3.5769388675689697,
      "learning_rate": 5.7185980880950426e-05,
      "loss": 2.498982238769531,
      "memory(GiB)": 77.56,
      "step": 52995,
      "token_acc": 0.4581151832460733,
      "train_speed(iter/s)": 1.438508
    },
    {
      "epoch": 2.270682490038987,
      "grad_norm": 7.290960311889648,
      "learning_rate": 5.717932090832621e-05,
      "loss": 2.399822807312012,
      "memory(GiB)": 77.56,
      "step": 53000,
      "token_acc": 0.5063694267515924,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 2.270682490038987,
      "eval_loss": 2.3298373222351074,
      "eval_runtime": 14.6176,
      "eval_samples_per_second": 6.841,
      "eval_steps_per_second": 6.841,
      "eval_token_acc": 0.45588235294117646,
      "step": 53000
    },
    {
      "epoch": 2.2708967053682363,
      "grad_norm": 5.588569641113281,
      "learning_rate": 5.7172660805641855e-05,
      "loss": 2.29166316986084,
      "memory(GiB)": 77.56,
      "step": 53005,
      "token_acc": 0.47610294117647056,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 2.271110920697485,
      "grad_norm": 5.524135112762451,
      "learning_rate": 5.716600057301802e-05,
      "loss": 2.526597023010254,
      "memory(GiB)": 77.56,
      "step": 53010,
      "token_acc": 0.4460431654676259,
      "train_speed(iter/s)": 1.437924
    },
    {
      "epoch": 2.271325136026734,
      "grad_norm": 6.729628562927246,
      "learning_rate": 5.7159340210575355e-05,
      "loss": 2.395654296875,
      "memory(GiB)": 77.56,
      "step": 53015,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 2.271539351355983,
      "grad_norm": 7.5427422523498535,
      "learning_rate": 5.715267971843453e-05,
      "loss": 2.386401557922363,
      "memory(GiB)": 77.56,
      "step": 53020,
      "token_acc": 0.4511784511784512,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 2.271753566685232,
      "grad_norm": 5.130465984344482,
      "learning_rate": 5.71460190967162e-05,
      "loss": 2.367441177368164,
      "memory(GiB)": 77.56,
      "step": 53025,
      "token_acc": 0.5056179775280899,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 2.271967782014481,
      "grad_norm": 4.470834732055664,
      "learning_rate": 5.713935834554104e-05,
      "loss": 2.4068119049072267,
      "memory(GiB)": 77.56,
      "step": 53030,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 2.27218199734373,
      "grad_norm": 7.210778713226318,
      "learning_rate": 5.713269746502971e-05,
      "loss": 2.5451162338256834,
      "memory(GiB)": 77.56,
      "step": 53035,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 2.272396212672979,
      "grad_norm": 5.885176181793213,
      "learning_rate": 5.7126036455302855e-05,
      "loss": 2.252902030944824,
      "memory(GiB)": 77.56,
      "step": 53040,
      "token_acc": 0.5129032258064516,
      "train_speed(iter/s)": 1.43803
    },
    {
      "epoch": 2.2726104280022277,
      "grad_norm": 6.645816326141357,
      "learning_rate": 5.7119375316481175e-05,
      "loss": 2.2691341400146485,
      "memory(GiB)": 77.56,
      "step": 53045,
      "token_acc": 0.52734375,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 2.272824643331477,
      "grad_norm": 4.853824138641357,
      "learning_rate": 5.7112714048685324e-05,
      "loss": 2.2286312103271486,
      "memory(GiB)": 77.56,
      "step": 53050,
      "token_acc": 0.5241157556270096,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 2.2730388586607257,
      "grad_norm": 6.165210247039795,
      "learning_rate": 5.710605265203599e-05,
      "loss": 2.574686813354492,
      "memory(GiB)": 77.56,
      "step": 53055,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 2.2732530739899746,
      "grad_norm": 6.995044231414795,
      "learning_rate": 5.7099391126653855e-05,
      "loss": 2.667811965942383,
      "memory(GiB)": 77.56,
      "step": 53060,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 2.273467289319224,
      "grad_norm": 5.122948169708252,
      "learning_rate": 5.709272947265959e-05,
      "loss": 2.2270503997802735,
      "memory(GiB)": 77.56,
      "step": 53065,
      "token_acc": 0.5876288659793815,
      "train_speed(iter/s)": 1.438112
    },
    {
      "epoch": 2.2736815046484726,
      "grad_norm": 6.087159156799316,
      "learning_rate": 5.708606769017386e-05,
      "loss": 2.7900115966796877,
      "memory(GiB)": 77.56,
      "step": 53070,
      "token_acc": 0.4380664652567976,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 2.2738957199777214,
      "grad_norm": 5.669824600219727,
      "learning_rate": 5.707940577931739e-05,
      "loss": 2.3901580810546874,
      "memory(GiB)": 77.56,
      "step": 53075,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 2.2741099353069707,
      "grad_norm": 5.867022514343262,
      "learning_rate": 5.707274374021082e-05,
      "loss": 2.693343925476074,
      "memory(GiB)": 77.56,
      "step": 53080,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.438082
    },
    {
      "epoch": 2.2743241506362195,
      "grad_norm": 4.98362922668457,
      "learning_rate": 5.706608157297488e-05,
      "loss": 2.6642425537109373,
      "memory(GiB)": 77.56,
      "step": 53085,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.438117
    },
    {
      "epoch": 2.2745383659654683,
      "grad_norm": 5.155498027801514,
      "learning_rate": 5.705941927773023e-05,
      "loss": 2.536794662475586,
      "memory(GiB)": 77.56,
      "step": 53090,
      "token_acc": 0.51171875,
      "train_speed(iter/s)": 1.438148
    },
    {
      "epoch": 2.2747525812947176,
      "grad_norm": 6.072821140289307,
      "learning_rate": 5.70527568545976e-05,
      "loss": 2.4132221221923826,
      "memory(GiB)": 77.56,
      "step": 53095,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.438142
    },
    {
      "epoch": 2.2749667966239664,
      "grad_norm": 5.548638820648193,
      "learning_rate": 5.704609430369764e-05,
      "loss": 2.2902013778686525,
      "memory(GiB)": 77.56,
      "step": 53100,
      "token_acc": 0.5338645418326693,
      "train_speed(iter/s)": 1.438136
    },
    {
      "epoch": 2.275181011953215,
      "grad_norm": 4.359196186065674,
      "learning_rate": 5.703943162515109e-05,
      "loss": 2.485875129699707,
      "memory(GiB)": 77.56,
      "step": 53105,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.438167
    },
    {
      "epoch": 2.2753952272824645,
      "grad_norm": 5.7630157470703125,
      "learning_rate": 5.703276881907861e-05,
      "loss": 2.366191101074219,
      "memory(GiB)": 77.56,
      "step": 53110,
      "token_acc": 0.523972602739726,
      "train_speed(iter/s)": 1.438167
    },
    {
      "epoch": 2.2756094426117133,
      "grad_norm": 5.62446403503418,
      "learning_rate": 5.7026105885600946e-05,
      "loss": 2.3020889282226564,
      "memory(GiB)": 77.56,
      "step": 53115,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.438175
    },
    {
      "epoch": 2.275823657940962,
      "grad_norm": 6.563262462615967,
      "learning_rate": 5.701944282483877e-05,
      "loss": 2.3872055053710937,
      "memory(GiB)": 77.56,
      "step": 53120,
      "token_acc": 0.4923547400611621,
      "train_speed(iter/s)": 1.438187
    },
    {
      "epoch": 2.2760378732702113,
      "grad_norm": 5.30178165435791,
      "learning_rate": 5.7012779636912804e-05,
      "loss": 2.105153465270996,
      "memory(GiB)": 77.56,
      "step": 53125,
      "token_acc": 0.5285171102661597,
      "train_speed(iter/s)": 1.438217
    },
    {
      "epoch": 2.27625208859946,
      "grad_norm": 5.890365123748779,
      "learning_rate": 5.700611632194377e-05,
      "loss": 2.4600345611572267,
      "memory(GiB)": 77.56,
      "step": 53130,
      "token_acc": 0.45348837209302323,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 2.276466303928709,
      "grad_norm": 4.762717247009277,
      "learning_rate": 5.699945288005235e-05,
      "loss": 2.375034713745117,
      "memory(GiB)": 77.56,
      "step": 53135,
      "token_acc": 0.44336569579288027,
      "train_speed(iter/s)": 1.438251
    },
    {
      "epoch": 2.2766805192579582,
      "grad_norm": 6.028038024902344,
      "learning_rate": 5.699278931135929e-05,
      "loss": 2.4622344970703125,
      "memory(GiB)": 77.56,
      "step": 53140,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.438238
    },
    {
      "epoch": 2.276894734587207,
      "grad_norm": 8.230001449584961,
      "learning_rate": 5.6986125615985285e-05,
      "loss": 2.2831315994262695,
      "memory(GiB)": 77.56,
      "step": 53145,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.438251
    },
    {
      "epoch": 2.277108949916456,
      "grad_norm": 6.929459095001221,
      "learning_rate": 5.6979461794051056e-05,
      "loss": 2.4925300598144533,
      "memory(GiB)": 77.56,
      "step": 53150,
      "token_acc": 0.445993031358885,
      "train_speed(iter/s)": 1.438234
    },
    {
      "epoch": 2.277323165245705,
      "grad_norm": 5.7662224769592285,
      "learning_rate": 5.6972797845677325e-05,
      "loss": 2.0039365768432615,
      "memory(GiB)": 77.56,
      "step": 53155,
      "token_acc": 0.5354330708661418,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 2.277537380574954,
      "grad_norm": 4.668109893798828,
      "learning_rate": 5.696613377098484e-05,
      "loss": 2.0935928344726564,
      "memory(GiB)": 77.56,
      "step": 53160,
      "token_acc": 0.549645390070922,
      "train_speed(iter/s)": 1.438257
    },
    {
      "epoch": 2.2777515959042027,
      "grad_norm": 8.795716285705566,
      "learning_rate": 5.695946957009429e-05,
      "loss": 2.2457677841186525,
      "memory(GiB)": 77.56,
      "step": 53165,
      "token_acc": 0.4871060171919771,
      "train_speed(iter/s)": 1.438251
    },
    {
      "epoch": 2.277965811233452,
      "grad_norm": 6.286344528198242,
      "learning_rate": 5.6952805243126416e-05,
      "loss": 2.5730365753173827,
      "memory(GiB)": 77.56,
      "step": 53170,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.438264
    },
    {
      "epoch": 2.278180026562701,
      "grad_norm": 6.191279411315918,
      "learning_rate": 5.694614079020197e-05,
      "loss": 2.0139749526977537,
      "memory(GiB)": 77.56,
      "step": 53175,
      "token_acc": 0.5425531914893617,
      "train_speed(iter/s)": 1.438263
    },
    {
      "epoch": 2.2783942418919496,
      "grad_norm": 7.085164546966553,
      "learning_rate": 5.6939476211441664e-05,
      "loss": 2.4792854309082033,
      "memory(GiB)": 77.56,
      "step": 53180,
      "token_acc": 0.5158730158730159,
      "train_speed(iter/s)": 1.438251
    },
    {
      "epoch": 2.278608457221199,
      "grad_norm": 8.756878852844238,
      "learning_rate": 5.6932811506966236e-05,
      "loss": 2.576771545410156,
      "memory(GiB)": 77.56,
      "step": 53185,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.438256
    },
    {
      "epoch": 2.2788226725504477,
      "grad_norm": 4.465390682220459,
      "learning_rate": 5.692614667689642e-05,
      "loss": 2.4743396759033205,
      "memory(GiB)": 77.56,
      "step": 53190,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.438251
    },
    {
      "epoch": 2.2790368878796965,
      "grad_norm": 5.911646366119385,
      "learning_rate": 5.6919481721352954e-05,
      "loss": 2.3559106826782226,
      "memory(GiB)": 77.56,
      "step": 53195,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438262
    },
    {
      "epoch": 2.2792511032089458,
      "grad_norm": 4.425171375274658,
      "learning_rate": 5.691281664045659e-05,
      "loss": 2.4223958969116213,
      "memory(GiB)": 77.56,
      "step": 53200,
      "token_acc": 0.49523809523809526,
      "train_speed(iter/s)": 1.438277
    },
    {
      "epoch": 2.2794653185381946,
      "grad_norm": 5.114206790924072,
      "learning_rate": 5.690615143432807e-05,
      "loss": 2.3967639923095705,
      "memory(GiB)": 77.56,
      "step": 53205,
      "token_acc": 0.508695652173913,
      "train_speed(iter/s)": 1.438255
    },
    {
      "epoch": 2.2796795338674434,
      "grad_norm": 6.245949745178223,
      "learning_rate": 5.689948610308815e-05,
      "loss": 2.6908233642578123,
      "memory(GiB)": 77.56,
      "step": 53210,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.43829
    },
    {
      "epoch": 2.2798937491966926,
      "grad_norm": 5.675736904144287,
      "learning_rate": 5.689282064685755e-05,
      "loss": 2.213981819152832,
      "memory(GiB)": 77.56,
      "step": 53215,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.43829
    },
    {
      "epoch": 2.2801079645259414,
      "grad_norm": 6.717597961425781,
      "learning_rate": 5.688615506575704e-05,
      "loss": 2.7104873657226562,
      "memory(GiB)": 77.56,
      "step": 53220,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.438329
    },
    {
      "epoch": 2.2803221798551903,
      "grad_norm": 5.387569427490234,
      "learning_rate": 5.687948935990737e-05,
      "loss": 2.4388608932495117,
      "memory(GiB)": 77.56,
      "step": 53225,
      "token_acc": 0.49538461538461537,
      "train_speed(iter/s)": 1.438356
    },
    {
      "epoch": 2.2805363951844395,
      "grad_norm": 6.406753063201904,
      "learning_rate": 5.687282352942931e-05,
      "loss": 2.838492012023926,
      "memory(GiB)": 77.56,
      "step": 53230,
      "token_acc": 0.4524590163934426,
      "train_speed(iter/s)": 1.43837
    },
    {
      "epoch": 2.2807506105136883,
      "grad_norm": 5.576780319213867,
      "learning_rate": 5.686615757444359e-05,
      "loss": 2.5071510314941405,
      "memory(GiB)": 77.56,
      "step": 53235,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.438328
    },
    {
      "epoch": 2.280964825842937,
      "grad_norm": 6.09677267074585,
      "learning_rate": 5.685949149507099e-05,
      "loss": 2.8217147827148437,
      "memory(GiB)": 77.56,
      "step": 53240,
      "token_acc": 0.44715447154471544,
      "train_speed(iter/s)": 1.438306
    },
    {
      "epoch": 2.2811790411721864,
      "grad_norm": 6.976754188537598,
      "learning_rate": 5.6852825291432264e-05,
      "loss": 2.2277990341186524,
      "memory(GiB)": 77.56,
      "step": 53245,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.438325
    },
    {
      "epoch": 2.281393256501435,
      "grad_norm": 5.044142246246338,
      "learning_rate": 5.684615896364819e-05,
      "loss": 2.4598684310913086,
      "memory(GiB)": 77.56,
      "step": 53250,
      "token_acc": 0.5182724252491694,
      "train_speed(iter/s)": 1.438303
    },
    {
      "epoch": 2.281607471830684,
      "grad_norm": 6.840868949890137,
      "learning_rate": 5.683949251183951e-05,
      "loss": 2.270075225830078,
      "memory(GiB)": 77.56,
      "step": 53255,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.438319
    },
    {
      "epoch": 2.2818216871599333,
      "grad_norm": 6.621948719024658,
      "learning_rate": 5.683282593612702e-05,
      "loss": 2.5731340408325196,
      "memory(GiB)": 77.56,
      "step": 53260,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.438351
    },
    {
      "epoch": 2.282035902489182,
      "grad_norm": 5.110633373260498,
      "learning_rate": 5.682615923663147e-05,
      "loss": 2.6413650512695312,
      "memory(GiB)": 77.56,
      "step": 53265,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.438365
    },
    {
      "epoch": 2.282250117818431,
      "grad_norm": 5.838475227355957,
      "learning_rate": 5.681949241347364e-05,
      "loss": 2.316972351074219,
      "memory(GiB)": 77.56,
      "step": 53270,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.438351
    },
    {
      "epoch": 2.28246433314768,
      "grad_norm": 4.8606414794921875,
      "learning_rate": 5.681282546677431e-05,
      "loss": 2.6067874908447264,
      "memory(GiB)": 77.56,
      "step": 53275,
      "token_acc": 0.4520123839009288,
      "train_speed(iter/s)": 1.438325
    },
    {
      "epoch": 2.282678548476929,
      "grad_norm": 6.489525318145752,
      "learning_rate": 5.6806158396654264e-05,
      "loss": 2.2418067932128904,
      "memory(GiB)": 77.56,
      "step": 53280,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.438362
    },
    {
      "epoch": 2.282892763806178,
      "grad_norm": 5.963095664978027,
      "learning_rate": 5.679949120323426e-05,
      "loss": 2.815349006652832,
      "memory(GiB)": 77.56,
      "step": 53285,
      "token_acc": 0.44144144144144143,
      "train_speed(iter/s)": 1.438403
    },
    {
      "epoch": 2.283106979135427,
      "grad_norm": 4.6756205558776855,
      "learning_rate": 5.6792823886635104e-05,
      "loss": 2.2031063079833983,
      "memory(GiB)": 77.56,
      "step": 53290,
      "token_acc": 0.5220338983050847,
      "train_speed(iter/s)": 1.438429
    },
    {
      "epoch": 2.283321194464676,
      "grad_norm": 5.816607475280762,
      "learning_rate": 5.678615644697758e-05,
      "loss": 2.2711997985839845,
      "memory(GiB)": 77.56,
      "step": 53295,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.438452
    },
    {
      "epoch": 2.2835354097939247,
      "grad_norm": 6.056549072265625,
      "learning_rate": 5.6779488884382446e-05,
      "loss": 2.533121109008789,
      "memory(GiB)": 77.56,
      "step": 53300,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.438469
    },
    {
      "epoch": 2.283749625123174,
      "grad_norm": 5.472777366638184,
      "learning_rate": 5.677282119897053e-05,
      "loss": 2.2757379531860353,
      "memory(GiB)": 77.56,
      "step": 53305,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.438497
    },
    {
      "epoch": 2.2839638404524227,
      "grad_norm": 7.357724189758301,
      "learning_rate": 5.6766153390862585e-05,
      "loss": 2.488249588012695,
      "memory(GiB)": 77.56,
      "step": 53310,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.438483
    },
    {
      "epoch": 2.2841780557816715,
      "grad_norm": 5.0392045974731445,
      "learning_rate": 5.675948546017943e-05,
      "loss": 2.2134862899780274,
      "memory(GiB)": 77.56,
      "step": 53315,
      "token_acc": 0.5483870967741935,
      "train_speed(iter/s)": 1.438481
    },
    {
      "epoch": 2.284392271110921,
      "grad_norm": 4.15414571762085,
      "learning_rate": 5.6752817407041855e-05,
      "loss": 2.2173389434814452,
      "memory(GiB)": 77.56,
      "step": 53320,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.438507
    },
    {
      "epoch": 2.2846064864401696,
      "grad_norm": 5.964130401611328,
      "learning_rate": 5.674614923157068e-05,
      "loss": 2.273541259765625,
      "memory(GiB)": 77.56,
      "step": 53325,
      "token_acc": 0.5175097276264592,
      "train_speed(iter/s)": 1.438465
    },
    {
      "epoch": 2.2848207017694184,
      "grad_norm": 5.023801803588867,
      "learning_rate": 5.6739480933886655e-05,
      "loss": 2.452076721191406,
      "memory(GiB)": 77.56,
      "step": 53330,
      "token_acc": 0.4748427672955975,
      "train_speed(iter/s)": 1.438479
    },
    {
      "epoch": 2.2850349170986677,
      "grad_norm": 4.366186618804932,
      "learning_rate": 5.6732812514110624e-05,
      "loss": 2.5886425018310546,
      "memory(GiB)": 77.56,
      "step": 53335,
      "token_acc": 0.48135593220338985,
      "train_speed(iter/s)": 1.438489
    },
    {
      "epoch": 2.2852491324279165,
      "grad_norm": 4.92492151260376,
      "learning_rate": 5.672614397236337e-05,
      "loss": 2.34570426940918,
      "memory(GiB)": 77.56,
      "step": 53340,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.438499
    },
    {
      "epoch": 2.2854633477571653,
      "grad_norm": 5.4748125076293945,
      "learning_rate": 5.671947530876571e-05,
      "loss": 2.4210758209228516,
      "memory(GiB)": 77.56,
      "step": 53345,
      "token_acc": 0.477124183006536,
      "train_speed(iter/s)": 1.438506
    },
    {
      "epoch": 2.2856775630864146,
      "grad_norm": 7.903741359710693,
      "learning_rate": 5.671280652343846e-05,
      "loss": 2.6645380020141602,
      "memory(GiB)": 77.56,
      "step": 53350,
      "token_acc": 0.45703125,
      "train_speed(iter/s)": 1.438504
    },
    {
      "epoch": 2.2858917784156634,
      "grad_norm": 5.288139820098877,
      "learning_rate": 5.670613761650242e-05,
      "loss": 2.6764604568481447,
      "memory(GiB)": 77.56,
      "step": 53355,
      "token_acc": 0.46285714285714286,
      "train_speed(iter/s)": 1.438525
    },
    {
      "epoch": 2.286105993744912,
      "grad_norm": 5.179421901702881,
      "learning_rate": 5.6699468588078394e-05,
      "loss": 2.383032035827637,
      "memory(GiB)": 77.56,
      "step": 53360,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438487
    },
    {
      "epoch": 2.2863202090741614,
      "grad_norm": 5.836581230163574,
      "learning_rate": 5.669279943828723e-05,
      "loss": 2.3547563552856445,
      "memory(GiB)": 77.56,
      "step": 53365,
      "token_acc": 0.5016611295681063,
      "train_speed(iter/s)": 1.438511
    },
    {
      "epoch": 2.2865344244034103,
      "grad_norm": 4.6228227615356445,
      "learning_rate": 5.6686130167249706e-05,
      "loss": 2.406845474243164,
      "memory(GiB)": 77.56,
      "step": 53370,
      "token_acc": 0.44368600682593856,
      "train_speed(iter/s)": 1.438545
    },
    {
      "epoch": 2.286748639732659,
      "grad_norm": 5.463676929473877,
      "learning_rate": 5.6679460775086676e-05,
      "loss": 2.379409980773926,
      "memory(GiB)": 77.56,
      "step": 53375,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 2.2869628550619083,
      "grad_norm": 5.259365558624268,
      "learning_rate": 5.667279126191896e-05,
      "loss": 2.4757621765136717,
      "memory(GiB)": 77.56,
      "step": 53380,
      "token_acc": 0.47039473684210525,
      "train_speed(iter/s)": 1.438585
    },
    {
      "epoch": 2.287177070391157,
      "grad_norm": 4.919700622558594,
      "learning_rate": 5.666612162786734e-05,
      "loss": 2.513957214355469,
      "memory(GiB)": 77.56,
      "step": 53385,
      "token_acc": 0.49372384937238495,
      "train_speed(iter/s)": 1.438641
    },
    {
      "epoch": 2.287391285720406,
      "grad_norm": 5.996565341949463,
      "learning_rate": 5.665945187305268e-05,
      "loss": 2.4554615020751953,
      "memory(GiB)": 77.56,
      "step": 53390,
      "token_acc": 0.4847457627118644,
      "train_speed(iter/s)": 1.438634
    },
    {
      "epoch": 2.287605501049655,
      "grad_norm": 6.9908061027526855,
      "learning_rate": 5.6652781997595814e-05,
      "loss": 2.5153329849243162,
      "memory(GiB)": 77.56,
      "step": 53395,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438654
    },
    {
      "epoch": 2.287819716378904,
      "grad_norm": 4.970003604888916,
      "learning_rate": 5.6646112001617554e-05,
      "loss": 2.6753328323364256,
      "memory(GiB)": 77.56,
      "step": 53400,
      "token_acc": 0.4585635359116022,
      "train_speed(iter/s)": 1.438685
    },
    {
      "epoch": 2.288033931708153,
      "grad_norm": 6.306012153625488,
      "learning_rate": 5.663944188523875e-05,
      "loss": 2.233827590942383,
      "memory(GiB)": 77.56,
      "step": 53405,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.438693
    },
    {
      "epoch": 2.288248147037402,
      "grad_norm": 4.609481334686279,
      "learning_rate": 5.663277164858023e-05,
      "loss": 2.8375007629394533,
      "memory(GiB)": 77.56,
      "step": 53410,
      "token_acc": 0.4478114478114478,
      "train_speed(iter/s)": 1.438714
    },
    {
      "epoch": 2.288462362366651,
      "grad_norm": 5.747157096862793,
      "learning_rate": 5.662610129176281e-05,
      "loss": 2.1186235427856444,
      "memory(GiB)": 77.56,
      "step": 53415,
      "token_acc": 0.5433962264150943,
      "train_speed(iter/s)": 1.438695
    },
    {
      "epoch": 2.2886765776958997,
      "grad_norm": 6.112963676452637,
      "learning_rate": 5.661943081490736e-05,
      "loss": 2.583347511291504,
      "memory(GiB)": 77.56,
      "step": 53420,
      "token_acc": 0.47440273037542663,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 2.288890793025149,
      "grad_norm": 6.028528213500977,
      "learning_rate": 5.661276021813472e-05,
      "loss": 2.379299354553223,
      "memory(GiB)": 77.56,
      "step": 53425,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.438716
    },
    {
      "epoch": 2.289105008354398,
      "grad_norm": 6.210702896118164,
      "learning_rate": 5.6606089501565704e-05,
      "loss": 2.4029861450195313,
      "memory(GiB)": 77.56,
      "step": 53430,
      "token_acc": 0.5033112582781457,
      "train_speed(iter/s)": 1.438736
    },
    {
      "epoch": 2.2893192236836466,
      "grad_norm": 6.106517791748047,
      "learning_rate": 5.65994186653212e-05,
      "loss": 2.525979423522949,
      "memory(GiB)": 77.56,
      "step": 53435,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.438762
    },
    {
      "epoch": 2.289533439012896,
      "grad_norm": 7.56550407409668,
      "learning_rate": 5.659274770952204e-05,
      "loss": 2.4921316146850585,
      "memory(GiB)": 77.56,
      "step": 53440,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.438775
    },
    {
      "epoch": 2.2897476543421447,
      "grad_norm": 4.833075523376465,
      "learning_rate": 5.658607663428905e-05,
      "loss": 2.3022693634033202,
      "memory(GiB)": 77.56,
      "step": 53445,
      "token_acc": 0.551829268292683,
      "train_speed(iter/s)": 1.438787
    },
    {
      "epoch": 2.2899618696713935,
      "grad_norm": 5.692952632904053,
      "learning_rate": 5.6579405439743114e-05,
      "loss": 2.270130157470703,
      "memory(GiB)": 77.56,
      "step": 53450,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.438816
    },
    {
      "epoch": 2.2901760850006427,
      "grad_norm": 7.176535606384277,
      "learning_rate": 5.6572734126005076e-05,
      "loss": 2.3104724884033203,
      "memory(GiB)": 77.56,
      "step": 53455,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.438808
    },
    {
      "epoch": 2.2903903003298915,
      "grad_norm": 3.950342893600464,
      "learning_rate": 5.656606269319579e-05,
      "loss": 2.116921806335449,
      "memory(GiB)": 77.56,
      "step": 53460,
      "token_acc": 0.5228070175438596,
      "train_speed(iter/s)": 1.438794
    },
    {
      "epoch": 2.2906045156591404,
      "grad_norm": 5.201592445373535,
      "learning_rate": 5.6559391141436126e-05,
      "loss": 2.420242691040039,
      "memory(GiB)": 77.56,
      "step": 53465,
      "token_acc": 0.5065359477124183,
      "train_speed(iter/s)": 1.438822
    },
    {
      "epoch": 2.2908187309883896,
      "grad_norm": 5.9489030838012695,
      "learning_rate": 5.655271947084694e-05,
      "loss": 2.26196231842041,
      "memory(GiB)": 77.56,
      "step": 53470,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.438803
    },
    {
      "epoch": 2.2910329463176384,
      "grad_norm": 6.96991491317749,
      "learning_rate": 5.654604768154909e-05,
      "loss": 2.6189952850341798,
      "memory(GiB)": 77.56,
      "step": 53475,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.438813
    },
    {
      "epoch": 2.2912471616468872,
      "grad_norm": 4.714742183685303,
      "learning_rate": 5.6539375773663436e-05,
      "loss": 2.2030481338500976,
      "memory(GiB)": 77.56,
      "step": 53480,
      "token_acc": 0.5463917525773195,
      "train_speed(iter/s)": 1.43878
    },
    {
      "epoch": 2.2914613769761365,
      "grad_norm": 4.81325626373291,
      "learning_rate": 5.653270374731087e-05,
      "loss": 2.750800895690918,
      "memory(GiB)": 77.56,
      "step": 53485,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.438783
    },
    {
      "epoch": 2.2916755923053853,
      "grad_norm": 5.346724033355713,
      "learning_rate": 5.652603160261224e-05,
      "loss": 2.5518165588378907,
      "memory(GiB)": 77.56,
      "step": 53490,
      "token_acc": 0.4880952380952381,
      "train_speed(iter/s)": 1.438761
    },
    {
      "epoch": 2.291889807634634,
      "grad_norm": 5.134522914886475,
      "learning_rate": 5.651935933968843e-05,
      "loss": 2.2650596618652346,
      "memory(GiB)": 77.56,
      "step": 53495,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.438746
    },
    {
      "epoch": 2.2921040229638834,
      "grad_norm": 4.741995811462402,
      "learning_rate": 5.65126869586603e-05,
      "loss": 2.345151901245117,
      "memory(GiB)": 77.56,
      "step": 53500,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.438763
    },
    {
      "epoch": 2.2921040229638834,
      "eval_loss": 2.2262728214263916,
      "eval_runtime": 13.6477,
      "eval_samples_per_second": 7.327,
      "eval_steps_per_second": 7.327,
      "eval_token_acc": 0.48289473684210527,
      "step": 53500
    },
    {
      "epoch": 2.292318238293132,
      "grad_norm": 4.873902320861816,
      "learning_rate": 5.650601445964874e-05,
      "loss": 2.298366355895996,
      "memory(GiB)": 77.56,
      "step": 53505,
      "token_acc": 0.47879359095193214,
      "train_speed(iter/s)": 1.438183
    },
    {
      "epoch": 2.292532453622381,
      "grad_norm": 4.590123176574707,
      "learning_rate": 5.6499341842774637e-05,
      "loss": 2.370613670349121,
      "memory(GiB)": 77.56,
      "step": 53510,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.438192
    },
    {
      "epoch": 2.2927466689516303,
      "grad_norm": 5.4586615562438965,
      "learning_rate": 5.649266910815885e-05,
      "loss": 2.103523063659668,
      "memory(GiB)": 77.56,
      "step": 53515,
      "token_acc": 0.49224806201550386,
      "train_speed(iter/s)": 1.438202
    },
    {
      "epoch": 2.292960884280879,
      "grad_norm": 5.6876420974731445,
      "learning_rate": 5.648599625592227e-05,
      "loss": 2.237200927734375,
      "memory(GiB)": 77.56,
      "step": 53520,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438218
    },
    {
      "epoch": 2.293175099610128,
      "grad_norm": 4.383730411529541,
      "learning_rate": 5.647932328618579e-05,
      "loss": 2.3002960205078127,
      "memory(GiB)": 77.56,
      "step": 53525,
      "token_acc": 0.5084175084175084,
      "train_speed(iter/s)": 1.438247
    },
    {
      "epoch": 2.293389314939377,
      "grad_norm": 5.919870376586914,
      "learning_rate": 5.647265019907027e-05,
      "loss": 2.341591644287109,
      "memory(GiB)": 77.56,
      "step": 53530,
      "token_acc": 0.504885993485342,
      "train_speed(iter/s)": 1.438284
    },
    {
      "epoch": 2.293603530268626,
      "grad_norm": 6.525994300842285,
      "learning_rate": 5.646597699469665e-05,
      "loss": 2.644367790222168,
      "memory(GiB)": 77.56,
      "step": 53535,
      "token_acc": 0.4551971326164875,
      "train_speed(iter/s)": 1.438278
    },
    {
      "epoch": 2.2938177455978748,
      "grad_norm": 5.621082305908203,
      "learning_rate": 5.645930367318577e-05,
      "loss": 2.302459716796875,
      "memory(GiB)": 77.56,
      "step": 53540,
      "token_acc": 0.5413533834586466,
      "train_speed(iter/s)": 1.438281
    },
    {
      "epoch": 2.294031960927124,
      "grad_norm": 4.449276447296143,
      "learning_rate": 5.645263023465854e-05,
      "loss": 2.276961326599121,
      "memory(GiB)": 77.56,
      "step": 53545,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.438298
    },
    {
      "epoch": 2.294246176256373,
      "grad_norm": 4.564559459686279,
      "learning_rate": 5.644595667923589e-05,
      "loss": 2.425223541259766,
      "memory(GiB)": 77.56,
      "step": 53550,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.438305
    },
    {
      "epoch": 2.2944603915856217,
      "grad_norm": 5.750235080718994,
      "learning_rate": 5.643928300703867e-05,
      "loss": 2.1752178192138674,
      "memory(GiB)": 77.56,
      "step": 53555,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 2.294674606914871,
      "grad_norm": 3.7093615531921387,
      "learning_rate": 5.643260921818779e-05,
      "loss": 2.690829849243164,
      "memory(GiB)": 77.56,
      "step": 53560,
      "token_acc": 0.48220064724919093,
      "train_speed(iter/s)": 1.438337
    },
    {
      "epoch": 2.2948888222441197,
      "grad_norm": 4.8255934715271,
      "learning_rate": 5.642593531280418e-05,
      "loss": 2.6003559112548826,
      "memory(GiB)": 77.56,
      "step": 53565,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.438345
    },
    {
      "epoch": 2.2951030375733685,
      "grad_norm": 5.6724042892456055,
      "learning_rate": 5.6419261291008696e-05,
      "loss": 2.417842483520508,
      "memory(GiB)": 77.56,
      "step": 53570,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.438361
    },
    {
      "epoch": 2.295317252902618,
      "grad_norm": 5.824110507965088,
      "learning_rate": 5.641258715292228e-05,
      "loss": 2.454248046875,
      "memory(GiB)": 77.56,
      "step": 53575,
      "token_acc": 0.5050847457627119,
      "train_speed(iter/s)": 1.438337
    },
    {
      "epoch": 2.2955314682318666,
      "grad_norm": 4.26109504699707,
      "learning_rate": 5.640591289866585e-05,
      "loss": 1.9606668472290039,
      "memory(GiB)": 77.56,
      "step": 53580,
      "token_acc": 0.5708154506437768,
      "train_speed(iter/s)": 1.43837
    },
    {
      "epoch": 2.2957456835611154,
      "grad_norm": 7.147060394287109,
      "learning_rate": 5.6399238528360285e-05,
      "loss": 2.374354362487793,
      "memory(GiB)": 77.56,
      "step": 53585,
      "token_acc": 0.5131964809384164,
      "train_speed(iter/s)": 1.438362
    },
    {
      "epoch": 2.2959598988903647,
      "grad_norm": 5.928290843963623,
      "learning_rate": 5.6392564042126514e-05,
      "loss": 2.417148208618164,
      "memory(GiB)": 77.56,
      "step": 53590,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.438366
    },
    {
      "epoch": 2.2961741142196135,
      "grad_norm": 4.662358283996582,
      "learning_rate": 5.638588944008546e-05,
      "loss": 2.7058742523193358,
      "memory(GiB)": 77.56,
      "step": 53595,
      "token_acc": 0.48598130841121495,
      "train_speed(iter/s)": 1.438382
    },
    {
      "epoch": 2.2963883295488623,
      "grad_norm": 4.462774276733398,
      "learning_rate": 5.637921472235801e-05,
      "loss": 2.414849281311035,
      "memory(GiB)": 77.56,
      "step": 53600,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.438381
    },
    {
      "epoch": 2.2966025448781116,
      "grad_norm": 5.896153926849365,
      "learning_rate": 5.63725398890651e-05,
      "loss": 2.1580427169799803,
      "memory(GiB)": 77.56,
      "step": 53605,
      "token_acc": 0.5020080321285141,
      "train_speed(iter/s)": 1.438382
    },
    {
      "epoch": 2.2968167602073604,
      "grad_norm": 5.841938018798828,
      "learning_rate": 5.636586494032766e-05,
      "loss": 2.544931983947754,
      "memory(GiB)": 77.56,
      "step": 53610,
      "token_acc": 0.49079754601226994,
      "train_speed(iter/s)": 1.438367
    },
    {
      "epoch": 2.297030975536609,
      "grad_norm": 6.33357048034668,
      "learning_rate": 5.635918987626661e-05,
      "loss": 2.297098731994629,
      "memory(GiB)": 77.56,
      "step": 53615,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438385
    },
    {
      "epoch": 2.2972451908658584,
      "grad_norm": 5.119426727294922,
      "learning_rate": 5.6352514697002865e-05,
      "loss": 2.330405616760254,
      "memory(GiB)": 77.56,
      "step": 53620,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.43839
    },
    {
      "epoch": 2.2974594061951072,
      "grad_norm": 5.214448928833008,
      "learning_rate": 5.6345839402657364e-05,
      "loss": 2.562021827697754,
      "memory(GiB)": 77.56,
      "step": 53625,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.438415
    },
    {
      "epoch": 2.297673621524356,
      "grad_norm": 6.473047733306885,
      "learning_rate": 5.633916399335102e-05,
      "loss": 2.627433204650879,
      "memory(GiB)": 77.56,
      "step": 53630,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.438396
    },
    {
      "epoch": 2.2978878368536053,
      "grad_norm": 4.60115385055542,
      "learning_rate": 5.633248846920478e-05,
      "loss": 2.236454391479492,
      "memory(GiB)": 77.56,
      "step": 53635,
      "token_acc": 0.5346153846153846,
      "train_speed(iter/s)": 1.438404
    },
    {
      "epoch": 2.298102052182854,
      "grad_norm": 5.444350242614746,
      "learning_rate": 5.632581283033956e-05,
      "loss": 2.5599668502807615,
      "memory(GiB)": 77.56,
      "step": 53640,
      "token_acc": 0.46703296703296704,
      "train_speed(iter/s)": 1.438377
    },
    {
      "epoch": 2.298316267512103,
      "grad_norm": 5.982178688049316,
      "learning_rate": 5.631913707687632e-05,
      "loss": 2.057514953613281,
      "memory(GiB)": 77.56,
      "step": 53645,
      "token_acc": 0.5444444444444444,
      "train_speed(iter/s)": 1.438383
    },
    {
      "epoch": 2.298530482841352,
      "grad_norm": 5.037120819091797,
      "learning_rate": 5.6312461208935964e-05,
      "loss": 2.0735027313232424,
      "memory(GiB)": 77.56,
      "step": 53650,
      "token_acc": 0.5284552845528455,
      "train_speed(iter/s)": 1.438401
    },
    {
      "epoch": 2.298744698170601,
      "grad_norm": 4.352396011352539,
      "learning_rate": 5.630578522663947e-05,
      "loss": 2.585700035095215,
      "memory(GiB)": 77.56,
      "step": 53655,
      "token_acc": 0.46646341463414637,
      "train_speed(iter/s)": 1.43841
    },
    {
      "epoch": 2.29895891349985,
      "grad_norm": 4.839389801025391,
      "learning_rate": 5.629910913010775e-05,
      "loss": 2.2515895843505858,
      "memory(GiB)": 77.56,
      "step": 53660,
      "token_acc": 0.5014164305949008,
      "train_speed(iter/s)": 1.43845
    },
    {
      "epoch": 2.299173128829099,
      "grad_norm": 5.365807056427002,
      "learning_rate": 5.6292432919461777e-05,
      "loss": 2.233770751953125,
      "memory(GiB)": 77.56,
      "step": 53665,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.438456
    },
    {
      "epoch": 2.299387344158348,
      "grad_norm": 5.616290092468262,
      "learning_rate": 5.628575659482247e-05,
      "loss": 2.2652149200439453,
      "memory(GiB)": 77.56,
      "step": 53670,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.438461
    },
    {
      "epoch": 2.2996015594875967,
      "grad_norm": 5.762845039367676,
      "learning_rate": 5.627908015631078e-05,
      "loss": 2.7932287216186524,
      "memory(GiB)": 77.56,
      "step": 53675,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.438453
    },
    {
      "epoch": 2.299815774816846,
      "grad_norm": 4.533267974853516,
      "learning_rate": 5.6272403604047663e-05,
      "loss": 2.4930322647094725,
      "memory(GiB)": 77.56,
      "step": 53680,
      "token_acc": 0.49869451697127937,
      "train_speed(iter/s)": 1.438464
    },
    {
      "epoch": 2.3000299901460948,
      "grad_norm": 6.187613010406494,
      "learning_rate": 5.626572693815407e-05,
      "loss": 2.6859153747558593,
      "memory(GiB)": 77.56,
      "step": 53685,
      "token_acc": 0.45484949832775917,
      "train_speed(iter/s)": 1.438468
    },
    {
      "epoch": 2.3002442054753436,
      "grad_norm": 7.417507171630859,
      "learning_rate": 5.625905015875096e-05,
      "loss": 2.473193359375,
      "memory(GiB)": 77.56,
      "step": 53690,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.438487
    },
    {
      "epoch": 2.300458420804593,
      "grad_norm": 6.058342456817627,
      "learning_rate": 5.625237326595929e-05,
      "loss": 2.296754837036133,
      "memory(GiB)": 77.56,
      "step": 53695,
      "token_acc": 0.5264705882352941,
      "train_speed(iter/s)": 1.43852
    },
    {
      "epoch": 2.3006726361338417,
      "grad_norm": 6.117280006408691,
      "learning_rate": 5.624569625990002e-05,
      "loss": 2.3953039169311525,
      "memory(GiB)": 77.56,
      "step": 53700,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.438517
    },
    {
      "epoch": 2.3008868514630905,
      "grad_norm": 4.356379985809326,
      "learning_rate": 5.623901914069407e-05,
      "loss": 2.2769742965698243,
      "memory(GiB)": 77.56,
      "step": 53705,
      "token_acc": 0.5082508250825083,
      "train_speed(iter/s)": 1.438535
    },
    {
      "epoch": 2.3011010667923397,
      "grad_norm": 4.994988918304443,
      "learning_rate": 5.623234190846247e-05,
      "loss": 2.143342971801758,
      "memory(GiB)": 77.56,
      "step": 53710,
      "token_acc": 0.5338345864661654,
      "train_speed(iter/s)": 1.438528
    },
    {
      "epoch": 2.3013152821215885,
      "grad_norm": 5.715671539306641,
      "learning_rate": 5.6225664563326134e-05,
      "loss": 2.5589460372924804,
      "memory(GiB)": 77.56,
      "step": 53715,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 2.3015294974508373,
      "grad_norm": 6.478508472442627,
      "learning_rate": 5.621898710540604e-05,
      "loss": 2.2853019714355467,
      "memory(GiB)": 77.56,
      "step": 53720,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.438538
    },
    {
      "epoch": 2.3017437127800866,
      "grad_norm": 6.701432228088379,
      "learning_rate": 5.621230953482317e-05,
      "loss": 2.1565780639648438,
      "memory(GiB)": 77.56,
      "step": 53725,
      "token_acc": 0.4957983193277311,
      "train_speed(iter/s)": 1.43856
    },
    {
      "epoch": 2.3019579281093354,
      "grad_norm": 5.394850730895996,
      "learning_rate": 5.620563185169848e-05,
      "loss": 2.4948917388916017,
      "memory(GiB)": 77.56,
      "step": 53730,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.438552
    },
    {
      "epoch": 2.3021721434385842,
      "grad_norm": 5.482250213623047,
      "learning_rate": 5.6198954056152954e-05,
      "loss": 2.3999359130859377,
      "memory(GiB)": 77.56,
      "step": 53735,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.438563
    },
    {
      "epoch": 2.3023863587678335,
      "grad_norm": 5.089041709899902,
      "learning_rate": 5.6192276148307556e-05,
      "loss": 2.28002815246582,
      "memory(GiB)": 77.56,
      "step": 53740,
      "token_acc": 0.5425531914893617,
      "train_speed(iter/s)": 1.4386
    },
    {
      "epoch": 2.3026005740970823,
      "grad_norm": 4.6152873039245605,
      "learning_rate": 5.618559812828327e-05,
      "loss": 2.472471237182617,
      "memory(GiB)": 77.56,
      "step": 53745,
      "token_acc": 0.46757679180887374,
      "train_speed(iter/s)": 1.438587
    },
    {
      "epoch": 2.302814789426331,
      "grad_norm": 5.413963794708252,
      "learning_rate": 5.6178919996201064e-05,
      "loss": 2.601222610473633,
      "memory(GiB)": 77.56,
      "step": 53750,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.438612
    },
    {
      "epoch": 2.3030290047555804,
      "grad_norm": 4.750934600830078,
      "learning_rate": 5.617224175218193e-05,
      "loss": 2.4426929473876955,
      "memory(GiB)": 77.56,
      "step": 53755,
      "token_acc": 0.4635036496350365,
      "train_speed(iter/s)": 1.438614
    },
    {
      "epoch": 2.303243220084829,
      "grad_norm": 5.06614351272583,
      "learning_rate": 5.616556339634686e-05,
      "loss": 2.2192386627197265,
      "memory(GiB)": 77.56,
      "step": 53760,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 2.303457435414078,
      "grad_norm": 5.729245662689209,
      "learning_rate": 5.61588849288168e-05,
      "loss": 2.384457015991211,
      "memory(GiB)": 77.56,
      "step": 53765,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.438607
    },
    {
      "epoch": 2.3036716507433272,
      "grad_norm": 6.668264865875244,
      "learning_rate": 5.6152206349712776e-05,
      "loss": 2.093819236755371,
      "memory(GiB)": 77.56,
      "step": 53770,
      "token_acc": 0.5675675675675675,
      "train_speed(iter/s)": 1.438601
    },
    {
      "epoch": 2.303885866072576,
      "grad_norm": 6.468653678894043,
      "learning_rate": 5.614552765915575e-05,
      "loss": 2.3463817596435548,
      "memory(GiB)": 77.56,
      "step": 53775,
      "token_acc": 0.523972602739726,
      "train_speed(iter/s)": 1.438624
    },
    {
      "epoch": 2.304100081401825,
      "grad_norm": 4.567581653594971,
      "learning_rate": 5.613884885726675e-05,
      "loss": 2.3168701171875,
      "memory(GiB)": 77.56,
      "step": 53780,
      "token_acc": 0.5140845070422535,
      "train_speed(iter/s)": 1.438617
    },
    {
      "epoch": 2.304314296731074,
      "grad_norm": 4.7725510597229,
      "learning_rate": 5.6132169944166735e-05,
      "loss": 2.2093624114990233,
      "memory(GiB)": 77.56,
      "step": 53785,
      "token_acc": 0.5307443365695793,
      "train_speed(iter/s)": 1.43864
    },
    {
      "epoch": 2.304528512060323,
      "grad_norm": 5.204513072967529,
      "learning_rate": 5.6125490919976696e-05,
      "loss": 2.4235612869262697,
      "memory(GiB)": 77.56,
      "step": 53790,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.438676
    },
    {
      "epoch": 2.3047427273895718,
      "grad_norm": 6.216206073760986,
      "learning_rate": 5.611881178481765e-05,
      "loss": 2.1947036743164063,
      "memory(GiB)": 77.56,
      "step": 53795,
      "token_acc": 0.5491071428571429,
      "train_speed(iter/s)": 1.438685
    },
    {
      "epoch": 2.304956942718821,
      "grad_norm": 6.209000110626221,
      "learning_rate": 5.611213253881059e-05,
      "loss": 2.575878715515137,
      "memory(GiB)": 77.56,
      "step": 53800,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.438691
    },
    {
      "epoch": 2.30517115804807,
      "grad_norm": 4.774401664733887,
      "learning_rate": 5.610545318207652e-05,
      "loss": 2.487649154663086,
      "memory(GiB)": 77.56,
      "step": 53805,
      "token_acc": 0.4621212121212121,
      "train_speed(iter/s)": 1.438695
    },
    {
      "epoch": 2.3053853733773186,
      "grad_norm": 5.229221820831299,
      "learning_rate": 5.609877371473643e-05,
      "loss": 2.395768737792969,
      "memory(GiB)": 77.56,
      "step": 53810,
      "token_acc": 0.4942084942084942,
      "train_speed(iter/s)": 1.438689
    },
    {
      "epoch": 2.305599588706568,
      "grad_norm": 5.733492374420166,
      "learning_rate": 5.6092094136911344e-05,
      "loss": 2.225937080383301,
      "memory(GiB)": 77.56,
      "step": 53815,
      "token_acc": 0.48627450980392156,
      "train_speed(iter/s)": 1.438713
    },
    {
      "epoch": 2.3058138040358167,
      "grad_norm": 5.653471946716309,
      "learning_rate": 5.608541444872224e-05,
      "loss": 2.3452863693237305,
      "memory(GiB)": 77.56,
      "step": 53820,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.43871
    },
    {
      "epoch": 2.3060280193650655,
      "grad_norm": 7.246216297149658,
      "learning_rate": 5.607873465029017e-05,
      "loss": 2.464826202392578,
      "memory(GiB)": 77.56,
      "step": 53825,
      "token_acc": 0.5276752767527675,
      "train_speed(iter/s)": 1.438745
    },
    {
      "epoch": 2.3062422346943148,
      "grad_norm": 4.9488372802734375,
      "learning_rate": 5.607205474173609e-05,
      "loss": 2.3327274322509766,
      "memory(GiB)": 77.56,
      "step": 53830,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.438777
    },
    {
      "epoch": 2.3064564500235636,
      "grad_norm": 5.391012191772461,
      "learning_rate": 5.606537472318105e-05,
      "loss": 2.6019224166870116,
      "memory(GiB)": 77.56,
      "step": 53835,
      "token_acc": 0.4548611111111111,
      "train_speed(iter/s)": 1.438783
    },
    {
      "epoch": 2.3066706653528124,
      "grad_norm": 4.587630271911621,
      "learning_rate": 5.605869459474608e-05,
      "loss": 2.727021408081055,
      "memory(GiB)": 77.56,
      "step": 53840,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438802
    },
    {
      "epoch": 2.3068848806820617,
      "grad_norm": 5.3732709884643555,
      "learning_rate": 5.6052014356552166e-05,
      "loss": 2.4661779403686523,
      "memory(GiB)": 77.56,
      "step": 53845,
      "token_acc": 0.45528455284552843,
      "train_speed(iter/s)": 1.43882
    },
    {
      "epoch": 2.3070990960113105,
      "grad_norm": 5.235447406768799,
      "learning_rate": 5.6045334008720316e-05,
      "loss": 2.272469902038574,
      "memory(GiB)": 77.56,
      "step": 53850,
      "token_acc": 0.5206611570247934,
      "train_speed(iter/s)": 1.438855
    },
    {
      "epoch": 2.3073133113405593,
      "grad_norm": 8.965286254882812,
      "learning_rate": 5.603865355137159e-05,
      "loss": 2.526051330566406,
      "memory(GiB)": 77.56,
      "step": 53855,
      "token_acc": 0.46567164179104475,
      "train_speed(iter/s)": 1.438882
    },
    {
      "epoch": 2.3075275266698085,
      "grad_norm": 5.878705024719238,
      "learning_rate": 5.603197298462697e-05,
      "loss": 2.586753082275391,
      "memory(GiB)": 77.56,
      "step": 53860,
      "token_acc": 0.4689265536723164,
      "train_speed(iter/s)": 1.43888
    },
    {
      "epoch": 2.3077417419990573,
      "grad_norm": 8.354619026184082,
      "learning_rate": 5.60252923086075e-05,
      "loss": 2.4950727462768554,
      "memory(GiB)": 77.56,
      "step": 53865,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.438901
    },
    {
      "epoch": 2.3079559573283066,
      "grad_norm": 5.443455219268799,
      "learning_rate": 5.601861152343423e-05,
      "loss": 2.5206817626953124,
      "memory(GiB)": 77.56,
      "step": 53870,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.438952
    },
    {
      "epoch": 2.3081701726575554,
      "grad_norm": 11.195306777954102,
      "learning_rate": 5.601193062922816e-05,
      "loss": 2.527469253540039,
      "memory(GiB)": 77.56,
      "step": 53875,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.438967
    },
    {
      "epoch": 2.3083843879868042,
      "grad_norm": 5.233806133270264,
      "learning_rate": 5.600524962611032e-05,
      "loss": 2.5457073211669923,
      "memory(GiB)": 77.56,
      "step": 53880,
      "token_acc": 0.46254071661237783,
      "train_speed(iter/s)": 1.438978
    },
    {
      "epoch": 2.3085986033160535,
      "grad_norm": 5.601393699645996,
      "learning_rate": 5.5998568514201754e-05,
      "loss": 2.3417791366577148,
      "memory(GiB)": 77.56,
      "step": 53885,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.439
    },
    {
      "epoch": 2.3088128186453023,
      "grad_norm": 5.6394805908203125,
      "learning_rate": 5.599188729362349e-05,
      "loss": 2.4958471298217773,
      "memory(GiB)": 77.56,
      "step": 53890,
      "token_acc": 0.478134110787172,
      "train_speed(iter/s)": 1.439003
    },
    {
      "epoch": 2.309027033974551,
      "grad_norm": 9.374964714050293,
      "learning_rate": 5.598520596449657e-05,
      "loss": 2.617146682739258,
      "memory(GiB)": 77.56,
      "step": 53895,
      "token_acc": 0.4264705882352941,
      "train_speed(iter/s)": 1.439027
    },
    {
      "epoch": 2.3092412493038004,
      "grad_norm": 5.910741806030273,
      "learning_rate": 5.597852452694202e-05,
      "loss": 2.5937618255615233,
      "memory(GiB)": 77.56,
      "step": 53900,
      "token_acc": 0.4907749077490775,
      "train_speed(iter/s)": 1.439026
    },
    {
      "epoch": 2.309455464633049,
      "grad_norm": 5.780880928039551,
      "learning_rate": 5.5971842981080905e-05,
      "loss": 2.5331489562988283,
      "memory(GiB)": 77.56,
      "step": 53905,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.439029
    },
    {
      "epoch": 2.309669679962298,
      "grad_norm": 7.117316722869873,
      "learning_rate": 5.5965161327034234e-05,
      "loss": 2.3115524291992187,
      "memory(GiB)": 77.56,
      "step": 53910,
      "token_acc": 0.5014925373134328,
      "train_speed(iter/s)": 1.439065
    },
    {
      "epoch": 2.3098838952915473,
      "grad_norm": 7.297818660736084,
      "learning_rate": 5.595847956492308e-05,
      "loss": 2.189962387084961,
      "memory(GiB)": 77.56,
      "step": 53915,
      "token_acc": 0.511864406779661,
      "train_speed(iter/s)": 1.439028
    },
    {
      "epoch": 2.310098110620796,
      "grad_norm": 5.504382610321045,
      "learning_rate": 5.595179769486848e-05,
      "loss": 2.1856462478637697,
      "memory(GiB)": 77.56,
      "step": 53920,
      "token_acc": 0.5748987854251012,
      "train_speed(iter/s)": 1.439026
    },
    {
      "epoch": 2.310312325950045,
      "grad_norm": 5.445213794708252,
      "learning_rate": 5.5945115716991484e-05,
      "loss": 2.5255428314208985,
      "memory(GiB)": 77.56,
      "step": 53925,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.439028
    },
    {
      "epoch": 2.310526541279294,
      "grad_norm": 7.495448112487793,
      "learning_rate": 5.5938433631413145e-05,
      "loss": 2.3789600372314452,
      "memory(GiB)": 77.56,
      "step": 53930,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.439017
    },
    {
      "epoch": 2.310740756608543,
      "grad_norm": 6.696556091308594,
      "learning_rate": 5.5931751438254486e-05,
      "loss": 2.7472387313842774,
      "memory(GiB)": 77.56,
      "step": 53935,
      "token_acc": 0.4398496240601504,
      "train_speed(iter/s)": 1.439029
    },
    {
      "epoch": 2.3109549719377918,
      "grad_norm": 6.056976318359375,
      "learning_rate": 5.59250691376366e-05,
      "loss": 2.533992958068848,
      "memory(GiB)": 77.56,
      "step": 53940,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.438985
    },
    {
      "epoch": 2.311169187267041,
      "grad_norm": 6.147352695465088,
      "learning_rate": 5.5918386729680535e-05,
      "loss": 2.479195785522461,
      "memory(GiB)": 77.56,
      "step": 53945,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.439008
    },
    {
      "epoch": 2.31138340259629,
      "grad_norm": 5.396799087524414,
      "learning_rate": 5.591170421450733e-05,
      "loss": 2.662628173828125,
      "memory(GiB)": 77.56,
      "step": 53950,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.439011
    },
    {
      "epoch": 2.3115976179255386,
      "grad_norm": 7.313360214233398,
      "learning_rate": 5.590502159223807e-05,
      "loss": 2.657781982421875,
      "memory(GiB)": 77.56,
      "step": 53955,
      "token_acc": 0.4409937888198758,
      "train_speed(iter/s)": 1.439032
    },
    {
      "epoch": 2.311811833254788,
      "grad_norm": 6.940757751464844,
      "learning_rate": 5.58983388629938e-05,
      "loss": 2.4785959243774416,
      "memory(GiB)": 77.56,
      "step": 53960,
      "token_acc": 0.5020408163265306,
      "train_speed(iter/s)": 1.439033
    },
    {
      "epoch": 2.3120260485840367,
      "grad_norm": 4.621875762939453,
      "learning_rate": 5.589165602689559e-05,
      "loss": 2.2554958343505858,
      "memory(GiB)": 77.56,
      "step": 53965,
      "token_acc": 0.5330882352941176,
      "train_speed(iter/s)": 1.439049
    },
    {
      "epoch": 2.3122402639132855,
      "grad_norm": 5.698229789733887,
      "learning_rate": 5.588497308406451e-05,
      "loss": 2.238688278198242,
      "memory(GiB)": 77.56,
      "step": 53970,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.438991
    },
    {
      "epoch": 2.3124544792425348,
      "grad_norm": 5.020229816436768,
      "learning_rate": 5.5878290034621593e-05,
      "loss": 2.2700672149658203,
      "memory(GiB)": 77.56,
      "step": 53975,
      "token_acc": 0.4695121951219512,
      "train_speed(iter/s)": 1.438962
    },
    {
      "epoch": 2.3126686945717836,
      "grad_norm": 6.037504196166992,
      "learning_rate": 5.5871606878687956e-05,
      "loss": 2.140109634399414,
      "memory(GiB)": 77.56,
      "step": 53980,
      "token_acc": 0.5463917525773195,
      "train_speed(iter/s)": 1.438961
    },
    {
      "epoch": 2.3128829099010324,
      "grad_norm": 5.255722522735596,
      "learning_rate": 5.586492361638466e-05,
      "loss": 2.4218265533447267,
      "memory(GiB)": 77.56,
      "step": 53985,
      "token_acc": 0.453125,
      "train_speed(iter/s)": 1.438954
    },
    {
      "epoch": 2.3130971252302817,
      "grad_norm": 7.433424949645996,
      "learning_rate": 5.585824024783277e-05,
      "loss": 2.3656471252441404,
      "memory(GiB)": 77.56,
      "step": 53990,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.438958
    },
    {
      "epoch": 2.3133113405595305,
      "grad_norm": 5.4691972732543945,
      "learning_rate": 5.585155677315336e-05,
      "loss": 2.3481334686279296,
      "memory(GiB)": 77.56,
      "step": 53995,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.438952
    },
    {
      "epoch": 2.3135255558887793,
      "grad_norm": 8.643807411193848,
      "learning_rate": 5.5844873192467506e-05,
      "loss": 2.1556068420410157,
      "memory(GiB)": 77.56,
      "step": 54000,
      "token_acc": 0.549645390070922,
      "train_speed(iter/s)": 1.438956
    },
    {
      "epoch": 2.3135255558887793,
      "eval_loss": 2.3864502906799316,
      "eval_runtime": 14.6584,
      "eval_samples_per_second": 6.822,
      "eval_steps_per_second": 6.822,
      "eval_token_acc": 0.47003994673768307,
      "step": 54000
    },
    {
      "epoch": 2.3137397712180285,
      "grad_norm": 4.805451393127441,
      "learning_rate": 5.583818950589629e-05,
      "loss": 2.264748764038086,
      "memory(GiB)": 77.56,
      "step": 54005,
      "token_acc": 0.4810964083175803,
      "train_speed(iter/s)": 1.438351
    },
    {
      "epoch": 2.3139539865472774,
      "grad_norm": 5.354040145874023,
      "learning_rate": 5.583150571356079e-05,
      "loss": 2.4373064041137695,
      "memory(GiB)": 77.56,
      "step": 54010,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.438399
    },
    {
      "epoch": 2.314168201876526,
      "grad_norm": 5.284209251403809,
      "learning_rate": 5.5824821815582104e-05,
      "loss": 2.2593448638916014,
      "memory(GiB)": 77.56,
      "step": 54015,
      "token_acc": 0.5512367491166078,
      "train_speed(iter/s)": 1.438422
    },
    {
      "epoch": 2.3143824172057754,
      "grad_norm": 4.4956278800964355,
      "learning_rate": 5.5818137812081306e-05,
      "loss": 2.4375816345214845,
      "memory(GiB)": 77.56,
      "step": 54020,
      "token_acc": 0.5029585798816568,
      "train_speed(iter/s)": 1.438436
    },
    {
      "epoch": 2.3145966325350242,
      "grad_norm": 8.219070434570312,
      "learning_rate": 5.581145370317948e-05,
      "loss": 2.1817834854125975,
      "memory(GiB)": 77.56,
      "step": 54025,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438432
    },
    {
      "epoch": 2.314810847864273,
      "grad_norm": 5.6516594886779785,
      "learning_rate": 5.5804769488997724e-05,
      "loss": 2.3674079895019533,
      "memory(GiB)": 77.56,
      "step": 54030,
      "token_acc": 0.5220125786163522,
      "train_speed(iter/s)": 1.438455
    },
    {
      "epoch": 2.3150250631935223,
      "grad_norm": 5.184658527374268,
      "learning_rate": 5.579808516965711e-05,
      "loss": 2.2186534881591795,
      "memory(GiB)": 77.56,
      "step": 54035,
      "token_acc": 0.49375,
      "train_speed(iter/s)": 1.438493
    },
    {
      "epoch": 2.315239278522771,
      "grad_norm": 5.937741756439209,
      "learning_rate": 5.579140074527877e-05,
      "loss": 2.2813316345214845,
      "memory(GiB)": 77.56,
      "step": 54040,
      "token_acc": 0.524,
      "train_speed(iter/s)": 1.438491
    },
    {
      "epoch": 2.31545349385202,
      "grad_norm": 4.384599685668945,
      "learning_rate": 5.578471621598376e-05,
      "loss": 2.4732698440551757,
      "memory(GiB)": 77.56,
      "step": 54045,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.43849
    },
    {
      "epoch": 2.315667709181269,
      "grad_norm": 5.9085774421691895,
      "learning_rate": 5.5778031581893175e-05,
      "loss": 2.7172569274902343,
      "memory(GiB)": 77.56,
      "step": 54050,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.438466
    },
    {
      "epoch": 2.315881924510518,
      "grad_norm": 6.28882360458374,
      "learning_rate": 5.5771346843128124e-05,
      "loss": 2.4833419799804686,
      "memory(GiB)": 77.56,
      "step": 54055,
      "token_acc": 0.4507936507936508,
      "train_speed(iter/s)": 1.438494
    },
    {
      "epoch": 2.316096139839767,
      "grad_norm": 7.162354946136475,
      "learning_rate": 5.576466199980973e-05,
      "loss": 2.6254657745361327,
      "memory(GiB)": 77.56,
      "step": 54060,
      "token_acc": 0.4351851851851852,
      "train_speed(iter/s)": 1.438504
    },
    {
      "epoch": 2.316310355169016,
      "grad_norm": 5.696046352386475,
      "learning_rate": 5.575797705205907e-05,
      "loss": 2.5408376693725585,
      "memory(GiB)": 77.56,
      "step": 54065,
      "token_acc": 0.4524590163934426,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 2.316524570498265,
      "grad_norm": 5.709765911102295,
      "learning_rate": 5.5751291999997256e-05,
      "loss": 2.3990570068359376,
      "memory(GiB)": 77.56,
      "step": 54070,
      "token_acc": 0.49846153846153846,
      "train_speed(iter/s)": 1.438515
    },
    {
      "epoch": 2.3167387858275137,
      "grad_norm": 5.901362419128418,
      "learning_rate": 5.574460684374541e-05,
      "loss": 2.7339168548583985,
      "memory(GiB)": 77.56,
      "step": 54075,
      "token_acc": 0.4475920679886686,
      "train_speed(iter/s)": 1.438525
    },
    {
      "epoch": 2.316953001156763,
      "grad_norm": 5.384810447692871,
      "learning_rate": 5.573792158342459e-05,
      "loss": 2.2162403106689452,
      "memory(GiB)": 77.56,
      "step": 54080,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.438516
    },
    {
      "epoch": 2.3171672164860118,
      "grad_norm": 4.498585224151611,
      "learning_rate": 5.573123621915595e-05,
      "loss": 2.4023914337158203,
      "memory(GiB)": 77.56,
      "step": 54085,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.438502
    },
    {
      "epoch": 2.3173814318152606,
      "grad_norm": 7.0708699226379395,
      "learning_rate": 5.572455075106059e-05,
      "loss": 2.5315731048583983,
      "memory(GiB)": 77.56,
      "step": 54090,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.438511
    },
    {
      "epoch": 2.31759564714451,
      "grad_norm": 6.007331371307373,
      "learning_rate": 5.5717865179259624e-05,
      "loss": 2.358246994018555,
      "memory(GiB)": 77.56,
      "step": 54095,
      "token_acc": 0.477124183006536,
      "train_speed(iter/s)": 1.438482
    },
    {
      "epoch": 2.3178098624737586,
      "grad_norm": 6.702929973602295,
      "learning_rate": 5.571117950387417e-05,
      "loss": 2.501856803894043,
      "memory(GiB)": 77.56,
      "step": 54100,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.438499
    },
    {
      "epoch": 2.3180240778030075,
      "grad_norm": 5.773906707763672,
      "learning_rate": 5.570449372502534e-05,
      "loss": 2.597435188293457,
      "memory(GiB)": 77.56,
      "step": 54105,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438501
    },
    {
      "epoch": 2.3182382931322567,
      "grad_norm": 5.605736255645752,
      "learning_rate": 5.5697807842834245e-05,
      "loss": 2.514650344848633,
      "memory(GiB)": 77.56,
      "step": 54110,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.438524
    },
    {
      "epoch": 2.3184525084615055,
      "grad_norm": 5.441324710845947,
      "learning_rate": 5.5691121857422034e-05,
      "loss": 2.143595886230469,
      "memory(GiB)": 77.56,
      "step": 54115,
      "token_acc": 0.5042016806722689,
      "train_speed(iter/s)": 1.438566
    },
    {
      "epoch": 2.3186667237907543,
      "grad_norm": 5.025784015655518,
      "learning_rate": 5.568443576890979e-05,
      "loss": 2.489344596862793,
      "memory(GiB)": 77.56,
      "step": 54120,
      "token_acc": 0.4794520547945205,
      "train_speed(iter/s)": 1.438548
    },
    {
      "epoch": 2.3188809391200036,
      "grad_norm": 6.301708698272705,
      "learning_rate": 5.567774957741867e-05,
      "loss": 2.4384708404541016,
      "memory(GiB)": 77.56,
      "step": 54125,
      "token_acc": 0.4309210526315789,
      "train_speed(iter/s)": 1.438567
    },
    {
      "epoch": 2.3190951544492524,
      "grad_norm": 5.335469722747803,
      "learning_rate": 5.56710632830698e-05,
      "loss": 1.951573944091797,
      "memory(GiB)": 77.56,
      "step": 54130,
      "token_acc": 0.576,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 2.319309369778501,
      "grad_norm": 5.782993316650391,
      "learning_rate": 5.5664376885984296e-05,
      "loss": 2.064581871032715,
      "memory(GiB)": 77.56,
      "step": 54135,
      "token_acc": 0.5271966527196653,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.3195235851077505,
      "grad_norm": 5.268125534057617,
      "learning_rate": 5.565769038628328e-05,
      "loss": 2.3723209381103514,
      "memory(GiB)": 77.56,
      "step": 54140,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.438558
    },
    {
      "epoch": 2.3197378004369993,
      "grad_norm": 8.005107879638672,
      "learning_rate": 5.5651003784087904e-05,
      "loss": 2.663521575927734,
      "memory(GiB)": 77.56,
      "step": 54145,
      "token_acc": 0.4728682170542636,
      "train_speed(iter/s)": 1.438569
    },
    {
      "epoch": 2.319952015766248,
      "grad_norm": 4.887722969055176,
      "learning_rate": 5.5644317079519296e-05,
      "loss": 2.5326534271240235,
      "memory(GiB)": 77.56,
      "step": 54150,
      "token_acc": 0.4724137931034483,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.3201662310954974,
      "grad_norm": 5.536297798156738,
      "learning_rate": 5.5637630272698584e-05,
      "loss": 2.433104705810547,
      "memory(GiB)": 77.56,
      "step": 54155,
      "token_acc": 0.4707692307692308,
      "train_speed(iter/s)": 1.438535
    },
    {
      "epoch": 2.320380446424746,
      "grad_norm": 5.90949010848999,
      "learning_rate": 5.56309433637469e-05,
      "loss": 2.585856056213379,
      "memory(GiB)": 77.56,
      "step": 54160,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 2.320594661753995,
      "grad_norm": 5.377594470977783,
      "learning_rate": 5.562425635278541e-05,
      "loss": 2.2658287048339845,
      "memory(GiB)": 77.56,
      "step": 54165,
      "token_acc": 0.4797507788161994,
      "train_speed(iter/s)": 1.43852
    },
    {
      "epoch": 2.3208088770832442,
      "grad_norm": 5.158398151397705,
      "learning_rate": 5.561756923993523e-05,
      "loss": 2.622241401672363,
      "memory(GiB)": 77.56,
      "step": 54170,
      "token_acc": 0.45151515151515154,
      "train_speed(iter/s)": 1.438526
    },
    {
      "epoch": 2.321023092412493,
      "grad_norm": 6.190441131591797,
      "learning_rate": 5.561088202531752e-05,
      "loss": 2.3491865158081056,
      "memory(GiB)": 77.56,
      "step": 54175,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.438532
    },
    {
      "epoch": 2.321237307741742,
      "grad_norm": 7.089932441711426,
      "learning_rate": 5.5604194709053424e-05,
      "loss": 2.508444595336914,
      "memory(GiB)": 77.56,
      "step": 54180,
      "token_acc": 0.5134099616858238,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 2.321451523070991,
      "grad_norm": 7.774629592895508,
      "learning_rate": 5.5597507291264087e-05,
      "loss": 2.468191909790039,
      "memory(GiB)": 77.56,
      "step": 54185,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.438607
    },
    {
      "epoch": 2.32166573840024,
      "grad_norm": 6.656830310821533,
      "learning_rate": 5.559081977207065e-05,
      "loss": 2.3286041259765624,
      "memory(GiB)": 77.56,
      "step": 54190,
      "token_acc": 0.4661016949152542,
      "train_speed(iter/s)": 1.438622
    },
    {
      "epoch": 2.3218799537294887,
      "grad_norm": 6.327261924743652,
      "learning_rate": 5.558413215159425e-05,
      "loss": 2.519520950317383,
      "memory(GiB)": 77.56,
      "step": 54195,
      "token_acc": 0.47435897435897434,
      "train_speed(iter/s)": 1.438628
    },
    {
      "epoch": 2.322094169058738,
      "grad_norm": 5.310946941375732,
      "learning_rate": 5.557744442995607e-05,
      "loss": 2.695193862915039,
      "memory(GiB)": 77.56,
      "step": 54200,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 2.322308384387987,
      "grad_norm": 9.110607147216797,
      "learning_rate": 5.5570756607277256e-05,
      "loss": 2.407896041870117,
      "memory(GiB)": 77.56,
      "step": 54205,
      "token_acc": 0.5539568345323741,
      "train_speed(iter/s)": 1.438627
    },
    {
      "epoch": 2.3225225997172356,
      "grad_norm": 4.725812911987305,
      "learning_rate": 5.556406868367895e-05,
      "loss": 2.484501075744629,
      "memory(GiB)": 77.56,
      "step": 54210,
      "token_acc": 0.4815950920245399,
      "train_speed(iter/s)": 1.438608
    },
    {
      "epoch": 2.322736815046485,
      "grad_norm": 6.263838291168213,
      "learning_rate": 5.555738065928233e-05,
      "loss": 2.5613006591796874,
      "memory(GiB)": 77.56,
      "step": 54215,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 2.3229510303757337,
      "grad_norm": 6.026326656341553,
      "learning_rate": 5.555069253420855e-05,
      "loss": 2.2715780258178713,
      "memory(GiB)": 77.56,
      "step": 54220,
      "token_acc": 0.5433333333333333,
      "train_speed(iter/s)": 1.43864
    },
    {
      "epoch": 2.3231652457049825,
      "grad_norm": 5.113918781280518,
      "learning_rate": 5.554400430857876e-05,
      "loss": 2.2215293884277343,
      "memory(GiB)": 77.56,
      "step": 54225,
      "token_acc": 0.5284810126582279,
      "train_speed(iter/s)": 1.438674
    },
    {
      "epoch": 2.3233794610342318,
      "grad_norm": 4.733516216278076,
      "learning_rate": 5.5537315982514135e-05,
      "loss": 2.378024101257324,
      "memory(GiB)": 77.56,
      "step": 54230,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.438675
    },
    {
      "epoch": 2.3235936763634806,
      "grad_norm": 4.240628242492676,
      "learning_rate": 5.5530627556135826e-05,
      "loss": 2.7475404739379883,
      "memory(GiB)": 77.56,
      "step": 54235,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.438685
    },
    {
      "epoch": 2.3238078916927294,
      "grad_norm": 7.404336929321289,
      "learning_rate": 5.552393902956502e-05,
      "loss": 2.373859977722168,
      "memory(GiB)": 77.56,
      "step": 54240,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.438728
    },
    {
      "epoch": 2.3240221070219786,
      "grad_norm": 5.088076114654541,
      "learning_rate": 5.551725040292288e-05,
      "loss": 2.458535003662109,
      "memory(GiB)": 77.56,
      "step": 54245,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.438709
    },
    {
      "epoch": 2.3242363223512275,
      "grad_norm": 5.420531749725342,
      "learning_rate": 5.551056167633058e-05,
      "loss": 2.508696937561035,
      "memory(GiB)": 77.56,
      "step": 54250,
      "token_acc": 0.48562300319488816,
      "train_speed(iter/s)": 1.438727
    },
    {
      "epoch": 2.3244505376804763,
      "grad_norm": 5.287556171417236,
      "learning_rate": 5.550387284990927e-05,
      "loss": 2.508671760559082,
      "memory(GiB)": 77.56,
      "step": 54255,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.438717
    },
    {
      "epoch": 2.3246647530097255,
      "grad_norm": 4.1852946281433105,
      "learning_rate": 5.549718392378014e-05,
      "loss": 2.6818719863891602,
      "memory(GiB)": 77.56,
      "step": 54260,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.438751
    },
    {
      "epoch": 2.3248789683389743,
      "grad_norm": 4.860667705535889,
      "learning_rate": 5.549049489806437e-05,
      "loss": 2.311024856567383,
      "memory(GiB)": 77.56,
      "step": 54265,
      "token_acc": 0.48881789137380194,
      "train_speed(iter/s)": 1.438779
    },
    {
      "epoch": 2.325093183668223,
      "grad_norm": 6.025043487548828,
      "learning_rate": 5.5483805772883133e-05,
      "loss": 2.485648536682129,
      "memory(GiB)": 77.56,
      "step": 54270,
      "token_acc": 0.48732394366197185,
      "train_speed(iter/s)": 1.438812
    },
    {
      "epoch": 2.3253073989974724,
      "grad_norm": 5.138810634613037,
      "learning_rate": 5.547711654835761e-05,
      "loss": 2.5689886093139647,
      "memory(GiB)": 77.56,
      "step": 54275,
      "token_acc": 0.4794007490636704,
      "train_speed(iter/s)": 1.438815
    },
    {
      "epoch": 2.325521614326721,
      "grad_norm": 5.626759052276611,
      "learning_rate": 5.5470427224609e-05,
      "loss": 2.79862060546875,
      "memory(GiB)": 77.56,
      "step": 54280,
      "token_acc": 0.4654255319148936,
      "train_speed(iter/s)": 1.438768
    },
    {
      "epoch": 2.32573582965597,
      "grad_norm": 4.4945478439331055,
      "learning_rate": 5.5463737801758443e-05,
      "loss": 2.5011838912963866,
      "memory(GiB)": 77.56,
      "step": 54285,
      "token_acc": 0.4752851711026616,
      "train_speed(iter/s)": 1.438747
    },
    {
      "epoch": 2.3259500449852193,
      "grad_norm": 6.036630153656006,
      "learning_rate": 5.545704827992717e-05,
      "loss": 2.304207611083984,
      "memory(GiB)": 77.56,
      "step": 54290,
      "token_acc": 0.518796992481203,
      "train_speed(iter/s)": 1.438728
    },
    {
      "epoch": 2.326164260314468,
      "grad_norm": 5.231227397918701,
      "learning_rate": 5.5450358659236336e-05,
      "loss": 2.514471435546875,
      "memory(GiB)": 77.56,
      "step": 54295,
      "token_acc": 0.46387832699619774,
      "train_speed(iter/s)": 1.438719
    },
    {
      "epoch": 2.326378475643717,
      "grad_norm": 4.567483425140381,
      "learning_rate": 5.5443668939807156e-05,
      "loss": 2.3522150039672853,
      "memory(GiB)": 77.56,
      "step": 54300,
      "token_acc": 0.5064516129032258,
      "train_speed(iter/s)": 1.438719
    },
    {
      "epoch": 2.326592690972966,
      "grad_norm": 5.361208915710449,
      "learning_rate": 5.5436979121760804e-05,
      "loss": 2.5086477279663084,
      "memory(GiB)": 77.56,
      "step": 54305,
      "token_acc": 0.5126582278481012,
      "train_speed(iter/s)": 1.438687
    },
    {
      "epoch": 2.326806906302215,
      "grad_norm": 5.930521011352539,
      "learning_rate": 5.543028920521846e-05,
      "loss": 2.255126190185547,
      "memory(GiB)": 77.56,
      "step": 54310,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.327021121631464,
      "grad_norm": 5.941011905670166,
      "learning_rate": 5.542359919030133e-05,
      "loss": 2.325210952758789,
      "memory(GiB)": 77.56,
      "step": 54315,
      "token_acc": 0.49691358024691357,
      "train_speed(iter/s)": 1.438641
    },
    {
      "epoch": 2.327235336960713,
      "grad_norm": 5.84672737121582,
      "learning_rate": 5.541690907713063e-05,
      "loss": 2.555367279052734,
      "memory(GiB)": 77.56,
      "step": 54320,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 2.327449552289962,
      "grad_norm": 7.087546348571777,
      "learning_rate": 5.5410218865827534e-05,
      "loss": 2.2478092193603514,
      "memory(GiB)": 77.56,
      "step": 54325,
      "token_acc": 0.536,
      "train_speed(iter/s)": 1.438684
    },
    {
      "epoch": 2.3276637676192107,
      "grad_norm": 4.922154903411865,
      "learning_rate": 5.540352855651325e-05,
      "loss": 2.2991874694824217,
      "memory(GiB)": 77.56,
      "step": 54330,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.438691
    },
    {
      "epoch": 2.32787798294846,
      "grad_norm": 4.60272741317749,
      "learning_rate": 5.539683814930898e-05,
      "loss": 2.4507495880126955,
      "memory(GiB)": 77.56,
      "step": 54335,
      "token_acc": 0.5315985130111525,
      "train_speed(iter/s)": 1.438688
    },
    {
      "epoch": 2.3280921982777087,
      "grad_norm": 4.2224931716918945,
      "learning_rate": 5.539014764433592e-05,
      "loss": 2.5097467422485353,
      "memory(GiB)": 77.56,
      "step": 54340,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.438691
    },
    {
      "epoch": 2.3283064136069576,
      "grad_norm": 4.0915751457214355,
      "learning_rate": 5.538345704171527e-05,
      "loss": 1.945652961730957,
      "memory(GiB)": 77.56,
      "step": 54345,
      "token_acc": 0.5351170568561873,
      "train_speed(iter/s)": 1.438678
    },
    {
      "epoch": 2.328520628936207,
      "grad_norm": 5.094549655914307,
      "learning_rate": 5.537676634156827e-05,
      "loss": 2.4953020095825194,
      "memory(GiB)": 77.56,
      "step": 54350,
      "token_acc": 0.49696969696969695,
      "train_speed(iter/s)": 1.438678
    },
    {
      "epoch": 2.3287348442654556,
      "grad_norm": 4.522862911224365,
      "learning_rate": 5.537007554401608e-05,
      "loss": 2.5908123016357423,
      "memory(GiB)": 77.56,
      "step": 54355,
      "token_acc": 0.45734597156398105,
      "train_speed(iter/s)": 1.438708
    },
    {
      "epoch": 2.3289490595947044,
      "grad_norm": 5.012341499328613,
      "learning_rate": 5.536338464917995e-05,
      "loss": 2.46358699798584,
      "memory(GiB)": 77.56,
      "step": 54360,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.438717
    },
    {
      "epoch": 2.3291632749239537,
      "grad_norm": 5.816551208496094,
      "learning_rate": 5.535669365718107e-05,
      "loss": 2.1036087036132813,
      "memory(GiB)": 77.56,
      "step": 54365,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.438754
    },
    {
      "epoch": 2.3293774902532025,
      "grad_norm": 6.4626030921936035,
      "learning_rate": 5.535000256814066e-05,
      "loss": 2.3061901092529298,
      "memory(GiB)": 77.56,
      "step": 54370,
      "token_acc": 0.4844290657439446,
      "train_speed(iter/s)": 1.438707
    },
    {
      "epoch": 2.3295917055824513,
      "grad_norm": 5.194098472595215,
      "learning_rate": 5.534331138217993e-05,
      "loss": 2.339299774169922,
      "memory(GiB)": 77.56,
      "step": 54375,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.438729
    },
    {
      "epoch": 2.3298059209117006,
      "grad_norm": 5.277906894683838,
      "learning_rate": 5.53366200994201e-05,
      "loss": 2.381866455078125,
      "memory(GiB)": 77.56,
      "step": 54380,
      "token_acc": 0.49693251533742333,
      "train_speed(iter/s)": 1.438709
    },
    {
      "epoch": 2.3300201362409494,
      "grad_norm": 4.18302583694458,
      "learning_rate": 5.532992871998239e-05,
      "loss": 2.344228744506836,
      "memory(GiB)": 77.56,
      "step": 54385,
      "token_acc": 0.5249169435215947,
      "train_speed(iter/s)": 1.438741
    },
    {
      "epoch": 2.330234351570198,
      "grad_norm": 4.742477893829346,
      "learning_rate": 5.532323724398804e-05,
      "loss": 2.58557243347168,
      "memory(GiB)": 77.56,
      "step": 54390,
      "token_acc": 0.4735099337748344,
      "train_speed(iter/s)": 1.438737
    },
    {
      "epoch": 2.3304485668994475,
      "grad_norm": 8.05929946899414,
      "learning_rate": 5.531654567155824e-05,
      "loss": 2.248756217956543,
      "memory(GiB)": 77.56,
      "step": 54395,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438754
    },
    {
      "epoch": 2.3306627822286963,
      "grad_norm": 3.982360601425171,
      "learning_rate": 5.530985400281422e-05,
      "loss": 2.4308753967285157,
      "memory(GiB)": 77.56,
      "step": 54400,
      "token_acc": 0.5016501650165016,
      "train_speed(iter/s)": 1.438787
    },
    {
      "epoch": 2.330876997557945,
      "grad_norm": 7.862496376037598,
      "learning_rate": 5.530316223787723e-05,
      "loss": 2.3568853378295898,
      "memory(GiB)": 77.56,
      "step": 54405,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.438724
    },
    {
      "epoch": 2.3310912128871943,
      "grad_norm": 6.096340179443359,
      "learning_rate": 5.529647037686847e-05,
      "loss": 2.3952186584472654,
      "memory(GiB)": 77.56,
      "step": 54410,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.438751
    },
    {
      "epoch": 2.331305428216443,
      "grad_norm": 5.931663990020752,
      "learning_rate": 5.528977841990919e-05,
      "loss": 2.5866153717041014,
      "memory(GiB)": 77.56,
      "step": 54415,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438757
    },
    {
      "epoch": 2.331519643545692,
      "grad_norm": 4.925633907318115,
      "learning_rate": 5.5283086367120594e-05,
      "loss": 2.492633819580078,
      "memory(GiB)": 77.56,
      "step": 54420,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.438751
    },
    {
      "epoch": 2.3317338588749412,
      "grad_norm": 4.742899417877197,
      "learning_rate": 5.527639421862395e-05,
      "loss": 2.5154991149902344,
      "memory(GiB)": 77.56,
      "step": 54425,
      "token_acc": 0.45695364238410596,
      "train_speed(iter/s)": 1.438777
    },
    {
      "epoch": 2.33194807420419,
      "grad_norm": 4.619522571563721,
      "learning_rate": 5.526970197454047e-05,
      "loss": 2.53272705078125,
      "memory(GiB)": 77.56,
      "step": 54430,
      "token_acc": 0.49848024316109424,
      "train_speed(iter/s)": 1.438776
    },
    {
      "epoch": 2.332162289533439,
      "grad_norm": 4.009924411773682,
      "learning_rate": 5.52630096349914e-05,
      "loss": 2.7802892684936524,
      "memory(GiB)": 77.56,
      "step": 54435,
      "token_acc": 0.44140625,
      "train_speed(iter/s)": 1.438774
    },
    {
      "epoch": 2.332376504862688,
      "grad_norm": 5.807552337646484,
      "learning_rate": 5.525631720009796e-05,
      "loss": 2.6330240249633787,
      "memory(GiB)": 77.56,
      "step": 54440,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.438798
    },
    {
      "epoch": 2.332590720191937,
      "grad_norm": 4.5297040939331055,
      "learning_rate": 5.524962466998142e-05,
      "loss": 2.5877969741821287,
      "memory(GiB)": 77.56,
      "step": 54445,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.438786
    },
    {
      "epoch": 2.3328049355211857,
      "grad_norm": 5.650115013122559,
      "learning_rate": 5.5242932044762994e-05,
      "loss": 2.4883726119995115,
      "memory(GiB)": 77.56,
      "step": 54450,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.438794
    },
    {
      "epoch": 2.333019150850435,
      "grad_norm": 5.331174850463867,
      "learning_rate": 5.523623932456394e-05,
      "loss": 2.5523576736450195,
      "memory(GiB)": 77.56,
      "step": 54455,
      "token_acc": 0.4855072463768116,
      "train_speed(iter/s)": 1.438801
    },
    {
      "epoch": 2.333233366179684,
      "grad_norm": 5.889141082763672,
      "learning_rate": 5.522954650950549e-05,
      "loss": 2.586617088317871,
      "memory(GiB)": 77.56,
      "step": 54460,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.438793
    },
    {
      "epoch": 2.3334475815089326,
      "grad_norm": 6.986037731170654,
      "learning_rate": 5.522285359970891e-05,
      "loss": 2.318703460693359,
      "memory(GiB)": 77.56,
      "step": 54465,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.438781
    },
    {
      "epoch": 2.333661796838182,
      "grad_norm": 5.666684150695801,
      "learning_rate": 5.521616059529543e-05,
      "loss": 2.489665985107422,
      "memory(GiB)": 77.56,
      "step": 54470,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.438814
    },
    {
      "epoch": 2.3338760121674307,
      "grad_norm": 4.9587788581848145,
      "learning_rate": 5.5209467496386315e-05,
      "loss": 2.844677543640137,
      "memory(GiB)": 77.56,
      "step": 54475,
      "token_acc": 0.4367469879518072,
      "train_speed(iter/s)": 1.438818
    },
    {
      "epoch": 2.3340902274966795,
      "grad_norm": 6.0969462394714355,
      "learning_rate": 5.5202774303102814e-05,
      "loss": 2.2558618545532227,
      "memory(GiB)": 77.56,
      "step": 54480,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.438814
    },
    {
      "epoch": 2.3343044428259287,
      "grad_norm": 5.694857120513916,
      "learning_rate": 5.519608101556616e-05,
      "loss": 2.535897636413574,
      "memory(GiB)": 77.56,
      "step": 54485,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.438826
    },
    {
      "epoch": 2.3345186581551776,
      "grad_norm": 5.296054840087891,
      "learning_rate": 5.518938763389764e-05,
      "loss": 2.5851299285888674,
      "memory(GiB)": 77.56,
      "step": 54490,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.438858
    },
    {
      "epoch": 2.3347328734844264,
      "grad_norm": 4.489163875579834,
      "learning_rate": 5.5182694158218476e-05,
      "loss": 2.463297462463379,
      "memory(GiB)": 77.56,
      "step": 54495,
      "token_acc": 0.498371335504886,
      "train_speed(iter/s)": 1.43885
    },
    {
      "epoch": 2.3349470888136756,
      "grad_norm": 5.169692039489746,
      "learning_rate": 5.5176000588649946e-05,
      "loss": 2.5896942138671877,
      "memory(GiB)": 77.56,
      "step": 54500,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.438851
    },
    {
      "epoch": 2.3349470888136756,
      "eval_loss": 2.114393472671509,
      "eval_runtime": 14.5742,
      "eval_samples_per_second": 6.861,
      "eval_steps_per_second": 6.861,
      "eval_token_acc": 0.49039692701664533,
      "step": 54500
    },
    {
      "epoch": 2.3351613041429244,
      "grad_norm": 5.3127522468566895,
      "learning_rate": 5.5169306925313324e-05,
      "loss": 2.268874168395996,
      "memory(GiB)": 77.56,
      "step": 54505,
      "token_acc": 0.49150943396226415,
      "train_speed(iter/s)": 1.438283
    },
    {
      "epoch": 2.3353755194721733,
      "grad_norm": 6.650249004364014,
      "learning_rate": 5.5162613168329856e-05,
      "loss": 2.3926322937011717,
      "memory(GiB)": 77.56,
      "step": 54510,
      "token_acc": 0.5160142348754448,
      "train_speed(iter/s)": 1.438293
    },
    {
      "epoch": 2.3355897348014225,
      "grad_norm": 5.038234710693359,
      "learning_rate": 5.5155919317820806e-05,
      "loss": 2.2601802825927733,
      "memory(GiB)": 77.56,
      "step": 54515,
      "token_acc": 0.5275590551181102,
      "train_speed(iter/s)": 1.43826
    },
    {
      "epoch": 2.3358039501306713,
      "grad_norm": 5.188998222351074,
      "learning_rate": 5.514922537390744e-05,
      "loss": 2.011192512512207,
      "memory(GiB)": 77.56,
      "step": 54520,
      "token_acc": 0.5362318840579711,
      "train_speed(iter/s)": 1.438245
    },
    {
      "epoch": 2.33601816545992,
      "grad_norm": 4.637081146240234,
      "learning_rate": 5.5142531336711034e-05,
      "loss": 2.211981773376465,
      "memory(GiB)": 77.56,
      "step": 54525,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438247
    },
    {
      "epoch": 2.3362323807891694,
      "grad_norm": 5.339656829833984,
      "learning_rate": 5.513583720635283e-05,
      "loss": 2.3581382751464846,
      "memory(GiB)": 77.56,
      "step": 54530,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.43827
    },
    {
      "epoch": 2.336446596118418,
      "grad_norm": 10.428476333618164,
      "learning_rate": 5.512914298295413e-05,
      "loss": 2.3907444000244142,
      "memory(GiB)": 77.56,
      "step": 54535,
      "token_acc": 0.5033333333333333,
      "train_speed(iter/s)": 1.438244
    },
    {
      "epoch": 2.336660811447667,
      "grad_norm": 4.768646717071533,
      "learning_rate": 5.51224486666362e-05,
      "loss": 2.534077262878418,
      "memory(GiB)": 77.56,
      "step": 54540,
      "token_acc": 0.4837758112094395,
      "train_speed(iter/s)": 1.438268
    },
    {
      "epoch": 2.3368750267769163,
      "grad_norm": 5.592228412628174,
      "learning_rate": 5.511575425752029e-05,
      "loss": 2.669137954711914,
      "memory(GiB)": 77.56,
      "step": 54545,
      "token_acc": 0.43272727272727274,
      "train_speed(iter/s)": 1.438253
    },
    {
      "epoch": 2.337089242106165,
      "grad_norm": 5.471279144287109,
      "learning_rate": 5.510905975572771e-05,
      "loss": 2.3823238372802735,
      "memory(GiB)": 77.56,
      "step": 54550,
      "token_acc": 0.5031847133757962,
      "train_speed(iter/s)": 1.438215
    },
    {
      "epoch": 2.337303457435414,
      "grad_norm": 6.329859256744385,
      "learning_rate": 5.51023651613797e-05,
      "loss": 2.399275779724121,
      "memory(GiB)": 77.56,
      "step": 54555,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.438215
    },
    {
      "epoch": 2.337517672764663,
      "grad_norm": 5.307493686676025,
      "learning_rate": 5.5095670474597574e-05,
      "loss": 2.191426658630371,
      "memory(GiB)": 77.56,
      "step": 54560,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.43823
    },
    {
      "epoch": 2.337731888093912,
      "grad_norm": 5.246840000152588,
      "learning_rate": 5.508897569550259e-05,
      "loss": 2.470229911804199,
      "memory(GiB)": 77.56,
      "step": 54565,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.438222
    },
    {
      "epoch": 2.337946103423161,
      "grad_norm": 6.284454822540283,
      "learning_rate": 5.5082280824216046e-05,
      "loss": 2.5420969009399412,
      "memory(GiB)": 77.56,
      "step": 54570,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.438217
    },
    {
      "epoch": 2.33816031875241,
      "grad_norm": 5.767074108123779,
      "learning_rate": 5.5075585860859206e-05,
      "loss": 2.346650505065918,
      "memory(GiB)": 77.56,
      "step": 54575,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.438231
    },
    {
      "epoch": 2.338374534081659,
      "grad_norm": 5.003377437591553,
      "learning_rate": 5.5068890805553374e-05,
      "loss": 2.387823486328125,
      "memory(GiB)": 77.56,
      "step": 54580,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.438271
    },
    {
      "epoch": 2.3385887494109077,
      "grad_norm": 6.539522171020508,
      "learning_rate": 5.506219565841983e-05,
      "loss": 2.4285747528076174,
      "memory(GiB)": 77.56,
      "step": 54585,
      "token_acc": 0.46184738955823296,
      "train_speed(iter/s)": 1.438274
    },
    {
      "epoch": 2.338802964740157,
      "grad_norm": 4.871034622192383,
      "learning_rate": 5.5055500419579866e-05,
      "loss": 2.1912952423095704,
      "memory(GiB)": 77.56,
      "step": 54590,
      "token_acc": 0.5412186379928315,
      "train_speed(iter/s)": 1.438295
    },
    {
      "epoch": 2.3390171800694057,
      "grad_norm": 4.646417617797852,
      "learning_rate": 5.5048805089154765e-05,
      "loss": 2.1723554611206053,
      "memory(GiB)": 77.56,
      "step": 54595,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.438293
    },
    {
      "epoch": 2.3392313953986545,
      "grad_norm": 5.2171783447265625,
      "learning_rate": 5.504210966726582e-05,
      "loss": 2.253154182434082,
      "memory(GiB)": 77.56,
      "step": 54600,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.438277
    },
    {
      "epoch": 2.339445610727904,
      "grad_norm": 5.858293056488037,
      "learning_rate": 5.503541415403433e-05,
      "loss": 2.4629806518554687,
      "memory(GiB)": 77.56,
      "step": 54605,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.438261
    },
    {
      "epoch": 2.3396598260571526,
      "grad_norm": 5.127858638763428,
      "learning_rate": 5.502871854958159e-05,
      "loss": 2.4169069290161134,
      "memory(GiB)": 77.56,
      "step": 54610,
      "token_acc": 0.4896755162241888,
      "train_speed(iter/s)": 1.438255
    },
    {
      "epoch": 2.3398740413864014,
      "grad_norm": 5.172540187835693,
      "learning_rate": 5.502202285402889e-05,
      "loss": 2.552465057373047,
      "memory(GiB)": 77.56,
      "step": 54615,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.438273
    },
    {
      "epoch": 2.3400882567156507,
      "grad_norm": 5.092914581298828,
      "learning_rate": 5.5015327067497544e-05,
      "loss": 2.1506404876708984,
      "memory(GiB)": 77.56,
      "step": 54620,
      "token_acc": 0.5308641975308642,
      "train_speed(iter/s)": 1.438274
    },
    {
      "epoch": 2.3403024720448995,
      "grad_norm": 5.406525135040283,
      "learning_rate": 5.500863119010885e-05,
      "loss": 2.4598031997680665,
      "memory(GiB)": 77.56,
      "step": 54625,
      "token_acc": 0.4608150470219436,
      "train_speed(iter/s)": 1.438277
    },
    {
      "epoch": 2.3405166873741483,
      "grad_norm": 5.842250347137451,
      "learning_rate": 5.500193522198409e-05,
      "loss": 2.4525663375854494,
      "memory(GiB)": 77.56,
      "step": 54630,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438292
    },
    {
      "epoch": 2.3407309027033976,
      "grad_norm": 8.250039100646973,
      "learning_rate": 5.499523916324459e-05,
      "loss": 2.4933700561523438,
      "memory(GiB)": 77.56,
      "step": 54635,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.438278
    },
    {
      "epoch": 2.3409451180326464,
      "grad_norm": 5.769161224365234,
      "learning_rate": 5.498854301401164e-05,
      "loss": 2.7864315032958986,
      "memory(GiB)": 77.56,
      "step": 54640,
      "token_acc": 0.4157303370786517,
      "train_speed(iter/s)": 1.438304
    },
    {
      "epoch": 2.341159333361895,
      "grad_norm": 5.727543354034424,
      "learning_rate": 5.4981846774406555e-05,
      "loss": 2.553101348876953,
      "memory(GiB)": 77.56,
      "step": 54645,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438319
    },
    {
      "epoch": 2.3413735486911444,
      "grad_norm": 6.9831767082214355,
      "learning_rate": 5.497515044455065e-05,
      "loss": 2.28961296081543,
      "memory(GiB)": 77.56,
      "step": 54650,
      "token_acc": 0.5202492211838006,
      "train_speed(iter/s)": 1.438345
    },
    {
      "epoch": 2.3415877640203933,
      "grad_norm": 4.3163161277771,
      "learning_rate": 5.496845402456522e-05,
      "loss": 2.5322153091430666,
      "memory(GiB)": 77.56,
      "step": 54655,
      "token_acc": 0.42950819672131146,
      "train_speed(iter/s)": 1.438369
    },
    {
      "epoch": 2.341801979349642,
      "grad_norm": 8.255607604980469,
      "learning_rate": 5.496175751457159e-05,
      "loss": 2.5266948699951173,
      "memory(GiB)": 77.56,
      "step": 54660,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.438341
    },
    {
      "epoch": 2.3420161946788913,
      "grad_norm": 5.314919471740723,
      "learning_rate": 5.495506091469106e-05,
      "loss": 2.6604766845703125,
      "memory(GiB)": 77.56,
      "step": 54665,
      "token_acc": 0.4690909090909091,
      "train_speed(iter/s)": 1.438367
    },
    {
      "epoch": 2.34223041000814,
      "grad_norm": 4.820956707000732,
      "learning_rate": 5.494836422504497e-05,
      "loss": 2.361622619628906,
      "memory(GiB)": 77.56,
      "step": 54670,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.438382
    },
    {
      "epoch": 2.342444625337389,
      "grad_norm": 6.008187294006348,
      "learning_rate": 5.49416674457546e-05,
      "loss": 2.5905376434326173,
      "memory(GiB)": 77.56,
      "step": 54675,
      "token_acc": 0.4479768786127168,
      "train_speed(iter/s)": 1.438377
    },
    {
      "epoch": 2.342658840666638,
      "grad_norm": 5.974514484405518,
      "learning_rate": 5.493497057694129e-05,
      "loss": 2.441915512084961,
      "memory(GiB)": 77.56,
      "step": 54680,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.438399
    },
    {
      "epoch": 2.342873055995887,
      "grad_norm": 5.400164604187012,
      "learning_rate": 5.492827361872637e-05,
      "loss": 2.4609085083007813,
      "memory(GiB)": 77.56,
      "step": 54685,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.438426
    },
    {
      "epoch": 2.343087271325136,
      "grad_norm": 7.050365924835205,
      "learning_rate": 5.492157657123114e-05,
      "loss": 2.0577001571655273,
      "memory(GiB)": 77.56,
      "step": 54690,
      "token_acc": 0.5298013245033113,
      "train_speed(iter/s)": 1.43839
    },
    {
      "epoch": 2.343301486654385,
      "grad_norm": 5.239218235015869,
      "learning_rate": 5.4914879434576936e-05,
      "loss": 2.534061813354492,
      "memory(GiB)": 77.56,
      "step": 54695,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.438404
    },
    {
      "epoch": 2.343515701983634,
      "grad_norm": 6.294853210449219,
      "learning_rate": 5.490818220888508e-05,
      "loss": 2.6373470306396483,
      "memory(GiB)": 77.56,
      "step": 54700,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.438413
    },
    {
      "epoch": 2.3437299173128827,
      "grad_norm": 7.120141983032227,
      "learning_rate": 5.4901484894276914e-05,
      "loss": 2.2757015228271484,
      "memory(GiB)": 77.56,
      "step": 54705,
      "token_acc": 0.5148148148148148,
      "train_speed(iter/s)": 1.43841
    },
    {
      "epoch": 2.343944132642132,
      "grad_norm": 7.0399088859558105,
      "learning_rate": 5.4894787490873746e-05,
      "loss": 2.6073095321655275,
      "memory(GiB)": 77.56,
      "step": 54710,
      "token_acc": 0.43703703703703706,
      "train_speed(iter/s)": 1.438411
    },
    {
      "epoch": 2.344158347971381,
      "grad_norm": 5.254524230957031,
      "learning_rate": 5.48880899987969e-05,
      "loss": 2.7771066665649413,
      "memory(GiB)": 77.56,
      "step": 54715,
      "token_acc": 0.43462897526501765,
      "train_speed(iter/s)": 1.438452
    },
    {
      "epoch": 2.34437256330063,
      "grad_norm": 5.308459758758545,
      "learning_rate": 5.488139241816771e-05,
      "loss": 2.1082611083984375,
      "memory(GiB)": 77.56,
      "step": 54720,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.438438
    },
    {
      "epoch": 2.344586778629879,
      "grad_norm": 4.598628044128418,
      "learning_rate": 5.487469474910754e-05,
      "loss": 2.3201196670532225,
      "memory(GiB)": 77.56,
      "step": 54725,
      "token_acc": 0.4779116465863454,
      "train_speed(iter/s)": 1.438463
    },
    {
      "epoch": 2.3448009939591277,
      "grad_norm": 4.571203708648682,
      "learning_rate": 5.486799699173768e-05,
      "loss": 2.4418289184570314,
      "memory(GiB)": 77.56,
      "step": 54730,
      "token_acc": 0.5101449275362319,
      "train_speed(iter/s)": 1.438482
    },
    {
      "epoch": 2.345015209288377,
      "grad_norm": 4.999051094055176,
      "learning_rate": 5.486129914617951e-05,
      "loss": 2.2475542068481444,
      "memory(GiB)": 77.56,
      "step": 54735,
      "token_acc": 0.5345911949685535,
      "train_speed(iter/s)": 1.438503
    },
    {
      "epoch": 2.3452294246176257,
      "grad_norm": 6.978096008300781,
      "learning_rate": 5.485460121255434e-05,
      "loss": 2.308015823364258,
      "memory(GiB)": 77.56,
      "step": 54740,
      "token_acc": 0.5214285714285715,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 2.3454436399468745,
      "grad_norm": 5.965931415557861,
      "learning_rate": 5.484790319098351e-05,
      "loss": 2.616992378234863,
      "memory(GiB)": 77.56,
      "step": 54745,
      "token_acc": 0.4734848484848485,
      "train_speed(iter/s)": 1.438508
    },
    {
      "epoch": 2.345657855276124,
      "grad_norm": 4.755771160125732,
      "learning_rate": 5.4841205081588354e-05,
      "loss": 2.4398662567138674,
      "memory(GiB)": 77.56,
      "step": 54750,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.438535
    },
    {
      "epoch": 2.3458720706053726,
      "grad_norm": 4.0919928550720215,
      "learning_rate": 5.483450688449024e-05,
      "loss": 2.4105281829833984,
      "memory(GiB)": 77.56,
      "step": 54755,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.438543
    },
    {
      "epoch": 2.3460862859346214,
      "grad_norm": 5.815858840942383,
      "learning_rate": 5.482780859981049e-05,
      "loss": 2.9770795822143556,
      "memory(GiB)": 77.56,
      "step": 54760,
      "token_acc": 0.41002949852507375,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 2.3463005012638707,
      "grad_norm": 4.6366190910339355,
      "learning_rate": 5.482111022767048e-05,
      "loss": 2.343940544128418,
      "memory(GiB)": 77.56,
      "step": 54765,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.438557
    },
    {
      "epoch": 2.3465147165931195,
      "grad_norm": 5.3660759925842285,
      "learning_rate": 5.481441176819152e-05,
      "loss": 2.5082046508789064,
      "memory(GiB)": 77.56,
      "step": 54770,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.438553
    },
    {
      "epoch": 2.3467289319223683,
      "grad_norm": 5.451185703277588,
      "learning_rate": 5.480771322149497e-05,
      "loss": 2.2944231033325195,
      "memory(GiB)": 77.56,
      "step": 54775,
      "token_acc": 0.501628664495114,
      "train_speed(iter/s)": 1.438558
    },
    {
      "epoch": 2.3469431472516176,
      "grad_norm": 6.541532516479492,
      "learning_rate": 5.480101458770221e-05,
      "loss": 2.389505386352539,
      "memory(GiB)": 77.56,
      "step": 54780,
      "token_acc": 0.4481707317073171,
      "train_speed(iter/s)": 1.438567
    },
    {
      "epoch": 2.3471573625808664,
      "grad_norm": 6.510070323944092,
      "learning_rate": 5.479431586693454e-05,
      "loss": 2.7133731842041016,
      "memory(GiB)": 77.56,
      "step": 54785,
      "token_acc": 0.4397163120567376,
      "train_speed(iter/s)": 1.438582
    },
    {
      "epoch": 2.347371577910115,
      "grad_norm": 4.8781914710998535,
      "learning_rate": 5.478761705931334e-05,
      "loss": 2.652548599243164,
      "memory(GiB)": 77.56,
      "step": 54790,
      "token_acc": 0.4309210526315789,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 2.3475857932393644,
      "grad_norm": 5.47180700302124,
      "learning_rate": 5.478091816495997e-05,
      "loss": 2.564466094970703,
      "memory(GiB)": 77.56,
      "step": 54795,
      "token_acc": 0.453125,
      "train_speed(iter/s)": 1.43863
    },
    {
      "epoch": 2.3478000085686133,
      "grad_norm": 5.245362281799316,
      "learning_rate": 5.4774219183995804e-05,
      "loss": 2.5847213745117186,
      "memory(GiB)": 77.56,
      "step": 54800,
      "token_acc": 0.4656084656084656,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 2.348014223897862,
      "grad_norm": 5.536609649658203,
      "learning_rate": 5.476752011654215e-05,
      "loss": 2.6254127502441404,
      "memory(GiB)": 77.56,
      "step": 54805,
      "token_acc": 0.4153225806451613,
      "train_speed(iter/s)": 1.438651
    },
    {
      "epoch": 2.3482284392271113,
      "grad_norm": 5.326805114746094,
      "learning_rate": 5.4760820962720416e-05,
      "loss": 2.354401206970215,
      "memory(GiB)": 77.56,
      "step": 54810,
      "token_acc": 0.4491525423728814,
      "train_speed(iter/s)": 1.438637
    },
    {
      "epoch": 2.34844265455636,
      "grad_norm": 5.349968910217285,
      "learning_rate": 5.475412172265193e-05,
      "loss": 2.3771642684936523,
      "memory(GiB)": 77.56,
      "step": 54815,
      "token_acc": 0.45791245791245794,
      "train_speed(iter/s)": 1.438645
    },
    {
      "epoch": 2.348656869885609,
      "grad_norm": 6.083472728729248,
      "learning_rate": 5.4747422396458085e-05,
      "loss": 2.7385623931884764,
      "memory(GiB)": 77.56,
      "step": 54820,
      "token_acc": 0.4509090909090909,
      "train_speed(iter/s)": 1.438638
    },
    {
      "epoch": 2.348871085214858,
      "grad_norm": 4.6936774253845215,
      "learning_rate": 5.4740722984260216e-05,
      "loss": 2.223306655883789,
      "memory(GiB)": 77.56,
      "step": 54825,
      "token_acc": 0.5431654676258992,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 2.349085300544107,
      "grad_norm": 5.431412696838379,
      "learning_rate": 5.473402348617971e-05,
      "loss": 2.5443208694458006,
      "memory(GiB)": 77.56,
      "step": 54830,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.43864
    },
    {
      "epoch": 2.349299515873356,
      "grad_norm": 6.527918338775635,
      "learning_rate": 5.472732390233792e-05,
      "loss": 2.254547119140625,
      "memory(GiB)": 77.56,
      "step": 54835,
      "token_acc": 0.4778481012658228,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 2.349513731202605,
      "grad_norm": 5.981904983520508,
      "learning_rate": 5.472062423285623e-05,
      "loss": 2.742825508117676,
      "memory(GiB)": 77.56,
      "step": 54840,
      "token_acc": 0.4669260700389105,
      "train_speed(iter/s)": 1.438626
    },
    {
      "epoch": 2.349727946531854,
      "grad_norm": 5.265769004821777,
      "learning_rate": 5.4713924477856006e-05,
      "loss": 2.4999710083007813,
      "memory(GiB)": 77.56,
      "step": 54845,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.438651
    },
    {
      "epoch": 2.3499421618611027,
      "grad_norm": 4.826220512390137,
      "learning_rate": 5.470722463745862e-05,
      "loss": 2.300009536743164,
      "memory(GiB)": 77.56,
      "step": 54850,
      "token_acc": 0.575091575091575,
      "train_speed(iter/s)": 1.438683
    },
    {
      "epoch": 2.350156377190352,
      "grad_norm": 4.447829723358154,
      "learning_rate": 5.4700524711785436e-05,
      "loss": 2.40640869140625,
      "memory(GiB)": 77.56,
      "step": 54855,
      "token_acc": 0.48036253776435045,
      "train_speed(iter/s)": 1.438659
    },
    {
      "epoch": 2.350370592519601,
      "grad_norm": 6.051206588745117,
      "learning_rate": 5.469382470095784e-05,
      "loss": 2.362230682373047,
      "memory(GiB)": 77.56,
      "step": 54860,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.438662
    },
    {
      "epoch": 2.3505848078488496,
      "grad_norm": 5.714086055755615,
      "learning_rate": 5.4687124605097196e-05,
      "loss": 2.525142478942871,
      "memory(GiB)": 77.56,
      "step": 54865,
      "token_acc": 0.4604519774011299,
      "train_speed(iter/s)": 1.438648
    },
    {
      "epoch": 2.350799023178099,
      "grad_norm": 6.7206010818481445,
      "learning_rate": 5.4680424424324914e-05,
      "loss": 2.4359636306762695,
      "memory(GiB)": 77.56,
      "step": 54870,
      "token_acc": 0.46496815286624205,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.3510132385073477,
      "grad_norm": 7.983884334564209,
      "learning_rate": 5.467372415876233e-05,
      "loss": 2.737460708618164,
      "memory(GiB)": 77.56,
      "step": 54875,
      "token_acc": 0.44086021505376344,
      "train_speed(iter/s)": 1.438676
    },
    {
      "epoch": 2.3512274538365965,
      "grad_norm": 5.137965202331543,
      "learning_rate": 5.466702380853087e-05,
      "loss": 2.4100505828857424,
      "memory(GiB)": 77.56,
      "step": 54880,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.438685
    },
    {
      "epoch": 2.3514416691658457,
      "grad_norm": 7.0209808349609375,
      "learning_rate": 5.4660323373751884e-05,
      "loss": 2.6817886352539064,
      "memory(GiB)": 77.56,
      "step": 54885,
      "token_acc": 0.4472573839662447,
      "train_speed(iter/s)": 1.438693
    },
    {
      "epoch": 2.3516558844950946,
      "grad_norm": 6.822183132171631,
      "learning_rate": 5.465362285454677e-05,
      "loss": 2.4985605239868165,
      "memory(GiB)": 77.56,
      "step": 54890,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.438725
    },
    {
      "epoch": 2.3518700998243434,
      "grad_norm": 7.360143184661865,
      "learning_rate": 5.4646922251036924e-05,
      "loss": 2.506234359741211,
      "memory(GiB)": 77.56,
      "step": 54895,
      "token_acc": 0.4767932489451477,
      "train_speed(iter/s)": 1.438694
    },
    {
      "epoch": 2.3520843151535926,
      "grad_norm": 5.007196426391602,
      "learning_rate": 5.464022156334371e-05,
      "loss": 2.3190681457519533,
      "memory(GiB)": 77.56,
      "step": 54900,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.4387
    },
    {
      "epoch": 2.3522985304828414,
      "grad_norm": 3.920491933822632,
      "learning_rate": 5.463352079158852e-05,
      "loss": 2.384654426574707,
      "memory(GiB)": 77.56,
      "step": 54905,
      "token_acc": 0.5132450331125827,
      "train_speed(iter/s)": 1.438695
    },
    {
      "epoch": 2.3525127458120902,
      "grad_norm": 5.728605270385742,
      "learning_rate": 5.462681993589277e-05,
      "loss": 2.4840839385986326,
      "memory(GiB)": 77.56,
      "step": 54910,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.438671
    },
    {
      "epoch": 2.3527269611413395,
      "grad_norm": 5.6435136795043945,
      "learning_rate": 5.462011899637783e-05,
      "loss": 2.40546875,
      "memory(GiB)": 77.56,
      "step": 54915,
      "token_acc": 0.47896440129449835,
      "train_speed(iter/s)": 1.438691
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 6.231410980224609,
      "learning_rate": 5.4613417973165106e-05,
      "loss": 2.3908317565917967,
      "memory(GiB)": 77.56,
      "step": 54920,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.438676
    },
    {
      "epoch": 2.353155391799837,
      "grad_norm": 5.761892795562744,
      "learning_rate": 5.4606716866375985e-05,
      "loss": 2.661233329772949,
      "memory(GiB)": 77.56,
      "step": 54925,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.438696
    },
    {
      "epoch": 2.3533696071290864,
      "grad_norm": 9.214360237121582,
      "learning_rate": 5.460001567613188e-05,
      "loss": 2.558022880554199,
      "memory(GiB)": 77.56,
      "step": 54930,
      "token_acc": 0.49836065573770494,
      "train_speed(iter/s)": 1.438717
    },
    {
      "epoch": 2.353583822458335,
      "grad_norm": 4.837754249572754,
      "learning_rate": 5.459331440255414e-05,
      "loss": 2.489152526855469,
      "memory(GiB)": 77.56,
      "step": 54935,
      "token_acc": 0.4554140127388535,
      "train_speed(iter/s)": 1.438706
    },
    {
      "epoch": 2.353798037787584,
      "grad_norm": 6.567314624786377,
      "learning_rate": 5.458661304576422e-05,
      "loss": 2.268810272216797,
      "memory(GiB)": 77.56,
      "step": 54940,
      "token_acc": 0.5064516129032258,
      "train_speed(iter/s)": 1.438703
    },
    {
      "epoch": 2.3540122531168333,
      "grad_norm": 6.083878040313721,
      "learning_rate": 5.457991160588351e-05,
      "loss": 2.4191219329833986,
      "memory(GiB)": 77.56,
      "step": 54945,
      "token_acc": 0.5060728744939271,
      "train_speed(iter/s)": 1.438708
    },
    {
      "epoch": 2.354226468446082,
      "grad_norm": 7.360652923583984,
      "learning_rate": 5.45732100830334e-05,
      "loss": 2.5673065185546875,
      "memory(GiB)": 77.56,
      "step": 54950,
      "token_acc": 0.45396825396825397,
      "train_speed(iter/s)": 1.438694
    },
    {
      "epoch": 2.354440683775331,
      "grad_norm": 5.104067325592041,
      "learning_rate": 5.4566508477335296e-05,
      "loss": 2.730603790283203,
      "memory(GiB)": 77.56,
      "step": 54955,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.438721
    },
    {
      "epoch": 2.35465489910458,
      "grad_norm": 4.436124801635742,
      "learning_rate": 5.4559806788910606e-05,
      "loss": 2.2991706848144533,
      "memory(GiB)": 77.56,
      "step": 54960,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.438722
    },
    {
      "epoch": 2.354869114433829,
      "grad_norm": 5.377289295196533,
      "learning_rate": 5.455310501788075e-05,
      "loss": 2.3624456405639647,
      "memory(GiB)": 77.56,
      "step": 54965,
      "token_acc": 0.5259259259259259,
      "train_speed(iter/s)": 1.438744
    },
    {
      "epoch": 2.3550833297630778,
      "grad_norm": 4.002988815307617,
      "learning_rate": 5.454640316436712e-05,
      "loss": 2.536465644836426,
      "memory(GiB)": 77.56,
      "step": 54970,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.438719
    },
    {
      "epoch": 2.355297545092327,
      "grad_norm": 5.537210464477539,
      "learning_rate": 5.453970122849113e-05,
      "loss": 2.722956657409668,
      "memory(GiB)": 77.56,
      "step": 54975,
      "token_acc": 0.453416149068323,
      "train_speed(iter/s)": 1.438715
    },
    {
      "epoch": 2.355511760421576,
      "grad_norm": 5.738343238830566,
      "learning_rate": 5.453299921037418e-05,
      "loss": 2.351471710205078,
      "memory(GiB)": 77.56,
      "step": 54980,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.438726
    },
    {
      "epoch": 2.3557259757508247,
      "grad_norm": 5.757815837860107,
      "learning_rate": 5.452629711013773e-05,
      "loss": 2.2772943496704103,
      "memory(GiB)": 77.56,
      "step": 54985,
      "token_acc": 0.504950495049505,
      "train_speed(iter/s)": 1.438709
    },
    {
      "epoch": 2.355940191080074,
      "grad_norm": 5.990734100341797,
      "learning_rate": 5.451959492790313e-05,
      "loss": 2.5603607177734373,
      "memory(GiB)": 77.56,
      "step": 54990,
      "token_acc": 0.4342857142857143,
      "train_speed(iter/s)": 1.438716
    },
    {
      "epoch": 2.3561544064093227,
      "grad_norm": 5.451921463012695,
      "learning_rate": 5.451289266379184e-05,
      "loss": 2.5363142013549806,
      "memory(GiB)": 77.56,
      "step": 54995,
      "token_acc": 0.4295774647887324,
      "train_speed(iter/s)": 1.438737
    },
    {
      "epoch": 2.3563686217385715,
      "grad_norm": 4.374578952789307,
      "learning_rate": 5.4506190317925275e-05,
      "loss": 2.2245977401733397,
      "memory(GiB)": 77.56,
      "step": 55000,
      "token_acc": 0.5340501792114696,
      "train_speed(iter/s)": 1.438748
    },
    {
      "epoch": 2.3563686217385715,
      "eval_loss": 2.2667043209075928,
      "eval_runtime": 13.8579,
      "eval_samples_per_second": 7.216,
      "eval_steps_per_second": 7.216,
      "eval_token_acc": 0.4754335260115607,
      "step": 55000
    },
    {
      "epoch": 2.356582837067821,
      "grad_norm": 6.08438777923584,
      "learning_rate": 5.4499487890424826e-05,
      "loss": 2.456856918334961,
      "memory(GiB)": 77.56,
      "step": 55005,
      "token_acc": 0.47685683530678147,
      "train_speed(iter/s)": 1.438183
    },
    {
      "epoch": 2.3567970523970696,
      "grad_norm": 4.270294666290283,
      "learning_rate": 5.4492785381411936e-05,
      "loss": 2.2999061584472655,
      "memory(GiB)": 77.56,
      "step": 55010,
      "token_acc": 0.5387205387205387,
      "train_speed(iter/s)": 1.438205
    },
    {
      "epoch": 2.3570112677263184,
      "grad_norm": 4.787764549255371,
      "learning_rate": 5.4486082791008044e-05,
      "loss": 2.6687618255615235,
      "memory(GiB)": 77.56,
      "step": 55015,
      "token_acc": 0.47854785478547857,
      "train_speed(iter/s)": 1.438226
    },
    {
      "epoch": 2.3572254830555677,
      "grad_norm": 7.941731929779053,
      "learning_rate": 5.447938011933453e-05,
      "loss": 2.556752014160156,
      "memory(GiB)": 77.56,
      "step": 55020,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.438239
    },
    {
      "epoch": 2.3574396983848165,
      "grad_norm": 4.8594512939453125,
      "learning_rate": 5.447267736651286e-05,
      "loss": 2.569650650024414,
      "memory(GiB)": 77.56,
      "step": 55025,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.438226
    },
    {
      "epoch": 2.3576539137140653,
      "grad_norm": 4.816797733306885,
      "learning_rate": 5.446597453266444e-05,
      "loss": 2.287433052062988,
      "memory(GiB)": 77.56,
      "step": 55030,
      "token_acc": 0.4954682779456193,
      "train_speed(iter/s)": 1.438228
    },
    {
      "epoch": 2.3578681290433146,
      "grad_norm": 6.369913578033447,
      "learning_rate": 5.44592716179107e-05,
      "loss": 2.4946224212646486,
      "memory(GiB)": 77.56,
      "step": 55035,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.438241
    },
    {
      "epoch": 2.3580823443725634,
      "grad_norm": 5.064686298370361,
      "learning_rate": 5.445256862237308e-05,
      "loss": 2.3456184387207033,
      "memory(GiB)": 77.56,
      "step": 55040,
      "token_acc": 0.5278810408921933,
      "train_speed(iter/s)": 1.438264
    },
    {
      "epoch": 2.358296559701812,
      "grad_norm": 5.750959396362305,
      "learning_rate": 5.4445865546172995e-05,
      "loss": 2.4515045166015623,
      "memory(GiB)": 77.56,
      "step": 55045,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.438268
    },
    {
      "epoch": 2.3585107750310614,
      "grad_norm": 6.908394813537598,
      "learning_rate": 5.4439162389431876e-05,
      "loss": 2.460306167602539,
      "memory(GiB)": 77.56,
      "step": 55050,
      "token_acc": 0.49127906976744184,
      "train_speed(iter/s)": 1.438269
    },
    {
      "epoch": 2.3587249903603102,
      "grad_norm": 6.866858959197998,
      "learning_rate": 5.443245915227117e-05,
      "loss": 2.4044116973876952,
      "memory(GiB)": 77.56,
      "step": 55055,
      "token_acc": 0.5171339563862928,
      "train_speed(iter/s)": 1.438286
    },
    {
      "epoch": 2.358939205689559,
      "grad_norm": 5.466826915740967,
      "learning_rate": 5.442575583481232e-05,
      "loss": 2.015051078796387,
      "memory(GiB)": 77.56,
      "step": 55060,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.438327
    },
    {
      "epoch": 2.3591534210188083,
      "grad_norm": 5.861602306365967,
      "learning_rate": 5.441905243717674e-05,
      "loss": 2.225111961364746,
      "memory(GiB)": 77.56,
      "step": 55065,
      "token_acc": 0.49514563106796117,
      "train_speed(iter/s)": 1.438302
    },
    {
      "epoch": 2.359367636348057,
      "grad_norm": 5.2651824951171875,
      "learning_rate": 5.4412348959485894e-05,
      "loss": 2.3751714706420897,
      "memory(GiB)": 77.56,
      "step": 55070,
      "token_acc": 0.4554794520547945,
      "train_speed(iter/s)": 1.438261
    },
    {
      "epoch": 2.359581851677306,
      "grad_norm": 7.595268726348877,
      "learning_rate": 5.4405645401861205e-05,
      "loss": 2.487693977355957,
      "memory(GiB)": 77.56,
      "step": 55075,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438275
    },
    {
      "epoch": 2.359796067006555,
      "grad_norm": 4.716050148010254,
      "learning_rate": 5.439894176442409e-05,
      "loss": 2.337527275085449,
      "memory(GiB)": 77.56,
      "step": 55080,
      "token_acc": 0.4942857142857143,
      "train_speed(iter/s)": 1.438258
    },
    {
      "epoch": 2.360010282335804,
      "grad_norm": 6.211947917938232,
      "learning_rate": 5.439223804729604e-05,
      "loss": 2.431791877746582,
      "memory(GiB)": 77.56,
      "step": 55085,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438287
    },
    {
      "epoch": 2.360224497665053,
      "grad_norm": 5.643962383270264,
      "learning_rate": 5.438553425059848e-05,
      "loss": 2.328583526611328,
      "memory(GiB)": 77.56,
      "step": 55090,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.438296
    },
    {
      "epoch": 2.360438712994302,
      "grad_norm": 4.780154228210449,
      "learning_rate": 5.437883037445284e-05,
      "loss": 2.3667251586914064,
      "memory(GiB)": 77.56,
      "step": 55095,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438292
    },
    {
      "epoch": 2.360652928323551,
      "grad_norm": 5.0685954093933105,
      "learning_rate": 5.43721264189806e-05,
      "loss": 2.3226537704467773,
      "memory(GiB)": 77.56,
      "step": 55100,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.438266
    },
    {
      "epoch": 2.3608671436527997,
      "grad_norm": 4.993478298187256,
      "learning_rate": 5.4365422384303164e-05,
      "loss": 2.565767288208008,
      "memory(GiB)": 77.56,
      "step": 55105,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.438273
    },
    {
      "epoch": 2.361081358982049,
      "grad_norm": 4.918037414550781,
      "learning_rate": 5.435871827054203e-05,
      "loss": 2.5779842376708983,
      "memory(GiB)": 77.56,
      "step": 55110,
      "token_acc": 0.4937888198757764,
      "train_speed(iter/s)": 1.438292
    },
    {
      "epoch": 2.3612955743112978,
      "grad_norm": 5.096170425415039,
      "learning_rate": 5.435201407781863e-05,
      "loss": 2.371634292602539,
      "memory(GiB)": 77.56,
      "step": 55115,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.438309
    },
    {
      "epoch": 2.3615097896405466,
      "grad_norm": 8.549903869628906,
      "learning_rate": 5.434530980625438e-05,
      "loss": 2.2073726654052734,
      "memory(GiB)": 77.56,
      "step": 55120,
      "token_acc": 0.5314685314685315,
      "train_speed(iter/s)": 1.438341
    },
    {
      "epoch": 2.361724004969796,
      "grad_norm": 5.753121852874756,
      "learning_rate": 5.433860545597078e-05,
      "loss": 2.4982673645019533,
      "memory(GiB)": 77.56,
      "step": 55125,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.438352
    },
    {
      "epoch": 2.3619382202990447,
      "grad_norm": 4.9193291664123535,
      "learning_rate": 5.433190102708928e-05,
      "loss": 2.4230201721191404,
      "memory(GiB)": 77.56,
      "step": 55130,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.438359
    },
    {
      "epoch": 2.3621524356282935,
      "grad_norm": 6.801519870758057,
      "learning_rate": 5.432519651973133e-05,
      "loss": 2.428292083740234,
      "memory(GiB)": 77.56,
      "step": 55135,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.43836
    },
    {
      "epoch": 2.3623666509575427,
      "grad_norm": 5.328106880187988,
      "learning_rate": 5.431849193401838e-05,
      "loss": 2.5010623931884766,
      "memory(GiB)": 77.56,
      "step": 55140,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.438357
    },
    {
      "epoch": 2.3625808662867915,
      "grad_norm": 7.334033489227295,
      "learning_rate": 5.431178727007191e-05,
      "loss": 2.5254135131835938,
      "memory(GiB)": 77.56,
      "step": 55145,
      "token_acc": 0.496,
      "train_speed(iter/s)": 1.438365
    },
    {
      "epoch": 2.3627950816160403,
      "grad_norm": 5.021880149841309,
      "learning_rate": 5.430508252801335e-05,
      "loss": 2.393972396850586,
      "memory(GiB)": 77.56,
      "step": 55150,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.438377
    },
    {
      "epoch": 2.3630092969452896,
      "grad_norm": 5.0176239013671875,
      "learning_rate": 5.42983777079642e-05,
      "loss": 2.3331554412841795,
      "memory(GiB)": 77.56,
      "step": 55155,
      "token_acc": 0.4534412955465587,
      "train_speed(iter/s)": 1.438348
    },
    {
      "epoch": 2.3632235122745384,
      "grad_norm": 6.08544397354126,
      "learning_rate": 5.429167281004589e-05,
      "loss": 2.344183921813965,
      "memory(GiB)": 77.56,
      "step": 55160,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438356
    },
    {
      "epoch": 2.3634377276037872,
      "grad_norm": 5.9323296546936035,
      "learning_rate": 5.42849678343799e-05,
      "loss": 2.5026529312133787,
      "memory(GiB)": 77.56,
      "step": 55165,
      "token_acc": 0.4746376811594203,
      "train_speed(iter/s)": 1.438356
    },
    {
      "epoch": 2.3636519429330365,
      "grad_norm": 4.6031060218811035,
      "learning_rate": 5.4278262781087706e-05,
      "loss": 2.4346088409423827,
      "memory(GiB)": 77.56,
      "step": 55170,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.438369
    },
    {
      "epoch": 2.3638661582622853,
      "grad_norm": 4.812386989593506,
      "learning_rate": 5.427155765029078e-05,
      "loss": 2.3574138641357423,
      "memory(GiB)": 77.56,
      "step": 55175,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.438357
    },
    {
      "epoch": 2.364080373591534,
      "grad_norm": 5.376172065734863,
      "learning_rate": 5.426485244211056e-05,
      "loss": 2.628916931152344,
      "memory(GiB)": 77.56,
      "step": 55180,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.438367
    },
    {
      "epoch": 2.3642945889207834,
      "grad_norm": 5.451033115386963,
      "learning_rate": 5.4258147156668545e-05,
      "loss": 2.5946929931640623,
      "memory(GiB)": 77.56,
      "step": 55185,
      "token_acc": 0.4735202492211838,
      "train_speed(iter/s)": 1.43835
    },
    {
      "epoch": 2.364508804250032,
      "grad_norm": 5.982585430145264,
      "learning_rate": 5.425144179408621e-05,
      "loss": 2.4256092071533204,
      "memory(GiB)": 77.56,
      "step": 55190,
      "token_acc": 0.46905537459283386,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 2.364723019579281,
      "grad_norm": 4.28377103805542,
      "learning_rate": 5.4244736354484994e-05,
      "loss": 2.2783056259155274,
      "memory(GiB)": 77.56,
      "step": 55195,
      "token_acc": 0.5299401197604791,
      "train_speed(iter/s)": 1.438349
    },
    {
      "epoch": 2.3649372349085303,
      "grad_norm": 5.632835388183594,
      "learning_rate": 5.423803083798642e-05,
      "loss": 2.5441762924194338,
      "memory(GiB)": 77.56,
      "step": 55200,
      "token_acc": 0.48338368580060426,
      "train_speed(iter/s)": 1.438358
    },
    {
      "epoch": 2.365151450237779,
      "grad_norm": 6.502086162567139,
      "learning_rate": 5.423132524471194e-05,
      "loss": 2.4038652420043944,
      "memory(GiB)": 77.56,
      "step": 55205,
      "token_acc": 0.4742268041237113,
      "train_speed(iter/s)": 1.438379
    },
    {
      "epoch": 2.365365665567028,
      "grad_norm": 4.80830717086792,
      "learning_rate": 5.422461957478302e-05,
      "loss": 2.2743370056152346,
      "memory(GiB)": 77.56,
      "step": 55210,
      "token_acc": 0.504885993485342,
      "train_speed(iter/s)": 1.438402
    },
    {
      "epoch": 2.365579880896277,
      "grad_norm": 4.352596759796143,
      "learning_rate": 5.4217913828321166e-05,
      "loss": 2.4236536026000977,
      "memory(GiB)": 77.56,
      "step": 55215,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.438393
    },
    {
      "epoch": 2.365794096225526,
      "grad_norm": 5.5595502853393555,
      "learning_rate": 5.421120800544785e-05,
      "loss": 2.4924606323242187,
      "memory(GiB)": 77.56,
      "step": 55220,
      "token_acc": 0.4727272727272727,
      "train_speed(iter/s)": 1.438414
    },
    {
      "epoch": 2.3660083115547748,
      "grad_norm": 4.732698917388916,
      "learning_rate": 5.420450210628454e-05,
      "loss": 2.037059783935547,
      "memory(GiB)": 77.56,
      "step": 55225,
      "token_acc": 0.572463768115942,
      "train_speed(iter/s)": 1.438438
    },
    {
      "epoch": 2.366222526884024,
      "grad_norm": 6.975305557250977,
      "learning_rate": 5.4197796130952724e-05,
      "loss": 2.4264499664306642,
      "memory(GiB)": 77.56,
      "step": 55230,
      "token_acc": 0.45555555555555555,
      "train_speed(iter/s)": 1.438466
    },
    {
      "epoch": 2.366436742213273,
      "grad_norm": 4.999119281768799,
      "learning_rate": 5.41910900795739e-05,
      "loss": 2.443581771850586,
      "memory(GiB)": 77.56,
      "step": 55235,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.438462
    },
    {
      "epoch": 2.3666509575425216,
      "grad_norm": 4.922730922698975,
      "learning_rate": 5.418438395226954e-05,
      "loss": 2.6420093536376954,
      "memory(GiB)": 77.56,
      "step": 55240,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.438466
    },
    {
      "epoch": 2.366865172871771,
      "grad_norm": 7.011460304260254,
      "learning_rate": 5.417767774916116e-05,
      "loss": 2.2443168640136717,
      "memory(GiB)": 77.56,
      "step": 55245,
      "token_acc": 0.5202952029520295,
      "train_speed(iter/s)": 1.438482
    },
    {
      "epoch": 2.3670793882010197,
      "grad_norm": 6.412450790405273,
      "learning_rate": 5.417097147037021e-05,
      "loss": 2.1097509384155275,
      "memory(GiB)": 77.56,
      "step": 55250,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.438488
    },
    {
      "epoch": 2.3672936035302685,
      "grad_norm": 7.007317543029785,
      "learning_rate": 5.416426511601822e-05,
      "loss": 2.6930639266967775,
      "memory(GiB)": 77.56,
      "step": 55255,
      "token_acc": 0.43666666666666665,
      "train_speed(iter/s)": 1.43851
    },
    {
      "epoch": 2.3675078188595178,
      "grad_norm": 5.313680171966553,
      "learning_rate": 5.415755868622665e-05,
      "loss": 2.5467241287231444,
      "memory(GiB)": 77.56,
      "step": 55260,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.438526
    },
    {
      "epoch": 2.3677220341887666,
      "grad_norm": 4.119755268096924,
      "learning_rate": 5.4150852181116995e-05,
      "loss": 2.4497085571289063,
      "memory(GiB)": 77.56,
      "step": 55265,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.438537
    },
    {
      "epoch": 2.3679362495180154,
      "grad_norm": 5.886566638946533,
      "learning_rate": 5.414414560081076e-05,
      "loss": 2.458090972900391,
      "memory(GiB)": 77.56,
      "step": 55270,
      "token_acc": 0.5326797385620915,
      "train_speed(iter/s)": 1.438564
    },
    {
      "epoch": 2.3681504648472647,
      "grad_norm": 4.560958385467529,
      "learning_rate": 5.413743894542945e-05,
      "loss": 2.3180269241333007,
      "memory(GiB)": 77.56,
      "step": 55275,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.438563
    },
    {
      "epoch": 2.3683646801765135,
      "grad_norm": 5.445955753326416,
      "learning_rate": 5.413073221509454e-05,
      "loss": 2.568151092529297,
      "memory(GiB)": 77.56,
      "step": 55280,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.438595
    },
    {
      "epoch": 2.3685788955057623,
      "grad_norm": 5.808968544006348,
      "learning_rate": 5.412402540992756e-05,
      "loss": 2.205615997314453,
      "memory(GiB)": 77.56,
      "step": 55285,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.438633
    },
    {
      "epoch": 2.3687931108350115,
      "grad_norm": 14.384493827819824,
      "learning_rate": 5.4117318530049976e-05,
      "loss": 2.3757465362548826,
      "memory(GiB)": 77.56,
      "step": 55290,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 2.3690073261642604,
      "grad_norm": 4.8841023445129395,
      "learning_rate": 5.41106115755833e-05,
      "loss": 2.3218557357788088,
      "memory(GiB)": 77.56,
      "step": 55295,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.438666
    },
    {
      "epoch": 2.369221541493509,
      "grad_norm": 7.7853288650512695,
      "learning_rate": 5.4103904546649063e-05,
      "loss": 2.515058898925781,
      "memory(GiB)": 77.56,
      "step": 55300,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 2.3694357568227584,
      "grad_norm": 4.586111545562744,
      "learning_rate": 5.4097197443368716e-05,
      "loss": 2.370487594604492,
      "memory(GiB)": 77.56,
      "step": 55305,
      "token_acc": 0.5071942446043165,
      "train_speed(iter/s)": 1.438657
    },
    {
      "epoch": 2.3696499721520072,
      "grad_norm": 4.423598766326904,
      "learning_rate": 5.4090490265863814e-05,
      "loss": 2.1970808029174806,
      "memory(GiB)": 77.56,
      "step": 55310,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.438685
    },
    {
      "epoch": 2.369864187481256,
      "grad_norm": 5.399542331695557,
      "learning_rate": 5.408378301425584e-05,
      "loss": 2.234607696533203,
      "memory(GiB)": 77.56,
      "step": 55315,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438681
    },
    {
      "epoch": 2.3700784028105053,
      "grad_norm": 4.355242729187012,
      "learning_rate": 5.40770756886663e-05,
      "loss": 2.564031982421875,
      "memory(GiB)": 77.56,
      "step": 55320,
      "token_acc": 0.4688427299703264,
      "train_speed(iter/s)": 1.438679
    },
    {
      "epoch": 2.370292618139754,
      "grad_norm": 6.180300235748291,
      "learning_rate": 5.407036828921671e-05,
      "loss": 2.691029739379883,
      "memory(GiB)": 77.56,
      "step": 55325,
      "token_acc": 0.47896440129449835,
      "train_speed(iter/s)": 1.438693
    },
    {
      "epoch": 2.370506833469003,
      "grad_norm": 5.43287467956543,
      "learning_rate": 5.406366081602859e-05,
      "loss": 2.409763526916504,
      "memory(GiB)": 77.56,
      "step": 55330,
      "token_acc": 0.4726775956284153,
      "train_speed(iter/s)": 1.438712
    },
    {
      "epoch": 2.370721048798252,
      "grad_norm": 5.2215986251831055,
      "learning_rate": 5.4056953269223445e-05,
      "loss": 2.6180233001708983,
      "memory(GiB)": 77.56,
      "step": 55335,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.438718
    },
    {
      "epoch": 2.370935264127501,
      "grad_norm": 5.596814155578613,
      "learning_rate": 5.405024564892277e-05,
      "loss": 2.3289669036865233,
      "memory(GiB)": 77.56,
      "step": 55340,
      "token_acc": 0.4859437751004016,
      "train_speed(iter/s)": 1.438732
    },
    {
      "epoch": 2.37114947945675,
      "grad_norm": 6.0897932052612305,
      "learning_rate": 5.4043537955248094e-05,
      "loss": 2.46077823638916,
      "memory(GiB)": 77.56,
      "step": 55345,
      "token_acc": 0.5071942446043165,
      "train_speed(iter/s)": 1.438754
    },
    {
      "epoch": 2.371363694785999,
      "grad_norm": 5.113719463348389,
      "learning_rate": 5.403683018832094e-05,
      "loss": 2.41677303314209,
      "memory(GiB)": 77.56,
      "step": 55350,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.438783
    },
    {
      "epoch": 2.371577910115248,
      "grad_norm": 4.7541093826293945,
      "learning_rate": 5.4030122348262824e-05,
      "loss": 2.305705451965332,
      "memory(GiB)": 77.56,
      "step": 55355,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.438799
    },
    {
      "epoch": 2.3717921254444967,
      "grad_norm": 5.01695442199707,
      "learning_rate": 5.402341443519526e-05,
      "loss": 2.7446502685546874,
      "memory(GiB)": 77.56,
      "step": 55360,
      "token_acc": 0.4965034965034965,
      "train_speed(iter/s)": 1.438797
    },
    {
      "epoch": 2.372006340773746,
      "grad_norm": 5.386935710906982,
      "learning_rate": 5.4016706449239764e-05,
      "loss": 2.620396614074707,
      "memory(GiB)": 77.56,
      "step": 55365,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.438796
    },
    {
      "epoch": 2.3722205561029948,
      "grad_norm": 4.951467037200928,
      "learning_rate": 5.400999839051788e-05,
      "loss": 2.551708221435547,
      "memory(GiB)": 77.56,
      "step": 55370,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.438798
    },
    {
      "epoch": 2.3724347714322436,
      "grad_norm": 4.838369846343994,
      "learning_rate": 5.40032902591511e-05,
      "loss": 2.6849252700805666,
      "memory(GiB)": 77.56,
      "step": 55375,
      "token_acc": 0.4360655737704918,
      "train_speed(iter/s)": 1.438819
    },
    {
      "epoch": 2.372648986761493,
      "grad_norm": 5.829607963562012,
      "learning_rate": 5.399658205526096e-05,
      "loss": 2.4711143493652346,
      "memory(GiB)": 77.56,
      "step": 55380,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438829
    },
    {
      "epoch": 2.3728632020907416,
      "grad_norm": 5.011562824249268,
      "learning_rate": 5.398987377896898e-05,
      "loss": 2.1905250549316406,
      "memory(GiB)": 77.56,
      "step": 55385,
      "token_acc": 0.5247148288973384,
      "train_speed(iter/s)": 1.438835
    },
    {
      "epoch": 2.3730774174199905,
      "grad_norm": 7.044679164886475,
      "learning_rate": 5.398316543039671e-05,
      "loss": 2.323820877075195,
      "memory(GiB)": 77.56,
      "step": 55390,
      "token_acc": 0.4805194805194805,
      "train_speed(iter/s)": 1.438845
    },
    {
      "epoch": 2.3732916327492397,
      "grad_norm": 6.043456554412842,
      "learning_rate": 5.397645700966565e-05,
      "loss": 2.5732213973999025,
      "memory(GiB)": 77.56,
      "step": 55395,
      "token_acc": 0.47491638795986624,
      "train_speed(iter/s)": 1.438835
    },
    {
      "epoch": 2.3735058480784885,
      "grad_norm": 6.244682788848877,
      "learning_rate": 5.3969748516897354e-05,
      "loss": 2.6469261169433596,
      "memory(GiB)": 77.56,
      "step": 55400,
      "token_acc": 0.44696969696969696,
      "train_speed(iter/s)": 1.438841
    },
    {
      "epoch": 2.3737200634077373,
      "grad_norm": 6.698449611663818,
      "learning_rate": 5.3963039952213336e-05,
      "loss": 2.2927406311035154,
      "memory(GiB)": 77.56,
      "step": 55405,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.438866
    },
    {
      "epoch": 2.3739342787369866,
      "grad_norm": 6.650759220123291,
      "learning_rate": 5.395633131573512e-05,
      "loss": 2.5409788131713866,
      "memory(GiB)": 77.56,
      "step": 55410,
      "token_acc": 0.48091603053435117,
      "train_speed(iter/s)": 1.438901
    },
    {
      "epoch": 2.3741484940662354,
      "grad_norm": 4.832799911499023,
      "learning_rate": 5.394962260758425e-05,
      "loss": 2.307061767578125,
      "memory(GiB)": 77.56,
      "step": 55415,
      "token_acc": 0.5523465703971119,
      "train_speed(iter/s)": 1.43889
    },
    {
      "epoch": 2.374362709395484,
      "grad_norm": 5.03985071182251,
      "learning_rate": 5.394291382788228e-05,
      "loss": 2.4070295333862304,
      "memory(GiB)": 77.56,
      "step": 55420,
      "token_acc": 0.4966442953020134,
      "train_speed(iter/s)": 1.438913
    },
    {
      "epoch": 2.3745769247247335,
      "grad_norm": 5.742072105407715,
      "learning_rate": 5.39362049767507e-05,
      "loss": 1.8717477798461915,
      "memory(GiB)": 77.56,
      "step": 55425,
      "token_acc": 0.5341880341880342,
      "train_speed(iter/s)": 1.438935
    },
    {
      "epoch": 2.3747911400539823,
      "grad_norm": 5.457836627960205,
      "learning_rate": 5.3929496054311104e-05,
      "loss": 2.23592586517334,
      "memory(GiB)": 77.56,
      "step": 55430,
      "token_acc": 0.48534201954397393,
      "train_speed(iter/s)": 1.438912
    },
    {
      "epoch": 2.375005355383231,
      "grad_norm": 7.373669147491455,
      "learning_rate": 5.3922787060685e-05,
      "loss": 2.3980010986328124,
      "memory(GiB)": 77.56,
      "step": 55435,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.43892
    },
    {
      "epoch": 2.3752195707124804,
      "grad_norm": 6.243663787841797,
      "learning_rate": 5.391607799599391e-05,
      "loss": 2.476382827758789,
      "memory(GiB)": 77.56,
      "step": 55440,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.438956
    },
    {
      "epoch": 2.375433786041729,
      "grad_norm": 4.847306251525879,
      "learning_rate": 5.39093688603594e-05,
      "loss": 2.197188377380371,
      "memory(GiB)": 77.56,
      "step": 55445,
      "token_acc": 0.5409252669039146,
      "train_speed(iter/s)": 1.438991
    },
    {
      "epoch": 2.375648001370978,
      "grad_norm": 6.375215530395508,
      "learning_rate": 5.390265965390301e-05,
      "loss": 2.798021697998047,
      "memory(GiB)": 77.56,
      "step": 55450,
      "token_acc": 0.44785276073619634,
      "train_speed(iter/s)": 1.439021
    },
    {
      "epoch": 2.3758622167002272,
      "grad_norm": 5.868282318115234,
      "learning_rate": 5.3895950376746266e-05,
      "loss": 2.2429452896118165,
      "memory(GiB)": 77.56,
      "step": 55455,
      "token_acc": 0.516320474777448,
      "train_speed(iter/s)": 1.438994
    },
    {
      "epoch": 2.376076432029476,
      "grad_norm": 5.980130672454834,
      "learning_rate": 5.388924102901074e-05,
      "loss": 2.48947811126709,
      "memory(GiB)": 77.56,
      "step": 55460,
      "token_acc": 0.48104956268221577,
      "train_speed(iter/s)": 1.438999
    },
    {
      "epoch": 2.376290647358725,
      "grad_norm": 4.3123931884765625,
      "learning_rate": 5.388253161081795e-05,
      "loss": 2.232387924194336,
      "memory(GiB)": 77.56,
      "step": 55465,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.439035
    },
    {
      "epoch": 2.376504862687974,
      "grad_norm": 4.940186977386475,
      "learning_rate": 5.387582212228948e-05,
      "loss": 2.635325241088867,
      "memory(GiB)": 77.56,
      "step": 55470,
      "token_acc": 0.4561933534743202,
      "train_speed(iter/s)": 1.439051
    },
    {
      "epoch": 2.376719078017223,
      "grad_norm": 5.646570682525635,
      "learning_rate": 5.3869112563546844e-05,
      "loss": 2.4602468490600584,
      "memory(GiB)": 77.56,
      "step": 55475,
      "token_acc": 0.4041095890410959,
      "train_speed(iter/s)": 1.439038
    },
    {
      "epoch": 2.3769332933464717,
      "grad_norm": 4.777437686920166,
      "learning_rate": 5.386240293471161e-05,
      "loss": 2.4900028228759767,
      "memory(GiB)": 77.56,
      "step": 55480,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.439032
    },
    {
      "epoch": 2.377147508675721,
      "grad_norm": 5.969933032989502,
      "learning_rate": 5.385569323590531e-05,
      "loss": 2.389388656616211,
      "memory(GiB)": 77.56,
      "step": 55485,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.439057
    },
    {
      "epoch": 2.37736172400497,
      "grad_norm": 4.918552398681641,
      "learning_rate": 5.3848983467249516e-05,
      "loss": 2.6371192932128906,
      "memory(GiB)": 77.56,
      "step": 55490,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.439046
    },
    {
      "epoch": 2.3775759393342186,
      "grad_norm": 4.467573642730713,
      "learning_rate": 5.384227362886578e-05,
      "loss": 2.3733142852783202,
      "memory(GiB)": 77.56,
      "step": 55495,
      "token_acc": 0.4921135646687697,
      "train_speed(iter/s)": 1.439046
    },
    {
      "epoch": 2.377790154663468,
      "grad_norm": 6.7481489181518555,
      "learning_rate": 5.3835563720875645e-05,
      "loss": 2.632244110107422,
      "memory(GiB)": 77.56,
      "step": 55500,
      "token_acc": 0.45394736842105265,
      "train_speed(iter/s)": 1.439057
    },
    {
      "epoch": 2.377790154663468,
      "eval_loss": 2.0676262378692627,
      "eval_runtime": 14.3769,
      "eval_samples_per_second": 6.956,
      "eval_steps_per_second": 6.956,
      "eval_token_acc": 0.494550408719346,
      "step": 55500
    },
    {
      "epoch": 2.3780043699927167,
      "grad_norm": 7.466429710388184,
      "learning_rate": 5.382885374340069e-05,
      "loss": 2.413881301879883,
      "memory(GiB)": 77.56,
      "step": 55505,
      "token_acc": 0.49607843137254903,
      "train_speed(iter/s)": 1.438486
    },
    {
      "epoch": 2.3782185853219655,
      "grad_norm": 5.201564788818359,
      "learning_rate": 5.382214369656244e-05,
      "loss": 2.23253173828125,
      "memory(GiB)": 77.56,
      "step": 55510,
      "token_acc": 0.5125448028673835,
      "train_speed(iter/s)": 1.438483
    },
    {
      "epoch": 2.3784328006512148,
      "grad_norm": 5.7796244621276855,
      "learning_rate": 5.381543358048249e-05,
      "loss": 2.617478370666504,
      "memory(GiB)": 77.56,
      "step": 55515,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.438457
    },
    {
      "epoch": 2.3786470159804636,
      "grad_norm": 9.110394477844238,
      "learning_rate": 5.380872339528237e-05,
      "loss": 2.659603500366211,
      "memory(GiB)": 77.56,
      "step": 55520,
      "token_acc": 0.4059040590405904,
      "train_speed(iter/s)": 1.438438
    },
    {
      "epoch": 2.3788612313097124,
      "grad_norm": 6.04946231842041,
      "learning_rate": 5.380201314108365e-05,
      "loss": 2.3853595733642576,
      "memory(GiB)": 77.56,
      "step": 55525,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.438452
    },
    {
      "epoch": 2.3790754466389616,
      "grad_norm": 5.455751895904541,
      "learning_rate": 5.3795302818007895e-05,
      "loss": 2.3142032623291016,
      "memory(GiB)": 77.56,
      "step": 55530,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.438438
    },
    {
      "epoch": 2.3792896619682105,
      "grad_norm": 5.232691287994385,
      "learning_rate": 5.378859242617668e-05,
      "loss": 2.18917293548584,
      "memory(GiB)": 77.56,
      "step": 55535,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.438436
    },
    {
      "epoch": 2.3795038772974593,
      "grad_norm": 5.232368469238281,
      "learning_rate": 5.378188196571154e-05,
      "loss": 2.3399673461914063,
      "memory(GiB)": 77.56,
      "step": 55540,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.43839
    },
    {
      "epoch": 2.3797180926267085,
      "grad_norm": 6.277401447296143,
      "learning_rate": 5.3775171436734084e-05,
      "loss": 2.7239477157592775,
      "memory(GiB)": 77.56,
      "step": 55545,
      "token_acc": 0.4522058823529412,
      "train_speed(iter/s)": 1.438418
    },
    {
      "epoch": 2.3799323079559573,
      "grad_norm": 6.038576602935791,
      "learning_rate": 5.376846083936585e-05,
      "loss": 2.38717098236084,
      "memory(GiB)": 77.56,
      "step": 55550,
      "token_acc": 0.46381578947368424,
      "train_speed(iter/s)": 1.43845
    },
    {
      "epoch": 2.380146523285206,
      "grad_norm": 7.51558780670166,
      "learning_rate": 5.376175017372841e-05,
      "loss": 2.617052459716797,
      "memory(GiB)": 77.56,
      "step": 55555,
      "token_acc": 0.50390625,
      "train_speed(iter/s)": 1.438469
    },
    {
      "epoch": 2.3803607386144554,
      "grad_norm": 4.218362808227539,
      "learning_rate": 5.375503943994333e-05,
      "loss": 2.5389705657958985,
      "memory(GiB)": 77.56,
      "step": 55560,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.438487
    },
    {
      "epoch": 2.380574953943704,
      "grad_norm": 5.294258117675781,
      "learning_rate": 5.37483286381322e-05,
      "loss": 2.4298730850219727,
      "memory(GiB)": 77.56,
      "step": 55565,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.438507
    },
    {
      "epoch": 2.380789169272953,
      "grad_norm": 6.186878681182861,
      "learning_rate": 5.374161776841656e-05,
      "loss": 2.632823944091797,
      "memory(GiB)": 77.56,
      "step": 55570,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.438528
    },
    {
      "epoch": 2.3810033846022023,
      "grad_norm": 4.771598815917969,
      "learning_rate": 5.373490683091802e-05,
      "loss": 2.436178970336914,
      "memory(GiB)": 77.56,
      "step": 55575,
      "token_acc": 0.5047021943573667,
      "train_speed(iter/s)": 1.438491
    },
    {
      "epoch": 2.381217599931451,
      "grad_norm": 6.250410079956055,
      "learning_rate": 5.372819582575814e-05,
      "loss": 2.494367790222168,
      "memory(GiB)": 77.56,
      "step": 55580,
      "token_acc": 0.4801587301587302,
      "train_speed(iter/s)": 1.438494
    },
    {
      "epoch": 2.3814318152607,
      "grad_norm": 5.710798263549805,
      "learning_rate": 5.372148475305849e-05,
      "loss": 2.6136795043945313,
      "memory(GiB)": 77.56,
      "step": 55585,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.438477
    },
    {
      "epoch": 2.381646030589949,
      "grad_norm": 3.6559946537017822,
      "learning_rate": 5.371477361294066e-05,
      "loss": 2.548026466369629,
      "memory(GiB)": 77.56,
      "step": 55590,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.438491
    },
    {
      "epoch": 2.381860245919198,
      "grad_norm": 6.05740213394165,
      "learning_rate": 5.370806240552623e-05,
      "loss": 2.6580345153808596,
      "memory(GiB)": 77.56,
      "step": 55595,
      "token_acc": 0.47717842323651455,
      "train_speed(iter/s)": 1.438514
    },
    {
      "epoch": 2.382074461248447,
      "grad_norm": 5.075231552124023,
      "learning_rate": 5.370135113093674e-05,
      "loss": 2.2091745376586913,
      "memory(GiB)": 77.56,
      "step": 55600,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 2.382288676577696,
      "grad_norm": 5.125804901123047,
      "learning_rate": 5.369463978929382e-05,
      "loss": 2.198167610168457,
      "memory(GiB)": 77.56,
      "step": 55605,
      "token_acc": 0.5204081632653061,
      "train_speed(iter/s)": 1.438528
    },
    {
      "epoch": 2.382502891906945,
      "grad_norm": 5.294960021972656,
      "learning_rate": 5.3687928380719044e-05,
      "loss": 2.4617355346679686,
      "memory(GiB)": 77.56,
      "step": 55610,
      "token_acc": 0.5060728744939271,
      "train_speed(iter/s)": 1.438543
    },
    {
      "epoch": 2.3827171072361937,
      "grad_norm": 6.396987438201904,
      "learning_rate": 5.368121690533396e-05,
      "loss": 2.5283481597900392,
      "memory(GiB)": 77.56,
      "step": 55615,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.438547
    },
    {
      "epoch": 2.382931322565443,
      "grad_norm": 6.050921440124512,
      "learning_rate": 5.3674505363260206e-05,
      "loss": 2.4809059143066405,
      "memory(GiB)": 77.56,
      "step": 55620,
      "token_acc": 0.5030864197530864,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.3831455378946917,
      "grad_norm": 5.0938849449157715,
      "learning_rate": 5.366779375461933e-05,
      "loss": 2.4185562133789062,
      "memory(GiB)": 77.56,
      "step": 55625,
      "token_acc": 0.4697986577181208,
      "train_speed(iter/s)": 1.438562
    },
    {
      "epoch": 2.3833597532239406,
      "grad_norm": 5.465989589691162,
      "learning_rate": 5.366108207953293e-05,
      "loss": 2.583892059326172,
      "memory(GiB)": 77.56,
      "step": 55630,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.438608
    },
    {
      "epoch": 2.38357396855319,
      "grad_norm": 5.359743118286133,
      "learning_rate": 5.365437033812259e-05,
      "loss": 2.3713050842285157,
      "memory(GiB)": 77.56,
      "step": 55635,
      "token_acc": 0.5204918032786885,
      "train_speed(iter/s)": 1.438613
    },
    {
      "epoch": 2.3837881838824386,
      "grad_norm": 6.497318744659424,
      "learning_rate": 5.3647658530509904e-05,
      "loss": 2.702520179748535,
      "memory(GiB)": 77.56,
      "step": 55640,
      "token_acc": 0.44761904761904764,
      "train_speed(iter/s)": 1.438628
    },
    {
      "epoch": 2.3840023992116874,
      "grad_norm": 4.790289402008057,
      "learning_rate": 5.364094665681646e-05,
      "loss": 2.127778244018555,
      "memory(GiB)": 77.56,
      "step": 55645,
      "token_acc": 0.5445205479452054,
      "train_speed(iter/s)": 1.438655
    },
    {
      "epoch": 2.3842166145409367,
      "grad_norm": 4.052952766418457,
      "learning_rate": 5.363423471716386e-05,
      "loss": 2.3164926528930665,
      "memory(GiB)": 77.56,
      "step": 55650,
      "token_acc": 0.4819277108433735,
      "train_speed(iter/s)": 1.438653
    },
    {
      "epoch": 2.3844308298701855,
      "grad_norm": 4.31178092956543,
      "learning_rate": 5.3627522711673675e-05,
      "loss": 2.3609615325927735,
      "memory(GiB)": 77.56,
      "step": 55655,
      "token_acc": 0.5062893081761006,
      "train_speed(iter/s)": 1.438659
    },
    {
      "epoch": 2.3846450451994343,
      "grad_norm": 3.9674179553985596,
      "learning_rate": 5.362081064046753e-05,
      "loss": 1.916794776916504,
      "memory(GiB)": 77.56,
      "step": 55660,
      "token_acc": 0.5886524822695035,
      "train_speed(iter/s)": 1.438643
    },
    {
      "epoch": 2.3848592605286836,
      "grad_norm": 7.026403903961182,
      "learning_rate": 5.3614098503667e-05,
      "loss": 2.5660417556762694,
      "memory(GiB)": 77.56,
      "step": 55665,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.438662
    },
    {
      "epoch": 2.3850734758579324,
      "grad_norm": 6.199244976043701,
      "learning_rate": 5.360738630139368e-05,
      "loss": 2.320438194274902,
      "memory(GiB)": 77.56,
      "step": 55670,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.438667
    },
    {
      "epoch": 2.385287691187181,
      "grad_norm": 5.9237165451049805,
      "learning_rate": 5.360067403376916e-05,
      "loss": 2.6910205841064454,
      "memory(GiB)": 77.56,
      "step": 55675,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.438664
    },
    {
      "epoch": 2.3855019065164305,
      "grad_norm": 4.771265983581543,
      "learning_rate": 5.359396170091508e-05,
      "loss": 2.4336774826049803,
      "memory(GiB)": 77.56,
      "step": 55680,
      "token_acc": 0.4619718309859155,
      "train_speed(iter/s)": 1.438654
    },
    {
      "epoch": 2.3857161218456793,
      "grad_norm": 7.338444232940674,
      "learning_rate": 5.358724930295299e-05,
      "loss": 2.59556884765625,
      "memory(GiB)": 77.56,
      "step": 55685,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.438656
    },
    {
      "epoch": 2.385930337174928,
      "grad_norm": 4.878612995147705,
      "learning_rate": 5.3580536840004524e-05,
      "loss": 2.4306787490844726,
      "memory(GiB)": 77.56,
      "step": 55690,
      "token_acc": 0.4967532467532468,
      "train_speed(iter/s)": 1.438688
    },
    {
      "epoch": 2.3861445525041773,
      "grad_norm": 5.07237434387207,
      "learning_rate": 5.3573824312191276e-05,
      "loss": 2.467372512817383,
      "memory(GiB)": 77.56,
      "step": 55695,
      "token_acc": 0.5132450331125827,
      "train_speed(iter/s)": 1.438706
    },
    {
      "epoch": 2.386358767833426,
      "grad_norm": 6.616938591003418,
      "learning_rate": 5.356711171963484e-05,
      "loss": 2.234323501586914,
      "memory(GiB)": 77.56,
      "step": 55700,
      "token_acc": 0.510548523206751,
      "train_speed(iter/s)": 1.438734
    },
    {
      "epoch": 2.386572983162675,
      "grad_norm": 5.262768268585205,
      "learning_rate": 5.356039906245684e-05,
      "loss": 2.4600982666015625,
      "memory(GiB)": 77.56,
      "step": 55705,
      "token_acc": 0.4895833333333333,
      "train_speed(iter/s)": 1.438742
    },
    {
      "epoch": 2.3867871984919242,
      "grad_norm": 4.559207439422607,
      "learning_rate": 5.3553686340778875e-05,
      "loss": 2.655509185791016,
      "memory(GiB)": 77.56,
      "step": 55710,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.438703
    },
    {
      "epoch": 2.387001413821173,
      "grad_norm": 6.9952921867370605,
      "learning_rate": 5.354697355472252e-05,
      "loss": 2.460745620727539,
      "memory(GiB)": 77.56,
      "step": 55715,
      "token_acc": 0.49049429657794674,
      "train_speed(iter/s)": 1.43873
    },
    {
      "epoch": 2.387215629150422,
      "grad_norm": 7.227347373962402,
      "learning_rate": 5.354026070440944e-05,
      "loss": 2.2894195556640624,
      "memory(GiB)": 77.56,
      "step": 55720,
      "token_acc": 0.49050632911392406,
      "train_speed(iter/s)": 1.438724
    },
    {
      "epoch": 2.387429844479671,
      "grad_norm": 5.833535671234131,
      "learning_rate": 5.35335477899612e-05,
      "loss": 2.538958740234375,
      "memory(GiB)": 77.56,
      "step": 55725,
      "token_acc": 0.45936395759717313,
      "train_speed(iter/s)": 1.438756
    },
    {
      "epoch": 2.38764405980892,
      "grad_norm": 5.033865451812744,
      "learning_rate": 5.352683481149944e-05,
      "loss": 2.5982614517211915,
      "memory(GiB)": 77.56,
      "step": 55730,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.438768
    },
    {
      "epoch": 2.3878582751381687,
      "grad_norm": 6.980336666107178,
      "learning_rate": 5.352012176914575e-05,
      "loss": 2.2374475479125975,
      "memory(GiB)": 77.56,
      "step": 55735,
      "token_acc": 0.5095785440613027,
      "train_speed(iter/s)": 1.438764
    },
    {
      "epoch": 2.388072490467418,
      "grad_norm": 6.426371097564697,
      "learning_rate": 5.351340866302176e-05,
      "loss": 2.369184684753418,
      "memory(GiB)": 77.56,
      "step": 55740,
      "token_acc": 0.5177865612648221,
      "train_speed(iter/s)": 1.438772
    },
    {
      "epoch": 2.388286705796667,
      "grad_norm": 6.667300701141357,
      "learning_rate": 5.350669549324907e-05,
      "loss": 2.8681800842285154,
      "memory(GiB)": 77.56,
      "step": 55745,
      "token_acc": 0.4470198675496689,
      "train_speed(iter/s)": 1.438773
    },
    {
      "epoch": 2.3885009211259156,
      "grad_norm": 5.618614196777344,
      "learning_rate": 5.3499982259949285e-05,
      "loss": 2.390421676635742,
      "memory(GiB)": 77.56,
      "step": 55750,
      "token_acc": 0.49224806201550386,
      "train_speed(iter/s)": 1.438778
    },
    {
      "epoch": 2.388715136455165,
      "grad_norm": 6.456734657287598,
      "learning_rate": 5.349326896324407e-05,
      "loss": 2.3189796447753905,
      "memory(GiB)": 77.56,
      "step": 55755,
      "token_acc": 0.4979253112033195,
      "train_speed(iter/s)": 1.43878
    },
    {
      "epoch": 2.3889293517844137,
      "grad_norm": 4.946966171264648,
      "learning_rate": 5.348655560325498e-05,
      "loss": 2.4424358367919923,
      "memory(GiB)": 77.56,
      "step": 55760,
      "token_acc": 0.504,
      "train_speed(iter/s)": 1.438792
    },
    {
      "epoch": 2.3891435671136625,
      "grad_norm": 4.634521007537842,
      "learning_rate": 5.347984218010369e-05,
      "loss": 2.392115020751953,
      "memory(GiB)": 77.56,
      "step": 55765,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.438803
    },
    {
      "epoch": 2.3893577824429117,
      "grad_norm": 6.2234907150268555,
      "learning_rate": 5.347312869391179e-05,
      "loss": 2.5383182525634767,
      "memory(GiB)": 77.56,
      "step": 55770,
      "token_acc": 0.46283783783783783,
      "train_speed(iter/s)": 1.438817
    },
    {
      "epoch": 2.3895719977721606,
      "grad_norm": 5.292355537414551,
      "learning_rate": 5.3466415144800893e-05,
      "loss": 2.629548263549805,
      "memory(GiB)": 77.56,
      "step": 55775,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.438817
    },
    {
      "epoch": 2.3897862131014094,
      "grad_norm": 5.502911567687988,
      "learning_rate": 5.345970153289266e-05,
      "loss": 2.5506542205810545,
      "memory(GiB)": 77.56,
      "step": 55780,
      "token_acc": 0.4623287671232877,
      "train_speed(iter/s)": 1.438837
    },
    {
      "epoch": 2.3900004284306586,
      "grad_norm": 5.6977949142456055,
      "learning_rate": 5.345298785830866e-05,
      "loss": 2.372870445251465,
      "memory(GiB)": 77.56,
      "step": 55785,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.438847
    },
    {
      "epoch": 2.3902146437599074,
      "grad_norm": 5.290041446685791,
      "learning_rate": 5.344627412117055e-05,
      "loss": 2.2906618118286133,
      "memory(GiB)": 77.56,
      "step": 55790,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.43887
    },
    {
      "epoch": 2.3904288590891563,
      "grad_norm": 5.064677715301514,
      "learning_rate": 5.343956032159996e-05,
      "loss": 2.596945381164551,
      "memory(GiB)": 77.56,
      "step": 55795,
      "token_acc": 0.486404833836858,
      "train_speed(iter/s)": 1.438899
    },
    {
      "epoch": 2.3906430744184055,
      "grad_norm": 6.323420524597168,
      "learning_rate": 5.34328464597185e-05,
      "loss": 2.092578887939453,
      "memory(GiB)": 77.56,
      "step": 55800,
      "token_acc": 0.5461254612546126,
      "train_speed(iter/s)": 1.438909
    },
    {
      "epoch": 2.3908572897476543,
      "grad_norm": 7.094272136688232,
      "learning_rate": 5.342613253564782e-05,
      "loss": 2.310749816894531,
      "memory(GiB)": 77.56,
      "step": 55805,
      "token_acc": 0.4831804281345566,
      "train_speed(iter/s)": 1.43892
    },
    {
      "epoch": 2.391071505076903,
      "grad_norm": 5.3603129386901855,
      "learning_rate": 5.341941854950952e-05,
      "loss": 2.3845823287963865,
      "memory(GiB)": 77.56,
      "step": 55810,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.438937
    },
    {
      "epoch": 2.3912857204061524,
      "grad_norm": 4.8593363761901855,
      "learning_rate": 5.341270450142526e-05,
      "loss": 2.562577819824219,
      "memory(GiB)": 77.56,
      "step": 55815,
      "token_acc": 0.51171875,
      "train_speed(iter/s)": 1.438956
    },
    {
      "epoch": 2.391499935735401,
      "grad_norm": 7.277611255645752,
      "learning_rate": 5.340599039151665e-05,
      "loss": 2.3729629516601562,
      "memory(GiB)": 77.56,
      "step": 55820,
      "token_acc": 0.5043103448275862,
      "train_speed(iter/s)": 1.438967
    },
    {
      "epoch": 2.39171415106465,
      "grad_norm": 5.727371692657471,
      "learning_rate": 5.339927621990533e-05,
      "loss": 2.5254974365234375,
      "memory(GiB)": 77.56,
      "step": 55825,
      "token_acc": 0.5067114093959731,
      "train_speed(iter/s)": 1.43898
    },
    {
      "epoch": 2.3919283663938993,
      "grad_norm": 6.238166809082031,
      "learning_rate": 5.3392561986712916e-05,
      "loss": 2.342428970336914,
      "memory(GiB)": 77.56,
      "step": 55830,
      "token_acc": 0.4835820895522388,
      "train_speed(iter/s)": 1.439015
    },
    {
      "epoch": 2.392142581723148,
      "grad_norm": 6.191690444946289,
      "learning_rate": 5.3385847692061075e-05,
      "loss": 2.622299385070801,
      "memory(GiB)": 77.56,
      "step": 55835,
      "token_acc": 0.41924398625429554,
      "train_speed(iter/s)": 1.439024
    },
    {
      "epoch": 2.392356797052397,
      "grad_norm": 7.068532943725586,
      "learning_rate": 5.337913333607143e-05,
      "loss": 2.555002784729004,
      "memory(GiB)": 77.56,
      "step": 55840,
      "token_acc": 0.44765342960288806,
      "train_speed(iter/s)": 1.439054
    },
    {
      "epoch": 2.392571012381646,
      "grad_norm": 5.620985984802246,
      "learning_rate": 5.337241891886561e-05,
      "loss": 2.413443946838379,
      "memory(GiB)": 77.56,
      "step": 55845,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.439062
    },
    {
      "epoch": 2.392785227710895,
      "grad_norm": 6.190371036529541,
      "learning_rate": 5.3365704440565255e-05,
      "loss": 2.46199893951416,
      "memory(GiB)": 77.56,
      "step": 55850,
      "token_acc": 0.45806451612903226,
      "train_speed(iter/s)": 1.439094
    },
    {
      "epoch": 2.392999443040144,
      "grad_norm": 5.548079490661621,
      "learning_rate": 5.335898990129202e-05,
      "loss": 2.48559684753418,
      "memory(GiB)": 77.56,
      "step": 55855,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.439113
    },
    {
      "epoch": 2.393213658369393,
      "grad_norm": 5.645583152770996,
      "learning_rate": 5.335227530116751e-05,
      "loss": 2.3503229141235353,
      "memory(GiB)": 77.56,
      "step": 55860,
      "token_acc": 0.5142045454545454,
      "train_speed(iter/s)": 1.439107
    },
    {
      "epoch": 2.393427873698642,
      "grad_norm": 5.426797389984131,
      "learning_rate": 5.3345560640313395e-05,
      "loss": 2.3297786712646484,
      "memory(GiB)": 77.56,
      "step": 55865,
      "token_acc": 0.4858490566037736,
      "train_speed(iter/s)": 1.439136
    },
    {
      "epoch": 2.3936420890278907,
      "grad_norm": 6.183347225189209,
      "learning_rate": 5.333884591885132e-05,
      "loss": 2.2710235595703123,
      "memory(GiB)": 77.56,
      "step": 55870,
      "token_acc": 0.5207667731629393,
      "train_speed(iter/s)": 1.439166
    },
    {
      "epoch": 2.39385630435714,
      "grad_norm": 5.147744655609131,
      "learning_rate": 5.333213113690291e-05,
      "loss": 2.2079469680786135,
      "memory(GiB)": 77.56,
      "step": 55875,
      "token_acc": 0.551094890510949,
      "train_speed(iter/s)": 1.439158
    },
    {
      "epoch": 2.3940705196863887,
      "grad_norm": 4.292358875274658,
      "learning_rate": 5.3325416294589826e-05,
      "loss": 2.3044912338256838,
      "memory(GiB)": 77.56,
      "step": 55880,
      "token_acc": 0.48427672955974843,
      "train_speed(iter/s)": 1.439178
    },
    {
      "epoch": 2.3942847350156375,
      "grad_norm": 6.845787048339844,
      "learning_rate": 5.331870139203371e-05,
      "loss": 2.157851791381836,
      "memory(GiB)": 77.56,
      "step": 55885,
      "token_acc": 0.5114754098360655,
      "train_speed(iter/s)": 1.439184
    },
    {
      "epoch": 2.394498950344887,
      "grad_norm": 5.979043006896973,
      "learning_rate": 5.33119864293562e-05,
      "loss": 2.2227333068847654,
      "memory(GiB)": 77.56,
      "step": 55890,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.439191
    },
    {
      "epoch": 2.3947131656741356,
      "grad_norm": 5.1875762939453125,
      "learning_rate": 5.3305271406678936e-05,
      "loss": 2.6128095626831054,
      "memory(GiB)": 77.56,
      "step": 55895,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.439211
    },
    {
      "epoch": 2.3949273810033844,
      "grad_norm": 4.127660751342773,
      "learning_rate": 5.329855632412359e-05,
      "loss": 2.5641387939453124,
      "memory(GiB)": 77.56,
      "step": 55900,
      "token_acc": 0.46973365617433416,
      "train_speed(iter/s)": 1.439206
    },
    {
      "epoch": 2.3951415963326337,
      "grad_norm": 11.877144813537598,
      "learning_rate": 5.329184118181181e-05,
      "loss": 2.5081594467163084,
      "memory(GiB)": 77.56,
      "step": 55905,
      "token_acc": 0.48328267477203646,
      "train_speed(iter/s)": 1.439213
    },
    {
      "epoch": 2.3953558116618825,
      "grad_norm": 5.12606954574585,
      "learning_rate": 5.3285125979865234e-05,
      "loss": 2.336347961425781,
      "memory(GiB)": 77.56,
      "step": 55910,
      "token_acc": 0.49377593360995853,
      "train_speed(iter/s)": 1.439221
    },
    {
      "epoch": 2.3955700269911313,
      "grad_norm": 5.319666385650635,
      "learning_rate": 5.3278410718405514e-05,
      "loss": 2.23387451171875,
      "memory(GiB)": 77.56,
      "step": 55915,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.439219
    },
    {
      "epoch": 2.3957842423203806,
      "grad_norm": 5.634885787963867,
      "learning_rate": 5.327169539755431e-05,
      "loss": 2.4125213623046875,
      "memory(GiB)": 77.56,
      "step": 55920,
      "token_acc": 0.5040322580645161,
      "train_speed(iter/s)": 1.439244
    },
    {
      "epoch": 2.3959984576496294,
      "grad_norm": 6.618666648864746,
      "learning_rate": 5.3264980017433284e-05,
      "loss": 2.662358856201172,
      "memory(GiB)": 77.56,
      "step": 55925,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.439276
    },
    {
      "epoch": 2.396212672978878,
      "grad_norm": 5.333796501159668,
      "learning_rate": 5.325826457816406e-05,
      "loss": 2.3391931533813475,
      "memory(GiB)": 77.56,
      "step": 55930,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.439303
    },
    {
      "epoch": 2.3964268883081274,
      "grad_norm": 4.858936309814453,
      "learning_rate": 5.325154907986833e-05,
      "loss": 2.212980270385742,
      "memory(GiB)": 77.56,
      "step": 55935,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.43932
    },
    {
      "epoch": 2.3966411036373763,
      "grad_norm": 4.452764511108398,
      "learning_rate": 5.324483352266775e-05,
      "loss": 2.5510013580322264,
      "memory(GiB)": 77.56,
      "step": 55940,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.439325
    },
    {
      "epoch": 2.396855318966625,
      "grad_norm": 5.3482584953308105,
      "learning_rate": 5.323811790668395e-05,
      "loss": 2.4425991058349608,
      "memory(GiB)": 77.56,
      "step": 55945,
      "token_acc": 0.458041958041958,
      "train_speed(iter/s)": 1.439319
    },
    {
      "epoch": 2.3970695342958743,
      "grad_norm": 4.145442962646484,
      "learning_rate": 5.323140223203862e-05,
      "loss": 2.3750329971313477,
      "memory(GiB)": 77.56,
      "step": 55950,
      "token_acc": 0.4868804664723032,
      "train_speed(iter/s)": 1.439342
    },
    {
      "epoch": 2.397283749625123,
      "grad_norm": 8.035231590270996,
      "learning_rate": 5.322468649885341e-05,
      "loss": 2.761701965332031,
      "memory(GiB)": 77.56,
      "step": 55955,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.439361
    },
    {
      "epoch": 2.397497964954372,
      "grad_norm": 7.0439910888671875,
      "learning_rate": 5.3217970707249955e-05,
      "loss": 2.406748580932617,
      "memory(GiB)": 77.56,
      "step": 55960,
      "token_acc": 0.43670886075949367,
      "train_speed(iter/s)": 1.439367
    },
    {
      "epoch": 2.397712180283621,
      "grad_norm": 5.987401485443115,
      "learning_rate": 5.3211254857349955e-05,
      "loss": 2.289975166320801,
      "memory(GiB)": 77.56,
      "step": 55965,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439374
    },
    {
      "epoch": 2.39792639561287,
      "grad_norm": 5.4337849617004395,
      "learning_rate": 5.320453894927506e-05,
      "loss": 2.20496826171875,
      "memory(GiB)": 77.56,
      "step": 55970,
      "token_acc": 0.5464684014869888,
      "train_speed(iter/s)": 1.439356
    },
    {
      "epoch": 2.398140610942119,
      "grad_norm": 7.567423343658447,
      "learning_rate": 5.319782298314694e-05,
      "loss": 2.1844202041625977,
      "memory(GiB)": 77.56,
      "step": 55975,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.439339
    },
    {
      "epoch": 2.398354826271368,
      "grad_norm": 5.335397243499756,
      "learning_rate": 5.319110695908726e-05,
      "loss": 2.394648551940918,
      "memory(GiB)": 77.56,
      "step": 55980,
      "token_acc": 0.48830409356725146,
      "train_speed(iter/s)": 1.439353
    },
    {
      "epoch": 2.398569041600617,
      "grad_norm": 4.841787815093994,
      "learning_rate": 5.318439087721768e-05,
      "loss": 2.3667648315429686,
      "memory(GiB)": 77.56,
      "step": 55985,
      "token_acc": 0.49700598802395207,
      "train_speed(iter/s)": 1.439369
    },
    {
      "epoch": 2.3987832569298657,
      "grad_norm": 6.294712066650391,
      "learning_rate": 5.3177674737659865e-05,
      "loss": 2.507879638671875,
      "memory(GiB)": 77.56,
      "step": 55990,
      "token_acc": 0.46258503401360546,
      "train_speed(iter/s)": 1.439357
    },
    {
      "epoch": 2.398997472259115,
      "grad_norm": 6.565338611602783,
      "learning_rate": 5.31709585405355e-05,
      "loss": 2.4787265777587892,
      "memory(GiB)": 77.56,
      "step": 55995,
      "token_acc": 0.5129310344827587,
      "train_speed(iter/s)": 1.439374
    },
    {
      "epoch": 2.399211687588364,
      "grad_norm": 6.02125358581543,
      "learning_rate": 5.316424228596625e-05,
      "loss": 2.3187360763549805,
      "memory(GiB)": 77.56,
      "step": 56000,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.439384
    },
    {
      "epoch": 2.399211687588364,
      "eval_loss": 2.0392301082611084,
      "eval_runtime": 14.4516,
      "eval_samples_per_second": 6.92,
      "eval_steps_per_second": 6.92,
      "eval_token_acc": 0.4645858343337335,
      "step": 56000
    },
    {
      "epoch": 2.3994259029176126,
      "grad_norm": 8.405716896057129,
      "learning_rate": 5.315752597407376e-05,
      "loss": 2.3897193908691405,
      "memory(GiB)": 77.56,
      "step": 56005,
      "token_acc": 0.46860782529572337,
      "train_speed(iter/s)": 1.43884
    },
    {
      "epoch": 2.399640118246862,
      "grad_norm": 8.381136894226074,
      "learning_rate": 5.315080960497975e-05,
      "loss": 2.7350940704345703,
      "memory(GiB)": 77.56,
      "step": 56010,
      "token_acc": 0.4344569288389513,
      "train_speed(iter/s)": 1.438849
    },
    {
      "epoch": 2.3998543335761107,
      "grad_norm": 7.539923667907715,
      "learning_rate": 5.3144093178805856e-05,
      "loss": 2.585416793823242,
      "memory(GiB)": 77.56,
      "step": 56015,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.438876
    },
    {
      "epoch": 2.4000685489053595,
      "grad_norm": 4.847936630249023,
      "learning_rate": 5.313737669567377e-05,
      "loss": 2.556245231628418,
      "memory(GiB)": 77.56,
      "step": 56020,
      "token_acc": 0.5059523809523809,
      "train_speed(iter/s)": 1.438876
    },
    {
      "epoch": 2.4002827642346087,
      "grad_norm": 6.339115142822266,
      "learning_rate": 5.313066015570516e-05,
      "loss": 2.512187957763672,
      "memory(GiB)": 77.56,
      "step": 56025,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.438861
    },
    {
      "epoch": 2.4004969795638575,
      "grad_norm": 7.122486114501953,
      "learning_rate": 5.312394355902171e-05,
      "loss": 2.4464868545532226,
      "memory(GiB)": 77.56,
      "step": 56030,
      "token_acc": 0.5245901639344263,
      "train_speed(iter/s)": 1.438869
    },
    {
      "epoch": 2.4007111948931064,
      "grad_norm": 4.212314128875732,
      "learning_rate": 5.3117226905745075e-05,
      "loss": 2.5223152160644533,
      "memory(GiB)": 77.56,
      "step": 56035,
      "token_acc": 0.48606811145510836,
      "train_speed(iter/s)": 1.438876
    },
    {
      "epoch": 2.4009254102223556,
      "grad_norm": 5.982733249664307,
      "learning_rate": 5.311051019599698e-05,
      "loss": 2.341153144836426,
      "memory(GiB)": 77.56,
      "step": 56040,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.438876
    },
    {
      "epoch": 2.4011396255516044,
      "grad_norm": 4.982411861419678,
      "learning_rate": 5.310379342989904e-05,
      "loss": 2.5095245361328127,
      "memory(GiB)": 77.56,
      "step": 56045,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.438879
    },
    {
      "epoch": 2.4013538408808532,
      "grad_norm": 6.8948140144348145,
      "learning_rate": 5.3097076607572984e-05,
      "loss": 2.175385665893555,
      "memory(GiB)": 77.56,
      "step": 56050,
      "token_acc": 0.5269709543568465,
      "train_speed(iter/s)": 1.43888
    },
    {
      "epoch": 2.4015680562101025,
      "grad_norm": 4.494455337524414,
      "learning_rate": 5.309035972914049e-05,
      "loss": 2.364501953125,
      "memory(GiB)": 77.56,
      "step": 56055,
      "token_acc": 0.5132743362831859,
      "train_speed(iter/s)": 1.438902
    },
    {
      "epoch": 2.4017822715393513,
      "grad_norm": 5.145925998687744,
      "learning_rate": 5.308364279472322e-05,
      "loss": 2.5412187576293945,
      "memory(GiB)": 77.56,
      "step": 56060,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.438884
    },
    {
      "epoch": 2.4019964868686,
      "grad_norm": 5.406068325042725,
      "learning_rate": 5.3076925804442865e-05,
      "loss": 2.3500289916992188,
      "memory(GiB)": 77.56,
      "step": 56065,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438883
    },
    {
      "epoch": 2.4022107021978494,
      "grad_norm": 8.811362266540527,
      "learning_rate": 5.307020875842114e-05,
      "loss": 2.3378273010253907,
      "memory(GiB)": 77.56,
      "step": 56070,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438892
    },
    {
      "epoch": 2.402424917527098,
      "grad_norm": 5.983698844909668,
      "learning_rate": 5.306349165677967e-05,
      "loss": 2.7025020599365233,
      "memory(GiB)": 77.56,
      "step": 56075,
      "token_acc": 0.43666666666666665,
      "train_speed(iter/s)": 1.438872
    },
    {
      "epoch": 2.402639132856347,
      "grad_norm": 5.231452465057373,
      "learning_rate": 5.3056774499640184e-05,
      "loss": 2.346792221069336,
      "memory(GiB)": 77.56,
      "step": 56080,
      "token_acc": 0.5119453924914675,
      "train_speed(iter/s)": 1.438878
    },
    {
      "epoch": 2.4028533481855963,
      "grad_norm": 4.943996906280518,
      "learning_rate": 5.305005728712437e-05,
      "loss": 2.5767940521240233,
      "memory(GiB)": 77.56,
      "step": 56085,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.438881
    },
    {
      "epoch": 2.403067563514845,
      "grad_norm": 6.586244106292725,
      "learning_rate": 5.304334001935389e-05,
      "loss": 2.310846519470215,
      "memory(GiB)": 77.56,
      "step": 56090,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.438901
    },
    {
      "epoch": 2.403281778844094,
      "grad_norm": 6.0121989250183105,
      "learning_rate": 5.3036622696450466e-05,
      "loss": 2.0009632110595703,
      "memory(GiB)": 77.56,
      "step": 56095,
      "token_acc": 0.550561797752809,
      "train_speed(iter/s)": 1.438919
    },
    {
      "epoch": 2.403495994173343,
      "grad_norm": 5.395315170288086,
      "learning_rate": 5.302990531853578e-05,
      "loss": 2.4876937866210938,
      "memory(GiB)": 77.56,
      "step": 56100,
      "token_acc": 0.4911660777385159,
      "train_speed(iter/s)": 1.438901
    },
    {
      "epoch": 2.403710209502592,
      "grad_norm": 4.957891464233398,
      "learning_rate": 5.302318788573149e-05,
      "loss": 2.266989517211914,
      "memory(GiB)": 77.56,
      "step": 56105,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438879
    },
    {
      "epoch": 2.4039244248318408,
      "grad_norm": 6.9275054931640625,
      "learning_rate": 5.3016470398159344e-05,
      "loss": 2.6653812408447264,
      "memory(GiB)": 77.56,
      "step": 56110,
      "token_acc": 0.5019305019305019,
      "train_speed(iter/s)": 1.438864
    },
    {
      "epoch": 2.40413864016109,
      "grad_norm": 5.8965582847595215,
      "learning_rate": 5.3009752855941e-05,
      "loss": 2.4196725845336915,
      "memory(GiB)": 77.56,
      "step": 56115,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.438861
    },
    {
      "epoch": 2.404352855490339,
      "grad_norm": 4.167325019836426,
      "learning_rate": 5.300303525919813e-05,
      "loss": 2.3755931854248047,
      "memory(GiB)": 77.56,
      "step": 56120,
      "token_acc": 0.49407114624505927,
      "train_speed(iter/s)": 1.438885
    },
    {
      "epoch": 2.4045670708195876,
      "grad_norm": 5.9489970207214355,
      "learning_rate": 5.2996317608052494e-05,
      "loss": 2.276514434814453,
      "memory(GiB)": 77.56,
      "step": 56125,
      "token_acc": 0.5302491103202847,
      "train_speed(iter/s)": 1.438875
    },
    {
      "epoch": 2.404781286148837,
      "grad_norm": 6.91579008102417,
      "learning_rate": 5.298959990262574e-05,
      "loss": 2.4200796127319335,
      "memory(GiB)": 77.56,
      "step": 56130,
      "token_acc": 0.4843205574912892,
      "train_speed(iter/s)": 1.438868
    },
    {
      "epoch": 2.4049955014780857,
      "grad_norm": 5.1532511711120605,
      "learning_rate": 5.298288214303958e-05,
      "loss": 2.5728235244750977,
      "memory(GiB)": 77.56,
      "step": 56135,
      "token_acc": 0.501466275659824,
      "train_speed(iter/s)": 1.438897
    },
    {
      "epoch": 2.4052097168073345,
      "grad_norm": 5.5746307373046875,
      "learning_rate": 5.2976164329415725e-05,
      "loss": 2.5000816345214845,
      "memory(GiB)": 77.56,
      "step": 56140,
      "token_acc": 0.4670846394984326,
      "train_speed(iter/s)": 1.438887
    },
    {
      "epoch": 2.405423932136584,
      "grad_norm": 5.644297122955322,
      "learning_rate": 5.296944646187585e-05,
      "loss": 2.333147430419922,
      "memory(GiB)": 77.56,
      "step": 56145,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438906
    },
    {
      "epoch": 2.4056381474658326,
      "grad_norm": 5.093411922454834,
      "learning_rate": 5.296272854054166e-05,
      "loss": 2.655217742919922,
      "memory(GiB)": 77.56,
      "step": 56150,
      "token_acc": 0.48338368580060426,
      "train_speed(iter/s)": 1.438915
    },
    {
      "epoch": 2.4058523627950814,
      "grad_norm": 5.415943622589111,
      "learning_rate": 5.295601056553486e-05,
      "loss": 2.397196960449219,
      "memory(GiB)": 77.56,
      "step": 56155,
      "token_acc": 0.478134110787172,
      "train_speed(iter/s)": 1.438934
    },
    {
      "epoch": 2.4060665781243307,
      "grad_norm": 4.815046310424805,
      "learning_rate": 5.294929253697718e-05,
      "loss": 2.4271930694580077,
      "memory(GiB)": 77.56,
      "step": 56160,
      "token_acc": 0.5119760479041916,
      "train_speed(iter/s)": 1.438949
    },
    {
      "epoch": 2.4062807934535795,
      "grad_norm": 6.341163635253906,
      "learning_rate": 5.294257445499028e-05,
      "loss": 2.531138229370117,
      "memory(GiB)": 77.56,
      "step": 56165,
      "token_acc": 0.468,
      "train_speed(iter/s)": 1.43898
    },
    {
      "epoch": 2.4064950087828283,
      "grad_norm": 4.485575199127197,
      "learning_rate": 5.29358563196959e-05,
      "loss": 2.299831771850586,
      "memory(GiB)": 77.56,
      "step": 56170,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.438958
    },
    {
      "epoch": 2.4067092241120775,
      "grad_norm": 5.203620433807373,
      "learning_rate": 5.2929138131215715e-05,
      "loss": 2.5045700073242188,
      "memory(GiB)": 77.56,
      "step": 56175,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.438995
    },
    {
      "epoch": 2.4069234394413264,
      "grad_norm": 4.303263187408447,
      "learning_rate": 5.292241988967144e-05,
      "loss": 2.6499711990356447,
      "memory(GiB)": 77.56,
      "step": 56180,
      "token_acc": 0.4383116883116883,
      "train_speed(iter/s)": 1.438977
    },
    {
      "epoch": 2.407137654770575,
      "grad_norm": 5.529666423797607,
      "learning_rate": 5.29157015951848e-05,
      "loss": 2.4012521743774413,
      "memory(GiB)": 77.56,
      "step": 56185,
      "token_acc": 0.5050167224080268,
      "train_speed(iter/s)": 1.438967
    },
    {
      "epoch": 2.4073518700998244,
      "grad_norm": 5.961589336395264,
      "learning_rate": 5.2908983247877475e-05,
      "loss": 2.374089813232422,
      "memory(GiB)": 77.56,
      "step": 56190,
      "token_acc": 0.492,
      "train_speed(iter/s)": 1.438979
    },
    {
      "epoch": 2.4075660854290732,
      "grad_norm": 5.217300891876221,
      "learning_rate": 5.2902264847871196e-05,
      "loss": 2.674641227722168,
      "memory(GiB)": 77.56,
      "step": 56195,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.438983
    },
    {
      "epoch": 2.407780300758322,
      "grad_norm": 5.262786865234375,
      "learning_rate": 5.289554639528768e-05,
      "loss": 2.3925403594970702,
      "memory(GiB)": 77.56,
      "step": 56200,
      "token_acc": 0.5043731778425656,
      "train_speed(iter/s)": 1.438967
    },
    {
      "epoch": 2.4079945160875713,
      "grad_norm": 5.554116725921631,
      "learning_rate": 5.2888827890248604e-05,
      "loss": 2.6348649978637697,
      "memory(GiB)": 77.56,
      "step": 56205,
      "token_acc": 0.4440677966101695,
      "train_speed(iter/s)": 1.438979
    },
    {
      "epoch": 2.40820873141682,
      "grad_norm": 6.3835906982421875,
      "learning_rate": 5.288210933287572e-05,
      "loss": 2.7073829650878904,
      "memory(GiB)": 77.56,
      "step": 56210,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.438979
    },
    {
      "epoch": 2.408422946746069,
      "grad_norm": 4.819346904754639,
      "learning_rate": 5.287539072329072e-05,
      "loss": 2.3233657836914063,
      "memory(GiB)": 77.56,
      "step": 56215,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.438986
    },
    {
      "epoch": 2.408637162075318,
      "grad_norm": 4.799155235290527,
      "learning_rate": 5.286867206161531e-05,
      "loss": 2.3258378982543944,
      "memory(GiB)": 77.56,
      "step": 56220,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.439
    },
    {
      "epoch": 2.408851377404567,
      "grad_norm": 4.620659828186035,
      "learning_rate": 5.286195334797122e-05,
      "loss": 2.4757110595703127,
      "memory(GiB)": 77.56,
      "step": 56225,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.439002
    },
    {
      "epoch": 2.409065592733816,
      "grad_norm": 5.389440059661865,
      "learning_rate": 5.285523458248015e-05,
      "loss": 2.359396553039551,
      "memory(GiB)": 77.56,
      "step": 56230,
      "token_acc": 0.4513888888888889,
      "train_speed(iter/s)": 1.439012
    },
    {
      "epoch": 2.409279808063065,
      "grad_norm": 5.467198848724365,
      "learning_rate": 5.284851576526383e-05,
      "loss": 2.6087230682373046,
      "memory(GiB)": 77.56,
      "step": 56235,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.439019
    },
    {
      "epoch": 2.409494023392314,
      "grad_norm": 5.533074378967285,
      "learning_rate": 5.2841796896443986e-05,
      "loss": 2.4206157684326173,
      "memory(GiB)": 77.56,
      "step": 56240,
      "token_acc": 0.4819277108433735,
      "train_speed(iter/s)": 1.43902
    },
    {
      "epoch": 2.4097082387215627,
      "grad_norm": 5.181727886199951,
      "learning_rate": 5.283507797614232e-05,
      "loss": 2.207934761047363,
      "memory(GiB)": 77.56,
      "step": 56245,
      "token_acc": 0.5427350427350427,
      "train_speed(iter/s)": 1.439021
    },
    {
      "epoch": 2.409922454050812,
      "grad_norm": 5.358429431915283,
      "learning_rate": 5.282835900448055e-05,
      "loss": 2.2529346466064455,
      "memory(GiB)": 77.56,
      "step": 56250,
      "token_acc": 0.5032467532467533,
      "train_speed(iter/s)": 1.439049
    },
    {
      "epoch": 2.4101366693800608,
      "grad_norm": 5.2248382568359375,
      "learning_rate": 5.282163998158042e-05,
      "loss": 2.5750003814697267,
      "memory(GiB)": 77.56,
      "step": 56255,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.439061
    },
    {
      "epoch": 2.4103508847093096,
      "grad_norm": 6.446933746337891,
      "learning_rate": 5.281492090756364e-05,
      "loss": 2.411317825317383,
      "memory(GiB)": 77.56,
      "step": 56260,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.439091
    },
    {
      "epoch": 2.410565100038559,
      "grad_norm": 4.414382457733154,
      "learning_rate": 5.280820178255188e-05,
      "loss": 2.3740716934204102,
      "memory(GiB)": 77.56,
      "step": 56265,
      "token_acc": 0.5331010452961672,
      "train_speed(iter/s)": 1.439098
    },
    {
      "epoch": 2.4107793153678077,
      "grad_norm": 6.324094295501709,
      "learning_rate": 5.2801482606666955e-05,
      "loss": 2.545774459838867,
      "memory(GiB)": 77.56,
      "step": 56270,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.43911
    },
    {
      "epoch": 2.4109935306970565,
      "grad_norm": 4.340853214263916,
      "learning_rate": 5.279476338003053e-05,
      "loss": 2.3693159103393553,
      "memory(GiB)": 77.56,
      "step": 56275,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.439108
    },
    {
      "epoch": 2.4112077460263057,
      "grad_norm": 6.0674262046813965,
      "learning_rate": 5.2788044102764345e-05,
      "loss": 2.3557533264160155,
      "memory(GiB)": 77.56,
      "step": 56280,
      "token_acc": 0.5239852398523985,
      "train_speed(iter/s)": 1.439105
    },
    {
      "epoch": 2.4114219613555545,
      "grad_norm": 5.321403980255127,
      "learning_rate": 5.278132477499014e-05,
      "loss": 2.6936790466308596,
      "memory(GiB)": 77.56,
      "step": 56285,
      "token_acc": 0.43137254901960786,
      "train_speed(iter/s)": 1.439147
    },
    {
      "epoch": 2.4116361766848033,
      "grad_norm": 6.133849143981934,
      "learning_rate": 5.277460539682961e-05,
      "loss": 2.4603178024291994,
      "memory(GiB)": 77.56,
      "step": 56290,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.439152
    },
    {
      "epoch": 2.4118503920140526,
      "grad_norm": 4.4758148193359375,
      "learning_rate": 5.276788596840451e-05,
      "loss": 2.671260643005371,
      "memory(GiB)": 77.56,
      "step": 56295,
      "token_acc": 0.4658753709198813,
      "train_speed(iter/s)": 1.439165
    },
    {
      "epoch": 2.4120646073433014,
      "grad_norm": 5.828741550445557,
      "learning_rate": 5.276116648983656e-05,
      "loss": 2.5580806732177734,
      "memory(GiB)": 77.56,
      "step": 56300,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.439175
    },
    {
      "epoch": 2.4122788226725502,
      "grad_norm": 5.495771884918213,
      "learning_rate": 5.275444696124747e-05,
      "loss": 2.572454833984375,
      "memory(GiB)": 77.56,
      "step": 56305,
      "token_acc": 0.5095541401273885,
      "train_speed(iter/s)": 1.439165
    },
    {
      "epoch": 2.4124930380017995,
      "grad_norm": 3.5633745193481445,
      "learning_rate": 5.2747727382758993e-05,
      "loss": 2.3117183685302733,
      "memory(GiB)": 77.56,
      "step": 56310,
      "token_acc": 0.5197368421052632,
      "train_speed(iter/s)": 1.43915
    },
    {
      "epoch": 2.4127072533310483,
      "grad_norm": 5.507272720336914,
      "learning_rate": 5.274100775449288e-05,
      "loss": 2.4593910217285155,
      "memory(GiB)": 77.56,
      "step": 56315,
      "token_acc": 0.47653429602888087,
      "train_speed(iter/s)": 1.439153
    },
    {
      "epoch": 2.412921468660297,
      "grad_norm": 4.739686012268066,
      "learning_rate": 5.2734288076570824e-05,
      "loss": 2.3920015335083007,
      "memory(GiB)": 77.56,
      "step": 56320,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.439153
    },
    {
      "epoch": 2.4131356839895464,
      "grad_norm": 5.539514541625977,
      "learning_rate": 5.2727568349114576e-05,
      "loss": 2.201377105712891,
      "memory(GiB)": 77.56,
      "step": 56325,
      "token_acc": 0.5288461538461539,
      "train_speed(iter/s)": 1.439135
    },
    {
      "epoch": 2.413349899318795,
      "grad_norm": 5.625189781188965,
      "learning_rate": 5.2720848572245874e-05,
      "loss": 2.710480880737305,
      "memory(GiB)": 77.56,
      "step": 56330,
      "token_acc": 0.4377224199288256,
      "train_speed(iter/s)": 1.439147
    },
    {
      "epoch": 2.413564114648044,
      "grad_norm": 7.14446496963501,
      "learning_rate": 5.271412874608644e-05,
      "loss": 2.3742630004882814,
      "memory(GiB)": 77.56,
      "step": 56335,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.439137
    },
    {
      "epoch": 2.4137783299772932,
      "grad_norm": 4.664318084716797,
      "learning_rate": 5.270740887075801e-05,
      "loss": 2.212387466430664,
      "memory(GiB)": 77.56,
      "step": 56340,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.439101
    },
    {
      "epoch": 2.413992545306542,
      "grad_norm": 6.09464168548584,
      "learning_rate": 5.2700688946382345e-05,
      "loss": 2.399026298522949,
      "memory(GiB)": 77.56,
      "step": 56345,
      "token_acc": 0.532,
      "train_speed(iter/s)": 1.439106
    },
    {
      "epoch": 2.414206760635791,
      "grad_norm": 4.864884376525879,
      "learning_rate": 5.269396897308117e-05,
      "loss": 2.4771289825439453,
      "memory(GiB)": 77.56,
      "step": 56350,
      "token_acc": 0.4641638225255973,
      "train_speed(iter/s)": 1.439107
    },
    {
      "epoch": 2.41442097596504,
      "grad_norm": 4.9332499504089355,
      "learning_rate": 5.2687248950976206e-05,
      "loss": 2.2081476211547852,
      "memory(GiB)": 77.56,
      "step": 56355,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.439111
    },
    {
      "epoch": 2.414635191294289,
      "grad_norm": 5.544666767120361,
      "learning_rate": 5.268052888018922e-05,
      "loss": 2.853831481933594,
      "memory(GiB)": 77.56,
      "step": 56360,
      "token_acc": 0.4186046511627907,
      "train_speed(iter/s)": 1.439141
    },
    {
      "epoch": 2.4148494066235378,
      "grad_norm": 4.919857025146484,
      "learning_rate": 5.267380876084192e-05,
      "loss": 2.4408655166625977,
      "memory(GiB)": 77.56,
      "step": 56365,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.439159
    },
    {
      "epoch": 2.415063621952787,
      "grad_norm": 5.91994047164917,
      "learning_rate": 5.266708859305608e-05,
      "loss": 2.340895080566406,
      "memory(GiB)": 77.56,
      "step": 56370,
      "token_acc": 0.5691823899371069,
      "train_speed(iter/s)": 1.439135
    },
    {
      "epoch": 2.415277837282036,
      "grad_norm": 3.824577808380127,
      "learning_rate": 5.266036837695344e-05,
      "loss": 2.38179931640625,
      "memory(GiB)": 77.56,
      "step": 56375,
      "token_acc": 0.5109717868338558,
      "train_speed(iter/s)": 1.439141
    },
    {
      "epoch": 2.4154920526112846,
      "grad_norm": 5.2734832763671875,
      "learning_rate": 5.26536481126557e-05,
      "loss": 2.5354665756225585,
      "memory(GiB)": 77.56,
      "step": 56380,
      "token_acc": 0.49731182795698925,
      "train_speed(iter/s)": 1.43916
    },
    {
      "epoch": 2.415706267940534,
      "grad_norm": 5.425694942474365,
      "learning_rate": 5.264692780028465e-05,
      "loss": 2.5977821350097656,
      "memory(GiB)": 77.56,
      "step": 56385,
      "token_acc": 0.45302013422818793,
      "train_speed(iter/s)": 1.43918
    },
    {
      "epoch": 2.4159204832697827,
      "grad_norm": 4.564004421234131,
      "learning_rate": 5.264020743996203e-05,
      "loss": 2.2885391235351564,
      "memory(GiB)": 77.56,
      "step": 56390,
      "token_acc": 0.5337423312883436,
      "train_speed(iter/s)": 1.439182
    },
    {
      "epoch": 2.4161346985990315,
      "grad_norm": 7.108612537384033,
      "learning_rate": 5.263348703180956e-05,
      "loss": 2.2892339706420897,
      "memory(GiB)": 77.56,
      "step": 56395,
      "token_acc": 0.516245487364621,
      "train_speed(iter/s)": 1.43917
    },
    {
      "epoch": 2.4163489139282808,
      "grad_norm": 4.639871120452881,
      "learning_rate": 5.262676657594901e-05,
      "loss": 2.316487503051758,
      "memory(GiB)": 77.56,
      "step": 56400,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.439195
    },
    {
      "epoch": 2.4165631292575296,
      "grad_norm": 6.653697967529297,
      "learning_rate": 5.2620046072502124e-05,
      "loss": 2.4122287750244142,
      "memory(GiB)": 77.56,
      "step": 56405,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.439211
    },
    {
      "epoch": 2.4167773445867784,
      "grad_norm": 5.839682579040527,
      "learning_rate": 5.2613325521590626e-05,
      "loss": 2.5080310821533205,
      "memory(GiB)": 77.56,
      "step": 56410,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.439188
    },
    {
      "epoch": 2.4169915599160277,
      "grad_norm": 4.128732681274414,
      "learning_rate": 5.26066049233363e-05,
      "loss": 2.1906267166137696,
      "memory(GiB)": 77.56,
      "step": 56415,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.439202
    },
    {
      "epoch": 2.4172057752452765,
      "grad_norm": 7.49148416519165,
      "learning_rate": 5.259988427786088e-05,
      "loss": 2.7538579940795898,
      "memory(GiB)": 77.56,
      "step": 56420,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.439206
    },
    {
      "epoch": 2.4174199905745253,
      "grad_norm": 6.461582183837891,
      "learning_rate": 5.259316358528611e-05,
      "loss": 2.402140426635742,
      "memory(GiB)": 77.56,
      "step": 56425,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.43923
    },
    {
      "epoch": 2.4176342059037745,
      "grad_norm": 5.6993207931518555,
      "learning_rate": 5.2586442845733765e-05,
      "loss": 2.5821022033691405,
      "memory(GiB)": 77.56,
      "step": 56430,
      "token_acc": 0.4672131147540984,
      "train_speed(iter/s)": 1.439221
    },
    {
      "epoch": 2.4178484212330233,
      "grad_norm": 5.566256999969482,
      "learning_rate": 5.257972205932558e-05,
      "loss": 2.52720947265625,
      "memory(GiB)": 77.56,
      "step": 56435,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.439227
    },
    {
      "epoch": 2.418062636562272,
      "grad_norm": 5.295451641082764,
      "learning_rate": 5.257300122618329e-05,
      "loss": 2.3013282775878907,
      "memory(GiB)": 77.56,
      "step": 56440,
      "token_acc": 0.4837758112094395,
      "train_speed(iter/s)": 1.439242
    },
    {
      "epoch": 2.4182768518915214,
      "grad_norm": 6.468355178833008,
      "learning_rate": 5.2566280346428687e-05,
      "loss": 2.83559513092041,
      "memory(GiB)": 77.56,
      "step": 56445,
      "token_acc": 0.4246987951807229,
      "train_speed(iter/s)": 1.439258
    },
    {
      "epoch": 2.4184910672207702,
      "grad_norm": 5.487728118896484,
      "learning_rate": 5.25595594201835e-05,
      "loss": 2.2627710342407226,
      "memory(GiB)": 77.56,
      "step": 56450,
      "token_acc": 0.5420875420875421,
      "train_speed(iter/s)": 1.439229
    },
    {
      "epoch": 2.418705282550019,
      "grad_norm": 4.941653251647949,
      "learning_rate": 5.2552838447569485e-05,
      "loss": 2.6452669143676757,
      "memory(GiB)": 77.56,
      "step": 56455,
      "token_acc": 0.4658753709198813,
      "train_speed(iter/s)": 1.439222
    },
    {
      "epoch": 2.4189194978792683,
      "grad_norm": 5.474538803100586,
      "learning_rate": 5.254611742870842e-05,
      "loss": 2.5214195251464844,
      "memory(GiB)": 77.56,
      "step": 56460,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.439258
    },
    {
      "epoch": 2.419133713208517,
      "grad_norm": 5.894259929656982,
      "learning_rate": 5.253939636372206e-05,
      "loss": 2.303129959106445,
      "memory(GiB)": 77.56,
      "step": 56465,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.439264
    },
    {
      "epoch": 2.419347928537766,
      "grad_norm": 5.728356838226318,
      "learning_rate": 5.253267525273213e-05,
      "loss": 2.2912128448486326,
      "memory(GiB)": 77.56,
      "step": 56470,
      "token_acc": 0.48417721518987344,
      "train_speed(iter/s)": 1.439263
    },
    {
      "epoch": 2.419562143867015,
      "grad_norm": 4.11015510559082,
      "learning_rate": 5.252595409586042e-05,
      "loss": 2.379174995422363,
      "memory(GiB)": 77.56,
      "step": 56475,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.439264
    },
    {
      "epoch": 2.419776359196264,
      "grad_norm": 6.598214626312256,
      "learning_rate": 5.251923289322868e-05,
      "loss": 2.0697322845458985,
      "memory(GiB)": 77.56,
      "step": 56480,
      "token_acc": 0.5134099616858238,
      "train_speed(iter/s)": 1.439272
    },
    {
      "epoch": 2.419990574525513,
      "grad_norm": 5.734899997711182,
      "learning_rate": 5.251251164495868e-05,
      "loss": 2.4696388244628906,
      "memory(GiB)": 77.56,
      "step": 56485,
      "token_acc": 0.4887459807073955,
      "train_speed(iter/s)": 1.439277
    },
    {
      "epoch": 2.420204789854762,
      "grad_norm": 9.856157302856445,
      "learning_rate": 5.250579035117217e-05,
      "loss": 2.546694755554199,
      "memory(GiB)": 77.56,
      "step": 56490,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.439247
    },
    {
      "epoch": 2.420419005184011,
      "grad_norm": 4.791378021240234,
      "learning_rate": 5.249906901199091e-05,
      "loss": 2.635769844055176,
      "memory(GiB)": 77.56,
      "step": 56495,
      "token_acc": 0.4261744966442953,
      "train_speed(iter/s)": 1.439223
    },
    {
      "epoch": 2.4206332205132597,
      "grad_norm": 6.565951824188232,
      "learning_rate": 5.2492347627536686e-05,
      "loss": 2.485715866088867,
      "memory(GiB)": 77.56,
      "step": 56500,
      "token_acc": 0.5232558139534884,
      "train_speed(iter/s)": 1.439203
    },
    {
      "epoch": 2.4206332205132597,
      "eval_loss": 2.3672709465026855,
      "eval_runtime": 14.3838,
      "eval_samples_per_second": 6.952,
      "eval_steps_per_second": 6.952,
      "eval_token_acc": 0.4813218390804598,
      "step": 56500
    },
    {
      "epoch": 2.420847435842509,
      "grad_norm": 5.80452823638916,
      "learning_rate": 5.248562619793124e-05,
      "loss": 2.574502182006836,
      "memory(GiB)": 77.56,
      "step": 56505,
      "token_acc": 0.48422090729783035,
      "train_speed(iter/s)": 1.438658
    },
    {
      "epoch": 2.4210616511717578,
      "grad_norm": 4.908410549163818,
      "learning_rate": 5.247890472329634e-05,
      "loss": 2.420794677734375,
      "memory(GiB)": 77.56,
      "step": 56510,
      "token_acc": 0.4954954954954955,
      "train_speed(iter/s)": 1.438666
    },
    {
      "epoch": 2.4212758665010066,
      "grad_norm": 5.8564043045043945,
      "learning_rate": 5.247218320375376e-05,
      "loss": 2.382805824279785,
      "memory(GiB)": 77.56,
      "step": 56515,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.438673
    },
    {
      "epoch": 2.421490081830256,
      "grad_norm": 5.025363922119141,
      "learning_rate": 5.246546163942526e-05,
      "loss": 2.334534454345703,
      "memory(GiB)": 77.56,
      "step": 56520,
      "token_acc": 0.4770992366412214,
      "train_speed(iter/s)": 1.43862
    },
    {
      "epoch": 2.4217042971595046,
      "grad_norm": 7.295405387878418,
      "learning_rate": 5.2458740030432595e-05,
      "loss": 2.5426416397094727,
      "memory(GiB)": 77.56,
      "step": 56525,
      "token_acc": 0.43006993006993005,
      "train_speed(iter/s)": 1.438625
    },
    {
      "epoch": 2.4219185124887534,
      "grad_norm": 3.918203115463257,
      "learning_rate": 5.245201837689757e-05,
      "loss": 2.103429985046387,
      "memory(GiB)": 77.56,
      "step": 56530,
      "token_acc": 0.5186440677966102,
      "train_speed(iter/s)": 1.438638
    },
    {
      "epoch": 2.4221327278180027,
      "grad_norm": 5.002352714538574,
      "learning_rate": 5.2445296678941935e-05,
      "loss": 2.5431236267089843,
      "memory(GiB)": 77.56,
      "step": 56535,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.438645
    },
    {
      "epoch": 2.4223469431472515,
      "grad_norm": 6.707949161529541,
      "learning_rate": 5.243857493668743e-05,
      "loss": 2.640879821777344,
      "memory(GiB)": 77.56,
      "step": 56540,
      "token_acc": 0.450920245398773,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.4225611584765003,
      "grad_norm": 4.90098237991333,
      "learning_rate": 5.2431853150255886e-05,
      "loss": 2.680977439880371,
      "memory(GiB)": 77.56,
      "step": 56545,
      "token_acc": 0.4554140127388535,
      "train_speed(iter/s)": 1.438666
    },
    {
      "epoch": 2.4227753738057496,
      "grad_norm": 6.017824172973633,
      "learning_rate": 5.2425131319769024e-05,
      "loss": 2.978159523010254,
      "memory(GiB)": 77.56,
      "step": 56550,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.438687
    },
    {
      "epoch": 2.4229895891349984,
      "grad_norm": 7.294832229614258,
      "learning_rate": 5.241840944534862e-05,
      "loss": 1.836798095703125,
      "memory(GiB)": 77.56,
      "step": 56555,
      "token_acc": 0.5097276264591439,
      "train_speed(iter/s)": 1.43869
    },
    {
      "epoch": 2.423203804464247,
      "grad_norm": 5.722614765167236,
      "learning_rate": 5.241168752711648e-05,
      "loss": 2.312526321411133,
      "memory(GiB)": 77.56,
      "step": 56560,
      "token_acc": 0.5019607843137255,
      "train_speed(iter/s)": 1.438706
    },
    {
      "epoch": 2.4234180197934965,
      "grad_norm": 6.166382312774658,
      "learning_rate": 5.240496556519435e-05,
      "loss": 2.3768888473510743,
      "memory(GiB)": 77.56,
      "step": 56565,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.438743
    },
    {
      "epoch": 2.4236322351227453,
      "grad_norm": 5.268062114715576,
      "learning_rate": 5.239824355970401e-05,
      "loss": 2.379104804992676,
      "memory(GiB)": 77.56,
      "step": 56570,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.438767
    },
    {
      "epoch": 2.4238464504519945,
      "grad_norm": 5.939847946166992,
      "learning_rate": 5.239152151076725e-05,
      "loss": 2.3494287490844727,
      "memory(GiB)": 77.56,
      "step": 56575,
      "token_acc": 0.5223880597014925,
      "train_speed(iter/s)": 1.43875
    },
    {
      "epoch": 2.4240606657812434,
      "grad_norm": 5.592379570007324,
      "learning_rate": 5.2384799418505824e-05,
      "loss": 2.7172195434570314,
      "memory(GiB)": 77.56,
      "step": 56580,
      "token_acc": 0.4541984732824427,
      "train_speed(iter/s)": 1.438787
    },
    {
      "epoch": 2.424274881110492,
      "grad_norm": 5.186420917510986,
      "learning_rate": 5.237807728304152e-05,
      "loss": 2.4347774505615236,
      "memory(GiB)": 77.56,
      "step": 56585,
      "token_acc": 0.5150214592274678,
      "train_speed(iter/s)": 1.438816
    },
    {
      "epoch": 2.4244890964397414,
      "grad_norm": 4.3312554359436035,
      "learning_rate": 5.2371355104496125e-05,
      "loss": 2.3911298751831054,
      "memory(GiB)": 77.56,
      "step": 56590,
      "token_acc": 0.4745222929936306,
      "train_speed(iter/s)": 1.438838
    },
    {
      "epoch": 2.4247033117689902,
      "grad_norm": 4.943553924560547,
      "learning_rate": 5.236463288299139e-05,
      "loss": 2.390595054626465,
      "memory(GiB)": 77.56,
      "step": 56595,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.438839
    },
    {
      "epoch": 2.424917527098239,
      "grad_norm": 4.774374485015869,
      "learning_rate": 5.235791061864912e-05,
      "loss": 2.2636978149414064,
      "memory(GiB)": 77.56,
      "step": 56600,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.438826
    },
    {
      "epoch": 2.4251317424274883,
      "grad_norm": 6.53861141204834,
      "learning_rate": 5.2351188311591093e-05,
      "loss": 2.4385959625244142,
      "memory(GiB)": 77.56,
      "step": 56605,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.438834
    },
    {
      "epoch": 2.425345957756737,
      "grad_norm": 4.330164909362793,
      "learning_rate": 5.234446596193907e-05,
      "loss": 2.604307174682617,
      "memory(GiB)": 77.56,
      "step": 56610,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.438849
    },
    {
      "epoch": 2.425560173085986,
      "grad_norm": 5.388030052185059,
      "learning_rate": 5.233774356981487e-05,
      "loss": 2.4409034729003904,
      "memory(GiB)": 77.56,
      "step": 56615,
      "token_acc": 0.506578947368421,
      "train_speed(iter/s)": 1.43885
    },
    {
      "epoch": 2.425774388415235,
      "grad_norm": 4.841846942901611,
      "learning_rate": 5.233102113534024e-05,
      "loss": 2.645443916320801,
      "memory(GiB)": 77.56,
      "step": 56620,
      "token_acc": 0.4406779661016949,
      "train_speed(iter/s)": 1.438847
    },
    {
      "epoch": 2.425988603744484,
      "grad_norm": 5.512052059173584,
      "learning_rate": 5.2324298658636974e-05,
      "loss": 2.2702354431152343,
      "memory(GiB)": 77.56,
      "step": 56625,
      "token_acc": 0.50814332247557,
      "train_speed(iter/s)": 1.438854
    },
    {
      "epoch": 2.426202819073733,
      "grad_norm": 6.381377220153809,
      "learning_rate": 5.231757613982686e-05,
      "loss": 2.223843574523926,
      "memory(GiB)": 77.56,
      "step": 56630,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438861
    },
    {
      "epoch": 2.426417034402982,
      "grad_norm": 4.889789581298828,
      "learning_rate": 5.2310853579031674e-05,
      "loss": 2.463627815246582,
      "memory(GiB)": 77.56,
      "step": 56635,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.438827
    },
    {
      "epoch": 2.426631249732231,
      "grad_norm": 6.635883808135986,
      "learning_rate": 5.23041309763732e-05,
      "loss": 2.3669376373291016,
      "memory(GiB)": 77.56,
      "step": 56640,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.43884
    },
    {
      "epoch": 2.4268454650614797,
      "grad_norm": 5.636005401611328,
      "learning_rate": 5.229740833197325e-05,
      "loss": 2.5137474060058596,
      "memory(GiB)": 77.56,
      "step": 56645,
      "token_acc": 0.44904458598726116,
      "train_speed(iter/s)": 1.438832
    },
    {
      "epoch": 2.427059680390729,
      "grad_norm": 5.689830303192139,
      "learning_rate": 5.2290685645953597e-05,
      "loss": 2.5631351470947266,
      "memory(GiB)": 77.56,
      "step": 56650,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.438847
    },
    {
      "epoch": 2.4272738957199778,
      "grad_norm": 4.363366603851318,
      "learning_rate": 5.2283962918436014e-05,
      "loss": 2.565142059326172,
      "memory(GiB)": 77.56,
      "step": 56655,
      "token_acc": 0.4908424908424908,
      "train_speed(iter/s)": 1.438872
    },
    {
      "epoch": 2.4274881110492266,
      "grad_norm": 7.218557357788086,
      "learning_rate": 5.227724014954231e-05,
      "loss": 1.9128873825073243,
      "memory(GiB)": 77.56,
      "step": 56660,
      "token_acc": 0.5481171548117155,
      "train_speed(iter/s)": 1.438892
    },
    {
      "epoch": 2.427702326378476,
      "grad_norm": 6.1528544425964355,
      "learning_rate": 5.227051733939425e-05,
      "loss": 2.5260448455810547,
      "memory(GiB)": 77.56,
      "step": 56665,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.438884
    },
    {
      "epoch": 2.4279165417077246,
      "grad_norm": 4.401196002960205,
      "learning_rate": 5.2263794488113635e-05,
      "loss": 2.6199535369873046,
      "memory(GiB)": 77.56,
      "step": 56670,
      "token_acc": 0.48265895953757226,
      "train_speed(iter/s)": 1.438881
    },
    {
      "epoch": 2.4281307570369735,
      "grad_norm": 6.6801838874816895,
      "learning_rate": 5.225707159582227e-05,
      "loss": 2.549765396118164,
      "memory(GiB)": 77.56,
      "step": 56675,
      "token_acc": 0.5138461538461538,
      "train_speed(iter/s)": 1.438884
    },
    {
      "epoch": 2.4283449723662227,
      "grad_norm": 5.841605186462402,
      "learning_rate": 5.225034866264193e-05,
      "loss": 2.4959110260009765,
      "memory(GiB)": 77.56,
      "step": 56680,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.438894
    },
    {
      "epoch": 2.4285591876954715,
      "grad_norm": 4.549417018890381,
      "learning_rate": 5.224362568869442e-05,
      "loss": 2.639225387573242,
      "memory(GiB)": 77.56,
      "step": 56685,
      "token_acc": 0.4806201550387597,
      "train_speed(iter/s)": 1.438891
    },
    {
      "epoch": 2.4287734030247203,
      "grad_norm": 4.565102577209473,
      "learning_rate": 5.223690267410153e-05,
      "loss": 2.490378189086914,
      "memory(GiB)": 77.56,
      "step": 56690,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438884
    },
    {
      "epoch": 2.4289876183539696,
      "grad_norm": 5.574451923370361,
      "learning_rate": 5.223017961898504e-05,
      "loss": 2.2715522766113283,
      "memory(GiB)": 77.56,
      "step": 56695,
      "token_acc": 0.48627450980392156,
      "train_speed(iter/s)": 1.438835
    },
    {
      "epoch": 2.4292018336832184,
      "grad_norm": 5.546135425567627,
      "learning_rate": 5.222345652346675e-05,
      "loss": 2.2152938842773438,
      "memory(GiB)": 77.56,
      "step": 56700,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438871
    },
    {
      "epoch": 2.429416049012467,
      "grad_norm": 4.191052436828613,
      "learning_rate": 5.221673338766847e-05,
      "loss": 2.5945505142211913,
      "memory(GiB)": 77.56,
      "step": 56705,
      "token_acc": 0.4899713467048711,
      "train_speed(iter/s)": 1.438887
    },
    {
      "epoch": 2.4296302643417165,
      "grad_norm": 10.171424865722656,
      "learning_rate": 5.221001021171198e-05,
      "loss": 2.2691066741943358,
      "memory(GiB)": 77.56,
      "step": 56710,
      "token_acc": 0.549800796812749,
      "train_speed(iter/s)": 1.43891
    },
    {
      "epoch": 2.4298444796709653,
      "grad_norm": 4.609151840209961,
      "learning_rate": 5.220328699571908e-05,
      "loss": 2.553591728210449,
      "memory(GiB)": 77.56,
      "step": 56715,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.438921
    },
    {
      "epoch": 2.430058695000214,
      "grad_norm": 6.131407737731934,
      "learning_rate": 5.219656373981158e-05,
      "loss": 2.3142824172973633,
      "memory(GiB)": 77.56,
      "step": 56720,
      "token_acc": 0.5101351351351351,
      "train_speed(iter/s)": 1.438902
    },
    {
      "epoch": 2.4302729103294634,
      "grad_norm": 5.306214332580566,
      "learning_rate": 5.218984044411126e-05,
      "loss": 2.3295936584472656,
      "memory(GiB)": 77.56,
      "step": 56725,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.438918
    },
    {
      "epoch": 2.430487125658712,
      "grad_norm": 4.491667747497559,
      "learning_rate": 5.218311710873992e-05,
      "loss": 2.4540481567382812,
      "memory(GiB)": 77.56,
      "step": 56730,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.438947
    },
    {
      "epoch": 2.430701340987961,
      "grad_norm": 6.126719951629639,
      "learning_rate": 5.2176393733819376e-05,
      "loss": 2.548937225341797,
      "memory(GiB)": 77.56,
      "step": 56735,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.438955
    },
    {
      "epoch": 2.4309155563172102,
      "grad_norm": 5.413206100463867,
      "learning_rate": 5.2169670319471406e-05,
      "loss": 2.453196144104004,
      "memory(GiB)": 77.56,
      "step": 56740,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.438978
    },
    {
      "epoch": 2.431129771646459,
      "grad_norm": 4.76425838470459,
      "learning_rate": 5.216294686581783e-05,
      "loss": 2.3705583572387696,
      "memory(GiB)": 77.56,
      "step": 56745,
      "token_acc": 0.49514563106796117,
      "train_speed(iter/s)": 1.438976
    },
    {
      "epoch": 2.431343986975708,
      "grad_norm": 6.396388053894043,
      "learning_rate": 5.215622337298044e-05,
      "loss": 2.5731929779052733,
      "memory(GiB)": 77.56,
      "step": 56750,
      "token_acc": 0.48135593220338985,
      "train_speed(iter/s)": 1.438967
    },
    {
      "epoch": 2.431558202304957,
      "grad_norm": 6.104854106903076,
      "learning_rate": 5.214949984108104e-05,
      "loss": 2.5788434982299804,
      "memory(GiB)": 77.56,
      "step": 56755,
      "token_acc": 0.44591029023746703,
      "train_speed(iter/s)": 1.438978
    },
    {
      "epoch": 2.431772417634206,
      "grad_norm": 4.484294414520264,
      "learning_rate": 5.214277627024144e-05,
      "loss": 2.128470230102539,
      "memory(GiB)": 77.56,
      "step": 56760,
      "token_acc": 0.5089605734767025,
      "train_speed(iter/s)": 1.439011
    },
    {
      "epoch": 2.4319866329634547,
      "grad_norm": 7.305736064910889,
      "learning_rate": 5.2136052660583444e-05,
      "loss": 2.4432231903076174,
      "memory(GiB)": 77.56,
      "step": 56765,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.43901
    },
    {
      "epoch": 2.432200848292704,
      "grad_norm": 6.269150733947754,
      "learning_rate": 5.212932901222883e-05,
      "loss": 2.531831169128418,
      "memory(GiB)": 77.56,
      "step": 56770,
      "token_acc": 0.4969512195121951,
      "train_speed(iter/s)": 1.439
    },
    {
      "epoch": 2.432415063621953,
      "grad_norm": 4.121421813964844,
      "learning_rate": 5.212260532529945e-05,
      "loss": 2.3903039932250976,
      "memory(GiB)": 77.56,
      "step": 56775,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.43901
    },
    {
      "epoch": 2.4326292789512016,
      "grad_norm": 4.335596561431885,
      "learning_rate": 5.211588159991707e-05,
      "loss": 2.445465087890625,
      "memory(GiB)": 77.56,
      "step": 56780,
      "token_acc": 0.4599406528189911,
      "train_speed(iter/s)": 1.438985
    },
    {
      "epoch": 2.432843494280451,
      "grad_norm": 4.750160217285156,
      "learning_rate": 5.210915783620349e-05,
      "loss": 2.7006240844726563,
      "memory(GiB)": 77.56,
      "step": 56785,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.439003
    },
    {
      "epoch": 2.4330577096096997,
      "grad_norm": 4.748763084411621,
      "learning_rate": 5.2102434034280566e-05,
      "loss": 2.5845464706420898,
      "memory(GiB)": 77.56,
      "step": 56790,
      "token_acc": 0.4774011299435028,
      "train_speed(iter/s)": 1.438998
    },
    {
      "epoch": 2.4332719249389485,
      "grad_norm": 4.500877380371094,
      "learning_rate": 5.2095710194270067e-05,
      "loss": 2.089221954345703,
      "memory(GiB)": 77.56,
      "step": 56795,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.439039
    },
    {
      "epoch": 2.4334861402681978,
      "grad_norm": 5.059268951416016,
      "learning_rate": 5.208898631629381e-05,
      "loss": 2.2982500076293944,
      "memory(GiB)": 77.56,
      "step": 56800,
      "token_acc": 0.5125448028673835,
      "train_speed(iter/s)": 1.439033
    },
    {
      "epoch": 2.4337003555974466,
      "grad_norm": 6.589064598083496,
      "learning_rate": 5.208226240047362e-05,
      "loss": 2.342428970336914,
      "memory(GiB)": 77.56,
      "step": 56805,
      "token_acc": 0.5433070866141733,
      "train_speed(iter/s)": 1.439044
    },
    {
      "epoch": 2.4339145709266954,
      "grad_norm": 7.438954830169678,
      "learning_rate": 5.207553844693128e-05,
      "loss": 2.438463592529297,
      "memory(GiB)": 77.56,
      "step": 56810,
      "token_acc": 0.44155844155844154,
      "train_speed(iter/s)": 1.439056
    },
    {
      "epoch": 2.4341287862559446,
      "grad_norm": 5.0491108894348145,
      "learning_rate": 5.206881445578861e-05,
      "loss": 2.3306510925292967,
      "memory(GiB)": 77.56,
      "step": 56815,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.439014
    },
    {
      "epoch": 2.4343430015851935,
      "grad_norm": 4.765693664550781,
      "learning_rate": 5.206209042716742e-05,
      "loss": 2.502008628845215,
      "memory(GiB)": 77.56,
      "step": 56820,
      "token_acc": 0.4983277591973244,
      "train_speed(iter/s)": 1.439021
    },
    {
      "epoch": 2.4345572169144423,
      "grad_norm": 5.881290912628174,
      "learning_rate": 5.205536636118955e-05,
      "loss": 2.302350616455078,
      "memory(GiB)": 77.56,
      "step": 56825,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.43904
    },
    {
      "epoch": 2.4347714322436915,
      "grad_norm": 5.645346164703369,
      "learning_rate": 5.204864225797676e-05,
      "loss": 2.7165205001831056,
      "memory(GiB)": 77.56,
      "step": 56830,
      "token_acc": 0.44785276073619634,
      "train_speed(iter/s)": 1.439037
    },
    {
      "epoch": 2.4349856475729403,
      "grad_norm": 5.8589653968811035,
      "learning_rate": 5.204191811765092e-05,
      "loss": 2.419597625732422,
      "memory(GiB)": 77.56,
      "step": 56835,
      "token_acc": 0.49074074074074076,
      "train_speed(iter/s)": 1.439038
    },
    {
      "epoch": 2.435199862902189,
      "grad_norm": 6.1219024658203125,
      "learning_rate": 5.203519394033382e-05,
      "loss": 2.6928516387939454,
      "memory(GiB)": 77.56,
      "step": 56840,
      "token_acc": 0.43795620437956206,
      "train_speed(iter/s)": 1.439057
    },
    {
      "epoch": 2.4354140782314384,
      "grad_norm": 8.250514030456543,
      "learning_rate": 5.202846972614726e-05,
      "loss": 2.279958724975586,
      "memory(GiB)": 77.56,
      "step": 56845,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.439059
    },
    {
      "epoch": 2.435628293560687,
      "grad_norm": 4.9896721839904785,
      "learning_rate": 5.2021745475213076e-05,
      "loss": 2.6677433013916017,
      "memory(GiB)": 77.56,
      "step": 56850,
      "token_acc": 0.46540880503144655,
      "train_speed(iter/s)": 1.439051
    },
    {
      "epoch": 2.435842508889936,
      "grad_norm": 5.202160835266113,
      "learning_rate": 5.201502118765307e-05,
      "loss": 2.9514438629150392,
      "memory(GiB)": 77.56,
      "step": 56855,
      "token_acc": 0.436950146627566,
      "train_speed(iter/s)": 1.439084
    },
    {
      "epoch": 2.4360567242191853,
      "grad_norm": 5.538449764251709,
      "learning_rate": 5.200829686358906e-05,
      "loss": 2.3525764465332033,
      "memory(GiB)": 77.56,
      "step": 56860,
      "token_acc": 0.44912280701754387,
      "train_speed(iter/s)": 1.439094
    },
    {
      "epoch": 2.436270939548434,
      "grad_norm": 4.442500114440918,
      "learning_rate": 5.2001572503142873e-05,
      "loss": 2.54138069152832,
      "memory(GiB)": 77.56,
      "step": 56865,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.439052
    },
    {
      "epoch": 2.436485154877683,
      "grad_norm": 6.37665319442749,
      "learning_rate": 5.1994848106436334e-05,
      "loss": 2.4878376007080076,
      "memory(GiB)": 77.56,
      "step": 56870,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.439055
    },
    {
      "epoch": 2.436699370206932,
      "grad_norm": 5.775146961212158,
      "learning_rate": 5.198812367359123e-05,
      "loss": 2.33713436126709,
      "memory(GiB)": 77.56,
      "step": 56875,
      "token_acc": 0.5469255663430421,
      "train_speed(iter/s)": 1.43904
    },
    {
      "epoch": 2.436913585536181,
      "grad_norm": 6.4485344886779785,
      "learning_rate": 5.198139920472942e-05,
      "loss": 2.4708410263061524,
      "memory(GiB)": 77.56,
      "step": 56880,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.439056
    },
    {
      "epoch": 2.43712780086543,
      "grad_norm": 5.370219707489014,
      "learning_rate": 5.1974674699972684e-05,
      "loss": 2.2273605346679686,
      "memory(GiB)": 77.56,
      "step": 56885,
      "token_acc": 0.5338645418326693,
      "train_speed(iter/s)": 1.439077
    },
    {
      "epoch": 2.437342016194679,
      "grad_norm": 5.101024627685547,
      "learning_rate": 5.196795015944288e-05,
      "loss": 2.219680404663086,
      "memory(GiB)": 77.56,
      "step": 56890,
      "token_acc": 0.501432664756447,
      "train_speed(iter/s)": 1.439104
    },
    {
      "epoch": 2.437556231523928,
      "grad_norm": 6.431314468383789,
      "learning_rate": 5.19612255832618e-05,
      "loss": 2.4655418395996094,
      "memory(GiB)": 77.56,
      "step": 56895,
      "token_acc": 0.5190839694656488,
      "train_speed(iter/s)": 1.439112
    },
    {
      "epoch": 2.4377704468531767,
      "grad_norm": 5.0001373291015625,
      "learning_rate": 5.195450097155128e-05,
      "loss": 2.5330574035644533,
      "memory(GiB)": 77.56,
      "step": 56900,
      "token_acc": 0.47126436781609193,
      "train_speed(iter/s)": 1.4391
    },
    {
      "epoch": 2.437984662182426,
      "grad_norm": 5.0970940589904785,
      "learning_rate": 5.194777632443315e-05,
      "loss": 2.1877840042114256,
      "memory(GiB)": 77.56,
      "step": 56905,
      "token_acc": 0.49673202614379086,
      "train_speed(iter/s)": 1.439057
    },
    {
      "epoch": 2.4381988775116747,
      "grad_norm": 5.015025615692139,
      "learning_rate": 5.194105164202924e-05,
      "loss": 2.518899917602539,
      "memory(GiB)": 77.56,
      "step": 56910,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.43904
    },
    {
      "epoch": 2.4384130928409236,
      "grad_norm": 5.5324883460998535,
      "learning_rate": 5.1934326924461326e-05,
      "loss": 2.56801700592041,
      "memory(GiB)": 77.56,
      "step": 56915,
      "token_acc": 0.47181008902077154,
      "train_speed(iter/s)": 1.43906
    },
    {
      "epoch": 2.438627308170173,
      "grad_norm": 6.703329086303711,
      "learning_rate": 5.192760217185129e-05,
      "loss": 2.689248275756836,
      "memory(GiB)": 77.56,
      "step": 56920,
      "token_acc": 0.4862068965517241,
      "train_speed(iter/s)": 1.43907
    },
    {
      "epoch": 2.4388415234994216,
      "grad_norm": 5.418503284454346,
      "learning_rate": 5.192087738432092e-05,
      "loss": 2.197000503540039,
      "memory(GiB)": 77.56,
      "step": 56925,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.439047
    },
    {
      "epoch": 2.4390557388286704,
      "grad_norm": 8.2334623336792,
      "learning_rate": 5.191415256199205e-05,
      "loss": 2.6109909057617187,
      "memory(GiB)": 77.56,
      "step": 56930,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.43908
    },
    {
      "epoch": 2.4392699541579197,
      "grad_norm": 5.991984844207764,
      "learning_rate": 5.190742770498652e-05,
      "loss": 2.408322334289551,
      "memory(GiB)": 77.56,
      "step": 56935,
      "token_acc": 0.47678018575851394,
      "train_speed(iter/s)": 1.439097
    },
    {
      "epoch": 2.4394841694871685,
      "grad_norm": 8.74606704711914,
      "learning_rate": 5.190070281342615e-05,
      "loss": 2.6613529205322264,
      "memory(GiB)": 77.56,
      "step": 56940,
      "token_acc": 0.4576719576719577,
      "train_speed(iter/s)": 1.4391
    },
    {
      "epoch": 2.4396983848164173,
      "grad_norm": 4.8893022537231445,
      "learning_rate": 5.189397788743275e-05,
      "loss": 2.3952220916748046,
      "memory(GiB)": 77.56,
      "step": 56945,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.43911
    },
    {
      "epoch": 2.4399126001456666,
      "grad_norm": 5.606250286102295,
      "learning_rate": 5.188725292712818e-05,
      "loss": 2.571638488769531,
      "memory(GiB)": 77.56,
      "step": 56950,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.439127
    },
    {
      "epoch": 2.4401268154749154,
      "grad_norm": 6.874665260314941,
      "learning_rate": 5.188052793263426e-05,
      "loss": 2.331051254272461,
      "memory(GiB)": 77.56,
      "step": 56955,
      "token_acc": 0.5325670498084292,
      "train_speed(iter/s)": 1.43912
    },
    {
      "epoch": 2.440341030804164,
      "grad_norm": 5.383266448974609,
      "learning_rate": 5.1873802904072786e-05,
      "loss": 2.647352600097656,
      "memory(GiB)": 77.56,
      "step": 56960,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.439116
    },
    {
      "epoch": 2.4405552461334135,
      "grad_norm": 4.835541725158691,
      "learning_rate": 5.1867077841565635e-05,
      "loss": 2.5778778076171873,
      "memory(GiB)": 77.56,
      "step": 56965,
      "token_acc": 0.49828178694158076,
      "train_speed(iter/s)": 1.43916
    },
    {
      "epoch": 2.4407694614626623,
      "grad_norm": 5.5413079261779785,
      "learning_rate": 5.186035274523461e-05,
      "loss": 2.4015180587768556,
      "memory(GiB)": 77.56,
      "step": 56970,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.439156
    },
    {
      "epoch": 2.440983676791911,
      "grad_norm": 4.444792747497559,
      "learning_rate": 5.1853627615201536e-05,
      "loss": 2.411921501159668,
      "memory(GiB)": 77.56,
      "step": 56975,
      "token_acc": 0.4733893557422969,
      "train_speed(iter/s)": 1.439144
    },
    {
      "epoch": 2.4411978921211603,
      "grad_norm": 7.380198001861572,
      "learning_rate": 5.184690245158829e-05,
      "loss": 2.441391372680664,
      "memory(GiB)": 77.56,
      "step": 56980,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.439137
    },
    {
      "epoch": 2.441412107450409,
      "grad_norm": 4.234613418579102,
      "learning_rate": 5.1840177254516666e-05,
      "loss": 2.4302005767822266,
      "memory(GiB)": 77.56,
      "step": 56985,
      "token_acc": 0.45671641791044776,
      "train_speed(iter/s)": 1.439167
    },
    {
      "epoch": 2.441626322779658,
      "grad_norm": 5.636434078216553,
      "learning_rate": 5.183345202410849e-05,
      "loss": 2.2529296875,
      "memory(GiB)": 77.56,
      "step": 56990,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.439184
    },
    {
      "epoch": 2.441840538108907,
      "grad_norm": 4.551118850708008,
      "learning_rate": 5.182672676048561e-05,
      "loss": 2.4376110076904296,
      "memory(GiB)": 77.56,
      "step": 56995,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.43918
    },
    {
      "epoch": 2.442054753438156,
      "grad_norm": 5.243435382843018,
      "learning_rate": 5.182000146376986e-05,
      "loss": 2.241717529296875,
      "memory(GiB)": 77.56,
      "step": 57000,
      "token_acc": 0.5340909090909091,
      "train_speed(iter/s)": 1.439178
    },
    {
      "epoch": 2.442054753438156,
      "eval_loss": 2.3264853954315186,
      "eval_runtime": 14.1731,
      "eval_samples_per_second": 7.056,
      "eval_steps_per_second": 7.056,
      "eval_token_acc": 0.4582763337893297,
      "step": 57000
    },
    {
      "epoch": 2.442268968767405,
      "grad_norm": 5.387640476226807,
      "learning_rate": 5.181327613408309e-05,
      "loss": 2.4150400161743164,
      "memory(GiB)": 77.56,
      "step": 57005,
      "token_acc": 0.4698318496538081,
      "train_speed(iter/s)": 1.438633
    },
    {
      "epoch": 2.442483184096654,
      "grad_norm": 5.350733757019043,
      "learning_rate": 5.1806550771547115e-05,
      "loss": 2.360303497314453,
      "memory(GiB)": 77.56,
      "step": 57010,
      "token_acc": 0.5196078431372549,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.442697399425903,
      "grad_norm": 4.9536943435668945,
      "learning_rate": 5.179982537628378e-05,
      "loss": 2.185806655883789,
      "memory(GiB)": 77.56,
      "step": 57015,
      "token_acc": 0.5490909090909091,
      "train_speed(iter/s)": 1.43866
    },
    {
      "epoch": 2.4429116147551517,
      "grad_norm": 4.514573097229004,
      "learning_rate": 5.1793099948414925e-05,
      "loss": 2.4682483673095703,
      "memory(GiB)": 77.56,
      "step": 57020,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.438666
    },
    {
      "epoch": 2.443125830084401,
      "grad_norm": 5.264124870300293,
      "learning_rate": 5.1786374488062375e-05,
      "loss": 2.4311029434204103,
      "memory(GiB)": 77.56,
      "step": 57025,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.438658
    },
    {
      "epoch": 2.44334004541365,
      "grad_norm": 9.807960510253906,
      "learning_rate": 5.1779648995347975e-05,
      "loss": 2.5561866760253906,
      "memory(GiB)": 77.56,
      "step": 57030,
      "token_acc": 0.4440993788819876,
      "train_speed(iter/s)": 1.438642
    },
    {
      "epoch": 2.4435542607428986,
      "grad_norm": 5.882599353790283,
      "learning_rate": 5.177292347039358e-05,
      "loss": 2.78738956451416,
      "memory(GiB)": 77.56,
      "step": 57035,
      "token_acc": 0.43137254901960786,
      "train_speed(iter/s)": 1.43862
    },
    {
      "epoch": 2.443768476072148,
      "grad_norm": 5.670618534088135,
      "learning_rate": 5.176619791332099e-05,
      "loss": 2.327910232543945,
      "memory(GiB)": 77.56,
      "step": 57040,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 2.4439826914013967,
      "grad_norm": 4.727508068084717,
      "learning_rate": 5.175947232425207e-05,
      "loss": 2.5686370849609377,
      "memory(GiB)": 77.56,
      "step": 57045,
      "token_acc": 0.5050505050505051,
      "train_speed(iter/s)": 1.43864
    },
    {
      "epoch": 2.4441969067306455,
      "grad_norm": 7.624876976013184,
      "learning_rate": 5.1752746703308664e-05,
      "loss": 2.6940095901489256,
      "memory(GiB)": 77.56,
      "step": 57050,
      "token_acc": 0.45244956772334294,
      "train_speed(iter/s)": 1.4386
    },
    {
      "epoch": 2.4444111220598947,
      "grad_norm": 6.475161075592041,
      "learning_rate": 5.174602105061262e-05,
      "loss": 2.572902297973633,
      "memory(GiB)": 77.56,
      "step": 57055,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.438609
    },
    {
      "epoch": 2.4446253373891436,
      "grad_norm": 4.965032577514648,
      "learning_rate": 5.1739295366285745e-05,
      "loss": 2.3567081451416017,
      "memory(GiB)": 77.56,
      "step": 57060,
      "token_acc": 0.49640287769784175,
      "train_speed(iter/s)": 1.438601
    },
    {
      "epoch": 2.4448395527183924,
      "grad_norm": 7.131523609161377,
      "learning_rate": 5.173256965044991e-05,
      "loss": 2.5855518341064454,
      "memory(GiB)": 77.56,
      "step": 57065,
      "token_acc": 0.4559748427672956,
      "train_speed(iter/s)": 1.438606
    },
    {
      "epoch": 2.4450537680476416,
      "grad_norm": 5.767808437347412,
      "learning_rate": 5.1725843903226966e-05,
      "loss": 2.420552062988281,
      "memory(GiB)": 77.56,
      "step": 57070,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.438609
    },
    {
      "epoch": 2.4452679833768904,
      "grad_norm": 6.030223846435547,
      "learning_rate": 5.171911812473872e-05,
      "loss": 2.4342273712158202,
      "memory(GiB)": 77.56,
      "step": 57075,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.438636
    },
    {
      "epoch": 2.4454821987061393,
      "grad_norm": 9.87091064453125,
      "learning_rate": 5.171239231510704e-05,
      "loss": 2.0547002792358398,
      "memory(GiB)": 77.56,
      "step": 57080,
      "token_acc": 0.5568181818181818,
      "train_speed(iter/s)": 1.438622
    },
    {
      "epoch": 2.4456964140353885,
      "grad_norm": 4.279675483703613,
      "learning_rate": 5.1705666474453785e-05,
      "loss": 2.2895599365234376,
      "memory(GiB)": 77.56,
      "step": 57085,
      "token_acc": 0.5063291139240507,
      "train_speed(iter/s)": 1.438598
    },
    {
      "epoch": 2.4459106293646373,
      "grad_norm": 7.536383628845215,
      "learning_rate": 5.1698940602900756e-05,
      "loss": 2.5207191467285157,
      "memory(GiB)": 77.56,
      "step": 57090,
      "token_acc": 0.452,
      "train_speed(iter/s)": 1.438621
    },
    {
      "epoch": 2.446124844693886,
      "grad_norm": 4.840969085693359,
      "learning_rate": 5.169221470056984e-05,
      "loss": 2.4955591201782226,
      "memory(GiB)": 77.56,
      "step": 57095,
      "token_acc": 0.46319018404907975,
      "train_speed(iter/s)": 1.438623
    },
    {
      "epoch": 2.4463390600231354,
      "grad_norm": 6.287506580352783,
      "learning_rate": 5.168548876758288e-05,
      "loss": 2.747912788391113,
      "memory(GiB)": 77.56,
      "step": 57100,
      "token_acc": 0.43288590604026844,
      "train_speed(iter/s)": 1.438627
    },
    {
      "epoch": 2.446553275352384,
      "grad_norm": 5.823217391967773,
      "learning_rate": 5.1678762804061685e-05,
      "loss": 2.7979211807250977,
      "memory(GiB)": 77.56,
      "step": 57105,
      "token_acc": 0.44285714285714284,
      "train_speed(iter/s)": 1.438647
    },
    {
      "epoch": 2.446767490681633,
      "grad_norm": 5.685962677001953,
      "learning_rate": 5.167203681012813e-05,
      "loss": 2.4895488739013674,
      "memory(GiB)": 77.56,
      "step": 57110,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.438635
    },
    {
      "epoch": 2.4469817060108823,
      "grad_norm": 6.21130895614624,
      "learning_rate": 5.1665310785904066e-05,
      "loss": 2.4648815155029298,
      "memory(GiB)": 77.56,
      "step": 57115,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.438612
    },
    {
      "epoch": 2.447195921340131,
      "grad_norm": 6.889520645141602,
      "learning_rate": 5.165858473151133e-05,
      "loss": 2.6206830978393554,
      "memory(GiB)": 77.56,
      "step": 57120,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438596
    },
    {
      "epoch": 2.44741013666938,
      "grad_norm": 5.61447811126709,
      "learning_rate": 5.165185864707178e-05,
      "loss": 2.7210128784179686,
      "memory(GiB)": 77.56,
      "step": 57125,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.438527
    },
    {
      "epoch": 2.447624351998629,
      "grad_norm": 5.406766891479492,
      "learning_rate": 5.164513253270727e-05,
      "loss": 2.4325815200805665,
      "memory(GiB)": 77.56,
      "step": 57130,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.438558
    },
    {
      "epoch": 2.447838567327878,
      "grad_norm": 15.234235763549805,
      "learning_rate": 5.163840638853963e-05,
      "loss": 2.159321403503418,
      "memory(GiB)": 77.56,
      "step": 57135,
      "token_acc": 0.5442622950819672,
      "train_speed(iter/s)": 1.438562
    },
    {
      "epoch": 2.448052782657127,
      "grad_norm": 5.185170650482178,
      "learning_rate": 5.163168021469073e-05,
      "loss": 2.4400049209594727,
      "memory(GiB)": 77.56,
      "step": 57140,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.438554
    },
    {
      "epoch": 2.448266997986376,
      "grad_norm": 4.750624179840088,
      "learning_rate": 5.16249540112824e-05,
      "loss": 2.6198116302490235,
      "memory(GiB)": 77.56,
      "step": 57145,
      "token_acc": 0.45484949832775917,
      "train_speed(iter/s)": 1.438551
    },
    {
      "epoch": 2.448481213315625,
      "grad_norm": 5.566971778869629,
      "learning_rate": 5.161822777843651e-05,
      "loss": 2.4444602966308593,
      "memory(GiB)": 77.56,
      "step": 57150,
      "token_acc": 0.48253968253968255,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 2.4486954286448737,
      "grad_norm": 5.943056583404541,
      "learning_rate": 5.1611501516274904e-05,
      "loss": 2.2202121734619142,
      "memory(GiB)": 77.56,
      "step": 57155,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.43857
    },
    {
      "epoch": 2.448909643974123,
      "grad_norm": 6.960949420928955,
      "learning_rate": 5.160477522491943e-05,
      "loss": 2.309294509887695,
      "memory(GiB)": 77.56,
      "step": 57160,
      "token_acc": 0.5078740157480315,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 2.4491238593033717,
      "grad_norm": 6.0811381340026855,
      "learning_rate": 5.159804890449196e-05,
      "loss": 2.507560920715332,
      "memory(GiB)": 77.56,
      "step": 57165,
      "token_acc": 0.4659090909090909,
      "train_speed(iter/s)": 1.438583
    },
    {
      "epoch": 2.4493380746326205,
      "grad_norm": 5.529671669006348,
      "learning_rate": 5.159132255511434e-05,
      "loss": 2.1746746063232423,
      "memory(GiB)": 77.56,
      "step": 57170,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 2.44955228996187,
      "grad_norm": 3.9784586429595947,
      "learning_rate": 5.15845961769084e-05,
      "loss": 2.395619583129883,
      "memory(GiB)": 77.56,
      "step": 57175,
      "token_acc": 0.4622356495468278,
      "train_speed(iter/s)": 1.438594
    },
    {
      "epoch": 2.4497665052911186,
      "grad_norm": 5.22333288192749,
      "learning_rate": 5.157786976999602e-05,
      "loss": 2.3220149993896486,
      "memory(GiB)": 77.56,
      "step": 57180,
      "token_acc": 0.5282258064516129,
      "train_speed(iter/s)": 1.438581
    },
    {
      "epoch": 2.4499807206203674,
      "grad_norm": 6.703789234161377,
      "learning_rate": 5.157114333449906e-05,
      "loss": 2.1752740859985353,
      "memory(GiB)": 77.56,
      "step": 57185,
      "token_acc": 0.5376712328767124,
      "train_speed(iter/s)": 1.43858
    },
    {
      "epoch": 2.4501949359496167,
      "grad_norm": 7.221778869628906,
      "learning_rate": 5.1564416870539346e-05,
      "loss": 2.0189912796020506,
      "memory(GiB)": 77.56,
      "step": 57190,
      "token_acc": 0.5091463414634146,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 2.4504091512788655,
      "grad_norm": 3.823230266571045,
      "learning_rate": 5.155769037823876e-05,
      "loss": 2.3497365951538085,
      "memory(GiB)": 77.56,
      "step": 57195,
      "token_acc": 0.47186147186147187,
      "train_speed(iter/s)": 1.438614
    },
    {
      "epoch": 2.4506233666081143,
      "grad_norm": 6.960144996643066,
      "learning_rate": 5.155096385771917e-05,
      "loss": 2.5806880950927735,
      "memory(GiB)": 77.56,
      "step": 57200,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.43865
    },
    {
      "epoch": 2.4508375819373636,
      "grad_norm": 5.065337657928467,
      "learning_rate": 5.1544237309102395e-05,
      "loss": 2.4673130035400392,
      "memory(GiB)": 77.56,
      "step": 57205,
      "token_acc": 0.503125,
      "train_speed(iter/s)": 1.438626
    },
    {
      "epoch": 2.4510517972666124,
      "grad_norm": 4.6179118156433105,
      "learning_rate": 5.153751073251032e-05,
      "loss": 2.3746482849121096,
      "memory(GiB)": 77.56,
      "step": 57210,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.438648
    },
    {
      "epoch": 2.451266012595861,
      "grad_norm": 4.876397609710693,
      "learning_rate": 5.15307841280648e-05,
      "loss": 2.256892967224121,
      "memory(GiB)": 77.56,
      "step": 57215,
      "token_acc": 0.5077399380804953,
      "train_speed(iter/s)": 1.438686
    },
    {
      "epoch": 2.4514802279251104,
      "grad_norm": 5.341527938842773,
      "learning_rate": 5.152405749588768e-05,
      "loss": 2.0741554260253907,
      "memory(GiB)": 77.56,
      "step": 57220,
      "token_acc": 0.5080645161290323,
      "train_speed(iter/s)": 1.43866
    },
    {
      "epoch": 2.4516944432543593,
      "grad_norm": 4.602362155914307,
      "learning_rate": 5.151733083610083e-05,
      "loss": 2.2884637832641603,
      "memory(GiB)": 77.56,
      "step": 57225,
      "token_acc": 0.5335820895522388,
      "train_speed(iter/s)": 1.438687
    },
    {
      "epoch": 2.451908658583608,
      "grad_norm": 4.572879791259766,
      "learning_rate": 5.15106041488261e-05,
      "loss": 2.3112060546875,
      "memory(GiB)": 77.56,
      "step": 57230,
      "token_acc": 0.5379939209726444,
      "train_speed(iter/s)": 1.438686
    },
    {
      "epoch": 2.4521228739128573,
      "grad_norm": 5.744421482086182,
      "learning_rate": 5.1503877434185366e-05,
      "loss": 2.58641357421875,
      "memory(GiB)": 77.56,
      "step": 57235,
      "token_acc": 0.4503105590062112,
      "train_speed(iter/s)": 1.438704
    },
    {
      "epoch": 2.452337089242106,
      "grad_norm": 4.397494792938232,
      "learning_rate": 5.149715069230049e-05,
      "loss": 2.3761240005493165,
      "memory(GiB)": 77.56,
      "step": 57240,
      "token_acc": 0.4783950617283951,
      "train_speed(iter/s)": 1.438708
    },
    {
      "epoch": 2.452551304571355,
      "grad_norm": 5.352940082550049,
      "learning_rate": 5.149042392329333e-05,
      "loss": 2.3313854217529295,
      "memory(GiB)": 77.56,
      "step": 57245,
      "token_acc": 0.49032258064516127,
      "train_speed(iter/s)": 1.438723
    },
    {
      "epoch": 2.452765519900604,
      "grad_norm": 6.408934593200684,
      "learning_rate": 5.148369712728572e-05,
      "loss": 2.553639221191406,
      "memory(GiB)": 77.56,
      "step": 57250,
      "token_acc": 0.45126353790613716,
      "train_speed(iter/s)": 1.438731
    },
    {
      "epoch": 2.452979735229853,
      "grad_norm": 4.5913238525390625,
      "learning_rate": 5.1476970304399565e-05,
      "loss": 2.281085968017578,
      "memory(GiB)": 77.56,
      "step": 57255,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438725
    },
    {
      "epoch": 2.453193950559102,
      "grad_norm": 5.944769859313965,
      "learning_rate": 5.1470243454756694e-05,
      "loss": 2.485824966430664,
      "memory(GiB)": 77.56,
      "step": 57260,
      "token_acc": 0.493006993006993,
      "train_speed(iter/s)": 1.438738
    },
    {
      "epoch": 2.453408165888351,
      "grad_norm": 5.902571201324463,
      "learning_rate": 5.146351657847898e-05,
      "loss": 2.5209325790405273,
      "memory(GiB)": 77.56,
      "step": 57265,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.438736
    },
    {
      "epoch": 2.4536223812176,
      "grad_norm": 5.643834590911865,
      "learning_rate": 5.14567896756883e-05,
      "loss": 2.4684677124023438,
      "memory(GiB)": 77.56,
      "step": 57270,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.438734
    },
    {
      "epoch": 2.4538365965468487,
      "grad_norm": 5.017353534698486,
      "learning_rate": 5.145006274650652e-05,
      "loss": 2.446546936035156,
      "memory(GiB)": 77.56,
      "step": 57275,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.438725
    },
    {
      "epoch": 2.454050811876098,
      "grad_norm": 8.103410720825195,
      "learning_rate": 5.144333579105547e-05,
      "loss": 2.547798919677734,
      "memory(GiB)": 77.56,
      "step": 57280,
      "token_acc": 0.4393939393939394,
      "train_speed(iter/s)": 1.43872
    },
    {
      "epoch": 2.454265027205347,
      "grad_norm": 5.171634674072266,
      "learning_rate": 5.143660880945705e-05,
      "loss": 2.5604633331298827,
      "memory(GiB)": 77.56,
      "step": 57285,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.438728
    },
    {
      "epoch": 2.4544792425345956,
      "grad_norm": 8.173238754272461,
      "learning_rate": 5.1429881801833116e-05,
      "loss": 2.5359878540039062,
      "memory(GiB)": 77.56,
      "step": 57290,
      "token_acc": 0.4810126582278481,
      "train_speed(iter/s)": 1.438712
    },
    {
      "epoch": 2.454693457863845,
      "grad_norm": 5.89133882522583,
      "learning_rate": 5.1423154768305524e-05,
      "loss": 2.49134521484375,
      "memory(GiB)": 77.56,
      "step": 57295,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.438722
    },
    {
      "epoch": 2.4549076731930937,
      "grad_norm": 4.935211181640625,
      "learning_rate": 5.1416427708996154e-05,
      "loss": 2.411406135559082,
      "memory(GiB)": 77.56,
      "step": 57300,
      "token_acc": 0.48089171974522293,
      "train_speed(iter/s)": 1.43874
    },
    {
      "epoch": 2.4551218885223425,
      "grad_norm": 6.120079040527344,
      "learning_rate": 5.1409700624026855e-05,
      "loss": 2.2102788925170898,
      "memory(GiB)": 77.56,
      "step": 57305,
      "token_acc": 0.5442477876106194,
      "train_speed(iter/s)": 1.438738
    },
    {
      "epoch": 2.4553361038515917,
      "grad_norm": 5.650693416595459,
      "learning_rate": 5.1402973513519527e-05,
      "loss": 2.5443172454833984,
      "memory(GiB)": 77.56,
      "step": 57310,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.438761
    },
    {
      "epoch": 2.4555503191808405,
      "grad_norm": 5.473184585571289,
      "learning_rate": 5.139624637759601e-05,
      "loss": 2.262398529052734,
      "memory(GiB)": 77.56,
      "step": 57315,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.43876
    },
    {
      "epoch": 2.4557645345100894,
      "grad_norm": 5.1242194175720215,
      "learning_rate": 5.138951921637817e-05,
      "loss": 2.3731449127197264,
      "memory(GiB)": 77.56,
      "step": 57320,
      "token_acc": 0.5127388535031847,
      "train_speed(iter/s)": 1.438765
    },
    {
      "epoch": 2.4559787498393386,
      "grad_norm": 8.556614875793457,
      "learning_rate": 5.1382792029987904e-05,
      "loss": 2.6999895095825197,
      "memory(GiB)": 77.56,
      "step": 57325,
      "token_acc": 0.44673539518900346,
      "train_speed(iter/s)": 1.438797
    },
    {
      "epoch": 2.4561929651685874,
      "grad_norm": 5.725161075592041,
      "learning_rate": 5.137606481854705e-05,
      "loss": 2.1397125244140627,
      "memory(GiB)": 77.56,
      "step": 57330,
      "token_acc": 0.555045871559633,
      "train_speed(iter/s)": 1.4388
    },
    {
      "epoch": 2.4564071804978362,
      "grad_norm": 4.597769737243652,
      "learning_rate": 5.136933758217749e-05,
      "loss": 2.4590587615966797,
      "memory(GiB)": 77.56,
      "step": 57335,
      "token_acc": 0.5127272727272727,
      "train_speed(iter/s)": 1.438797
    },
    {
      "epoch": 2.4566213958270855,
      "grad_norm": 5.495504379272461,
      "learning_rate": 5.13626103210011e-05,
      "loss": 2.698875617980957,
      "memory(GiB)": 77.56,
      "step": 57340,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.438833
    },
    {
      "epoch": 2.4568356111563343,
      "grad_norm": 5.055111408233643,
      "learning_rate": 5.135588303513975e-05,
      "loss": 2.6068859100341797,
      "memory(GiB)": 77.56,
      "step": 57345,
      "token_acc": 0.5251572327044025,
      "train_speed(iter/s)": 1.438833
    },
    {
      "epoch": 2.457049826485583,
      "grad_norm": 6.525391101837158,
      "learning_rate": 5.1349155724715294e-05,
      "loss": 2.4812475204467774,
      "memory(GiB)": 77.56,
      "step": 57350,
      "token_acc": 0.49823321554770317,
      "train_speed(iter/s)": 1.438851
    },
    {
      "epoch": 2.4572640418148324,
      "grad_norm": 5.498703956604004,
      "learning_rate": 5.1342428389849626e-05,
      "loss": 2.586172866821289,
      "memory(GiB)": 77.56,
      "step": 57355,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.438869
    },
    {
      "epoch": 2.457478257144081,
      "grad_norm": 5.428562641143799,
      "learning_rate": 5.133570103066462e-05,
      "loss": 2.24481201171875,
      "memory(GiB)": 77.56,
      "step": 57360,
      "token_acc": 0.486404833836858,
      "train_speed(iter/s)": 1.438866
    },
    {
      "epoch": 2.45769247247333,
      "grad_norm": 5.332465648651123,
      "learning_rate": 5.1328973647282116e-05,
      "loss": 2.378965377807617,
      "memory(GiB)": 77.56,
      "step": 57365,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.438882
    },
    {
      "epoch": 2.4579066878025793,
      "grad_norm": 6.097894668579102,
      "learning_rate": 5.1322246239824024e-05,
      "loss": 2.6744949340820314,
      "memory(GiB)": 77.56,
      "step": 57370,
      "token_acc": 0.46774193548387094,
      "train_speed(iter/s)": 1.438893
    },
    {
      "epoch": 2.458120903131828,
      "grad_norm": 5.567917823791504,
      "learning_rate": 5.131551880841219e-05,
      "loss": 2.3654218673706056,
      "memory(GiB)": 77.56,
      "step": 57375,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.438905
    },
    {
      "epoch": 2.458335118461077,
      "grad_norm": 4.319338321685791,
      "learning_rate": 5.1308791353168484e-05,
      "loss": 2.2937015533447265,
      "memory(GiB)": 77.56,
      "step": 57380,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438933
    },
    {
      "epoch": 2.458549333790326,
      "grad_norm": 4.3087897300720215,
      "learning_rate": 5.130206387421482e-05,
      "loss": 2.769311714172363,
      "memory(GiB)": 77.56,
      "step": 57385,
      "token_acc": 0.44518272425249167,
      "train_speed(iter/s)": 1.438954
    },
    {
      "epoch": 2.458763549119575,
      "grad_norm": 5.06441068649292,
      "learning_rate": 5.1295336371673045e-05,
      "loss": 2.5599733352661134,
      "memory(GiB)": 77.56,
      "step": 57390,
      "token_acc": 0.44518272425249167,
      "train_speed(iter/s)": 1.438952
    },
    {
      "epoch": 2.4589777644488238,
      "grad_norm": 5.019257068634033,
      "learning_rate": 5.1288608845665034e-05,
      "loss": 2.318602752685547,
      "memory(GiB)": 77.56,
      "step": 57395,
      "token_acc": 0.5437956204379562,
      "train_speed(iter/s)": 1.438967
    },
    {
      "epoch": 2.459191979778073,
      "grad_norm": 6.315002918243408,
      "learning_rate": 5.128188129631266e-05,
      "loss": 2.4263404846191405,
      "memory(GiB)": 77.56,
      "step": 57400,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.438995
    },
    {
      "epoch": 2.459406195107322,
      "grad_norm": 5.129228591918945,
      "learning_rate": 5.12751537237378e-05,
      "loss": 2.459724235534668,
      "memory(GiB)": 77.56,
      "step": 57405,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.439039
    },
    {
      "epoch": 2.4596204104365706,
      "grad_norm": 7.693178653717041,
      "learning_rate": 5.126842612806234e-05,
      "loss": 2.2444610595703125,
      "memory(GiB)": 77.56,
      "step": 57410,
      "token_acc": 0.5399239543726235,
      "train_speed(iter/s)": 1.439052
    },
    {
      "epoch": 2.45983462576582,
      "grad_norm": 7.329493999481201,
      "learning_rate": 5.1261698509408154e-05,
      "loss": 2.370320510864258,
      "memory(GiB)": 77.56,
      "step": 57415,
      "token_acc": 0.45526315789473687,
      "train_speed(iter/s)": 1.439072
    },
    {
      "epoch": 2.4600488410950687,
      "grad_norm": 5.697206497192383,
      "learning_rate": 5.125497086789711e-05,
      "loss": 2.334388542175293,
      "memory(GiB)": 77.56,
      "step": 57420,
      "token_acc": 0.4979919678714859,
      "train_speed(iter/s)": 1.439084
    },
    {
      "epoch": 2.460263056424318,
      "grad_norm": 4.582203388214111,
      "learning_rate": 5.1248243203651094e-05,
      "loss": 2.381041145324707,
      "memory(GiB)": 77.56,
      "step": 57425,
      "token_acc": 0.5387096774193548,
      "train_speed(iter/s)": 1.439076
    },
    {
      "epoch": 2.460477271753567,
      "grad_norm": 5.91025447845459,
      "learning_rate": 5.124151551679198e-05,
      "loss": 2.4503894805908204,
      "memory(GiB)": 77.56,
      "step": 57430,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.439093
    },
    {
      "epoch": 2.4606914870828156,
      "grad_norm": 5.3099894523620605,
      "learning_rate": 5.1234787807441655e-05,
      "loss": 2.023750114440918,
      "memory(GiB)": 77.56,
      "step": 57435,
      "token_acc": 0.55859375,
      "train_speed(iter/s)": 1.439109
    },
    {
      "epoch": 2.460905702412065,
      "grad_norm": 5.285356044769287,
      "learning_rate": 5.122806007572198e-05,
      "loss": 2.383597564697266,
      "memory(GiB)": 77.56,
      "step": 57440,
      "token_acc": 0.476038338658147,
      "train_speed(iter/s)": 1.43909
    },
    {
      "epoch": 2.4611199177413137,
      "grad_norm": 5.4618120193481445,
      "learning_rate": 5.1221332321754855e-05,
      "loss": 2.4647523880004885,
      "memory(GiB)": 77.56,
      "step": 57445,
      "token_acc": 0.45874587458745875,
      "train_speed(iter/s)": 1.439097
    },
    {
      "epoch": 2.4613341330705625,
      "grad_norm": 4.592566013336182,
      "learning_rate": 5.1214604545662135e-05,
      "loss": 2.218570327758789,
      "memory(GiB)": 77.56,
      "step": 57450,
      "token_acc": 0.5038461538461538,
      "train_speed(iter/s)": 1.439115
    },
    {
      "epoch": 2.4615483483998117,
      "grad_norm": 5.485901355743408,
      "learning_rate": 5.120787674756573e-05,
      "loss": 2.2112655639648438,
      "memory(GiB)": 77.56,
      "step": 57455,
      "token_acc": 0.5682656826568265,
      "train_speed(iter/s)": 1.439121
    },
    {
      "epoch": 2.4617625637290605,
      "grad_norm": 4.783279895782471,
      "learning_rate": 5.120114892758749e-05,
      "loss": 2.4660833358764647,
      "memory(GiB)": 77.56,
      "step": 57460,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.439138
    },
    {
      "epoch": 2.4619767790583094,
      "grad_norm": 4.32682466506958,
      "learning_rate": 5.119442108584932e-05,
      "loss": 2.0927169799804686,
      "memory(GiB)": 77.56,
      "step": 57465,
      "token_acc": 0.5773584905660377,
      "train_speed(iter/s)": 1.439118
    },
    {
      "epoch": 2.4621909943875586,
      "grad_norm": 5.225218772888184,
      "learning_rate": 5.11876932224731e-05,
      "loss": 2.220818519592285,
      "memory(GiB)": 77.56,
      "step": 57470,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.439116
    },
    {
      "epoch": 2.4624052097168074,
      "grad_norm": 5.393348693847656,
      "learning_rate": 5.11809653375807e-05,
      "loss": 2.4311344146728517,
      "memory(GiB)": 77.56,
      "step": 57475,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.439126
    },
    {
      "epoch": 2.4626194250460562,
      "grad_norm": 4.696230888366699,
      "learning_rate": 5.1174237431293994e-05,
      "loss": 2.447711181640625,
      "memory(GiB)": 77.56,
      "step": 57480,
      "token_acc": 0.4969512195121951,
      "train_speed(iter/s)": 1.439154
    },
    {
      "epoch": 2.4628336403753055,
      "grad_norm": 6.5843353271484375,
      "learning_rate": 5.116750950373487e-05,
      "loss": 2.7485700607299806,
      "memory(GiB)": 77.56,
      "step": 57485,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.439167
    },
    {
      "epoch": 2.4630478557045543,
      "grad_norm": 4.670713424682617,
      "learning_rate": 5.1160781555025225e-05,
      "loss": 2.2367443084716796,
      "memory(GiB)": 77.56,
      "step": 57490,
      "token_acc": 0.51875,
      "train_speed(iter/s)": 1.43917
    },
    {
      "epoch": 2.463262071033803,
      "grad_norm": 5.058983325958252,
      "learning_rate": 5.115405358528693e-05,
      "loss": 2.2882545471191404,
      "memory(GiB)": 77.56,
      "step": 57495,
      "token_acc": 0.5409252669039146,
      "train_speed(iter/s)": 1.43917
    },
    {
      "epoch": 2.4634762863630524,
      "grad_norm": 6.193905353546143,
      "learning_rate": 5.114732559464188e-05,
      "loss": 2.2321853637695312,
      "memory(GiB)": 77.56,
      "step": 57500,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.439199
    },
    {
      "epoch": 2.4634762863630524,
      "eval_loss": 2.181269884109497,
      "eval_runtime": 13.2963,
      "eval_samples_per_second": 7.521,
      "eval_steps_per_second": 7.521,
      "eval_token_acc": 0.48021828103683495,
      "step": 57500
    },
    {
      "epoch": 2.463690501692301,
      "grad_norm": 4.969756603240967,
      "learning_rate": 5.114059758321196e-05,
      "loss": 2.528905487060547,
      "memory(GiB)": 77.56,
      "step": 57505,
      "token_acc": 0.4846449136276392,
      "train_speed(iter/s)": 1.438707
    },
    {
      "epoch": 2.46390471702155,
      "grad_norm": 6.251031875610352,
      "learning_rate": 5.1133869551119016e-05,
      "loss": 2.0313507080078126,
      "memory(GiB)": 77.56,
      "step": 57510,
      "token_acc": 0.5367647058823529,
      "train_speed(iter/s)": 1.438716
    },
    {
      "epoch": 2.4641189323507993,
      "grad_norm": 6.339770793914795,
      "learning_rate": 5.112714149848499e-05,
      "loss": 2.6249263763427733,
      "memory(GiB)": 77.56,
      "step": 57515,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.438712
    },
    {
      "epoch": 2.464333147680048,
      "grad_norm": 4.177573204040527,
      "learning_rate": 5.112041342543171e-05,
      "loss": 2.6119308471679688,
      "memory(GiB)": 77.56,
      "step": 57520,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.438713
    },
    {
      "epoch": 2.464547363009297,
      "grad_norm": 6.631214618682861,
      "learning_rate": 5.1113685332081094e-05,
      "loss": 2.4411493301391602,
      "memory(GiB)": 77.56,
      "step": 57525,
      "token_acc": 0.4983277591973244,
      "train_speed(iter/s)": 1.438738
    },
    {
      "epoch": 2.464761578338546,
      "grad_norm": 4.378689765930176,
      "learning_rate": 5.110695721855505e-05,
      "loss": 2.7039278030395506,
      "memory(GiB)": 77.56,
      "step": 57530,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.438764
    },
    {
      "epoch": 2.464975793667795,
      "grad_norm": 8.13716983795166,
      "learning_rate": 5.1100229084975424e-05,
      "loss": 2.189742851257324,
      "memory(GiB)": 77.56,
      "step": 57535,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438778
    },
    {
      "epoch": 2.4651900089970438,
      "grad_norm": 4.951664924621582,
      "learning_rate": 5.109350093146411e-05,
      "loss": 2.170023536682129,
      "memory(GiB)": 77.56,
      "step": 57540,
      "token_acc": 0.5394736842105263,
      "train_speed(iter/s)": 1.438818
    },
    {
      "epoch": 2.465404224326293,
      "grad_norm": 6.1526923179626465,
      "learning_rate": 5.108677275814301e-05,
      "loss": 2.6115438461303713,
      "memory(GiB)": 77.56,
      "step": 57545,
      "token_acc": 0.43137254901960786,
      "train_speed(iter/s)": 1.438806
    },
    {
      "epoch": 2.465618439655542,
      "grad_norm": 8.08043384552002,
      "learning_rate": 5.108004456513399e-05,
      "loss": 2.3119247436523436,
      "memory(GiB)": 77.56,
      "step": 57550,
      "token_acc": 0.4978723404255319,
      "train_speed(iter/s)": 1.438816
    },
    {
      "epoch": 2.4658326549847907,
      "grad_norm": 5.128757953643799,
      "learning_rate": 5.107331635255895e-05,
      "loss": 2.3346920013427734,
      "memory(GiB)": 77.56,
      "step": 57555,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.438814
    },
    {
      "epoch": 2.46604687031404,
      "grad_norm": 8.985196113586426,
      "learning_rate": 5.1066588120539785e-05,
      "loss": 2.39608211517334,
      "memory(GiB)": 77.56,
      "step": 57560,
      "token_acc": 0.4978540772532189,
      "train_speed(iter/s)": 1.438858
    },
    {
      "epoch": 2.4662610856432887,
      "grad_norm": 5.544095516204834,
      "learning_rate": 5.105985986919838e-05,
      "loss": 2.2582118988037108,
      "memory(GiB)": 77.56,
      "step": 57565,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.438869
    },
    {
      "epoch": 2.4664753009725375,
      "grad_norm": 5.5937066078186035,
      "learning_rate": 5.1053131598656614e-05,
      "loss": 2.600021171569824,
      "memory(GiB)": 77.56,
      "step": 57570,
      "token_acc": 0.46006389776357826,
      "train_speed(iter/s)": 1.438824
    },
    {
      "epoch": 2.466689516301787,
      "grad_norm": 5.682216644287109,
      "learning_rate": 5.104640330903638e-05,
      "loss": 2.501003646850586,
      "memory(GiB)": 77.56,
      "step": 57575,
      "token_acc": 0.4750733137829912,
      "train_speed(iter/s)": 1.438818
    },
    {
      "epoch": 2.4669037316310356,
      "grad_norm": 5.798671722412109,
      "learning_rate": 5.103967500045956e-05,
      "loss": 2.7489118576049805,
      "memory(GiB)": 77.56,
      "step": 57580,
      "token_acc": 0.466403162055336,
      "train_speed(iter/s)": 1.4388
    },
    {
      "epoch": 2.4671179469602844,
      "grad_norm": 5.5647382736206055,
      "learning_rate": 5.1032946673048067e-05,
      "loss": 2.5334585189819334,
      "memory(GiB)": 77.56,
      "step": 57585,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43882
    },
    {
      "epoch": 2.4673321622895337,
      "grad_norm": 5.787646293640137,
      "learning_rate": 5.102621832692378e-05,
      "loss": 2.200487518310547,
      "memory(GiB)": 77.56,
      "step": 57590,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.438787
    },
    {
      "epoch": 2.4675463776187825,
      "grad_norm": 5.420490264892578,
      "learning_rate": 5.1019489962208555e-05,
      "loss": 2.5062013626098634,
      "memory(GiB)": 77.56,
      "step": 57595,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.438791
    },
    {
      "epoch": 2.4677605929480313,
      "grad_norm": 6.231259346008301,
      "learning_rate": 5.101276157902434e-05,
      "loss": 2.5847116470336915,
      "memory(GiB)": 77.56,
      "step": 57600,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.43882
    },
    {
      "epoch": 2.4679748082772806,
      "grad_norm": 5.21881103515625,
      "learning_rate": 5.100603317749299e-05,
      "loss": 2.63265380859375,
      "memory(GiB)": 77.56,
      "step": 57605,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.438853
    },
    {
      "epoch": 2.4681890236065294,
      "grad_norm": 5.555099010467529,
      "learning_rate": 5.099930475773641e-05,
      "loss": 2.415752410888672,
      "memory(GiB)": 77.56,
      "step": 57610,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.438851
    },
    {
      "epoch": 2.468403238935778,
      "grad_norm": 4.607346534729004,
      "learning_rate": 5.099257631987648e-05,
      "loss": 2.609634017944336,
      "memory(GiB)": 77.56,
      "step": 57615,
      "token_acc": 0.44363636363636366,
      "train_speed(iter/s)": 1.438851
    },
    {
      "epoch": 2.4686174542650274,
      "grad_norm": 4.508793354034424,
      "learning_rate": 5.098584786403512e-05,
      "loss": 2.289145851135254,
      "memory(GiB)": 77.56,
      "step": 57620,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.438852
    },
    {
      "epoch": 2.4688316695942762,
      "grad_norm": 6.425941467285156,
      "learning_rate": 5.0979119390334175e-05,
      "loss": 2.796823501586914,
      "memory(GiB)": 77.56,
      "step": 57625,
      "token_acc": 0.4491017964071856,
      "train_speed(iter/s)": 1.438838
    },
    {
      "epoch": 2.469045884923525,
      "grad_norm": 5.5024518966674805,
      "learning_rate": 5.097239089889558e-05,
      "loss": 2.4876276016235352,
      "memory(GiB)": 77.56,
      "step": 57630,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.438831
    },
    {
      "epoch": 2.4692601002527743,
      "grad_norm": 5.610123157501221,
      "learning_rate": 5.0965662389841196e-05,
      "loss": 2.4639196395874023,
      "memory(GiB)": 77.56,
      "step": 57635,
      "token_acc": 0.49603174603174605,
      "train_speed(iter/s)": 1.43885
    },
    {
      "epoch": 2.469474315582023,
      "grad_norm": 6.002385139465332,
      "learning_rate": 5.095893386329293e-05,
      "loss": 2.4816234588623045,
      "memory(GiB)": 77.56,
      "step": 57640,
      "token_acc": 0.5080645161290323,
      "train_speed(iter/s)": 1.438881
    },
    {
      "epoch": 2.469688530911272,
      "grad_norm": 3.819467306137085,
      "learning_rate": 5.0952205319372706e-05,
      "loss": 2.254865264892578,
      "memory(GiB)": 77.56,
      "step": 57645,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.438904
    },
    {
      "epoch": 2.469902746240521,
      "grad_norm": 6.105497360229492,
      "learning_rate": 5.094547675820237e-05,
      "loss": 2.43424072265625,
      "memory(GiB)": 77.56,
      "step": 57650,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.438925
    },
    {
      "epoch": 2.47011696156977,
      "grad_norm": 4.623422622680664,
      "learning_rate": 5.093874817990383e-05,
      "loss": 2.5657581329345702,
      "memory(GiB)": 77.56,
      "step": 57655,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.438949
    },
    {
      "epoch": 2.470331176899019,
      "grad_norm": 5.294508457183838,
      "learning_rate": 5.0932019584599e-05,
      "loss": 2.4200065612792967,
      "memory(GiB)": 77.56,
      "step": 57660,
      "token_acc": 0.49107142857142855,
      "train_speed(iter/s)": 1.438907
    },
    {
      "epoch": 2.470545392228268,
      "grad_norm": 4.910347938537598,
      "learning_rate": 5.092529097240976e-05,
      "loss": 2.3683277130126954,
      "memory(GiB)": 77.56,
      "step": 57665,
      "token_acc": 0.525974025974026,
      "train_speed(iter/s)": 1.438925
    },
    {
      "epoch": 2.470759607557517,
      "grad_norm": 5.081131458282471,
      "learning_rate": 5.091856234345799e-05,
      "loss": 2.343279266357422,
      "memory(GiB)": 77.56,
      "step": 57670,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438911
    },
    {
      "epoch": 2.4709738228867657,
      "grad_norm": 4.529219627380371,
      "learning_rate": 5.0911833697865607e-05,
      "loss": 2.6960355758666994,
      "memory(GiB)": 77.56,
      "step": 57675,
      "token_acc": 0.4306784660766962,
      "train_speed(iter/s)": 1.438917
    },
    {
      "epoch": 2.471188038216015,
      "grad_norm": 6.560049057006836,
      "learning_rate": 5.0905105035754516e-05,
      "loss": 2.308742332458496,
      "memory(GiB)": 77.56,
      "step": 57680,
      "token_acc": 0.524390243902439,
      "train_speed(iter/s)": 1.438932
    },
    {
      "epoch": 2.4714022535452638,
      "grad_norm": 6.020293712615967,
      "learning_rate": 5.089837635724658e-05,
      "loss": 2.517816925048828,
      "memory(GiB)": 77.56,
      "step": 57685,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.438961
    },
    {
      "epoch": 2.4716164688745126,
      "grad_norm": 5.475345611572266,
      "learning_rate": 5.0891647662463724e-05,
      "loss": 2.476761817932129,
      "memory(GiB)": 77.56,
      "step": 57690,
      "token_acc": 0.528052805280528,
      "train_speed(iter/s)": 1.438983
    },
    {
      "epoch": 2.471830684203762,
      "grad_norm": 6.823609828948975,
      "learning_rate": 5.088491895152784e-05,
      "loss": 2.4553321838378905,
      "memory(GiB)": 77.56,
      "step": 57695,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.438987
    },
    {
      "epoch": 2.4720448995330107,
      "grad_norm": 5.711009979248047,
      "learning_rate": 5.0878190224560804e-05,
      "loss": 2.364780807495117,
      "memory(GiB)": 77.56,
      "step": 57700,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.438972
    },
    {
      "epoch": 2.4722591148622595,
      "grad_norm": 4.534488201141357,
      "learning_rate": 5.0871461481684546e-05,
      "loss": 2.149310302734375,
      "memory(GiB)": 77.56,
      "step": 57705,
      "token_acc": 0.5392857142857143,
      "train_speed(iter/s)": 1.438975
    },
    {
      "epoch": 2.4724733301915087,
      "grad_norm": 5.547414779663086,
      "learning_rate": 5.0864732723020915e-05,
      "loss": 2.160359191894531,
      "memory(GiB)": 77.56,
      "step": 57710,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.438949
    },
    {
      "epoch": 2.4726875455207575,
      "grad_norm": 6.99166202545166,
      "learning_rate": 5.085800394869187e-05,
      "loss": 2.383811187744141,
      "memory(GiB)": 77.56,
      "step": 57715,
      "token_acc": 0.5257731958762887,
      "train_speed(iter/s)": 1.438924
    },
    {
      "epoch": 2.4729017608500063,
      "grad_norm": 5.011475086212158,
      "learning_rate": 5.0851275158819264e-05,
      "loss": 2.288974380493164,
      "memory(GiB)": 77.56,
      "step": 57720,
      "token_acc": 0.535483870967742,
      "train_speed(iter/s)": 1.438909
    },
    {
      "epoch": 2.4731159761792556,
      "grad_norm": 6.043013095855713,
      "learning_rate": 5.084454635352501e-05,
      "loss": 2.4874574661254885,
      "memory(GiB)": 77.56,
      "step": 57725,
      "token_acc": 0.4984894259818731,
      "train_speed(iter/s)": 1.438906
    },
    {
      "epoch": 2.4733301915085044,
      "grad_norm": 6.45123291015625,
      "learning_rate": 5.083781753293102e-05,
      "loss": 2.1487701416015623,
      "memory(GiB)": 77.56,
      "step": 57730,
      "token_acc": 0.5543071161048689,
      "train_speed(iter/s)": 1.438903
    },
    {
      "epoch": 2.4735444068377532,
      "grad_norm": 5.536190509796143,
      "learning_rate": 5.083108869715918e-05,
      "loss": 2.208243179321289,
      "memory(GiB)": 77.56,
      "step": 57735,
      "token_acc": 0.5518518518518518,
      "train_speed(iter/s)": 1.438911
    },
    {
      "epoch": 2.4737586221670025,
      "grad_norm": 4.991488456726074,
      "learning_rate": 5.082435984633137e-05,
      "loss": 2.4614936828613283,
      "memory(GiB)": 77.56,
      "step": 57740,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.43894
    },
    {
      "epoch": 2.4739728374962513,
      "grad_norm": 6.942436695098877,
      "learning_rate": 5.081763098056951e-05,
      "loss": 2.446458625793457,
      "memory(GiB)": 77.56,
      "step": 57745,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.438936
    },
    {
      "epoch": 2.4741870528255,
      "grad_norm": 4.558429718017578,
      "learning_rate": 5.08109020999955e-05,
      "loss": 2.6614004135131837,
      "memory(GiB)": 77.56,
      "step": 57750,
      "token_acc": 0.4670846394984326,
      "train_speed(iter/s)": 1.438959
    },
    {
      "epoch": 2.4744012681547494,
      "grad_norm": 4.911935806274414,
      "learning_rate": 5.080417320473124e-05,
      "loss": 2.335498809814453,
      "memory(GiB)": 77.56,
      "step": 57755,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.43895
    },
    {
      "epoch": 2.474615483483998,
      "grad_norm": 4.802637100219727,
      "learning_rate": 5.0797444294898644e-05,
      "loss": 2.1629831314086916,
      "memory(GiB)": 77.56,
      "step": 57760,
      "token_acc": 0.5162241887905604,
      "train_speed(iter/s)": 1.438906
    },
    {
      "epoch": 2.474829698813247,
      "grad_norm": 4.477532863616943,
      "learning_rate": 5.079071537061959e-05,
      "loss": 2.095111274719238,
      "memory(GiB)": 77.56,
      "step": 57765,
      "token_acc": 0.5464684014869888,
      "train_speed(iter/s)": 1.438918
    },
    {
      "epoch": 2.4750439141424962,
      "grad_norm": 5.740736484527588,
      "learning_rate": 5.078398643201597e-05,
      "loss": 2.5655784606933594,
      "memory(GiB)": 77.56,
      "step": 57770,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.43891
    },
    {
      "epoch": 2.475258129471745,
      "grad_norm": 4.7918901443481445,
      "learning_rate": 5.077725747920972e-05,
      "loss": 2.198044013977051,
      "memory(GiB)": 77.56,
      "step": 57775,
      "token_acc": 0.5404411764705882,
      "train_speed(iter/s)": 1.438938
    },
    {
      "epoch": 2.475472344800994,
      "grad_norm": 4.255255222320557,
      "learning_rate": 5.07705285123227e-05,
      "loss": 2.214427185058594,
      "memory(GiB)": 77.56,
      "step": 57780,
      "token_acc": 0.5271565495207667,
      "train_speed(iter/s)": 1.438929
    },
    {
      "epoch": 2.475686560130243,
      "grad_norm": 5.639316558837891,
      "learning_rate": 5.076379953147684e-05,
      "loss": 2.607181167602539,
      "memory(GiB)": 77.56,
      "step": 57785,
      "token_acc": 0.4984709480122324,
      "train_speed(iter/s)": 1.438902
    },
    {
      "epoch": 2.475900775459492,
      "grad_norm": 4.517269611358643,
      "learning_rate": 5.075707053679404e-05,
      "loss": 2.305590057373047,
      "memory(GiB)": 77.56,
      "step": 57790,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.438886
    },
    {
      "epoch": 2.4761149907887408,
      "grad_norm": 5.247602462768555,
      "learning_rate": 5.075034152839621e-05,
      "loss": 2.4419702529907226,
      "memory(GiB)": 77.56,
      "step": 57795,
      "token_acc": 0.49842271293375395,
      "train_speed(iter/s)": 1.438863
    },
    {
      "epoch": 2.47632920611799,
      "grad_norm": 3.173375368118286,
      "learning_rate": 5.074361250640521e-05,
      "loss": 2.157380294799805,
      "memory(GiB)": 77.56,
      "step": 57800,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.438886
    },
    {
      "epoch": 2.476543421447239,
      "grad_norm": 5.00565242767334,
      "learning_rate": 5.0736883470942986e-05,
      "loss": 2.3380767822265627,
      "memory(GiB)": 77.56,
      "step": 57805,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.438891
    },
    {
      "epoch": 2.4767576367764876,
      "grad_norm": 6.09058141708374,
      "learning_rate": 5.0730154422131424e-05,
      "loss": 2.4289512634277344,
      "memory(GiB)": 77.56,
      "step": 57810,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.438896
    },
    {
      "epoch": 2.476971852105737,
      "grad_norm": 4.935365676879883,
      "learning_rate": 5.072342536009245e-05,
      "loss": 2.480707550048828,
      "memory(GiB)": 77.56,
      "step": 57815,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43889
    },
    {
      "epoch": 2.4771860674349857,
      "grad_norm": 6.091104507446289,
      "learning_rate": 5.071669628494792e-05,
      "loss": 2.5238800048828125,
      "memory(GiB)": 77.56,
      "step": 57820,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438905
    },
    {
      "epoch": 2.4774002827642345,
      "grad_norm": 8.003154754638672,
      "learning_rate": 5.070996719681977e-05,
      "loss": 2.52032527923584,
      "memory(GiB)": 77.56,
      "step": 57825,
      "token_acc": 0.5070921985815603,
      "train_speed(iter/s)": 1.438923
    },
    {
      "epoch": 2.4776144980934838,
      "grad_norm": 5.366401672363281,
      "learning_rate": 5.070323809582991e-05,
      "loss": 2.2390411376953123,
      "memory(GiB)": 77.56,
      "step": 57830,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.438885
    },
    {
      "epoch": 2.4778287134227326,
      "grad_norm": 5.645700931549072,
      "learning_rate": 5.0696508982100225e-05,
      "loss": 2.2358333587646486,
      "memory(GiB)": 77.56,
      "step": 57835,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.438897
    },
    {
      "epoch": 2.4780429287519814,
      "grad_norm": 5.364466667175293,
      "learning_rate": 5.0689779855752616e-05,
      "loss": 2.5054306030273437,
      "memory(GiB)": 77.56,
      "step": 57840,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.438897
    },
    {
      "epoch": 2.4782571440812307,
      "grad_norm": 6.162137508392334,
      "learning_rate": 5.068305071690902e-05,
      "loss": 2.4064870834350587,
      "memory(GiB)": 77.56,
      "step": 57845,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.438912
    },
    {
      "epoch": 2.4784713594104795,
      "grad_norm": 4.874151229858398,
      "learning_rate": 5.067632156569131e-05,
      "loss": 2.454629898071289,
      "memory(GiB)": 77.56,
      "step": 57850,
      "token_acc": 0.5057915057915058,
      "train_speed(iter/s)": 1.438915
    },
    {
      "epoch": 2.4786855747397283,
      "grad_norm": 5.7840800285339355,
      "learning_rate": 5.066959240222138e-05,
      "loss": 2.785805892944336,
      "memory(GiB)": 77.56,
      "step": 57855,
      "token_acc": 0.4630225080385852,
      "train_speed(iter/s)": 1.438928
    },
    {
      "epoch": 2.4788997900689775,
      "grad_norm": 7.42020845413208,
      "learning_rate": 5.066286322662118e-05,
      "loss": 2.225681495666504,
      "memory(GiB)": 77.56,
      "step": 57860,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.438926
    },
    {
      "epoch": 2.4791140053982264,
      "grad_norm": 5.325389862060547,
      "learning_rate": 5.0656134039012593e-05,
      "loss": 2.7070350646972656,
      "memory(GiB)": 77.56,
      "step": 57865,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.438955
    },
    {
      "epoch": 2.479328220727475,
      "grad_norm": 6.985469818115234,
      "learning_rate": 5.06494048395175e-05,
      "loss": 2.329922676086426,
      "memory(GiB)": 77.56,
      "step": 57870,
      "token_acc": 0.47580645161290325,
      "train_speed(iter/s)": 1.438964
    },
    {
      "epoch": 2.4795424360567244,
      "grad_norm": 5.885095119476318,
      "learning_rate": 5.064267562825785e-05,
      "loss": 2.2059085845947264,
      "memory(GiB)": 77.56,
      "step": 57875,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.438932
    },
    {
      "epoch": 2.4797566513859732,
      "grad_norm": 5.26485538482666,
      "learning_rate": 5.0635946405355525e-05,
      "loss": 2.2520315170288088,
      "memory(GiB)": 77.56,
      "step": 57880,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.438942
    },
    {
      "epoch": 2.479970866715222,
      "grad_norm": 5.791314125061035,
      "learning_rate": 5.062921717093243e-05,
      "loss": 2.424143600463867,
      "memory(GiB)": 77.56,
      "step": 57885,
      "token_acc": 0.45182724252491696,
      "train_speed(iter/s)": 1.438966
    },
    {
      "epoch": 2.4801850820444713,
      "grad_norm": 6.202103614807129,
      "learning_rate": 5.062248792511048e-05,
      "loss": 2.5702686309814453,
      "memory(GiB)": 77.56,
      "step": 57890,
      "token_acc": 0.44692737430167595,
      "train_speed(iter/s)": 1.438951
    },
    {
      "epoch": 2.48039929737372,
      "grad_norm": 4.706606388092041,
      "learning_rate": 5.061575866801157e-05,
      "loss": 2.291838836669922,
      "memory(GiB)": 77.56,
      "step": 57895,
      "token_acc": 0.5374449339207048,
      "train_speed(iter/s)": 1.438964
    },
    {
      "epoch": 2.480613512702969,
      "grad_norm": 6.094140529632568,
      "learning_rate": 5.0609029399757615e-05,
      "loss": 2.5490455627441406,
      "memory(GiB)": 77.56,
      "step": 57900,
      "token_acc": 0.42662116040955633,
      "train_speed(iter/s)": 1.438975
    },
    {
      "epoch": 2.480827728032218,
      "grad_norm": 5.127205848693848,
      "learning_rate": 5.060230012047052e-05,
      "loss": 2.4769092559814454,
      "memory(GiB)": 77.56,
      "step": 57905,
      "token_acc": 0.4533333333333333,
      "train_speed(iter/s)": 1.438986
    },
    {
      "epoch": 2.481041943361467,
      "grad_norm": 6.49498987197876,
      "learning_rate": 5.059557083027221e-05,
      "loss": 2.4379310607910156,
      "memory(GiB)": 77.56,
      "step": 57910,
      "token_acc": 0.46745562130177515,
      "train_speed(iter/s)": 1.438978
    },
    {
      "epoch": 2.481256158690716,
      "grad_norm": 5.46910285949707,
      "learning_rate": 5.058884152928455e-05,
      "loss": 2.576978302001953,
      "memory(GiB)": 77.56,
      "step": 57915,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.439019
    },
    {
      "epoch": 2.481470374019965,
      "grad_norm": 6.400634765625,
      "learning_rate": 5.0582112217629494e-05,
      "loss": 2.4977977752685545,
      "memory(GiB)": 77.56,
      "step": 57920,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.439024
    },
    {
      "epoch": 2.481684589349214,
      "grad_norm": 4.879184722900391,
      "learning_rate": 5.0575382895428914e-05,
      "loss": 2.3481157302856444,
      "memory(GiB)": 77.56,
      "step": 57925,
      "token_acc": 0.4830188679245283,
      "train_speed(iter/s)": 1.439019
    },
    {
      "epoch": 2.4818988046784627,
      "grad_norm": 5.49393892288208,
      "learning_rate": 5.0568653562804734e-05,
      "loss": 2.4221954345703125,
      "memory(GiB)": 77.56,
      "step": 57930,
      "token_acc": 0.46688741721854304,
      "train_speed(iter/s)": 1.438996
    },
    {
      "epoch": 2.482113020007712,
      "grad_norm": 5.23515510559082,
      "learning_rate": 5.056192421987888e-05,
      "loss": 2.602817916870117,
      "memory(GiB)": 77.56,
      "step": 57935,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.439021
    },
    {
      "epoch": 2.4823272353369608,
      "grad_norm": 4.713934421539307,
      "learning_rate": 5.055519486677322e-05,
      "loss": 2.351328468322754,
      "memory(GiB)": 77.56,
      "step": 57940,
      "token_acc": 0.5346938775510204,
      "train_speed(iter/s)": 1.43903
    },
    {
      "epoch": 2.4825414506662096,
      "grad_norm": 4.903232097625732,
      "learning_rate": 5.0548465503609697e-05,
      "loss": 2.4050979614257812,
      "memory(GiB)": 77.56,
      "step": 57945,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.438975
    },
    {
      "epoch": 2.482755665995459,
      "grad_norm": 6.948683738708496,
      "learning_rate": 5.054173613051021e-05,
      "loss": 2.4426015853881835,
      "memory(GiB)": 77.56,
      "step": 57950,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.439008
    },
    {
      "epoch": 2.4829698813247076,
      "grad_norm": 4.461533069610596,
      "learning_rate": 5.0535006747596646e-05,
      "loss": 2.7285572052001954,
      "memory(GiB)": 77.56,
      "step": 57955,
      "token_acc": 0.43922651933701656,
      "train_speed(iter/s)": 1.439014
    },
    {
      "epoch": 2.4831840966539565,
      "grad_norm": 4.471543788909912,
      "learning_rate": 5.052827735499095e-05,
      "loss": 2.9877140045166017,
      "memory(GiB)": 77.56,
      "step": 57960,
      "token_acc": 0.4010152284263959,
      "train_speed(iter/s)": 1.439014
    },
    {
      "epoch": 2.4833983119832057,
      "grad_norm": 6.1984543800354,
      "learning_rate": 5.0521547952815005e-05,
      "loss": 2.3776714324951174,
      "memory(GiB)": 77.56,
      "step": 57965,
      "token_acc": 0.4742268041237113,
      "train_speed(iter/s)": 1.43905
    },
    {
      "epoch": 2.4836125273124545,
      "grad_norm": 4.951435565948486,
      "learning_rate": 5.0514818541190745e-05,
      "loss": 2.098245620727539,
      "memory(GiB)": 77.56,
      "step": 57970,
      "token_acc": 0.5335463258785943,
      "train_speed(iter/s)": 1.439056
    },
    {
      "epoch": 2.4838267426417033,
      "grad_norm": 6.0008625984191895,
      "learning_rate": 5.050808912024004e-05,
      "loss": 2.408772659301758,
      "memory(GiB)": 77.56,
      "step": 57975,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.439033
    },
    {
      "epoch": 2.4840409579709526,
      "grad_norm": 6.2969770431518555,
      "learning_rate": 5.050135969008485e-05,
      "loss": 2.5903215408325195,
      "memory(GiB)": 77.56,
      "step": 57980,
      "token_acc": 0.4625,
      "train_speed(iter/s)": 1.439057
    },
    {
      "epoch": 2.4842551733002014,
      "grad_norm": 4.6040263175964355,
      "learning_rate": 5.049463025084703e-05,
      "loss": 2.398129463195801,
      "memory(GiB)": 77.56,
      "step": 57985,
      "token_acc": 0.49691358024691357,
      "train_speed(iter/s)": 1.439081
    },
    {
      "epoch": 2.48446938862945,
      "grad_norm": 5.366873741149902,
      "learning_rate": 5.0487900802648544e-05,
      "loss": 2.522391128540039,
      "memory(GiB)": 77.56,
      "step": 57990,
      "token_acc": 0.5340909090909091,
      "train_speed(iter/s)": 1.439076
    },
    {
      "epoch": 2.4846836039586995,
      "grad_norm": 4.175821304321289,
      "learning_rate": 5.048117134561128e-05,
      "loss": 2.3147741317749024,
      "memory(GiB)": 77.56,
      "step": 57995,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.439074
    },
    {
      "epoch": 2.4848978192879483,
      "grad_norm": 5.270333766937256,
      "learning_rate": 5.0474441879857125e-05,
      "loss": 2.7343297958374024,
      "memory(GiB)": 77.56,
      "step": 58000,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.439077
    },
    {
      "epoch": 2.4848978192879483,
      "eval_loss": 2.177473306655884,
      "eval_runtime": 13.7372,
      "eval_samples_per_second": 7.279,
      "eval_steps_per_second": 7.279,
      "eval_token_acc": 0.47198879551820727,
      "step": 58000
    },
    {
      "epoch": 2.485112034617197,
      "grad_norm": 7.586592674255371,
      "learning_rate": 5.046771240550801e-05,
      "loss": 2.7471054077148436,
      "memory(GiB)": 77.56,
      "step": 58005,
      "token_acc": 0.46255924170616114,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 2.4853262499464464,
      "grad_norm": 4.5493645668029785,
      "learning_rate": 5.0460982922685865e-05,
      "loss": 2.4920017242431642,
      "memory(GiB)": 77.56,
      "step": 58010,
      "token_acc": 0.4803921568627451,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.485540465275695,
      "grad_norm": 5.337832450866699,
      "learning_rate": 5.045425343151255e-05,
      "loss": 2.594310760498047,
      "memory(GiB)": 77.56,
      "step": 58015,
      "token_acc": 0.4509090909090909,
      "train_speed(iter/s)": 1.43856
    },
    {
      "epoch": 2.485754680604944,
      "grad_norm": 9.150239944458008,
      "learning_rate": 5.044752393211004e-05,
      "loss": 2.4332067489624025,
      "memory(GiB)": 77.56,
      "step": 58020,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.438579
    },
    {
      "epoch": 2.4859688959341932,
      "grad_norm": 6.049983501434326,
      "learning_rate": 5.044079442460021e-05,
      "loss": 2.490147018432617,
      "memory(GiB)": 77.56,
      "step": 58025,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.438566
    },
    {
      "epoch": 2.486183111263442,
      "grad_norm": 5.0744452476501465,
      "learning_rate": 5.043406490910497e-05,
      "loss": 2.259611129760742,
      "memory(GiB)": 77.56,
      "step": 58030,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.438555
    },
    {
      "epoch": 2.486397326592691,
      "grad_norm": 7.369129657745361,
      "learning_rate": 5.0427335385746244e-05,
      "loss": 2.3613285064697265,
      "memory(GiB)": 77.56,
      "step": 58035,
      "token_acc": 0.5105740181268882,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 2.48661154192194,
      "grad_norm": 6.428347587585449,
      "learning_rate": 5.042060585464592e-05,
      "loss": 2.57977352142334,
      "memory(GiB)": 77.56,
      "step": 58040,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 2.486825757251189,
      "grad_norm": 4.430943965911865,
      "learning_rate": 5.0413876315925925e-05,
      "loss": 2.448247718811035,
      "memory(GiB)": 77.56,
      "step": 58045,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.438599
    },
    {
      "epoch": 2.4870399725804377,
      "grad_norm": 5.167415142059326,
      "learning_rate": 5.040714676970819e-05,
      "loss": 2.2805667877197267,
      "memory(GiB)": 77.56,
      "step": 58050,
      "token_acc": 0.5148514851485149,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 2.487254187909687,
      "grad_norm": 5.407455921173096,
      "learning_rate": 5.04004172161146e-05,
      "loss": 2.453420639038086,
      "memory(GiB)": 77.56,
      "step": 58055,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.43861
    },
    {
      "epoch": 2.487468403238936,
      "grad_norm": 4.259662628173828,
      "learning_rate": 5.039368765526708e-05,
      "loss": 1.986313247680664,
      "memory(GiB)": 77.56,
      "step": 58060,
      "token_acc": 0.5276752767527675,
      "train_speed(iter/s)": 1.438629
    },
    {
      "epoch": 2.4876826185681846,
      "grad_norm": 5.050387382507324,
      "learning_rate": 5.038695808728754e-05,
      "loss": 2.3218318939208986,
      "memory(GiB)": 77.56,
      "step": 58065,
      "token_acc": 0.5307692307692308,
      "train_speed(iter/s)": 1.438643
    },
    {
      "epoch": 2.487896833897434,
      "grad_norm": 7.325018882751465,
      "learning_rate": 5.038022851229789e-05,
      "loss": 2.463140678405762,
      "memory(GiB)": 77.56,
      "step": 58070,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438634
    },
    {
      "epoch": 2.4881110492266827,
      "grad_norm": 6.819101810455322,
      "learning_rate": 5.037349893042005e-05,
      "loss": 2.5707046508789064,
      "memory(GiB)": 77.56,
      "step": 58075,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.43866
    },
    {
      "epoch": 2.4883252645559315,
      "grad_norm": 5.634852409362793,
      "learning_rate": 5.036676934177591e-05,
      "loss": 2.4254791259765627,
      "memory(GiB)": 77.56,
      "step": 58080,
      "token_acc": 0.45857988165680474,
      "train_speed(iter/s)": 1.438679
    },
    {
      "epoch": 2.4885394798851808,
      "grad_norm": 5.327877521514893,
      "learning_rate": 5.036003974648741e-05,
      "loss": 2.4051029205322267,
      "memory(GiB)": 77.56,
      "step": 58085,
      "token_acc": 0.47674418604651164,
      "train_speed(iter/s)": 1.438726
    },
    {
      "epoch": 2.4887536952144296,
      "grad_norm": 7.217409133911133,
      "learning_rate": 5.0353310144676445e-05,
      "loss": 2.490331268310547,
      "memory(GiB)": 77.56,
      "step": 58090,
      "token_acc": 0.5193798449612403,
      "train_speed(iter/s)": 1.43869
    },
    {
      "epoch": 2.4889679105436784,
      "grad_norm": 5.765705585479736,
      "learning_rate": 5.034658053646495e-05,
      "loss": 2.113507843017578,
      "memory(GiB)": 77.56,
      "step": 58095,
      "token_acc": 0.5283018867924528,
      "train_speed(iter/s)": 1.438682
    },
    {
      "epoch": 2.4891821258729276,
      "grad_norm": 6.187605381011963,
      "learning_rate": 5.033985092197481e-05,
      "loss": 2.472217559814453,
      "memory(GiB)": 77.56,
      "step": 58100,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.438693
    },
    {
      "epoch": 2.4893963412021765,
      "grad_norm": 4.89593505859375,
      "learning_rate": 5.033312130132796e-05,
      "loss": 2.4233102798461914,
      "memory(GiB)": 77.56,
      "step": 58105,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.438707
    },
    {
      "epoch": 2.4896105565314253,
      "grad_norm": 5.415012836456299,
      "learning_rate": 5.03263916746463e-05,
      "loss": 2.6598560333251955,
      "memory(GiB)": 77.56,
      "step": 58110,
      "token_acc": 0.4523076923076923,
      "train_speed(iter/s)": 1.438724
    },
    {
      "epoch": 2.4898247718606745,
      "grad_norm": 4.77933931350708,
      "learning_rate": 5.031966204205175e-05,
      "loss": 2.501139450073242,
      "memory(GiB)": 77.56,
      "step": 58115,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.438742
    },
    {
      "epoch": 2.4900389871899233,
      "grad_norm": 6.370707035064697,
      "learning_rate": 5.0312932403666214e-05,
      "loss": 2.31909294128418,
      "memory(GiB)": 77.56,
      "step": 58120,
      "token_acc": 0.5321888412017167,
      "train_speed(iter/s)": 1.438773
    },
    {
      "epoch": 2.490253202519172,
      "grad_norm": 4.952884197235107,
      "learning_rate": 5.0306202759611634e-05,
      "loss": 2.438155746459961,
      "memory(GiB)": 77.56,
      "step": 58125,
      "token_acc": 0.49171270718232046,
      "train_speed(iter/s)": 1.438796
    },
    {
      "epoch": 2.4904674178484214,
      "grad_norm": 4.658510684967041,
      "learning_rate": 5.0299473110009876e-05,
      "loss": 2.148599624633789,
      "memory(GiB)": 77.56,
      "step": 58130,
      "token_acc": 0.524,
      "train_speed(iter/s)": 1.438824
    },
    {
      "epoch": 2.49068163317767,
      "grad_norm": 6.677910804748535,
      "learning_rate": 5.0292743454982905e-05,
      "loss": 2.5059249877929686,
      "memory(GiB)": 77.56,
      "step": 58135,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.438836
    },
    {
      "epoch": 2.490895848506919,
      "grad_norm": 6.639410972595215,
      "learning_rate": 5.028601379465261e-05,
      "loss": 2.759622573852539,
      "memory(GiB)": 77.56,
      "step": 58140,
      "token_acc": 0.45723684210526316,
      "train_speed(iter/s)": 1.438838
    },
    {
      "epoch": 2.4911100638361683,
      "grad_norm": 5.123559951782227,
      "learning_rate": 5.0279284129140894e-05,
      "loss": 2.5407032012939452,
      "memory(GiB)": 77.56,
      "step": 58145,
      "token_acc": 0.46113989637305697,
      "train_speed(iter/s)": 1.438817
    },
    {
      "epoch": 2.491324279165417,
      "grad_norm": 4.856555461883545,
      "learning_rate": 5.027255445856969e-05,
      "loss": 2.4736461639404297,
      "memory(GiB)": 77.56,
      "step": 58150,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.438835
    },
    {
      "epoch": 2.491538494494666,
      "grad_norm": 5.092394828796387,
      "learning_rate": 5.0265824783060914e-05,
      "loss": 2.3582536697387697,
      "memory(GiB)": 77.56,
      "step": 58155,
      "token_acc": 0.5138339920948617,
      "train_speed(iter/s)": 1.438825
    },
    {
      "epoch": 2.491752709823915,
      "grad_norm": 5.648548126220703,
      "learning_rate": 5.025909510273645e-05,
      "loss": 2.356985092163086,
      "memory(GiB)": 77.56,
      "step": 58160,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.438816
    },
    {
      "epoch": 2.491966925153164,
      "grad_norm": 6.463406085968018,
      "learning_rate": 5.025236541771826e-05,
      "loss": 2.3251243591308595,
      "memory(GiB)": 77.56,
      "step": 58165,
      "token_acc": 0.5204918032786885,
      "train_speed(iter/s)": 1.438768
    },
    {
      "epoch": 2.492181140482413,
      "grad_norm": 6.251805782318115,
      "learning_rate": 5.0245635728128225e-05,
      "loss": 2.4452255249023436,
      "memory(GiB)": 77.56,
      "step": 58170,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.438792
    },
    {
      "epoch": 2.492395355811662,
      "grad_norm": 4.523138046264648,
      "learning_rate": 5.023890603408827e-05,
      "loss": 2.1374338150024412,
      "memory(GiB)": 77.56,
      "step": 58175,
      "token_acc": 0.5181518151815182,
      "train_speed(iter/s)": 1.438803
    },
    {
      "epoch": 2.492609571140911,
      "grad_norm": 4.479569435119629,
      "learning_rate": 5.0232176335720314e-05,
      "loss": 2.4535398483276367,
      "memory(GiB)": 77.56,
      "step": 58180,
      "token_acc": 0.4823943661971831,
      "train_speed(iter/s)": 1.438813
    },
    {
      "epoch": 2.4928237864701597,
      "grad_norm": 5.842418670654297,
      "learning_rate": 5.0225446633146247e-05,
      "loss": 2.688613510131836,
      "memory(GiB)": 77.56,
      "step": 58185,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.438798
    },
    {
      "epoch": 2.493038001799409,
      "grad_norm": 5.621906280517578,
      "learning_rate": 5.021871692648801e-05,
      "loss": 2.476247215270996,
      "memory(GiB)": 77.56,
      "step": 58190,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.438827
    },
    {
      "epoch": 2.4932522171286577,
      "grad_norm": 6.361431121826172,
      "learning_rate": 5.0211987215867516e-05,
      "loss": 2.403704833984375,
      "memory(GiB)": 77.56,
      "step": 58195,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.438834
    },
    {
      "epoch": 2.4934664324579066,
      "grad_norm": 4.427335262298584,
      "learning_rate": 5.020525750140668e-05,
      "loss": 2.3656896591186523,
      "memory(GiB)": 77.56,
      "step": 58200,
      "token_acc": 0.45689655172413796,
      "train_speed(iter/s)": 1.438856
    },
    {
      "epoch": 2.493680647787156,
      "grad_norm": 4.050919532775879,
      "learning_rate": 5.01985277832274e-05,
      "loss": 2.3997989654541017,
      "memory(GiB)": 77.56,
      "step": 58205,
      "token_acc": 0.5106382978723404,
      "train_speed(iter/s)": 1.438855
    },
    {
      "epoch": 2.4938948631164046,
      "grad_norm": 6.055706977844238,
      "learning_rate": 5.0191798061451614e-05,
      "loss": 2.4805131912231446,
      "memory(GiB)": 77.56,
      "step": 58210,
      "token_acc": 0.483739837398374,
      "train_speed(iter/s)": 1.438839
    },
    {
      "epoch": 2.4941090784456534,
      "grad_norm": 5.819148063659668,
      "learning_rate": 5.0185068336201215e-05,
      "loss": 2.405121612548828,
      "memory(GiB)": 77.56,
      "step": 58215,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.438802
    },
    {
      "epoch": 2.4943232937749027,
      "grad_norm": 4.424496650695801,
      "learning_rate": 5.0178338607598144e-05,
      "loss": 2.3622785568237306,
      "memory(GiB)": 77.56,
      "step": 58220,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.438797
    },
    {
      "epoch": 2.4945375091041515,
      "grad_norm": 6.676568031311035,
      "learning_rate": 5.0171608875764286e-05,
      "loss": 2.3665557861328126,
      "memory(GiB)": 77.56,
      "step": 58225,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.438766
    },
    {
      "epoch": 2.4947517244334003,
      "grad_norm": 5.545111179351807,
      "learning_rate": 5.0164879140821586e-05,
      "loss": 2.52616024017334,
      "memory(GiB)": 77.56,
      "step": 58230,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.438747
    },
    {
      "epoch": 2.4949659397626496,
      "grad_norm": 5.15902853012085,
      "learning_rate": 5.015814940289193e-05,
      "loss": 2.2114841461181642,
      "memory(GiB)": 77.56,
      "step": 58235,
      "token_acc": 0.5377643504531722,
      "train_speed(iter/s)": 1.438779
    },
    {
      "epoch": 2.4951801550918984,
      "grad_norm": 3.7115330696105957,
      "learning_rate": 5.015141966209726e-05,
      "loss": 2.3777944564819338,
      "memory(GiB)": 77.56,
      "step": 58240,
      "token_acc": 0.519774011299435,
      "train_speed(iter/s)": 1.438804
    },
    {
      "epoch": 2.495394370421147,
      "grad_norm": 5.075016021728516,
      "learning_rate": 5.014468991855947e-05,
      "loss": 2.677511978149414,
      "memory(GiB)": 77.56,
      "step": 58245,
      "token_acc": 0.4676258992805755,
      "train_speed(iter/s)": 1.438808
    },
    {
      "epoch": 2.4956085857503965,
      "grad_norm": 4.5342278480529785,
      "learning_rate": 5.01379601724005e-05,
      "loss": 2.2680641174316407,
      "memory(GiB)": 77.56,
      "step": 58250,
      "token_acc": 0.5115606936416185,
      "train_speed(iter/s)": 1.438831
    },
    {
      "epoch": 2.4958228010796453,
      "grad_norm": 4.905433177947998,
      "learning_rate": 5.013123042374226e-05,
      "loss": 2.5868595123291014,
      "memory(GiB)": 77.56,
      "step": 58255,
      "token_acc": 0.4689265536723164,
      "train_speed(iter/s)": 1.438847
    },
    {
      "epoch": 2.496037016408894,
      "grad_norm": 7.08261251449585,
      "learning_rate": 5.012450067270663e-05,
      "loss": 2.3131925582885744,
      "memory(GiB)": 77.56,
      "step": 58260,
      "token_acc": 0.47634069400630913,
      "train_speed(iter/s)": 1.438859
    },
    {
      "epoch": 2.4962512317381433,
      "grad_norm": 5.00227689743042,
      "learning_rate": 5.0117770919415566e-05,
      "loss": 2.2290515899658203,
      "memory(GiB)": 77.56,
      "step": 58265,
      "token_acc": 0.5625,
      "train_speed(iter/s)": 1.438847
    },
    {
      "epoch": 2.496465447067392,
      "grad_norm": 4.830747604370117,
      "learning_rate": 5.011104116399098e-05,
      "loss": 2.41119441986084,
      "memory(GiB)": 77.56,
      "step": 58270,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.43882
    },
    {
      "epoch": 2.496679662396641,
      "grad_norm": 5.60421895980835,
      "learning_rate": 5.010431140655477e-05,
      "loss": 2.232265281677246,
      "memory(GiB)": 77.56,
      "step": 58275,
      "token_acc": 0.5193798449612403,
      "train_speed(iter/s)": 1.438806
    },
    {
      "epoch": 2.49689387772589,
      "grad_norm": 7.05345344543457,
      "learning_rate": 5.009758164722887e-05,
      "loss": 2.3105337142944338,
      "memory(GiB)": 77.56,
      "step": 58280,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.438793
    },
    {
      "epoch": 2.497108093055139,
      "grad_norm": 6.987645149230957,
      "learning_rate": 5.0090851886135184e-05,
      "loss": 2.398975372314453,
      "memory(GiB)": 77.56,
      "step": 58285,
      "token_acc": 0.4495798319327731,
      "train_speed(iter/s)": 1.438755
    },
    {
      "epoch": 2.497322308384388,
      "grad_norm": 4.815450191497803,
      "learning_rate": 5.0084122123395615e-05,
      "loss": 2.3575220108032227,
      "memory(GiB)": 77.56,
      "step": 58290,
      "token_acc": 0.4837662337662338,
      "train_speed(iter/s)": 1.438779
    },
    {
      "epoch": 2.497536523713637,
      "grad_norm": 5.71390962600708,
      "learning_rate": 5.007739235913211e-05,
      "loss": 2.4685995101928713,
      "memory(GiB)": 77.56,
      "step": 58295,
      "token_acc": 0.4592833876221498,
      "train_speed(iter/s)": 1.438766
    },
    {
      "epoch": 2.497750739042886,
      "grad_norm": 5.189290523529053,
      "learning_rate": 5.0070662593466564e-05,
      "loss": 2.3445167541503906,
      "memory(GiB)": 77.56,
      "step": 58300,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.438776
    },
    {
      "epoch": 2.4979649543721347,
      "grad_norm": 8.077359199523926,
      "learning_rate": 5.00639328265209e-05,
      "loss": 2.5039722442626955,
      "memory(GiB)": 77.56,
      "step": 58305,
      "token_acc": 0.497907949790795,
      "train_speed(iter/s)": 1.438741
    },
    {
      "epoch": 2.498179169701384,
      "grad_norm": 7.4290242195129395,
      "learning_rate": 5.0057203058417044e-05,
      "loss": 2.2886486053466797,
      "memory(GiB)": 77.56,
      "step": 58310,
      "token_acc": 0.48091603053435117,
      "train_speed(iter/s)": 1.438759
    },
    {
      "epoch": 2.498393385030633,
      "grad_norm": 6.249179840087891,
      "learning_rate": 5.00504732892769e-05,
      "loss": 2.6114885330200197,
      "memory(GiB)": 77.56,
      "step": 58315,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.438774
    },
    {
      "epoch": 2.4986076003598816,
      "grad_norm": 3.9913127422332764,
      "learning_rate": 5.004374351922236e-05,
      "loss": 2.4197288513183595,
      "memory(GiB)": 77.56,
      "step": 58320,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.438807
    },
    {
      "epoch": 2.498821815689131,
      "grad_norm": 5.795289039611816,
      "learning_rate": 5.003701374837539e-05,
      "loss": 2.378787612915039,
      "memory(GiB)": 77.56,
      "step": 58325,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.438796
    },
    {
      "epoch": 2.4990360310183797,
      "grad_norm": 11.66122817993164,
      "learning_rate": 5.003028397685787e-05,
      "loss": 2.6223936080932617,
      "memory(GiB)": 77.56,
      "step": 58330,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.438783
    },
    {
      "epoch": 2.4992502463476285,
      "grad_norm": 5.806344985961914,
      "learning_rate": 5.0023554204791724e-05,
      "loss": 2.4009864807128904,
      "memory(GiB)": 77.56,
      "step": 58335,
      "token_acc": 0.4690909090909091,
      "train_speed(iter/s)": 1.438818
    },
    {
      "epoch": 2.4994644616768777,
      "grad_norm": 5.129970550537109,
      "learning_rate": 5.001682443229888e-05,
      "loss": 2.718728446960449,
      "memory(GiB)": 77.56,
      "step": 58340,
      "token_acc": 0.45819397993311034,
      "train_speed(iter/s)": 1.438846
    },
    {
      "epoch": 2.4996786770061266,
      "grad_norm": 5.388007640838623,
      "learning_rate": 5.001009465950125e-05,
      "loss": 2.529027557373047,
      "memory(GiB)": 77.56,
      "step": 58345,
      "token_acc": 0.4740061162079511,
      "train_speed(iter/s)": 1.438881
    },
    {
      "epoch": 2.4998928923353754,
      "grad_norm": 5.799760341644287,
      "learning_rate": 5.000336488652074e-05,
      "loss": 2.488451385498047,
      "memory(GiB)": 77.56,
      "step": 58350,
      "token_acc": 0.4489051094890511,
      "train_speed(iter/s)": 1.438867
    },
    {
      "epoch": 2.5001071076646246,
      "grad_norm": 5.171080589294434,
      "learning_rate": 4.9996635113479265e-05,
      "loss": 2.423663330078125,
      "memory(GiB)": 77.56,
      "step": 58355,
      "token_acc": 0.5107296137339056,
      "train_speed(iter/s)": 1.438878
    },
    {
      "epoch": 2.5003213229938734,
      "grad_norm": 5.024204254150391,
      "learning_rate": 4.9989905340498755e-05,
      "loss": 2.700656700134277,
      "memory(GiB)": 77.56,
      "step": 58360,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.438893
    },
    {
      "epoch": 2.5005355383231223,
      "grad_norm": 5.59718656539917,
      "learning_rate": 4.998317556770111e-05,
      "loss": 2.100053596496582,
      "memory(GiB)": 77.56,
      "step": 58365,
      "token_acc": 0.5300353356890459,
      "train_speed(iter/s)": 1.438878
    },
    {
      "epoch": 2.5007497536523715,
      "grad_norm": 5.861362457275391,
      "learning_rate": 4.9976445795208274e-05,
      "loss": 2.3406686782836914,
      "memory(GiB)": 77.56,
      "step": 58370,
      "token_acc": 0.5064102564102564,
      "train_speed(iter/s)": 1.438885
    },
    {
      "epoch": 2.5009639689816203,
      "grad_norm": 6.280579090118408,
      "learning_rate": 4.996971602314214e-05,
      "loss": 2.453501892089844,
      "memory(GiB)": 77.56,
      "step": 58375,
      "token_acc": 0.47093023255813954,
      "train_speed(iter/s)": 1.438899
    },
    {
      "epoch": 2.501178184310869,
      "grad_norm": 4.300484657287598,
      "learning_rate": 4.996298625162462e-05,
      "loss": 2.253561592102051,
      "memory(GiB)": 77.56,
      "step": 58380,
      "token_acc": 0.5032467532467533,
      "train_speed(iter/s)": 1.438865
    },
    {
      "epoch": 2.5013923996401184,
      "grad_norm": 6.742525100708008,
      "learning_rate": 4.9956256480777643e-05,
      "loss": 2.060250473022461,
      "memory(GiB)": 77.56,
      "step": 58385,
      "token_acc": 0.5444444444444444,
      "train_speed(iter/s)": 1.438854
    },
    {
      "epoch": 2.501606614969367,
      "grad_norm": 6.936786651611328,
      "learning_rate": 4.994952671072312e-05,
      "loss": 2.442148780822754,
      "memory(GiB)": 77.56,
      "step": 58390,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.438881
    },
    {
      "epoch": 2.501820830298616,
      "grad_norm": 6.741944313049316,
      "learning_rate": 4.994279694158297e-05,
      "loss": 2.4010562896728516,
      "memory(GiB)": 77.56,
      "step": 58395,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.438877
    },
    {
      "epoch": 2.5020350456278653,
      "grad_norm": 6.655197620391846,
      "learning_rate": 4.993606717347911e-05,
      "loss": 2.3350038528442383,
      "memory(GiB)": 77.56,
      "step": 58400,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.438872
    },
    {
      "epoch": 2.502249260957114,
      "grad_norm": 5.165590286254883,
      "learning_rate": 4.992933740653345e-05,
      "loss": 2.0496925354003905,
      "memory(GiB)": 77.56,
      "step": 58405,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.438891
    },
    {
      "epoch": 2.502463476286363,
      "grad_norm": 6.217377185821533,
      "learning_rate": 4.99226076408679e-05,
      "loss": 2.4300582885742186,
      "memory(GiB)": 77.56,
      "step": 58410,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.438891
    },
    {
      "epoch": 2.502677691615612,
      "grad_norm": 5.3018293380737305,
      "learning_rate": 4.99158778766044e-05,
      "loss": 2.6122314453125,
      "memory(GiB)": 77.56,
      "step": 58415,
      "token_acc": 0.4522058823529412,
      "train_speed(iter/s)": 1.438888
    },
    {
      "epoch": 2.502891906944861,
      "grad_norm": 4.7727460861206055,
      "learning_rate": 4.990914811386484e-05,
      "loss": 2.798413848876953,
      "memory(GiB)": 77.56,
      "step": 58420,
      "token_acc": 0.45054945054945056,
      "train_speed(iter/s)": 1.43891
    },
    {
      "epoch": 2.50310612227411,
      "grad_norm": 6.41102409362793,
      "learning_rate": 4.990241835277116e-05,
      "loss": 2.8175573348999023,
      "memory(GiB)": 77.56,
      "step": 58425,
      "token_acc": 0.4487534626038781,
      "train_speed(iter/s)": 1.438908
    },
    {
      "epoch": 2.503320337603359,
      "grad_norm": 4.640605926513672,
      "learning_rate": 4.989568859344523e-05,
      "loss": 2.5508893966674804,
      "memory(GiB)": 77.56,
      "step": 58430,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.438906
    },
    {
      "epoch": 2.503534552932608,
      "grad_norm": 5.336583137512207,
      "learning_rate": 4.988895883600902e-05,
      "loss": 2.527036666870117,
      "memory(GiB)": 77.56,
      "step": 58435,
      "token_acc": 0.48398576512455516,
      "train_speed(iter/s)": 1.438922
    },
    {
      "epoch": 2.5037487682618567,
      "grad_norm": 5.696075439453125,
      "learning_rate": 4.988222908058443e-05,
      "loss": 2.3428720474243163,
      "memory(GiB)": 77.56,
      "step": 58440,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.438934
    },
    {
      "epoch": 2.503962983591106,
      "grad_norm": 4.8912248611450195,
      "learning_rate": 4.987549932729337e-05,
      "loss": 2.1608455657958983,
      "memory(GiB)": 77.56,
      "step": 58445,
      "token_acc": 0.5457627118644067,
      "train_speed(iter/s)": 1.438891
    },
    {
      "epoch": 2.5041771989203547,
      "grad_norm": 5.038296699523926,
      "learning_rate": 4.986876957625776e-05,
      "loss": 2.2645145416259767,
      "memory(GiB)": 77.56,
      "step": 58450,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.438916
    },
    {
      "epoch": 2.5043914142496035,
      "grad_norm": 5.156894207000732,
      "learning_rate": 4.98620398275995e-05,
      "loss": 2.403422164916992,
      "memory(GiB)": 77.56,
      "step": 58455,
      "token_acc": 0.48135593220338985,
      "train_speed(iter/s)": 1.43891
    },
    {
      "epoch": 2.504605629578853,
      "grad_norm": 5.4597859382629395,
      "learning_rate": 4.9855310081440535e-05,
      "loss": 2.6115291595458983,
      "memory(GiB)": 77.56,
      "step": 58460,
      "token_acc": 0.46417445482866043,
      "train_speed(iter/s)": 1.438935
    },
    {
      "epoch": 2.5048198449081016,
      "grad_norm": 6.157292366027832,
      "learning_rate": 4.984858033790275e-05,
      "loss": 2.4855825424194338,
      "memory(GiB)": 77.56,
      "step": 58465,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.438922
    },
    {
      "epoch": 2.5050340602373504,
      "grad_norm": 6.037020206451416,
      "learning_rate": 4.984185059710808e-05,
      "loss": 2.2975700378417967,
      "memory(GiB)": 77.56,
      "step": 58470,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438932
    },
    {
      "epoch": 2.5052482755665997,
      "grad_norm": 5.616464138031006,
      "learning_rate": 4.983512085917843e-05,
      "loss": 2.3390806198120115,
      "memory(GiB)": 77.56,
      "step": 58475,
      "token_acc": 0.4607142857142857,
      "train_speed(iter/s)": 1.4389
    },
    {
      "epoch": 2.5054624908958485,
      "grad_norm": 5.004324913024902,
      "learning_rate": 4.9828391124235726e-05,
      "loss": 2.354347801208496,
      "memory(GiB)": 77.56,
      "step": 58480,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.438869
    },
    {
      "epoch": 2.5056767062250973,
      "grad_norm": 6.0675950050354,
      "learning_rate": 4.982166139240188e-05,
      "loss": 2.6349502563476563,
      "memory(GiB)": 77.56,
      "step": 58485,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.438883
    },
    {
      "epoch": 2.5058909215543466,
      "grad_norm": 5.169259548187256,
      "learning_rate": 4.98149316637988e-05,
      "loss": 2.3559337615966798,
      "memory(GiB)": 77.56,
      "step": 58490,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.438876
    },
    {
      "epoch": 2.5061051368835954,
      "grad_norm": 5.194831371307373,
      "learning_rate": 4.9808201938548404e-05,
      "loss": 2.4086750030517576,
      "memory(GiB)": 77.56,
      "step": 58495,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.438893
    },
    {
      "epoch": 2.506319352212844,
      "grad_norm": 8.391009330749512,
      "learning_rate": 4.980147221677262e-05,
      "loss": 2.6264013290405273,
      "memory(GiB)": 77.56,
      "step": 58500,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.438906
    },
    {
      "epoch": 2.506319352212844,
      "eval_loss": 2.080378770828247,
      "eval_runtime": 15.1383,
      "eval_samples_per_second": 6.606,
      "eval_steps_per_second": 6.606,
      "eval_token_acc": 0.4725433526011561,
      "step": 58500
    },
    {
      "epoch": 2.5065335675420934,
      "grad_norm": 4.968241214752197,
      "learning_rate": 4.979474249859333e-05,
      "loss": 2.5486387252807616,
      "memory(GiB)": 77.56,
      "step": 58505,
      "token_acc": 0.4665314401622718,
      "train_speed(iter/s)": 1.438341
    },
    {
      "epoch": 2.5067477828713423,
      "grad_norm": 4.82818078994751,
      "learning_rate": 4.978801278413248e-05,
      "loss": 2.262587547302246,
      "memory(GiB)": 77.56,
      "step": 58510,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 2.506961998200591,
      "grad_norm": 6.097710609436035,
      "learning_rate": 4.9781283073511995e-05,
      "loss": 2.1054759979248048,
      "memory(GiB)": 77.56,
      "step": 58515,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.438401
    },
    {
      "epoch": 2.5071762135298403,
      "grad_norm": 6.493237018585205,
      "learning_rate": 4.977455336685376e-05,
      "loss": 2.6580795288085937,
      "memory(GiB)": 77.56,
      "step": 58520,
      "token_acc": 0.43882978723404253,
      "train_speed(iter/s)": 1.438392
    },
    {
      "epoch": 2.507390428859089,
      "grad_norm": 5.894067764282227,
      "learning_rate": 4.97678236642797e-05,
      "loss": 2.515596389770508,
      "memory(GiB)": 77.56,
      "step": 58525,
      "token_acc": 0.4986376021798365,
      "train_speed(iter/s)": 1.438414
    },
    {
      "epoch": 2.507604644188338,
      "grad_norm": 6.168473243713379,
      "learning_rate": 4.9761093965911746e-05,
      "loss": 2.4529163360595705,
      "memory(GiB)": 77.56,
      "step": 58530,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.438418
    },
    {
      "epoch": 2.507818859517587,
      "grad_norm": 5.022265434265137,
      "learning_rate": 4.975436427187178e-05,
      "loss": 2.247749137878418,
      "memory(GiB)": 77.56,
      "step": 58535,
      "token_acc": 0.512987012987013,
      "train_speed(iter/s)": 1.438442
    },
    {
      "epoch": 2.508033074846836,
      "grad_norm": 4.225148677825928,
      "learning_rate": 4.974763458228175e-05,
      "loss": 2.4482269287109375,
      "memory(GiB)": 77.56,
      "step": 58540,
      "token_acc": 0.4831804281345566,
      "train_speed(iter/s)": 1.438423
    },
    {
      "epoch": 2.508247290176085,
      "grad_norm": 6.861063480377197,
      "learning_rate": 4.9740904897263554e-05,
      "loss": 2.2667058944702148,
      "memory(GiB)": 77.56,
      "step": 58545,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.438421
    },
    {
      "epoch": 2.508461505505334,
      "grad_norm": 5.056663990020752,
      "learning_rate": 4.97341752169391e-05,
      "loss": 2.3969274520874024,
      "memory(GiB)": 77.56,
      "step": 58550,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.438431
    },
    {
      "epoch": 2.508675720834583,
      "grad_norm": 5.804452896118164,
      "learning_rate": 4.972744554143032e-05,
      "loss": 2.471655082702637,
      "memory(GiB)": 77.56,
      "step": 58555,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.438441
    },
    {
      "epoch": 2.5088899361638317,
      "grad_norm": 5.051337718963623,
      "learning_rate": 4.9720715870859124e-05,
      "loss": 2.5114559173583983,
      "memory(GiB)": 77.56,
      "step": 58560,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.438455
    },
    {
      "epoch": 2.509104151493081,
      "grad_norm": 5.521915435791016,
      "learning_rate": 4.971398620534741e-05,
      "loss": 2.2452049255371094,
      "memory(GiB)": 77.56,
      "step": 58565,
      "token_acc": 0.5461847389558233,
      "train_speed(iter/s)": 1.438445
    },
    {
      "epoch": 2.50931836682233,
      "grad_norm": 6.709741592407227,
      "learning_rate": 4.970725654501712e-05,
      "loss": 2.4361576080322265,
      "memory(GiB)": 77.56,
      "step": 58570,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.438415
    },
    {
      "epoch": 2.5095325821515786,
      "grad_norm": 9.513143539428711,
      "learning_rate": 4.970052688999012e-05,
      "loss": 2.603710746765137,
      "memory(GiB)": 77.56,
      "step": 58575,
      "token_acc": 0.4778761061946903,
      "train_speed(iter/s)": 1.438409
    },
    {
      "epoch": 2.509746797480828,
      "grad_norm": 6.703118801116943,
      "learning_rate": 4.969379724038837e-05,
      "loss": 2.8177309036254883,
      "memory(GiB)": 77.56,
      "step": 58580,
      "token_acc": 0.4725274725274725,
      "train_speed(iter/s)": 1.438419
    },
    {
      "epoch": 2.5099610128100767,
      "grad_norm": 6.20990514755249,
      "learning_rate": 4.9687067596333784e-05,
      "loss": 2.4423900604248048,
      "memory(GiB)": 77.56,
      "step": 58585,
      "token_acc": 0.4927007299270073,
      "train_speed(iter/s)": 1.438425
    },
    {
      "epoch": 2.5101752281393255,
      "grad_norm": 6.054320335388184,
      "learning_rate": 4.968033795794827e-05,
      "loss": 2.502968406677246,
      "memory(GiB)": 77.56,
      "step": 58590,
      "token_acc": 0.4367816091954023,
      "train_speed(iter/s)": 1.438433
    },
    {
      "epoch": 2.5103894434685747,
      "grad_norm": 6.248667240142822,
      "learning_rate": 4.967360832535371e-05,
      "loss": 2.3279033660888673,
      "memory(GiB)": 77.56,
      "step": 58595,
      "token_acc": 0.5095541401273885,
      "train_speed(iter/s)": 1.438413
    },
    {
      "epoch": 2.5106036587978235,
      "grad_norm": 7.26310396194458,
      "learning_rate": 4.9666878698672054e-05,
      "loss": 2.630337142944336,
      "memory(GiB)": 77.56,
      "step": 58600,
      "token_acc": 0.44696969696969696,
      "train_speed(iter/s)": 1.438407
    },
    {
      "epoch": 2.5108178741270724,
      "grad_norm": 6.325886249542236,
      "learning_rate": 4.96601490780252e-05,
      "loss": 2.371360015869141,
      "memory(GiB)": 77.56,
      "step": 58605,
      "token_acc": 0.4672897196261682,
      "train_speed(iter/s)": 1.438407
    },
    {
      "epoch": 2.5110320894563216,
      "grad_norm": 4.498351573944092,
      "learning_rate": 4.965341946353506e-05,
      "loss": 2.154536247253418,
      "memory(GiB)": 77.56,
      "step": 58610,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.438412
    },
    {
      "epoch": 2.5112463047855704,
      "grad_norm": 6.498818874359131,
      "learning_rate": 4.964668985532357e-05,
      "loss": 2.3352787017822267,
      "memory(GiB)": 77.56,
      "step": 58615,
      "token_acc": 0.4793650793650794,
      "train_speed(iter/s)": 1.438424
    },
    {
      "epoch": 2.5114605201148192,
      "grad_norm": 6.107295513153076,
      "learning_rate": 4.96399602535126e-05,
      "loss": 2.575478363037109,
      "memory(GiB)": 77.56,
      "step": 58620,
      "token_acc": 0.45741324921135645,
      "train_speed(iter/s)": 1.438429
    },
    {
      "epoch": 2.5116747354440685,
      "grad_norm": 4.7984771728515625,
      "learning_rate": 4.9633230658224106e-05,
      "loss": 2.307478904724121,
      "memory(GiB)": 77.56,
      "step": 58625,
      "token_acc": 0.5157894736842106,
      "train_speed(iter/s)": 1.438417
    },
    {
      "epoch": 2.5118889507733173,
      "grad_norm": 5.4650068283081055,
      "learning_rate": 4.9626501069579975e-05,
      "loss": 2.367713737487793,
      "memory(GiB)": 77.56,
      "step": 58630,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.438419
    },
    {
      "epoch": 2.512103166102566,
      "grad_norm": 5.911473274230957,
      "learning_rate": 4.961977148770212e-05,
      "loss": 2.188696098327637,
      "memory(GiB)": 77.56,
      "step": 58635,
      "token_acc": 0.5346938775510204,
      "train_speed(iter/s)": 1.438446
    },
    {
      "epoch": 2.5123173814318154,
      "grad_norm": 4.255577087402344,
      "learning_rate": 4.961304191271248e-05,
      "loss": 2.444392776489258,
      "memory(GiB)": 77.56,
      "step": 58640,
      "token_acc": 0.5316901408450704,
      "train_speed(iter/s)": 1.438466
    },
    {
      "epoch": 2.512531596761064,
      "grad_norm": 7.421138286590576,
      "learning_rate": 4.960631234473294e-05,
      "loss": 2.5387012481689455,
      "memory(GiB)": 77.56,
      "step": 58645,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.438485
    },
    {
      "epoch": 2.512745812090313,
      "grad_norm": 6.129209041595459,
      "learning_rate": 4.9599582783885404e-05,
      "loss": 2.654239273071289,
      "memory(GiB)": 77.56,
      "step": 58650,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.438497
    },
    {
      "epoch": 2.5129600274195623,
      "grad_norm": 5.452489376068115,
      "learning_rate": 4.959285323029181e-05,
      "loss": 2.670383071899414,
      "memory(GiB)": 77.56,
      "step": 58655,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.438533
    },
    {
      "epoch": 2.513174242748811,
      "grad_norm": 5.070569038391113,
      "learning_rate": 4.958612368407407e-05,
      "loss": 2.357744598388672,
      "memory(GiB)": 77.56,
      "step": 58660,
      "token_acc": 0.4597315436241611,
      "train_speed(iter/s)": 1.438482
    },
    {
      "epoch": 2.51338845807806,
      "grad_norm": 5.5211334228515625,
      "learning_rate": 4.957939414535409e-05,
      "loss": 2.39716796875,
      "memory(GiB)": 77.56,
      "step": 58665,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.438466
    },
    {
      "epoch": 2.513602673407309,
      "grad_norm": 5.426631927490234,
      "learning_rate": 4.9572664614253774e-05,
      "loss": 2.3903064727783203,
      "memory(GiB)": 77.56,
      "step": 58670,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.438488
    },
    {
      "epoch": 2.513816888736558,
      "grad_norm": 4.342236042022705,
      "learning_rate": 4.9565935090895046e-05,
      "loss": 2.535909080505371,
      "memory(GiB)": 77.56,
      "step": 58675,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.438498
    },
    {
      "epoch": 2.5140311040658068,
      "grad_norm": 5.142080307006836,
      "learning_rate": 4.95592055753998e-05,
      "loss": 2.4534364700317384,
      "memory(GiB)": 77.56,
      "step": 58680,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.438468
    },
    {
      "epoch": 2.514245319395056,
      "grad_norm": 6.049755573272705,
      "learning_rate": 4.9552476067889967e-05,
      "loss": 2.5266448974609377,
      "memory(GiB)": 77.56,
      "step": 58685,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.438448
    },
    {
      "epoch": 2.514459534724305,
      "grad_norm": 5.624327182769775,
      "learning_rate": 4.954574656848745e-05,
      "loss": 2.4393081665039062,
      "memory(GiB)": 77.56,
      "step": 58690,
      "token_acc": 0.50814332247557,
      "train_speed(iter/s)": 1.438432
    },
    {
      "epoch": 2.5146737500535536,
      "grad_norm": 5.702749252319336,
      "learning_rate": 4.953901707731415e-05,
      "loss": 2.182010269165039,
      "memory(GiB)": 77.56,
      "step": 58695,
      "token_acc": 0.5212464589235127,
      "train_speed(iter/s)": 1.438419
    },
    {
      "epoch": 2.514887965382803,
      "grad_norm": 4.576151371002197,
      "learning_rate": 4.9532287594491996e-05,
      "loss": 2.0450227737426756,
      "memory(GiB)": 77.56,
      "step": 58700,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.438422
    },
    {
      "epoch": 2.5151021807120517,
      "grad_norm": 8.027562141418457,
      "learning_rate": 4.95255581201429e-05,
      "loss": 2.2008514404296875,
      "memory(GiB)": 77.56,
      "step": 58705,
      "token_acc": 0.5241379310344828,
      "train_speed(iter/s)": 1.438431
    },
    {
      "epoch": 2.5153163960413005,
      "grad_norm": 5.362934112548828,
      "learning_rate": 4.951882865438875e-05,
      "loss": 2.3463413238525392,
      "memory(GiB)": 77.56,
      "step": 58710,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.438459
    },
    {
      "epoch": 2.51553061137055,
      "grad_norm": 7.179898262023926,
      "learning_rate": 4.951209919735148e-05,
      "loss": 2.3349565505981444,
      "memory(GiB)": 77.56,
      "step": 58715,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.438472
    },
    {
      "epoch": 2.5157448266997986,
      "grad_norm": 6.435906887054443,
      "learning_rate": 4.9505369749152975e-05,
      "loss": 2.452790451049805,
      "memory(GiB)": 77.56,
      "step": 58720,
      "token_acc": 0.4555984555984556,
      "train_speed(iter/s)": 1.438487
    },
    {
      "epoch": 2.5159590420290474,
      "grad_norm": 16.25479507446289,
      "learning_rate": 4.9498640309915156e-05,
      "loss": 2.348832702636719,
      "memory(GiB)": 77.56,
      "step": 58725,
      "token_acc": 0.5160142348754448,
      "train_speed(iter/s)": 1.438497
    },
    {
      "epoch": 2.5161732573582967,
      "grad_norm": 5.01768159866333,
      "learning_rate": 4.9491910879759956e-05,
      "loss": 2.618852424621582,
      "memory(GiB)": 77.56,
      "step": 58730,
      "token_acc": 0.49240121580547114,
      "train_speed(iter/s)": 1.438502
    },
    {
      "epoch": 2.5163874726875455,
      "grad_norm": 7.3222832679748535,
      "learning_rate": 4.9485181458809273e-05,
      "loss": 2.6504459381103516,
      "memory(GiB)": 77.56,
      "step": 58735,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 2.5166016880167943,
      "grad_norm": 4.788215160369873,
      "learning_rate": 4.947845204718499e-05,
      "loss": 2.475088882446289,
      "memory(GiB)": 77.56,
      "step": 58740,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.438514
    },
    {
      "epoch": 2.5168159033460435,
      "grad_norm": 4.636257171630859,
      "learning_rate": 4.947172264500905e-05,
      "loss": 2.0276056289672852,
      "memory(GiB)": 77.56,
      "step": 58745,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.438505
    },
    {
      "epoch": 2.5170301186752924,
      "grad_norm": 5.67156457901001,
      "learning_rate": 4.9464993252403366e-05,
      "loss": 2.4555709838867186,
      "memory(GiB)": 77.56,
      "step": 58750,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.438501
    },
    {
      "epoch": 2.517244334004541,
      "grad_norm": 4.645724296569824,
      "learning_rate": 4.945826386948981e-05,
      "loss": 2.2215145111083983,
      "memory(GiB)": 77.56,
      "step": 58755,
      "token_acc": 0.5655172413793104,
      "train_speed(iter/s)": 1.438513
    },
    {
      "epoch": 2.5174585493337904,
      "grad_norm": 5.968151092529297,
      "learning_rate": 4.9451534496390315e-05,
      "loss": 2.5536806106567385,
      "memory(GiB)": 77.56,
      "step": 58760,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.438517
    },
    {
      "epoch": 2.5176727646630392,
      "grad_norm": 5.72506046295166,
      "learning_rate": 4.944480513322678e-05,
      "loss": 2.3667137145996096,
      "memory(GiB)": 77.56,
      "step": 58765,
      "token_acc": 0.4831804281345566,
      "train_speed(iter/s)": 1.438527
    },
    {
      "epoch": 2.517886979992288,
      "grad_norm": 6.006704807281494,
      "learning_rate": 4.943807578012114e-05,
      "loss": 2.3640342712402345,
      "memory(GiB)": 77.56,
      "step": 58770,
      "token_acc": 0.5176991150442478,
      "train_speed(iter/s)": 1.43852
    },
    {
      "epoch": 2.5181011953215373,
      "grad_norm": 5.237598419189453,
      "learning_rate": 4.943134643719528e-05,
      "loss": 2.5975671768188477,
      "memory(GiB)": 77.56,
      "step": 58775,
      "token_acc": 0.46048109965635736,
      "train_speed(iter/s)": 1.438477
    },
    {
      "epoch": 2.518315410650786,
      "grad_norm": 4.4248809814453125,
      "learning_rate": 4.94246171045711e-05,
      "loss": 1.9326534271240234,
      "memory(GiB)": 77.56,
      "step": 58780,
      "token_acc": 0.5758754863813229,
      "train_speed(iter/s)": 1.43848
    },
    {
      "epoch": 2.518529625980035,
      "grad_norm": 5.686248779296875,
      "learning_rate": 4.941788778237053e-05,
      "loss": 2.3820608139038084,
      "memory(GiB)": 77.56,
      "step": 58785,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.438497
    },
    {
      "epoch": 2.518743841309284,
      "grad_norm": 8.44089412689209,
      "learning_rate": 4.941115847071547e-05,
      "loss": 2.6586257934570314,
      "memory(GiB)": 77.56,
      "step": 58790,
      "token_acc": 0.43283582089552236,
      "train_speed(iter/s)": 1.438534
    },
    {
      "epoch": 2.518958056638533,
      "grad_norm": 6.452279567718506,
      "learning_rate": 4.9404429169727815e-05,
      "loss": 2.2759790420532227,
      "memory(GiB)": 77.56,
      "step": 58795,
      "token_acc": 0.5275590551181102,
      "train_speed(iter/s)": 1.438536
    },
    {
      "epoch": 2.519172271967782,
      "grad_norm": 6.500096321105957,
      "learning_rate": 4.9397699879529475e-05,
      "loss": 2.3621746063232423,
      "memory(GiB)": 77.56,
      "step": 58800,
      "token_acc": 0.5126582278481012,
      "train_speed(iter/s)": 1.438552
    },
    {
      "epoch": 2.519386487297031,
      "grad_norm": 4.290776252746582,
      "learning_rate": 4.939097060024239e-05,
      "loss": 2.379546546936035,
      "memory(GiB)": 77.56,
      "step": 58805,
      "token_acc": 0.4900662251655629,
      "train_speed(iter/s)": 1.438545
    },
    {
      "epoch": 2.51960070262628,
      "grad_norm": 4.893869400024414,
      "learning_rate": 4.9384241331988444e-05,
      "loss": 2.309023857116699,
      "memory(GiB)": 77.56,
      "step": 58810,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.438538
    },
    {
      "epoch": 2.5198149179555287,
      "grad_norm": 3.77557373046875,
      "learning_rate": 4.937751207488953e-05,
      "loss": 2.4117080688476564,
      "memory(GiB)": 77.56,
      "step": 58815,
      "token_acc": 0.5015197568389058,
      "train_speed(iter/s)": 1.438538
    },
    {
      "epoch": 2.520029133284778,
      "grad_norm": 6.054540634155273,
      "learning_rate": 4.937078282906758e-05,
      "loss": 2.5684844970703127,
      "memory(GiB)": 77.56,
      "step": 58820,
      "token_acc": 0.4652567975830816,
      "train_speed(iter/s)": 1.438527
    },
    {
      "epoch": 2.5202433486140268,
      "grad_norm": 5.331477165222168,
      "learning_rate": 4.936405359464448e-05,
      "loss": 2.0903430938720704,
      "memory(GiB)": 77.56,
      "step": 58825,
      "token_acc": 0.5514705882352942,
      "train_speed(iter/s)": 1.438546
    },
    {
      "epoch": 2.5204575639432756,
      "grad_norm": 4.416264057159424,
      "learning_rate": 4.9357324371742156e-05,
      "loss": 2.2944374084472656,
      "memory(GiB)": 77.56,
      "step": 58830,
      "token_acc": 0.4940119760479042,
      "train_speed(iter/s)": 1.438541
    },
    {
      "epoch": 2.520671779272525,
      "grad_norm": 6.2284417152404785,
      "learning_rate": 4.93505951604825e-05,
      "loss": 2.346688461303711,
      "memory(GiB)": 77.56,
      "step": 58835,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.438552
    },
    {
      "epoch": 2.5208859946017736,
      "grad_norm": 5.001190662384033,
      "learning_rate": 4.9343865960987425e-05,
      "loss": 2.892074775695801,
      "memory(GiB)": 77.56,
      "step": 58840,
      "token_acc": 0.428169014084507,
      "train_speed(iter/s)": 1.438568
    },
    {
      "epoch": 2.5211002099310225,
      "grad_norm": 5.005980014801025,
      "learning_rate": 4.9337136773378836e-05,
      "loss": 2.4520015716552734,
      "memory(GiB)": 77.56,
      "step": 58845,
      "token_acc": 0.5047318611987381,
      "train_speed(iter/s)": 1.438573
    },
    {
      "epoch": 2.5213144252602717,
      "grad_norm": 7.85048246383667,
      "learning_rate": 4.933040759777863e-05,
      "loss": 2.4986574172973635,
      "memory(GiB)": 77.56,
      "step": 58850,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 2.5215286405895205,
      "grad_norm": 5.1033034324646,
      "learning_rate": 4.9323678434308713e-05,
      "loss": 2.619414520263672,
      "memory(GiB)": 77.56,
      "step": 58855,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.438595
    },
    {
      "epoch": 2.5217428559187693,
      "grad_norm": 6.098699569702148,
      "learning_rate": 4.9316949283091004e-05,
      "loss": 1.9633358001708985,
      "memory(GiB)": 77.56,
      "step": 58860,
      "token_acc": 0.5625,
      "train_speed(iter/s)": 1.438613
    },
    {
      "epoch": 2.5219570712480186,
      "grad_norm": 7.399806022644043,
      "learning_rate": 4.9310220144247396e-05,
      "loss": 2.4731571197509767,
      "memory(GiB)": 77.56,
      "step": 58865,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 2.5221712865772674,
      "grad_norm": 4.403747081756592,
      "learning_rate": 4.930349101789978e-05,
      "loss": 2.3901674270629885,
      "memory(GiB)": 77.56,
      "step": 58870,
      "token_acc": 0.4948186528497409,
      "train_speed(iter/s)": 1.438562
    },
    {
      "epoch": 2.5223855019065162,
      "grad_norm": 5.965559959411621,
      "learning_rate": 4.9296761904170095e-05,
      "loss": 2.225765037536621,
      "memory(GiB)": 77.56,
      "step": 58875,
      "token_acc": 0.536,
      "train_speed(iter/s)": 1.438533
    },
    {
      "epoch": 2.5225997172357655,
      "grad_norm": 5.793490886688232,
      "learning_rate": 4.9290032803180234e-05,
      "loss": 2.4884992599487306,
      "memory(GiB)": 77.56,
      "step": 58880,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438551
    },
    {
      "epoch": 2.5228139325650143,
      "grad_norm": 5.5262908935546875,
      "learning_rate": 4.928330371505208e-05,
      "loss": 2.624837112426758,
      "memory(GiB)": 77.56,
      "step": 58885,
      "token_acc": 0.425531914893617,
      "train_speed(iter/s)": 1.438544
    },
    {
      "epoch": 2.523028147894263,
      "grad_norm": 4.926614761352539,
      "learning_rate": 4.927657463990757e-05,
      "loss": 2.2829132080078125,
      "memory(GiB)": 77.56,
      "step": 58890,
      "token_acc": 0.5207667731629393,
      "train_speed(iter/s)": 1.438508
    },
    {
      "epoch": 2.5232423632235124,
      "grad_norm": 8.304628372192383,
      "learning_rate": 4.926984557786859e-05,
      "loss": 2.3865482330322267,
      "memory(GiB)": 77.56,
      "step": 58895,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.438513
    },
    {
      "epoch": 2.523456578552761,
      "grad_norm": 6.951370716094971,
      "learning_rate": 4.926311652905702e-05,
      "loss": 2.285795211791992,
      "memory(GiB)": 77.56,
      "step": 58900,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438512
    },
    {
      "epoch": 2.52367079388201,
      "grad_norm": 4.3239569664001465,
      "learning_rate": 4.9256387493594805e-05,
      "loss": 2.4147138595581055,
      "memory(GiB)": 77.56,
      "step": 58905,
      "token_acc": 0.49226006191950467,
      "train_speed(iter/s)": 1.438478
    },
    {
      "epoch": 2.5238850092112592,
      "grad_norm": 7.33284854888916,
      "learning_rate": 4.924965847160381e-05,
      "loss": 2.5570903778076173,
      "memory(GiB)": 77.56,
      "step": 58910,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.438502
    },
    {
      "epoch": 2.524099224540508,
      "grad_norm": 6.5535888671875,
      "learning_rate": 4.924292946320597e-05,
      "loss": 2.5323068618774416,
      "memory(GiB)": 77.56,
      "step": 58915,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.438507
    },
    {
      "epoch": 2.524313439869757,
      "grad_norm": 6.4172186851501465,
      "learning_rate": 4.923620046852318e-05,
      "loss": 2.78980712890625,
      "memory(GiB)": 77.56,
      "step": 58920,
      "token_acc": 0.45695364238410596,
      "train_speed(iter/s)": 1.438526
    },
    {
      "epoch": 2.524527655199006,
      "grad_norm": 6.0758514404296875,
      "learning_rate": 4.9229471487677316e-05,
      "loss": 2.3035104751586912,
      "memory(GiB)": 77.56,
      "step": 58925,
      "token_acc": 0.49609375,
      "train_speed(iter/s)": 1.43853
    },
    {
      "epoch": 2.524741870528255,
      "grad_norm": 5.2978034019470215,
      "learning_rate": 4.9222742520790314e-05,
      "loss": 2.121658515930176,
      "memory(GiB)": 77.56,
      "step": 58930,
      "token_acc": 0.534965034965035,
      "train_speed(iter/s)": 1.438561
    },
    {
      "epoch": 2.5249560858575038,
      "grad_norm": 5.900257110595703,
      "learning_rate": 4.9216013567984055e-05,
      "loss": 2.3860401153564452,
      "memory(GiB)": 77.56,
      "step": 58935,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438577
    },
    {
      "epoch": 2.525170301186753,
      "grad_norm": 5.098506450653076,
      "learning_rate": 4.920928462938044e-05,
      "loss": 2.3442935943603516,
      "memory(GiB)": 77.56,
      "step": 58940,
      "token_acc": 0.5250836120401338,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 2.525384516516002,
      "grad_norm": 5.645188808441162,
      "learning_rate": 4.9202555705101355e-05,
      "loss": 2.7175121307373047,
      "memory(GiB)": 77.56,
      "step": 58945,
      "token_acc": 0.44193548387096776,
      "train_speed(iter/s)": 1.438612
    },
    {
      "epoch": 2.5255987318452506,
      "grad_norm": 3.989086151123047,
      "learning_rate": 4.919582679526875e-05,
      "loss": 2.1617315292358397,
      "memory(GiB)": 77.56,
      "step": 58950,
      "token_acc": 0.5372168284789643,
      "train_speed(iter/s)": 1.438601
    },
    {
      "epoch": 2.5258129471745,
      "grad_norm": 4.649693012237549,
      "learning_rate": 4.918909790000449e-05,
      "loss": 2.3297706604003907,
      "memory(GiB)": 77.56,
      "step": 58955,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.43861
    },
    {
      "epoch": 2.5260271625037487,
      "grad_norm": 4.8881964683532715,
      "learning_rate": 4.918236901943049e-05,
      "loss": 2.4629371643066404,
      "memory(GiB)": 77.56,
      "step": 58960,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.438599
    },
    {
      "epoch": 2.5262413778329975,
      "grad_norm": 4.827664375305176,
      "learning_rate": 4.917564015366864e-05,
      "loss": 2.3383283615112305,
      "memory(GiB)": 77.56,
      "step": 58965,
      "token_acc": 0.5504201680672269,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 2.5264555931622468,
      "grad_norm": 14.114139556884766,
      "learning_rate": 4.9168911302840834e-05,
      "loss": 2.650992774963379,
      "memory(GiB)": 77.56,
      "step": 58970,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.438603
    },
    {
      "epoch": 2.5266698084914956,
      "grad_norm": 5.02547025680542,
      "learning_rate": 4.916218246706899e-05,
      "loss": 2.34653377532959,
      "memory(GiB)": 77.56,
      "step": 58975,
      "token_acc": 0.49777777777777776,
      "train_speed(iter/s)": 1.43863
    },
    {
      "epoch": 2.5268840238207444,
      "grad_norm": 6.5236735343933105,
      "learning_rate": 4.9155453646475e-05,
      "loss": 2.6372867584228517,
      "memory(GiB)": 77.56,
      "step": 58980,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 2.5270982391499937,
      "grad_norm": 6.692519187927246,
      "learning_rate": 4.914872484118074e-05,
      "loss": 2.363043212890625,
      "memory(GiB)": 77.56,
      "step": 58985,
      "token_acc": 0.49609375,
      "train_speed(iter/s)": 1.438651
    },
    {
      "epoch": 2.5273124544792425,
      "grad_norm": 7.264876365661621,
      "learning_rate": 4.914199605130814e-05,
      "loss": 2.435413360595703,
      "memory(GiB)": 77.56,
      "step": 58990,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.438655
    },
    {
      "epoch": 2.5275266698084913,
      "grad_norm": 5.1589274406433105,
      "learning_rate": 4.913526727697909e-05,
      "loss": 2.659071159362793,
      "memory(GiB)": 77.56,
      "step": 58995,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.438655
    },
    {
      "epoch": 2.5277408851377405,
      "grad_norm": 5.5924859046936035,
      "learning_rate": 4.912853851831547e-05,
      "loss": 2.331832504272461,
      "memory(GiB)": 77.56,
      "step": 59000,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.5277408851377405,
      "eval_loss": 2.3112387657165527,
      "eval_runtime": 14.4884,
      "eval_samples_per_second": 6.902,
      "eval_steps_per_second": 6.902,
      "eval_token_acc": 0.4672657252888318,
      "step": 59000
    },
    {
      "epoch": 2.5279551004669893,
      "grad_norm": 5.729962348937988,
      "learning_rate": 4.9121809775439214e-05,
      "loss": 2.224778938293457,
      "memory(GiB)": 77.56,
      "step": 59005,
      "token_acc": 0.46866485013623976,
      "train_speed(iter/s)": 1.438039
    },
    {
      "epoch": 2.528169315796238,
      "grad_norm": 6.025215148925781,
      "learning_rate": 4.911508104847218e-05,
      "loss": 2.5950698852539062,
      "memory(GiB)": 77.56,
      "step": 59010,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.438044
    },
    {
      "epoch": 2.5283835311254874,
      "grad_norm": 6.581197261810303,
      "learning_rate": 4.910835233753629e-05,
      "loss": 2.428325653076172,
      "memory(GiB)": 77.56,
      "step": 59015,
      "token_acc": 0.4797507788161994,
      "train_speed(iter/s)": 1.438058
    },
    {
      "epoch": 2.5285977464547362,
      "grad_norm": 4.956717491149902,
      "learning_rate": 4.9101623642753416e-05,
      "loss": 2.385212707519531,
      "memory(GiB)": 77.56,
      "step": 59020,
      "token_acc": 0.5139318885448917,
      "train_speed(iter/s)": 1.438095
    },
    {
      "epoch": 2.528811961783985,
      "grad_norm": 6.194469451904297,
      "learning_rate": 4.909489496424549e-05,
      "loss": 2.761299896240234,
      "memory(GiB)": 77.56,
      "step": 59025,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 2.5290261771132343,
      "grad_norm": 6.688918113708496,
      "learning_rate": 4.9088166302134385e-05,
      "loss": 2.44656925201416,
      "memory(GiB)": 77.56,
      "step": 59030,
      "token_acc": 0.423728813559322,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 2.529240392442483,
      "grad_norm": 9.883824348449707,
      "learning_rate": 4.908143765654201e-05,
      "loss": 2.012375259399414,
      "memory(GiB)": 77.56,
      "step": 59035,
      "token_acc": 0.5720524017467249,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 2.529454607771732,
      "grad_norm": 5.347978591918945,
      "learning_rate": 4.907470902759026e-05,
      "loss": 2.5401554107666016,
      "memory(GiB)": 77.56,
      "step": 59040,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 2.529668823100981,
      "grad_norm": 5.202302932739258,
      "learning_rate": 4.906798041540101e-05,
      "loss": 2.3805755615234374,
      "memory(GiB)": 77.56,
      "step": 59045,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.438108
    },
    {
      "epoch": 2.52988303843023,
      "grad_norm": 4.343328952789307,
      "learning_rate": 4.906125182009618e-05,
      "loss": 2.103287696838379,
      "memory(GiB)": 77.56,
      "step": 59050,
      "token_acc": 0.5631399317406144,
      "train_speed(iter/s)": 1.438085
    },
    {
      "epoch": 2.530097253759479,
      "grad_norm": 4.7397613525390625,
      "learning_rate": 4.905452324179764e-05,
      "loss": 2.289626693725586,
      "memory(GiB)": 77.56,
      "step": 59055,
      "token_acc": 0.5034965034965035,
      "train_speed(iter/s)": 1.438104
    },
    {
      "epoch": 2.530311469088728,
      "grad_norm": 5.467718601226807,
      "learning_rate": 4.904779468062731e-05,
      "loss": 2.3340198516845705,
      "memory(GiB)": 77.56,
      "step": 59060,
      "token_acc": 0.47232472324723246,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 2.530525684417977,
      "grad_norm": 5.405391693115234,
      "learning_rate": 4.904106613670707e-05,
      "loss": 2.423992156982422,
      "memory(GiB)": 77.56,
      "step": 59065,
      "token_acc": 0.47468354430379744,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 2.5307398997472257,
      "grad_norm": 6.097036838531494,
      "learning_rate": 4.9034337610158815e-05,
      "loss": 2.3963991165161134,
      "memory(GiB)": 77.56,
      "step": 59070,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.438093
    },
    {
      "epoch": 2.530954115076475,
      "grad_norm": 4.077427387237549,
      "learning_rate": 4.9027609101104444e-05,
      "loss": 2.5260807037353517,
      "memory(GiB)": 77.56,
      "step": 59075,
      "token_acc": 0.45564516129032256,
      "train_speed(iter/s)": 1.43813
    },
    {
      "epoch": 2.5311683304057238,
      "grad_norm": 5.249317169189453,
      "learning_rate": 4.902088060966585e-05,
      "loss": 2.912622833251953,
      "memory(GiB)": 77.56,
      "step": 59080,
      "token_acc": 0.45348837209302323,
      "train_speed(iter/s)": 1.438163
    },
    {
      "epoch": 2.5313825457349726,
      "grad_norm": 4.795521259307861,
      "learning_rate": 4.90141521359649e-05,
      "loss": 2.4073673248291017,
      "memory(GiB)": 77.56,
      "step": 59085,
      "token_acc": 0.49166666666666664,
      "train_speed(iter/s)": 1.438162
    },
    {
      "epoch": 2.531596761064222,
      "grad_norm": 4.793128490447998,
      "learning_rate": 4.900742368012353e-05,
      "loss": 2.4595111846923827,
      "memory(GiB)": 77.56,
      "step": 59090,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.43817
    },
    {
      "epoch": 2.5318109763934706,
      "grad_norm": 4.526932239532471,
      "learning_rate": 4.9000695242263586e-05,
      "loss": 2.0980030059814454,
      "memory(GiB)": 77.56,
      "step": 59095,
      "token_acc": 0.5302013422818792,
      "train_speed(iter/s)": 1.438188
    },
    {
      "epoch": 2.5320251917227194,
      "grad_norm": 7.1503987312316895,
      "learning_rate": 4.8993966822507006e-05,
      "loss": 2.7005409240722655,
      "memory(GiB)": 77.56,
      "step": 59100,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.438204
    },
    {
      "epoch": 2.5322394070519687,
      "grad_norm": 4.8963470458984375,
      "learning_rate": 4.898723842097566e-05,
      "loss": 2.420587921142578,
      "memory(GiB)": 77.56,
      "step": 59105,
      "token_acc": 0.4954954954954955,
      "train_speed(iter/s)": 1.438216
    },
    {
      "epoch": 2.5324536223812175,
      "grad_norm": 7.978202819824219,
      "learning_rate": 4.898051003779144e-05,
      "loss": 2.306772994995117,
      "memory(GiB)": 77.56,
      "step": 59110,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.438249
    },
    {
      "epoch": 2.5326678377104663,
      "grad_norm": 8.484978675842285,
      "learning_rate": 4.8973781673076236e-05,
      "loss": 2.2961393356323243,
      "memory(GiB)": 77.56,
      "step": 59115,
      "token_acc": 0.44569288389513106,
      "train_speed(iter/s)": 1.438239
    },
    {
      "epoch": 2.5328820530397156,
      "grad_norm": 5.341443061828613,
      "learning_rate": 4.896705332695194e-05,
      "loss": 2.437510681152344,
      "memory(GiB)": 77.56,
      "step": 59120,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.438242
    },
    {
      "epoch": 2.5330962683689644,
      "grad_norm": 5.369494915008545,
      "learning_rate": 4.896032499954045e-05,
      "loss": 2.5108789443969726,
      "memory(GiB)": 77.56,
      "step": 59125,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 2.533310483698213,
      "grad_norm": 5.534130096435547,
      "learning_rate": 4.895359669096363e-05,
      "loss": 2.6540061950683596,
      "memory(GiB)": 77.56,
      "step": 59130,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.438274
    },
    {
      "epoch": 2.5335246990274625,
      "grad_norm": 6.556541442871094,
      "learning_rate": 4.8946868401343404e-05,
      "loss": 2.3918256759643555,
      "memory(GiB)": 77.56,
      "step": 59135,
      "token_acc": 0.49814126394052044,
      "train_speed(iter/s)": 1.438284
    },
    {
      "epoch": 2.5337389143567113,
      "grad_norm": 5.085676670074463,
      "learning_rate": 4.894014013080163e-05,
      "loss": 2.6551151275634766,
      "memory(GiB)": 77.56,
      "step": 59140,
      "token_acc": 0.44871794871794873,
      "train_speed(iter/s)": 1.438302
    },
    {
      "epoch": 2.53395312968596,
      "grad_norm": 6.968482971191406,
      "learning_rate": 4.893341187946022e-05,
      "loss": 2.5774843215942385,
      "memory(GiB)": 77.56,
      "step": 59145,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.438303
    },
    {
      "epoch": 2.5341673450152093,
      "grad_norm": 7.608434200286865,
      "learning_rate": 4.8926683647441065e-05,
      "loss": 2.232361602783203,
      "memory(GiB)": 77.56,
      "step": 59150,
      "token_acc": 0.4758364312267658,
      "train_speed(iter/s)": 1.438313
    },
    {
      "epoch": 2.534381560344458,
      "grad_norm": 11.061235427856445,
      "learning_rate": 4.891995543486602e-05,
      "loss": 2.3544471740722654,
      "memory(GiB)": 77.56,
      "step": 59155,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.438346
    },
    {
      "epoch": 2.534595775673707,
      "grad_norm": 5.840290546417236,
      "learning_rate": 4.8913227241857016e-05,
      "loss": 2.585592269897461,
      "memory(GiB)": 77.56,
      "step": 59160,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.438366
    },
    {
      "epoch": 2.5348099910029562,
      "grad_norm": 5.152430534362793,
      "learning_rate": 4.8906499068535917e-05,
      "loss": 2.5015968322753905,
      "memory(GiB)": 77.56,
      "step": 59165,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.438348
    },
    {
      "epoch": 2.535024206332205,
      "grad_norm": 5.306188583374023,
      "learning_rate": 4.889977091502459e-05,
      "loss": 2.322525978088379,
      "memory(GiB)": 77.56,
      "step": 59170,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.438367
    },
    {
      "epoch": 2.535238421661454,
      "grad_norm": 5.255798816680908,
      "learning_rate": 4.889304278144495e-05,
      "loss": 2.4219472885131834,
      "memory(GiB)": 77.56,
      "step": 59175,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.438362
    },
    {
      "epoch": 2.535452636990703,
      "grad_norm": 5.514500141143799,
      "learning_rate": 4.88863146679189e-05,
      "loss": 2.479305648803711,
      "memory(GiB)": 77.56,
      "step": 59180,
      "token_acc": 0.49538461538461537,
      "train_speed(iter/s)": 1.438359
    },
    {
      "epoch": 2.535666852319952,
      "grad_norm": 5.697058200836182,
      "learning_rate": 4.88795865745683e-05,
      "loss": 2.5944629669189454,
      "memory(GiB)": 77.56,
      "step": 59185,
      "token_acc": 0.4579710144927536,
      "train_speed(iter/s)": 1.438359
    },
    {
      "epoch": 2.5358810676492007,
      "grad_norm": 5.072600841522217,
      "learning_rate": 4.887285850151503e-05,
      "loss": 2.4342037200927735,
      "memory(GiB)": 77.56,
      "step": 59190,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.438371
    },
    {
      "epoch": 2.53609528297845,
      "grad_norm": 5.58681583404541,
      "learning_rate": 4.8866130448880995e-05,
      "loss": 2.531167411804199,
      "memory(GiB)": 77.56,
      "step": 59195,
      "token_acc": 0.44625407166123776,
      "train_speed(iter/s)": 1.438391
    },
    {
      "epoch": 2.536309498307699,
      "grad_norm": 5.864629745483398,
      "learning_rate": 4.885940241678806e-05,
      "loss": 2.3252824783325194,
      "memory(GiB)": 77.56,
      "step": 59200,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.438387
    },
    {
      "epoch": 2.5365237136369476,
      "grad_norm": 4.608503341674805,
      "learning_rate": 4.885267440535813e-05,
      "loss": 2.4484243392944336,
      "memory(GiB)": 77.56,
      "step": 59205,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.438406
    },
    {
      "epoch": 2.536737928966197,
      "grad_norm": 6.551125526428223,
      "learning_rate": 4.8845946414713076e-05,
      "loss": 2.5869518280029298,
      "memory(GiB)": 77.56,
      "step": 59210,
      "token_acc": 0.4370860927152318,
      "train_speed(iter/s)": 1.438374
    },
    {
      "epoch": 2.5369521442954457,
      "grad_norm": 7.950067043304443,
      "learning_rate": 4.883921844497478e-05,
      "loss": 2.903058624267578,
      "memory(GiB)": 77.56,
      "step": 59215,
      "token_acc": 0.4393939393939394,
      "train_speed(iter/s)": 1.438382
    },
    {
      "epoch": 2.5371663596246945,
      "grad_norm": 4.788400173187256,
      "learning_rate": 4.883249049626514e-05,
      "loss": 2.566099166870117,
      "memory(GiB)": 77.56,
      "step": 59220,
      "token_acc": 0.43508771929824563,
      "train_speed(iter/s)": 1.438411
    },
    {
      "epoch": 2.5373805749539438,
      "grad_norm": 5.352656841278076,
      "learning_rate": 4.882576256870604e-05,
      "loss": 2.491599464416504,
      "memory(GiB)": 77.56,
      "step": 59225,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.438414
    },
    {
      "epoch": 2.5375947902831926,
      "grad_norm": 4.482700824737549,
      "learning_rate": 4.8819034662419326e-05,
      "loss": 2.4838624954223634,
      "memory(GiB)": 77.56,
      "step": 59230,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.438432
    },
    {
      "epoch": 2.5378090056124414,
      "grad_norm": 5.484973907470703,
      "learning_rate": 4.881230677752693e-05,
      "loss": 2.3808292388916015,
      "memory(GiB)": 77.56,
      "step": 59235,
      "token_acc": 0.50390625,
      "train_speed(iter/s)": 1.438449
    },
    {
      "epoch": 2.5380232209416906,
      "grad_norm": 4.7214274406433105,
      "learning_rate": 4.880557891415068e-05,
      "loss": 2.483547019958496,
      "memory(GiB)": 77.56,
      "step": 59240,
      "token_acc": 0.47305389221556887,
      "train_speed(iter/s)": 1.438464
    },
    {
      "epoch": 2.5382374362709395,
      "grad_norm": 5.173149585723877,
      "learning_rate": 4.87988510724125e-05,
      "loss": 2.4541904449462892,
      "memory(GiB)": 77.56,
      "step": 59245,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.438492
    },
    {
      "epoch": 2.5384516516001883,
      "grad_norm": 4.171067237854004,
      "learning_rate": 4.8792123252434276e-05,
      "loss": 2.2743024826049805,
      "memory(GiB)": 77.56,
      "step": 59250,
      "token_acc": 0.4983922829581994,
      "train_speed(iter/s)": 1.438462
    },
    {
      "epoch": 2.5386658669294375,
      "grad_norm": 5.42086124420166,
      "learning_rate": 4.878539545433787e-05,
      "loss": 2.404249572753906,
      "memory(GiB)": 77.56,
      "step": 59255,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.438469
    },
    {
      "epoch": 2.5388800822586863,
      "grad_norm": 4.652543544769287,
      "learning_rate": 4.877866767824515e-05,
      "loss": 2.578704833984375,
      "memory(GiB)": 77.56,
      "step": 59260,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.438484
    },
    {
      "epoch": 2.539094297587935,
      "grad_norm": 5.320056915283203,
      "learning_rate": 4.877193992427803e-05,
      "loss": 2.273621940612793,
      "memory(GiB)": 77.56,
      "step": 59265,
      "token_acc": 0.5441176470588235,
      "train_speed(iter/s)": 1.438477
    },
    {
      "epoch": 2.5393085129171844,
      "grad_norm": 6.450484275817871,
      "learning_rate": 4.876521219255836e-05,
      "loss": 2.215576171875,
      "memory(GiB)": 77.56,
      "step": 59270,
      "token_acc": 0.5265017667844523,
      "train_speed(iter/s)": 1.438473
    },
    {
      "epoch": 2.539522728246433,
      "grad_norm": 6.637783527374268,
      "learning_rate": 4.8758484483208024e-05,
      "loss": 2.595779609680176,
      "memory(GiB)": 77.56,
      "step": 59275,
      "token_acc": 0.4448979591836735,
      "train_speed(iter/s)": 1.438445
    },
    {
      "epoch": 2.539736943575682,
      "grad_norm": 6.064037799835205,
      "learning_rate": 4.875175679634892e-05,
      "loss": 2.6027334213256834,
      "memory(GiB)": 77.56,
      "step": 59280,
      "token_acc": 0.5095057034220533,
      "train_speed(iter/s)": 1.438457
    },
    {
      "epoch": 2.5399511589049313,
      "grad_norm": 5.9925665855407715,
      "learning_rate": 4.87450291321029e-05,
      "loss": 2.3716007232666017,
      "memory(GiB)": 77.56,
      "step": 59285,
      "token_acc": 0.5313807531380753,
      "train_speed(iter/s)": 1.438492
    },
    {
      "epoch": 2.54016537423418,
      "grad_norm": 4.940703868865967,
      "learning_rate": 4.873830149059186e-05,
      "loss": 2.00594539642334,
      "memory(GiB)": 77.56,
      "step": 59290,
      "token_acc": 0.5764705882352941,
      "train_speed(iter/s)": 1.438499
    },
    {
      "epoch": 2.540379589563429,
      "grad_norm": 5.072874546051025,
      "learning_rate": 4.873157387193767e-05,
      "loss": 2.4332725524902346,
      "memory(GiB)": 77.56,
      "step": 59295,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.438513
    },
    {
      "epoch": 2.540593804892678,
      "grad_norm": 4.554154872894287,
      "learning_rate": 4.872484627626221e-05,
      "loss": 2.604878234863281,
      "memory(GiB)": 77.56,
      "step": 59300,
      "token_acc": 0.4797507788161994,
      "train_speed(iter/s)": 1.438504
    },
    {
      "epoch": 2.540808020221927,
      "grad_norm": 4.922318935394287,
      "learning_rate": 4.871811870368736e-05,
      "loss": 2.366122817993164,
      "memory(GiB)": 77.56,
      "step": 59305,
      "token_acc": 0.5231788079470199,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 2.541022235551176,
      "grad_norm": 5.363475322723389,
      "learning_rate": 4.8711391154335e-05,
      "loss": 2.1453617095947264,
      "memory(GiB)": 77.56,
      "step": 59310,
      "token_acc": 0.5170940170940171,
      "train_speed(iter/s)": 1.438537
    },
    {
      "epoch": 2.541236450880425,
      "grad_norm": 4.4472222328186035,
      "learning_rate": 4.870466362832696e-05,
      "loss": 2.1647071838378906,
      "memory(GiB)": 77.56,
      "step": 59315,
      "token_acc": 0.5348837209302325,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.541450666209674,
      "grad_norm": 4.510188102722168,
      "learning_rate": 4.869793612578518e-05,
      "loss": 2.268606948852539,
      "memory(GiB)": 77.56,
      "step": 59320,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.43856
    },
    {
      "epoch": 2.5416648815389227,
      "grad_norm": 5.476666450500488,
      "learning_rate": 4.869120864683151e-05,
      "loss": 2.5754816055297853,
      "memory(GiB)": 77.56,
      "step": 59325,
      "token_acc": 0.5103448275862069,
      "train_speed(iter/s)": 1.438536
    },
    {
      "epoch": 2.541879096868172,
      "grad_norm": 5.3883442878723145,
      "learning_rate": 4.868448119158783e-05,
      "loss": 2.504313087463379,
      "memory(GiB)": 77.56,
      "step": 59330,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.438537
    },
    {
      "epoch": 2.5420933121974207,
      "grad_norm": 5.293647289276123,
      "learning_rate": 4.8677753760175995e-05,
      "loss": 2.352847862243652,
      "memory(GiB)": 77.56,
      "step": 59335,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438554
    },
    {
      "epoch": 2.5423075275266696,
      "grad_norm": 7.011762619018555,
      "learning_rate": 4.8671026352717895e-05,
      "loss": 2.2629510879516603,
      "memory(GiB)": 77.56,
      "step": 59340,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.438562
    },
    {
      "epoch": 2.542521742855919,
      "grad_norm": 5.031600475311279,
      "learning_rate": 4.866429896933539e-05,
      "loss": 2.171701431274414,
      "memory(GiB)": 77.56,
      "step": 59345,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.43855
    },
    {
      "epoch": 2.5427359581851676,
      "grad_norm": 6.941433429718018,
      "learning_rate": 4.865757161015038e-05,
      "loss": 2.555876541137695,
      "memory(GiB)": 77.56,
      "step": 59350,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.438547
    },
    {
      "epoch": 2.5429501735144164,
      "grad_norm": 5.078488349914551,
      "learning_rate": 4.865084427528471e-05,
      "loss": 2.546540451049805,
      "memory(GiB)": 77.56,
      "step": 59355,
      "token_acc": 0.4609756097560976,
      "train_speed(iter/s)": 1.438548
    },
    {
      "epoch": 2.5431643888436657,
      "grad_norm": 9.364160537719727,
      "learning_rate": 4.864411696486026e-05,
      "loss": 2.586537742614746,
      "memory(GiB)": 77.56,
      "step": 59360,
      "token_acc": 0.5104602510460251,
      "train_speed(iter/s)": 1.438535
    },
    {
      "epoch": 2.5433786041729145,
      "grad_norm": 5.819259166717529,
      "learning_rate": 4.863738967899891e-05,
      "loss": 2.5130224227905273,
      "memory(GiB)": 77.56,
      "step": 59365,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.438569
    },
    {
      "epoch": 2.5435928195021633,
      "grad_norm": 5.531962871551514,
      "learning_rate": 4.8630662417822526e-05,
      "loss": 2.374894905090332,
      "memory(GiB)": 77.56,
      "step": 59370,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.43857
    },
    {
      "epoch": 2.5438070348314126,
      "grad_norm": 5.3213911056518555,
      "learning_rate": 4.8623935181452966e-05,
      "loss": 2.6349231719970705,
      "memory(GiB)": 77.56,
      "step": 59375,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.438568
    },
    {
      "epoch": 2.5440212501606614,
      "grad_norm": 5.56639289855957,
      "learning_rate": 4.861720797001212e-05,
      "loss": 2.4288457870483398,
      "memory(GiB)": 77.56,
      "step": 59380,
      "token_acc": 0.4471299093655589,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 2.54423546548991,
      "grad_norm": 4.515190124511719,
      "learning_rate": 4.8610480783621835e-05,
      "loss": 2.5948448181152344,
      "memory(GiB)": 77.56,
      "step": 59385,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.43855
    },
    {
      "epoch": 2.5444496808191595,
      "grad_norm": 6.551082611083984,
      "learning_rate": 4.860375362240399e-05,
      "loss": 2.078593063354492,
      "memory(GiB)": 77.56,
      "step": 59390,
      "token_acc": 0.5461254612546126,
      "train_speed(iter/s)": 1.43857
    },
    {
      "epoch": 2.5446638961484083,
      "grad_norm": 5.5749897956848145,
      "learning_rate": 4.859702648648047e-05,
      "loss": 2.096351432800293,
      "memory(GiB)": 77.56,
      "step": 59395,
      "token_acc": 0.5291828793774319,
      "train_speed(iter/s)": 1.438577
    },
    {
      "epoch": 2.544878111477657,
      "grad_norm": 6.020618915557861,
      "learning_rate": 4.859029937597314e-05,
      "loss": 2.20284423828125,
      "memory(GiB)": 77.56,
      "step": 59400,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.438566
    },
    {
      "epoch": 2.5450923268069063,
      "grad_norm": 5.776437282562256,
      "learning_rate": 4.858357229100385e-05,
      "loss": 2.394856643676758,
      "memory(GiB)": 77.56,
      "step": 59405,
      "token_acc": 0.46629213483146065,
      "train_speed(iter/s)": 1.438596
    },
    {
      "epoch": 2.545306542136155,
      "grad_norm": 6.168442726135254,
      "learning_rate": 4.857684523169449e-05,
      "loss": 2.3220544815063477,
      "memory(GiB)": 77.56,
      "step": 59410,
      "token_acc": 0.47876447876447875,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 2.545520757465404,
      "grad_norm": 4.487974643707275,
      "learning_rate": 4.8570118198166896e-05,
      "loss": 2.2447771072387694,
      "memory(GiB)": 77.56,
      "step": 59415,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.438602
    },
    {
      "epoch": 2.545734972794653,
      "grad_norm": 5.289567470550537,
      "learning_rate": 4.8563391190542954e-05,
      "loss": 2.6859954833984374,
      "memory(GiB)": 77.56,
      "step": 59420,
      "token_acc": 0.49538461538461537,
      "train_speed(iter/s)": 1.43861
    },
    {
      "epoch": 2.545949188123902,
      "grad_norm": 5.069399356842041,
      "learning_rate": 4.855666420894454e-05,
      "loss": 2.771007537841797,
      "memory(GiB)": 77.56,
      "step": 59425,
      "token_acc": 0.4295774647887324,
      "train_speed(iter/s)": 1.438642
    },
    {
      "epoch": 2.546163403453151,
      "grad_norm": 6.1787428855896,
      "learning_rate": 4.85499372534935e-05,
      "loss": 2.628916549682617,
      "memory(GiB)": 77.56,
      "step": 59430,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 2.5463776187824,
      "grad_norm": 5.285228252410889,
      "learning_rate": 4.8543210324311704e-05,
      "loss": 2.485353469848633,
      "memory(GiB)": 77.56,
      "step": 59435,
      "token_acc": 0.45664739884393063,
      "train_speed(iter/s)": 1.438637
    },
    {
      "epoch": 2.546591834111649,
      "grad_norm": 7.288173198699951,
      "learning_rate": 4.8536483421521025e-05,
      "loss": 2.538277435302734,
      "memory(GiB)": 77.56,
      "step": 59440,
      "token_acc": 0.46449704142011833,
      "train_speed(iter/s)": 1.438629
    },
    {
      "epoch": 2.5468060494408977,
      "grad_norm": 5.360102653503418,
      "learning_rate": 4.852975654524332e-05,
      "loss": 2.5608760833740236,
      "memory(GiB)": 77.56,
      "step": 59445,
      "token_acc": 0.4507042253521127,
      "train_speed(iter/s)": 1.438636
    },
    {
      "epoch": 2.547020264770147,
      "grad_norm": 4.4294233322143555,
      "learning_rate": 4.852302969560046e-05,
      "loss": 2.1616743087768553,
      "memory(GiB)": 77.56,
      "step": 59450,
      "token_acc": 0.54296875,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 2.547234480099396,
      "grad_norm": 5.301448345184326,
      "learning_rate": 4.8516302872714295e-05,
      "loss": 2.5592079162597656,
      "memory(GiB)": 77.56,
      "step": 59455,
      "token_acc": 0.49185667752442996,
      "train_speed(iter/s)": 1.438679
    },
    {
      "epoch": 2.5474486954286446,
      "grad_norm": 6.969396114349365,
      "learning_rate": 4.8509576076706695e-05,
      "loss": 2.2596336364746095,
      "memory(GiB)": 77.56,
      "step": 59460,
      "token_acc": 0.509090909090909,
      "train_speed(iter/s)": 1.43867
    },
    {
      "epoch": 2.547662910757894,
      "grad_norm": 5.3324666023254395,
      "learning_rate": 4.8502849307699504e-05,
      "loss": 2.3390174865722657,
      "memory(GiB)": 77.56,
      "step": 59465,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 2.5478771260871427,
      "grad_norm": 4.63497257232666,
      "learning_rate": 4.849612256581463e-05,
      "loss": 2.6836269378662108,
      "memory(GiB)": 77.56,
      "step": 59470,
      "token_acc": 0.4503311258278146,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 2.5480913414163915,
      "grad_norm": 5.863549709320068,
      "learning_rate": 4.8489395851173905e-05,
      "loss": 2.3295476913452147,
      "memory(GiB)": 77.56,
      "step": 59475,
      "token_acc": 0.45724907063197023,
      "train_speed(iter/s)": 1.438666
    },
    {
      "epoch": 2.5483055567456407,
      "grad_norm": 9.32987117767334,
      "learning_rate": 4.848266916389918e-05,
      "loss": 2.429655075073242,
      "memory(GiB)": 77.56,
      "step": 59480,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.438659
    },
    {
      "epoch": 2.5485197720748896,
      "grad_norm": 4.570631504058838,
      "learning_rate": 4.847594250411234e-05,
      "loss": 2.41324405670166,
      "memory(GiB)": 77.56,
      "step": 59485,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.4387
    },
    {
      "epoch": 2.5487339874041384,
      "grad_norm": 5.507359981536865,
      "learning_rate": 4.8469215871935216e-05,
      "loss": 2.5094242095947266,
      "memory(GiB)": 77.56,
      "step": 59490,
      "token_acc": 0.4819672131147541,
      "train_speed(iter/s)": 1.438703
    },
    {
      "epoch": 2.5489482027333876,
      "grad_norm": 4.875783443450928,
      "learning_rate": 4.846248926748969e-05,
      "loss": 2.235984039306641,
      "memory(GiB)": 77.56,
      "step": 59495,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.438641
    },
    {
      "epoch": 2.5491624180626364,
      "grad_norm": 5.424518585205078,
      "learning_rate": 4.845576269089762e-05,
      "loss": 2.3595706939697267,
      "memory(GiB)": 77.56,
      "step": 59500,
      "token_acc": 0.5100864553314121,
      "train_speed(iter/s)": 1.438651
    },
    {
      "epoch": 2.5491624180626364,
      "eval_loss": 2.287466049194336,
      "eval_runtime": 14.7182,
      "eval_samples_per_second": 6.794,
      "eval_steps_per_second": 6.794,
      "eval_token_acc": 0.44106463878326996,
      "step": 59500
    },
    {
      "epoch": 2.5493766333918852,
      "grad_norm": 5.486857891082764,
      "learning_rate": 4.844903614228084e-05,
      "loss": 2.2575931549072266,
      "memory(GiB)": 77.56,
      "step": 59505,
      "token_acc": 0.4548022598870056,
      "train_speed(iter/s)": 1.43809
    },
    {
      "epoch": 2.5495908487211345,
      "grad_norm": 4.657067775726318,
      "learning_rate": 4.844230962176124e-05,
      "loss": 2.5744064331054686,
      "memory(GiB)": 77.56,
      "step": 59510,
      "token_acc": 0.4438040345821326,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 2.5498050640503833,
      "grad_norm": 5.722229957580566,
      "learning_rate": 4.8435583129460666e-05,
      "loss": 2.1946325302124023,
      "memory(GiB)": 77.56,
      "step": 59515,
      "token_acc": 0.5063291139240507,
      "train_speed(iter/s)": 1.438123
    },
    {
      "epoch": 2.550019279379632,
      "grad_norm": 6.230782985687256,
      "learning_rate": 4.842885666550095e-05,
      "loss": 2.3860937118530274,
      "memory(GiB)": 77.56,
      "step": 59520,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.43813
    },
    {
      "epoch": 2.5502334947088814,
      "grad_norm": 6.758782863616943,
      "learning_rate": 4.842213023000399e-05,
      "loss": 2.551443862915039,
      "memory(GiB)": 77.56,
      "step": 59525,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 2.55044771003813,
      "grad_norm": 5.023636341094971,
      "learning_rate": 4.841540382309161e-05,
      "loss": 2.397336959838867,
      "memory(GiB)": 77.56,
      "step": 59530,
      "token_acc": 0.48120300751879697,
      "train_speed(iter/s)": 1.438151
    },
    {
      "epoch": 2.550661925367379,
      "grad_norm": 7.535745143890381,
      "learning_rate": 4.8408677444885685e-05,
      "loss": 2.4807538986206055,
      "memory(GiB)": 77.56,
      "step": 59535,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.43817
    },
    {
      "epoch": 2.5508761406966283,
      "grad_norm": 6.641665458679199,
      "learning_rate": 4.840195109550804e-05,
      "loss": 2.4462207794189452,
      "memory(GiB)": 77.56,
      "step": 59540,
      "token_acc": 0.46037735849056605,
      "train_speed(iter/s)": 1.438168
    },
    {
      "epoch": 2.551090356025877,
      "grad_norm": 5.368220329284668,
      "learning_rate": 4.8395224775080574e-05,
      "loss": 2.412137031555176,
      "memory(GiB)": 77.56,
      "step": 59545,
      "token_acc": 0.4708029197080292,
      "train_speed(iter/s)": 1.438162
    },
    {
      "epoch": 2.5513045713551263,
      "grad_norm": 5.6271281242370605,
      "learning_rate": 4.83884984837251e-05,
      "loss": 2.4368988037109376,
      "memory(GiB)": 77.56,
      "step": 59550,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.438156
    },
    {
      "epoch": 2.551518786684375,
      "grad_norm": 5.452853202819824,
      "learning_rate": 4.83817722215635e-05,
      "loss": 2.67132511138916,
      "memory(GiB)": 77.56,
      "step": 59555,
      "token_acc": 0.47093023255813954,
      "train_speed(iter/s)": 1.438173
    },
    {
      "epoch": 2.551733002013624,
      "grad_norm": 5.387850761413574,
      "learning_rate": 4.837504598871762e-05,
      "loss": 2.4648250579833983,
      "memory(GiB)": 77.56,
      "step": 59560,
      "token_acc": 0.4482758620689655,
      "train_speed(iter/s)": 1.438154
    },
    {
      "epoch": 2.551947217342873,
      "grad_norm": 5.006322860717773,
      "learning_rate": 4.8368319785309285e-05,
      "loss": 2.3904516220092775,
      "memory(GiB)": 77.56,
      "step": 59565,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438188
    },
    {
      "epoch": 2.552161432672122,
      "grad_norm": 6.988188743591309,
      "learning_rate": 4.836159361146038e-05,
      "loss": 2.312815475463867,
      "memory(GiB)": 77.56,
      "step": 59570,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.438205
    },
    {
      "epoch": 2.552375648001371,
      "grad_norm": 5.220654010772705,
      "learning_rate": 4.835486746729274e-05,
      "loss": 2.6267271041870117,
      "memory(GiB)": 77.56,
      "step": 59575,
      "token_acc": 0.4221556886227545,
      "train_speed(iter/s)": 1.438235
    },
    {
      "epoch": 2.55258986333062,
      "grad_norm": 4.633787631988525,
      "learning_rate": 4.834814135292822e-05,
      "loss": 2.1964237213134767,
      "memory(GiB)": 77.56,
      "step": 59580,
      "token_acc": 0.5419354838709678,
      "train_speed(iter/s)": 1.438249
    },
    {
      "epoch": 2.552804078659869,
      "grad_norm": 6.835556507110596,
      "learning_rate": 4.834141526848868e-05,
      "loss": 2.3077939987182616,
      "memory(GiB)": 77.56,
      "step": 59585,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.438274
    },
    {
      "epoch": 2.5530182939891177,
      "grad_norm": 6.0026373863220215,
      "learning_rate": 4.833468921409594e-05,
      "loss": 2.52010555267334,
      "memory(GiB)": 77.56,
      "step": 59590,
      "token_acc": 0.5058823529411764,
      "train_speed(iter/s)": 1.438293
    },
    {
      "epoch": 2.553232509318367,
      "grad_norm": 8.448450088500977,
      "learning_rate": 4.832796318987188e-05,
      "loss": 2.5625141143798826,
      "memory(GiB)": 77.56,
      "step": 59595,
      "token_acc": 0.4517374517374517,
      "train_speed(iter/s)": 1.438253
    },
    {
      "epoch": 2.553446724647616,
      "grad_norm": 6.984602451324463,
      "learning_rate": 4.832123719593834e-05,
      "loss": 2.6494930267333983,
      "memory(GiB)": 77.56,
      "step": 59600,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.438278
    },
    {
      "epoch": 2.5536609399768646,
      "grad_norm": 6.8267741203308105,
      "learning_rate": 4.831451123241715e-05,
      "loss": 2.034275436401367,
      "memory(GiB)": 77.56,
      "step": 59605,
      "token_acc": 0.5678391959798995,
      "train_speed(iter/s)": 1.438309
    },
    {
      "epoch": 2.553875155306114,
      "grad_norm": 6.508211135864258,
      "learning_rate": 4.8307785299430156e-05,
      "loss": 2.752101516723633,
      "memory(GiB)": 77.56,
      "step": 59610,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.438319
    },
    {
      "epoch": 2.5540893706353627,
      "grad_norm": 4.957455158233643,
      "learning_rate": 4.830105939709924e-05,
      "loss": 2.447592544555664,
      "memory(GiB)": 77.56,
      "step": 59615,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 2.5543035859646115,
      "grad_norm": 6.223200798034668,
      "learning_rate": 4.8294333525546234e-05,
      "loss": 2.5762271881103516,
      "memory(GiB)": 77.56,
      "step": 59620,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.438277
    },
    {
      "epoch": 2.5545178012938607,
      "grad_norm": 6.365777015686035,
      "learning_rate": 4.828760768489295e-05,
      "loss": 2.6135543823242187,
      "memory(GiB)": 77.56,
      "step": 59625,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.4383
    },
    {
      "epoch": 2.5547320166231096,
      "grad_norm": 5.369986057281494,
      "learning_rate": 4.8280881875261284e-05,
      "loss": 2.4641145706176757,
      "memory(GiB)": 77.56,
      "step": 59630,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.438298
    },
    {
      "epoch": 2.5549462319523584,
      "grad_norm": 6.556378364562988,
      "learning_rate": 4.8274156096773046e-05,
      "loss": 2.4437116622924804,
      "memory(GiB)": 77.56,
      "step": 59635,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.438305
    },
    {
      "epoch": 2.5551604472816076,
      "grad_norm": 5.670872211456299,
      "learning_rate": 4.8267430349550094e-05,
      "loss": 2.2682395935058595,
      "memory(GiB)": 77.56,
      "step": 59640,
      "token_acc": 0.5280898876404494,
      "train_speed(iter/s)": 1.438307
    },
    {
      "epoch": 2.5553746626108564,
      "grad_norm": 5.879903316497803,
      "learning_rate": 4.826070463371427e-05,
      "loss": 2.502425193786621,
      "memory(GiB)": 77.56,
      "step": 59645,
      "token_acc": 0.4613003095975232,
      "train_speed(iter/s)": 1.438301
    },
    {
      "epoch": 2.5555888779401053,
      "grad_norm": 6.359103679656982,
      "learning_rate": 4.8253978949387394e-05,
      "loss": 2.4154285430908202,
      "memory(GiB)": 77.56,
      "step": 59650,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.438327
    },
    {
      "epoch": 2.5558030932693545,
      "grad_norm": 4.669216156005859,
      "learning_rate": 4.824725329669135e-05,
      "loss": 2.3376068115234374,
      "memory(GiB)": 77.56,
      "step": 59655,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.438346
    },
    {
      "epoch": 2.5560173085986033,
      "grad_norm": 5.696174621582031,
      "learning_rate": 4.824052767574795e-05,
      "loss": 2.300503158569336,
      "memory(GiB)": 77.56,
      "step": 59660,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.438346
    },
    {
      "epoch": 2.556231523927852,
      "grad_norm": 4.010333061218262,
      "learning_rate": 4.823380208667903e-05,
      "loss": 2.602652168273926,
      "memory(GiB)": 77.56,
      "step": 59665,
      "token_acc": 0.45609065155807366,
      "train_speed(iter/s)": 1.43832
    },
    {
      "epoch": 2.5564457392571014,
      "grad_norm": 6.473670959472656,
      "learning_rate": 4.8227076529606455e-05,
      "loss": 2.2469112396240236,
      "memory(GiB)": 77.56,
      "step": 59670,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.438321
    },
    {
      "epoch": 2.55665995458635,
      "grad_norm": 4.375396251678467,
      "learning_rate": 4.8220351004652036e-05,
      "loss": 2.400933265686035,
      "memory(GiB)": 77.56,
      "step": 59675,
      "token_acc": 0.48286604361370716,
      "train_speed(iter/s)": 1.438346
    },
    {
      "epoch": 2.556874169915599,
      "grad_norm": 5.402980804443359,
      "learning_rate": 4.8213625511937644e-05,
      "loss": 2.8342254638671873,
      "memory(GiB)": 77.56,
      "step": 59680,
      "token_acc": 0.436426116838488,
      "train_speed(iter/s)": 1.438372
    },
    {
      "epoch": 2.5570883852448483,
      "grad_norm": 6.275370121002197,
      "learning_rate": 4.820690005158508e-05,
      "loss": 2.266942596435547,
      "memory(GiB)": 77.56,
      "step": 59685,
      "token_acc": 0.5309446254071661,
      "train_speed(iter/s)": 1.438365
    },
    {
      "epoch": 2.557302600574097,
      "grad_norm": 4.4692888259887695,
      "learning_rate": 4.820017462371622e-05,
      "loss": 2.4160953521728517,
      "memory(GiB)": 77.56,
      "step": 59690,
      "token_acc": 0.45353159851301117,
      "train_speed(iter/s)": 1.438365
    },
    {
      "epoch": 2.557516815903346,
      "grad_norm": 8.16158390045166,
      "learning_rate": 4.819344922845288e-05,
      "loss": 2.6106136322021483,
      "memory(GiB)": 77.56,
      "step": 59695,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.438352
    },
    {
      "epoch": 2.557731031232595,
      "grad_norm": 7.145231246948242,
      "learning_rate": 4.818672386591691e-05,
      "loss": 2.588006019592285,
      "memory(GiB)": 77.56,
      "step": 59700,
      "token_acc": 0.4440894568690096,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 2.557945246561844,
      "grad_norm": 5.022751331329346,
      "learning_rate": 4.817999853623014e-05,
      "loss": 2.312425231933594,
      "memory(GiB)": 77.56,
      "step": 59705,
      "token_acc": 0.461038961038961,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 2.558159461891093,
      "grad_norm": 7.166477203369141,
      "learning_rate": 4.8173273239514396e-05,
      "loss": 2.4700672149658205,
      "memory(GiB)": 77.56,
      "step": 59710,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.438366
    },
    {
      "epoch": 2.558373677220342,
      "grad_norm": 5.713319778442383,
      "learning_rate": 4.816654797589153e-05,
      "loss": 2.461689758300781,
      "memory(GiB)": 77.56,
      "step": 59715,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.438363
    },
    {
      "epoch": 2.558587892549591,
      "grad_norm": 5.759060859680176,
      "learning_rate": 4.815982274548335e-05,
      "loss": 2.366672134399414,
      "memory(GiB)": 77.56,
      "step": 59720,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.438377
    },
    {
      "epoch": 2.5588021078788397,
      "grad_norm": 7.476055145263672,
      "learning_rate": 4.815309754841172e-05,
      "loss": 2.295948600769043,
      "memory(GiB)": 77.56,
      "step": 59725,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.438382
    },
    {
      "epoch": 2.559016323208089,
      "grad_norm": 6.503139495849609,
      "learning_rate": 4.814637238479847e-05,
      "loss": 2.4775360107421873,
      "memory(GiB)": 77.56,
      "step": 59730,
      "token_acc": 0.509375,
      "train_speed(iter/s)": 1.438402
    },
    {
      "epoch": 2.5592305385373377,
      "grad_norm": 4.851894855499268,
      "learning_rate": 4.8139647254765404e-05,
      "loss": 2.321548843383789,
      "memory(GiB)": 77.56,
      "step": 59735,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.43843
    },
    {
      "epoch": 2.5594447538665865,
      "grad_norm": 5.576775074005127,
      "learning_rate": 4.8132922158434384e-05,
      "loss": 2.3857646942138673,
      "memory(GiB)": 77.56,
      "step": 59740,
      "token_acc": 0.4956140350877193,
      "train_speed(iter/s)": 1.438444
    },
    {
      "epoch": 2.559658969195836,
      "grad_norm": 4.1467132568359375,
      "learning_rate": 4.812619709592723e-05,
      "loss": 2.2559581756591798,
      "memory(GiB)": 77.56,
      "step": 59745,
      "token_acc": 0.508,
      "train_speed(iter/s)": 1.438444
    },
    {
      "epoch": 2.5598731845250846,
      "grad_norm": 7.181698799133301,
      "learning_rate": 4.8119472067365766e-05,
      "loss": 2.2758745193481444,
      "memory(GiB)": 77.56,
      "step": 59750,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.438452
    },
    {
      "epoch": 2.5600873998543334,
      "grad_norm": 4.9263129234313965,
      "learning_rate": 4.8112747072871836e-05,
      "loss": 2.543910789489746,
      "memory(GiB)": 77.56,
      "step": 59755,
      "token_acc": 0.45051194539249145,
      "train_speed(iter/s)": 1.438467
    },
    {
      "epoch": 2.5603016151835827,
      "grad_norm": 5.002185821533203,
      "learning_rate": 4.8106022112567247e-05,
      "loss": 2.6644874572753907,
      "memory(GiB)": 77.56,
      "step": 59760,
      "token_acc": 0.43354430379746833,
      "train_speed(iter/s)": 1.438482
    },
    {
      "epoch": 2.5605158305128315,
      "grad_norm": 4.545617580413818,
      "learning_rate": 4.809929718657386e-05,
      "loss": 2.468119430541992,
      "memory(GiB)": 77.56,
      "step": 59765,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.438496
    },
    {
      "epoch": 2.5607300458420803,
      "grad_norm": 7.644782066345215,
      "learning_rate": 4.809257229501348e-05,
      "loss": 2.3550495147705077,
      "memory(GiB)": 77.56,
      "step": 59770,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.438471
    },
    {
      "epoch": 2.5609442611713296,
      "grad_norm": 7.732504844665527,
      "learning_rate": 4.8085847438007955e-05,
      "loss": 2.4434947967529297,
      "memory(GiB)": 77.56,
      "step": 59775,
      "token_acc": 0.4528985507246377,
      "train_speed(iter/s)": 1.438483
    },
    {
      "epoch": 2.5611584765005784,
      "grad_norm": 4.603659152984619,
      "learning_rate": 4.807912261567908e-05,
      "loss": 2.3427825927734376,
      "memory(GiB)": 77.56,
      "step": 59780,
      "token_acc": 0.5070921985815603,
      "train_speed(iter/s)": 1.438471
    },
    {
      "epoch": 2.561372691829827,
      "grad_norm": 6.309761047363281,
      "learning_rate": 4.807239782814872e-05,
      "loss": 2.404462432861328,
      "memory(GiB)": 77.56,
      "step": 59785,
      "token_acc": 0.5234375,
      "train_speed(iter/s)": 1.438483
    },
    {
      "epoch": 2.5615869071590764,
      "grad_norm": 5.627111434936523,
      "learning_rate": 4.8065673075538685e-05,
      "loss": 2.213104248046875,
      "memory(GiB)": 77.56,
      "step": 59790,
      "token_acc": 0.5191082802547771,
      "train_speed(iter/s)": 1.438469
    },
    {
      "epoch": 2.5618011224883253,
      "grad_norm": 4.7075276374816895,
      "learning_rate": 4.805894835797078e-05,
      "loss": 2.0538761138916017,
      "memory(GiB)": 77.56,
      "step": 59795,
      "token_acc": 0.540650406504065,
      "train_speed(iter/s)": 1.438466
    },
    {
      "epoch": 2.562015337817574,
      "grad_norm": 5.505980491638184,
      "learning_rate": 4.805222367556685e-05,
      "loss": 2.2942991256713867,
      "memory(GiB)": 77.56,
      "step": 59800,
      "token_acc": 0.4823943661971831,
      "train_speed(iter/s)": 1.438501
    },
    {
      "epoch": 2.5622295531468233,
      "grad_norm": 5.822419166564941,
      "learning_rate": 4.804549902844873e-05,
      "loss": 2.2477752685546877,
      "memory(GiB)": 77.56,
      "step": 59805,
      "token_acc": 0.48672566371681414,
      "train_speed(iter/s)": 1.438498
    },
    {
      "epoch": 2.562443768476072,
      "grad_norm": 5.29340124130249,
      "learning_rate": 4.8038774416738205e-05,
      "loss": 2.4582977294921875,
      "memory(GiB)": 77.56,
      "step": 59810,
      "token_acc": 0.47307692307692306,
      "train_speed(iter/s)": 1.43848
    },
    {
      "epoch": 2.562657983805321,
      "grad_norm": 5.51577091217041,
      "learning_rate": 4.803204984055714e-05,
      "loss": 2.402615547180176,
      "memory(GiB)": 77.56,
      "step": 59815,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.438483
    },
    {
      "epoch": 2.56287219913457,
      "grad_norm": 5.877311706542969,
      "learning_rate": 4.802532530002733e-05,
      "loss": 2.339146041870117,
      "memory(GiB)": 77.56,
      "step": 59820,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.438484
    },
    {
      "epoch": 2.563086414463819,
      "grad_norm": 5.680993556976318,
      "learning_rate": 4.801860079527061e-05,
      "loss": 2.393800735473633,
      "memory(GiB)": 77.56,
      "step": 59825,
      "token_acc": 0.4881656804733728,
      "train_speed(iter/s)": 1.43849
    },
    {
      "epoch": 2.563300629793068,
      "grad_norm": 5.882232666015625,
      "learning_rate": 4.8011876326408796e-05,
      "loss": 2.3937700271606444,
      "memory(GiB)": 77.56,
      "step": 59830,
      "token_acc": 0.4785992217898833,
      "train_speed(iter/s)": 1.438508
    },
    {
      "epoch": 2.563514845122317,
      "grad_norm": 5.29835319519043,
      "learning_rate": 4.8005151893563684e-05,
      "loss": 2.4504329681396486,
      "memory(GiB)": 77.56,
      "step": 59835,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438518
    },
    {
      "epoch": 2.563729060451566,
      "grad_norm": 5.1285400390625,
      "learning_rate": 4.799842749685713e-05,
      "loss": 2.3803762435913085,
      "memory(GiB)": 77.56,
      "step": 59840,
      "token_acc": 0.4314868804664723,
      "train_speed(iter/s)": 1.438549
    },
    {
      "epoch": 2.5639432757808147,
      "grad_norm": 4.771817207336426,
      "learning_rate": 4.799170313641095e-05,
      "loss": 2.4962512969970705,
      "memory(GiB)": 77.56,
      "step": 59845,
      "token_acc": 0.4871060171919771,
      "train_speed(iter/s)": 1.438557
    },
    {
      "epoch": 2.564157491110064,
      "grad_norm": 5.648887634277344,
      "learning_rate": 4.798497881234695e-05,
      "loss": 2.5638856887817383,
      "memory(GiB)": 77.56,
      "step": 59850,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.43857
    },
    {
      "epoch": 2.564371706439313,
      "grad_norm": 4.494454860687256,
      "learning_rate": 4.7978254524786935e-05,
      "loss": 2.489668273925781,
      "memory(GiB)": 77.56,
      "step": 59855,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.438552
    },
    {
      "epoch": 2.5645859217685616,
      "grad_norm": 7.119904518127441,
      "learning_rate": 4.7971530273852754e-05,
      "loss": 2.4390853881835937,
      "memory(GiB)": 77.56,
      "step": 59860,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 2.564800137097811,
      "grad_norm": 7.292238712310791,
      "learning_rate": 4.796480605966619e-05,
      "loss": 2.300970458984375,
      "memory(GiB)": 77.56,
      "step": 59865,
      "token_acc": 0.5321888412017167,
      "train_speed(iter/s)": 1.438528
    },
    {
      "epoch": 2.5650143524270597,
      "grad_norm": 7.616230487823486,
      "learning_rate": 4.795808188234909e-05,
      "loss": 2.079450988769531,
      "memory(GiB)": 77.56,
      "step": 59870,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.438536
    },
    {
      "epoch": 2.5652285677563085,
      "grad_norm": 4.937198638916016,
      "learning_rate": 4.795135774202324e-05,
      "loss": 2.3414304733276365,
      "memory(GiB)": 77.56,
      "step": 59875,
      "token_acc": 0.512,
      "train_speed(iter/s)": 1.438518
    },
    {
      "epoch": 2.5654427830855577,
      "grad_norm": 5.350150108337402,
      "learning_rate": 4.794463363881047e-05,
      "loss": 2.444717025756836,
      "memory(GiB)": 77.56,
      "step": 59880,
      "token_acc": 0.5322033898305085,
      "train_speed(iter/s)": 1.438497
    },
    {
      "epoch": 2.5656569984148065,
      "grad_norm": 4.827812671661377,
      "learning_rate": 4.793790957283259e-05,
      "loss": 2.355134201049805,
      "memory(GiB)": 77.56,
      "step": 59885,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.438499
    },
    {
      "epoch": 2.565871213744056,
      "grad_norm": 6.479953765869141,
      "learning_rate": 4.7931185544211416e-05,
      "loss": 2.437975311279297,
      "memory(GiB)": 77.56,
      "step": 59890,
      "token_acc": 0.46494464944649444,
      "train_speed(iter/s)": 1.438505
    },
    {
      "epoch": 2.5660854290733046,
      "grad_norm": 5.186941146850586,
      "learning_rate": 4.7924461553068745e-05,
      "loss": 2.648612213134766,
      "memory(GiB)": 77.56,
      "step": 59895,
      "token_acc": 0.43630573248407645,
      "train_speed(iter/s)": 1.438468
    },
    {
      "epoch": 2.5662996444025534,
      "grad_norm": 4.517250061035156,
      "learning_rate": 4.7917737599526415e-05,
      "loss": 2.4635990142822264,
      "memory(GiB)": 77.56,
      "step": 59900,
      "token_acc": 0.4574468085106383,
      "train_speed(iter/s)": 1.438478
    },
    {
      "epoch": 2.5665138597318027,
      "grad_norm": 9.733177185058594,
      "learning_rate": 4.791101368370619e-05,
      "loss": 2.624422073364258,
      "memory(GiB)": 77.56,
      "step": 59905,
      "token_acc": 0.5056179775280899,
      "train_speed(iter/s)": 1.438486
    },
    {
      "epoch": 2.5667280750610515,
      "grad_norm": 4.182031631469727,
      "learning_rate": 4.790428980572994e-05,
      "loss": 2.271896743774414,
      "memory(GiB)": 77.56,
      "step": 59910,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.438428
    },
    {
      "epoch": 2.5669422903903003,
      "grad_norm": 6.960338115692139,
      "learning_rate": 4.789756596571944e-05,
      "loss": 2.3902454376220703,
      "memory(GiB)": 77.56,
      "step": 59915,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.438447
    },
    {
      "epoch": 2.5671565057195496,
      "grad_norm": 5.9755659103393555,
      "learning_rate": 4.789084216379651e-05,
      "loss": 2.0333044052124025,
      "memory(GiB)": 77.56,
      "step": 59920,
      "token_acc": 0.5815602836879432,
      "train_speed(iter/s)": 1.438457
    },
    {
      "epoch": 2.5673707210487984,
      "grad_norm": 6.6166510581970215,
      "learning_rate": 4.788411840008294e-05,
      "loss": 2.3599395751953125,
      "memory(GiB)": 77.56,
      "step": 59925,
      "token_acc": 0.5124555160142349,
      "train_speed(iter/s)": 1.43846
    },
    {
      "epoch": 2.567584936378047,
      "grad_norm": 5.649675369262695,
      "learning_rate": 4.7877394674700564e-05,
      "loss": 2.254487228393555,
      "memory(GiB)": 77.56,
      "step": 59930,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.438468
    },
    {
      "epoch": 2.5677991517072964,
      "grad_norm": 5.2218403816223145,
      "learning_rate": 4.787067098777117e-05,
      "loss": 2.483497428894043,
      "memory(GiB)": 77.56,
      "step": 59935,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.438477
    },
    {
      "epoch": 2.5680133670365453,
      "grad_norm": 4.606977462768555,
      "learning_rate": 4.786394733941657e-05,
      "loss": 2.644822883605957,
      "memory(GiB)": 77.56,
      "step": 59940,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.438479
    },
    {
      "epoch": 2.568227582365794,
      "grad_norm": 5.572807788848877,
      "learning_rate": 4.785722372975857e-05,
      "loss": 2.6316604614257812,
      "memory(GiB)": 77.56,
      "step": 59945,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.438486
    },
    {
      "epoch": 2.5684417976950433,
      "grad_norm": 5.559742450714111,
      "learning_rate": 4.785050015891897e-05,
      "loss": 2.5154964447021486,
      "memory(GiB)": 77.56,
      "step": 59950,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438492
    },
    {
      "epoch": 2.568656013024292,
      "grad_norm": 6.341352462768555,
      "learning_rate": 4.7843776627019567e-05,
      "loss": 2.443520164489746,
      "memory(GiB)": 77.56,
      "step": 59955,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.438499
    },
    {
      "epoch": 2.568870228353541,
      "grad_norm": 4.311911106109619,
      "learning_rate": 4.7837053134182183e-05,
      "loss": 2.665169525146484,
      "memory(GiB)": 77.56,
      "step": 59960,
      "token_acc": 0.47987616099071206,
      "train_speed(iter/s)": 1.43847
    },
    {
      "epoch": 2.56908444368279,
      "grad_norm": 10.258307456970215,
      "learning_rate": 4.78303296805286e-05,
      "loss": 2.4120904922485353,
      "memory(GiB)": 77.56,
      "step": 59965,
      "token_acc": 0.4967532467532468,
      "train_speed(iter/s)": 1.438451
    },
    {
      "epoch": 2.569298659012039,
      "grad_norm": 5.7733283042907715,
      "learning_rate": 4.782360626618064e-05,
      "loss": 2.4882480621337892,
      "memory(GiB)": 77.56,
      "step": 59970,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.438452
    },
    {
      "epoch": 2.569512874341288,
      "grad_norm": 5.327549457550049,
      "learning_rate": 4.78168828912601e-05,
      "loss": 2.0749256134033205,
      "memory(GiB)": 77.56,
      "step": 59975,
      "token_acc": 0.5335968379446641,
      "train_speed(iter/s)": 1.438476
    },
    {
      "epoch": 2.569727089670537,
      "grad_norm": 6.815774917602539,
      "learning_rate": 4.781015955588875e-05,
      "loss": 2.417924690246582,
      "memory(GiB)": 77.56,
      "step": 59980,
      "token_acc": 0.5108225108225108,
      "train_speed(iter/s)": 1.438496
    },
    {
      "epoch": 2.569941304999786,
      "grad_norm": 7.461811065673828,
      "learning_rate": 4.7803436260188425e-05,
      "loss": 2.2282024383544923,
      "memory(GiB)": 77.56,
      "step": 59985,
      "token_acc": 0.556420233463035,
      "train_speed(iter/s)": 1.438505
    },
    {
      "epoch": 2.5701555203290347,
      "grad_norm": 5.774947166442871,
      "learning_rate": 4.779671300428092e-05,
      "loss": 2.406715202331543,
      "memory(GiB)": 77.56,
      "step": 59990,
      "token_acc": 0.541095890410959,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 2.570369735658284,
      "grad_norm": 5.347567081451416,
      "learning_rate": 4.778998978828803e-05,
      "loss": 2.5186132431030273,
      "memory(GiB)": 77.56,
      "step": 59995,
      "token_acc": 0.45663265306122447,
      "train_speed(iter/s)": 1.438496
    },
    {
      "epoch": 2.570583950987533,
      "grad_norm": 6.227933883666992,
      "learning_rate": 4.7783266612331536e-05,
      "loss": 2.5292888641357423,
      "memory(GiB)": 77.56,
      "step": 60000,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 2.570583950987533,
      "eval_loss": 2.0648553371429443,
      "eval_runtime": 14.8514,
      "eval_samples_per_second": 6.733,
      "eval_steps_per_second": 6.733,
      "eval_token_acc": 0.4709066305818674,
      "step": 60000
    },
    {
      "epoch": 2.5707981663167816,
      "grad_norm": 4.6119384765625,
      "learning_rate": 4.777654347653326e-05,
      "loss": 2.4082363128662108,
      "memory(GiB)": 77.56,
      "step": 60005,
      "token_acc": 0.47517039922103216,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 2.571012381646031,
      "grad_norm": 5.563006401062012,
      "learning_rate": 4.776982038101497e-05,
      "loss": 2.075536918640137,
      "memory(GiB)": 77.56,
      "step": 60010,
      "token_acc": 0.5698529411764706,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 2.5712265969752797,
      "grad_norm": 5.174965858459473,
      "learning_rate": 4.776309732589849e-05,
      "loss": 2.38769474029541,
      "memory(GiB)": 77.56,
      "step": 60015,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.438007
    },
    {
      "epoch": 2.5714408123045285,
      "grad_norm": 4.748098373413086,
      "learning_rate": 4.775637431130559e-05,
      "loss": 2.555836486816406,
      "memory(GiB)": 77.56,
      "step": 60020,
      "token_acc": 0.4892966360856269,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 2.5716550276337777,
      "grad_norm": 6.356767654418945,
      "learning_rate": 4.774965133735808e-05,
      "loss": 2.361519622802734,
      "memory(GiB)": 77.56,
      "step": 60025,
      "token_acc": 0.4908424908424908,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 2.5718692429630265,
      "grad_norm": 5.02504825592041,
      "learning_rate": 4.7742928404177746e-05,
      "loss": 2.428948974609375,
      "memory(GiB)": 77.56,
      "step": 60030,
      "token_acc": 0.4492307692307692,
      "train_speed(iter/s)": 1.437969
    },
    {
      "epoch": 2.5720834582922754,
      "grad_norm": 6.378479480743408,
      "learning_rate": 4.773620551188638e-05,
      "loss": 2.5078826904296876,
      "memory(GiB)": 77.56,
      "step": 60035,
      "token_acc": 0.4657534246575342,
      "train_speed(iter/s)": 1.437953
    },
    {
      "epoch": 2.5722976736215246,
      "grad_norm": 4.581421375274658,
      "learning_rate": 4.772948266060577e-05,
      "loss": 2.4601423263549806,
      "memory(GiB)": 77.56,
      "step": 60040,
      "token_acc": 0.46779661016949153,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 2.5725118889507734,
      "grad_norm": 7.045839309692383,
      "learning_rate": 4.772275985045772e-05,
      "loss": 2.8986183166503907,
      "memory(GiB)": 77.56,
      "step": 60045,
      "token_acc": 0.42662116040955633,
      "train_speed(iter/s)": 1.437985
    },
    {
      "epoch": 2.5727261042800222,
      "grad_norm": 6.260333061218262,
      "learning_rate": 4.7716037081564004e-05,
      "loss": 2.5616737365722657,
      "memory(GiB)": 77.56,
      "step": 60050,
      "token_acc": 0.4429530201342282,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 2.5729403196092715,
      "grad_norm": 4.881079196929932,
      "learning_rate": 4.7709314354046415e-05,
      "loss": 2.420989990234375,
      "memory(GiB)": 77.56,
      "step": 60055,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 2.5731545349385203,
      "grad_norm": 4.75499963760376,
      "learning_rate": 4.7702591668026745e-05,
      "loss": 2.6270772933959963,
      "memory(GiB)": 77.56,
      "step": 60060,
      "token_acc": 0.453125,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 2.573368750267769,
      "grad_norm": 5.190093994140625,
      "learning_rate": 4.769586902362679e-05,
      "loss": 2.395723342895508,
      "memory(GiB)": 77.56,
      "step": 60065,
      "token_acc": 0.47257383966244726,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 2.5735829655970184,
      "grad_norm": 4.898736953735352,
      "learning_rate": 4.768914642096833e-05,
      "loss": 2.443906784057617,
      "memory(GiB)": 77.56,
      "step": 60070,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 2.573797180926267,
      "grad_norm": 7.562080383300781,
      "learning_rate": 4.768242386017315e-05,
      "loss": 2.615232467651367,
      "memory(GiB)": 77.56,
      "step": 60075,
      "token_acc": 0.5038759689922481,
      "train_speed(iter/s)": 1.4381
    },
    {
      "epoch": 2.574011396255516,
      "grad_norm": 4.500131607055664,
      "learning_rate": 4.767570134136304e-05,
      "loss": 2.378614807128906,
      "memory(GiB)": 77.56,
      "step": 60080,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.438115
    },
    {
      "epoch": 2.5742256115847653,
      "grad_norm": 5.240380764007568,
      "learning_rate": 4.766897886465977e-05,
      "loss": 2.3287225723266602,
      "memory(GiB)": 77.56,
      "step": 60085,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 2.574439826914014,
      "grad_norm": 4.375026226043701,
      "learning_rate": 4.766225643018514e-05,
      "loss": 2.0763872146606444,
      "memory(GiB)": 77.56,
      "step": 60090,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438114
    },
    {
      "epoch": 2.574654042243263,
      "grad_norm": 5.3339667320251465,
      "learning_rate": 4.765553403806094e-05,
      "loss": 2.3074771881103517,
      "memory(GiB)": 77.56,
      "step": 60095,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.438112
    },
    {
      "epoch": 2.574868257572512,
      "grad_norm": 6.190136909484863,
      "learning_rate": 4.764881168840892e-05,
      "loss": 2.3066503524780275,
      "memory(GiB)": 77.56,
      "step": 60100,
      "token_acc": 0.4942084942084942,
      "train_speed(iter/s)": 1.438097
    },
    {
      "epoch": 2.575082472901761,
      "grad_norm": 4.96804141998291,
      "learning_rate": 4.7642089381350895e-05,
      "loss": 2.587931442260742,
      "memory(GiB)": 77.56,
      "step": 60105,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.438109
    },
    {
      "epoch": 2.5752966882310098,
      "grad_norm": 4.135900020599365,
      "learning_rate": 4.763536711700862e-05,
      "loss": 2.215572738647461,
      "memory(GiB)": 77.56,
      "step": 60110,
      "token_acc": 0.5047021943573667,
      "train_speed(iter/s)": 1.438078
    },
    {
      "epoch": 2.575510903560259,
      "grad_norm": 8.227154731750488,
      "learning_rate": 4.76286448955039e-05,
      "loss": 2.502964210510254,
      "memory(GiB)": 77.56,
      "step": 60115,
      "token_acc": 0.4796747967479675,
      "train_speed(iter/s)": 1.438093
    },
    {
      "epoch": 2.575725118889508,
      "grad_norm": 5.9069976806640625,
      "learning_rate": 4.76219227169585e-05,
      "loss": 2.428532028198242,
      "memory(GiB)": 77.56,
      "step": 60120,
      "token_acc": 0.5234375,
      "train_speed(iter/s)": 1.438082
    },
    {
      "epoch": 2.5759393342187566,
      "grad_norm": 17.34946060180664,
      "learning_rate": 4.7615200581494194e-05,
      "loss": 2.37042121887207,
      "memory(GiB)": 77.56,
      "step": 60125,
      "token_acc": 0.464,
      "train_speed(iter/s)": 1.438088
    },
    {
      "epoch": 2.576153549548006,
      "grad_norm": 6.491145610809326,
      "learning_rate": 4.7608478489232756e-05,
      "loss": 2.754975128173828,
      "memory(GiB)": 77.56,
      "step": 60130,
      "token_acc": 0.45323741007194246,
      "train_speed(iter/s)": 1.43811
    },
    {
      "epoch": 2.5763677648772547,
      "grad_norm": 5.394059181213379,
      "learning_rate": 4.760175644029599e-05,
      "loss": 2.623165512084961,
      "memory(GiB)": 77.56,
      "step": 60135,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438133
    },
    {
      "epoch": 2.5765819802065035,
      "grad_norm": 5.3159379959106445,
      "learning_rate": 4.759503443480566e-05,
      "loss": 2.5075891494750975,
      "memory(GiB)": 77.56,
      "step": 60140,
      "token_acc": 0.449685534591195,
      "train_speed(iter/s)": 1.438139
    },
    {
      "epoch": 2.576796195535753,
      "grad_norm": 5.586711406707764,
      "learning_rate": 4.758831247288353e-05,
      "loss": 2.483822250366211,
      "memory(GiB)": 77.56,
      "step": 60145,
      "token_acc": 0.4758364312267658,
      "train_speed(iter/s)": 1.438153
    },
    {
      "epoch": 2.5770104108650016,
      "grad_norm": 5.024467468261719,
      "learning_rate": 4.758159055465138e-05,
      "loss": 2.680333709716797,
      "memory(GiB)": 77.56,
      "step": 60150,
      "token_acc": 0.4501347708894879,
      "train_speed(iter/s)": 1.438174
    },
    {
      "epoch": 2.5772246261942504,
      "grad_norm": 6.300375461578369,
      "learning_rate": 4.757486868023099e-05,
      "loss": 2.555211639404297,
      "memory(GiB)": 77.56,
      "step": 60155,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.438191
    },
    {
      "epoch": 2.5774388415234997,
      "grad_norm": 4.312414169311523,
      "learning_rate": 4.756814684974413e-05,
      "loss": 2.3841041564941405,
      "memory(GiB)": 77.56,
      "step": 60160,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.438207
    },
    {
      "epoch": 2.5776530568527485,
      "grad_norm": 9.535069465637207,
      "learning_rate": 4.756142506331258e-05,
      "loss": 2.7303104400634766,
      "memory(GiB)": 77.56,
      "step": 60165,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.438196
    },
    {
      "epoch": 2.5778672721819973,
      "grad_norm": 5.826118469238281,
      "learning_rate": 4.755470332105808e-05,
      "loss": 2.127614974975586,
      "memory(GiB)": 77.56,
      "step": 60170,
      "token_acc": 0.550561797752809,
      "train_speed(iter/s)": 1.438206
    },
    {
      "epoch": 2.5780814875112466,
      "grad_norm": 5.456142425537109,
      "learning_rate": 4.754798162310244e-05,
      "loss": 2.158995246887207,
      "memory(GiB)": 77.56,
      "step": 60175,
      "token_acc": 0.5503355704697986,
      "train_speed(iter/s)": 1.438178
    },
    {
      "epoch": 2.5782957028404954,
      "grad_norm": 4.853738784790039,
      "learning_rate": 4.7541259969567416e-05,
      "loss": 2.2282285690307617,
      "memory(GiB)": 77.56,
      "step": 60180,
      "token_acc": 0.47540983606557374,
      "train_speed(iter/s)": 1.43816
    },
    {
      "epoch": 2.578509918169744,
      "grad_norm": 5.5688395500183105,
      "learning_rate": 4.753453836057476e-05,
      "loss": 2.418585014343262,
      "memory(GiB)": 77.56,
      "step": 60185,
      "token_acc": 0.48535564853556484,
      "train_speed(iter/s)": 1.43815
    },
    {
      "epoch": 2.5787241334989934,
      "grad_norm": 5.165946960449219,
      "learning_rate": 4.752781679624626e-05,
      "loss": 2.448345947265625,
      "memory(GiB)": 77.56,
      "step": 60190,
      "token_acc": 0.4966442953020134,
      "train_speed(iter/s)": 1.438182
    },
    {
      "epoch": 2.5789383488282422,
      "grad_norm": 4.71059513092041,
      "learning_rate": 4.7521095276703676e-05,
      "loss": 2.491571807861328,
      "memory(GiB)": 77.56,
      "step": 60195,
      "token_acc": 0.5078864353312302,
      "train_speed(iter/s)": 1.438156
    },
    {
      "epoch": 2.579152564157491,
      "grad_norm": 5.885578155517578,
      "learning_rate": 4.7514373802068786e-05,
      "loss": 2.4665386199951174,
      "memory(GiB)": 77.56,
      "step": 60200,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.438173
    },
    {
      "epoch": 2.5793667794867403,
      "grad_norm": 4.911962985992432,
      "learning_rate": 4.750765237246332e-05,
      "loss": 2.236262893676758,
      "memory(GiB)": 77.56,
      "step": 60205,
      "token_acc": 0.5146579804560261,
      "train_speed(iter/s)": 1.438166
    },
    {
      "epoch": 2.579580994815989,
      "grad_norm": 7.053640365600586,
      "learning_rate": 4.750093098800909e-05,
      "loss": 2.3783876419067385,
      "memory(GiB)": 77.56,
      "step": 60210,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43821
    },
    {
      "epoch": 2.579795210145238,
      "grad_norm": 5.45408296585083,
      "learning_rate": 4.749420964882783e-05,
      "loss": 2.621144104003906,
      "memory(GiB)": 77.56,
      "step": 60215,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.438213
    },
    {
      "epoch": 2.580009425474487,
      "grad_norm": 5.28698205947876,
      "learning_rate": 4.748748835504133e-05,
      "loss": 2.3261960983276366,
      "memory(GiB)": 77.56,
      "step": 60220,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.438247
    },
    {
      "epoch": 2.580223640803736,
      "grad_norm": 6.509244918823242,
      "learning_rate": 4.7480767106771326e-05,
      "loss": 2.5394302368164063,
      "memory(GiB)": 77.56,
      "step": 60225,
      "token_acc": 0.5092936802973977,
      "train_speed(iter/s)": 1.438232
    },
    {
      "epoch": 2.580437856132985,
      "grad_norm": 5.842759132385254,
      "learning_rate": 4.7474045904139586e-05,
      "loss": 2.6135311126708984,
      "memory(GiB)": 77.56,
      "step": 60230,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.438224
    },
    {
      "epoch": 2.580652071462234,
      "grad_norm": 5.242397308349609,
      "learning_rate": 4.746732474726788e-05,
      "loss": 2.456760025024414,
      "memory(GiB)": 77.56,
      "step": 60235,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.438212
    },
    {
      "epoch": 2.580866286791483,
      "grad_norm": 4.919493198394775,
      "learning_rate": 4.7460603636277956e-05,
      "loss": 2.716153144836426,
      "memory(GiB)": 77.56,
      "step": 60240,
      "token_acc": 0.4919093851132686,
      "train_speed(iter/s)": 1.43822
    },
    {
      "epoch": 2.5810805021207317,
      "grad_norm": 4.709676265716553,
      "learning_rate": 4.7453882571291584e-05,
      "loss": 2.7681114196777346,
      "memory(GiB)": 77.56,
      "step": 60245,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.438224
    },
    {
      "epoch": 2.581294717449981,
      "grad_norm": 5.668246746063232,
      "learning_rate": 4.7447161552430526e-05,
      "loss": 2.4259979248046877,
      "memory(GiB)": 77.56,
      "step": 60250,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.438246
    },
    {
      "epoch": 2.5815089327792298,
      "grad_norm": 4.389259338378906,
      "learning_rate": 4.744044057981651e-05,
      "loss": 2.4198646545410156,
      "memory(GiB)": 77.56,
      "step": 60255,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.438237
    },
    {
      "epoch": 2.5817231481084786,
      "grad_norm": 5.31933069229126,
      "learning_rate": 4.743371965357133e-05,
      "loss": 2.5894342422485352,
      "memory(GiB)": 77.56,
      "step": 60260,
      "token_acc": 0.45980707395498394,
      "train_speed(iter/s)": 1.43824
    },
    {
      "epoch": 2.581937363437728,
      "grad_norm": 5.5895490646362305,
      "learning_rate": 4.742699877381673e-05,
      "loss": 2.382550811767578,
      "memory(GiB)": 77.56,
      "step": 60265,
      "token_acc": 0.5057034220532319,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 2.5821515787669767,
      "grad_norm": 6.026521682739258,
      "learning_rate": 4.7420277940674446e-05,
      "loss": 2.5565954208374024,
      "memory(GiB)": 77.56,
      "step": 60270,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.43826
    },
    {
      "epoch": 2.5823657940962255,
      "grad_norm": 4.510000228881836,
      "learning_rate": 4.741355715426623e-05,
      "loss": 2.3520702362060546,
      "memory(GiB)": 77.56,
      "step": 60275,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.438262
    },
    {
      "epoch": 2.5825800094254747,
      "grad_norm": 7.214466094970703,
      "learning_rate": 4.7406836414713884e-05,
      "loss": 2.511739730834961,
      "memory(GiB)": 77.56,
      "step": 60280,
      "token_acc": 0.4759036144578313,
      "train_speed(iter/s)": 1.438287
    },
    {
      "epoch": 2.5827942247547235,
      "grad_norm": 6.1277971267700195,
      "learning_rate": 4.7400115722139126e-05,
      "loss": 2.3304912567138674,
      "memory(GiB)": 77.56,
      "step": 60285,
      "token_acc": 0.49523809523809526,
      "train_speed(iter/s)": 1.438274
    },
    {
      "epoch": 2.5830084400839723,
      "grad_norm": 6.150754928588867,
      "learning_rate": 4.73933950766637e-05,
      "loss": 2.687809181213379,
      "memory(GiB)": 77.56,
      "step": 60290,
      "token_acc": 0.4859154929577465,
      "train_speed(iter/s)": 1.438296
    },
    {
      "epoch": 2.5832226554132216,
      "grad_norm": 14.555551528930664,
      "learning_rate": 4.738667447840938e-05,
      "loss": 2.4184549331665037,
      "memory(GiB)": 77.56,
      "step": 60295,
      "token_acc": 0.49612403100775193,
      "train_speed(iter/s)": 1.438306
    },
    {
      "epoch": 2.5834368707424704,
      "grad_norm": 9.330744743347168,
      "learning_rate": 4.737995392749789e-05,
      "loss": 2.428777313232422,
      "memory(GiB)": 77.56,
      "step": 60300,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.43833
    },
    {
      "epoch": 2.5836510860717192,
      "grad_norm": 9.24018383026123,
      "learning_rate": 4.7373233424051e-05,
      "loss": 2.525027847290039,
      "memory(GiB)": 77.56,
      "step": 60305,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.438332
    },
    {
      "epoch": 2.5838653014009685,
      "grad_norm": 6.798473358154297,
      "learning_rate": 4.7366512968190454e-05,
      "loss": 2.5032781600952148,
      "memory(GiB)": 77.56,
      "step": 60310,
      "token_acc": 0.5148514851485149,
      "train_speed(iter/s)": 1.438304
    },
    {
      "epoch": 2.5840795167302173,
      "grad_norm": 6.8903727531433105,
      "learning_rate": 4.735979256003798e-05,
      "loss": 2.4880092620849608,
      "memory(GiB)": 77.56,
      "step": 60315,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.438306
    },
    {
      "epoch": 2.584293732059466,
      "grad_norm": 5.280789852142334,
      "learning_rate": 4.735307219971536e-05,
      "loss": 2.2187263488769533,
      "memory(GiB)": 77.56,
      "step": 60320,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438301
    },
    {
      "epoch": 2.5845079473887154,
      "grad_norm": 4.505428791046143,
      "learning_rate": 4.734635188734432e-05,
      "loss": 2.210480880737305,
      "memory(GiB)": 77.56,
      "step": 60325,
      "token_acc": 0.5508196721311476,
      "train_speed(iter/s)": 1.43831
    },
    {
      "epoch": 2.584722162717964,
      "grad_norm": 8.387556076049805,
      "learning_rate": 4.7339631623046585e-05,
      "loss": 2.3935256958007813,
      "memory(GiB)": 77.56,
      "step": 60330,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438333
    },
    {
      "epoch": 2.584936378047213,
      "grad_norm": 5.168647289276123,
      "learning_rate": 4.7332911406943934e-05,
      "loss": 2.561665725708008,
      "memory(GiB)": 77.56,
      "step": 60335,
      "token_acc": 0.50625,
      "train_speed(iter/s)": 1.438365
    },
    {
      "epoch": 2.5851505933764622,
      "grad_norm": 6.293690204620361,
      "learning_rate": 4.732619123915809e-05,
      "loss": 2.3551496505737304,
      "memory(GiB)": 77.56,
      "step": 60340,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.438358
    },
    {
      "epoch": 2.585364808705711,
      "grad_norm": 5.55072546005249,
      "learning_rate": 4.7319471119810805e-05,
      "loss": 2.4079263687133787,
      "memory(GiB)": 77.56,
      "step": 60345,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.438393
    },
    {
      "epoch": 2.58557902403496,
      "grad_norm": 5.686210632324219,
      "learning_rate": 4.731275104902379e-05,
      "loss": 2.4750595092773438,
      "memory(GiB)": 77.56,
      "step": 60350,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.438432
    },
    {
      "epoch": 2.585793239364209,
      "grad_norm": 5.239920139312744,
      "learning_rate": 4.730603102691884e-05,
      "loss": 2.549484634399414,
      "memory(GiB)": 77.56,
      "step": 60355,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.438454
    },
    {
      "epoch": 2.586007454693458,
      "grad_norm": 5.484700679779053,
      "learning_rate": 4.729931105361765e-05,
      "loss": 2.3309253692626952,
      "memory(GiB)": 77.56,
      "step": 60360,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.438472
    },
    {
      "epoch": 2.5862216700227068,
      "grad_norm": 6.028470993041992,
      "learning_rate": 4.7292591129241985e-05,
      "loss": 2.398227500915527,
      "memory(GiB)": 77.56,
      "step": 60365,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.438474
    },
    {
      "epoch": 2.586435885351956,
      "grad_norm": 15.866666793823242,
      "learning_rate": 4.728587125391357e-05,
      "loss": 2.642198944091797,
      "memory(GiB)": 77.56,
      "step": 60370,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.438502
    },
    {
      "epoch": 2.586650100681205,
      "grad_norm": 5.291264533996582,
      "learning_rate": 4.727915142775414e-05,
      "loss": 2.6901432037353517,
      "memory(GiB)": 77.56,
      "step": 60375,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.438483
    },
    {
      "epoch": 2.5868643160104536,
      "grad_norm": 5.216635704040527,
      "learning_rate": 4.7272431650885436e-05,
      "loss": 2.340800094604492,
      "memory(GiB)": 77.56,
      "step": 60380,
      "token_acc": 0.4697406340057637,
      "train_speed(iter/s)": 1.438503
    },
    {
      "epoch": 2.587078531339703,
      "grad_norm": 6.237765312194824,
      "learning_rate": 4.726571192342919e-05,
      "loss": 2.367164421081543,
      "memory(GiB)": 77.56,
      "step": 60385,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.438507
    },
    {
      "epoch": 2.5872927466689517,
      "grad_norm": 5.036973476409912,
      "learning_rate": 4.7258992245507134e-05,
      "loss": 2.3952657699584963,
      "memory(GiB)": 77.56,
      "step": 60390,
      "token_acc": 0.5276872964169381,
      "train_speed(iter/s)": 1.438541
    },
    {
      "epoch": 2.5875069619982005,
      "grad_norm": 4.255127429962158,
      "learning_rate": 4.725227261724101e-05,
      "loss": 2.1614418029785156,
      "memory(GiB)": 77.56,
      "step": 60395,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.438546
    },
    {
      "epoch": 2.5877211773274498,
      "grad_norm": 7.340925693511963,
      "learning_rate": 4.7245553038752535e-05,
      "loss": 2.442327880859375,
      "memory(GiB)": 77.56,
      "step": 60400,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438533
    },
    {
      "epoch": 2.5879353926566986,
      "grad_norm": 5.364330768585205,
      "learning_rate": 4.7238833510163475e-05,
      "loss": 2.4522726058959963,
      "memory(GiB)": 77.56,
      "step": 60405,
      "token_acc": 0.47530864197530864,
      "train_speed(iter/s)": 1.438518
    },
    {
      "epoch": 2.5881496079859474,
      "grad_norm": 6.723532199859619,
      "learning_rate": 4.723211403159552e-05,
      "loss": 2.831590461730957,
      "memory(GiB)": 77.56,
      "step": 60410,
      "token_acc": 0.4307692307692308,
      "train_speed(iter/s)": 1.43851
    },
    {
      "epoch": 2.5883638233151967,
      "grad_norm": 4.7872843742370605,
      "learning_rate": 4.722539460317041e-05,
      "loss": 2.6963634490966797,
      "memory(GiB)": 77.56,
      "step": 60415,
      "token_acc": 0.4967532467532468,
      "train_speed(iter/s)": 1.43849
    },
    {
      "epoch": 2.5885780386444455,
      "grad_norm": 4.921646595001221,
      "learning_rate": 4.721867522500989e-05,
      "loss": 2.4428171157836913,
      "memory(GiB)": 77.56,
      "step": 60420,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.438497
    },
    {
      "epoch": 2.5887922539736943,
      "grad_norm": 8.170104026794434,
      "learning_rate": 4.721195589723565e-05,
      "loss": 2.6428401947021483,
      "memory(GiB)": 77.56,
      "step": 60425,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.438493
    },
    {
      "epoch": 2.5890064693029435,
      "grad_norm": 7.618386268615723,
      "learning_rate": 4.7205236619969474e-05,
      "loss": 2.7102468490600584,
      "memory(GiB)": 77.56,
      "step": 60430,
      "token_acc": 0.4307116104868914,
      "train_speed(iter/s)": 1.438479
    },
    {
      "epoch": 2.5892206846321923,
      "grad_norm": 4.510335922241211,
      "learning_rate": 4.719851739333305e-05,
      "loss": 2.538511848449707,
      "memory(GiB)": 77.56,
      "step": 60435,
      "token_acc": 0.4980392156862745,
      "train_speed(iter/s)": 1.438484
    },
    {
      "epoch": 2.589434899961441,
      "grad_norm": 4.9730377197265625,
      "learning_rate": 4.7191798217448115e-05,
      "loss": 2.6503627777099608,
      "memory(GiB)": 77.56,
      "step": 60440,
      "token_acc": 0.44807121661721067,
      "train_speed(iter/s)": 1.438453
    },
    {
      "epoch": 2.5896491152906904,
      "grad_norm": 5.63424015045166,
      "learning_rate": 4.718507909243638e-05,
      "loss": 2.681383514404297,
      "memory(GiB)": 77.56,
      "step": 60445,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.438464
    },
    {
      "epoch": 2.5898633306199392,
      "grad_norm": 5.386106014251709,
      "learning_rate": 4.7178360018419585e-05,
      "loss": 2.538774871826172,
      "memory(GiB)": 77.56,
      "step": 60450,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.43846
    },
    {
      "epoch": 2.590077545949188,
      "grad_norm": 8.20904541015625,
      "learning_rate": 4.717164099551945e-05,
      "loss": 2.440165710449219,
      "memory(GiB)": 77.56,
      "step": 60455,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.43847
    },
    {
      "epoch": 2.5902917612784373,
      "grad_norm": 6.0574727058410645,
      "learning_rate": 4.7164922023857686e-05,
      "loss": 2.4994409561157225,
      "memory(GiB)": 77.56,
      "step": 60460,
      "token_acc": 0.4938650306748466,
      "train_speed(iter/s)": 1.438502
    },
    {
      "epoch": 2.590505976607686,
      "grad_norm": 4.768818378448486,
      "learning_rate": 4.7158203103556026e-05,
      "loss": 2.383126640319824,
      "memory(GiB)": 77.56,
      "step": 60465,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.438526
    },
    {
      "epoch": 2.590720191936935,
      "grad_norm": 4.0426506996154785,
      "learning_rate": 4.715148423473618e-05,
      "loss": 2.438172149658203,
      "memory(GiB)": 77.56,
      "step": 60470,
      "token_acc": 0.5229357798165137,
      "train_speed(iter/s)": 1.438541
    },
    {
      "epoch": 2.590934407266184,
      "grad_norm": 5.394260406494141,
      "learning_rate": 4.714476541751986e-05,
      "loss": 2.3842021942138674,
      "memory(GiB)": 77.56,
      "step": 60475,
      "token_acc": 0.4312977099236641,
      "train_speed(iter/s)": 1.438557
    },
    {
      "epoch": 2.591148622595433,
      "grad_norm": 4.433017253875732,
      "learning_rate": 4.71380466520288e-05,
      "loss": 2.3090087890625,
      "memory(GiB)": 77.56,
      "step": 60480,
      "token_acc": 0.5215686274509804,
      "train_speed(iter/s)": 1.438586
    },
    {
      "epoch": 2.591362837924682,
      "grad_norm": 6.5087409019470215,
      "learning_rate": 4.7131327938384706e-05,
      "loss": 2.400865364074707,
      "memory(GiB)": 77.56,
      "step": 60485,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.438578
    },
    {
      "epoch": 2.591577053253931,
      "grad_norm": 6.493954181671143,
      "learning_rate": 4.71246092767093e-05,
      "loss": 2.2848079681396483,
      "memory(GiB)": 77.56,
      "step": 60490,
      "token_acc": 0.4959677419354839,
      "train_speed(iter/s)": 1.438588
    },
    {
      "epoch": 2.59179126858318,
      "grad_norm": 4.80222225189209,
      "learning_rate": 4.7117890667124306e-05,
      "loss": 2.416494941711426,
      "memory(GiB)": 77.56,
      "step": 60495,
      "token_acc": 0.46417445482866043,
      "train_speed(iter/s)": 1.438617
    },
    {
      "epoch": 2.5920054839124287,
      "grad_norm": 7.05763053894043,
      "learning_rate": 4.7111172109751394e-05,
      "loss": 2.438814163208008,
      "memory(GiB)": 77.56,
      "step": 60500,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.438613
    },
    {
      "epoch": 2.5920054839124287,
      "eval_loss": 2.1299960613250732,
      "eval_runtime": 14.3991,
      "eval_samples_per_second": 6.945,
      "eval_steps_per_second": 6.945,
      "eval_token_acc": 0.4797297297297297,
      "step": 60500
    },
    {
      "epoch": 2.592219699241678,
      "grad_norm": 7.089644432067871,
      "learning_rate": 4.7104453604712326e-05,
      "loss": 2.1959226608276365,
      "memory(GiB)": 77.56,
      "step": 60505,
      "token_acc": 0.4916911045943304,
      "train_speed(iter/s)": 1.438085
    },
    {
      "epoch": 2.5924339145709268,
      "grad_norm": 4.846078395843506,
      "learning_rate": 4.70977351521288e-05,
      "loss": 2.626007080078125,
      "memory(GiB)": 77.56,
      "step": 60510,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 2.5926481299001756,
      "grad_norm": 5.694186687469482,
      "learning_rate": 4.709101675212253e-05,
      "loss": 2.5185497283935545,
      "memory(GiB)": 77.56,
      "step": 60515,
      "token_acc": 0.41114982578397213,
      "train_speed(iter/s)": 1.43811
    },
    {
      "epoch": 2.592862345229425,
      "grad_norm": 5.390349864959717,
      "learning_rate": 4.7084298404815206e-05,
      "loss": 3.103775405883789,
      "memory(GiB)": 77.56,
      "step": 60520,
      "token_acc": 0.40069686411149824,
      "train_speed(iter/s)": 1.438115
    },
    {
      "epoch": 2.5930765605586736,
      "grad_norm": 6.129212856292725,
      "learning_rate": 4.7077580110328566e-05,
      "loss": 2.274595260620117,
      "memory(GiB)": 77.56,
      "step": 60525,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.438131
    },
    {
      "epoch": 2.5932907758879225,
      "grad_norm": 6.412266254425049,
      "learning_rate": 4.7070861868784296e-05,
      "loss": 2.5591609954833983,
      "memory(GiB)": 77.56,
      "step": 60530,
      "token_acc": 0.4503311258278146,
      "train_speed(iter/s)": 1.438152
    },
    {
      "epoch": 2.5935049912171717,
      "grad_norm": 5.946508407592773,
      "learning_rate": 4.706414368030412e-05,
      "loss": 2.6155929565429688,
      "memory(GiB)": 77.56,
      "step": 60535,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.438135
    },
    {
      "epoch": 2.5937192065464205,
      "grad_norm": 5.697239398956299,
      "learning_rate": 4.705742554500973e-05,
      "loss": 2.376565933227539,
      "memory(GiB)": 77.56,
      "step": 60540,
      "token_acc": 0.48641304347826086,
      "train_speed(iter/s)": 1.438134
    },
    {
      "epoch": 2.5939334218756693,
      "grad_norm": 5.533055782318115,
      "learning_rate": 4.705070746302283e-05,
      "loss": 2.2665977478027344,
      "memory(GiB)": 77.56,
      "step": 60545,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.438112
    },
    {
      "epoch": 2.5941476372049186,
      "grad_norm": 7.01714563369751,
      "learning_rate": 4.704398943446514e-05,
      "loss": 2.5848180770874025,
      "memory(GiB)": 77.56,
      "step": 60550,
      "token_acc": 0.4340836012861736,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 2.5943618525341674,
      "grad_norm": 6.462080478668213,
      "learning_rate": 4.703727145945836e-05,
      "loss": 2.3493200302124024,
      "memory(GiB)": 77.56,
      "step": 60555,
      "token_acc": 0.519650655021834,
      "train_speed(iter/s)": 1.438062
    },
    {
      "epoch": 2.594576067863416,
      "grad_norm": 4.913819313049316,
      "learning_rate": 4.7030553538124166e-05,
      "loss": 2.3986820220947265,
      "memory(GiB)": 77.56,
      "step": 60560,
      "token_acc": 0.4896142433234421,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 2.5947902831926655,
      "grad_norm": 5.035233974456787,
      "learning_rate": 4.70238356705843e-05,
      "loss": 2.5434823989868165,
      "memory(GiB)": 77.56,
      "step": 60565,
      "token_acc": 0.45555555555555555,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 2.5950044985219143,
      "grad_norm": 6.447807788848877,
      "learning_rate": 4.701711785696042e-05,
      "loss": 2.892926025390625,
      "memory(GiB)": 77.56,
      "step": 60570,
      "token_acc": 0.42201834862385323,
      "train_speed(iter/s)": 1.438027
    },
    {
      "epoch": 2.595218713851163,
      "grad_norm": 3.8857436180114746,
      "learning_rate": 4.7010400097374264e-05,
      "loss": 2.411061668395996,
      "memory(GiB)": 77.56,
      "step": 60575,
      "token_acc": 0.5045045045045045,
      "train_speed(iter/s)": 1.438026
    },
    {
      "epoch": 2.5954329291804124,
      "grad_norm": 5.358659744262695,
      "learning_rate": 4.7003682391947504e-05,
      "loss": 2.3542449951171873,
      "memory(GiB)": 77.56,
      "step": 60580,
      "token_acc": 0.4847457627118644,
      "train_speed(iter/s)": 1.43804
    },
    {
      "epoch": 2.595647144509661,
      "grad_norm": 8.36040210723877,
      "learning_rate": 4.699696474080186e-05,
      "loss": 2.6073287963867187,
      "memory(GiB)": 77.56,
      "step": 60585,
      "token_acc": 0.4503311258278146,
      "train_speed(iter/s)": 1.438053
    },
    {
      "epoch": 2.59586135983891,
      "grad_norm": 4.727888107299805,
      "learning_rate": 4.699024714405901e-05,
      "loss": 2.234549331665039,
      "memory(GiB)": 77.56,
      "step": 60590,
      "token_acc": 0.5177304964539007,
      "train_speed(iter/s)": 1.43802
    },
    {
      "epoch": 2.5960755751681592,
      "grad_norm": 4.970827102661133,
      "learning_rate": 4.698352960184067e-05,
      "loss": 2.5970645904541017,
      "memory(GiB)": 77.56,
      "step": 60595,
      "token_acc": 0.4429530201342282,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 2.596289790497408,
      "grad_norm": 5.5648722648620605,
      "learning_rate": 4.697681211426851e-05,
      "loss": 2.184271240234375,
      "memory(GiB)": 77.56,
      "step": 60600,
      "token_acc": 0.518796992481203,
      "train_speed(iter/s)": 1.438039
    },
    {
      "epoch": 2.596504005826657,
      "grad_norm": 6.2273054122924805,
      "learning_rate": 4.697009468146423e-05,
      "loss": 2.3865848541259767,
      "memory(GiB)": 77.56,
      "step": 60605,
      "token_acc": 0.5036764705882353,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 2.596718221155906,
      "grad_norm": 5.652369976043701,
      "learning_rate": 4.6963377303549546e-05,
      "loss": 2.4741628646850584,
      "memory(GiB)": 77.56,
      "step": 60610,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 2.596932436485155,
      "grad_norm": 4.808661937713623,
      "learning_rate": 4.6956659980646125e-05,
      "loss": 2.6206716537475585,
      "memory(GiB)": 77.56,
      "step": 60615,
      "token_acc": 0.4536741214057508,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 2.5971466518144037,
      "grad_norm": 5.790530681610107,
      "learning_rate": 4.6949942712875645e-05,
      "loss": 2.71263427734375,
      "memory(GiB)": 77.56,
      "step": 60620,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.438052
    },
    {
      "epoch": 2.597360867143653,
      "grad_norm": 8.878464698791504,
      "learning_rate": 4.6943225500359834e-05,
      "loss": 2.418523597717285,
      "memory(GiB)": 77.56,
      "step": 60625,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 2.597575082472902,
      "grad_norm": 6.090363025665283,
      "learning_rate": 4.693650834322034e-05,
      "loss": 2.2794145584106444,
      "memory(GiB)": 77.56,
      "step": 60630,
      "token_acc": 0.4790874524714829,
      "train_speed(iter/s)": 1.438025
    },
    {
      "epoch": 2.5977892978021506,
      "grad_norm": 4.402954578399658,
      "learning_rate": 4.6929791241578894e-05,
      "loss": 2.333298110961914,
      "memory(GiB)": 77.56,
      "step": 60635,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.43802
    },
    {
      "epoch": 2.5980035131314,
      "grad_norm": 5.152929782867432,
      "learning_rate": 4.6923074195557146e-05,
      "loss": 2.3224788665771485,
      "memory(GiB)": 77.56,
      "step": 60640,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 2.5982177284606487,
      "grad_norm": 5.916562557220459,
      "learning_rate": 4.691635720527679e-05,
      "loss": 2.704825592041016,
      "memory(GiB)": 77.56,
      "step": 60645,
      "token_acc": 0.445993031358885,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 2.5984319437898975,
      "grad_norm": 6.161414623260498,
      "learning_rate": 4.690964027085951e-05,
      "loss": 2.561298942565918,
      "memory(GiB)": 77.56,
      "step": 60650,
      "token_acc": 0.4707792207792208,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 2.5986461591191468,
      "grad_norm": 8.037162780761719,
      "learning_rate": 4.6902923392427014e-05,
      "loss": 2.4153738021850586,
      "memory(GiB)": 77.56,
      "step": 60655,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.43804
    },
    {
      "epoch": 2.5988603744483956,
      "grad_norm": 6.481609344482422,
      "learning_rate": 4.689620657010097e-05,
      "loss": 2.7959033966064455,
      "memory(GiB)": 77.56,
      "step": 60660,
      "token_acc": 0.43977591036414565,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 2.5990745897776444,
      "grad_norm": 6.540483474731445,
      "learning_rate": 4.688948980400304e-05,
      "loss": 2.541012191772461,
      "memory(GiB)": 77.56,
      "step": 60665,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 2.5992888051068936,
      "grad_norm": 5.331500053405762,
      "learning_rate": 4.688277309425494e-05,
      "loss": 2.5811416625976564,
      "memory(GiB)": 77.56,
      "step": 60670,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.438034
    },
    {
      "epoch": 2.5995030204361425,
      "grad_norm": 4.370889663696289,
      "learning_rate": 4.68760564409783e-05,
      "loss": 2.7431217193603517,
      "memory(GiB)": 77.56,
      "step": 60675,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.438058
    },
    {
      "epoch": 2.5997172357653913,
      "grad_norm": 5.083352565765381,
      "learning_rate": 4.686933984429485e-05,
      "loss": 2.5576873779296876,
      "memory(GiB)": 77.56,
      "step": 60680,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 2.5999314510946405,
      "grad_norm": 6.1396379470825195,
      "learning_rate": 4.686262330432625e-05,
      "loss": 2.4654163360595702,
      "memory(GiB)": 77.56,
      "step": 60685,
      "token_acc": 0.45185185185185184,
      "train_speed(iter/s)": 1.438041
    },
    {
      "epoch": 2.6001456664238893,
      "grad_norm": 5.844450950622559,
      "learning_rate": 4.685590682119415e-05,
      "loss": 2.307367134094238,
      "memory(GiB)": 77.56,
      "step": 60690,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.43806
    },
    {
      "epoch": 2.600359881753138,
      "grad_norm": 8.420538902282715,
      "learning_rate": 4.684919039502027e-05,
      "loss": 2.718204689025879,
      "memory(GiB)": 77.56,
      "step": 60695,
      "token_acc": 0.4573170731707317,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 2.6005740970823874,
      "grad_norm": 4.86892557144165,
      "learning_rate": 4.6842474025926255e-05,
      "loss": 2.6100723266601564,
      "memory(GiB)": 77.56,
      "step": 60700,
      "token_acc": 0.44223107569721115,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 2.600788312411636,
      "grad_norm": 4.864750385284424,
      "learning_rate": 4.683575771403377e-05,
      "loss": 2.171669578552246,
      "memory(GiB)": 77.56,
      "step": 60705,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 2.601002527740885,
      "grad_norm": 7.040539264678955,
      "learning_rate": 4.6829041459464525e-05,
      "loss": 2.4126054763793947,
      "memory(GiB)": 77.56,
      "step": 60710,
      "token_acc": 0.4632768361581921,
      "train_speed(iter/s)": 1.438036
    },
    {
      "epoch": 2.6012167430701343,
      "grad_norm": 7.751967906951904,
      "learning_rate": 4.6822325262340147e-05,
      "loss": 2.6023113250732424,
      "memory(GiB)": 77.56,
      "step": 60715,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 2.601430958399383,
      "grad_norm": 6.601855754852295,
      "learning_rate": 4.6815609122782336e-05,
      "loss": 2.6646862030029297,
      "memory(GiB)": 77.56,
      "step": 60720,
      "token_acc": 0.47653429602888087,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 2.601645173728632,
      "grad_norm": 9.612900733947754,
      "learning_rate": 4.680889304091275e-05,
      "loss": 2.2821449279785155,
      "memory(GiB)": 77.56,
      "step": 60725,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.438044
    },
    {
      "epoch": 2.601859389057881,
      "grad_norm": 4.928402900695801,
      "learning_rate": 4.680217701685306e-05,
      "loss": 2.264250373840332,
      "memory(GiB)": 77.56,
      "step": 60730,
      "token_acc": 0.5079872204472844,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 2.60207360438713,
      "grad_norm": 5.766315460205078,
      "learning_rate": 4.679546105072494e-05,
      "loss": 2.488662338256836,
      "memory(GiB)": 77.56,
      "step": 60735,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 2.602287819716379,
      "grad_norm": 5.843890190124512,
      "learning_rate": 4.678874514265005e-05,
      "loss": 2.6733932495117188,
      "memory(GiB)": 77.56,
      "step": 60740,
      "token_acc": 0.46647230320699706,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 2.602502035045628,
      "grad_norm": 4.9181342124938965,
      "learning_rate": 4.6782029292750056e-05,
      "loss": 2.436018180847168,
      "memory(GiB)": 77.56,
      "step": 60745,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 2.602716250374877,
      "grad_norm": 7.540224075317383,
      "learning_rate": 4.677531350114661e-05,
      "loss": 2.3421604156494142,
      "memory(GiB)": 77.56,
      "step": 60750,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 2.6029304657041257,
      "grad_norm": 5.905830383300781,
      "learning_rate": 4.6768597767961395e-05,
      "loss": 2.4692893981933595,
      "memory(GiB)": 77.56,
      "step": 60755,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 2.603144681033375,
      "grad_norm": 5.923783302307129,
      "learning_rate": 4.676188209331606e-05,
      "loss": 2.3923463821411133,
      "memory(GiB)": 77.56,
      "step": 60760,
      "token_acc": 0.5030674846625767,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 2.6033588963626237,
      "grad_norm": 5.391059398651123,
      "learning_rate": 4.675516647733226e-05,
      "loss": 2.248657989501953,
      "memory(GiB)": 77.56,
      "step": 60765,
      "token_acc": 0.501628664495114,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 2.6035731116918726,
      "grad_norm": 5.573025226593018,
      "learning_rate": 4.674845092013168e-05,
      "loss": 2.987510871887207,
      "memory(GiB)": 77.56,
      "step": 60770,
      "token_acc": 0.3877551020408163,
      "train_speed(iter/s)": 1.437982
    },
    {
      "epoch": 2.603787327021122,
      "grad_norm": 7.309691905975342,
      "learning_rate": 4.674173542183594e-05,
      "loss": 2.3911584854125976,
      "memory(GiB)": 77.56,
      "step": 60775,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 2.6040015423503706,
      "grad_norm": 4.779590606689453,
      "learning_rate": 4.673501998256674e-05,
      "loss": 2.313121795654297,
      "memory(GiB)": 77.56,
      "step": 60780,
      "token_acc": 0.501577287066246,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 2.6042157576796194,
      "grad_norm": 5.138645648956299,
      "learning_rate": 4.672830460244571e-05,
      "loss": 2.380880355834961,
      "memory(GiB)": 77.56,
      "step": 60785,
      "token_acc": 0.46963562753036436,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 2.6044299730088687,
      "grad_norm": 5.897324562072754,
      "learning_rate": 4.6721589281594504e-05,
      "loss": 2.8266084671020506,
      "memory(GiB)": 77.56,
      "step": 60790,
      "token_acc": 0.4201954397394137,
      "train_speed(iter/s)": 1.437967
    },
    {
      "epoch": 2.6046441883381175,
      "grad_norm": 5.31672477722168,
      "learning_rate": 4.671487402013477e-05,
      "loss": 2.5437637329101563,
      "memory(GiB)": 77.56,
      "step": 60795,
      "token_acc": 0.49201277955271566,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 2.6048584036673663,
      "grad_norm": 4.853329658508301,
      "learning_rate": 4.6708158818188195e-05,
      "loss": 2.4282676696777346,
      "memory(GiB)": 77.56,
      "step": 60800,
      "token_acc": 0.431438127090301,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 2.6050726189966156,
      "grad_norm": 6.074598789215088,
      "learning_rate": 4.670144367587642e-05,
      "loss": 2.583574104309082,
      "memory(GiB)": 77.56,
      "step": 60805,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 2.6052868343258644,
      "grad_norm": 7.673786640167236,
      "learning_rate": 4.669472859332106e-05,
      "loss": 2.4890289306640625,
      "memory(GiB)": 77.56,
      "step": 60810,
      "token_acc": 0.45374449339207046,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 2.605501049655113,
      "grad_norm": 5.606259822845459,
      "learning_rate": 4.668801357064382e-05,
      "loss": 2.273537826538086,
      "memory(GiB)": 77.56,
      "step": 60815,
      "token_acc": 0.5149501661129569,
      "train_speed(iter/s)": 1.438007
    },
    {
      "epoch": 2.6057152649843625,
      "grad_norm": 7.482728481292725,
      "learning_rate": 4.66812986079663e-05,
      "loss": 2.5690435409545898,
      "memory(GiB)": 77.56,
      "step": 60820,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 2.6059294803136113,
      "grad_norm": 5.658421039581299,
      "learning_rate": 4.667458370541018e-05,
      "loss": 2.479115104675293,
      "memory(GiB)": 77.56,
      "step": 60825,
      "token_acc": 0.43243243243243246,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 2.60614369564286,
      "grad_norm": 6.401666164398193,
      "learning_rate": 4.6667868863097096e-05,
      "loss": 2.646041488647461,
      "memory(GiB)": 77.56,
      "step": 60830,
      "token_acc": 0.456973293768546,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 2.6063579109721093,
      "grad_norm": 6.190518379211426,
      "learning_rate": 4.666115408114869e-05,
      "loss": 2.1837493896484377,
      "memory(GiB)": 77.56,
      "step": 60835,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 2.606572126301358,
      "grad_norm": 5.382422924041748,
      "learning_rate": 4.665443935968662e-05,
      "loss": 2.594824028015137,
      "memory(GiB)": 77.56,
      "step": 60840,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.437964
    },
    {
      "epoch": 2.606786341630607,
      "grad_norm": 5.528724193572998,
      "learning_rate": 4.664772469883251e-05,
      "loss": 2.204884147644043,
      "memory(GiB)": 77.56,
      "step": 60845,
      "token_acc": 0.5283018867924528,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 2.607000556959856,
      "grad_norm": 5.810295581817627,
      "learning_rate": 4.6641010098708006e-05,
      "loss": 2.2973960876464843,
      "memory(GiB)": 77.56,
      "step": 60850,
      "token_acc": 0.5223880597014925,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 2.607214772289105,
      "grad_norm": 6.444004535675049,
      "learning_rate": 4.6634295559434756e-05,
      "loss": 2.540684127807617,
      "memory(GiB)": 77.56,
      "step": 60855,
      "token_acc": 0.44405594405594406,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 2.607428987618354,
      "grad_norm": 5.819235324859619,
      "learning_rate": 4.66275810811344e-05,
      "loss": 2.5616165161132813,
      "memory(GiB)": 77.56,
      "step": 60860,
      "token_acc": 0.43333333333333335,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 2.607643202947603,
      "grad_norm": 8.05027961730957,
      "learning_rate": 4.662086666392859e-05,
      "loss": 2.5462581634521486,
      "memory(GiB)": 77.56,
      "step": 60865,
      "token_acc": 0.4843205574912892,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 2.607857418276852,
      "grad_norm": 4.803243637084961,
      "learning_rate": 4.661415230793892e-05,
      "loss": 2.3517572402954103,
      "memory(GiB)": 77.56,
      "step": 60870,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 2.6080716336061007,
      "grad_norm": 5.180236339569092,
      "learning_rate": 4.660743801328709e-05,
      "loss": 2.2150434494018554,
      "memory(GiB)": 77.56,
      "step": 60875,
      "token_acc": 0.5358649789029536,
      "train_speed(iter/s)": 1.438026
    },
    {
      "epoch": 2.60828584893535,
      "grad_norm": 5.502531051635742,
      "learning_rate": 4.6600723780094677e-05,
      "loss": 2.5581836700439453,
      "memory(GiB)": 77.56,
      "step": 60880,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 2.608500064264599,
      "grad_norm": 4.951906204223633,
      "learning_rate": 4.659400960848336e-05,
      "loss": 2.3677539825439453,
      "memory(GiB)": 77.56,
      "step": 60885,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 2.6087142795938476,
      "grad_norm": 6.366088390350342,
      "learning_rate": 4.658729549857476e-05,
      "loss": 2.3809818267822265,
      "memory(GiB)": 77.56,
      "step": 60890,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 2.608928494923097,
      "grad_norm": 6.7157979011535645,
      "learning_rate": 4.658058145049048e-05,
      "loss": 2.658820152282715,
      "memory(GiB)": 77.56,
      "step": 60895,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 2.6091427102523457,
      "grad_norm": 4.269170761108398,
      "learning_rate": 4.65738674643522e-05,
      "loss": 2.461532211303711,
      "memory(GiB)": 77.56,
      "step": 60900,
      "token_acc": 0.48881789137380194,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 2.6093569255815945,
      "grad_norm": 6.531954288482666,
      "learning_rate": 4.656715354028151e-05,
      "loss": 2.31856689453125,
      "memory(GiB)": 77.56,
      "step": 60905,
      "token_acc": 0.5338345864661654,
      "train_speed(iter/s)": 1.438007
    },
    {
      "epoch": 2.6095711409108437,
      "grad_norm": 7.257207870483398,
      "learning_rate": 4.656043967840005e-05,
      "loss": 2.3338733673095704,
      "memory(GiB)": 77.56,
      "step": 60910,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 2.6097853562400926,
      "grad_norm": 5.517190933227539,
      "learning_rate": 4.655372587882946e-05,
      "loss": 2.529966354370117,
      "memory(GiB)": 77.56,
      "step": 60915,
      "token_acc": 0.4830508474576271,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 2.6099995715693414,
      "grad_norm": 5.898299694061279,
      "learning_rate": 4.654701214169135e-05,
      "loss": 2.1048030853271484,
      "memory(GiB)": 77.56,
      "step": 60920,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 2.6102137868985906,
      "grad_norm": 7.8402276039123535,
      "learning_rate": 4.654029846710737e-05,
      "loss": 2.2793750762939453,
      "memory(GiB)": 77.56,
      "step": 60925,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 2.6104280022278394,
      "grad_norm": 5.904280185699463,
      "learning_rate": 4.653358485519912e-05,
      "loss": 2.3322324752807617,
      "memory(GiB)": 77.56,
      "step": 60930,
      "token_acc": 0.5445205479452054,
      "train_speed(iter/s)": 1.438046
    },
    {
      "epoch": 2.6106422175570883,
      "grad_norm": 4.554354667663574,
      "learning_rate": 4.652687130608823e-05,
      "loss": 2.506477928161621,
      "memory(GiB)": 77.56,
      "step": 60935,
      "token_acc": 0.46005509641873277,
      "train_speed(iter/s)": 1.43807
    },
    {
      "epoch": 2.6108564328863375,
      "grad_norm": 6.028464317321777,
      "learning_rate": 4.652015781989631e-05,
      "loss": 2.6493663787841797,
      "memory(GiB)": 77.56,
      "step": 60940,
      "token_acc": 0.42124542124542125,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 2.6110706482155863,
      "grad_norm": 5.519216537475586,
      "learning_rate": 4.651344439674501e-05,
      "loss": 2.3362178802490234,
      "memory(GiB)": 77.56,
      "step": 60945,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 2.611284863544835,
      "grad_norm": 6.624106407165527,
      "learning_rate": 4.650673103675594e-05,
      "loss": 2.3710700988769533,
      "memory(GiB)": 77.56,
      "step": 60950,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 2.6114990788740844,
      "grad_norm": 4.884666919708252,
      "learning_rate": 4.6500017740050706e-05,
      "loss": 2.240732192993164,
      "memory(GiB)": 77.56,
      "step": 60955,
      "token_acc": 0.5255972696245734,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 2.611713294203333,
      "grad_norm": 4.90848970413208,
      "learning_rate": 4.649330450675095e-05,
      "loss": 2.2751197814941406,
      "memory(GiB)": 77.56,
      "step": 60960,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.438121
    },
    {
      "epoch": 2.611927509532582,
      "grad_norm": 5.464402675628662,
      "learning_rate": 4.6486591336978246e-05,
      "loss": 2.6248870849609376,
      "memory(GiB)": 77.56,
      "step": 60965,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.438134
    },
    {
      "epoch": 2.6121417248618313,
      "grad_norm": 5.042157173156738,
      "learning_rate": 4.647987823085426e-05,
      "loss": 2.4178579330444334,
      "memory(GiB)": 77.56,
      "step": 60970,
      "token_acc": 0.4915254237288136,
      "train_speed(iter/s)": 1.438131
    },
    {
      "epoch": 2.61235594019108,
      "grad_norm": 5.9759039878845215,
      "learning_rate": 4.647316518850058e-05,
      "loss": 2.470847320556641,
      "memory(GiB)": 77.56,
      "step": 60975,
      "token_acc": 0.4731182795698925,
      "train_speed(iter/s)": 1.438122
    },
    {
      "epoch": 2.612570155520329,
      "grad_norm": 5.451035499572754,
      "learning_rate": 4.64664522100388e-05,
      "loss": 2.3520626068115233,
      "memory(GiB)": 77.56,
      "step": 60980,
      "token_acc": 0.511400651465798,
      "train_speed(iter/s)": 1.438139
    },
    {
      "epoch": 2.612784370849578,
      "grad_norm": 5.501289367675781,
      "learning_rate": 4.645973929559058e-05,
      "loss": 2.528302001953125,
      "memory(GiB)": 77.56,
      "step": 60985,
      "token_acc": 0.45733788395904434,
      "train_speed(iter/s)": 1.438157
    },
    {
      "epoch": 2.612998586178827,
      "grad_norm": 5.421691417694092,
      "learning_rate": 4.645302644527749e-05,
      "loss": 2.1132137298583986,
      "memory(GiB)": 77.56,
      "step": 60990,
      "token_acc": 0.541958041958042,
      "train_speed(iter/s)": 1.438178
    },
    {
      "epoch": 2.6132128015080758,
      "grad_norm": 5.147592544555664,
      "learning_rate": 4.644631365922114e-05,
      "loss": 2.375047492980957,
      "memory(GiB)": 77.56,
      "step": 60995,
      "token_acc": 0.47337278106508873,
      "train_speed(iter/s)": 1.438181
    },
    {
      "epoch": 2.613427016837325,
      "grad_norm": 5.489688873291016,
      "learning_rate": 4.643960093754318e-05,
      "loss": 2.533391571044922,
      "memory(GiB)": 77.56,
      "step": 61000,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43819
    },
    {
      "epoch": 2.613427016837325,
      "eval_loss": 2.2321438789367676,
      "eval_runtime": 14.7761,
      "eval_samples_per_second": 6.768,
      "eval_steps_per_second": 6.768,
      "eval_token_acc": 0.47619047619047616,
      "step": 61000
    },
    {
      "epoch": 2.613641232166574,
      "grad_norm": 6.466928005218506,
      "learning_rate": 4.643288828036517e-05,
      "loss": 2.4165807723999024,
      "memory(GiB)": 77.56,
      "step": 61005,
      "token_acc": 0.48466257668711654,
      "train_speed(iter/s)": 1.437649
    },
    {
      "epoch": 2.6138554474958227,
      "grad_norm": 6.679676532745361,
      "learning_rate": 4.642617568780874e-05,
      "loss": 2.5052881240844727,
      "memory(GiB)": 77.56,
      "step": 61010,
      "token_acc": 0.5064377682403434,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 2.614069662825072,
      "grad_norm": 5.5163655281066895,
      "learning_rate": 4.6419463159995474e-05,
      "loss": 2.6603378295898437,
      "memory(GiB)": 77.56,
      "step": 61015,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 2.6142838781543207,
      "grad_norm": 5.466887950897217,
      "learning_rate": 4.6412750697047014e-05,
      "loss": 2.2969329833984373,
      "memory(GiB)": 77.56,
      "step": 61020,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 2.6144980934835695,
      "grad_norm": 5.1457014083862305,
      "learning_rate": 4.640603829908493e-05,
      "loss": 2.3360012054443358,
      "memory(GiB)": 77.56,
      "step": 61025,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.437622
    },
    {
      "epoch": 2.614712308812819,
      "grad_norm": 4.956437110900879,
      "learning_rate": 4.6399325966230835e-05,
      "loss": 2.8091278076171875,
      "memory(GiB)": 77.56,
      "step": 61030,
      "token_acc": 0.4077669902912621,
      "train_speed(iter/s)": 1.43761
    },
    {
      "epoch": 2.6149265241420676,
      "grad_norm": 4.8004279136657715,
      "learning_rate": 4.6392613698606336e-05,
      "loss": 2.5653026580810545,
      "memory(GiB)": 77.56,
      "step": 61035,
      "token_acc": 0.45864661654135336,
      "train_speed(iter/s)": 1.437584
    },
    {
      "epoch": 2.6151407394713164,
      "grad_norm": 6.565741062164307,
      "learning_rate": 4.638590149633301e-05,
      "loss": 2.6528976440429686,
      "memory(GiB)": 77.56,
      "step": 61040,
      "token_acc": 0.4456140350877193,
      "train_speed(iter/s)": 1.437572
    },
    {
      "epoch": 2.6153549548005657,
      "grad_norm": 4.783234119415283,
      "learning_rate": 4.637918935953248e-05,
      "loss": 2.322017860412598,
      "memory(GiB)": 77.56,
      "step": 61045,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.437588
    },
    {
      "epoch": 2.6155691701298145,
      "grad_norm": 3.8038084506988525,
      "learning_rate": 4.637247728832633e-05,
      "loss": 2.1964120864868164,
      "memory(GiB)": 77.56,
      "step": 61050,
      "token_acc": 0.5487364620938628,
      "train_speed(iter/s)": 1.437616
    },
    {
      "epoch": 2.6157833854590633,
      "grad_norm": 6.103489875793457,
      "learning_rate": 4.636576528283615e-05,
      "loss": 2.6924789428710936,
      "memory(GiB)": 77.56,
      "step": 61055,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.43761
    },
    {
      "epoch": 2.6159976007883126,
      "grad_norm": 7.4280595779418945,
      "learning_rate": 4.635905334318355e-05,
      "loss": 2.347287178039551,
      "memory(GiB)": 77.56,
      "step": 61060,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.437644
    },
    {
      "epoch": 2.6162118161175614,
      "grad_norm": 4.4463396072387695,
      "learning_rate": 4.635234146949011e-05,
      "loss": 2.0726484298706054,
      "memory(GiB)": 77.56,
      "step": 61065,
      "token_acc": 0.5149253731343284,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 2.61642603144681,
      "grad_norm": 7.577883243560791,
      "learning_rate": 4.6345629661877425e-05,
      "loss": 2.6312582015991213,
      "memory(GiB)": 77.56,
      "step": 61070,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 2.6166402467760594,
      "grad_norm": 5.872193813323975,
      "learning_rate": 4.6338917920467096e-05,
      "loss": 2.2583131790161133,
      "memory(GiB)": 77.56,
      "step": 61075,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.437601
    },
    {
      "epoch": 2.6168544621053083,
      "grad_norm": 5.659163951873779,
      "learning_rate": 4.633220624538069e-05,
      "loss": 2.408833122253418,
      "memory(GiB)": 77.56,
      "step": 61080,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 2.617068677434557,
      "grad_norm": 6.097976207733154,
      "learning_rate": 4.632549463673981e-05,
      "loss": 2.4263397216796876,
      "memory(GiB)": 77.56,
      "step": 61085,
      "token_acc": 0.531496062992126,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 2.6172828927638063,
      "grad_norm": 5.783149242401123,
      "learning_rate": 4.631878309466603e-05,
      "loss": 2.5575687408447267,
      "memory(GiB)": 77.56,
      "step": 61090,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.437671
    },
    {
      "epoch": 2.617497108093055,
      "grad_norm": 5.717090606689453,
      "learning_rate": 4.6312071619280974e-05,
      "loss": 2.3968442916870116,
      "memory(GiB)": 77.56,
      "step": 61095,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 2.617711323422304,
      "grad_norm": 4.587832927703857,
      "learning_rate": 4.630536021070618e-05,
      "loss": 2.7261463165283204,
      "memory(GiB)": 77.56,
      "step": 61100,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.617925538751553,
      "grad_norm": 5.032992839813232,
      "learning_rate": 4.629864886906326e-05,
      "loss": 2.287972640991211,
      "memory(GiB)": 77.56,
      "step": 61105,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.437696
    },
    {
      "epoch": 2.618139754080802,
      "grad_norm": 5.303487300872803,
      "learning_rate": 4.629193759447378e-05,
      "loss": 2.5813934326171877,
      "memory(GiB)": 77.56,
      "step": 61110,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 2.618353969410051,
      "grad_norm": 3.700289249420166,
      "learning_rate": 4.628522638705934e-05,
      "loss": 2.2632728576660157,
      "memory(GiB)": 77.56,
      "step": 61115,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.437677
    },
    {
      "epoch": 2.6185681847393,
      "grad_norm": 6.38515567779541,
      "learning_rate": 4.6278515246941514e-05,
      "loss": 2.355577850341797,
      "memory(GiB)": 77.56,
      "step": 61120,
      "token_acc": 0.5176848874598071,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 2.618782400068549,
      "grad_norm": 4.75387716293335,
      "learning_rate": 4.6271804174241864e-05,
      "loss": 2.401355743408203,
      "memory(GiB)": 77.56,
      "step": 61125,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.437676
    },
    {
      "epoch": 2.6189966153977977,
      "grad_norm": 4.666396617889404,
      "learning_rate": 4.6265093169081986e-05,
      "loss": 2.46710205078125,
      "memory(GiB)": 77.56,
      "step": 61130,
      "token_acc": 0.45962732919254656,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 2.619210830727047,
      "grad_norm": 5.302369594573975,
      "learning_rate": 4.625838223158345e-05,
      "loss": 2.417184829711914,
      "memory(GiB)": 77.56,
      "step": 61135,
      "token_acc": 0.4597315436241611,
      "train_speed(iter/s)": 1.437665
    },
    {
      "epoch": 2.619425046056296,
      "grad_norm": 8.818601608276367,
      "learning_rate": 4.6251671361867813e-05,
      "loss": 2.4843074798583986,
      "memory(GiB)": 77.56,
      "step": 61140,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 2.6196392613855446,
      "grad_norm": 5.414271354675293,
      "learning_rate": 4.6244960560056686e-05,
      "loss": 2.4678627014160157,
      "memory(GiB)": 77.56,
      "step": 61145,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.437687
    },
    {
      "epoch": 2.619853476714794,
      "grad_norm": 5.477501392364502,
      "learning_rate": 4.623824982627161e-05,
      "loss": 2.617709159851074,
      "memory(GiB)": 77.56,
      "step": 61150,
      "token_acc": 0.4742268041237113,
      "train_speed(iter/s)": 1.437682
    },
    {
      "epoch": 2.6200676920440427,
      "grad_norm": 4.945050239562988,
      "learning_rate": 4.623153916063417e-05,
      "loss": 2.4587818145751954,
      "memory(GiB)": 77.56,
      "step": 61155,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437671
    },
    {
      "epoch": 2.6202819073732915,
      "grad_norm": 5.390034198760986,
      "learning_rate": 4.6224828563265934e-05,
      "loss": 2.411081886291504,
      "memory(GiB)": 77.56,
      "step": 61160,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 2.6204961227025407,
      "grad_norm": 4.585055351257324,
      "learning_rate": 4.6218118034288456e-05,
      "loss": 2.637013053894043,
      "memory(GiB)": 77.56,
      "step": 61165,
      "token_acc": 0.49557522123893805,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 2.6207103380317895,
      "grad_norm": 5.090357780456543,
      "learning_rate": 4.6211407573823326e-05,
      "loss": 2.5173492431640625,
      "memory(GiB)": 77.56,
      "step": 61170,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.437702
    },
    {
      "epoch": 2.6209245533610384,
      "grad_norm": 9.149224281311035,
      "learning_rate": 4.620469718199211e-05,
      "loss": 2.3824291229248047,
      "memory(GiB)": 77.56,
      "step": 61175,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437726
    },
    {
      "epoch": 2.6211387686902876,
      "grad_norm": 6.559986591339111,
      "learning_rate": 4.619798685891637e-05,
      "loss": 2.543759346008301,
      "memory(GiB)": 77.56,
      "step": 61180,
      "token_acc": 0.4699248120300752,
      "train_speed(iter/s)": 1.43775
    },
    {
      "epoch": 2.6213529840195364,
      "grad_norm": 5.382274150848389,
      "learning_rate": 4.6191276604717645e-05,
      "loss": 2.148107719421387,
      "memory(GiB)": 77.56,
      "step": 61185,
      "token_acc": 0.5447761194029851,
      "train_speed(iter/s)": 1.437737
    },
    {
      "epoch": 2.6215671993487852,
      "grad_norm": 5.556448459625244,
      "learning_rate": 4.618456641951753e-05,
      "loss": 2.2009441375732424,
      "memory(GiB)": 77.56,
      "step": 61190,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 2.6217814146780345,
      "grad_norm": 6.283977031707764,
      "learning_rate": 4.617785630343757e-05,
      "loss": 2.6974502563476563,
      "memory(GiB)": 77.56,
      "step": 61195,
      "token_acc": 0.48615384615384616,
      "train_speed(iter/s)": 1.437769
    },
    {
      "epoch": 2.6219956300072833,
      "grad_norm": 4.915924072265625,
      "learning_rate": 4.617114625659932e-05,
      "loss": 2.227311134338379,
      "memory(GiB)": 77.56,
      "step": 61200,
      "token_acc": 0.5142045454545454,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 2.622209845336532,
      "grad_norm": 5.53706693649292,
      "learning_rate": 4.6164436279124366e-05,
      "loss": 2.6401288986206053,
      "memory(GiB)": 77.56,
      "step": 61205,
      "token_acc": 0.4828897338403042,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 2.6224240606657814,
      "grad_norm": 5.931461334228516,
      "learning_rate": 4.615772637113423e-05,
      "loss": 2.558811378479004,
      "memory(GiB)": 77.56,
      "step": 61210,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 2.62263827599503,
      "grad_norm": 5.994172096252441,
      "learning_rate": 4.61510165327505e-05,
      "loss": 2.4454769134521483,
      "memory(GiB)": 77.56,
      "step": 61215,
      "token_acc": 0.4911660777385159,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 2.622852491324279,
      "grad_norm": 6.3840155601501465,
      "learning_rate": 4.6144306764094714e-05,
      "loss": 2.450403594970703,
      "memory(GiB)": 77.56,
      "step": 61220,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.437773
    },
    {
      "epoch": 2.6230667066535283,
      "grad_norm": 4.7752885818481445,
      "learning_rate": 4.613759706528842e-05,
      "loss": 2.4518911361694338,
      "memory(GiB)": 77.56,
      "step": 61225,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 2.623280921982777,
      "grad_norm": 4.849348068237305,
      "learning_rate": 4.613088743645318e-05,
      "loss": 2.8188608169555662,
      "memory(GiB)": 77.56,
      "step": 61230,
      "token_acc": 0.4558011049723757,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 2.623495137312026,
      "grad_norm": 5.336507797241211,
      "learning_rate": 4.612417787771055e-05,
      "loss": 2.0962846755981444,
      "memory(GiB)": 77.56,
      "step": 61235,
      "token_acc": 0.5393700787401575,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 2.623709352641275,
      "grad_norm": 5.035750865936279,
      "learning_rate": 4.6117468389182053e-05,
      "loss": 2.576386260986328,
      "memory(GiB)": 77.56,
      "step": 61240,
      "token_acc": 0.45907473309608543,
      "train_speed(iter/s)": 1.437812
    },
    {
      "epoch": 2.623923567970524,
      "grad_norm": 6.402904510498047,
      "learning_rate": 4.6110758970989264e-05,
      "loss": 2.5084495544433594,
      "memory(GiB)": 77.56,
      "step": 61245,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.437814
    },
    {
      "epoch": 2.6241377832997728,
      "grad_norm": 5.828314304351807,
      "learning_rate": 4.610404962325374e-05,
      "loss": 2.374681282043457,
      "memory(GiB)": 77.56,
      "step": 61250,
      "token_acc": 0.5434782608695652,
      "train_speed(iter/s)": 1.437797
    },
    {
      "epoch": 2.624351998629022,
      "grad_norm": 4.399845123291016,
      "learning_rate": 4.6097340346097004e-05,
      "loss": 2.3787761688232423,
      "memory(GiB)": 77.56,
      "step": 61255,
      "token_acc": 0.47075208913649025,
      "train_speed(iter/s)": 1.4378
    },
    {
      "epoch": 2.624566213958271,
      "grad_norm": 5.238890647888184,
      "learning_rate": 4.609063113964061e-05,
      "loss": 2.248393249511719,
      "memory(GiB)": 77.56,
      "step": 61260,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 2.6247804292875196,
      "grad_norm": 6.462149620056152,
      "learning_rate": 4.608392200400611e-05,
      "loss": 2.5526473999023436,
      "memory(GiB)": 77.56,
      "step": 61265,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.437803
    },
    {
      "epoch": 2.624994644616769,
      "grad_norm": 6.404357433319092,
      "learning_rate": 4.607721293931502e-05,
      "loss": 2.203322982788086,
      "memory(GiB)": 77.56,
      "step": 61270,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.437803
    },
    {
      "epoch": 2.6252088599460177,
      "grad_norm": 8.24066162109375,
      "learning_rate": 4.607050394568891e-05,
      "loss": 2.4650421142578125,
      "memory(GiB)": 77.56,
      "step": 61275,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 2.6254230752752665,
      "grad_norm": 7.238137245178223,
      "learning_rate": 4.6063795023249303e-05,
      "loss": 2.476858901977539,
      "memory(GiB)": 77.56,
      "step": 61280,
      "token_acc": 0.4908424908424908,
      "train_speed(iter/s)": 1.437812
    },
    {
      "epoch": 2.625637290604516,
      "grad_norm": 5.264199733734131,
      "learning_rate": 4.605708617211774e-05,
      "loss": 2.3938180923461916,
      "memory(GiB)": 77.56,
      "step": 61285,
      "token_acc": 0.5067114093959731,
      "train_speed(iter/s)": 1.437815
    },
    {
      "epoch": 2.6258515059337646,
      "grad_norm": 5.499978542327881,
      "learning_rate": 4.605037739241576e-05,
      "loss": 2.7696823120117187,
      "memory(GiB)": 77.56,
      "step": 61290,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.437825
    },
    {
      "epoch": 2.6260657212630134,
      "grad_norm": 5.290322780609131,
      "learning_rate": 4.604366868426489e-05,
      "loss": 2.4920318603515623,
      "memory(GiB)": 77.56,
      "step": 61295,
      "token_acc": 0.4851190476190476,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 2.6262799365922627,
      "grad_norm": 6.1877522468566895,
      "learning_rate": 4.603696004778669e-05,
      "loss": 2.492643928527832,
      "memory(GiB)": 77.56,
      "step": 61300,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 2.6264941519215115,
      "grad_norm": 5.980542182922363,
      "learning_rate": 4.603025148310267e-05,
      "loss": 2.43652286529541,
      "memory(GiB)": 77.56,
      "step": 61305,
      "token_acc": 0.5057034220532319,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 2.6267083672507603,
      "grad_norm": 6.2820563316345215,
      "learning_rate": 4.602354299033435e-05,
      "loss": 2.477032470703125,
      "memory(GiB)": 77.56,
      "step": 61310,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 2.6269225825800095,
      "grad_norm": 5.314938545227051,
      "learning_rate": 4.601683456960329e-05,
      "loss": 2.8025466918945314,
      "memory(GiB)": 77.56,
      "step": 61315,
      "token_acc": 0.43820224719101125,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 2.6271367979092584,
      "grad_norm": 5.161302089691162,
      "learning_rate": 4.601012622103102e-05,
      "loss": 2.4552543640136717,
      "memory(GiB)": 77.56,
      "step": 61320,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 2.627351013238507,
      "grad_norm": 5.536771297454834,
      "learning_rate": 4.600341794473905e-05,
      "loss": 2.3944297790527345,
      "memory(GiB)": 77.56,
      "step": 61325,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.437904
    },
    {
      "epoch": 2.6275652285677564,
      "grad_norm": 4.604583263397217,
      "learning_rate": 4.599670974084891e-05,
      "loss": 2.487743377685547,
      "memory(GiB)": 77.56,
      "step": 61330,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.43791
    },
    {
      "epoch": 2.6277794438970052,
      "grad_norm": 5.908392429351807,
      "learning_rate": 4.599000160948214e-05,
      "loss": 2.3267688751220703,
      "memory(GiB)": 77.56,
      "step": 61335,
      "token_acc": 0.5533333333333333,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 2.627993659226254,
      "grad_norm": 5.070243835449219,
      "learning_rate": 4.598329355076024e-05,
      "loss": 2.486347961425781,
      "memory(GiB)": 77.56,
      "step": 61340,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 2.6282078745555033,
      "grad_norm": 6.921378135681152,
      "learning_rate": 4.597658556480475e-05,
      "loss": 2.50708065032959,
      "memory(GiB)": 77.56,
      "step": 61345,
      "token_acc": 0.460431654676259,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 2.628422089884752,
      "grad_norm": 7.598796844482422,
      "learning_rate": 4.5969877651737194e-05,
      "loss": 2.3890256881713867,
      "memory(GiB)": 77.56,
      "step": 61350,
      "token_acc": 0.5117370892018779,
      "train_speed(iter/s)": 1.43794
    },
    {
      "epoch": 2.628636305214001,
      "grad_norm": 5.354578971862793,
      "learning_rate": 4.596316981167906e-05,
      "loss": 2.3807161331176756,
      "memory(GiB)": 77.56,
      "step": 61355,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 2.62885052054325,
      "grad_norm": 7.167765140533447,
      "learning_rate": 4.595646204475192e-05,
      "loss": 2.4626785278320313,
      "memory(GiB)": 77.56,
      "step": 61360,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 2.629064735872499,
      "grad_norm": 5.551894187927246,
      "learning_rate": 4.594975435107726e-05,
      "loss": 2.3933115005493164,
      "memory(GiB)": 77.56,
      "step": 61365,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.437942
    },
    {
      "epoch": 2.629278951201748,
      "grad_norm": 6.5892205238342285,
      "learning_rate": 4.594304673077658e-05,
      "loss": 2.3925952911376953,
      "memory(GiB)": 77.56,
      "step": 61370,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.437949
    },
    {
      "epoch": 2.629493166530997,
      "grad_norm": 5.637641429901123,
      "learning_rate": 4.593633918397143e-05,
      "loss": 2.6355018615722656,
      "memory(GiB)": 77.56,
      "step": 61375,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 2.629707381860246,
      "grad_norm": 5.442171096801758,
      "learning_rate": 4.592963171078331e-05,
      "loss": 2.4948026657104494,
      "memory(GiB)": 77.56,
      "step": 61380,
      "token_acc": 0.4707792207792208,
      "train_speed(iter/s)": 1.437973
    },
    {
      "epoch": 2.6299215971894947,
      "grad_norm": 4.8702263832092285,
      "learning_rate": 4.59229243113337e-05,
      "loss": 2.388945960998535,
      "memory(GiB)": 77.56,
      "step": 61385,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.437987
    },
    {
      "epoch": 2.630135812518744,
      "grad_norm": 5.89458703994751,
      "learning_rate": 4.5916216985744164e-05,
      "loss": 2.391023063659668,
      "memory(GiB)": 77.56,
      "step": 61390,
      "token_acc": 0.4810126582278481,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 2.6303500278479928,
      "grad_norm": 7.389822006225586,
      "learning_rate": 4.590950973413619e-05,
      "loss": 2.1155399322509765,
      "memory(GiB)": 77.56,
      "step": 61395,
      "token_acc": 0.5446428571428571,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 2.6305642431772416,
      "grad_norm": 5.502264976501465,
      "learning_rate": 4.5902802556631275e-05,
      "loss": 2.4201284408569337,
      "memory(GiB)": 77.56,
      "step": 61400,
      "token_acc": 0.48639455782312924,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 2.630778458506491,
      "grad_norm": 7.2376532554626465,
      "learning_rate": 4.589609545335095e-05,
      "loss": 2.499235153198242,
      "memory(GiB)": 77.56,
      "step": 61405,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.438035
    },
    {
      "epoch": 2.6309926738357396,
      "grad_norm": 5.329863548278809,
      "learning_rate": 4.58893884244167e-05,
      "loss": 2.4541032791137694,
      "memory(GiB)": 77.56,
      "step": 61410,
      "token_acc": 0.4831804281345566,
      "train_speed(iter/s)": 1.438027
    },
    {
      "epoch": 2.6312068891649885,
      "grad_norm": 5.4700236320495605,
      "learning_rate": 4.588268146995003e-05,
      "loss": 2.7074026107788085,
      "memory(GiB)": 77.56,
      "step": 61415,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.438021
    },
    {
      "epoch": 2.6314211044942377,
      "grad_norm": 4.2026495933532715,
      "learning_rate": 4.587597459007246e-05,
      "loss": 2.3165355682373048,
      "memory(GiB)": 77.56,
      "step": 61420,
      "token_acc": 0.5065359477124183,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 2.6316353198234865,
      "grad_norm": 6.1931047439575195,
      "learning_rate": 4.586926778490547e-05,
      "loss": 2.3176557540893556,
      "memory(GiB)": 77.56,
      "step": 61425,
      "token_acc": 0.5148148148148148,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 2.6318495351527353,
      "grad_norm": 7.530332565307617,
      "learning_rate": 4.586256105457056e-05,
      "loss": 2.2246816635131834,
      "memory(GiB)": 77.56,
      "step": 61430,
      "token_acc": 0.4915254237288136,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 2.6320637504819846,
      "grad_norm": 5.485375881195068,
      "learning_rate": 4.585585439918925e-05,
      "loss": 2.5426197052001953,
      "memory(GiB)": 77.56,
      "step": 61435,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 2.6322779658112334,
      "grad_norm": 6.360501289367676,
      "learning_rate": 4.5849147818883017e-05,
      "loss": 2.617303657531738,
      "memory(GiB)": 77.56,
      "step": 61440,
      "token_acc": 0.48297213622291024,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 2.6324921811404822,
      "grad_norm": 8.146564483642578,
      "learning_rate": 4.584244131377338e-05,
      "loss": 2.1984540939331056,
      "memory(GiB)": 77.56,
      "step": 61445,
      "token_acc": 0.5258964143426295,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 2.6327063964697315,
      "grad_norm": 5.190566062927246,
      "learning_rate": 4.5835734883981804e-05,
      "loss": 2.370341491699219,
      "memory(GiB)": 77.56,
      "step": 61450,
      "token_acc": 0.4820846905537459,
      "train_speed(iter/s)": 1.437982
    },
    {
      "epoch": 2.6329206117989803,
      "grad_norm": 5.975463390350342,
      "learning_rate": 4.5829028529629794e-05,
      "loss": 2.45679931640625,
      "memory(GiB)": 77.56,
      "step": 61455,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 2.633134827128229,
      "grad_norm": 5.87574577331543,
      "learning_rate": 4.5822322250838836e-05,
      "loss": 2.689578628540039,
      "memory(GiB)": 77.56,
      "step": 61460,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 2.6333490424574784,
      "grad_norm": 6.860673904418945,
      "learning_rate": 4.5815616047730455e-05,
      "loss": 2.4837966918945313,
      "memory(GiB)": 77.56,
      "step": 61465,
      "token_acc": 0.5046153846153846,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 2.633563257786727,
      "grad_norm": 5.0913190841674805,
      "learning_rate": 4.580890992042611e-05,
      "loss": 2.35849666595459,
      "memory(GiB)": 77.56,
      "step": 61470,
      "token_acc": 0.48220064724919093,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 2.633777473115976,
      "grad_norm": 7.2061967849731445,
      "learning_rate": 4.580220386904728e-05,
      "loss": 2.5140594482421874,
      "memory(GiB)": 77.56,
      "step": 61475,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437997
    },
    {
      "epoch": 2.6339916884452252,
      "grad_norm": 5.278740882873535,
      "learning_rate": 4.579549789371548e-05,
      "loss": 2.097747802734375,
      "memory(GiB)": 77.56,
      "step": 61480,
      "token_acc": 0.5269709543568465,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 2.634205903774474,
      "grad_norm": 6.1236395835876465,
      "learning_rate": 4.578879199455216e-05,
      "loss": 2.319572639465332,
      "memory(GiB)": 77.56,
      "step": 61485,
      "token_acc": 0.5120967741935484,
      "train_speed(iter/s)": 1.438034
    },
    {
      "epoch": 2.634420119103723,
      "grad_norm": 5.959428787231445,
      "learning_rate": 4.5782086171678845e-05,
      "loss": 2.5790782928466798,
      "memory(GiB)": 77.56,
      "step": 61490,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 2.634634334432972,
      "grad_norm": 6.366353511810303,
      "learning_rate": 4.577538042521699e-05,
      "loss": 2.2309345245361327,
      "memory(GiB)": 77.56,
      "step": 61495,
      "token_acc": 0.5223880597014925,
      "train_speed(iter/s)": 1.43803
    },
    {
      "epoch": 2.634848549762221,
      "grad_norm": 5.365630626678467,
      "learning_rate": 4.5768674755288074e-05,
      "loss": 2.4895315170288086,
      "memory(GiB)": 77.56,
      "step": 61500,
      "token_acc": 0.4844290657439446,
      "train_speed(iter/s)": 1.438036
    },
    {
      "epoch": 2.634848549762221,
      "eval_loss": 2.0978097915649414,
      "eval_runtime": 14.3868,
      "eval_samples_per_second": 6.951,
      "eval_steps_per_second": 6.951,
      "eval_token_acc": 0.4894179894179894,
      "step": 61500
    },
    {
      "epoch": 2.6350627650914697,
      "grad_norm": 6.450132846832275,
      "learning_rate": 4.576196916201359e-05,
      "loss": 2.5573665618896486,
      "memory(GiB)": 77.56,
      "step": 61505,
      "token_acc": 0.4896810506566604,
      "train_speed(iter/s)": 1.437526
    },
    {
      "epoch": 2.635276980420719,
      "grad_norm": 4.668859958648682,
      "learning_rate": 4.575526364551501e-05,
      "loss": 2.3783327102661134,
      "memory(GiB)": 77.56,
      "step": 61510,
      "token_acc": 0.46229508196721314,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 2.635491195749968,
      "grad_norm": 4.791419982910156,
      "learning_rate": 4.5748558205913814e-05,
      "loss": 2.3759929656982424,
      "memory(GiB)": 77.56,
      "step": 61515,
      "token_acc": 0.49085365853658536,
      "train_speed(iter/s)": 1.437546
    },
    {
      "epoch": 2.6357054110792166,
      "grad_norm": 4.853916168212891,
      "learning_rate": 4.5741852843331466e-05,
      "loss": 2.403456687927246,
      "memory(GiB)": 77.56,
      "step": 61520,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 2.635919626408466,
      "grad_norm": 7.256995677947998,
      "learning_rate": 4.5735147557889445e-05,
      "loss": 2.435834503173828,
      "memory(GiB)": 77.56,
      "step": 61525,
      "token_acc": 0.4639175257731959,
      "train_speed(iter/s)": 1.437576
    },
    {
      "epoch": 2.6361338417377147,
      "grad_norm": 5.986945152282715,
      "learning_rate": 4.5728442349709244e-05,
      "loss": 2.2936466217041014,
      "memory(GiB)": 77.56,
      "step": 61530,
      "token_acc": 0.50390625,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 2.6363480570669635,
      "grad_norm": 5.918588161468506,
      "learning_rate": 4.5721737218912285e-05,
      "loss": 2.402658462524414,
      "memory(GiB)": 77.56,
      "step": 61535,
      "token_acc": 0.46325878594249204,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 2.6365622723962128,
      "grad_norm": 5.93505859375,
      "learning_rate": 4.57150321656201e-05,
      "loss": 2.6621875762939453,
      "memory(GiB)": 77.56,
      "step": 61540,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.437608
    },
    {
      "epoch": 2.6367764877254616,
      "grad_norm": 5.0371222496032715,
      "learning_rate": 4.5708327189954116e-05,
      "loss": 2.270490837097168,
      "memory(GiB)": 77.56,
      "step": 61545,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437624
    },
    {
      "epoch": 2.6369907030547104,
      "grad_norm": 6.274292945861816,
      "learning_rate": 4.5701622292035815e-05,
      "loss": 2.655232620239258,
      "memory(GiB)": 77.56,
      "step": 61550,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437609
    },
    {
      "epoch": 2.6372049183839597,
      "grad_norm": 6.238415718078613,
      "learning_rate": 4.569491747198666e-05,
      "loss": 2.353529930114746,
      "memory(GiB)": 77.56,
      "step": 61555,
      "token_acc": 0.5346534653465347,
      "train_speed(iter/s)": 1.437613
    },
    {
      "epoch": 2.6374191337132085,
      "grad_norm": 4.50010347366333,
      "learning_rate": 4.56882127299281e-05,
      "loss": 2.352578353881836,
      "memory(GiB)": 77.56,
      "step": 61560,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 2.6376333490424573,
      "grad_norm": 6.645480155944824,
      "learning_rate": 4.568150806598163e-05,
      "loss": 2.5335971832275392,
      "memory(GiB)": 77.56,
      "step": 61565,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.437635
    },
    {
      "epoch": 2.6378475643717065,
      "grad_norm": 6.403446197509766,
      "learning_rate": 4.567480348026869e-05,
      "loss": 2.4612537384033204,
      "memory(GiB)": 77.56,
      "step": 61570,
      "token_acc": 0.41935483870967744,
      "train_speed(iter/s)": 1.437663
    },
    {
      "epoch": 2.6380617797009553,
      "grad_norm": 6.063022136688232,
      "learning_rate": 4.566809897291073e-05,
      "loss": 2.363648796081543,
      "memory(GiB)": 77.56,
      "step": 61575,
      "token_acc": 0.545816733067729,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 2.638275995030204,
      "grad_norm": 6.214544773101807,
      "learning_rate": 4.566139454402923e-05,
      "loss": 2.4042648315429687,
      "memory(GiB)": 77.56,
      "step": 61580,
      "token_acc": 0.4599406528189911,
      "train_speed(iter/s)": 1.437677
    },
    {
      "epoch": 2.6384902103594534,
      "grad_norm": 6.261946201324463,
      "learning_rate": 4.565469019374563e-05,
      "loss": 2.2694778442382812,
      "memory(GiB)": 77.56,
      "step": 61585,
      "token_acc": 0.5330578512396694,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 2.6387044256887022,
      "grad_norm": 6.364415168762207,
      "learning_rate": 4.5647985922181405e-05,
      "loss": 2.2851587295532227,
      "memory(GiB)": 77.56,
      "step": 61590,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 2.638918641017951,
      "grad_norm": 5.5399322509765625,
      "learning_rate": 4.564128172945799e-05,
      "loss": 2.3585536956787108,
      "memory(GiB)": 77.56,
      "step": 61595,
      "token_acc": 0.5244444444444445,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 2.6391328563472003,
      "grad_norm": 7.091579914093018,
      "learning_rate": 4.563457761569685e-05,
      "loss": 2.488679122924805,
      "memory(GiB)": 77.56,
      "step": 61600,
      "token_acc": 0.5163636363636364,
      "train_speed(iter/s)": 1.437635
    },
    {
      "epoch": 2.639347071676449,
      "grad_norm": 5.665006160736084,
      "learning_rate": 4.56278735810194e-05,
      "loss": 2.534504508972168,
      "memory(GiB)": 77.56,
      "step": 61605,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.437657
    },
    {
      "epoch": 2.639561287005698,
      "grad_norm": 5.5218353271484375,
      "learning_rate": 4.562116962554716e-05,
      "loss": 2.477473258972168,
      "memory(GiB)": 77.56,
      "step": 61610,
      "token_acc": 0.48616600790513836,
      "train_speed(iter/s)": 1.437657
    },
    {
      "epoch": 2.639775502334947,
      "grad_norm": 4.8111653327941895,
      "learning_rate": 4.561446574940153e-05,
      "loss": 2.4436132431030275,
      "memory(GiB)": 77.56,
      "step": 61615,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 2.639989717664196,
      "grad_norm": 5.70447301864624,
      "learning_rate": 4.560776195270396e-05,
      "loss": 2.5345947265625,
      "memory(GiB)": 77.56,
      "step": 61620,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.437661
    },
    {
      "epoch": 2.640203932993445,
      "grad_norm": 4.418641090393066,
      "learning_rate": 4.5601058235575914e-05,
      "loss": 2.4715129852294924,
      "memory(GiB)": 77.56,
      "step": 61625,
      "token_acc": 0.4629080118694362,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 2.640418148322694,
      "grad_norm": 4.811949253082275,
      "learning_rate": 4.559435459813881e-05,
      "loss": 2.4463916778564454,
      "memory(GiB)": 77.56,
      "step": 61630,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.640632363651943,
      "grad_norm": 7.582095146179199,
      "learning_rate": 4.558765104051412e-05,
      "loss": 2.20712890625,
      "memory(GiB)": 77.56,
      "step": 61635,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437705
    },
    {
      "epoch": 2.6408465789811917,
      "grad_norm": 6.083032608032227,
      "learning_rate": 4.558094756282327e-05,
      "loss": 2.3288381576538084,
      "memory(GiB)": 77.56,
      "step": 61640,
      "token_acc": 0.521875,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 2.641060794310441,
      "grad_norm": 6.091094017028809,
      "learning_rate": 4.557424416518768e-05,
      "loss": 2.507367706298828,
      "memory(GiB)": 77.56,
      "step": 61645,
      "token_acc": 0.5399239543726235,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 2.6412750096396898,
      "grad_norm": 7.4312591552734375,
      "learning_rate": 4.556754084772883e-05,
      "loss": 2.6219932556152346,
      "memory(GiB)": 77.56,
      "step": 61650,
      "token_acc": 0.5061349693251533,
      "train_speed(iter/s)": 1.437718
    },
    {
      "epoch": 2.6414892249689386,
      "grad_norm": 5.042299270629883,
      "learning_rate": 4.556083761056814e-05,
      "loss": 2.2487071990966796,
      "memory(GiB)": 77.56,
      "step": 61655,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 2.641703440298188,
      "grad_norm": 5.684948921203613,
      "learning_rate": 4.5554134453827024e-05,
      "loss": 2.4063066482543944,
      "memory(GiB)": 77.56,
      "step": 61660,
      "token_acc": 0.4763779527559055,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 2.6419176556274366,
      "grad_norm": 6.47034215927124,
      "learning_rate": 4.554743137762694e-05,
      "loss": 2.3203659057617188,
      "memory(GiB)": 77.56,
      "step": 61665,
      "token_acc": 0.47896440129449835,
      "train_speed(iter/s)": 1.437751
    },
    {
      "epoch": 2.6421318709566854,
      "grad_norm": 5.349266529083252,
      "learning_rate": 4.554072838208931e-05,
      "loss": 2.536144828796387,
      "memory(GiB)": 77.56,
      "step": 61670,
      "token_acc": 0.46357615894039733,
      "train_speed(iter/s)": 1.437767
    },
    {
      "epoch": 2.6423460862859347,
      "grad_norm": 5.05371618270874,
      "learning_rate": 4.553402546733557e-05,
      "loss": 2.061013412475586,
      "memory(GiB)": 77.56,
      "step": 61675,
      "token_acc": 0.5400696864111498,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 2.6425603016151835,
      "grad_norm": 6.922592639923096,
      "learning_rate": 4.552732263348713e-05,
      "loss": 2.3230724334716797,
      "memory(GiB)": 77.56,
      "step": 61680,
      "token_acc": 0.532258064516129,
      "train_speed(iter/s)": 1.437765
    },
    {
      "epoch": 2.6427745169444323,
      "grad_norm": 6.067020416259766,
      "learning_rate": 4.5520619880665466e-05,
      "loss": 2.3619842529296875,
      "memory(GiB)": 77.56,
      "step": 61685,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 2.6429887322736816,
      "grad_norm": 5.453634738922119,
      "learning_rate": 4.551391720899196e-05,
      "loss": 2.785741424560547,
      "memory(GiB)": 77.56,
      "step": 61690,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.437777
    },
    {
      "epoch": 2.6432029476029304,
      "grad_norm": 5.473423957824707,
      "learning_rate": 4.5507214618588055e-05,
      "loss": 2.5167659759521483,
      "memory(GiB)": 77.56,
      "step": 61695,
      "token_acc": 0.48825065274151436,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 2.643417162932179,
      "grad_norm": 4.922832489013672,
      "learning_rate": 4.5500512109575186e-05,
      "loss": 2.6155630111694337,
      "memory(GiB)": 77.56,
      "step": 61700,
      "token_acc": 0.4773413897280967,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 2.6436313782614285,
      "grad_norm": 5.397421836853027,
      "learning_rate": 4.549380968207474e-05,
      "loss": 2.2809940338134767,
      "memory(GiB)": 77.56,
      "step": 61705,
      "token_acc": 0.484251968503937,
      "train_speed(iter/s)": 1.437794
    },
    {
      "epoch": 2.6438455935906773,
      "grad_norm": 6.049432754516602,
      "learning_rate": 4.5487107336208166e-05,
      "loss": 2.534933853149414,
      "memory(GiB)": 77.56,
      "step": 61710,
      "token_acc": 0.44482758620689655,
      "train_speed(iter/s)": 1.437834
    },
    {
      "epoch": 2.644059808919926,
      "grad_norm": 5.0446553230285645,
      "learning_rate": 4.548040507209689e-05,
      "loss": 2.156915855407715,
      "memory(GiB)": 77.56,
      "step": 61715,
      "token_acc": 0.532258064516129,
      "train_speed(iter/s)": 1.437863
    },
    {
      "epoch": 2.6442740242491753,
      "grad_norm": 4.818816184997559,
      "learning_rate": 4.547370288986229e-05,
      "loss": 2.4247295379638674,
      "memory(GiB)": 77.56,
      "step": 61720,
      "token_acc": 0.5016501650165016,
      "train_speed(iter/s)": 1.437867
    },
    {
      "epoch": 2.644488239578424,
      "grad_norm": 7.382740020751953,
      "learning_rate": 4.546700078962582e-05,
      "loss": 2.592110252380371,
      "memory(GiB)": 77.56,
      "step": 61725,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 2.644702454907673,
      "grad_norm": 7.030486106872559,
      "learning_rate": 4.5460298771508885e-05,
      "loss": 2.628032684326172,
      "memory(GiB)": 77.56,
      "step": 61730,
      "token_acc": 0.4219858156028369,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 2.6449166702369222,
      "grad_norm": 5.9276556968688965,
      "learning_rate": 4.54535968356329e-05,
      "loss": 2.621383857727051,
      "memory(GiB)": 77.56,
      "step": 61735,
      "token_acc": 0.4336569579288026,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 2.645130885566171,
      "grad_norm": 5.162145614624023,
      "learning_rate": 4.544689498211927e-05,
      "loss": 2.523931694030762,
      "memory(GiB)": 77.56,
      "step": 61740,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 2.64534510089542,
      "grad_norm": 4.634915828704834,
      "learning_rate": 4.54401932110894e-05,
      "loss": 2.373402404785156,
      "memory(GiB)": 77.56,
      "step": 61745,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.437936
    },
    {
      "epoch": 2.645559316224669,
      "grad_norm": 4.609366416931152,
      "learning_rate": 4.543349152266472e-05,
      "loss": 2.430101776123047,
      "memory(GiB)": 77.56,
      "step": 61750,
      "token_acc": 0.4483695652173913,
      "train_speed(iter/s)": 1.43794
    },
    {
      "epoch": 2.645773531553918,
      "grad_norm": 4.220828533172607,
      "learning_rate": 4.54267899169666e-05,
      "loss": 2.3142263412475588,
      "memory(GiB)": 77.56,
      "step": 61755,
      "token_acc": 0.4858156028368794,
      "train_speed(iter/s)": 1.43795
    },
    {
      "epoch": 2.6459877468831667,
      "grad_norm": 9.216361045837402,
      "learning_rate": 4.54200883941165e-05,
      "loss": 2.341102409362793,
      "memory(GiB)": 77.56,
      "step": 61760,
      "token_acc": 0.492,
      "train_speed(iter/s)": 1.437921
    },
    {
      "epoch": 2.646201962212416,
      "grad_norm": 4.543700218200684,
      "learning_rate": 4.541338695423578e-05,
      "loss": 2.056321907043457,
      "memory(GiB)": 77.56,
      "step": 61765,
      "token_acc": 0.5645756457564576,
      "train_speed(iter/s)": 1.437933
    },
    {
      "epoch": 2.646416177541665,
      "grad_norm": 7.84173059463501,
      "learning_rate": 4.5406685597445866e-05,
      "loss": 2.462577056884766,
      "memory(GiB)": 77.56,
      "step": 61770,
      "token_acc": 0.496,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 2.6466303928709136,
      "grad_norm": 6.8672261238098145,
      "learning_rate": 4.539998432386814e-05,
      "loss": 2.4416786193847657,
      "memory(GiB)": 77.56,
      "step": 61775,
      "token_acc": 0.4506578947368421,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 2.646844608200163,
      "grad_norm": 4.976284027099609,
      "learning_rate": 4.5393283133624026e-05,
      "loss": 2.443865203857422,
      "memory(GiB)": 77.56,
      "step": 61780,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.437973
    },
    {
      "epoch": 2.6470588235294117,
      "grad_norm": 5.521240711212158,
      "learning_rate": 4.5386582026834906e-05,
      "loss": 2.5446725845336915,
      "memory(GiB)": 77.56,
      "step": 61785,
      "token_acc": 0.4697986577181208,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 2.6472730388586605,
      "grad_norm": 5.984615802764893,
      "learning_rate": 4.5379881003622174e-05,
      "loss": 2.3382713317871096,
      "memory(GiB)": 77.56,
      "step": 61790,
      "token_acc": 0.5310344827586206,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 2.6474872541879098,
      "grad_norm": 6.743816375732422,
      "learning_rate": 4.537318006410724e-05,
      "loss": 2.3853715896606444,
      "memory(GiB)": 77.56,
      "step": 61795,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.437997
    },
    {
      "epoch": 2.6477014695171586,
      "grad_norm": 8.244396209716797,
      "learning_rate": 4.53664792084115e-05,
      "loss": 2.377880668640137,
      "memory(GiB)": 77.56,
      "step": 61800,
      "token_acc": 0.519163763066202,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 2.6479156848464074,
      "grad_norm": 5.245119094848633,
      "learning_rate": 4.535977843665631e-05,
      "loss": 2.7398296356201173,
      "memory(GiB)": 77.56,
      "step": 61805,
      "token_acc": 0.45938375350140054,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 2.6481299001756566,
      "grad_norm": 9.776571273803711,
      "learning_rate": 4.53530777489631e-05,
      "loss": 2.6974952697753904,
      "memory(GiB)": 77.56,
      "step": 61810,
      "token_acc": 0.4816326530612245,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 2.6483441155049054,
      "grad_norm": 7.274689674377441,
      "learning_rate": 4.534637714545324e-05,
      "loss": 2.5024799346923827,
      "memory(GiB)": 77.56,
      "step": 61815,
      "token_acc": 0.4645669291338583,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 2.6485583308341543,
      "grad_norm": 4.626080513000488,
      "learning_rate": 4.533967662624813e-05,
      "loss": 2.526080322265625,
      "memory(GiB)": 77.56,
      "step": 61820,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 2.6487725461634035,
      "grad_norm": 4.9425787925720215,
      "learning_rate": 4.533297619146915e-05,
      "loss": 2.480988311767578,
      "memory(GiB)": 77.56,
      "step": 61825,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 2.6489867614926523,
      "grad_norm": 4.877227783203125,
      "learning_rate": 4.532627584123766e-05,
      "loss": 2.669286346435547,
      "memory(GiB)": 77.56,
      "step": 61830,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.438058
    },
    {
      "epoch": 2.649200976821901,
      "grad_norm": 7.359621524810791,
      "learning_rate": 4.531957557567509e-05,
      "loss": 2.4072750091552733,
      "memory(GiB)": 77.56,
      "step": 61835,
      "token_acc": 0.48563218390804597,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 2.6494151921511504,
      "grad_norm": 6.575071811676025,
      "learning_rate": 4.5312875394902795e-05,
      "loss": 2.6007226943969726,
      "memory(GiB)": 77.56,
      "step": 61840,
      "token_acc": 0.4440677966101695,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 2.649629407480399,
      "grad_norm": 4.844421863555908,
      "learning_rate": 4.530617529904218e-05,
      "loss": 2.315010833740234,
      "memory(GiB)": 77.56,
      "step": 61845,
      "token_acc": 0.4954682779456193,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 2.649843622809648,
      "grad_norm": 4.905552864074707,
      "learning_rate": 4.529947528821457e-05,
      "loss": 2.1814138412475588,
      "memory(GiB)": 77.56,
      "step": 61850,
      "token_acc": 0.45724907063197023,
      "train_speed(iter/s)": 1.43803
    },
    {
      "epoch": 2.6500578381388973,
      "grad_norm": 6.575028896331787,
      "learning_rate": 4.52927753625414e-05,
      "loss": 2.420262908935547,
      "memory(GiB)": 77.56,
      "step": 61855,
      "token_acc": 0.5017064846416383,
      "train_speed(iter/s)": 1.438046
    },
    {
      "epoch": 2.650272053468146,
      "grad_norm": 5.843136310577393,
      "learning_rate": 4.528607552214401e-05,
      "loss": 2.747515106201172,
      "memory(GiB)": 77.56,
      "step": 61860,
      "token_acc": 0.44281524926686217,
      "train_speed(iter/s)": 1.438063
    },
    {
      "epoch": 2.650486268797395,
      "grad_norm": 4.591243743896484,
      "learning_rate": 4.527937576714378e-05,
      "loss": 2.6525651931762697,
      "memory(GiB)": 77.56,
      "step": 61865,
      "token_acc": 0.4560810810810811,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 2.650700484126644,
      "grad_norm": 5.79725980758667,
      "learning_rate": 4.527267609766209e-05,
      "loss": 2.37323112487793,
      "memory(GiB)": 77.56,
      "step": 61870,
      "token_acc": 0.5104602510460251,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 2.650914699455893,
      "grad_norm": 4.817253112792969,
      "learning_rate": 4.52659765138203e-05,
      "loss": 1.9315811157226563,
      "memory(GiB)": 77.56,
      "step": 61875,
      "token_acc": 0.555984555984556,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 2.651128914785142,
      "grad_norm": 5.264191627502441,
      "learning_rate": 4.5259277015739796e-05,
      "loss": 2.183505630493164,
      "memory(GiB)": 77.56,
      "step": 61880,
      "token_acc": 0.5418060200668896,
      "train_speed(iter/s)": 1.438021
    },
    {
      "epoch": 2.651343130114391,
      "grad_norm": 6.933333396911621,
      "learning_rate": 4.525257760354194e-05,
      "loss": 2.2891656875610353,
      "memory(GiB)": 77.56,
      "step": 61885,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.438034
    },
    {
      "epoch": 2.65155734544364,
      "grad_norm": 4.850844383239746,
      "learning_rate": 4.5245878277348084e-05,
      "loss": 2.208728790283203,
      "memory(GiB)": 77.56,
      "step": 61890,
      "token_acc": 0.4910394265232975,
      "train_speed(iter/s)": 1.438041
    },
    {
      "epoch": 2.6517715607728887,
      "grad_norm": 6.755412578582764,
      "learning_rate": 4.523917903727961e-05,
      "loss": 2.4329196929931642,
      "memory(GiB)": 77.56,
      "step": 61895,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 2.651985776102138,
      "grad_norm": 5.404649257659912,
      "learning_rate": 4.5232479883457865e-05,
      "loss": 2.2061956405639647,
      "memory(GiB)": 77.56,
      "step": 61900,
      "token_acc": 0.5136986301369864,
      "train_speed(iter/s)": 1.438057
    },
    {
      "epoch": 2.6521999914313867,
      "grad_norm": 4.8161845207214355,
      "learning_rate": 4.522578081600421e-05,
      "loss": 2.187418556213379,
      "memory(GiB)": 77.56,
      "step": 61905,
      "token_acc": 0.5364238410596026,
      "train_speed(iter/s)": 1.43807
    },
    {
      "epoch": 2.6524142067606356,
      "grad_norm": 5.173804759979248,
      "learning_rate": 4.521908183504002e-05,
      "loss": 2.567978858947754,
      "memory(GiB)": 77.56,
      "step": 61910,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 2.652628422089885,
      "grad_norm": 5.253247261047363,
      "learning_rate": 4.521238294068667e-05,
      "loss": 2.547380828857422,
      "memory(GiB)": 77.56,
      "step": 61915,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.438071
    },
    {
      "epoch": 2.6528426374191336,
      "grad_norm": 5.067152500152588,
      "learning_rate": 4.520568413306547e-05,
      "loss": 2.4333213806152343,
      "memory(GiB)": 77.56,
      "step": 61920,
      "token_acc": 0.5060606060606061,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 2.6530568527483824,
      "grad_norm": 4.9168925285339355,
      "learning_rate": 4.519898541229781e-05,
      "loss": 2.2088905334472657,
      "memory(GiB)": 77.56,
      "step": 61925,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.438042
    },
    {
      "epoch": 2.6532710680776317,
      "grad_norm": 5.341547966003418,
      "learning_rate": 4.519228677850504e-05,
      "loss": 2.274029350280762,
      "memory(GiB)": 77.56,
      "step": 61930,
      "token_acc": 0.5183823529411765,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 2.6534852834068805,
      "grad_norm": 5.764420509338379,
      "learning_rate": 4.5185588231808486e-05,
      "loss": 2.488014984130859,
      "memory(GiB)": 77.56,
      "step": 61935,
      "token_acc": 0.47076023391812866,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 2.6536994987361293,
      "grad_norm": 5.769437313079834,
      "learning_rate": 4.517888977232953e-05,
      "loss": 2.4321130752563476,
      "memory(GiB)": 77.56,
      "step": 61940,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 2.6539137140653786,
      "grad_norm": 5.0661516189575195,
      "learning_rate": 4.5172191400189515e-05,
      "loss": 2.268540382385254,
      "memory(GiB)": 77.56,
      "step": 61945,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 2.6541279293946274,
      "grad_norm": 5.428701400756836,
      "learning_rate": 4.516549311550977e-05,
      "loss": 2.3401361465454102,
      "memory(GiB)": 77.56,
      "step": 61950,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.438115
    },
    {
      "epoch": 2.654342144723876,
      "grad_norm": 6.1041789054870605,
      "learning_rate": 4.515879491841166e-05,
      "loss": 2.6630428314208983,
      "memory(GiB)": 77.56,
      "step": 61955,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.438141
    },
    {
      "epoch": 2.6545563600531255,
      "grad_norm": 5.632446765899658,
      "learning_rate": 4.515209680901651e-05,
      "loss": 2.4645896911621095,
      "memory(GiB)": 77.56,
      "step": 61960,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.438165
    },
    {
      "epoch": 2.6547705753823743,
      "grad_norm": 5.658358097076416,
      "learning_rate": 4.514539878744568e-05,
      "loss": 2.551421356201172,
      "memory(GiB)": 77.56,
      "step": 61965,
      "token_acc": 0.45321637426900585,
      "train_speed(iter/s)": 1.438161
    },
    {
      "epoch": 2.654984790711623,
      "grad_norm": 5.636961460113525,
      "learning_rate": 4.5138700853820516e-05,
      "loss": 2.4709659576416017,
      "memory(GiB)": 77.56,
      "step": 61970,
      "token_acc": 0.4527687296416938,
      "train_speed(iter/s)": 1.43818
    },
    {
      "epoch": 2.6551990060408723,
      "grad_norm": 5.3820600509643555,
      "learning_rate": 4.513200300826232e-05,
      "loss": 2.4653305053710937,
      "memory(GiB)": 77.56,
      "step": 61975,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.438189
    },
    {
      "epoch": 2.655413221370121,
      "grad_norm": 7.468258380889893,
      "learning_rate": 4.512530525089246e-05,
      "loss": 2.3487545013427735,
      "memory(GiB)": 77.56,
      "step": 61980,
      "token_acc": 0.5020080321285141,
      "train_speed(iter/s)": 1.438223
    },
    {
      "epoch": 2.65562743669937,
      "grad_norm": 4.721065044403076,
      "learning_rate": 4.511860758183229e-05,
      "loss": 2.338658905029297,
      "memory(GiB)": 77.56,
      "step": 61985,
      "token_acc": 0.48360655737704916,
      "train_speed(iter/s)": 1.438223
    },
    {
      "epoch": 2.655841652028619,
      "grad_norm": 4.905104637145996,
      "learning_rate": 4.511191000120312e-05,
      "loss": 2.575343704223633,
      "memory(GiB)": 77.56,
      "step": 61990,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.438233
    },
    {
      "epoch": 2.656055867357868,
      "grad_norm": 5.322173595428467,
      "learning_rate": 4.510521250912627e-05,
      "loss": 2.432086944580078,
      "memory(GiB)": 77.56,
      "step": 61995,
      "token_acc": 0.4703703703703704,
      "train_speed(iter/s)": 1.438216
    },
    {
      "epoch": 2.656270082687117,
      "grad_norm": 8.139036178588867,
      "learning_rate": 4.50985151057231e-05,
      "loss": 2.178491973876953,
      "memory(GiB)": 77.56,
      "step": 62000,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.43821
    },
    {
      "epoch": 2.656270082687117,
      "eval_loss": 2.1532678604125977,
      "eval_runtime": 14.7965,
      "eval_samples_per_second": 6.758,
      "eval_steps_per_second": 6.758,
      "eval_token_acc": 0.4946808510638298,
      "step": 62000
    },
    {
      "epoch": 2.656484298016366,
      "grad_norm": 5.055727005004883,
      "learning_rate": 4.509181779111493e-05,
      "loss": 2.136933135986328,
      "memory(GiB)": 77.56,
      "step": 62005,
      "token_acc": 0.507537688442211,
      "train_speed(iter/s)": 1.437696
    },
    {
      "epoch": 2.656698513345615,
      "grad_norm": 7.198328971862793,
      "learning_rate": 4.508512056542307e-05,
      "loss": 2.575959014892578,
      "memory(GiB)": 77.56,
      "step": 62010,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.437694
    },
    {
      "epoch": 2.6569127286748637,
      "grad_norm": 4.462030410766602,
      "learning_rate": 4.507842342876887e-05,
      "loss": 2.244471549987793,
      "memory(GiB)": 77.56,
      "step": 62015,
      "token_acc": 0.53,
      "train_speed(iter/s)": 1.437709
    },
    {
      "epoch": 2.657126944004113,
      "grad_norm": 7.001450538635254,
      "learning_rate": 4.507172638127364e-05,
      "loss": 2.293403434753418,
      "memory(GiB)": 77.56,
      "step": 62020,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 2.657341159333362,
      "grad_norm": 6.439095973968506,
      "learning_rate": 4.5065029423058726e-05,
      "loss": 2.484415054321289,
      "memory(GiB)": 77.56,
      "step": 62025,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.437707
    },
    {
      "epoch": 2.6575553746626106,
      "grad_norm": 5.766871929168701,
      "learning_rate": 4.505833255424543e-05,
      "loss": 2.247924041748047,
      "memory(GiB)": 77.56,
      "step": 62030,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.437709
    },
    {
      "epoch": 2.65776958999186,
      "grad_norm": 4.9292426109313965,
      "learning_rate": 4.505163577495506e-05,
      "loss": 2.292584037780762,
      "memory(GiB)": 77.56,
      "step": 62035,
      "token_acc": 0.5183946488294314,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 2.6579838053211087,
      "grad_norm": 6.00905179977417,
      "learning_rate": 4.504493908530896e-05,
      "loss": 2.2320985794067383,
      "memory(GiB)": 77.56,
      "step": 62040,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.437718
    },
    {
      "epoch": 2.6581980206503575,
      "grad_norm": 4.733302593231201,
      "learning_rate": 4.5038242485428436e-05,
      "loss": 2.4438005447387696,
      "memory(GiB)": 77.56,
      "step": 62045,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 2.6584122359796067,
      "grad_norm": 5.3019118309021,
      "learning_rate": 4.503154597543479e-05,
      "loss": 2.6854190826416016,
      "memory(GiB)": 77.56,
      "step": 62050,
      "token_acc": 0.4075342465753425,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 2.6586264513088556,
      "grad_norm": 6.637152671813965,
      "learning_rate": 4.5024849555449353e-05,
      "loss": 2.1898777008056642,
      "memory(GiB)": 77.56,
      "step": 62055,
      "token_acc": 0.5203252032520326,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 2.6588406666381044,
      "grad_norm": 6.405393123626709,
      "learning_rate": 4.501815322559345e-05,
      "loss": 2.4268672943115233,
      "memory(GiB)": 77.56,
      "step": 62060,
      "token_acc": 0.48441926345609065,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 2.6590548819673536,
      "grad_norm": 5.337629318237305,
      "learning_rate": 4.501145698598836e-05,
      "loss": 2.5698728561401367,
      "memory(GiB)": 77.56,
      "step": 62065,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 2.6592690972966024,
      "grad_norm": 6.5373005867004395,
      "learning_rate": 4.500476083675542e-05,
      "loss": 2.2987171173095704,
      "memory(GiB)": 77.56,
      "step": 62070,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.6594833126258512,
      "grad_norm": 5.635282516479492,
      "learning_rate": 4.499806477801592e-05,
      "loss": 2.2451248168945312,
      "memory(GiB)": 77.56,
      "step": 62075,
      "token_acc": 0.5191082802547771,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 2.6596975279551005,
      "grad_norm": 7.797072887420654,
      "learning_rate": 4.499136880989116e-05,
      "loss": 2.1330049514770506,
      "memory(GiB)": 77.56,
      "step": 62080,
      "token_acc": 0.5390946502057613,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 2.6599117432843493,
      "grad_norm": 5.674398422241211,
      "learning_rate": 4.498467293250246e-05,
      "loss": 2.315528678894043,
      "memory(GiB)": 77.56,
      "step": 62085,
      "token_acc": 0.5261044176706827,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 2.660125958613598,
      "grad_norm": 5.611018180847168,
      "learning_rate": 4.497797714597112e-05,
      "loss": 2.7278715133666993,
      "memory(GiB)": 77.56,
      "step": 62090,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.6603401739428474,
      "grad_norm": 8.207868576049805,
      "learning_rate": 4.4971281450418425e-05,
      "loss": 2.547300338745117,
      "memory(GiB)": 77.56,
      "step": 62095,
      "token_acc": 0.4675324675324675,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 2.660554389272096,
      "grad_norm": 4.00620698928833,
      "learning_rate": 4.496458584596569e-05,
      "loss": 2.5608427047729494,
      "memory(GiB)": 77.56,
      "step": 62100,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.437787
    },
    {
      "epoch": 2.660768604601345,
      "grad_norm": 5.188326835632324,
      "learning_rate": 4.495789033273419e-05,
      "loss": 2.6191925048828124,
      "memory(GiB)": 77.56,
      "step": 62105,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 2.6609828199305943,
      "grad_norm": 5.956536769866943,
      "learning_rate": 4.495119491084526e-05,
      "loss": 2.505153846740723,
      "memory(GiB)": 77.56,
      "step": 62110,
      "token_acc": 0.4521452145214521,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 2.661197035259843,
      "grad_norm": 5.652467250823975,
      "learning_rate": 4.4944499580420166e-05,
      "loss": 2.4883447647094727,
      "memory(GiB)": 77.56,
      "step": 62115,
      "token_acc": 0.48732394366197185,
      "train_speed(iter/s)": 1.437792
    },
    {
      "epoch": 2.661411250589092,
      "grad_norm": 5.945432662963867,
      "learning_rate": 4.4937804341580184e-05,
      "loss": 2.289636421203613,
      "memory(GiB)": 77.56,
      "step": 62120,
      "token_acc": 0.5102639296187683,
      "train_speed(iter/s)": 1.437768
    },
    {
      "epoch": 2.661625465918341,
      "grad_norm": 6.544475078582764,
      "learning_rate": 4.4931109194446624e-05,
      "loss": 2.480554962158203,
      "memory(GiB)": 77.56,
      "step": 62125,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.437778
    },
    {
      "epoch": 2.66183968124759,
      "grad_norm": 5.793374538421631,
      "learning_rate": 4.49244141391408e-05,
      "loss": 2.2614669799804688,
      "memory(GiB)": 77.56,
      "step": 62130,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437769
    },
    {
      "epoch": 2.6620538965768388,
      "grad_norm": 5.474088191986084,
      "learning_rate": 4.4917719175783965e-05,
      "loss": 2.6035057067871095,
      "memory(GiB)": 77.56,
      "step": 62135,
      "token_acc": 0.43952802359882004,
      "train_speed(iter/s)": 1.437788
    },
    {
      "epoch": 2.662268111906088,
      "grad_norm": 5.417370319366455,
      "learning_rate": 4.491102430449741e-05,
      "loss": 2.304451751708984,
      "memory(GiB)": 77.56,
      "step": 62140,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.437782
    },
    {
      "epoch": 2.662482327235337,
      "grad_norm": 5.341798782348633,
      "learning_rate": 4.490432952540243e-05,
      "loss": 2.488572692871094,
      "memory(GiB)": 77.56,
      "step": 62145,
      "token_acc": 0.4506578947368421,
      "train_speed(iter/s)": 1.437811
    },
    {
      "epoch": 2.6626965425645857,
      "grad_norm": 5.721759796142578,
      "learning_rate": 4.489763483862031e-05,
      "loss": 2.3949277877807615,
      "memory(GiB)": 77.56,
      "step": 62150,
      "token_acc": 0.4757834757834758,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 2.662910757893835,
      "grad_norm": 5.802772045135498,
      "learning_rate": 4.4890940244272305e-05,
      "loss": 2.3115299224853514,
      "memory(GiB)": 77.56,
      "step": 62155,
      "token_acc": 0.5159235668789809,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 2.6631249732230837,
      "grad_norm": 4.604318141937256,
      "learning_rate": 4.488424574247972e-05,
      "loss": 2.3293811798095705,
      "memory(GiB)": 77.56,
      "step": 62160,
      "token_acc": 0.486013986013986,
      "train_speed(iter/s)": 1.437799
    },
    {
      "epoch": 2.6633391885523325,
      "grad_norm": 5.474299907684326,
      "learning_rate": 4.4877551333363814e-05,
      "loss": 2.415744400024414,
      "memory(GiB)": 77.56,
      "step": 62165,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 2.663553403881582,
      "grad_norm": 4.148676872253418,
      "learning_rate": 4.487085701704588e-05,
      "loss": 2.5797075271606444,
      "memory(GiB)": 77.56,
      "step": 62170,
      "token_acc": 0.48771929824561405,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 2.6637676192108306,
      "grad_norm": 5.646249771118164,
      "learning_rate": 4.4864162793647184e-05,
      "loss": 2.4013315200805665,
      "memory(GiB)": 77.56,
      "step": 62175,
      "token_acc": 0.5077881619937694,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 2.6639818345400794,
      "grad_norm": 5.800014495849609,
      "learning_rate": 4.4857468663288985e-05,
      "loss": 2.3350486755371094,
      "memory(GiB)": 77.56,
      "step": 62180,
      "token_acc": 0.4674329501915709,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 2.6641960498693287,
      "grad_norm": 6.78893518447876,
      "learning_rate": 4.485077462609258e-05,
      "loss": 2.4084039688110352,
      "memory(GiB)": 77.56,
      "step": 62185,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 2.6644102651985775,
      "grad_norm": 5.667250156402588,
      "learning_rate": 4.484408068217922e-05,
      "loss": 2.7562252044677735,
      "memory(GiB)": 77.56,
      "step": 62190,
      "token_acc": 0.47904191616766467,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 2.6646244805278263,
      "grad_norm": 5.6670026779174805,
      "learning_rate": 4.4837386831670155e-05,
      "loss": 2.5700937271118165,
      "memory(GiB)": 77.56,
      "step": 62195,
      "token_acc": 0.4554794520547945,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 2.6648386958570756,
      "grad_norm": 5.202577590942383,
      "learning_rate": 4.4830693074686675e-05,
      "loss": 2.094849395751953,
      "memory(GiB)": 77.56,
      "step": 62200,
      "token_acc": 0.5546875,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 2.6650529111863244,
      "grad_norm": 4.83437442779541,
      "learning_rate": 4.482399941135005e-05,
      "loss": 2.7122657775878904,
      "memory(GiB)": 77.56,
      "step": 62205,
      "token_acc": 0.4273972602739726,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 2.665267126515573,
      "grad_norm": 5.236632347106934,
      "learning_rate": 4.481730584178153e-05,
      "loss": 2.6442386627197267,
      "memory(GiB)": 77.56,
      "step": 62210,
      "token_acc": 0.4452054794520548,
      "train_speed(iter/s)": 1.4379
    },
    {
      "epoch": 2.6654813418448224,
      "grad_norm": 5.442342758178711,
      "learning_rate": 4.481061236610238e-05,
      "loss": 2.339546966552734,
      "memory(GiB)": 77.56,
      "step": 62215,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.437904
    },
    {
      "epoch": 2.6656955571740713,
      "grad_norm": 5.601112365722656,
      "learning_rate": 4.480391898443386e-05,
      "loss": 2.450676918029785,
      "memory(GiB)": 77.56,
      "step": 62220,
      "token_acc": 0.4915254237288136,
      "train_speed(iter/s)": 1.437891
    },
    {
      "epoch": 2.66590977250332,
      "grad_norm": 5.787229537963867,
      "learning_rate": 4.4797225696897205e-05,
      "loss": 2.6089290618896483,
      "memory(GiB)": 77.56,
      "step": 62225,
      "token_acc": 0.46863468634686345,
      "train_speed(iter/s)": 1.437902
    },
    {
      "epoch": 2.6661239878325693,
      "grad_norm": 4.809157371520996,
      "learning_rate": 4.4790532503613696e-05,
      "loss": 2.4545944213867186,
      "memory(GiB)": 77.56,
      "step": 62230,
      "token_acc": 0.5207667731629393,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 2.666338203161818,
      "grad_norm": 4.896027088165283,
      "learning_rate": 4.4783839404704587e-05,
      "loss": 2.4091516494750977,
      "memory(GiB)": 77.56,
      "step": 62235,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.437849
    },
    {
      "epoch": 2.666552418491067,
      "grad_norm": 4.648581504821777,
      "learning_rate": 4.47771464002911e-05,
      "loss": 2.604655075073242,
      "memory(GiB)": 77.56,
      "step": 62240,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 2.666766633820316,
      "grad_norm": 4.681258678436279,
      "learning_rate": 4.4770453490494526e-05,
      "loss": 2.5066070556640625,
      "memory(GiB)": 77.56,
      "step": 62245,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.437903
    },
    {
      "epoch": 2.666980849149565,
      "grad_norm": 7.053329944610596,
      "learning_rate": 4.476376067543607e-05,
      "loss": 2.432002639770508,
      "memory(GiB)": 77.56,
      "step": 62250,
      "token_acc": 0.4539877300613497,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 2.6671950644788143,
      "grad_norm": 6.050266265869141,
      "learning_rate": 4.475706795523702e-05,
      "loss": 2.459771728515625,
      "memory(GiB)": 77.56,
      "step": 62255,
      "token_acc": 0.509009009009009,
      "train_speed(iter/s)": 1.437914
    },
    {
      "epoch": 2.667409279808063,
      "grad_norm": 4.887913703918457,
      "learning_rate": 4.47503753300186e-05,
      "loss": 2.7142181396484375,
      "memory(GiB)": 77.56,
      "step": 62260,
      "token_acc": 0.4417808219178082,
      "train_speed(iter/s)": 1.437915
    },
    {
      "epoch": 2.667623495137312,
      "grad_norm": 6.596017837524414,
      "learning_rate": 4.474368279990205e-05,
      "loss": 2.752261734008789,
      "memory(GiB)": 77.56,
      "step": 62265,
      "token_acc": 0.44664031620553357,
      "train_speed(iter/s)": 1.437916
    },
    {
      "epoch": 2.667837710466561,
      "grad_norm": 4.929872989654541,
      "learning_rate": 4.47369903650086e-05,
      "loss": 2.210143280029297,
      "memory(GiB)": 77.56,
      "step": 62270,
      "token_acc": 0.5409252669039146,
      "train_speed(iter/s)": 1.437936
    },
    {
      "epoch": 2.66805192579581,
      "grad_norm": 6.227293968200684,
      "learning_rate": 4.4730298025459536e-05,
      "loss": 2.450564956665039,
      "memory(GiB)": 77.56,
      "step": 62275,
      "token_acc": 0.49466192170818507,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 2.6682661411250588,
      "grad_norm": 5.108763694763184,
      "learning_rate": 4.4723605781376056e-05,
      "loss": 2.659554290771484,
      "memory(GiB)": 77.56,
      "step": 62280,
      "token_acc": 0.4208754208754209,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 2.668480356454308,
      "grad_norm": 5.252298831939697,
      "learning_rate": 4.47169136328794e-05,
      "loss": 2.1697967529296873,
      "memory(GiB)": 77.56,
      "step": 62285,
      "token_acc": 0.5477178423236515,
      "train_speed(iter/s)": 1.437981
    },
    {
      "epoch": 2.668694571783557,
      "grad_norm": 5.540384769439697,
      "learning_rate": 4.471022158009082e-05,
      "loss": 2.4978921890258787,
      "memory(GiB)": 77.56,
      "step": 62290,
      "token_acc": 0.4637223974763407,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 2.6689087871128057,
      "grad_norm": 5.051617622375488,
      "learning_rate": 4.470352962313154e-05,
      "loss": 2.589203643798828,
      "memory(GiB)": 77.56,
      "step": 62295,
      "token_acc": 0.4403183023872679,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 2.669123002442055,
      "grad_norm": 5.171085357666016,
      "learning_rate": 4.4696837762122777e-05,
      "loss": 2.4283390045166016,
      "memory(GiB)": 77.56,
      "step": 62300,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 2.6693372177713037,
      "grad_norm": 4.92645788192749,
      "learning_rate": 4.469014599718579e-05,
      "loss": 2.110988998413086,
      "memory(GiB)": 77.56,
      "step": 62305,
      "token_acc": 0.5288135593220339,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 2.6695514331005525,
      "grad_norm": 6.298330783843994,
      "learning_rate": 4.468345432844177e-05,
      "loss": 2.4649415969848634,
      "memory(GiB)": 77.56,
      "step": 62310,
      "token_acc": 0.5133333333333333,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 2.669765648429802,
      "grad_norm": 5.4456467628479,
      "learning_rate": 4.4676762756011974e-05,
      "loss": 2.4625049591064454,
      "memory(GiB)": 77.56,
      "step": 62315,
      "token_acc": 0.49146757679180886,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 2.6699798637590506,
      "grad_norm": 4.191608905792236,
      "learning_rate": 4.467007128001762e-05,
      "loss": 2.3561885833740233,
      "memory(GiB)": 77.56,
      "step": 62320,
      "token_acc": 0.5235294117647059,
      "train_speed(iter/s)": 1.438016
    },
    {
      "epoch": 2.6701940790882994,
      "grad_norm": 4.823205471038818,
      "learning_rate": 4.466337990057991e-05,
      "loss": 2.571339416503906,
      "memory(GiB)": 77.56,
      "step": 62325,
      "token_acc": 0.44281524926686217,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 2.6704082944175487,
      "grad_norm": 7.227558135986328,
      "learning_rate": 4.4656688617820095e-05,
      "loss": 2.380377006530762,
      "memory(GiB)": 77.56,
      "step": 62330,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 2.6706225097467975,
      "grad_norm": 4.9831156730651855,
      "learning_rate": 4.464999743185937e-05,
      "loss": 2.476052665710449,
      "memory(GiB)": 77.56,
      "step": 62335,
      "token_acc": 0.4923547400611621,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 2.6708367250760463,
      "grad_norm": 5.775318622589111,
      "learning_rate": 4.464330634281895e-05,
      "loss": 2.7013622283935548,
      "memory(GiB)": 77.56,
      "step": 62340,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.437985
    },
    {
      "epoch": 2.6710509404052956,
      "grad_norm": 6.507359504699707,
      "learning_rate": 4.4636615350820054e-05,
      "loss": 2.245701217651367,
      "memory(GiB)": 77.56,
      "step": 62345,
      "token_acc": 0.5232558139534884,
      "train_speed(iter/s)": 1.437959
    },
    {
      "epoch": 2.6712651557345444,
      "grad_norm": 5.102236747741699,
      "learning_rate": 4.462992445598392e-05,
      "loss": 2.418789863586426,
      "memory(GiB)": 77.56,
      "step": 62350,
      "token_acc": 0.49642857142857144,
      "train_speed(iter/s)": 1.437969
    },
    {
      "epoch": 2.671479371063793,
      "grad_norm": 5.49685525894165,
      "learning_rate": 4.462323365843174e-05,
      "loss": 2.376914405822754,
      "memory(GiB)": 77.56,
      "step": 62355,
      "token_acc": 0.5175097276264592,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 2.6716935863930424,
      "grad_norm": 5.1924848556518555,
      "learning_rate": 4.4616542958284725e-05,
      "loss": 2.1029449462890626,
      "memory(GiB)": 77.56,
      "step": 62360,
      "token_acc": 0.5611510791366906,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 2.6719078017222913,
      "grad_norm": 5.0350751876831055,
      "learning_rate": 4.460985235566409e-05,
      "loss": 2.2838520050048827,
      "memory(GiB)": 77.56,
      "step": 62365,
      "token_acc": 0.5190839694656488,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 2.67212201705154,
      "grad_norm": 4.944079875946045,
      "learning_rate": 4.4603161850691025e-05,
      "loss": 2.1518455505371095,
      "memory(GiB)": 77.56,
      "step": 62370,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 2.6723362323807893,
      "grad_norm": 4.947826862335205,
      "learning_rate": 4.459647144348675e-05,
      "loss": 2.3024925231933593,
      "memory(GiB)": 77.56,
      "step": 62375,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437987
    },
    {
      "epoch": 2.672550447710038,
      "grad_norm": 5.7744526863098145,
      "learning_rate": 4.458978113417248e-05,
      "loss": 2.4686527252197266,
      "memory(GiB)": 77.56,
      "step": 62380,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.437985
    },
    {
      "epoch": 2.672764663039287,
      "grad_norm": 6.173766613006592,
      "learning_rate": 4.4583090922869375e-05,
      "loss": 2.5134342193603514,
      "memory(GiB)": 77.56,
      "step": 62385,
      "token_acc": 0.44481605351170567,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 2.672978878368536,
      "grad_norm": 4.811943531036377,
      "learning_rate": 4.457640080969868e-05,
      "loss": 2.3879085540771485,
      "memory(GiB)": 77.56,
      "step": 62390,
      "token_acc": 0.46703296703296704,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 2.673193093697785,
      "grad_norm": 5.159696578979492,
      "learning_rate": 4.456971079478155e-05,
      "loss": 2.2175457000732424,
      "memory(GiB)": 77.56,
      "step": 62395,
      "token_acc": 0.5542635658914729,
      "train_speed(iter/s)": 1.437957
    },
    {
      "epoch": 2.673407309027034,
      "grad_norm": 4.60494327545166,
      "learning_rate": 4.456302087823922e-05,
      "loss": 2.1543155670166017,
      "memory(GiB)": 77.56,
      "step": 62400,
      "token_acc": 0.5233644859813084,
      "train_speed(iter/s)": 1.437961
    },
    {
      "epoch": 2.673621524356283,
      "grad_norm": 8.032392501831055,
      "learning_rate": 4.455633106019287e-05,
      "loss": 2.5719888687133787,
      "memory(GiB)": 77.56,
      "step": 62405,
      "token_acc": 0.43609022556390975,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 2.673835739685532,
      "grad_norm": 6.6925153732299805,
      "learning_rate": 4.4549641340763676e-05,
      "loss": 2.2387340545654295,
      "memory(GiB)": 77.56,
      "step": 62410,
      "token_acc": 0.5083333333333333,
      "train_speed(iter/s)": 1.437982
    },
    {
      "epoch": 2.6740499550147807,
      "grad_norm": 5.219361782073975,
      "learning_rate": 4.454295172007285e-05,
      "loss": 2.3546146392822265,
      "memory(GiB)": 77.56,
      "step": 62415,
      "token_acc": 0.4980694980694981,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 2.67426417034403,
      "grad_norm": 4.447869300842285,
      "learning_rate": 4.4536262198241555e-05,
      "loss": 2.368247222900391,
      "memory(GiB)": 77.56,
      "step": 62420,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437975
    },
    {
      "epoch": 2.674478385673279,
      "grad_norm": 5.033359527587891,
      "learning_rate": 4.4529572775391014e-05,
      "loss": 2.441751480102539,
      "memory(GiB)": 77.56,
      "step": 62425,
      "token_acc": 0.4576271186440678,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 2.6746926010025276,
      "grad_norm": 5.163084506988525,
      "learning_rate": 4.4522883451642386e-05,
      "loss": 2.264231491088867,
      "memory(GiB)": 77.56,
      "step": 62430,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437949
    },
    {
      "epoch": 2.674906816331777,
      "grad_norm": 5.881134510040283,
      "learning_rate": 4.451619422711687e-05,
      "loss": 2.434559631347656,
      "memory(GiB)": 77.56,
      "step": 62435,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.437942
    },
    {
      "epoch": 2.6751210316610257,
      "grad_norm": 5.882541179656982,
      "learning_rate": 4.4509505101935636e-05,
      "loss": 2.1831966400146485,
      "memory(GiB)": 77.56,
      "step": 62440,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 2.6753352469902745,
      "grad_norm": 6.888217926025391,
      "learning_rate": 4.450281607621987e-05,
      "loss": 2.733083724975586,
      "memory(GiB)": 77.56,
      "step": 62445,
      "token_acc": 0.4311594202898551,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 2.6755494623195237,
      "grad_norm": 4.711943626403809,
      "learning_rate": 4.449612715009075e-05,
      "loss": 2.7587135314941404,
      "memory(GiB)": 77.56,
      "step": 62450,
      "token_acc": 0.4517241379310345,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 2.6757636776487725,
      "grad_norm": 5.918360233306885,
      "learning_rate": 4.4489438323669435e-05,
      "loss": 2.2518150329589846,
      "memory(GiB)": 77.56,
      "step": 62455,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 2.6759778929780214,
      "grad_norm": 5.371405601501465,
      "learning_rate": 4.448274959707713e-05,
      "loss": 2.458642768859863,
      "memory(GiB)": 77.56,
      "step": 62460,
      "token_acc": 0.4539249146757679,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 2.6761921083072706,
      "grad_norm": 4.860164642333984,
      "learning_rate": 4.447606097043499e-05,
      "loss": 2.6802932739257814,
      "memory(GiB)": 77.56,
      "step": 62465,
      "token_acc": 0.48427672955974843,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 2.6764063236365194,
      "grad_norm": 7.362837791442871,
      "learning_rate": 4.4469372443864185e-05,
      "loss": 2.1227832794189454,
      "memory(GiB)": 77.56,
      "step": 62470,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.438027
    },
    {
      "epoch": 2.6766205389657682,
      "grad_norm": 5.705780506134033,
      "learning_rate": 4.4462684017485884e-05,
      "loss": 2.1604549407958986,
      "memory(GiB)": 77.56,
      "step": 62475,
      "token_acc": 0.5298245614035088,
      "train_speed(iter/s)": 1.438049
    },
    {
      "epoch": 2.6768347542950175,
      "grad_norm": 6.227565288543701,
      "learning_rate": 4.445599569142127e-05,
      "loss": 2.438285255432129,
      "memory(GiB)": 77.56,
      "step": 62480,
      "token_acc": 0.4894894894894895,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 2.6770489696242663,
      "grad_norm": 6.081276893615723,
      "learning_rate": 4.444930746579147e-05,
      "loss": 2.418336868286133,
      "memory(GiB)": 77.56,
      "step": 62485,
      "token_acc": 0.5173745173745173,
      "train_speed(iter/s)": 1.438046
    },
    {
      "epoch": 2.677263184953515,
      "grad_norm": 6.587379455566406,
      "learning_rate": 4.444261934071769e-05,
      "loss": 2.5576358795166017,
      "memory(GiB)": 77.56,
      "step": 62490,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.438074
    },
    {
      "epoch": 2.6774774002827644,
      "grad_norm": 4.905750274658203,
      "learning_rate": 4.443593131632105e-05,
      "loss": 2.571126937866211,
      "memory(GiB)": 77.56,
      "step": 62495,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 2.677691615612013,
      "grad_norm": 7.74746561050415,
      "learning_rate": 4.442924339272275e-05,
      "loss": 2.303331184387207,
      "memory(GiB)": 77.56,
      "step": 62500,
      "token_acc": 0.5340501792114696,
      "train_speed(iter/s)": 1.438088
    },
    {
      "epoch": 2.677691615612013,
      "eval_loss": 2.4737586975097656,
      "eval_runtime": 14.4274,
      "eval_samples_per_second": 6.931,
      "eval_steps_per_second": 6.931,
      "eval_token_acc": 0.45298013245033114,
      "step": 62500
    },
    {
      "epoch": 2.677905830941262,
      "grad_norm": 5.322314739227295,
      "learning_rate": 4.442255557004393e-05,
      "loss": 2.4849563598632813,
      "memory(GiB)": 77.56,
      "step": 62505,
      "token_acc": 0.4603024574669187,
      "train_speed(iter/s)": 1.437579
    },
    {
      "epoch": 2.6781200462705113,
      "grad_norm": 8.372267723083496,
      "learning_rate": 4.441586784840576e-05,
      "loss": 2.3722812652587892,
      "memory(GiB)": 77.56,
      "step": 62510,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.43759
    },
    {
      "epoch": 2.67833426159976,
      "grad_norm": 6.033510208129883,
      "learning_rate": 4.440918022792937e-05,
      "loss": 2.172313690185547,
      "memory(GiB)": 77.56,
      "step": 62515,
      "token_acc": 0.5551601423487544,
      "train_speed(iter/s)": 1.437601
    },
    {
      "epoch": 2.678548476929009,
      "grad_norm": 6.373674392700195,
      "learning_rate": 4.440249270873593e-05,
      "loss": 2.399858856201172,
      "memory(GiB)": 77.56,
      "step": 62520,
      "token_acc": 0.48757763975155277,
      "train_speed(iter/s)": 1.437611
    },
    {
      "epoch": 2.678762692258258,
      "grad_norm": 6.440563201904297,
      "learning_rate": 4.439580529094659e-05,
      "loss": 2.6191905975341796,
      "memory(GiB)": 77.56,
      "step": 62525,
      "token_acc": 0.46366782006920415,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 2.678976907587507,
      "grad_norm": 7.462244987487793,
      "learning_rate": 4.4389117974682484e-05,
      "loss": 2.4368354797363283,
      "memory(GiB)": 77.56,
      "step": 62530,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.437586
    },
    {
      "epoch": 2.6791911229167558,
      "grad_norm": 4.401798725128174,
      "learning_rate": 4.4382430760064774e-05,
      "loss": 2.2819873809814455,
      "memory(GiB)": 77.56,
      "step": 62535,
      "token_acc": 0.5191082802547771,
      "train_speed(iter/s)": 1.437604
    },
    {
      "epoch": 2.679405338246005,
      "grad_norm": 6.273928165435791,
      "learning_rate": 4.4375743647214596e-05,
      "loss": 2.8158090591430662,
      "memory(GiB)": 77.56,
      "step": 62540,
      "token_acc": 0.4397590361445783,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 2.679619553575254,
      "grad_norm": 9.880330085754395,
      "learning_rate": 4.436905663625311e-05,
      "loss": 2.5445882797241213,
      "memory(GiB)": 77.56,
      "step": 62545,
      "token_acc": 0.45112781954887216,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 2.6798337689045026,
      "grad_norm": 6.399016857147217,
      "learning_rate": 4.436236972730144e-05,
      "loss": 2.120593452453613,
      "memory(GiB)": 77.56,
      "step": 62550,
      "token_acc": 0.540084388185654,
      "train_speed(iter/s)": 1.437617
    },
    {
      "epoch": 2.680047984233752,
      "grad_norm": 4.407811164855957,
      "learning_rate": 4.435568292048072e-05,
      "loss": 2.6315799713134767,
      "memory(GiB)": 77.56,
      "step": 62555,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.437637
    },
    {
      "epoch": 2.6802621995630007,
      "grad_norm": 5.339437961578369,
      "learning_rate": 4.4348996215912114e-05,
      "loss": 2.5039394378662108,
      "memory(GiB)": 77.56,
      "step": 62560,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 2.6804764148922495,
      "grad_norm": 5.569267749786377,
      "learning_rate": 4.434230961371674e-05,
      "loss": 2.558920478820801,
      "memory(GiB)": 77.56,
      "step": 62565,
      "token_acc": 0.48546511627906974,
      "train_speed(iter/s)": 1.437614
    },
    {
      "epoch": 2.680690630221499,
      "grad_norm": 7.515854835510254,
      "learning_rate": 4.433562311401571e-05,
      "loss": 2.613993453979492,
      "memory(GiB)": 77.56,
      "step": 62570,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.437632
    },
    {
      "epoch": 2.6809048455507476,
      "grad_norm": 5.23411750793457,
      "learning_rate": 4.43289367169302e-05,
      "loss": 2.749032211303711,
      "memory(GiB)": 77.56,
      "step": 62575,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 2.6811190608799964,
      "grad_norm": 4.750558376312256,
      "learning_rate": 4.4322250422581326e-05,
      "loss": 2.3014715194702147,
      "memory(GiB)": 77.56,
      "step": 62580,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 2.6813332762092457,
      "grad_norm": 4.996963024139404,
      "learning_rate": 4.431556423109021e-05,
      "loss": 2.4371225357055666,
      "memory(GiB)": 77.56,
      "step": 62585,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.437661
    },
    {
      "epoch": 2.6815474915384945,
      "grad_norm": 4.832967758178711,
      "learning_rate": 4.430887814257798e-05,
      "loss": 2.337788391113281,
      "memory(GiB)": 77.56,
      "step": 62590,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.437659
    },
    {
      "epoch": 2.6817617068677437,
      "grad_norm": 7.261024475097656,
      "learning_rate": 4.430219215716576e-05,
      "loss": 2.333908462524414,
      "memory(GiB)": 77.56,
      "step": 62595,
      "token_acc": 0.5220883534136547,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 2.6819759221969925,
      "grad_norm": 5.736026763916016,
      "learning_rate": 4.429550627497467e-05,
      "loss": 2.5036090850830077,
      "memory(GiB)": 77.56,
      "step": 62600,
      "token_acc": 0.46863468634686345,
      "train_speed(iter/s)": 1.437644
    },
    {
      "epoch": 2.6821901375262414,
      "grad_norm": 5.654018878936768,
      "learning_rate": 4.428882049612584e-05,
      "loss": 2.2484457015991213,
      "memory(GiB)": 77.56,
      "step": 62605,
      "token_acc": 0.5345454545454545,
      "train_speed(iter/s)": 1.437623
    },
    {
      "epoch": 2.6824043528554906,
      "grad_norm": 5.278985500335693,
      "learning_rate": 4.428213482074039e-05,
      "loss": 2.1761598587036133,
      "memory(GiB)": 77.56,
      "step": 62610,
      "token_acc": 0.5373134328358209,
      "train_speed(iter/s)": 1.437603
    },
    {
      "epoch": 2.6826185681847394,
      "grad_norm": 6.985358715057373,
      "learning_rate": 4.427544924893941e-05,
      "loss": 2.569799041748047,
      "memory(GiB)": 77.56,
      "step": 62615,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.437617
    },
    {
      "epoch": 2.6828327835139882,
      "grad_norm": 5.450619697570801,
      "learning_rate": 4.426876378084406e-05,
      "loss": 2.458124542236328,
      "memory(GiB)": 77.56,
      "step": 62620,
      "token_acc": 0.4745222929936306,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 2.6830469988432375,
      "grad_norm": 6.4383931159973145,
      "learning_rate": 4.426207841657543e-05,
      "loss": 1.9995841979980469,
      "memory(GiB)": 77.56,
      "step": 62625,
      "token_acc": 0.5748031496062992,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 2.6832612141724863,
      "grad_norm": 5.315460681915283,
      "learning_rate": 4.425539315625462e-05,
      "loss": 2.5346023559570314,
      "memory(GiB)": 77.56,
      "step": 62630,
      "token_acc": 0.45864661654135336,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 2.683475429501735,
      "grad_norm": 5.402333736419678,
      "learning_rate": 4.424870800000276e-05,
      "loss": 2.337423324584961,
      "memory(GiB)": 77.56,
      "step": 62635,
      "token_acc": 0.5230125523012552,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 2.6836896448309844,
      "grad_norm": 4.58209228515625,
      "learning_rate": 4.424202294794093e-05,
      "loss": 2.636511039733887,
      "memory(GiB)": 77.56,
      "step": 62640,
      "token_acc": 0.46794871794871795,
      "train_speed(iter/s)": 1.437677
    },
    {
      "epoch": 2.683903860160233,
      "grad_norm": 5.8528151512146,
      "learning_rate": 4.423533800019026e-05,
      "loss": 2.838317108154297,
      "memory(GiB)": 77.56,
      "step": 62645,
      "token_acc": 0.430635838150289,
      "train_speed(iter/s)": 1.437664
    },
    {
      "epoch": 2.684118075489482,
      "grad_norm": 7.27495813369751,
      "learning_rate": 4.422865315687187e-05,
      "loss": 2.3046009063720705,
      "memory(GiB)": 77.56,
      "step": 62650,
      "token_acc": 0.5239852398523985,
      "train_speed(iter/s)": 1.437682
    },
    {
      "epoch": 2.6843322908187313,
      "grad_norm": 5.070964336395264,
      "learning_rate": 4.4221968418106844e-05,
      "loss": 2.4182600021362304,
      "memory(GiB)": 77.56,
      "step": 62655,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.437696
    },
    {
      "epoch": 2.68454650614798,
      "grad_norm": 6.140015125274658,
      "learning_rate": 4.421528378401626e-05,
      "loss": 2.3498199462890623,
      "memory(GiB)": 77.56,
      "step": 62660,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 2.684760721477229,
      "grad_norm": 4.644944667816162,
      "learning_rate": 4.420859925472125e-05,
      "loss": 2.1466753005981447,
      "memory(GiB)": 77.56,
      "step": 62665,
      "token_acc": 0.5344827586206896,
      "train_speed(iter/s)": 1.437726
    },
    {
      "epoch": 2.684974936806478,
      "grad_norm": 5.87173318862915,
      "learning_rate": 4.42019148303429e-05,
      "loss": 2.2788885116577147,
      "memory(GiB)": 77.56,
      "step": 62670,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.43773
    },
    {
      "epoch": 2.685189152135727,
      "grad_norm": 4.484790325164795,
      "learning_rate": 4.419523051100229e-05,
      "loss": 2.597868537902832,
      "memory(GiB)": 77.56,
      "step": 62675,
      "token_acc": 0.46397694524495675,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 2.6854033674649758,
      "grad_norm": 5.3702712059021,
      "learning_rate": 4.418854629682053e-05,
      "loss": 2.550777816772461,
      "memory(GiB)": 77.56,
      "step": 62680,
      "token_acc": 0.44654088050314467,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 2.685617582794225,
      "grad_norm": 6.263823986053467,
      "learning_rate": 4.41818621879187e-05,
      "loss": 2.556705665588379,
      "memory(GiB)": 77.56,
      "step": 62685,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 2.685831798123474,
      "grad_norm": 6.265450954437256,
      "learning_rate": 4.41751781844179e-05,
      "loss": 2.4886608123779297,
      "memory(GiB)": 77.56,
      "step": 62690,
      "token_acc": 0.5015015015015015,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 2.6860460134527226,
      "grad_norm": 4.281060218811035,
      "learning_rate": 4.416849428643922e-05,
      "loss": 2.6439865112304686,
      "memory(GiB)": 77.56,
      "step": 62695,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437777
    },
    {
      "epoch": 2.686260228781972,
      "grad_norm": 5.494462013244629,
      "learning_rate": 4.416181049410372e-05,
      "loss": 2.4629398345947267,
      "memory(GiB)": 77.56,
      "step": 62700,
      "token_acc": 0.49606299212598426,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 2.6864744441112207,
      "grad_norm": 5.727767467498779,
      "learning_rate": 4.415512680753251e-05,
      "loss": 2.2903635025024416,
      "memory(GiB)": 77.56,
      "step": 62705,
      "token_acc": 0.5096525096525096,
      "train_speed(iter/s)": 1.437807
    },
    {
      "epoch": 2.6866886594404695,
      "grad_norm": 5.721604824066162,
      "learning_rate": 4.414844322684667e-05,
      "loss": 2.3975643157958983,
      "memory(GiB)": 77.56,
      "step": 62710,
      "token_acc": 0.4646840148698885,
      "train_speed(iter/s)": 1.437816
    },
    {
      "epoch": 2.686902874769719,
      "grad_norm": 5.7855305671691895,
      "learning_rate": 4.414175975216724e-05,
      "loss": 2.6450050354003904,
      "memory(GiB)": 77.56,
      "step": 62715,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.437826
    },
    {
      "epoch": 2.6871170900989676,
      "grad_norm": 6.601729393005371,
      "learning_rate": 4.413507638361534e-05,
      "loss": 2.285299873352051,
      "memory(GiB)": 77.56,
      "step": 62720,
      "token_acc": 0.48639455782312924,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 2.6873313054282164,
      "grad_norm": 5.127752304077148,
      "learning_rate": 4.412839312131204e-05,
      "loss": 2.51846866607666,
      "memory(GiB)": 77.56,
      "step": 62725,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 2.6875455207574657,
      "grad_norm": 5.674526214599609,
      "learning_rate": 4.41217099653784e-05,
      "loss": 2.241515350341797,
      "memory(GiB)": 77.56,
      "step": 62730,
      "token_acc": 0.5180327868852459,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 2.6877597360867145,
      "grad_norm": 5.848512172698975,
      "learning_rate": 4.411502691593551e-05,
      "loss": 2.693959426879883,
      "memory(GiB)": 77.56,
      "step": 62735,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 2.6879739514159633,
      "grad_norm": 4.746672630310059,
      "learning_rate": 4.410834397310443e-05,
      "loss": 2.333148193359375,
      "memory(GiB)": 77.56,
      "step": 62740,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 2.6881881667452125,
      "grad_norm": 5.210576057434082,
      "learning_rate": 4.410166113700621e-05,
      "loss": 2.4908187866210936,
      "memory(GiB)": 77.56,
      "step": 62745,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 2.6884023820744614,
      "grad_norm": 7.094858169555664,
      "learning_rate": 4.4094978407761936e-05,
      "loss": 2.3643184661865235,
      "memory(GiB)": 77.56,
      "step": 62750,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.437833
    },
    {
      "epoch": 2.68861659740371,
      "grad_norm": 4.565588474273682,
      "learning_rate": 4.408829578549268e-05,
      "loss": 2.207187461853027,
      "memory(GiB)": 77.56,
      "step": 62755,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.437835
    },
    {
      "epoch": 2.6888308127329594,
      "grad_norm": 7.889364719390869,
      "learning_rate": 4.4081613270319476e-05,
      "loss": 2.643632698059082,
      "memory(GiB)": 77.56,
      "step": 62760,
      "token_acc": 0.46564885496183206,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 2.6890450280622082,
      "grad_norm": 5.547224521636963,
      "learning_rate": 4.407493086236341e-05,
      "loss": 2.2921451568603515,
      "memory(GiB)": 77.56,
      "step": 62765,
      "token_acc": 0.464,
      "train_speed(iter/s)": 1.437825
    },
    {
      "epoch": 2.689259243391457,
      "grad_norm": 3.637024402618408,
      "learning_rate": 4.406824856174552e-05,
      "loss": 2.4423154830932616,
      "memory(GiB)": 77.56,
      "step": 62770,
      "token_acc": 0.49193548387096775,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 2.6894734587207063,
      "grad_norm": 6.107388019561768,
      "learning_rate": 4.406156636858688e-05,
      "loss": 2.5040639877319335,
      "memory(GiB)": 77.56,
      "step": 62775,
      "token_acc": 0.447098976109215,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 2.689687674049955,
      "grad_norm": 5.72420597076416,
      "learning_rate": 4.4054884283008534e-05,
      "loss": 2.4143774032592775,
      "memory(GiB)": 77.56,
      "step": 62780,
      "token_acc": 0.4813664596273292,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 2.689901889379204,
      "grad_norm": 4.765916347503662,
      "learning_rate": 4.404820230513153e-05,
      "loss": 2.19256591796875,
      "memory(GiB)": 77.56,
      "step": 62785,
      "token_acc": 0.5038461538461538,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 2.690116104708453,
      "grad_norm": 5.38768196105957,
      "learning_rate": 4.404152043507692e-05,
      "loss": 2.513363075256348,
      "memory(GiB)": 77.56,
      "step": 62790,
      "token_acc": 0.4673202614379085,
      "train_speed(iter/s)": 1.437861
    },
    {
      "epoch": 2.690330320037702,
      "grad_norm": 4.33722448348999,
      "learning_rate": 4.4034838672965764e-05,
      "loss": 2.576100730895996,
      "memory(GiB)": 77.56,
      "step": 62795,
      "token_acc": 0.48125,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 2.690544535366951,
      "grad_norm": 5.493600845336914,
      "learning_rate": 4.4028157018919106e-05,
      "loss": 2.598653793334961,
      "memory(GiB)": 77.56,
      "step": 62800,
      "token_acc": 0.4637223974763407,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 2.6907587506962,
      "grad_norm": 5.1063690185546875,
      "learning_rate": 4.4021475473057984e-05,
      "loss": 2.5322696685791017,
      "memory(GiB)": 77.56,
      "step": 62805,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 2.690972966025449,
      "grad_norm": 4.514760494232178,
      "learning_rate": 4.401479403550344e-05,
      "loss": 2.22237606048584,
      "memory(GiB)": 77.56,
      "step": 62810,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 2.6911871813546977,
      "grad_norm": 4.8120856285095215,
      "learning_rate": 4.400811270637652e-05,
      "loss": 2.5820423126220704,
      "memory(GiB)": 77.56,
      "step": 62815,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 2.691401396683947,
      "grad_norm": 5.443692207336426,
      "learning_rate": 4.400143148579826e-05,
      "loss": 2.8748239517211913,
      "memory(GiB)": 77.56,
      "step": 62820,
      "token_acc": 0.43973941368078173,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 2.6916156120131958,
      "grad_norm": 5.166743755340576,
      "learning_rate": 4.399475037388969e-05,
      "loss": 2.3774322509765624,
      "memory(GiB)": 77.56,
      "step": 62825,
      "token_acc": 0.4778481012658228,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 2.6918298273424446,
      "grad_norm": 5.580989837646484,
      "learning_rate": 4.398806937077185e-05,
      "loss": 2.3642290115356444,
      "memory(GiB)": 77.56,
      "step": 62830,
      "token_acc": 0.49050632911392406,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 2.692044042671694,
      "grad_norm": 5.6126322746276855,
      "learning_rate": 4.398138847656578e-05,
      "loss": 2.585008239746094,
      "memory(GiB)": 77.56,
      "step": 62835,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.437934
    },
    {
      "epoch": 2.6922582580009427,
      "grad_norm": 6.069372177124023,
      "learning_rate": 4.3974707691392503e-05,
      "loss": 2.3440319061279298,
      "memory(GiB)": 77.56,
      "step": 62840,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 2.6924724733301915,
      "grad_norm": 4.513391494750977,
      "learning_rate": 4.396802701537304e-05,
      "loss": 2.3978302001953127,
      "memory(GiB)": 77.56,
      "step": 62845,
      "token_acc": 0.4784172661870504,
      "train_speed(iter/s)": 1.437948
    },
    {
      "epoch": 2.6926866886594407,
      "grad_norm": 6.225036144256592,
      "learning_rate": 4.396134644862844e-05,
      "loss": 2.280904769897461,
      "memory(GiB)": 77.56,
      "step": 62850,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.437944
    },
    {
      "epoch": 2.6929009039886895,
      "grad_norm": 4.990926742553711,
      "learning_rate": 4.395466599127971e-05,
      "loss": 2.6861850738525392,
      "memory(GiB)": 77.56,
      "step": 62855,
      "token_acc": 0.4805194805194805,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 2.6931151193179383,
      "grad_norm": 5.945162296295166,
      "learning_rate": 4.3947985643447866e-05,
      "loss": 2.314472770690918,
      "memory(GiB)": 77.56,
      "step": 62860,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 2.6933293346471876,
      "grad_norm": 4.823639869689941,
      "learning_rate": 4.394130540525392e-05,
      "loss": 2.3924217224121094,
      "memory(GiB)": 77.56,
      "step": 62865,
      "token_acc": 0.4984423676012461,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 2.6935435499764364,
      "grad_norm": 4.460719108581543,
      "learning_rate": 4.393462527681894e-05,
      "loss": 2.4931190490722654,
      "memory(GiB)": 77.56,
      "step": 62870,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 2.6937577653056852,
      "grad_norm": 5.305558681488037,
      "learning_rate": 4.39279452582639e-05,
      "loss": 2.4298730850219727,
      "memory(GiB)": 77.56,
      "step": 62875,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 2.6939719806349345,
      "grad_norm": 4.590268611907959,
      "learning_rate": 4.3921265349709844e-05,
      "loss": 2.3436609268188477,
      "memory(GiB)": 77.56,
      "step": 62880,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.437967
    },
    {
      "epoch": 2.6941861959641833,
      "grad_norm": 5.058292865753174,
      "learning_rate": 4.391458555127777e-05,
      "loss": 2.3872493743896483,
      "memory(GiB)": 77.56,
      "step": 62885,
      "token_acc": 0.4855072463768116,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 2.694400411293432,
      "grad_norm": 9.457488059997559,
      "learning_rate": 4.390790586308867e-05,
      "loss": 2.352023696899414,
      "memory(GiB)": 77.56,
      "step": 62890,
      "token_acc": 0.5131578947368421,
      "train_speed(iter/s)": 1.437975
    },
    {
      "epoch": 2.6946146266226814,
      "grad_norm": 6.278934001922607,
      "learning_rate": 4.390122628526358e-05,
      "loss": 2.2947227478027346,
      "memory(GiB)": 77.56,
      "step": 62895,
      "token_acc": 0.48757763975155277,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 2.69482884195193,
      "grad_norm": 5.026858806610107,
      "learning_rate": 4.38945468179235e-05,
      "loss": 2.3993967056274412,
      "memory(GiB)": 77.56,
      "step": 62900,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.437997
    },
    {
      "epoch": 2.695043057281179,
      "grad_norm": 7.493849754333496,
      "learning_rate": 4.3887867461189416e-05,
      "loss": 2.624139976501465,
      "memory(GiB)": 77.56,
      "step": 62905,
      "token_acc": 0.47876447876447875,
      "train_speed(iter/s)": 1.438004
    },
    {
      "epoch": 2.6952572726104282,
      "grad_norm": 7.232560634613037,
      "learning_rate": 4.388118821518236e-05,
      "loss": 2.421573829650879,
      "memory(GiB)": 77.56,
      "step": 62910,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.438039
    },
    {
      "epoch": 2.695471487939677,
      "grad_norm": 5.298771381378174,
      "learning_rate": 4.3874509080023315e-05,
      "loss": 2.3434898376464846,
      "memory(GiB)": 77.56,
      "step": 62915,
      "token_acc": 0.5159010600706714,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 2.695685703268926,
      "grad_norm": 5.103635787963867,
      "learning_rate": 4.3867830055833284e-05,
      "loss": 2.5847805023193358,
      "memory(GiB)": 77.56,
      "step": 62920,
      "token_acc": 0.48286604361370716,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 2.695899918598175,
      "grad_norm": 6.296849250793457,
      "learning_rate": 4.386115114273328e-05,
      "loss": 2.135004234313965,
      "memory(GiB)": 77.56,
      "step": 62925,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 2.696114133927424,
      "grad_norm": 4.00954008102417,
      "learning_rate": 4.385447234084426e-05,
      "loss": 2.176651191711426,
      "memory(GiB)": 77.56,
      "step": 62930,
      "token_acc": 0.5444444444444444,
      "train_speed(iter/s)": 1.438044
    },
    {
      "epoch": 2.6963283492566728,
      "grad_norm": 5.4145121574401855,
      "learning_rate": 4.384779365028722e-05,
      "loss": 2.4199140548706053,
      "memory(GiB)": 77.56,
      "step": 62935,
      "token_acc": 0.49642857142857144,
      "train_speed(iter/s)": 1.438062
    },
    {
      "epoch": 2.696542564585922,
      "grad_norm": 6.146902561187744,
      "learning_rate": 4.38411150711832e-05,
      "loss": 2.628597640991211,
      "memory(GiB)": 77.56,
      "step": 62940,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 2.696756779915171,
      "grad_norm": 4.526706695556641,
      "learning_rate": 4.383443660365316e-05,
      "loss": 2.359404754638672,
      "memory(GiB)": 77.56,
      "step": 62945,
      "token_acc": 0.4707692307692308,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 2.6969709952444196,
      "grad_norm": 5.5696821212768555,
      "learning_rate": 4.3827758247818075e-05,
      "loss": 2.440283012390137,
      "memory(GiB)": 77.56,
      "step": 62950,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 2.697185210573669,
      "grad_norm": 5.297787666320801,
      "learning_rate": 4.382108000379894e-05,
      "loss": 2.4412647247314454,
      "memory(GiB)": 77.56,
      "step": 62955,
      "token_acc": 0.5095057034220533,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 2.6973994259029177,
      "grad_norm": 5.234545707702637,
      "learning_rate": 4.381440187171675e-05,
      "loss": 2.5680448532104494,
      "memory(GiB)": 77.56,
      "step": 62960,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 2.6976136412321665,
      "grad_norm": 5.247034549713135,
      "learning_rate": 4.380772385169245e-05,
      "loss": 2.402149963378906,
      "memory(GiB)": 77.56,
      "step": 62965,
      "token_acc": 0.4703703703703704,
      "train_speed(iter/s)": 1.438046
    },
    {
      "epoch": 2.6978278565614158,
      "grad_norm": 6.02849817276001,
      "learning_rate": 4.3801045943847064e-05,
      "loss": 2.156708526611328,
      "memory(GiB)": 77.56,
      "step": 62970,
      "token_acc": 0.5354609929078015,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 2.6980420718906646,
      "grad_norm": 5.664326190948486,
      "learning_rate": 4.3794368148301525e-05,
      "loss": 2.359878730773926,
      "memory(GiB)": 77.56,
      "step": 62975,
      "token_acc": 0.48036253776435045,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 2.6982562872199134,
      "grad_norm": 5.418706893920898,
      "learning_rate": 4.378769046517685e-05,
      "loss": 2.3388364791870115,
      "memory(GiB)": 77.56,
      "step": 62980,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438044
    },
    {
      "epoch": 2.6984705025491627,
      "grad_norm": 5.228435039520264,
      "learning_rate": 4.3781012894593975e-05,
      "loss": 2.275046157836914,
      "memory(GiB)": 77.56,
      "step": 62985,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.438074
    },
    {
      "epoch": 2.6986847178784115,
      "grad_norm": 6.260488510131836,
      "learning_rate": 4.377433543667388e-05,
      "loss": 2.421467590332031,
      "memory(GiB)": 77.56,
      "step": 62990,
      "token_acc": 0.5175718849840255,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 2.6988989332076603,
      "grad_norm": 5.846724987030029,
      "learning_rate": 4.376765809153755e-05,
      "loss": 2.5110095977783202,
      "memory(GiB)": 77.56,
      "step": 62995,
      "token_acc": 0.4748427672955975,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 2.6991131485369095,
      "grad_norm": 5.498667240142822,
      "learning_rate": 4.376098085930594e-05,
      "loss": 2.3920913696289063,
      "memory(GiB)": 77.56,
      "step": 63000,
      "token_acc": 0.4968553459119497,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 2.6991131485369095,
      "eval_loss": 2.1162877082824707,
      "eval_runtime": 13.5205,
      "eval_samples_per_second": 7.396,
      "eval_steps_per_second": 7.396,
      "eval_token_acc": 0.47831632653061223,
      "step": 63000
    },
    {
      "epoch": 2.6993273638661583,
      "grad_norm": 8.477178573608398,
      "learning_rate": 4.375430374010001e-05,
      "loss": 2.34906005859375,
      "memory(GiB)": 77.56,
      "step": 63005,
      "token_acc": 0.491362763915547,
      "train_speed(iter/s)": 1.437637
    },
    {
      "epoch": 2.699541579195407,
      "grad_norm": 5.616421222686768,
      "learning_rate": 4.3747626734040716e-05,
      "loss": 2.306650924682617,
      "memory(GiB)": 77.56,
      "step": 63010,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.437622
    },
    {
      "epoch": 2.6997557945246564,
      "grad_norm": 3.6965794563293457,
      "learning_rate": 4.374094984124904e-05,
      "loss": 2.4355173110961914,
      "memory(GiB)": 77.56,
      "step": 63015,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.437606
    },
    {
      "epoch": 2.6999700098539052,
      "grad_norm": 5.530742168426514,
      "learning_rate": 4.3734273061845926e-05,
      "loss": 2.4598361968994142,
      "memory(GiB)": 77.56,
      "step": 63020,
      "token_acc": 0.5095541401273885,
      "train_speed(iter/s)": 1.437609
    },
    {
      "epoch": 2.700184225183154,
      "grad_norm": 6.5386834144592285,
      "learning_rate": 4.372759639595234e-05,
      "loss": 2.47520751953125,
      "memory(GiB)": 77.56,
      "step": 63025,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 2.7003984405124033,
      "grad_norm": 4.548661231994629,
      "learning_rate": 4.3720919843689236e-05,
      "loss": 2.329376220703125,
      "memory(GiB)": 77.56,
      "step": 63030,
      "token_acc": 0.5214285714285715,
      "train_speed(iter/s)": 1.437605
    },
    {
      "epoch": 2.700612655841652,
      "grad_norm": 4.839725494384766,
      "learning_rate": 4.371424340517754e-05,
      "loss": 2.5615768432617188,
      "memory(GiB)": 77.56,
      "step": 63035,
      "token_acc": 0.4638888888888889,
      "train_speed(iter/s)": 1.437604
    },
    {
      "epoch": 2.700826871170901,
      "grad_norm": 5.1556267738342285,
      "learning_rate": 4.3707567080538235e-05,
      "loss": 2.4451314926147463,
      "memory(GiB)": 77.56,
      "step": 63040,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 2.70104108650015,
      "grad_norm": 6.217810153961182,
      "learning_rate": 4.370089086989225e-05,
      "loss": 2.330446243286133,
      "memory(GiB)": 77.56,
      "step": 63045,
      "token_acc": 0.512280701754386,
      "train_speed(iter/s)": 1.437637
    },
    {
      "epoch": 2.701255301829399,
      "grad_norm": 4.29569149017334,
      "learning_rate": 4.369421477336054e-05,
      "loss": 2.2924964904785154,
      "memory(GiB)": 77.56,
      "step": 63050,
      "token_acc": 0.5145631067961165,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 2.701469517158648,
      "grad_norm": 4.126863956451416,
      "learning_rate": 4.368753879106404e-05,
      "loss": 2.3577011108398436,
      "memory(GiB)": 77.56,
      "step": 63055,
      "token_acc": 0.4790996784565916,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 2.701683732487897,
      "grad_norm": 5.330383777618408,
      "learning_rate": 4.368086292312369e-05,
      "loss": 2.2391632080078123,
      "memory(GiB)": 77.56,
      "step": 63060,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 2.701897947817146,
      "grad_norm": 5.941152095794678,
      "learning_rate": 4.367418716966045e-05,
      "loss": 2.3233346939086914,
      "memory(GiB)": 77.56,
      "step": 63065,
      "token_acc": 0.5346153846153846,
      "train_speed(iter/s)": 1.43761
    },
    {
      "epoch": 2.7021121631463947,
      "grad_norm": 4.539067268371582,
      "learning_rate": 4.366751153079525e-05,
      "loss": 2.4829776763916014,
      "memory(GiB)": 77.56,
      "step": 63070,
      "token_acc": 0.4592274678111588,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 2.702326378475644,
      "grad_norm": 5.581842422485352,
      "learning_rate": 4.3660836006649e-05,
      "loss": 2.586964797973633,
      "memory(GiB)": 77.56,
      "step": 63075,
      "token_acc": 0.453125,
      "train_speed(iter/s)": 1.437636
    },
    {
      "epoch": 2.7025405938048928,
      "grad_norm": 5.822049140930176,
      "learning_rate": 4.365416059734266e-05,
      "loss": 2.448215675354004,
      "memory(GiB)": 77.56,
      "step": 63080,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 2.7027548091341416,
      "grad_norm": 5.618698596954346,
      "learning_rate": 4.364748530299714e-05,
      "loss": 2.464839744567871,
      "memory(GiB)": 77.56,
      "step": 63085,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 2.702969024463391,
      "grad_norm": 6.2707390785217285,
      "learning_rate": 4.364081012373339e-05,
      "loss": 2.436688041687012,
      "memory(GiB)": 77.56,
      "step": 63090,
      "token_acc": 0.4828571428571429,
      "train_speed(iter/s)": 1.437654
    },
    {
      "epoch": 2.7031832397926396,
      "grad_norm": 5.044200420379639,
      "learning_rate": 4.363413505967233e-05,
      "loss": 2.5676612854003906,
      "memory(GiB)": 77.56,
      "step": 63095,
      "token_acc": 0.4557377049180328,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 2.7033974551218884,
      "grad_norm": 5.146721839904785,
      "learning_rate": 4.36274601109349e-05,
      "loss": 2.6601850509643556,
      "memory(GiB)": 77.56,
      "step": 63100,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 2.7036116704511377,
      "grad_norm": 6.288225173950195,
      "learning_rate": 4.3620785277642004e-05,
      "loss": 2.7035140991210938,
      "memory(GiB)": 77.56,
      "step": 63105,
      "token_acc": 0.44155844155844154,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 2.7038258857803865,
      "grad_norm": 5.237624168395996,
      "learning_rate": 4.3614110559914555e-05,
      "loss": 2.430165481567383,
      "memory(GiB)": 77.56,
      "step": 63110,
      "token_acc": 0.5209003215434084,
      "train_speed(iter/s)": 1.437623
    },
    {
      "epoch": 2.7040401011096353,
      "grad_norm": 6.375870704650879,
      "learning_rate": 4.360743595787349e-05,
      "loss": 2.487293243408203,
      "memory(GiB)": 77.56,
      "step": 63115,
      "token_acc": 0.48632218844984804,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 2.7042543164388846,
      "grad_norm": 6.199286460876465,
      "learning_rate": 4.360076147163972e-05,
      "loss": 2.5808658599853516,
      "memory(GiB)": 77.56,
      "step": 63120,
      "token_acc": 0.4533762057877814,
      "train_speed(iter/s)": 1.43763
    },
    {
      "epoch": 2.7044685317681334,
      "grad_norm": 5.016519069671631,
      "learning_rate": 4.3594087101334164e-05,
      "loss": 2.2785139083862305,
      "memory(GiB)": 77.56,
      "step": 63125,
      "token_acc": 0.517799352750809,
      "train_speed(iter/s)": 1.437596
    },
    {
      "epoch": 2.704682747097382,
      "grad_norm": 5.4700117111206055,
      "learning_rate": 4.3587412847077726e-05,
      "loss": 2.3348194122314454,
      "memory(GiB)": 77.56,
      "step": 63130,
      "token_acc": 0.5053763440860215,
      "train_speed(iter/s)": 1.437598
    },
    {
      "epoch": 2.7048969624266315,
      "grad_norm": 5.129421710968018,
      "learning_rate": 4.358073870899131e-05,
      "loss": 2.6222465515136717,
      "memory(GiB)": 77.56,
      "step": 63135,
      "token_acc": 0.47904191616766467,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 2.7051111777558803,
      "grad_norm": 7.872922420501709,
      "learning_rate": 4.3574064687195846e-05,
      "loss": 2.609187698364258,
      "memory(GiB)": 77.56,
      "step": 63140,
      "token_acc": 0.47653429602888087,
      "train_speed(iter/s)": 1.437609
    },
    {
      "epoch": 2.705325393085129,
      "grad_norm": 5.309908390045166,
      "learning_rate": 4.356739078181223e-05,
      "loss": 2.4413629531860352,
      "memory(GiB)": 77.56,
      "step": 63145,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 2.7055396084143783,
      "grad_norm": 5.252475738525391,
      "learning_rate": 4.356071699296135e-05,
      "loss": 2.2228572845458983,
      "memory(GiB)": 77.56,
      "step": 63150,
      "token_acc": 0.5584905660377358,
      "train_speed(iter/s)": 1.437575
    },
    {
      "epoch": 2.705753823743627,
      "grad_norm": 6.663409233093262,
      "learning_rate": 4.3554043320764134e-05,
      "loss": 2.2703807830810545,
      "memory(GiB)": 77.56,
      "step": 63155,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.437585
    },
    {
      "epoch": 2.705968039072876,
      "grad_norm": 5.493904113769531,
      "learning_rate": 4.354736976534145e-05,
      "loss": 2.488634490966797,
      "memory(GiB)": 77.56,
      "step": 63160,
      "token_acc": 0.47794117647058826,
      "train_speed(iter/s)": 1.437586
    },
    {
      "epoch": 2.7061822544021252,
      "grad_norm": 6.335967063903809,
      "learning_rate": 4.354069632681423e-05,
      "loss": 2.3626171112060548,
      "memory(GiB)": 77.56,
      "step": 63165,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.43762
    },
    {
      "epoch": 2.706396469731374,
      "grad_norm": 7.072081089019775,
      "learning_rate": 4.353402300530336e-05,
      "loss": 2.2816446304321287,
      "memory(GiB)": 77.56,
      "step": 63170,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 2.706610685060623,
      "grad_norm": 4.870462894439697,
      "learning_rate": 4.352734980092973e-05,
      "loss": 2.2714717864990233,
      "memory(GiB)": 77.56,
      "step": 63175,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 2.706824900389872,
      "grad_norm": 5.844668388366699,
      "learning_rate": 4.352067671381422e-05,
      "loss": 2.396867561340332,
      "memory(GiB)": 77.56,
      "step": 63180,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.437596
    },
    {
      "epoch": 2.707039115719121,
      "grad_norm": 4.8983283042907715,
      "learning_rate": 4.3514003744077745e-05,
      "loss": 2.678304672241211,
      "memory(GiB)": 77.56,
      "step": 63185,
      "token_acc": 0.45224719101123595,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 2.7072533310483697,
      "grad_norm": 7.347422122955322,
      "learning_rate": 4.350733089184117e-05,
      "loss": 2.2833236694335937,
      "memory(GiB)": 77.56,
      "step": 63190,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.437559
    },
    {
      "epoch": 2.707467546377619,
      "grad_norm": 7.245707988739014,
      "learning_rate": 4.3500658157225375e-05,
      "loss": 2.4425127029418947,
      "memory(GiB)": 77.56,
      "step": 63195,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.437576
    },
    {
      "epoch": 2.707681761706868,
      "grad_norm": 5.923473358154297,
      "learning_rate": 4.3493985540351265e-05,
      "loss": 2.3449058532714844,
      "memory(GiB)": 77.56,
      "step": 63200,
      "token_acc": 0.49712643678160917,
      "train_speed(iter/s)": 1.437592
    },
    {
      "epoch": 2.7078959770361166,
      "grad_norm": 6.155705451965332,
      "learning_rate": 4.34873130413397e-05,
      "loss": 2.5253669738769533,
      "memory(GiB)": 77.56,
      "step": 63205,
      "token_acc": 0.4519230769230769,
      "train_speed(iter/s)": 1.437598
    },
    {
      "epoch": 2.708110192365366,
      "grad_norm": 6.043651580810547,
      "learning_rate": 4.348064066031159e-05,
      "loss": 2.5645116806030273,
      "memory(GiB)": 77.56,
      "step": 63210,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 2.7083244076946147,
      "grad_norm": 5.626470565795898,
      "learning_rate": 4.3473968397387774e-05,
      "loss": 2.606903839111328,
      "memory(GiB)": 77.56,
      "step": 63215,
      "token_acc": 0.44755244755244755,
      "train_speed(iter/s)": 1.437589
    },
    {
      "epoch": 2.7085386230238635,
      "grad_norm": 4.792459487915039,
      "learning_rate": 4.3467296252689144e-05,
      "loss": 2.524639892578125,
      "memory(GiB)": 77.56,
      "step": 63220,
      "token_acc": 0.4231974921630094,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 2.7087528383531128,
      "grad_norm": 5.940973281860352,
      "learning_rate": 4.3460624226336576e-05,
      "loss": 2.463256072998047,
      "memory(GiB)": 77.56,
      "step": 63225,
      "token_acc": 0.4849624060150376,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 2.7089670536823616,
      "grad_norm": 6.710156440734863,
      "learning_rate": 4.345395231845094e-05,
      "loss": 2.6621768951416014,
      "memory(GiB)": 77.56,
      "step": 63230,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 2.7091812690116104,
      "grad_norm": 7.585979461669922,
      "learning_rate": 4.344728052915307e-05,
      "loss": 2.262676239013672,
      "memory(GiB)": 77.56,
      "step": 63235,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 2.7093954843408596,
      "grad_norm": 5.6849236488342285,
      "learning_rate": 4.344060885856387e-05,
      "loss": 2.1216243743896483,
      "memory(GiB)": 77.56,
      "step": 63240,
      "token_acc": 0.5272727272727272,
      "train_speed(iter/s)": 1.437587
    },
    {
      "epoch": 2.7096096996701085,
      "grad_norm": 4.63074254989624,
      "learning_rate": 4.343393730680421e-05,
      "loss": 2.4104248046875,
      "memory(GiB)": 77.56,
      "step": 63245,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 2.7098239149993573,
      "grad_norm": 6.267821788787842,
      "learning_rate": 4.3427265873994935e-05,
      "loss": 2.1667816162109377,
      "memory(GiB)": 77.56,
      "step": 63250,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.43757
    },
    {
      "epoch": 2.7100381303286065,
      "grad_norm": 5.555212497711182,
      "learning_rate": 4.342059456025689e-05,
      "loss": 2.2661861419677733,
      "memory(GiB)": 77.56,
      "step": 63255,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437572
    },
    {
      "epoch": 2.7102523456578553,
      "grad_norm": 5.554281234741211,
      "learning_rate": 4.341392336571096e-05,
      "loss": 2.500975799560547,
      "memory(GiB)": 77.56,
      "step": 63260,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.4376
    },
    {
      "epoch": 2.710466560987104,
      "grad_norm": 5.827896595001221,
      "learning_rate": 4.340725229047797e-05,
      "loss": 2.834640693664551,
      "memory(GiB)": 77.56,
      "step": 63265,
      "token_acc": 0.431438127090301,
      "train_speed(iter/s)": 1.437635
    },
    {
      "epoch": 2.7106807763163534,
      "grad_norm": 4.751279830932617,
      "learning_rate": 4.3400581334678805e-05,
      "loss": 2.234446334838867,
      "memory(GiB)": 77.56,
      "step": 63270,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 2.710894991645602,
      "grad_norm": 4.170665740966797,
      "learning_rate": 4.33939104984343e-05,
      "loss": 2.3763450622558593,
      "memory(GiB)": 77.56,
      "step": 63275,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 2.711109206974851,
      "grad_norm": 7.812934875488281,
      "learning_rate": 4.338723978186529e-05,
      "loss": 2.5575469970703124,
      "memory(GiB)": 77.56,
      "step": 63280,
      "token_acc": 0.4558303886925795,
      "train_speed(iter/s)": 1.437636
    },
    {
      "epoch": 2.7113234223041003,
      "grad_norm": 5.3547492027282715,
      "learning_rate": 4.338056918509265e-05,
      "loss": 2.553681182861328,
      "memory(GiB)": 77.56,
      "step": 63285,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.437619
    },
    {
      "epoch": 2.711537637633349,
      "grad_norm": 4.910192966461182,
      "learning_rate": 4.33738987082372e-05,
      "loss": 2.466324234008789,
      "memory(GiB)": 77.56,
      "step": 63290,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.437619
    },
    {
      "epoch": 2.711751852962598,
      "grad_norm": 4.390270709991455,
      "learning_rate": 4.336722835141979e-05,
      "loss": 2.21740779876709,
      "memory(GiB)": 77.56,
      "step": 63295,
      "token_acc": 0.5444444444444444,
      "train_speed(iter/s)": 1.43762
    },
    {
      "epoch": 2.711966068291847,
      "grad_norm": 4.178674697875977,
      "learning_rate": 4.3360558114761266e-05,
      "loss": 2.0102195739746094,
      "memory(GiB)": 77.56,
      "step": 63300,
      "token_acc": 0.5701754385964912,
      "train_speed(iter/s)": 1.437638
    },
    {
      "epoch": 2.712180283621096,
      "grad_norm": 5.572151184082031,
      "learning_rate": 4.3353887998382444e-05,
      "loss": 2.4363922119140624,
      "memory(GiB)": 77.56,
      "step": 63305,
      "token_acc": 0.4861878453038674,
      "train_speed(iter/s)": 1.437632
    },
    {
      "epoch": 2.712394498950345,
      "grad_norm": 7.255897521972656,
      "learning_rate": 4.334721800240418e-05,
      "loss": 2.480462837219238,
      "memory(GiB)": 77.56,
      "step": 63310,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 2.712608714279594,
      "grad_norm": 6.7720112800598145,
      "learning_rate": 4.3340548126947316e-05,
      "loss": 2.4891117095947264,
      "memory(GiB)": 77.56,
      "step": 63315,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 2.712822929608843,
      "grad_norm": 5.638824462890625,
      "learning_rate": 4.333387837213267e-05,
      "loss": 2.2209213256835936,
      "memory(GiB)": 77.56,
      "step": 63320,
      "token_acc": 0.5174603174603175,
      "train_speed(iter/s)": 1.43763
    },
    {
      "epoch": 2.7130371449380917,
      "grad_norm": 5.645864009857178,
      "learning_rate": 4.332720873808106e-05,
      "loss": 2.323262023925781,
      "memory(GiB)": 77.56,
      "step": 63325,
      "token_acc": 0.4921259842519685,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 2.713251360267341,
      "grad_norm": 5.4597063064575195,
      "learning_rate": 4.332053922491333e-05,
      "loss": 2.7017515182495115,
      "memory(GiB)": 77.56,
      "step": 63330,
      "token_acc": 0.43272727272727274,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 2.7134655755965897,
      "grad_norm": 6.1688313484191895,
      "learning_rate": 4.33138698327503e-05,
      "loss": 2.5099727630615236,
      "memory(GiB)": 77.56,
      "step": 63335,
      "token_acc": 0.48026315789473684,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 2.7136797909258386,
      "grad_norm": 5.5754313468933105,
      "learning_rate": 4.3307200561712777e-05,
      "loss": 2.4527713775634767,
      "memory(GiB)": 77.56,
      "step": 63340,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.437646
    },
    {
      "epoch": 2.713894006255088,
      "grad_norm": 6.005540370941162,
      "learning_rate": 4.330053141192161e-05,
      "loss": 2.424937438964844,
      "memory(GiB)": 77.56,
      "step": 63345,
      "token_acc": 0.5127272727272727,
      "train_speed(iter/s)": 1.437629
    },
    {
      "epoch": 2.7141082215843366,
      "grad_norm": 4.311441421508789,
      "learning_rate": 4.3293862383497594e-05,
      "loss": 2.2430381774902344,
      "memory(GiB)": 77.56,
      "step": 63350,
      "token_acc": 0.46987951807228917,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 2.7143224369135854,
      "grad_norm": 5.606351852416992,
      "learning_rate": 4.3287193476561555e-05,
      "loss": 2.2092987060546876,
      "memory(GiB)": 77.56,
      "step": 63355,
      "token_acc": 0.4980392156862745,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 2.7145366522428347,
      "grad_norm": 7.537847995758057,
      "learning_rate": 4.32805246912343e-05,
      "loss": 2.292195129394531,
      "memory(GiB)": 77.56,
      "step": 63360,
      "token_acc": 0.47720364741641336,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 2.7147508675720835,
      "grad_norm": 6.258525848388672,
      "learning_rate": 4.327385602763664e-05,
      "loss": 2.2040334701538087,
      "memory(GiB)": 77.56,
      "step": 63365,
      "token_acc": 0.5140845070422535,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 2.7149650829013323,
      "grad_norm": 5.4620232582092285,
      "learning_rate": 4.32671874858894e-05,
      "loss": 2.3868431091308593,
      "memory(GiB)": 77.56,
      "step": 63370,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.437604
    },
    {
      "epoch": 2.7151792982305816,
      "grad_norm": 4.934838771820068,
      "learning_rate": 4.326051906611337e-05,
      "loss": 2.5465763092041014,
      "memory(GiB)": 77.56,
      "step": 63375,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 2.7153935135598304,
      "grad_norm": 5.326876163482666,
      "learning_rate": 4.325385076842934e-05,
      "loss": 2.3296792984008787,
      "memory(GiB)": 77.56,
      "step": 63380,
      "token_acc": 0.5215686274509804,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 2.715607728889079,
      "grad_norm": 5.654313564300537,
      "learning_rate": 4.3247182592958136e-05,
      "loss": 2.5130685806274413,
      "memory(GiB)": 77.56,
      "step": 63385,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.437672
    },
    {
      "epoch": 2.7158219442183285,
      "grad_norm": 5.089864253997803,
      "learning_rate": 4.3240514539820574e-05,
      "loss": 2.4864452362060545,
      "memory(GiB)": 77.56,
      "step": 63390,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.437657
    },
    {
      "epoch": 2.7160361595475773,
      "grad_norm": 5.473681926727295,
      "learning_rate": 4.323384660913743e-05,
      "loss": 2.273955154418945,
      "memory(GiB)": 77.56,
      "step": 63395,
      "token_acc": 0.4962962962962963,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 2.716250374876826,
      "grad_norm": 6.059061050415039,
      "learning_rate": 4.322717880102949e-05,
      "loss": 2.6015626907348635,
      "memory(GiB)": 77.56,
      "step": 63400,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.437651
    },
    {
      "epoch": 2.7164645902060753,
      "grad_norm": 4.939112663269043,
      "learning_rate": 4.3220511115617565e-05,
      "loss": 2.588299369812012,
      "memory(GiB)": 77.56,
      "step": 63405,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.437652
    },
    {
      "epoch": 2.716678805535324,
      "grad_norm": 6.9835662841796875,
      "learning_rate": 4.321384355302244e-05,
      "loss": 2.371767044067383,
      "memory(GiB)": 77.56,
      "step": 63410,
      "token_acc": 0.4779116465863454,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 2.716893020864573,
      "grad_norm": 4.72863245010376,
      "learning_rate": 4.320717611336491e-05,
      "loss": 2.433430290222168,
      "memory(GiB)": 77.56,
      "step": 63415,
      "token_acc": 0.4670846394984326,
      "train_speed(iter/s)": 1.437638
    },
    {
      "epoch": 2.717107236193822,
      "grad_norm": 5.178319931030273,
      "learning_rate": 4.320050879676575e-05,
      "loss": 2.5596729278564454,
      "memory(GiB)": 77.56,
      "step": 63420,
      "token_acc": 0.48286604361370716,
      "train_speed(iter/s)": 1.437651
    },
    {
      "epoch": 2.717321451523071,
      "grad_norm": 6.3593339920043945,
      "learning_rate": 4.3193841603345755e-05,
      "loss": 2.3459300994873047,
      "memory(GiB)": 77.56,
      "step": 63425,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.71753566685232,
      "grad_norm": 5.314087867736816,
      "learning_rate": 4.318717453322571e-05,
      "loss": 2.1478271484375,
      "memory(GiB)": 77.56,
      "step": 63430,
      "token_acc": 0.5403225806451613,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 2.717749882181569,
      "grad_norm": 5.62764835357666,
      "learning_rate": 4.318050758652638e-05,
      "loss": 2.523588180541992,
      "memory(GiB)": 77.56,
      "step": 63435,
      "token_acc": 0.47896440129449835,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 2.717964097510818,
      "grad_norm": 6.044641971588135,
      "learning_rate": 4.317384076336855e-05,
      "loss": 2.4947792053222657,
      "memory(GiB)": 77.56,
      "step": 63440,
      "token_acc": 0.49240121580547114,
      "train_speed(iter/s)": 1.437709
    },
    {
      "epoch": 2.7181783128400667,
      "grad_norm": 5.518644332885742,
      "learning_rate": 4.3167174063873004e-05,
      "loss": 2.3688568115234374,
      "memory(GiB)": 77.56,
      "step": 63445,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.43771
    },
    {
      "epoch": 2.718392528169316,
      "grad_norm": 5.159210205078125,
      "learning_rate": 4.3160507488160504e-05,
      "loss": 2.248431396484375,
      "memory(GiB)": 77.56,
      "step": 63450,
      "token_acc": 0.5032894736842105,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 2.718606743498565,
      "grad_norm": 5.272059440612793,
      "learning_rate": 4.3153841036351814e-05,
      "loss": 2.185207176208496,
      "memory(GiB)": 77.56,
      "step": 63455,
      "token_acc": 0.5084033613445378,
      "train_speed(iter/s)": 1.437699
    },
    {
      "epoch": 2.7188209588278136,
      "grad_norm": 4.720504283905029,
      "learning_rate": 4.314717470856774e-05,
      "loss": 2.5009529113769533,
      "memory(GiB)": 77.56,
      "step": 63460,
      "token_acc": 0.5057915057915058,
      "train_speed(iter/s)": 1.437717
    },
    {
      "epoch": 2.719035174157063,
      "grad_norm": 4.2666707038879395,
      "learning_rate": 4.314050850492902e-05,
      "loss": 2.318477249145508,
      "memory(GiB)": 77.56,
      "step": 63465,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 2.7192493894863117,
      "grad_norm": 9.12672233581543,
      "learning_rate": 4.313384242555641e-05,
      "loss": 2.3794715881347654,
      "memory(GiB)": 77.56,
      "step": 63470,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 2.7194636048155605,
      "grad_norm": 5.946536064147949,
      "learning_rate": 4.31271764705707e-05,
      "loss": 2.5504430770874023,
      "memory(GiB)": 77.56,
      "step": 63475,
      "token_acc": 0.45384615384615384,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 2.7196778201448097,
      "grad_norm": 4.6696391105651855,
      "learning_rate": 4.3120510640092636e-05,
      "loss": 2.590966987609863,
      "memory(GiB)": 77.56,
      "step": 63480,
      "token_acc": 0.44135802469135804,
      "train_speed(iter/s)": 1.437733
    },
    {
      "epoch": 2.7198920354740586,
      "grad_norm": 5.773374557495117,
      "learning_rate": 4.311384493424297e-05,
      "loss": 2.4556367874145506,
      "memory(GiB)": 77.56,
      "step": 63485,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.437748
    },
    {
      "epoch": 2.7201062508033074,
      "grad_norm": 6.427194118499756,
      "learning_rate": 4.310717935314247e-05,
      "loss": 2.672300910949707,
      "memory(GiB)": 77.56,
      "step": 63490,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.437748
    },
    {
      "epoch": 2.7203204661325566,
      "grad_norm": 8.000069618225098,
      "learning_rate": 4.3100513896911865e-05,
      "loss": 2.444048309326172,
      "memory(GiB)": 77.56,
      "step": 63495,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 2.7205346814618054,
      "grad_norm": 4.372310161590576,
      "learning_rate": 4.309384856567194e-05,
      "loss": 2.7593429565429686,
      "memory(GiB)": 77.56,
      "step": 63500,
      "token_acc": 0.4400871459694989,
      "train_speed(iter/s)": 1.437706
    },
    {
      "epoch": 2.7205346814618054,
      "eval_loss": 2.2863733768463135,
      "eval_runtime": 14.3387,
      "eval_samples_per_second": 6.974,
      "eval_steps_per_second": 6.974,
      "eval_token_acc": 0.47322540473225405,
      "step": 63500
    },
    {
      "epoch": 2.7207488967910542,
      "grad_norm": 5.951817512512207,
      "learning_rate": 4.3087183359543426e-05,
      "loss": 2.231358528137207,
      "memory(GiB)": 77.56,
      "step": 63505,
      "token_acc": 0.487758945386064,
      "train_speed(iter/s)": 1.43721
    },
    {
      "epoch": 2.7209631121203035,
      "grad_norm": 4.930519104003906,
      "learning_rate": 4.308051827864705e-05,
      "loss": 2.52340087890625,
      "memory(GiB)": 77.56,
      "step": 63510,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.437217
    },
    {
      "epoch": 2.7211773274495523,
      "grad_norm": 4.144283294677734,
      "learning_rate": 4.3073853323103604e-05,
      "loss": 2.7167423248291014,
      "memory(GiB)": 77.56,
      "step": 63515,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.437192
    },
    {
      "epoch": 2.721391542778801,
      "grad_norm": 4.923089504241943,
      "learning_rate": 4.3067188493033796e-05,
      "loss": 2.500599670410156,
      "memory(GiB)": 77.56,
      "step": 63520,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.437224
    },
    {
      "epoch": 2.7216057581080504,
      "grad_norm": 5.032780170440674,
      "learning_rate": 4.3060523788558355e-05,
      "loss": 2.7366724014282227,
      "memory(GiB)": 77.56,
      "step": 63525,
      "token_acc": 0.46325878594249204,
      "train_speed(iter/s)": 1.437226
    },
    {
      "epoch": 2.721819973437299,
      "grad_norm": 5.736341953277588,
      "learning_rate": 4.3053859209798025e-05,
      "loss": 2.6113428115844726,
      "memory(GiB)": 77.56,
      "step": 63530,
      "token_acc": 0.44876325088339225,
      "train_speed(iter/s)": 1.437223
    },
    {
      "epoch": 2.722034188766548,
      "grad_norm": 5.133331775665283,
      "learning_rate": 4.3047194756873575e-05,
      "loss": 2.786575508117676,
      "memory(GiB)": 77.56,
      "step": 63535,
      "token_acc": 0.43018867924528303,
      "train_speed(iter/s)": 1.437218
    },
    {
      "epoch": 2.7222484040957973,
      "grad_norm": 4.850512504577637,
      "learning_rate": 4.3040530429905715e-05,
      "loss": 2.3773895263671876,
      "memory(GiB)": 77.56,
      "step": 63540,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.437241
    },
    {
      "epoch": 2.722462619425046,
      "grad_norm": 4.3101348876953125,
      "learning_rate": 4.3033866229015166e-05,
      "loss": 2.4509647369384764,
      "memory(GiB)": 77.56,
      "step": 63545,
      "token_acc": 0.47645429362880887,
      "train_speed(iter/s)": 1.437226
    },
    {
      "epoch": 2.722676834754295,
      "grad_norm": 4.423937797546387,
      "learning_rate": 4.302720215432268e-05,
      "loss": 2.192299461364746,
      "memory(GiB)": 77.56,
      "step": 63550,
      "token_acc": 0.5490196078431373,
      "train_speed(iter/s)": 1.437213
    },
    {
      "epoch": 2.722891050083544,
      "grad_norm": 5.230189323425293,
      "learning_rate": 4.302053820594895e-05,
      "loss": 2.5490650177001952,
      "memory(GiB)": 77.56,
      "step": 63555,
      "token_acc": 0.45222929936305734,
      "train_speed(iter/s)": 1.437241
    },
    {
      "epoch": 2.723105265412793,
      "grad_norm": 6.743096828460693,
      "learning_rate": 4.301387438401473e-05,
      "loss": 2.626717376708984,
      "memory(GiB)": 77.56,
      "step": 63560,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.437181
    },
    {
      "epoch": 2.7233194807420418,
      "grad_norm": 4.990501880645752,
      "learning_rate": 4.300721068864073e-05,
      "loss": 2.580089569091797,
      "memory(GiB)": 77.56,
      "step": 63565,
      "token_acc": 0.5062893081761006,
      "train_speed(iter/s)": 1.437205
    },
    {
      "epoch": 2.723533696071291,
      "grad_norm": 4.702508449554443,
      "learning_rate": 4.3000547119947656e-05,
      "loss": 2.6212514877319335,
      "memory(GiB)": 77.56,
      "step": 63570,
      "token_acc": 0.4434250764525994,
      "train_speed(iter/s)": 1.437221
    },
    {
      "epoch": 2.72374791140054,
      "grad_norm": 5.278894901275635,
      "learning_rate": 4.2993883678056246e-05,
      "loss": 2.4061717987060547,
      "memory(GiB)": 77.56,
      "step": 63575,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.437194
    },
    {
      "epoch": 2.7239621267297887,
      "grad_norm": 6.516336441040039,
      "learning_rate": 4.298722036308721e-05,
      "loss": 2.6168821334838865,
      "memory(GiB)": 77.56,
      "step": 63580,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 2.724176342059038,
      "grad_norm": 5.345509052276611,
      "learning_rate": 4.298055717516124e-05,
      "loss": 2.461959457397461,
      "memory(GiB)": 77.56,
      "step": 63585,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.437232
    },
    {
      "epoch": 2.7243905573882867,
      "grad_norm": 4.316685676574707,
      "learning_rate": 4.297389411439908e-05,
      "loss": 1.944647979736328,
      "memory(GiB)": 77.56,
      "step": 63590,
      "token_acc": 0.5584905660377358,
      "train_speed(iter/s)": 1.437216
    },
    {
      "epoch": 2.7246047727175355,
      "grad_norm": 5.405136585235596,
      "learning_rate": 4.2967231180921395e-05,
      "loss": 2.5211957931518554,
      "memory(GiB)": 77.56,
      "step": 63595,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.437242
    },
    {
      "epoch": 2.724818988046785,
      "grad_norm": 4.766284465789795,
      "learning_rate": 4.296056837484894e-05,
      "loss": 2.6702812194824217,
      "memory(GiB)": 77.56,
      "step": 63600,
      "token_acc": 0.43909348441926344,
      "train_speed(iter/s)": 1.437219
    },
    {
      "epoch": 2.7250332033760336,
      "grad_norm": 6.10042142868042,
      "learning_rate": 4.295390569630236e-05,
      "loss": 2.390985107421875,
      "memory(GiB)": 77.56,
      "step": 63605,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.437209
    },
    {
      "epoch": 2.7252474187052824,
      "grad_norm": 6.727659702301025,
      "learning_rate": 4.294724314540241e-05,
      "loss": 2.349884605407715,
      "memory(GiB)": 77.56,
      "step": 63610,
      "token_acc": 0.5302491103202847,
      "train_speed(iter/s)": 1.437209
    },
    {
      "epoch": 2.7254616340345317,
      "grad_norm": 6.050721168518066,
      "learning_rate": 4.294058072226976e-05,
      "loss": 2.5393613815307616,
      "memory(GiB)": 77.56,
      "step": 63615,
      "token_acc": 0.46131805157593125,
      "train_speed(iter/s)": 1.437223
    },
    {
      "epoch": 2.7256758493637805,
      "grad_norm": 7.031256675720215,
      "learning_rate": 4.293391842702513e-05,
      "loss": 2.5156442642211916,
      "memory(GiB)": 77.56,
      "step": 63620,
      "token_acc": 0.4148148148148148,
      "train_speed(iter/s)": 1.437222
    },
    {
      "epoch": 2.7258900646930293,
      "grad_norm": 6.383103370666504,
      "learning_rate": 4.2927256259789184e-05,
      "loss": 2.409423828125,
      "memory(GiB)": 77.56,
      "step": 63625,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.437234
    },
    {
      "epoch": 2.7261042800222786,
      "grad_norm": 6.660313606262207,
      "learning_rate": 4.292059422068262e-05,
      "loss": 2.379970359802246,
      "memory(GiB)": 77.56,
      "step": 63630,
      "token_acc": 0.48221343873517786,
      "train_speed(iter/s)": 1.437207
    },
    {
      "epoch": 2.7263184953515274,
      "grad_norm": 4.641581058502197,
      "learning_rate": 4.291393230982614e-05,
      "loss": 2.2225364685058593,
      "memory(GiB)": 77.56,
      "step": 63635,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.437208
    },
    {
      "epoch": 2.726532710680776,
      "grad_norm": 5.122830390930176,
      "learning_rate": 4.290727052734042e-05,
      "loss": 2.3288331985473634,
      "memory(GiB)": 77.56,
      "step": 63640,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.437214
    },
    {
      "epoch": 2.7267469260100254,
      "grad_norm": 4.4000067710876465,
      "learning_rate": 4.290060887334616e-05,
      "loss": 2.3637311935424803,
      "memory(GiB)": 77.56,
      "step": 63645,
      "token_acc": 0.4820359281437126,
      "train_speed(iter/s)": 1.437241
    },
    {
      "epoch": 2.7269611413392743,
      "grad_norm": 5.200026512145996,
      "learning_rate": 4.289394734796402e-05,
      "loss": 2.624894714355469,
      "memory(GiB)": 77.56,
      "step": 63650,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.437275
    },
    {
      "epoch": 2.727175356668523,
      "grad_norm": 5.261202812194824,
      "learning_rate": 4.288728595131469e-05,
      "loss": 2.2805709838867188,
      "memory(GiB)": 77.56,
      "step": 63655,
      "token_acc": 0.5222672064777328,
      "train_speed(iter/s)": 1.437291
    },
    {
      "epoch": 2.7273895719977723,
      "grad_norm": 4.581358909606934,
      "learning_rate": 4.2880624683518844e-05,
      "loss": 2.2480937957763674,
      "memory(GiB)": 77.56,
      "step": 63660,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.437323
    },
    {
      "epoch": 2.727603787327021,
      "grad_norm": 5.376256942749023,
      "learning_rate": 4.287396354469717e-05,
      "loss": 2.4637903213500976,
      "memory(GiB)": 77.56,
      "step": 63665,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.437346
    },
    {
      "epoch": 2.72781800265627,
      "grad_norm": 4.416100978851318,
      "learning_rate": 4.286730253497032e-05,
      "loss": 2.713575744628906,
      "memory(GiB)": 77.56,
      "step": 63670,
      "token_acc": 0.4495677233429395,
      "train_speed(iter/s)": 1.437357
    },
    {
      "epoch": 2.728032217985519,
      "grad_norm": 4.978917598724365,
      "learning_rate": 4.286064165445896e-05,
      "loss": 2.523843002319336,
      "memory(GiB)": 77.56,
      "step": 63675,
      "token_acc": 0.49382716049382713,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 2.728246433314768,
      "grad_norm": 5.200705528259277,
      "learning_rate": 4.28539809032838e-05,
      "loss": 2.511026382446289,
      "memory(GiB)": 77.56,
      "step": 63680,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.437392
    },
    {
      "epoch": 2.728460648644017,
      "grad_norm": 5.517684459686279,
      "learning_rate": 4.284732028156548e-05,
      "loss": 2.3788078308105467,
      "memory(GiB)": 77.56,
      "step": 63685,
      "token_acc": 0.5042016806722689,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 2.728674863973266,
      "grad_norm": 5.72015380859375,
      "learning_rate": 4.284065978942465e-05,
      "loss": 2.31542911529541,
      "memory(GiB)": 77.56,
      "step": 63690,
      "token_acc": 0.5066079295154186,
      "train_speed(iter/s)": 1.437442
    },
    {
      "epoch": 2.728889079302515,
      "grad_norm": 7.0121612548828125,
      "learning_rate": 4.2833999426982e-05,
      "loss": 2.4734703063964845,
      "memory(GiB)": 77.56,
      "step": 63695,
      "token_acc": 0.4735099337748344,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 2.7291032946317637,
      "grad_norm": 5.633831977844238,
      "learning_rate": 4.282733919435815e-05,
      "loss": 2.3416704177856444,
      "memory(GiB)": 77.56,
      "step": 63700,
      "token_acc": 0.5046439628482973,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 2.729317509961013,
      "grad_norm": 5.1880388259887695,
      "learning_rate": 4.28206790916738e-05,
      "loss": 2.6888330459594725,
      "memory(GiB)": 77.56,
      "step": 63705,
      "token_acc": 0.4628975265017668,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 2.729531725290262,
      "grad_norm": 5.1609392166137695,
      "learning_rate": 4.281401911904958e-05,
      "loss": 2.3305335998535157,
      "memory(GiB)": 77.56,
      "step": 63710,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.437442
    },
    {
      "epoch": 2.7297459406195106,
      "grad_norm": 6.756219387054443,
      "learning_rate": 4.280735927660613e-05,
      "loss": 2.470112609863281,
      "memory(GiB)": 77.56,
      "step": 63715,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.437476
    },
    {
      "epoch": 2.72996015594876,
      "grad_norm": 5.60759973526001,
      "learning_rate": 4.2800699564464134e-05,
      "loss": 2.5540462493896485,
      "memory(GiB)": 77.56,
      "step": 63720,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.437495
    },
    {
      "epoch": 2.7301743712780087,
      "grad_norm": 4.942457675933838,
      "learning_rate": 4.279403998274421e-05,
      "loss": 2.2493419647216797,
      "memory(GiB)": 77.56,
      "step": 63725,
      "token_acc": 0.5082508250825083,
      "train_speed(iter/s)": 1.437519
    },
    {
      "epoch": 2.7303885866072575,
      "grad_norm": 6.366645812988281,
      "learning_rate": 4.2787380531567e-05,
      "loss": 2.4286148071289064,
      "memory(GiB)": 77.56,
      "step": 63730,
      "token_acc": 0.44891640866873067,
      "train_speed(iter/s)": 1.437498
    },
    {
      "epoch": 2.7306028019365067,
      "grad_norm": 5.511943817138672,
      "learning_rate": 4.278072121105318e-05,
      "loss": 2.598671722412109,
      "memory(GiB)": 77.56,
      "step": 63735,
      "token_acc": 0.4265232974910394,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 2.7308170172657555,
      "grad_norm": 4.579773902893066,
      "learning_rate": 4.277406202132335e-05,
      "loss": 2.3932857513427734,
      "memory(GiB)": 77.56,
      "step": 63740,
      "token_acc": 0.5340909090909091,
      "train_speed(iter/s)": 1.437492
    },
    {
      "epoch": 2.7310312325950044,
      "grad_norm": 5.3524346351623535,
      "learning_rate": 4.276740296249817e-05,
      "loss": 2.3440357208251954,
      "memory(GiB)": 77.56,
      "step": 63745,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 2.7312454479242536,
      "grad_norm": 5.517590045928955,
      "learning_rate": 4.276074403469825e-05,
      "loss": 2.7323217391967773,
      "memory(GiB)": 77.56,
      "step": 63750,
      "token_acc": 0.44680851063829785,
      "train_speed(iter/s)": 1.43754
    },
    {
      "epoch": 2.7314596632535024,
      "grad_norm": 5.0154314041137695,
      "learning_rate": 4.275408523804427e-05,
      "loss": 2.7700546264648436,
      "memory(GiB)": 77.56,
      "step": 63755,
      "token_acc": 0.44410876132930516,
      "train_speed(iter/s)": 1.437554
    },
    {
      "epoch": 2.7316738785827512,
      "grad_norm": 5.464852333068848,
      "learning_rate": 4.274742657265682e-05,
      "loss": 2.411619186401367,
      "memory(GiB)": 77.56,
      "step": 63760,
      "token_acc": 0.45075757575757575,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 2.7318880939120005,
      "grad_norm": 5.425018310546875,
      "learning_rate": 4.2740768038656546e-05,
      "loss": 2.21066837310791,
      "memory(GiB)": 77.56,
      "step": 63765,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 2.7321023092412493,
      "grad_norm": 7.191006183624268,
      "learning_rate": 4.2734109636164074e-05,
      "loss": 2.612629699707031,
      "memory(GiB)": 77.56,
      "step": 63770,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.437609
    },
    {
      "epoch": 2.732316524570498,
      "grad_norm": 5.584221363067627,
      "learning_rate": 4.2727451365300014e-05,
      "loss": 2.433461380004883,
      "memory(GiB)": 77.56,
      "step": 63775,
      "token_acc": 0.5093167701863354,
      "train_speed(iter/s)": 1.437619
    },
    {
      "epoch": 2.7325307398997474,
      "grad_norm": 4.80769681930542,
      "learning_rate": 4.272079322618501e-05,
      "loss": 2.5371467590332033,
      "memory(GiB)": 77.56,
      "step": 63780,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 2.732744955228996,
      "grad_norm": 4.261141300201416,
      "learning_rate": 4.271413521893965e-05,
      "loss": 2.3539379119873045,
      "memory(GiB)": 77.56,
      "step": 63785,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 2.732959170558245,
      "grad_norm": 5.126184940338135,
      "learning_rate": 4.270747734368457e-05,
      "loss": 2.2294429779052733,
      "memory(GiB)": 77.56,
      "step": 63790,
      "token_acc": 0.5376712328767124,
      "train_speed(iter/s)": 1.437677
    },
    {
      "epoch": 2.7331733858874943,
      "grad_norm": 5.520096778869629,
      "learning_rate": 4.270081960054038e-05,
      "loss": 2.525926971435547,
      "memory(GiB)": 77.56,
      "step": 63795,
      "token_acc": 0.4847328244274809,
      "train_speed(iter/s)": 1.437681
    },
    {
      "epoch": 2.733387601216743,
      "grad_norm": 4.697901248931885,
      "learning_rate": 4.269416198962768e-05,
      "loss": 2.1814714431762696,
      "memory(GiB)": 77.56,
      "step": 63800,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.437649
    },
    {
      "epoch": 2.733601816545992,
      "grad_norm": 4.715577602386475,
      "learning_rate": 4.26875045110671e-05,
      "loss": 2.3868234634399412,
      "memory(GiB)": 77.56,
      "step": 63805,
      "token_acc": 0.5186246418338109,
      "train_speed(iter/s)": 1.437658
    },
    {
      "epoch": 2.733816031875241,
      "grad_norm": 5.891027927398682,
      "learning_rate": 4.268084716497924e-05,
      "loss": 2.3933685302734373,
      "memory(GiB)": 77.56,
      "step": 63810,
      "token_acc": 0.5022222222222222,
      "train_speed(iter/s)": 1.437692
    },
    {
      "epoch": 2.73403024720449,
      "grad_norm": 5.315133571624756,
      "learning_rate": 4.267418995148468e-05,
      "loss": 2.706546401977539,
      "memory(GiB)": 77.56,
      "step": 63815,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 2.7342444625337388,
      "grad_norm": 5.708217620849609,
      "learning_rate": 4.266753287070406e-05,
      "loss": 2.252176284790039,
      "memory(GiB)": 77.56,
      "step": 63820,
      "token_acc": 0.5351170568561873,
      "train_speed(iter/s)": 1.437707
    },
    {
      "epoch": 2.734458677862988,
      "grad_norm": 5.75045919418335,
      "learning_rate": 4.266087592275794e-05,
      "loss": 2.4530160903930662,
      "memory(GiB)": 77.56,
      "step": 63825,
      "token_acc": 0.49025069637883006,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 2.734672893192237,
      "grad_norm": 5.411227226257324,
      "learning_rate": 4.265421910776694e-05,
      "loss": 2.471903991699219,
      "memory(GiB)": 77.56,
      "step": 63830,
      "token_acc": 0.519163763066202,
      "train_speed(iter/s)": 1.437709
    },
    {
      "epoch": 2.7348871085214856,
      "grad_norm": 5.818785190582275,
      "learning_rate": 4.2647562425851666e-05,
      "loss": 2.489049530029297,
      "memory(GiB)": 77.56,
      "step": 63835,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.43771
    },
    {
      "epoch": 2.735101323850735,
      "grad_norm": 5.858545303344727,
      "learning_rate": 4.26409058771327e-05,
      "loss": 2.4503189086914063,
      "memory(GiB)": 77.56,
      "step": 63840,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.437717
    },
    {
      "epoch": 2.7353155391799837,
      "grad_norm": 4.953150749206543,
      "learning_rate": 4.2634249461730616e-05,
      "loss": 2.2386812210083007,
      "memory(GiB)": 77.56,
      "step": 63845,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 2.7355297545092325,
      "grad_norm": 6.272424221038818,
      "learning_rate": 4.262759317976602e-05,
      "loss": 2.668625259399414,
      "memory(GiB)": 77.56,
      "step": 63850,
      "token_acc": 0.4732824427480916,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 2.735743969838482,
      "grad_norm": 6.13362455368042,
      "learning_rate": 4.262093703135949e-05,
      "loss": 2.479934310913086,
      "memory(GiB)": 77.56,
      "step": 63855,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 2.7359581851677306,
      "grad_norm": 7.060302734375,
      "learning_rate": 4.26142810166316e-05,
      "loss": 2.2734760284423827,
      "memory(GiB)": 77.56,
      "step": 63860,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 2.7361724004969794,
      "grad_norm": 4.511002540588379,
      "learning_rate": 4.260762513570294e-05,
      "loss": 2.16192569732666,
      "memory(GiB)": 77.56,
      "step": 63865,
      "token_acc": 0.5209125475285171,
      "train_speed(iter/s)": 1.43772
    },
    {
      "epoch": 2.7363866158262287,
      "grad_norm": 5.379841327667236,
      "learning_rate": 4.260096938869409e-05,
      "loss": 2.4815832138061524,
      "memory(GiB)": 77.56,
      "step": 63870,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 2.7366008311554775,
      "grad_norm": 4.512026786804199,
      "learning_rate": 4.2594313775725616e-05,
      "loss": 2.228351593017578,
      "memory(GiB)": 77.56,
      "step": 63875,
      "token_acc": 0.5348101265822784,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 2.7368150464847263,
      "grad_norm": 4.49551248550415,
      "learning_rate": 4.25876582969181e-05,
      "loss": 2.3812911987304686,
      "memory(GiB)": 77.56,
      "step": 63880,
      "token_acc": 0.49538461538461537,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.7370292618139755,
      "grad_norm": 6.15917444229126,
      "learning_rate": 4.258100295239209e-05,
      "loss": 2.1320409774780273,
      "memory(GiB)": 77.56,
      "step": 63885,
      "token_acc": 0.5114503816793893,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.7372434771432244,
      "grad_norm": 6.878716468811035,
      "learning_rate": 4.25743477422682e-05,
      "loss": 2.1629804611206054,
      "memory(GiB)": 77.56,
      "step": 63890,
      "token_acc": 0.509493670886076,
      "train_speed(iter/s)": 1.437765
    },
    {
      "epoch": 2.737457692472473,
      "grad_norm": 5.495079040527344,
      "learning_rate": 4.2567692666666945e-05,
      "loss": 2.4723489761352537,
      "memory(GiB)": 77.56,
      "step": 63895,
      "token_acc": 0.48089171974522293,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.7376719078017224,
      "grad_norm": 4.677679538726807,
      "learning_rate": 4.2561037725708904e-05,
      "loss": 2.304025650024414,
      "memory(GiB)": 77.56,
      "step": 63900,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 2.7378861231309712,
      "grad_norm": 8.411466598510742,
      "learning_rate": 4.2554382919514645e-05,
      "loss": 2.4469631195068358,
      "memory(GiB)": 77.56,
      "step": 63905,
      "token_acc": 0.4793388429752066,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 2.73810033846022,
      "grad_norm": 6.134751796722412,
      "learning_rate": 4.254772824820474e-05,
      "loss": 2.371603012084961,
      "memory(GiB)": 77.56,
      "step": 63910,
      "token_acc": 0.5652173913043478,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 2.7383145537894693,
      "grad_norm": 6.018167495727539,
      "learning_rate": 4.254107371189973e-05,
      "loss": 2.400075340270996,
      "memory(GiB)": 77.56,
      "step": 63915,
      "token_acc": 0.5054945054945055,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 2.738528769118718,
      "grad_norm": 6.0654706954956055,
      "learning_rate": 4.253441931072015e-05,
      "loss": 2.488100051879883,
      "memory(GiB)": 77.56,
      "step": 63920,
      "token_acc": 0.48616600790513836,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 2.738742984447967,
      "grad_norm": 5.200046539306641,
      "learning_rate": 4.2527765044786576e-05,
      "loss": 2.4781816482543944,
      "memory(GiB)": 77.56,
      "step": 63925,
      "token_acc": 0.4789156626506024,
      "train_speed(iter/s)": 1.437775
    },
    {
      "epoch": 2.738957199777216,
      "grad_norm": 6.048184871673584,
      "learning_rate": 4.252111091421954e-05,
      "loss": 2.5228322982788085,
      "memory(GiB)": 77.56,
      "step": 63930,
      "token_acc": 0.4555984555984556,
      "train_speed(iter/s)": 1.4378
    },
    {
      "epoch": 2.739171415106465,
      "grad_norm": 13.278603553771973,
      "learning_rate": 4.251445691913961e-05,
      "loss": 2.2331220626831056,
      "memory(GiB)": 77.56,
      "step": 63935,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.437803
    },
    {
      "epoch": 2.739385630435714,
      "grad_norm": 5.795896530151367,
      "learning_rate": 4.250780305966731e-05,
      "loss": 2.5971269607543945,
      "memory(GiB)": 77.56,
      "step": 63940,
      "token_acc": 0.4651898734177215,
      "train_speed(iter/s)": 1.4378
    },
    {
      "epoch": 2.739599845764963,
      "grad_norm": 6.14264440536499,
      "learning_rate": 4.2501149335923176e-05,
      "loss": 2.2010746002197266,
      "memory(GiB)": 77.56,
      "step": 63945,
      "token_acc": 0.5149253731343284,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 2.739814061094212,
      "grad_norm": 7.9233927726745605,
      "learning_rate": 4.2494495748027776e-05,
      "loss": 2.227177619934082,
      "memory(GiB)": 77.56,
      "step": 63950,
      "token_acc": 0.5349794238683128,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 2.7400282764234607,
      "grad_norm": 4.575939655303955,
      "learning_rate": 4.2487842296101615e-05,
      "loss": 2.452601432800293,
      "memory(GiB)": 77.56,
      "step": 63955,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 2.74024249175271,
      "grad_norm": 8.174505233764648,
      "learning_rate": 4.248118898026523e-05,
      "loss": 2.5592151641845704,
      "memory(GiB)": 77.56,
      "step": 63960,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.437849
    },
    {
      "epoch": 2.7404567070819588,
      "grad_norm": 5.447796821594238,
      "learning_rate": 4.247453580063917e-05,
      "loss": 2.167592239379883,
      "memory(GiB)": 77.56,
      "step": 63965,
      "token_acc": 0.5014005602240896,
      "train_speed(iter/s)": 1.437823
    },
    {
      "epoch": 2.7406709224112076,
      "grad_norm": 5.853092193603516,
      "learning_rate": 4.246788275734393e-05,
      "loss": 2.3901538848876953,
      "memory(GiB)": 77.56,
      "step": 63970,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 2.740885137740457,
      "grad_norm": 6.94603157043457,
      "learning_rate": 4.2461229850500075e-05,
      "loss": 2.5193914413452148,
      "memory(GiB)": 77.56,
      "step": 63975,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 2.7410993530697056,
      "grad_norm": 4.713394641876221,
      "learning_rate": 4.245457708022813e-05,
      "loss": 2.1621885299682617,
      "memory(GiB)": 77.56,
      "step": 63980,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 2.7413135683989545,
      "grad_norm": 5.517740249633789,
      "learning_rate": 4.244792444664859e-05,
      "loss": 2.488557243347168,
      "memory(GiB)": 77.56,
      "step": 63985,
      "token_acc": 0.47126436781609193,
      "train_speed(iter/s)": 1.43786
    },
    {
      "epoch": 2.7415277837282037,
      "grad_norm": 5.591786861419678,
      "learning_rate": 4.2441271949881975e-05,
      "loss": 2.33306999206543,
      "memory(GiB)": 77.56,
      "step": 63990,
      "token_acc": 0.49429657794676807,
      "train_speed(iter/s)": 1.437858
    },
    {
      "epoch": 2.7417419990574525,
      "grad_norm": 4.84340763092041,
      "learning_rate": 4.2434619590048817e-05,
      "loss": 2.554062843322754,
      "memory(GiB)": 77.56,
      "step": 63995,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 2.7419562143867013,
      "grad_norm": 5.150165557861328,
      "learning_rate": 4.242796736726963e-05,
      "loss": 2.116430473327637,
      "memory(GiB)": 77.56,
      "step": 64000,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.437901
    },
    {
      "epoch": 2.7419562143867013,
      "eval_loss": 2.249814033508301,
      "eval_runtime": 13.9151,
      "eval_samples_per_second": 7.186,
      "eval_steps_per_second": 7.186,
      "eval_token_acc": 0.4489528795811518,
      "step": 64000
    },
    {
      "epoch": 2.7421704297159506,
      "grad_norm": 7.5258378982543945,
      "learning_rate": 4.24213152816649e-05,
      "loss": 2.838523292541504,
      "memory(GiB)": 77.56,
      "step": 64005,
      "token_acc": 0.4566420664206642,
      "train_speed(iter/s)": 1.437436
    },
    {
      "epoch": 2.7423846450451994,
      "grad_norm": 6.42828369140625,
      "learning_rate": 4.241466333335517e-05,
      "loss": 2.504629135131836,
      "memory(GiB)": 77.56,
      "step": 64010,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.437439
    },
    {
      "epoch": 2.742598860374448,
      "grad_norm": 5.749969005584717,
      "learning_rate": 4.240801152246091e-05,
      "loss": 2.305723190307617,
      "memory(GiB)": 77.56,
      "step": 64015,
      "token_acc": 0.5173745173745173,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 2.7428130757036975,
      "grad_norm": 5.637388706207275,
      "learning_rate": 4.2401359849102653e-05,
      "loss": 2.6274797439575197,
      "memory(GiB)": 77.56,
      "step": 64020,
      "token_acc": 0.5122699386503068,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 2.7430272910329463,
      "grad_norm": 4.276328086853027,
      "learning_rate": 4.2394708313400894e-05,
      "loss": 2.275105857849121,
      "memory(GiB)": 77.56,
      "step": 64025,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.437416
    },
    {
      "epoch": 2.743241506362195,
      "grad_norm": 5.042388916015625,
      "learning_rate": 4.238805691547612e-05,
      "loss": 2.4059112548828123,
      "memory(GiB)": 77.56,
      "step": 64030,
      "token_acc": 0.5274725274725275,
      "train_speed(iter/s)": 1.437442
    },
    {
      "epoch": 2.7434557216914444,
      "grad_norm": 5.682839870452881,
      "learning_rate": 4.238140565544885e-05,
      "loss": 2.565266418457031,
      "memory(GiB)": 77.56,
      "step": 64035,
      "token_acc": 0.4322033898305085,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 2.743669937020693,
      "grad_norm": 5.284014701843262,
      "learning_rate": 4.237475453343955e-05,
      "loss": 2.4839141845703123,
      "memory(GiB)": 77.56,
      "step": 64040,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.437432
    },
    {
      "epoch": 2.743884152349942,
      "grad_norm": 8.432207107543945,
      "learning_rate": 4.236810354956872e-05,
      "loss": 2.3649349212646484,
      "memory(GiB)": 77.56,
      "step": 64045,
      "token_acc": 0.525096525096525,
      "train_speed(iter/s)": 1.437432
    },
    {
      "epoch": 2.7440983676791912,
      "grad_norm": 5.532547950744629,
      "learning_rate": 4.236145270395685e-05,
      "loss": 2.35306396484375,
      "memory(GiB)": 77.56,
      "step": 64050,
      "token_acc": 0.5189873417721519,
      "train_speed(iter/s)": 1.437452
    },
    {
      "epoch": 2.74431258300844,
      "grad_norm": 5.355996131896973,
      "learning_rate": 4.2354801996724444e-05,
      "loss": 2.484995651245117,
      "memory(GiB)": 77.56,
      "step": 64055,
      "token_acc": 0.5232974910394266,
      "train_speed(iter/s)": 1.437448
    },
    {
      "epoch": 2.744526798337689,
      "grad_norm": 6.381491184234619,
      "learning_rate": 4.2348151427991974e-05,
      "loss": 2.394114875793457,
      "memory(GiB)": 77.56,
      "step": 64060,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437478
    },
    {
      "epoch": 2.744741013666938,
      "grad_norm": 4.961148738861084,
      "learning_rate": 4.234150099787991e-05,
      "loss": 2.376239013671875,
      "memory(GiB)": 77.56,
      "step": 64065,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.437486
    },
    {
      "epoch": 2.744955228996187,
      "grad_norm": 5.538020133972168,
      "learning_rate": 4.233485070650874e-05,
      "loss": 1.9315692901611328,
      "memory(GiB)": 77.56,
      "step": 64070,
      "token_acc": 0.5396825396825397,
      "train_speed(iter/s)": 1.437457
    },
    {
      "epoch": 2.7451694443254357,
      "grad_norm": 5.472782611846924,
      "learning_rate": 4.2328200553998944e-05,
      "loss": 2.3290735244750977,
      "memory(GiB)": 77.56,
      "step": 64075,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.437454
    },
    {
      "epoch": 2.745383659654685,
      "grad_norm": 6.293210506439209,
      "learning_rate": 4.2321550540470996e-05,
      "loss": 2.475215530395508,
      "memory(GiB)": 77.56,
      "step": 64080,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.437477
    },
    {
      "epoch": 2.745597874983934,
      "grad_norm": 5.869363784790039,
      "learning_rate": 4.231490066604536e-05,
      "loss": 2.6260141372680663,
      "memory(GiB)": 77.56,
      "step": 64085,
      "token_acc": 0.4565916398713826,
      "train_speed(iter/s)": 1.437478
    },
    {
      "epoch": 2.7458120903131826,
      "grad_norm": 5.941280841827393,
      "learning_rate": 4.23082509308425e-05,
      "loss": 2.4243059158325195,
      "memory(GiB)": 77.56,
      "step": 64090,
      "token_acc": 0.4365079365079365,
      "train_speed(iter/s)": 1.437496
    },
    {
      "epoch": 2.746026305642432,
      "grad_norm": 5.775561809539795,
      "learning_rate": 4.23016013349829e-05,
      "loss": 2.7795574188232424,
      "memory(GiB)": 77.56,
      "step": 64095,
      "token_acc": 0.46200607902735563,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 2.7462405209716807,
      "grad_norm": 5.151354789733887,
      "learning_rate": 4.229495187858701e-05,
      "loss": 2.2424667358398436,
      "memory(GiB)": 77.56,
      "step": 64100,
      "token_acc": 0.5360501567398119,
      "train_speed(iter/s)": 1.437498
    },
    {
      "epoch": 2.7464547363009295,
      "grad_norm": 6.758477210998535,
      "learning_rate": 4.2288302561775295e-05,
      "loss": 2.7601896286010743,
      "memory(GiB)": 77.56,
      "step": 64105,
      "token_acc": 0.4041095890410959,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 2.7466689516301788,
      "grad_norm": 5.916769504547119,
      "learning_rate": 4.228165338466821e-05,
      "loss": 2.557953643798828,
      "memory(GiB)": 77.56,
      "step": 64110,
      "token_acc": 0.4610169491525424,
      "train_speed(iter/s)": 1.437483
    },
    {
      "epoch": 2.7468831669594276,
      "grad_norm": 7.3369550704956055,
      "learning_rate": 4.227500434738622e-05,
      "loss": 2.4919652938842773,
      "memory(GiB)": 77.56,
      "step": 64115,
      "token_acc": 0.4781144781144781,
      "train_speed(iter/s)": 1.437472
    },
    {
      "epoch": 2.7470973822886764,
      "grad_norm": 6.129856109619141,
      "learning_rate": 4.226835545004975e-05,
      "loss": 2.297022247314453,
      "memory(GiB)": 77.56,
      "step": 64120,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.437452
    },
    {
      "epoch": 2.7473115976179256,
      "grad_norm": 4.017956256866455,
      "learning_rate": 4.226170669277929e-05,
      "loss": 2.086667060852051,
      "memory(GiB)": 77.56,
      "step": 64125,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.437434
    },
    {
      "epoch": 2.7475258129471745,
      "grad_norm": 4.9504475593566895,
      "learning_rate": 4.225505807569529e-05,
      "loss": 2.4380519866943358,
      "memory(GiB)": 77.56,
      "step": 64130,
      "token_acc": 0.48184818481848185,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 2.7477400282764233,
      "grad_norm": 6.795463562011719,
      "learning_rate": 4.224840959891815e-05,
      "loss": 2.3949138641357424,
      "memory(GiB)": 77.56,
      "step": 64135,
      "token_acc": 0.5241379310344828,
      "train_speed(iter/s)": 1.437439
    },
    {
      "epoch": 2.7479542436056725,
      "grad_norm": 5.56437873840332,
      "learning_rate": 4.224176126256836e-05,
      "loss": 2.289603424072266,
      "memory(GiB)": 77.56,
      "step": 64140,
      "token_acc": 0.5097276264591439,
      "train_speed(iter/s)": 1.437443
    },
    {
      "epoch": 2.7481684589349213,
      "grad_norm": 6.308046340942383,
      "learning_rate": 4.223511306676634e-05,
      "loss": 2.2859668731689453,
      "memory(GiB)": 77.56,
      "step": 64145,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 2.74838267426417,
      "grad_norm": 5.176784992218018,
      "learning_rate": 4.222846501163253e-05,
      "loss": 2.1252552032470704,
      "memory(GiB)": 77.56,
      "step": 64150,
      "token_acc": 0.5122699386503068,
      "train_speed(iter/s)": 1.437434
    },
    {
      "epoch": 2.7485968895934194,
      "grad_norm": 6.133218765258789,
      "learning_rate": 4.222181709728736e-05,
      "loss": 1.989995574951172,
      "memory(GiB)": 77.56,
      "step": 64155,
      "token_acc": 0.5465587044534413,
      "train_speed(iter/s)": 1.437467
    },
    {
      "epoch": 2.7488111049226682,
      "grad_norm": 5.641963958740234,
      "learning_rate": 4.221516932385128e-05,
      "loss": 2.3942235946655273,
      "memory(GiB)": 77.56,
      "step": 64160,
      "token_acc": 0.5223880597014925,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 2.749025320251917,
      "grad_norm": 6.726781368255615,
      "learning_rate": 4.220852169144471e-05,
      "loss": 2.5887020111083983,
      "memory(GiB)": 77.56,
      "step": 64165,
      "token_acc": 0.46037735849056605,
      "train_speed(iter/s)": 1.437464
    },
    {
      "epoch": 2.7492395355811663,
      "grad_norm": 6.003223419189453,
      "learning_rate": 4.2201874200188074e-05,
      "loss": 2.3801013946533205,
      "memory(GiB)": 77.56,
      "step": 64170,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.437451
    },
    {
      "epoch": 2.749453750910415,
      "grad_norm": 5.951838970184326,
      "learning_rate": 4.2195226850201796e-05,
      "loss": 2.3997867584228514,
      "memory(GiB)": 77.56,
      "step": 64175,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.437449
    },
    {
      "epoch": 2.749667966239664,
      "grad_norm": 5.521102428436279,
      "learning_rate": 4.2188579641606316e-05,
      "loss": 2.394172477722168,
      "memory(GiB)": 77.56,
      "step": 64180,
      "token_acc": 0.4790996784565916,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 2.749882181568913,
      "grad_norm": 6.232408046722412,
      "learning_rate": 4.218193257452204e-05,
      "loss": 2.4417022705078124,
      "memory(GiB)": 77.56,
      "step": 64185,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.437476
    },
    {
      "epoch": 2.750096396898162,
      "grad_norm": 4.895714282989502,
      "learning_rate": 4.217528564906938e-05,
      "loss": 2.3579416275024414,
      "memory(GiB)": 77.56,
      "step": 64190,
      "token_acc": 0.5202702702702703,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 2.750310612227411,
      "grad_norm": 5.193429946899414,
      "learning_rate": 4.2168638865368756e-05,
      "loss": 2.474519157409668,
      "memory(GiB)": 77.56,
      "step": 64195,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.437522
    },
    {
      "epoch": 2.75052482755666,
      "grad_norm": 6.1212687492370605,
      "learning_rate": 4.2161992223540606e-05,
      "loss": 2.344174385070801,
      "memory(GiB)": 77.56,
      "step": 64200,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 2.750739042885909,
      "grad_norm": 6.556796073913574,
      "learning_rate": 4.215534572370531e-05,
      "loss": 2.169703483581543,
      "memory(GiB)": 77.56,
      "step": 64205,
      "token_acc": 0.509375,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 2.7509532582151577,
      "grad_norm": 5.754509925842285,
      "learning_rate": 4.214869936598329e-05,
      "loss": 2.6595726013183594,
      "memory(GiB)": 77.56,
      "step": 64210,
      "token_acc": 0.48328267477203646,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 2.751167473544407,
      "grad_norm": 4.662178993225098,
      "learning_rate": 4.214205315049494e-05,
      "loss": 2.3331153869628904,
      "memory(GiB)": 77.56,
      "step": 64215,
      "token_acc": 0.5215686274509804,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 2.7513816888736558,
      "grad_norm": 4.8971266746521,
      "learning_rate": 4.2135407077360667e-05,
      "loss": 2.5479814529418947,
      "memory(GiB)": 77.56,
      "step": 64220,
      "token_acc": 0.4702194357366771,
      "train_speed(iter/s)": 1.437539
    },
    {
      "epoch": 2.7515959042029046,
      "grad_norm": 5.7266621589660645,
      "learning_rate": 4.2128761146700884e-05,
      "loss": 2.296100616455078,
      "memory(GiB)": 77.56,
      "step": 64225,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.437538
    },
    {
      "epoch": 2.751810119532154,
      "grad_norm": 4.668185234069824,
      "learning_rate": 4.2122115358635975e-05,
      "loss": 2.357310485839844,
      "memory(GiB)": 77.56,
      "step": 64230,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.437556
    },
    {
      "epoch": 2.7520243348614026,
      "grad_norm": 4.7295145988464355,
      "learning_rate": 4.2115469713286325e-05,
      "loss": 2.4320003509521486,
      "memory(GiB)": 77.56,
      "step": 64235,
      "token_acc": 0.48872180451127817,
      "train_speed(iter/s)": 1.437583
    },
    {
      "epoch": 2.7522385501906514,
      "grad_norm": 5.936611175537109,
      "learning_rate": 4.210882421077235e-05,
      "loss": 2.478118133544922,
      "memory(GiB)": 77.56,
      "step": 64240,
      "token_acc": 0.46494464944649444,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 2.7524527655199007,
      "grad_norm": 4.063772201538086,
      "learning_rate": 4.210217885121442e-05,
      "loss": 2.463674545288086,
      "memory(GiB)": 77.56,
      "step": 64245,
      "token_acc": 0.46407185628742514,
      "train_speed(iter/s)": 1.437555
    },
    {
      "epoch": 2.7526669808491495,
      "grad_norm": 6.379580974578857,
      "learning_rate": 4.209553363473293e-05,
      "loss": 2.376224327087402,
      "memory(GiB)": 77.56,
      "step": 64250,
      "token_acc": 0.5145631067961165,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 2.7528811961783983,
      "grad_norm": 9.669970512390137,
      "learning_rate": 4.208888856144826e-05,
      "loss": 2.440271186828613,
      "memory(GiB)": 77.56,
      "step": 64255,
      "token_acc": 0.4818941504178273,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 2.7530954115076476,
      "grad_norm": 6.004137992858887,
      "learning_rate": 4.2082243631480795e-05,
      "loss": 2.288914108276367,
      "memory(GiB)": 77.56,
      "step": 64260,
      "token_acc": 0.47794117647058826,
      "train_speed(iter/s)": 1.437573
    },
    {
      "epoch": 2.7533096268368964,
      "grad_norm": 4.116734981536865,
      "learning_rate": 4.207559884495092e-05,
      "loss": 2.394124412536621,
      "memory(GiB)": 77.56,
      "step": 64265,
      "token_acc": 0.49836065573770494,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 2.753523842166145,
      "grad_norm": 5.865310192108154,
      "learning_rate": 4.2068954201978985e-05,
      "loss": 2.4204483032226562,
      "memory(GiB)": 77.56,
      "step": 64270,
      "token_acc": 0.4967532467532468,
      "train_speed(iter/s)": 1.437599
    },
    {
      "epoch": 2.7537380574953945,
      "grad_norm": 6.250423431396484,
      "learning_rate": 4.2062309702685404e-05,
      "loss": 2.1397994995117187,
      "memory(GiB)": 77.56,
      "step": 64275,
      "token_acc": 0.5429553264604811,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 2.7539522728246433,
      "grad_norm": 6.483864784240723,
      "learning_rate": 4.205566534719052e-05,
      "loss": 2.4461437225341798,
      "memory(GiB)": 77.56,
      "step": 64280,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 2.754166488153892,
      "grad_norm": 6.371665000915527,
      "learning_rate": 4.204902113561472e-05,
      "loss": 2.339692306518555,
      "memory(GiB)": 77.56,
      "step": 64285,
      "token_acc": 0.5243055555555556,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 2.7543807034831413,
      "grad_norm": 4.810792922973633,
      "learning_rate": 4.204237706807836e-05,
      "loss": 2.317988967895508,
      "memory(GiB)": 77.56,
      "step": 64290,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 2.75459491881239,
      "grad_norm": 6.8076982498168945,
      "learning_rate": 4.20357331447018e-05,
      "loss": 2.5038827896118163,
      "memory(GiB)": 77.56,
      "step": 64295,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.437551
    },
    {
      "epoch": 2.754809134141639,
      "grad_norm": 4.727570533752441,
      "learning_rate": 4.202908936560541e-05,
      "loss": 2.75933723449707,
      "memory(GiB)": 77.56,
      "step": 64300,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.437552
    },
    {
      "epoch": 2.7550233494708882,
      "grad_norm": 5.943518161773682,
      "learning_rate": 4.202244573090954e-05,
      "loss": 2.4613155364990233,
      "memory(GiB)": 77.56,
      "step": 64305,
      "token_acc": 0.4734848484848485,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 2.755237564800137,
      "grad_norm": 4.886484622955322,
      "learning_rate": 4.2015802240734554e-05,
      "loss": 2.343560791015625,
      "memory(GiB)": 77.56,
      "step": 64310,
      "token_acc": 0.5015015015015015,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 2.755451780129386,
      "grad_norm": 5.043753147125244,
      "learning_rate": 4.200915889520079e-05,
      "loss": 2.4613529205322267,
      "memory(GiB)": 77.56,
      "step": 64315,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.437564
    },
    {
      "epoch": 2.755665995458635,
      "grad_norm": 6.139881134033203,
      "learning_rate": 4.2002515694428616e-05,
      "loss": 2.395285224914551,
      "memory(GiB)": 77.56,
      "step": 64320,
      "token_acc": 0.4847328244274809,
      "train_speed(iter/s)": 1.437592
    },
    {
      "epoch": 2.755880210787884,
      "grad_norm": 6.566813945770264,
      "learning_rate": 4.1995872638538375e-05,
      "loss": 2.3253076553344725,
      "memory(GiB)": 77.56,
      "step": 64325,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.43761
    },
    {
      "epoch": 2.7560944261171327,
      "grad_norm": 5.818182468414307,
      "learning_rate": 4.1989229727650416e-05,
      "loss": 2.640835189819336,
      "memory(GiB)": 77.56,
      "step": 64330,
      "token_acc": 0.4535809018567639,
      "train_speed(iter/s)": 1.437602
    },
    {
      "epoch": 2.756308641446382,
      "grad_norm": 5.255789756774902,
      "learning_rate": 4.1982586961885054e-05,
      "loss": 2.444224739074707,
      "memory(GiB)": 77.56,
      "step": 64335,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.437584
    },
    {
      "epoch": 2.756522856775631,
      "grad_norm": 4.549758434295654,
      "learning_rate": 4.1975944341362646e-05,
      "loss": 2.3875503540039062,
      "memory(GiB)": 77.56,
      "step": 64340,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.437601
    },
    {
      "epoch": 2.7567370721048796,
      "grad_norm": 5.681735038757324,
      "learning_rate": 4.196930186620355e-05,
      "loss": 2.4830114364624025,
      "memory(GiB)": 77.56,
      "step": 64345,
      "token_acc": 0.43854748603351956,
      "train_speed(iter/s)": 1.437591
    },
    {
      "epoch": 2.756951287434129,
      "grad_norm": 6.80520486831665,
      "learning_rate": 4.196265953652809e-05,
      "loss": 2.4649124145507812,
      "memory(GiB)": 77.56,
      "step": 64350,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 2.7571655027633777,
      "grad_norm": 4.921061992645264,
      "learning_rate": 4.195601735245658e-05,
      "loss": 2.4463294982910155,
      "memory(GiB)": 77.56,
      "step": 64355,
      "token_acc": 0.5050167224080268,
      "train_speed(iter/s)": 1.437606
    },
    {
      "epoch": 2.7573797180926265,
      "grad_norm": 6.679776668548584,
      "learning_rate": 4.194937531410937e-05,
      "loss": 2.5255287170410154,
      "memory(GiB)": 77.56,
      "step": 64360,
      "token_acc": 0.45317220543806647,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 2.7575939334218758,
      "grad_norm": 4.265938758850098,
      "learning_rate": 4.1942733421606764e-05,
      "loss": 2.481113052368164,
      "memory(GiB)": 77.56,
      "step": 64365,
      "token_acc": 0.4670846394984326,
      "train_speed(iter/s)": 1.437636
    },
    {
      "epoch": 2.7578081487511246,
      "grad_norm": 6.029409885406494,
      "learning_rate": 4.193609167506912e-05,
      "loss": 2.729705047607422,
      "memory(GiB)": 77.56,
      "step": 64370,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 2.7580223640803734,
      "grad_norm": 5.590997219085693,
      "learning_rate": 4.1929450074616734e-05,
      "loss": 2.3650768280029295,
      "memory(GiB)": 77.56,
      "step": 64375,
      "token_acc": 0.4662576687116564,
      "train_speed(iter/s)": 1.437665
    },
    {
      "epoch": 2.7582365794096226,
      "grad_norm": 4.810853004455566,
      "learning_rate": 4.192280862036992e-05,
      "loss": 2.2588579177856447,
      "memory(GiB)": 77.56,
      "step": 64380,
      "token_acc": 0.5152542372881356,
      "train_speed(iter/s)": 1.437671
    },
    {
      "epoch": 2.7584507947388714,
      "grad_norm": 5.563048839569092,
      "learning_rate": 4.1916167312449014e-05,
      "loss": 2.512114715576172,
      "memory(GiB)": 77.56,
      "step": 64385,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.437686
    },
    {
      "epoch": 2.7586650100681203,
      "grad_norm": 6.968373775482178,
      "learning_rate": 4.1909526150974326e-05,
      "loss": 2.564522552490234,
      "memory(GiB)": 77.56,
      "step": 64390,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.437674
    },
    {
      "epoch": 2.7588792253973695,
      "grad_norm": 5.360578536987305,
      "learning_rate": 4.190288513606615e-05,
      "loss": 2.151266670227051,
      "memory(GiB)": 77.56,
      "step": 64395,
      "token_acc": 0.5527156549520766,
      "train_speed(iter/s)": 1.437687
    },
    {
      "epoch": 2.7590934407266183,
      "grad_norm": 4.924160480499268,
      "learning_rate": 4.189624426784481e-05,
      "loss": 2.451461410522461,
      "memory(GiB)": 77.56,
      "step": 64400,
      "token_acc": 0.5196078431372549,
      "train_speed(iter/s)": 1.437665
    },
    {
      "epoch": 2.759307656055867,
      "grad_norm": 6.2384724617004395,
      "learning_rate": 4.18896035464306e-05,
      "loss": 2.4778888702392576,
      "memory(GiB)": 77.56,
      "step": 64405,
      "token_acc": 0.4892966360856269,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 2.7595218713851164,
      "grad_norm": 5.335391998291016,
      "learning_rate": 4.188296297194384e-05,
      "loss": 2.6137153625488283,
      "memory(GiB)": 77.56,
      "step": 64410,
      "token_acc": 0.4601449275362319,
      "train_speed(iter/s)": 1.437697
    },
    {
      "epoch": 2.759736086714365,
      "grad_norm": 5.631296157836914,
      "learning_rate": 4.18763225445048e-05,
      "loss": 2.6373186111450195,
      "memory(GiB)": 77.56,
      "step": 64415,
      "token_acc": 0.44297082228116713,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 2.759950302043614,
      "grad_norm": 5.207281589508057,
      "learning_rate": 4.1869682264233826e-05,
      "loss": 2.363142204284668,
      "memory(GiB)": 77.56,
      "step": 64420,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.437705
    },
    {
      "epoch": 2.7601645173728633,
      "grad_norm": 5.0512871742248535,
      "learning_rate": 4.186304213125116e-05,
      "loss": 2.500602149963379,
      "memory(GiB)": 77.56,
      "step": 64425,
      "token_acc": 0.5098814229249012,
      "train_speed(iter/s)": 1.43772
    },
    {
      "epoch": 2.760378732702112,
      "grad_norm": 7.788637638092041,
      "learning_rate": 4.185640214567714e-05,
      "loss": 2.4444881439208985,
      "memory(GiB)": 77.56,
      "step": 64430,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 2.760592948031361,
      "grad_norm": 7.587400913238525,
      "learning_rate": 4.184976230763203e-05,
      "loss": 2.3888885498046877,
      "memory(GiB)": 77.56,
      "step": 64435,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 2.76080716336061,
      "grad_norm": 5.943777561187744,
      "learning_rate": 4.1843122617236106e-05,
      "loss": 2.4553855895996093,
      "memory(GiB)": 77.56,
      "step": 64440,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.437693
    },
    {
      "epoch": 2.761021378689859,
      "grad_norm": 7.957814693450928,
      "learning_rate": 4.183648307460969e-05,
      "loss": 2.4153724670410157,
      "memory(GiB)": 77.56,
      "step": 64445,
      "token_acc": 0.4567901234567901,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 2.761235594019108,
      "grad_norm": 5.246782302856445,
      "learning_rate": 4.182984367987302e-05,
      "loss": 2.5522727966308594,
      "memory(GiB)": 77.56,
      "step": 64450,
      "token_acc": 0.43312101910828027,
      "train_speed(iter/s)": 1.437763
    },
    {
      "epoch": 2.761449809348357,
      "grad_norm": 5.906622409820557,
      "learning_rate": 4.182320443314641e-05,
      "loss": 2.1670705795288088,
      "memory(GiB)": 77.56,
      "step": 64455,
      "token_acc": 0.5395189003436426,
      "train_speed(iter/s)": 1.437782
    },
    {
      "epoch": 2.761664024677606,
      "grad_norm": 5.36324405670166,
      "learning_rate": 4.181656533455013e-05,
      "loss": 2.5967132568359377,
      "memory(GiB)": 77.56,
      "step": 64460,
      "token_acc": 0.47474747474747475,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 2.7618782400068547,
      "grad_norm": 7.413851261138916,
      "learning_rate": 4.1809926384204426e-05,
      "loss": 2.175606536865234,
      "memory(GiB)": 77.56,
      "step": 64465,
      "token_acc": 0.5568181818181818,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 2.762092455336104,
      "grad_norm": 4.746966361999512,
      "learning_rate": 4.18032875822296e-05,
      "loss": 2.280744743347168,
      "memory(GiB)": 77.56,
      "step": 64470,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 2.7623066706653527,
      "grad_norm": 8.898507118225098,
      "learning_rate": 4.179664892874591e-05,
      "loss": 2.388660430908203,
      "memory(GiB)": 77.56,
      "step": 64475,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.437767
    },
    {
      "epoch": 2.7625208859946015,
      "grad_norm": 6.157951354980469,
      "learning_rate": 4.17900104238736e-05,
      "loss": 2.45493106842041,
      "memory(GiB)": 77.56,
      "step": 64480,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 2.762735101323851,
      "grad_norm": 4.3204665184021,
      "learning_rate": 4.1783372067732977e-05,
      "loss": 2.4240743637084963,
      "memory(GiB)": 77.56,
      "step": 64485,
      "token_acc": 0.493006993006993,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 2.7629493166530996,
      "grad_norm": 5.623483180999756,
      "learning_rate": 4.177673386044425e-05,
      "loss": 2.080682945251465,
      "memory(GiB)": 77.56,
      "step": 64490,
      "token_acc": 0.5186721991701245,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 2.7631635319823484,
      "grad_norm": 5.1229963302612305,
      "learning_rate": 4.177009580212773e-05,
      "loss": 2.451407051086426,
      "memory(GiB)": 77.56,
      "step": 64495,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.437796
    },
    {
      "epoch": 2.7633777473115977,
      "grad_norm": 6.0577497482299805,
      "learning_rate": 4.176345789290363e-05,
      "loss": 2.3344985961914064,
      "memory(GiB)": 77.56,
      "step": 64500,
      "token_acc": 0.5431309904153354,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 2.7633777473115977,
      "eval_loss": 2.1342391967773438,
      "eval_runtime": 14.1941,
      "eval_samples_per_second": 7.045,
      "eval_steps_per_second": 7.045,
      "eval_token_acc": 0.48824343015214383,
      "step": 64500
    },
    {
      "epoch": 2.7635919626408465,
      "grad_norm": 5.9025959968566895,
      "learning_rate": 4.175682013289223e-05,
      "loss": 2.546734619140625,
      "memory(GiB)": 77.56,
      "step": 64505,
      "token_acc": 0.4811800610376399,
      "train_speed(iter/s)": 1.437345
    },
    {
      "epoch": 2.7638061779700953,
      "grad_norm": 5.714148044586182,
      "learning_rate": 4.1750182522213745e-05,
      "loss": 2.6803285598754885,
      "memory(GiB)": 77.56,
      "step": 64510,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.43735
    },
    {
      "epoch": 2.7640203932993446,
      "grad_norm": 5.070879936218262,
      "learning_rate": 4.174354506098847e-05,
      "loss": 2.26236572265625,
      "memory(GiB)": 77.56,
      "step": 64515,
      "token_acc": 0.5422077922077922,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 2.7642346086285934,
      "grad_norm": 5.840444564819336,
      "learning_rate": 4.1736907749336603e-05,
      "loss": 2.308195686340332,
      "memory(GiB)": 77.56,
      "step": 64520,
      "token_acc": 0.4900662251655629,
      "train_speed(iter/s)": 1.437349
    },
    {
      "epoch": 2.764448823957842,
      "grad_norm": 6.4930596351623535,
      "learning_rate": 4.17302705873784e-05,
      "loss": 2.465194320678711,
      "memory(GiB)": 77.56,
      "step": 64525,
      "token_acc": 0.45671641791044776,
      "train_speed(iter/s)": 1.437328
    },
    {
      "epoch": 2.7646630392870915,
      "grad_norm": 5.947774887084961,
      "learning_rate": 4.172363357523412e-05,
      "loss": 2.484974479675293,
      "memory(GiB)": 77.56,
      "step": 64530,
      "token_acc": 0.5016501650165016,
      "train_speed(iter/s)": 1.43732
    },
    {
      "epoch": 2.7648772546163403,
      "grad_norm": 5.339717388153076,
      "learning_rate": 4.1716996713023976e-05,
      "loss": 2.369556427001953,
      "memory(GiB)": 77.56,
      "step": 64535,
      "token_acc": 0.5018181818181818,
      "train_speed(iter/s)": 1.437314
    },
    {
      "epoch": 2.765091469945589,
      "grad_norm": 7.05649471282959,
      "learning_rate": 4.17103600008682e-05,
      "loss": 2.4731819152832033,
      "memory(GiB)": 77.56,
      "step": 64540,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.437331
    },
    {
      "epoch": 2.7653056852748383,
      "grad_norm": 5.375307559967041,
      "learning_rate": 4.170372343888703e-05,
      "loss": 2.378491020202637,
      "memory(GiB)": 77.56,
      "step": 64545,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 2.765519900604087,
      "grad_norm": 6.3528594970703125,
      "learning_rate": 4.169708702720069e-05,
      "loss": 2.4149532318115234,
      "memory(GiB)": 77.56,
      "step": 64550,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.437374
    },
    {
      "epoch": 2.765734115933336,
      "grad_norm": 6.189089298248291,
      "learning_rate": 4.169045076592942e-05,
      "loss": 2.469185447692871,
      "memory(GiB)": 77.56,
      "step": 64555,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.437391
    },
    {
      "epoch": 2.765948331262585,
      "grad_norm": 5.242286205291748,
      "learning_rate": 4.168381465519342e-05,
      "loss": 2.5613691329956056,
      "memory(GiB)": 77.56,
      "step": 64560,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 2.766162546591834,
      "grad_norm": 4.933730125427246,
      "learning_rate": 4.167717869511291e-05,
      "loss": 2.4313411712646484,
      "memory(GiB)": 77.56,
      "step": 64565,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.437407
    },
    {
      "epoch": 2.766376761921083,
      "grad_norm": 6.2235798835754395,
      "learning_rate": 4.167054288580812e-05,
      "loss": 2.370328903198242,
      "memory(GiB)": 77.56,
      "step": 64570,
      "token_acc": 0.5160142348754448,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 2.766590977250332,
      "grad_norm": 6.177143573760986,
      "learning_rate": 4.166390722739926e-05,
      "loss": 2.471444320678711,
      "memory(GiB)": 77.56,
      "step": 64575,
      "token_acc": 0.49606299212598426,
      "train_speed(iter/s)": 1.437406
    },
    {
      "epoch": 2.766805192579581,
      "grad_norm": 4.598169326782227,
      "learning_rate": 4.165727172000655e-05,
      "loss": 2.2622312545776366,
      "memory(GiB)": 77.56,
      "step": 64580,
      "token_acc": 0.5218978102189781,
      "train_speed(iter/s)": 1.437416
    },
    {
      "epoch": 2.7670194079088297,
      "grad_norm": 6.439113616943359,
      "learning_rate": 4.165063636375018e-05,
      "loss": 2.3918380737304688,
      "memory(GiB)": 77.56,
      "step": 64585,
      "token_acc": 0.4954682779456193,
      "train_speed(iter/s)": 1.437414
    },
    {
      "epoch": 2.767233623238079,
      "grad_norm": 5.608087062835693,
      "learning_rate": 4.164400115875037e-05,
      "loss": 2.027151107788086,
      "memory(GiB)": 77.56,
      "step": 64590,
      "token_acc": 0.5418502202643172,
      "train_speed(iter/s)": 1.437364
    },
    {
      "epoch": 2.767447838567328,
      "grad_norm": 5.899224281311035,
      "learning_rate": 4.1637366105127315e-05,
      "loss": 2.553853988647461,
      "memory(GiB)": 77.56,
      "step": 64595,
      "token_acc": 0.4728682170542636,
      "train_speed(iter/s)": 1.437376
    },
    {
      "epoch": 2.7676620538965766,
      "grad_norm": 6.398929119110107,
      "learning_rate": 4.163073120300122e-05,
      "loss": 2.4647218704223635,
      "memory(GiB)": 77.56,
      "step": 64600,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.437362
    },
    {
      "epoch": 2.767876269225826,
      "grad_norm": 4.618252277374268,
      "learning_rate": 4.162409645249228e-05,
      "loss": 2.3526901245117187,
      "memory(GiB)": 77.56,
      "step": 64605,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 2.7680904845550747,
      "grad_norm": 4.611629009246826,
      "learning_rate": 4.1617461853720685e-05,
      "loss": 2.2983558654785154,
      "memory(GiB)": 77.56,
      "step": 64610,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437391
    },
    {
      "epoch": 2.7683046998843235,
      "grad_norm": 7.5736541748046875,
      "learning_rate": 4.161082740680664e-05,
      "loss": 2.33463134765625,
      "memory(GiB)": 77.56,
      "step": 64615,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.437413
    },
    {
      "epoch": 2.7685189152135727,
      "grad_norm": 6.545660018920898,
      "learning_rate": 4.160419311187033e-05,
      "loss": 2.5211572647094727,
      "memory(GiB)": 77.56,
      "step": 64620,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.43741
    },
    {
      "epoch": 2.7687331305428216,
      "grad_norm": 7.074057102203369,
      "learning_rate": 4.1597558969031924e-05,
      "loss": 2.0580442428588865,
      "memory(GiB)": 77.56,
      "step": 64625,
      "token_acc": 0.5336134453781513,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 2.7689473458720704,
      "grad_norm": 7.0091328620910645,
      "learning_rate": 4.159092497841163e-05,
      "loss": 2.3606201171875,
      "memory(GiB)": 77.56,
      "step": 64630,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.437359
    },
    {
      "epoch": 2.7691615612013196,
      "grad_norm": 5.173372268676758,
      "learning_rate": 4.15842911401296e-05,
      "loss": 2.7328081130981445,
      "memory(GiB)": 77.56,
      "step": 64635,
      "token_acc": 0.47249190938511326,
      "train_speed(iter/s)": 1.43738
    },
    {
      "epoch": 2.7693757765305684,
      "grad_norm": 4.831366062164307,
      "learning_rate": 4.157765745430605e-05,
      "loss": 2.5863250732421874,
      "memory(GiB)": 77.56,
      "step": 64640,
      "token_acc": 0.43352601156069365,
      "train_speed(iter/s)": 1.437405
    },
    {
      "epoch": 2.7695899918598172,
      "grad_norm": 7.955633163452148,
      "learning_rate": 4.157102392106112e-05,
      "loss": 2.4766910552978514,
      "memory(GiB)": 77.56,
      "step": 64645,
      "token_acc": 0.4662576687116564,
      "train_speed(iter/s)": 1.437425
    },
    {
      "epoch": 2.7698042071890665,
      "grad_norm": 5.130234718322754,
      "learning_rate": 4.156439054051501e-05,
      "loss": 2.302436637878418,
      "memory(GiB)": 77.56,
      "step": 64650,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 2.7700184225183153,
      "grad_norm": 5.311408996582031,
      "learning_rate": 4.1557757312787874e-05,
      "loss": 2.4721141815185548,
      "memory(GiB)": 77.56,
      "step": 64655,
      "token_acc": 0.4697508896797153,
      "train_speed(iter/s)": 1.437418
    },
    {
      "epoch": 2.770232637847564,
      "grad_norm": 6.5296711921691895,
      "learning_rate": 4.1551124237999895e-05,
      "loss": 2.4906278610229493,
      "memory(GiB)": 77.56,
      "step": 64660,
      "token_acc": 0.5117056856187291,
      "train_speed(iter/s)": 1.437444
    },
    {
      "epoch": 2.7704468531768134,
      "grad_norm": 5.803679466247559,
      "learning_rate": 4.1544491316271224e-05,
      "loss": 2.27764835357666,
      "memory(GiB)": 77.56,
      "step": 64665,
      "token_acc": 0.4937106918238994,
      "train_speed(iter/s)": 1.437437
    },
    {
      "epoch": 2.770661068506062,
      "grad_norm": 4.957920074462891,
      "learning_rate": 4.153785854772201e-05,
      "loss": 2.3398094177246094,
      "memory(GiB)": 77.56,
      "step": 64670,
      "token_acc": 0.5150501672240803,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 2.770875283835311,
      "grad_norm": 6.413315773010254,
      "learning_rate": 4.153122593247244e-05,
      "loss": 2.6030054092407227,
      "memory(GiB)": 77.56,
      "step": 64675,
      "token_acc": 0.4375,
      "train_speed(iter/s)": 1.437441
    },
    {
      "epoch": 2.7710894991645603,
      "grad_norm": 6.683399677276611,
      "learning_rate": 4.1524593470642656e-05,
      "loss": 2.6472394943237303,
      "memory(GiB)": 77.56,
      "step": 64680,
      "token_acc": 0.4492307692307692,
      "train_speed(iter/s)": 1.437458
    },
    {
      "epoch": 2.771303714493809,
      "grad_norm": 5.128006458282471,
      "learning_rate": 4.15179611623528e-05,
      "loss": 2.2427282333374023,
      "memory(GiB)": 77.56,
      "step": 64685,
      "token_acc": 0.47560975609756095,
      "train_speed(iter/s)": 1.437467
    },
    {
      "epoch": 2.771517929823058,
      "grad_norm": 5.724479675292969,
      "learning_rate": 4.1511329007723046e-05,
      "loss": 2.698308563232422,
      "memory(GiB)": 77.56,
      "step": 64690,
      "token_acc": 0.4513888888888889,
      "train_speed(iter/s)": 1.437456
    },
    {
      "epoch": 2.771732145152307,
      "grad_norm": 6.818203449249268,
      "learning_rate": 4.1504697006873524e-05,
      "loss": 2.5303741455078126,
      "memory(GiB)": 77.56,
      "step": 64695,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.437472
    },
    {
      "epoch": 2.771946360481556,
      "grad_norm": 4.720088958740234,
      "learning_rate": 4.1498065159924394e-05,
      "loss": 2.213025665283203,
      "memory(GiB)": 77.56,
      "step": 64700,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.437469
    },
    {
      "epoch": 2.7721605758108048,
      "grad_norm": 3.9661316871643066,
      "learning_rate": 4.149143346699579e-05,
      "loss": 2.248960494995117,
      "memory(GiB)": 77.56,
      "step": 64705,
      "token_acc": 0.5687022900763359,
      "train_speed(iter/s)": 1.437486
    },
    {
      "epoch": 2.772374791140054,
      "grad_norm": 8.90654468536377,
      "learning_rate": 4.1484801928207824e-05,
      "loss": 2.401824188232422,
      "memory(GiB)": 77.56,
      "step": 64710,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.43746
    },
    {
      "epoch": 2.772589006469303,
      "grad_norm": 11.235576629638672,
      "learning_rate": 4.1478170543680664e-05,
      "loss": 2.220451545715332,
      "memory(GiB)": 77.56,
      "step": 64715,
      "token_acc": 0.5503875968992248,
      "train_speed(iter/s)": 1.437453
    },
    {
      "epoch": 2.7728032217985517,
      "grad_norm": 5.220694541931152,
      "learning_rate": 4.147153931353446e-05,
      "loss": 2.2256542205810548,
      "memory(GiB)": 77.56,
      "step": 64720,
      "token_acc": 0.4911660777385159,
      "train_speed(iter/s)": 1.437454
    },
    {
      "epoch": 2.773017437127801,
      "grad_norm": 6.951962471008301,
      "learning_rate": 4.1464908237889324e-05,
      "loss": 2.3189970016479493,
      "memory(GiB)": 77.56,
      "step": 64725,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43748
    },
    {
      "epoch": 2.7732316524570497,
      "grad_norm": 5.977963924407959,
      "learning_rate": 4.145827731686536e-05,
      "loss": 2.4968643188476562,
      "memory(GiB)": 77.56,
      "step": 64730,
      "token_acc": 0.4520547945205479,
      "train_speed(iter/s)": 1.437491
    },
    {
      "epoch": 2.7734458677862985,
      "grad_norm": 8.230284690856934,
      "learning_rate": 4.145164655058273e-05,
      "loss": 2.6562217712402343,
      "memory(GiB)": 77.56,
      "step": 64735,
      "token_acc": 0.44654088050314467,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 2.773660083115548,
      "grad_norm": 5.399183750152588,
      "learning_rate": 4.144501593916154e-05,
      "loss": 2.4795648574829103,
      "memory(GiB)": 77.56,
      "step": 64740,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 2.7738742984447966,
      "grad_norm": 4.590023040771484,
      "learning_rate": 4.1438385482721913e-05,
      "loss": 2.0696184158325197,
      "memory(GiB)": 77.56,
      "step": 64745,
      "token_acc": 0.5963636363636363,
      "train_speed(iter/s)": 1.437562
    },
    {
      "epoch": 2.7740885137740454,
      "grad_norm": 4.3344502449035645,
      "learning_rate": 4.143175518138397e-05,
      "loss": 2.6216949462890624,
      "memory(GiB)": 77.56,
      "step": 64750,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.437559
    },
    {
      "epoch": 2.7743027291032947,
      "grad_norm": 6.110363483428955,
      "learning_rate": 4.14251250352678e-05,
      "loss": 2.2581365585327147,
      "memory(GiB)": 77.56,
      "step": 64755,
      "token_acc": 0.4789272030651341,
      "train_speed(iter/s)": 1.437572
    },
    {
      "epoch": 2.7745169444325435,
      "grad_norm": 4.645518779754639,
      "learning_rate": 4.141849504449355e-05,
      "loss": 2.5854907989501954,
      "memory(GiB)": 77.56,
      "step": 64760,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.437591
    },
    {
      "epoch": 2.7747311597617923,
      "grad_norm": 4.550884246826172,
      "learning_rate": 4.141186520918132e-05,
      "loss": 2.709996223449707,
      "memory(GiB)": 77.56,
      "step": 64765,
      "token_acc": 0.445859872611465,
      "train_speed(iter/s)": 1.437575
    },
    {
      "epoch": 2.7749453750910416,
      "grad_norm": 5.34348201751709,
      "learning_rate": 4.140523552945118e-05,
      "loss": 2.696388816833496,
      "memory(GiB)": 77.56,
      "step": 64770,
      "token_acc": 0.45664739884393063,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 2.7751595904202904,
      "grad_norm": 5.4660964012146,
      "learning_rate": 4.1398606005423284e-05,
      "loss": 2.664356994628906,
      "memory(GiB)": 77.56,
      "step": 64775,
      "token_acc": 0.4218289085545723,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 2.775373805749539,
      "grad_norm": 7.453643798828125,
      "learning_rate": 4.13919766372177e-05,
      "loss": 2.270530891418457,
      "memory(GiB)": 77.56,
      "step": 64780,
      "token_acc": 0.5043478260869565,
      "train_speed(iter/s)": 1.437605
    },
    {
      "epoch": 2.7755880210787884,
      "grad_norm": 8.337597846984863,
      "learning_rate": 4.1385347424954526e-05,
      "loss": 2.6300540924072267,
      "memory(GiB)": 77.56,
      "step": 64785,
      "token_acc": 0.4405797101449275,
      "train_speed(iter/s)": 1.437622
    },
    {
      "epoch": 2.7758022364080372,
      "grad_norm": 5.461379528045654,
      "learning_rate": 4.137871836875387e-05,
      "loss": 2.4108423233032226,
      "memory(GiB)": 77.56,
      "step": 64790,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 2.776016451737286,
      "grad_norm": 5.3231024742126465,
      "learning_rate": 4.137208946873582e-05,
      "loss": 2.436798095703125,
      "memory(GiB)": 77.56,
      "step": 64795,
      "token_acc": 0.46863468634686345,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 2.7762306670665353,
      "grad_norm": 7.851663112640381,
      "learning_rate": 4.1365460725020466e-05,
      "loss": 2.346063995361328,
      "memory(GiB)": 77.56,
      "step": 64800,
      "token_acc": 0.4495798319327731,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 2.776444882395784,
      "grad_norm": 5.869938850402832,
      "learning_rate": 4.135883213772789e-05,
      "loss": 2.8061729431152345,
      "memory(GiB)": 77.56,
      "step": 64805,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 2.776659097725033,
      "grad_norm": 4.127633571624756,
      "learning_rate": 4.1352203706978186e-05,
      "loss": 2.3608001708984374,
      "memory(GiB)": 77.56,
      "step": 64810,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437653
    },
    {
      "epoch": 2.776873313054282,
      "grad_norm": 5.486953258514404,
      "learning_rate": 4.134557543289141e-05,
      "loss": 2.616672897338867,
      "memory(GiB)": 77.56,
      "step": 64815,
      "token_acc": 0.48632218844984804,
      "train_speed(iter/s)": 1.437655
    },
    {
      "epoch": 2.777087528383531,
      "grad_norm": 6.279545307159424,
      "learning_rate": 4.1338947315587664e-05,
      "loss": 2.276936721801758,
      "memory(GiB)": 77.56,
      "step": 64820,
      "token_acc": 0.5202952029520295,
      "train_speed(iter/s)": 1.437659
    },
    {
      "epoch": 2.77730174371278,
      "grad_norm": 5.216151714324951,
      "learning_rate": 4.133231935518701e-05,
      "loss": 2.5696372985839844,
      "memory(GiB)": 77.56,
      "step": 64825,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.437638
    },
    {
      "epoch": 2.777515959042029,
      "grad_norm": 4.620060920715332,
      "learning_rate": 4.132569155180951e-05,
      "loss": 2.619928550720215,
      "memory(GiB)": 77.56,
      "step": 64830,
      "token_acc": 0.44358974358974357,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 2.777730174371278,
      "grad_norm": 5.995078086853027,
      "learning_rate": 4.131906390557526e-05,
      "loss": 2.3185092926025392,
      "memory(GiB)": 77.56,
      "step": 64835,
      "token_acc": 0.4584837545126354,
      "train_speed(iter/s)": 1.43763
    },
    {
      "epoch": 2.7779443897005267,
      "grad_norm": 4.878668308258057,
      "learning_rate": 4.131243641660429e-05,
      "loss": 2.600627326965332,
      "memory(GiB)": 77.56,
      "step": 64840,
      "token_acc": 0.4740061162079511,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 2.778158605029776,
      "grad_norm": 4.560103416442871,
      "learning_rate": 4.130580908501671e-05,
      "loss": 2.4108642578125,
      "memory(GiB)": 77.56,
      "step": 64845,
      "token_acc": 0.5111940298507462,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 2.7783728203590248,
      "grad_norm": 6.4313578605651855,
      "learning_rate": 4.129918191093254e-05,
      "loss": 2.3842390060424803,
      "memory(GiB)": 77.56,
      "step": 64850,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 2.7785870356882736,
      "grad_norm": 5.557571887969971,
      "learning_rate": 4.1292554894471847e-05,
      "loss": 2.424895095825195,
      "memory(GiB)": 77.56,
      "step": 64855,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.437694
    },
    {
      "epoch": 2.778801251017523,
      "grad_norm": 5.516140937805176,
      "learning_rate": 4.1285928035754684e-05,
      "loss": 2.4183223724365233,
      "memory(GiB)": 77.56,
      "step": 64860,
      "token_acc": 0.4521072796934866,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 2.7790154663467717,
      "grad_norm": 7.888944149017334,
      "learning_rate": 4.127930133490112e-05,
      "loss": 2.245655822753906,
      "memory(GiB)": 77.56,
      "step": 64865,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.437746
    },
    {
      "epoch": 2.7792296816760205,
      "grad_norm": 4.838222026824951,
      "learning_rate": 4.12726747920312e-05,
      "loss": 2.3545743942260744,
      "memory(GiB)": 77.56,
      "step": 64870,
      "token_acc": 0.47572815533980584,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 2.7794438970052697,
      "grad_norm": 5.020430088043213,
      "learning_rate": 4.126604840726496e-05,
      "loss": 2.4572126388549806,
      "memory(GiB)": 77.56,
      "step": 64875,
      "token_acc": 0.512,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 2.7796581123345185,
      "grad_norm": 5.911706924438477,
      "learning_rate": 4.125942218072244e-05,
      "loss": 2.3762020111083983,
      "memory(GiB)": 77.56,
      "step": 64880,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 2.7798723276637674,
      "grad_norm": 4.535211563110352,
      "learning_rate": 4.125279611252369e-05,
      "loss": 2.378687286376953,
      "memory(GiB)": 77.56,
      "step": 64885,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.437767
    },
    {
      "epoch": 2.7800865429930166,
      "grad_norm": 5.900091171264648,
      "learning_rate": 4.124617020278875e-05,
      "loss": 2.5336162567138674,
      "memory(GiB)": 77.56,
      "step": 64890,
      "token_acc": 0.4732824427480916,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 2.7803007583222654,
      "grad_norm": 4.068755149841309,
      "learning_rate": 4.1239544451637646e-05,
      "loss": 2.3729990005493162,
      "memory(GiB)": 77.56,
      "step": 64895,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 2.7805149736515142,
      "grad_norm": 6.493914604187012,
      "learning_rate": 4.12329188591904e-05,
      "loss": 2.2886421203613283,
      "memory(GiB)": 77.56,
      "step": 64900,
      "token_acc": 0.5086505190311419,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 2.7807291889807635,
      "grad_norm": 8.492538452148438,
      "learning_rate": 4.122629342556706e-05,
      "loss": 2.275677490234375,
      "memory(GiB)": 77.56,
      "step": 64905,
      "token_acc": 0.498371335504886,
      "train_speed(iter/s)": 1.437771
    },
    {
      "epoch": 2.7809434043100123,
      "grad_norm": 5.051513195037842,
      "learning_rate": 4.121966815088766e-05,
      "loss": 2.4003904342651365,
      "memory(GiB)": 77.56,
      "step": 64910,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 2.781157619639261,
      "grad_norm": 5.041954040527344,
      "learning_rate": 4.1213043035272184e-05,
      "loss": 2.3659652709960937,
      "memory(GiB)": 77.56,
      "step": 64915,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 2.7813718349685104,
      "grad_norm": 6.168399810791016,
      "learning_rate": 4.12064180788407e-05,
      "loss": 2.443279838562012,
      "memory(GiB)": 77.56,
      "step": 64920,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.437744
    },
    {
      "epoch": 2.781586050297759,
      "grad_norm": 5.1472673416137695,
      "learning_rate": 4.1199793281713176e-05,
      "loss": 2.4265264511108398,
      "memory(GiB)": 77.56,
      "step": 64925,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 2.781800265627008,
      "grad_norm": 6.6545586585998535,
      "learning_rate": 4.119316864400967e-05,
      "loss": 2.37560977935791,
      "memory(GiB)": 77.56,
      "step": 64930,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 2.7820144809562573,
      "grad_norm": 4.588661193847656,
      "learning_rate": 4.118654416585015e-05,
      "loss": 2.3643335342407226,
      "memory(GiB)": 77.56,
      "step": 64935,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.43774
    },
    {
      "epoch": 2.782228696285506,
      "grad_norm": 6.41018533706665,
      "learning_rate": 4.117991984735468e-05,
      "loss": 2.5099355697631838,
      "memory(GiB)": 77.56,
      "step": 64940,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 2.782442911614755,
      "grad_norm": 4.999509811401367,
      "learning_rate": 4.117329568864322e-05,
      "loss": 2.3964168548583986,
      "memory(GiB)": 77.56,
      "step": 64945,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 2.782657126944004,
      "grad_norm": 5.6082868576049805,
      "learning_rate": 4.11666716898358e-05,
      "loss": 2.533946418762207,
      "memory(GiB)": 77.56,
      "step": 64950,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 2.782871342273253,
      "grad_norm": 6.538122653961182,
      "learning_rate": 4.116004785105241e-05,
      "loss": 2.4954124450683595,
      "memory(GiB)": 77.56,
      "step": 64955,
      "token_acc": 0.49615384615384617,
      "train_speed(iter/s)": 1.43772
    },
    {
      "epoch": 2.7830855576025018,
      "grad_norm": 4.925257682800293,
      "learning_rate": 4.115342417241304e-05,
      "loss": 2.2166458129882813,
      "memory(GiB)": 77.56,
      "step": 64960,
      "token_acc": 0.5363984674329502,
      "train_speed(iter/s)": 1.437751
    },
    {
      "epoch": 2.783299772931751,
      "grad_norm": 6.454108238220215,
      "learning_rate": 4.114680065403769e-05,
      "loss": 2.402741050720215,
      "memory(GiB)": 77.56,
      "step": 64965,
      "token_acc": 0.46996466431095407,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.783513988261,
      "grad_norm": 5.182466506958008,
      "learning_rate": 4.114017729604635e-05,
      "loss": 2.4790927886962892,
      "memory(GiB)": 77.56,
      "step": 64970,
      "token_acc": 0.4628975265017668,
      "train_speed(iter/s)": 1.437753
    },
    {
      "epoch": 2.783728203590249,
      "grad_norm": 6.383825302124023,
      "learning_rate": 4.113355409855901e-05,
      "loss": 2.2500200271606445,
      "memory(GiB)": 77.56,
      "step": 64975,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 2.783942418919498,
      "grad_norm": 4.710538864135742,
      "learning_rate": 4.1126931061695656e-05,
      "loss": 2.3343473434448243,
      "memory(GiB)": 77.56,
      "step": 64980,
      "token_acc": 0.5223367697594502,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 2.7841566342487467,
      "grad_norm": 6.233351230621338,
      "learning_rate": 4.112030818557626e-05,
      "loss": 2.7330272674560545,
      "memory(GiB)": 77.56,
      "step": 64985,
      "token_acc": 0.40418118466898956,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 2.784370849577996,
      "grad_norm": 6.382244110107422,
      "learning_rate": 4.111368547032083e-05,
      "loss": 2.7764728546142576,
      "memory(GiB)": 77.56,
      "step": 64990,
      "token_acc": 0.4460431654676259,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 2.784585064907245,
      "grad_norm": 6.608416557312012,
      "learning_rate": 4.110706291604931e-05,
      "loss": 2.515018653869629,
      "memory(GiB)": 77.56,
      "step": 64995,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.437751
    },
    {
      "epoch": 2.7847992802364936,
      "grad_norm": 6.346921443939209,
      "learning_rate": 4.110044052288169e-05,
      "loss": 2.5534263610839845,
      "memory(GiB)": 77.56,
      "step": 65000,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 2.7847992802364936,
      "eval_loss": 2.2311465740203857,
      "eval_runtime": 14.4638,
      "eval_samples_per_second": 6.914,
      "eval_steps_per_second": 6.914,
      "eval_token_acc": 0.4702774108322325,
      "step": 65000
    },
    {
      "epoch": 2.785013495565743,
      "grad_norm": 5.451372146606445,
      "learning_rate": 4.109381829093792e-05,
      "loss": 2.2569353103637697,
      "memory(GiB)": 77.56,
      "step": 65005,
      "token_acc": 0.48096192384769537,
      "train_speed(iter/s)": 1.437266
    },
    {
      "epoch": 2.7852277108949917,
      "grad_norm": 5.394570827484131,
      "learning_rate": 4.108719622033801e-05,
      "loss": 2.6775018692016603,
      "memory(GiB)": 77.56,
      "step": 65010,
      "token_acc": 0.4454828660436137,
      "train_speed(iter/s)": 1.437295
    },
    {
      "epoch": 2.7854419262242405,
      "grad_norm": 5.483050346374512,
      "learning_rate": 4.10805743112019e-05,
      "loss": 2.6770463943481446,
      "memory(GiB)": 77.56,
      "step": 65015,
      "token_acc": 0.45075757575757575,
      "train_speed(iter/s)": 1.437307
    },
    {
      "epoch": 2.7856561415534897,
      "grad_norm": 6.613809585571289,
      "learning_rate": 4.1073952563649546e-05,
      "loss": 2.2119678497314452,
      "memory(GiB)": 77.56,
      "step": 65020,
      "token_acc": 0.5387596899224806,
      "train_speed(iter/s)": 1.437274
    },
    {
      "epoch": 2.7858703568827385,
      "grad_norm": 4.505990028381348,
      "learning_rate": 4.1067330977800924e-05,
      "loss": 2.5043289184570314,
      "memory(GiB)": 77.56,
      "step": 65025,
      "token_acc": 0.4456140350877193,
      "train_speed(iter/s)": 1.437284
    },
    {
      "epoch": 2.7860845722119874,
      "grad_norm": 6.0927324295043945,
      "learning_rate": 4.106070955377597e-05,
      "loss": 2.2570865631103514,
      "memory(GiB)": 77.56,
      "step": 65030,
      "token_acc": 0.5149253731343284,
      "train_speed(iter/s)": 1.437256
    },
    {
      "epoch": 2.7862987875412366,
      "grad_norm": 5.915740489959717,
      "learning_rate": 4.105408829169466e-05,
      "loss": 2.4408477783203124,
      "memory(GiB)": 77.56,
      "step": 65035,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.437251
    },
    {
      "epoch": 2.7865130028704854,
      "grad_norm": 4.612392902374268,
      "learning_rate": 4.104746719167693e-05,
      "loss": 2.6966320037841798,
      "memory(GiB)": 77.56,
      "step": 65040,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.437253
    },
    {
      "epoch": 2.7867272181997342,
      "grad_norm": 5.180809497833252,
      "learning_rate": 4.104084625384272e-05,
      "loss": 2.607496452331543,
      "memory(GiB)": 77.56,
      "step": 65045,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.437274
    },
    {
      "epoch": 2.7869414335289835,
      "grad_norm": 4.020688056945801,
      "learning_rate": 4.103422547831199e-05,
      "loss": 2.6246448516845704,
      "memory(GiB)": 77.56,
      "step": 65050,
      "token_acc": 0.44039735099337746,
      "train_speed(iter/s)": 1.43729
    },
    {
      "epoch": 2.7871556488582323,
      "grad_norm": 4.693151473999023,
      "learning_rate": 4.102760486520468e-05,
      "loss": 2.650675582885742,
      "memory(GiB)": 77.56,
      "step": 65055,
      "token_acc": 0.4641638225255973,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 2.787369864187481,
      "grad_norm": 6.454774856567383,
      "learning_rate": 4.1020984414640716e-05,
      "loss": 2.5119747161865233,
      "memory(GiB)": 77.56,
      "step": 65060,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.43732
    },
    {
      "epoch": 2.7875840795167304,
      "grad_norm": 5.322110176086426,
      "learning_rate": 4.1014364126740056e-05,
      "loss": 2.2648176193237304,
      "memory(GiB)": 77.56,
      "step": 65065,
      "token_acc": 0.5177304964539007,
      "train_speed(iter/s)": 1.437304
    },
    {
      "epoch": 2.787798294845979,
      "grad_norm": 9.234334945678711,
      "learning_rate": 4.100774400162261e-05,
      "loss": 2.3098682403564452,
      "memory(GiB)": 77.56,
      "step": 65070,
      "token_acc": 0.49828178694158076,
      "train_speed(iter/s)": 1.437321
    },
    {
      "epoch": 2.788012510175228,
      "grad_norm": 6.000372409820557,
      "learning_rate": 4.100112403940832e-05,
      "loss": 2.0317359924316407,
      "memory(GiB)": 77.56,
      "step": 65075,
      "token_acc": 0.5755395683453237,
      "train_speed(iter/s)": 1.437326
    },
    {
      "epoch": 2.7882267255044773,
      "grad_norm": 6.502902030944824,
      "learning_rate": 4.099450424021709e-05,
      "loss": 2.38585205078125,
      "memory(GiB)": 77.56,
      "step": 65080,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.437344
    },
    {
      "epoch": 2.788440940833726,
      "grad_norm": 7.568596839904785,
      "learning_rate": 4.0987884604168886e-05,
      "loss": 2.4225059509277345,
      "memory(GiB)": 77.56,
      "step": 65085,
      "token_acc": 0.5015105740181269,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 2.788655156162975,
      "grad_norm": 4.692899703979492,
      "learning_rate": 4.09812651313836e-05,
      "loss": 2.3116535186767577,
      "memory(GiB)": 77.56,
      "step": 65090,
      "token_acc": 0.4657534246575342,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 2.788869371492224,
      "grad_norm": 4.729672908782959,
      "learning_rate": 4.097464582198116e-05,
      "loss": 2.1429588317871096,
      "memory(GiB)": 77.56,
      "step": 65095,
      "token_acc": 0.5666666666666667,
      "train_speed(iter/s)": 1.43737
    },
    {
      "epoch": 2.789083586821473,
      "grad_norm": 4.973909854888916,
      "learning_rate": 4.0968026676081474e-05,
      "loss": 2.6030345916748048,
      "memory(GiB)": 77.56,
      "step": 65100,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.43736
    },
    {
      "epoch": 2.7892978021507218,
      "grad_norm": 6.733554840087891,
      "learning_rate": 4.096140769380445e-05,
      "loss": 2.4143299102783202,
      "memory(GiB)": 77.56,
      "step": 65105,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 2.789512017479971,
      "grad_norm": 4.743870735168457,
      "learning_rate": 4.095478887527002e-05,
      "loss": 2.7167331695556642,
      "memory(GiB)": 77.56,
      "step": 65110,
      "token_acc": 0.40869565217391307,
      "train_speed(iter/s)": 1.437346
    },
    {
      "epoch": 2.78972623280922,
      "grad_norm": 5.542399883270264,
      "learning_rate": 4.094817022059806e-05,
      "loss": 2.4073665618896483,
      "memory(GiB)": 77.56,
      "step": 65115,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.437346
    },
    {
      "epoch": 2.7899404481384686,
      "grad_norm": 5.528306484222412,
      "learning_rate": 4.09415517299085e-05,
      "loss": 2.686728858947754,
      "memory(GiB)": 77.56,
      "step": 65120,
      "token_acc": 0.44816053511705684,
      "train_speed(iter/s)": 1.437352
    },
    {
      "epoch": 2.790154663467718,
      "grad_norm": 4.568327903747559,
      "learning_rate": 4.0934933403321226e-05,
      "loss": 2.0923236846923827,
      "memory(GiB)": 77.56,
      "step": 65125,
      "token_acc": 0.5505226480836237,
      "train_speed(iter/s)": 1.437362
    },
    {
      "epoch": 2.7903688787969667,
      "grad_norm": 6.15056848526001,
      "learning_rate": 4.0928315240956134e-05,
      "loss": 2.405792236328125,
      "memory(GiB)": 77.56,
      "step": 65130,
      "token_acc": 0.5190476190476191,
      "train_speed(iter/s)": 1.437342
    },
    {
      "epoch": 2.7905830941262155,
      "grad_norm": 5.395101070404053,
      "learning_rate": 4.0921697242933125e-05,
      "loss": 2.421052932739258,
      "memory(GiB)": 77.56,
      "step": 65135,
      "token_acc": 0.49240121580547114,
      "train_speed(iter/s)": 1.437368
    },
    {
      "epoch": 2.790797309455465,
      "grad_norm": 5.541062355041504,
      "learning_rate": 4.0915079409372094e-05,
      "loss": 2.3892696380615233,
      "memory(GiB)": 77.56,
      "step": 65140,
      "token_acc": 0.46006389776357826,
      "train_speed(iter/s)": 1.437367
    },
    {
      "epoch": 2.7910115247847136,
      "grad_norm": 5.859163284301758,
      "learning_rate": 4.09084617403929e-05,
      "loss": 2.734787368774414,
      "memory(GiB)": 77.56,
      "step": 65145,
      "token_acc": 0.45317220543806647,
      "train_speed(iter/s)": 1.437355
    },
    {
      "epoch": 2.7912257401139624,
      "grad_norm": 5.762321949005127,
      "learning_rate": 4.0901844236115464e-05,
      "loss": 2.5448539733886717,
      "memory(GiB)": 77.56,
      "step": 65150,
      "token_acc": 0.4662576687116564,
      "train_speed(iter/s)": 1.437345
    },
    {
      "epoch": 2.7914399554432117,
      "grad_norm": 6.067520618438721,
      "learning_rate": 4.089522689665964e-05,
      "loss": 2.402146911621094,
      "memory(GiB)": 77.56,
      "step": 65155,
      "token_acc": 0.48297213622291024,
      "train_speed(iter/s)": 1.437336
    },
    {
      "epoch": 2.7916541707724605,
      "grad_norm": 4.517601490020752,
      "learning_rate": 4.088860972214534e-05,
      "loss": 2.391473388671875,
      "memory(GiB)": 77.56,
      "step": 65160,
      "token_acc": 0.45251396648044695,
      "train_speed(iter/s)": 1.43735
    },
    {
      "epoch": 2.7918683861017093,
      "grad_norm": 5.006802558898926,
      "learning_rate": 4.088199271269241e-05,
      "loss": 2.1704885482788088,
      "memory(GiB)": 77.56,
      "step": 65165,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.437341
    },
    {
      "epoch": 2.7920826014309585,
      "grad_norm": 4.754733085632324,
      "learning_rate": 4.087537586842074e-05,
      "loss": 2.5851518630981447,
      "memory(GiB)": 77.56,
      "step": 65170,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.43734
    },
    {
      "epoch": 2.7922968167602074,
      "grad_norm": 4.961844444274902,
      "learning_rate": 4.086875918945019e-05,
      "loss": 2.3648906707763673,
      "memory(GiB)": 77.56,
      "step": 65175,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 2.792511032089456,
      "grad_norm": 5.78552770614624,
      "learning_rate": 4.0862142675900645e-05,
      "loss": 2.576050567626953,
      "memory(GiB)": 77.56,
      "step": 65180,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.43739
    },
    {
      "epoch": 2.7927252474187054,
      "grad_norm": 5.701693534851074,
      "learning_rate": 4.0855526327891956e-05,
      "loss": 2.504083442687988,
      "memory(GiB)": 77.56,
      "step": 65185,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.437397
    },
    {
      "epoch": 2.7929394627479542,
      "grad_norm": 6.094283580780029,
      "learning_rate": 4.084891014554398e-05,
      "loss": 2.5823450088500977,
      "memory(GiB)": 77.56,
      "step": 65190,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 2.793153678077203,
      "grad_norm": 4.283876419067383,
      "learning_rate": 4.0842294128976586e-05,
      "loss": 2.4167831420898436,
      "memory(GiB)": 77.56,
      "step": 65195,
      "token_acc": 0.5046439628482973,
      "train_speed(iter/s)": 1.437396
    },
    {
      "epoch": 2.7933678934064523,
      "grad_norm": 6.271384239196777,
      "learning_rate": 4.083567827830962e-05,
      "loss": 2.320393371582031,
      "memory(GiB)": 77.56,
      "step": 65200,
      "token_acc": 0.5143769968051118,
      "train_speed(iter/s)": 1.437432
    },
    {
      "epoch": 2.793582108735701,
      "grad_norm": 7.535311698913574,
      "learning_rate": 4.0829062593662944e-05,
      "loss": 2.4804813385009767,
      "memory(GiB)": 77.56,
      "step": 65205,
      "token_acc": 0.43388429752066116,
      "train_speed(iter/s)": 1.437431
    },
    {
      "epoch": 2.79379632406495,
      "grad_norm": 7.255420207977295,
      "learning_rate": 4.08224470751564e-05,
      "loss": 2.4209873199462892,
      "memory(GiB)": 77.56,
      "step": 65210,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 2.794010539394199,
      "grad_norm": 5.833639621734619,
      "learning_rate": 4.081583172290983e-05,
      "loss": 2.108981895446777,
      "memory(GiB)": 77.56,
      "step": 65215,
      "token_acc": 0.49034749034749037,
      "train_speed(iter/s)": 1.437426
    },
    {
      "epoch": 2.794224754723448,
      "grad_norm": 5.483102798461914,
      "learning_rate": 4.080921653704309e-05,
      "loss": 2.21221809387207,
      "memory(GiB)": 77.56,
      "step": 65220,
      "token_acc": 0.49794238683127573,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 2.794438970052697,
      "grad_norm": 5.654956817626953,
      "learning_rate": 4.080260151767602e-05,
      "loss": 2.4338405609130858,
      "memory(GiB)": 77.56,
      "step": 65225,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 2.794653185381946,
      "grad_norm": 5.657629489898682,
      "learning_rate": 4.079598666492843e-05,
      "loss": 2.381987380981445,
      "memory(GiB)": 77.56,
      "step": 65230,
      "token_acc": 0.5036231884057971,
      "train_speed(iter/s)": 1.437441
    },
    {
      "epoch": 2.794867400711195,
      "grad_norm": 4.7143096923828125,
      "learning_rate": 4.0789371978920185e-05,
      "loss": 2.6112834930419924,
      "memory(GiB)": 77.56,
      "step": 65235,
      "token_acc": 0.49363057324840764,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 2.7950816160404437,
      "grad_norm": 5.765244007110596,
      "learning_rate": 4.078275745977112e-05,
      "loss": 2.4692554473876953,
      "memory(GiB)": 77.56,
      "step": 65240,
      "token_acc": 0.4910394265232975,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 2.795295831369693,
      "grad_norm": 6.623595714569092,
      "learning_rate": 4.0776143107601037e-05,
      "loss": 2.2529712677001954,
      "memory(GiB)": 77.56,
      "step": 65245,
      "token_acc": 0.5490909090909091,
      "train_speed(iter/s)": 1.437409
    },
    {
      "epoch": 2.7955100466989418,
      "grad_norm": 4.823298931121826,
      "learning_rate": 4.076952892252977e-05,
      "loss": 2.239645004272461,
      "memory(GiB)": 77.56,
      "step": 65250,
      "token_acc": 0.5176056338028169,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 2.7957242620281906,
      "grad_norm": 5.661194801330566,
      "learning_rate": 4.0762914904677165e-05,
      "loss": 2.4171648025512695,
      "memory(GiB)": 77.56,
      "step": 65255,
      "token_acc": 0.4425287356321839,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 2.79593847735744,
      "grad_norm": 5.11198616027832,
      "learning_rate": 4.0756301054163004e-05,
      "loss": 2.558170700073242,
      "memory(GiB)": 77.56,
      "step": 65260,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.43742
    },
    {
      "epoch": 2.7961526926866886,
      "grad_norm": 4.3902153968811035,
      "learning_rate": 4.074968737110713e-05,
      "loss": 2.3321956634521483,
      "memory(GiB)": 77.56,
      "step": 65265,
      "token_acc": 0.5015479876160991,
      "train_speed(iter/s)": 1.437399
    },
    {
      "epoch": 2.7963669080159375,
      "grad_norm": 5.167755126953125,
      "learning_rate": 4.0743073855629355e-05,
      "loss": 2.502057647705078,
      "memory(GiB)": 77.56,
      "step": 65270,
      "token_acc": 0.45645645645645644,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 2.7965811233451867,
      "grad_norm": 5.482974529266357,
      "learning_rate": 4.073646050784946e-05,
      "loss": 2.311526298522949,
      "memory(GiB)": 77.56,
      "step": 65275,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.437403
    },
    {
      "epoch": 2.7967953386744355,
      "grad_norm": 5.12534761428833,
      "learning_rate": 4.072984732788729e-05,
      "loss": 2.6396320343017576,
      "memory(GiB)": 77.56,
      "step": 65280,
      "token_acc": 0.45645645645645644,
      "train_speed(iter/s)": 1.437412
    },
    {
      "epoch": 2.7970095540036843,
      "grad_norm": 7.030798435211182,
      "learning_rate": 4.072323431586263e-05,
      "loss": 2.56494026184082,
      "memory(GiB)": 77.56,
      "step": 65285,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.437414
    },
    {
      "epoch": 2.7972237693329336,
      "grad_norm": 5.001982688903809,
      "learning_rate": 4.0716621471895275e-05,
      "loss": 2.4871023178100584,
      "memory(GiB)": 77.56,
      "step": 65290,
      "token_acc": 0.5073529411764706,
      "train_speed(iter/s)": 1.437415
    },
    {
      "epoch": 2.7974379846621824,
      "grad_norm": 4.885438919067383,
      "learning_rate": 4.0710008796105034e-05,
      "loss": 2.155305099487305,
      "memory(GiB)": 77.56,
      "step": 65295,
      "token_acc": 0.5690376569037657,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 2.797652199991431,
      "grad_norm": 6.164182662963867,
      "learning_rate": 4.0703396288611694e-05,
      "loss": 2.4674415588378906,
      "memory(GiB)": 77.56,
      "step": 65300,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 2.7978664153206805,
      "grad_norm": 5.3449249267578125,
      "learning_rate": 4.069678394953505e-05,
      "loss": 2.2658706665039063,
      "memory(GiB)": 77.56,
      "step": 65305,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.43743
    },
    {
      "epoch": 2.7980806306499293,
      "grad_norm": 7.248142719268799,
      "learning_rate": 4.069017177899489e-05,
      "loss": 2.6356481552124023,
      "memory(GiB)": 77.56,
      "step": 65310,
      "token_acc": 0.43859649122807015,
      "train_speed(iter/s)": 1.437432
    },
    {
      "epoch": 2.7982948459791785,
      "grad_norm": 6.9197869300842285,
      "learning_rate": 4.0683559777111014e-05,
      "loss": 2.498550224304199,
      "memory(GiB)": 77.56,
      "step": 65315,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 2.7985090613084274,
      "grad_norm": 4.77265739440918,
      "learning_rate": 4.0676947944003175e-05,
      "loss": 2.079168701171875,
      "memory(GiB)": 77.56,
      "step": 65320,
      "token_acc": 0.5611814345991561,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 2.798723276637676,
      "grad_norm": 6.321204662322998,
      "learning_rate": 4.0670336279791186e-05,
      "loss": 2.6338876724243163,
      "memory(GiB)": 77.56,
      "step": 65325,
      "token_acc": 0.4461538461538462,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 2.7989374919669254,
      "grad_norm": 6.061026096343994,
      "learning_rate": 4.066372478459481e-05,
      "loss": 2.3017366409301756,
      "memory(GiB)": 77.56,
      "step": 65330,
      "token_acc": 0.5247148288973384,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 2.7991517072961742,
      "grad_norm": 4.916214466094971,
      "learning_rate": 4.06571134585338e-05,
      "loss": 2.3075334548950197,
      "memory(GiB)": 77.56,
      "step": 65335,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.437431
    },
    {
      "epoch": 2.799365922625423,
      "grad_norm": 5.356903076171875,
      "learning_rate": 4.065050230172796e-05,
      "loss": 2.2832189559936524,
      "memory(GiB)": 77.56,
      "step": 65340,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437436
    },
    {
      "epoch": 2.7995801379546723,
      "grad_norm": 5.1805315017700195,
      "learning_rate": 4.064389131429704e-05,
      "loss": 2.3636260986328126,
      "memory(GiB)": 77.56,
      "step": 65345,
      "token_acc": 0.5134328358208955,
      "train_speed(iter/s)": 1.437442
    },
    {
      "epoch": 2.799794353283921,
      "grad_norm": 5.781987190246582,
      "learning_rate": 4.0637280496360795e-05,
      "loss": 2.7125370025634767,
      "memory(GiB)": 77.56,
      "step": 65350,
      "token_acc": 0.4319526627218935,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 2.80000856861317,
      "grad_norm": 6.091648578643799,
      "learning_rate": 4.0630669848039005e-05,
      "loss": 2.214900016784668,
      "memory(GiB)": 77.56,
      "step": 65355,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.437444
    },
    {
      "epoch": 2.800222783942419,
      "grad_norm": 4.333321571350098,
      "learning_rate": 4.0624059369451415e-05,
      "loss": 2.4089054107666015,
      "memory(GiB)": 77.56,
      "step": 65360,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.437449
    },
    {
      "epoch": 2.800436999271668,
      "grad_norm": 6.672257423400879,
      "learning_rate": 4.061744906071779e-05,
      "loss": 2.5361865997314452,
      "memory(GiB)": 77.56,
      "step": 65365,
      "token_acc": 0.4682080924855491,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 2.800651214600917,
      "grad_norm": 5.3495025634765625,
      "learning_rate": 4.061083892195788e-05,
      "loss": 2.330461883544922,
      "memory(GiB)": 77.56,
      "step": 65370,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.437455
    },
    {
      "epoch": 2.800865429930166,
      "grad_norm": 8.395328521728516,
      "learning_rate": 4.0604228953291404e-05,
      "loss": 2.4021318435668944,
      "memory(GiB)": 77.56,
      "step": 65375,
      "token_acc": 0.4901185770750988,
      "train_speed(iter/s)": 1.437436
    },
    {
      "epoch": 2.801079645259415,
      "grad_norm": 12.81687068939209,
      "learning_rate": 4.059761915483815e-05,
      "loss": 2.1439666748046875,
      "memory(GiB)": 77.56,
      "step": 65380,
      "token_acc": 0.5141700404858299,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 2.8012938605886637,
      "grad_norm": 4.305462837219238,
      "learning_rate": 4.059100952671786e-05,
      "loss": 2.4263351440429686,
      "memory(GiB)": 77.56,
      "step": 65385,
      "token_acc": 0.5045592705167173,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 2.801508075917913,
      "grad_norm": 8.798970222473145,
      "learning_rate": 4.058440006905025e-05,
      "loss": 2.534514236450195,
      "memory(GiB)": 77.56,
      "step": 65390,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.437478
    },
    {
      "epoch": 2.8017222912471618,
      "grad_norm": 6.869740962982178,
      "learning_rate": 4.057779078195506e-05,
      "loss": 2.173324966430664,
      "memory(GiB)": 77.56,
      "step": 65395,
      "token_acc": 0.5037037037037037,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 2.8019365065764106,
      "grad_norm": 4.36246395111084,
      "learning_rate": 4.0571181665552035e-05,
      "loss": 2.2432683944702148,
      "memory(GiB)": 77.56,
      "step": 65400,
      "token_acc": 0.5131086142322098,
      "train_speed(iter/s)": 1.437474
    },
    {
      "epoch": 2.80215072190566,
      "grad_norm": 4.669100284576416,
      "learning_rate": 4.05645727199609e-05,
      "loss": 2.628772163391113,
      "memory(GiB)": 77.56,
      "step": 65405,
      "token_acc": 0.45925925925925926,
      "train_speed(iter/s)": 1.437476
    },
    {
      "epoch": 2.8023649372349086,
      "grad_norm": 5.992696762084961,
      "learning_rate": 4.055796394530138e-05,
      "loss": 2.3717302322387694,
      "memory(GiB)": 77.56,
      "step": 65410,
      "token_acc": 0.49673202614379086,
      "train_speed(iter/s)": 1.437499
    },
    {
      "epoch": 2.8025791525641575,
      "grad_norm": 6.627432346343994,
      "learning_rate": 4.05513553416932e-05,
      "loss": 2.2462282180786133,
      "memory(GiB)": 77.56,
      "step": 65415,
      "token_acc": 0.5175438596491229,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 2.8027933678934067,
      "grad_norm": 5.615395545959473,
      "learning_rate": 4.054474690925607e-05,
      "loss": 2.4067811965942383,
      "memory(GiB)": 77.56,
      "step": 65420,
      "token_acc": 0.4794007490636704,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 2.8030075832226555,
      "grad_norm": 3.8676412105560303,
      "learning_rate": 4.053813864810974e-05,
      "loss": 2.3095130920410156,
      "memory(GiB)": 77.56,
      "step": 65425,
      "token_acc": 0.4811594202898551,
      "train_speed(iter/s)": 1.437499
    },
    {
      "epoch": 2.8032217985519043,
      "grad_norm": 4.624801158905029,
      "learning_rate": 4.05315305583739e-05,
      "loss": 2.5628631591796873,
      "memory(GiB)": 77.56,
      "step": 65430,
      "token_acc": 0.4584615384615385,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 2.8034360138811536,
      "grad_norm": 7.322268009185791,
      "learning_rate": 4.052492264016825e-05,
      "loss": 2.681002616882324,
      "memory(GiB)": 77.56,
      "step": 65435,
      "token_acc": 0.44376899696048633,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 2.8036502292104024,
      "grad_norm": 5.505902290344238,
      "learning_rate": 4.0518314893612535e-05,
      "loss": 2.4549976348876954,
      "memory(GiB)": 77.56,
      "step": 65440,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.437527
    },
    {
      "epoch": 2.8038644445396512,
      "grad_norm": 6.477055072784424,
      "learning_rate": 4.0511707318826426e-05,
      "loss": 2.3828807830810548,
      "memory(GiB)": 77.56,
      "step": 65445,
      "token_acc": 0.5034722222222222,
      "train_speed(iter/s)": 1.437539
    },
    {
      "epoch": 2.8040786598689005,
      "grad_norm": 5.5579752922058105,
      "learning_rate": 4.050509991592964e-05,
      "loss": 2.4007511138916016,
      "memory(GiB)": 77.56,
      "step": 65450,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.437539
    },
    {
      "epoch": 2.8042928751981493,
      "grad_norm": 6.419618129730225,
      "learning_rate": 4.049849268504187e-05,
      "loss": 2.0940181732177736,
      "memory(GiB)": 77.56,
      "step": 65455,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.437531
    },
    {
      "epoch": 2.804507090527398,
      "grad_norm": 6.002536773681641,
      "learning_rate": 4.0491885626282836e-05,
      "loss": 2.3335418701171875,
      "memory(GiB)": 77.56,
      "step": 65460,
      "token_acc": 0.5101351351351351,
      "train_speed(iter/s)": 1.437534
    },
    {
      "epoch": 2.8047213058566474,
      "grad_norm": 6.772785186767578,
      "learning_rate": 4.04852787397722e-05,
      "loss": 2.3547508239746096,
      "memory(GiB)": 77.56,
      "step": 65465,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437537
    },
    {
      "epoch": 2.804935521185896,
      "grad_norm": 5.774203300476074,
      "learning_rate": 4.047867202562967e-05,
      "loss": 2.6899650573730467,
      "memory(GiB)": 77.56,
      "step": 65470,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 2.805149736515145,
      "grad_norm": 6.2780561447143555,
      "learning_rate": 4.0472065483974933e-05,
      "loss": 2.2603769302368164,
      "memory(GiB)": 77.56,
      "step": 65475,
      "token_acc": 0.5111940298507462,
      "train_speed(iter/s)": 1.437578
    },
    {
      "epoch": 2.8053639518443942,
      "grad_norm": 5.402710437774658,
      "learning_rate": 4.046545911492766e-05,
      "loss": 2.4195978164672853,
      "memory(GiB)": 77.56,
      "step": 65480,
      "token_acc": 0.49142857142857144,
      "train_speed(iter/s)": 1.437545
    },
    {
      "epoch": 2.805578167173643,
      "grad_norm": 5.430166244506836,
      "learning_rate": 4.0458852918607545e-05,
      "loss": 2.3975383758544924,
      "memory(GiB)": 77.56,
      "step": 65485,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437558
    },
    {
      "epoch": 2.805792382502892,
      "grad_norm": 6.7078633308410645,
      "learning_rate": 4.0452246895134266e-05,
      "loss": 2.6554012298583984,
      "memory(GiB)": 77.56,
      "step": 65490,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.437502
    },
    {
      "epoch": 2.806006597832141,
      "grad_norm": 4.563215255737305,
      "learning_rate": 4.044564104462747e-05,
      "loss": 2.5749319076538084,
      "memory(GiB)": 77.56,
      "step": 65495,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437526
    },
    {
      "epoch": 2.80622081316139,
      "grad_norm": 5.335965156555176,
      "learning_rate": 4.0439035367206875e-05,
      "loss": 2.300836372375488,
      "memory(GiB)": 77.56,
      "step": 65500,
      "token_acc": 0.50390625,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 2.80622081316139,
      "eval_loss": 2.1007239818573,
      "eval_runtime": 14.5777,
      "eval_samples_per_second": 6.86,
      "eval_steps_per_second": 6.86,
      "eval_token_acc": 0.4951321279554937,
      "step": 65500
    },
    {
      "epoch": 2.8064350284906388,
      "grad_norm": 6.0830817222595215,
      "learning_rate": 4.04324298629921e-05,
      "loss": 2.215840530395508,
      "memory(GiB)": 77.56,
      "step": 65505,
      "token_acc": 0.49278152069297404,
      "train_speed(iter/s)": 1.437018
    },
    {
      "epoch": 2.806649243819888,
      "grad_norm": 4.430671691894531,
      "learning_rate": 4.042582453210285e-05,
      "loss": 2.4220972061157227,
      "memory(GiB)": 77.56,
      "step": 65510,
      "token_acc": 0.5220125786163522,
      "train_speed(iter/s)": 1.437021
    },
    {
      "epoch": 2.806863459149137,
      "grad_norm": 7.196639060974121,
      "learning_rate": 4.0419219374658766e-05,
      "loss": 2.4943925857543947,
      "memory(GiB)": 77.56,
      "step": 65515,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.437042
    },
    {
      "epoch": 2.8070776744783856,
      "grad_norm": 6.069049835205078,
      "learning_rate": 4.04126143907795e-05,
      "loss": 2.7378908157348634,
      "memory(GiB)": 77.56,
      "step": 65520,
      "token_acc": 0.44525547445255476,
      "train_speed(iter/s)": 1.437009
    },
    {
      "epoch": 2.807291889807635,
      "grad_norm": 6.978514671325684,
      "learning_rate": 4.040600958058471e-05,
      "loss": 2.1614658355712892,
      "memory(GiB)": 77.56,
      "step": 65525,
      "token_acc": 0.5316901408450704,
      "train_speed(iter/s)": 1.437019
    },
    {
      "epoch": 2.8075061051368837,
      "grad_norm": 7.327071666717529,
      "learning_rate": 4.039940494419407e-05,
      "loss": 2.451673889160156,
      "memory(GiB)": 77.56,
      "step": 65530,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.437018
    },
    {
      "epoch": 2.8077203204661325,
      "grad_norm": 7.619434833526611,
      "learning_rate": 4.0392800481727224e-05,
      "loss": 2.153657913208008,
      "memory(GiB)": 77.56,
      "step": 65535,
      "token_acc": 0.51171875,
      "train_speed(iter/s)": 1.437017
    },
    {
      "epoch": 2.8079345357953818,
      "grad_norm": 6.128969192504883,
      "learning_rate": 4.03861961933038e-05,
      "loss": 2.597476577758789,
      "memory(GiB)": 77.56,
      "step": 65540,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.437035
    },
    {
      "epoch": 2.8081487511246306,
      "grad_norm": 5.445517539978027,
      "learning_rate": 4.037959207904346e-05,
      "loss": 2.3313583374023437,
      "memory(GiB)": 77.56,
      "step": 65545,
      "token_acc": 0.4727272727272727,
      "train_speed(iter/s)": 1.437057
    },
    {
      "epoch": 2.8083629664538794,
      "grad_norm": 6.074429988861084,
      "learning_rate": 4.0372988139065824e-05,
      "loss": 2.5054134368896483,
      "memory(GiB)": 77.56,
      "step": 65550,
      "token_acc": 0.4923547400611621,
      "train_speed(iter/s)": 1.437056
    },
    {
      "epoch": 2.8085771817831287,
      "grad_norm": 4.698348522186279,
      "learning_rate": 4.036638437349054e-05,
      "loss": 2.0462865829467773,
      "memory(GiB)": 77.56,
      "step": 65555,
      "token_acc": 0.5846153846153846,
      "train_speed(iter/s)": 1.437055
    },
    {
      "epoch": 2.8087913971123775,
      "grad_norm": 5.37380838394165,
      "learning_rate": 4.035978078243725e-05,
      "loss": 2.6219940185546875,
      "memory(GiB)": 77.56,
      "step": 65560,
      "token_acc": 0.45535714285714285,
      "train_speed(iter/s)": 1.437077
    },
    {
      "epoch": 2.8090056124416263,
      "grad_norm": 5.415415287017822,
      "learning_rate": 4.0353177366025565e-05,
      "loss": 2.3958534240722655,
      "memory(GiB)": 77.56,
      "step": 65565,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437099
    },
    {
      "epoch": 2.8092198277708755,
      "grad_norm": 4.243507385253906,
      "learning_rate": 4.0346574124375126e-05,
      "loss": 2.2483856201171877,
      "memory(GiB)": 77.56,
      "step": 65570,
      "token_acc": 0.5480427046263345,
      "train_speed(iter/s)": 1.437122
    },
    {
      "epoch": 2.8094340431001243,
      "grad_norm": 5.420349597930908,
      "learning_rate": 4.033997105760555e-05,
      "loss": 2.6823713302612306,
      "memory(GiB)": 77.56,
      "step": 65575,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.437118
    },
    {
      "epoch": 2.809648258429373,
      "grad_norm": 6.035757064819336,
      "learning_rate": 4.0333368165836456e-05,
      "loss": 2.3820026397705076,
      "memory(GiB)": 77.56,
      "step": 65580,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.437103
    },
    {
      "epoch": 2.8098624737586224,
      "grad_norm": 6.843135356903076,
      "learning_rate": 4.032676544918747e-05,
      "loss": 2.3901126861572264,
      "memory(GiB)": 77.56,
      "step": 65585,
      "token_acc": 0.4602076124567474,
      "train_speed(iter/s)": 1.43714
    },
    {
      "epoch": 2.8100766890878712,
      "grad_norm": 4.060319900512695,
      "learning_rate": 4.0320162907778196e-05,
      "loss": 2.5534442901611327,
      "memory(GiB)": 77.56,
      "step": 65590,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.437172
    },
    {
      "epoch": 2.81029090441712,
      "grad_norm": 4.8427252769470215,
      "learning_rate": 4.031356054172826e-05,
      "loss": 2.5128955841064453,
      "memory(GiB)": 77.56,
      "step": 65595,
      "token_acc": 0.48026315789473684,
      "train_speed(iter/s)": 1.437204
    },
    {
      "epoch": 2.8105051197463693,
      "grad_norm": 5.334048748016357,
      "learning_rate": 4.0306958351157245e-05,
      "loss": 2.6606060028076173,
      "memory(GiB)": 77.56,
      "step": 65600,
      "token_acc": 0.4524590163934426,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 2.810719335075618,
      "grad_norm": 5.572875499725342,
      "learning_rate": 4.0300356336184786e-05,
      "loss": 2.3118499755859374,
      "memory(GiB)": 77.56,
      "step": 65605,
      "token_acc": 0.49865951742627346,
      "train_speed(iter/s)": 1.437253
    },
    {
      "epoch": 2.810933550404867,
      "grad_norm": 5.52056884765625,
      "learning_rate": 4.029375449693047e-05,
      "loss": 2.3903099060058595,
      "memory(GiB)": 77.56,
      "step": 65610,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.437256
    },
    {
      "epoch": 2.811147765734116,
      "grad_norm": 5.154854774475098,
      "learning_rate": 4.02871528335139e-05,
      "loss": 2.4831390380859375,
      "memory(GiB)": 77.56,
      "step": 65615,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.437279
    },
    {
      "epoch": 2.811361981063365,
      "grad_norm": 5.662505626678467,
      "learning_rate": 4.028055134605467e-05,
      "loss": 2.4398040771484375,
      "memory(GiB)": 77.56,
      "step": 65620,
      "token_acc": 0.5057034220532319,
      "train_speed(iter/s)": 1.437247
    },
    {
      "epoch": 2.811576196392614,
      "grad_norm": 6.19303035736084,
      "learning_rate": 4.0273950034672356e-05,
      "loss": 2.620880126953125,
      "memory(GiB)": 77.56,
      "step": 65625,
      "token_acc": 0.44660194174757284,
      "train_speed(iter/s)": 1.437253
    },
    {
      "epoch": 2.811790411721863,
      "grad_norm": 5.007617473602295,
      "learning_rate": 4.026734889948657e-05,
      "loss": 2.4864192962646485,
      "memory(GiB)": 77.56,
      "step": 65630,
      "token_acc": 0.4872611464968153,
      "train_speed(iter/s)": 1.437254
    },
    {
      "epoch": 2.812004627051112,
      "grad_norm": 5.41358757019043,
      "learning_rate": 4.02607479406169e-05,
      "loss": 2.369548034667969,
      "memory(GiB)": 77.56,
      "step": 65635,
      "token_acc": 0.5283582089552239,
      "train_speed(iter/s)": 1.437231
    },
    {
      "epoch": 2.8122188423803607,
      "grad_norm": 6.951308250427246,
      "learning_rate": 4.0254147158182895e-05,
      "loss": 2.2678949356079103,
      "memory(GiB)": 77.56,
      "step": 65640,
      "token_acc": 0.48444444444444446,
      "train_speed(iter/s)": 1.437239
    },
    {
      "epoch": 2.81243305770961,
      "grad_norm": 5.569390773773193,
      "learning_rate": 4.024754655230417e-05,
      "loss": 2.464281463623047,
      "memory(GiB)": 77.56,
      "step": 65645,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.437243
    },
    {
      "epoch": 2.8126472730388588,
      "grad_norm": 5.123871326446533,
      "learning_rate": 4.024094612310028e-05,
      "loss": 2.1402820587158202,
      "memory(GiB)": 77.56,
      "step": 65650,
      "token_acc": 0.5315985130111525,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 2.8128614883681076,
      "grad_norm": 5.817910671234131,
      "learning_rate": 4.023434587069081e-05,
      "loss": 2.4309709548950194,
      "memory(GiB)": 77.56,
      "step": 65655,
      "token_acc": 0.4584615384615385,
      "train_speed(iter/s)": 1.437261
    },
    {
      "epoch": 2.813075703697357,
      "grad_norm": 8.321303367614746,
      "learning_rate": 4.0227745795195335e-05,
      "loss": 2.6245563507080076,
      "memory(GiB)": 77.56,
      "step": 65660,
      "token_acc": 0.4584527220630373,
      "train_speed(iter/s)": 1.437287
    },
    {
      "epoch": 2.8132899190266056,
      "grad_norm": 8.671968460083008,
      "learning_rate": 4.02211458967334e-05,
      "loss": 2.7484920501708983,
      "memory(GiB)": 77.56,
      "step": 65665,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.437301
    },
    {
      "epoch": 2.8135041343558544,
      "grad_norm": 13.474469184875488,
      "learning_rate": 4.021454617542457e-05,
      "loss": 2.464711570739746,
      "memory(GiB)": 77.56,
      "step": 65670,
      "token_acc": 0.5474452554744526,
      "train_speed(iter/s)": 1.437334
    },
    {
      "epoch": 2.8137183496851037,
      "grad_norm": 4.670197010040283,
      "learning_rate": 4.0207946631388426e-05,
      "loss": 2.3289325714111326,
      "memory(GiB)": 77.56,
      "step": 65675,
      "token_acc": 0.5070422535211268,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 2.8139325650143525,
      "grad_norm": 5.20142936706543,
      "learning_rate": 4.0201347264744524e-05,
      "loss": 2.1882888793945314,
      "memory(GiB)": 77.56,
      "step": 65680,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.437352
    },
    {
      "epoch": 2.8141467803436013,
      "grad_norm": 8.296268463134766,
      "learning_rate": 4.0194748075612396e-05,
      "loss": 2.2669937133789064,
      "memory(GiB)": 77.56,
      "step": 65685,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.437385
    },
    {
      "epoch": 2.8143609956728506,
      "grad_norm": 4.953884124755859,
      "learning_rate": 4.0188149064111615e-05,
      "loss": 2.7729183197021485,
      "memory(GiB)": 77.56,
      "step": 65690,
      "token_acc": 0.46616541353383456,
      "train_speed(iter/s)": 1.437414
    },
    {
      "epoch": 2.8145752110020994,
      "grad_norm": 6.948336124420166,
      "learning_rate": 4.018155023036171e-05,
      "loss": 2.230704116821289,
      "memory(GiB)": 77.56,
      "step": 65695,
      "token_acc": 0.5407725321888412,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 2.814789426331348,
      "grad_norm": 6.491889476776123,
      "learning_rate": 4.017495157448224e-05,
      "loss": 2.6713748931884767,
      "memory(GiB)": 77.56,
      "step": 65700,
      "token_acc": 0.4541832669322709,
      "train_speed(iter/s)": 1.437444
    },
    {
      "epoch": 2.8150036416605975,
      "grad_norm": 5.530081748962402,
      "learning_rate": 4.0168353096592735e-05,
      "loss": 2.574834442138672,
      "memory(GiB)": 77.56,
      "step": 65705,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.437461
    },
    {
      "epoch": 2.8152178569898463,
      "grad_norm": 4.453324317932129,
      "learning_rate": 4.0161754796812736e-05,
      "loss": 2.532196807861328,
      "memory(GiB)": 77.56,
      "step": 65710,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.437453
    },
    {
      "epoch": 2.815432072319095,
      "grad_norm": 11.306683540344238,
      "learning_rate": 4.0155156675261785e-05,
      "loss": 2.567127990722656,
      "memory(GiB)": 77.56,
      "step": 65715,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.437452
    },
    {
      "epoch": 2.8156462876483443,
      "grad_norm": 4.425283908843994,
      "learning_rate": 4.01485587320594e-05,
      "loss": 2.338490104675293,
      "memory(GiB)": 77.56,
      "step": 65720,
      "token_acc": 0.5083612040133779,
      "train_speed(iter/s)": 1.437451
    },
    {
      "epoch": 2.815860502977593,
      "grad_norm": 4.620877265930176,
      "learning_rate": 4.014196096732511e-05,
      "loss": 2.109731674194336,
      "memory(GiB)": 77.56,
      "step": 65725,
      "token_acc": 0.5291828793774319,
      "train_speed(iter/s)": 1.437472
    },
    {
      "epoch": 2.816074718306842,
      "grad_norm": 4.456051349639893,
      "learning_rate": 4.0135363381178454e-05,
      "loss": 2.3745370864868165,
      "memory(GiB)": 77.56,
      "step": 65730,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.437487
    },
    {
      "epoch": 2.8162889336360912,
      "grad_norm": 6.78685188293457,
      "learning_rate": 4.012876597373893e-05,
      "loss": 2.2988656997680663,
      "memory(GiB)": 77.56,
      "step": 65735,
      "token_acc": 0.4980694980694981,
      "train_speed(iter/s)": 1.437486
    },
    {
      "epoch": 2.81650314896534,
      "grad_norm": 4.214529037475586,
      "learning_rate": 4.012216874512609e-05,
      "loss": 2.0138893127441406,
      "memory(GiB)": 77.56,
      "step": 65740,
      "token_acc": 0.5268456375838926,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 2.816717364294589,
      "grad_norm": 6.482390403747559,
      "learning_rate": 4.0115571695459396e-05,
      "loss": 2.7807937622070313,
      "memory(GiB)": 77.56,
      "step": 65745,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 2.816931579623838,
      "grad_norm": 5.318929195404053,
      "learning_rate": 4.0108974824858425e-05,
      "loss": 2.42476806640625,
      "memory(GiB)": 77.56,
      "step": 65750,
      "token_acc": 0.4689922480620155,
      "train_speed(iter/s)": 1.437519
    },
    {
      "epoch": 2.817145794953087,
      "grad_norm": 5.66624641418457,
      "learning_rate": 4.010237813344264e-05,
      "loss": 2.466521453857422,
      "memory(GiB)": 77.56,
      "step": 65755,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.43754
    },
    {
      "epoch": 2.8173600102823357,
      "grad_norm": 6.177026271820068,
      "learning_rate": 4.0095781621331563e-05,
      "loss": 2.2636268615722654,
      "memory(GiB)": 77.56,
      "step": 65760,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.437565
    },
    {
      "epoch": 2.817574225611585,
      "grad_norm": 5.124622344970703,
      "learning_rate": 4.0089185288644706e-05,
      "loss": 1.9913835525512695,
      "memory(GiB)": 77.56,
      "step": 65765,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 2.817788440940834,
      "grad_norm": 5.099691390991211,
      "learning_rate": 4.008258913550153e-05,
      "loss": 2.2811561584472657,
      "memory(GiB)": 77.56,
      "step": 65770,
      "token_acc": 0.486013986013986,
      "train_speed(iter/s)": 1.437545
    },
    {
      "epoch": 2.8180026562700826,
      "grad_norm": 4.510668754577637,
      "learning_rate": 4.0075993162021575e-05,
      "loss": 2.6661176681518555,
      "memory(GiB)": 77.56,
      "step": 65775,
      "token_acc": 0.4437869822485207,
      "train_speed(iter/s)": 1.437569
    },
    {
      "epoch": 2.818216871599332,
      "grad_norm": 7.073869228363037,
      "learning_rate": 4.006939736832431e-05,
      "loss": 2.379596710205078,
      "memory(GiB)": 77.56,
      "step": 65780,
      "token_acc": 0.5016835016835017,
      "train_speed(iter/s)": 1.437564
    },
    {
      "epoch": 2.8184310869285807,
      "grad_norm": 5.8068108558654785,
      "learning_rate": 4.006280175452922e-05,
      "loss": 2.5193090438842773,
      "memory(GiB)": 77.56,
      "step": 65785,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 2.8186453022578295,
      "grad_norm": 7.517391204833984,
      "learning_rate": 4.0056206320755806e-05,
      "loss": 2.311432647705078,
      "memory(GiB)": 77.56,
      "step": 65790,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 2.8188595175870788,
      "grad_norm": 6.91351842880249,
      "learning_rate": 4.0049611067123526e-05,
      "loss": 2.487359619140625,
      "memory(GiB)": 77.56,
      "step": 65795,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.437575
    },
    {
      "epoch": 2.8190737329163276,
      "grad_norm": 4.8192362785339355,
      "learning_rate": 4.00430159937519e-05,
      "loss": 2.3111480712890624,
      "memory(GiB)": 77.56,
      "step": 65800,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.437594
    },
    {
      "epoch": 2.8192879482455764,
      "grad_norm": 6.4928789138793945,
      "learning_rate": 4.003642110076037e-05,
      "loss": 2.295087432861328,
      "memory(GiB)": 77.56,
      "step": 65805,
      "token_acc": 0.48132780082987553,
      "train_speed(iter/s)": 1.437623
    },
    {
      "epoch": 2.8195021635748256,
      "grad_norm": 6.54436731338501,
      "learning_rate": 4.002982638826841e-05,
      "loss": 2.430122947692871,
      "memory(GiB)": 77.56,
      "step": 65810,
      "token_acc": 0.48742138364779874,
      "train_speed(iter/s)": 1.43761
    },
    {
      "epoch": 2.8197163789040744,
      "grad_norm": 6.262338161468506,
      "learning_rate": 4.0023231856395505e-05,
      "loss": 2.304804229736328,
      "memory(GiB)": 77.56,
      "step": 65815,
      "token_acc": 0.5127118644067796,
      "train_speed(iter/s)": 1.43762
    },
    {
      "epoch": 2.8199305942333233,
      "grad_norm": 5.558399677276611,
      "learning_rate": 4.00166375052611e-05,
      "loss": 2.3864105224609373,
      "memory(GiB)": 77.56,
      "step": 65820,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 2.8201448095625725,
      "grad_norm": 5.8830037117004395,
      "learning_rate": 4.001004333498469e-05,
      "loss": 2.6569293975830077,
      "memory(GiB)": 77.56,
      "step": 65825,
      "token_acc": 0.44696969696969696,
      "train_speed(iter/s)": 1.437651
    },
    {
      "epoch": 2.8203590248918213,
      "grad_norm": 4.427763938903809,
      "learning_rate": 4.0003449345685704e-05,
      "loss": 2.5883785247802735,
      "memory(GiB)": 77.56,
      "step": 65830,
      "token_acc": 0.4515235457063712,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 2.82057324022107,
      "grad_norm": 4.4692277908325195,
      "learning_rate": 3.999685553748362e-05,
      "loss": 2.401324653625488,
      "memory(GiB)": 77.56,
      "step": 65835,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.437675
    },
    {
      "epoch": 2.8207874555503194,
      "grad_norm": 5.024598121643066,
      "learning_rate": 3.9990261910497876e-05,
      "loss": 2.576115036010742,
      "memory(GiB)": 77.56,
      "step": 65840,
      "token_acc": 0.48299319727891155,
      "train_speed(iter/s)": 1.43769
    },
    {
      "epoch": 2.821001670879568,
      "grad_norm": 5.368990421295166,
      "learning_rate": 3.9983668464847935e-05,
      "loss": 2.5567007064819336,
      "memory(GiB)": 77.56,
      "step": 65845,
      "token_acc": 0.5177865612648221,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 2.821215886208817,
      "grad_norm": 5.691743850708008,
      "learning_rate": 3.9977075200653234e-05,
      "loss": 2.290756607055664,
      "memory(GiB)": 77.56,
      "step": 65850,
      "token_acc": 0.45484949832775917,
      "train_speed(iter/s)": 1.437733
    },
    {
      "epoch": 2.8214301015380663,
      "grad_norm": 5.7045698165893555,
      "learning_rate": 3.997048211803321e-05,
      "loss": 2.4875391006469725,
      "memory(GiB)": 77.56,
      "step": 65855,
      "token_acc": 0.49846153846153846,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 2.821644316867315,
      "grad_norm": 4.722874641418457,
      "learning_rate": 3.996388921710732e-05,
      "loss": 2.12872200012207,
      "memory(GiB)": 77.56,
      "step": 65860,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 2.821858532196564,
      "grad_norm": 5.208227157592773,
      "learning_rate": 3.995729649799499e-05,
      "loss": 2.4433000564575194,
      "memory(GiB)": 77.56,
      "step": 65865,
      "token_acc": 0.4573170731707317,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 2.822072747525813,
      "grad_norm": 5.113282203674316,
      "learning_rate": 3.995070396081565e-05,
      "loss": 2.0730361938476562,
      "memory(GiB)": 77.56,
      "step": 65870,
      "token_acc": 0.55078125,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 2.822286962855062,
      "grad_norm": 5.095364570617676,
      "learning_rate": 3.994411160568874e-05,
      "loss": 2.5483612060546874,
      "memory(GiB)": 77.56,
      "step": 65875,
      "token_acc": 0.4668769716088328,
      "train_speed(iter/s)": 1.437755
    },
    {
      "epoch": 2.822501178184311,
      "grad_norm": 5.002376079559326,
      "learning_rate": 3.993751943273367e-05,
      "loss": 2.248806953430176,
      "memory(GiB)": 77.56,
      "step": 65880,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 2.82271539351356,
      "grad_norm": 4.768962860107422,
      "learning_rate": 3.9930927442069885e-05,
      "loss": 2.433235740661621,
      "memory(GiB)": 77.56,
      "step": 65885,
      "token_acc": 0.4968553459119497,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 2.822929608842809,
      "grad_norm": 5.688992977142334,
      "learning_rate": 3.99243356338168e-05,
      "loss": 2.554449462890625,
      "memory(GiB)": 77.56,
      "step": 65890,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 2.8231438241720577,
      "grad_norm": 7.391767501831055,
      "learning_rate": 3.9917744008093806e-05,
      "loss": 2.4067935943603516,
      "memory(GiB)": 77.56,
      "step": 65895,
      "token_acc": 0.47674418604651164,
      "train_speed(iter/s)": 1.437746
    },
    {
      "epoch": 2.823358039501307,
      "grad_norm": 6.47195291519165,
      "learning_rate": 3.991115256502034e-05,
      "loss": 2.3028024673461913,
      "memory(GiB)": 77.56,
      "step": 65900,
      "token_acc": 0.5328467153284672,
      "train_speed(iter/s)": 1.437748
    },
    {
      "epoch": 2.8235722548305557,
      "grad_norm": 5.738492012023926,
      "learning_rate": 3.9904561304715824e-05,
      "loss": 2.4111894607543944,
      "memory(GiB)": 77.56,
      "step": 65905,
      "token_acc": 0.4689655172413793,
      "train_speed(iter/s)": 1.437742
    },
    {
      "epoch": 2.8237864701598046,
      "grad_norm": 5.683606147766113,
      "learning_rate": 3.989797022729966e-05,
      "loss": 2.4715497970581053,
      "memory(GiB)": 77.56,
      "step": 65910,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.437763
    },
    {
      "epoch": 2.824000685489054,
      "grad_norm": 5.9431304931640625,
      "learning_rate": 3.9891379332891224e-05,
      "loss": 2.3763616561889647,
      "memory(GiB)": 77.56,
      "step": 65915,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.437789
    },
    {
      "epoch": 2.8242149008183026,
      "grad_norm": 5.847568035125732,
      "learning_rate": 3.9884788621609936e-05,
      "loss": 2.5116508483886717,
      "memory(GiB)": 77.56,
      "step": 65920,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 2.8244291161475514,
      "grad_norm": 5.4715256690979,
      "learning_rate": 3.98781980935752e-05,
      "loss": 2.3955821990966797,
      "memory(GiB)": 77.56,
      "step": 65925,
      "token_acc": 0.5209003215434084,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 2.8246433314768007,
      "grad_norm": 6.0657501220703125,
      "learning_rate": 3.9871607748906395e-05,
      "loss": 2.707859992980957,
      "memory(GiB)": 77.56,
      "step": 65930,
      "token_acc": 0.42896174863387976,
      "train_speed(iter/s)": 1.437808
    },
    {
      "epoch": 2.8248575468060495,
      "grad_norm": 4.782413959503174,
      "learning_rate": 3.9865017587722916e-05,
      "loss": 2.365995407104492,
      "memory(GiB)": 77.56,
      "step": 65935,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 2.8250717621352983,
      "grad_norm": 5.763640880584717,
      "learning_rate": 3.985842761014414e-05,
      "loss": 2.3232954025268553,
      "memory(GiB)": 77.56,
      "step": 65940,
      "token_acc": 0.5292096219931272,
      "train_speed(iter/s)": 1.437779
    },
    {
      "epoch": 2.8252859774645476,
      "grad_norm": 5.335360050201416,
      "learning_rate": 3.9851837816289485e-05,
      "loss": 2.410564422607422,
      "memory(GiB)": 77.56,
      "step": 65945,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.437788
    },
    {
      "epoch": 2.8255001927937964,
      "grad_norm": 5.844241142272949,
      "learning_rate": 3.984524820627829e-05,
      "loss": 2.3132436752319334,
      "memory(GiB)": 77.56,
      "step": 65950,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437799
    },
    {
      "epoch": 2.825714408123045,
      "grad_norm": 6.3975510597229,
      "learning_rate": 3.983865878022995e-05,
      "loss": 2.2847755432128904,
      "memory(GiB)": 77.56,
      "step": 65955,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 2.8259286234522945,
      "grad_norm": 4.587973117828369,
      "learning_rate": 3.983206953826385e-05,
      "loss": 2.316034507751465,
      "memory(GiB)": 77.56,
      "step": 65960,
      "token_acc": 0.5153374233128835,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 2.8261428387815433,
      "grad_norm": 5.519717693328857,
      "learning_rate": 3.982548048049935e-05,
      "loss": 2.412979316711426,
      "memory(GiB)": 77.56,
      "step": 65965,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.437847
    },
    {
      "epoch": 2.826357054110792,
      "grad_norm": 6.584933757781982,
      "learning_rate": 3.981889160705579e-05,
      "loss": 2.2350576400756834,
      "memory(GiB)": 77.56,
      "step": 65970,
      "token_acc": 0.5315985130111525,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 2.8265712694400413,
      "grad_norm": 9.50361156463623,
      "learning_rate": 3.981230291805257e-05,
      "loss": 2.6061330795288087,
      "memory(GiB)": 77.56,
      "step": 65975,
      "token_acc": 0.4619883040935672,
      "train_speed(iter/s)": 1.437867
    },
    {
      "epoch": 2.82678548476929,
      "grad_norm": 6.705776691436768,
      "learning_rate": 3.980571441360904e-05,
      "loss": 2.3313167572021483,
      "memory(GiB)": 77.56,
      "step": 65980,
      "token_acc": 0.5099601593625498,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 2.826999700098539,
      "grad_norm": 6.451496601104736,
      "learning_rate": 3.979912609384456e-05,
      "loss": 2.455087089538574,
      "memory(GiB)": 77.56,
      "step": 65985,
      "token_acc": 0.4700854700854701,
      "train_speed(iter/s)": 1.437858
    },
    {
      "epoch": 2.827213915427788,
      "grad_norm": 4.885879039764404,
      "learning_rate": 3.979253795887849e-05,
      "loss": 2.2485420227050783,
      "memory(GiB)": 77.56,
      "step": 65990,
      "token_acc": 0.5292207792207793,
      "train_speed(iter/s)": 1.437872
    },
    {
      "epoch": 2.827428130757037,
      "grad_norm": 5.241494178771973,
      "learning_rate": 3.978595000883017e-05,
      "loss": 2.3914384841918945,
      "memory(GiB)": 77.56,
      "step": 65995,
      "token_acc": 0.5126582278481012,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 2.827642346086286,
      "grad_norm": 7.1190032958984375,
      "learning_rate": 3.977936224381893e-05,
      "loss": 2.4144733428955076,
      "memory(GiB)": 77.56,
      "step": 66000,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.437883
    },
    {
      "epoch": 2.827642346086286,
      "eval_loss": 2.2393407821655273,
      "eval_runtime": 13.5598,
      "eval_samples_per_second": 7.375,
      "eval_steps_per_second": 7.375,
      "eval_token_acc": 0.4613259668508287,
      "step": 66000
    },
    {
      "epoch": 2.827856561415535,
      "grad_norm": 5.540735244750977,
      "learning_rate": 3.9772774663964145e-05,
      "loss": 2.4929662704467774,
      "memory(GiB)": 77.56,
      "step": 66005,
      "token_acc": 0.47952047952047955,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 2.828070776744784,
      "grad_norm": 4.505105495452881,
      "learning_rate": 3.9766187269385144e-05,
      "loss": 2.415993309020996,
      "memory(GiB)": 77.56,
      "step": 66010,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.437463
    },
    {
      "epoch": 2.8282849920740327,
      "grad_norm": 5.809741497039795,
      "learning_rate": 3.9759600060201245e-05,
      "loss": 2.7046775817871094,
      "memory(GiB)": 77.56,
      "step": 66015,
      "token_acc": 0.44696969696969696,
      "train_speed(iter/s)": 1.437485
    },
    {
      "epoch": 2.828499207403282,
      "grad_norm": 6.980930328369141,
      "learning_rate": 3.975301303653181e-05,
      "loss": 2.5878875732421873,
      "memory(GiB)": 77.56,
      "step": 66020,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.437498
    },
    {
      "epoch": 2.828713422732531,
      "grad_norm": 4.306469917297363,
      "learning_rate": 3.974642619849615e-05,
      "loss": 2.4797847747802733,
      "memory(GiB)": 77.56,
      "step": 66025,
      "token_acc": 0.48138297872340424,
      "train_speed(iter/s)": 1.437475
    },
    {
      "epoch": 2.8289276380617796,
      "grad_norm": 4.987720489501953,
      "learning_rate": 3.9739839546213596e-05,
      "loss": 2.5715408325195312,
      "memory(GiB)": 77.56,
      "step": 66030,
      "token_acc": 0.46774193548387094,
      "train_speed(iter/s)": 1.437507
    },
    {
      "epoch": 2.829141853391029,
      "grad_norm": 5.3655195236206055,
      "learning_rate": 3.9733253079803486e-05,
      "loss": 2.1925025939941407,
      "memory(GiB)": 77.56,
      "step": 66035,
      "token_acc": 0.5354609929078015,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 2.8293560687202777,
      "grad_norm": 4.701647758483887,
      "learning_rate": 3.9726666799385095e-05,
      "loss": 2.303322982788086,
      "memory(GiB)": 77.56,
      "step": 66040,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.437485
    },
    {
      "epoch": 2.8295702840495265,
      "grad_norm": 4.967202186584473,
      "learning_rate": 3.972008070507779e-05,
      "loss": 2.211300086975098,
      "memory(GiB)": 77.56,
      "step": 66045,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.437461
    },
    {
      "epoch": 2.8297844993787757,
      "grad_norm": 4.112099647521973,
      "learning_rate": 3.971349479700088e-05,
      "loss": 2.4259616851806642,
      "memory(GiB)": 77.56,
      "step": 66050,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.437451
    },
    {
      "epoch": 2.8299987147080246,
      "grad_norm": 4.565462589263916,
      "learning_rate": 3.970690907527366e-05,
      "loss": 2.73983211517334,
      "memory(GiB)": 77.56,
      "step": 66055,
      "token_acc": 0.45819397993311034,
      "train_speed(iter/s)": 1.437469
    },
    {
      "epoch": 2.8302129300372734,
      "grad_norm": 6.556137561798096,
      "learning_rate": 3.970032354001542e-05,
      "loss": 2.639791488647461,
      "memory(GiB)": 77.56,
      "step": 66060,
      "token_acc": 0.4316546762589928,
      "train_speed(iter/s)": 1.437449
    },
    {
      "epoch": 2.8304271453665226,
      "grad_norm": 5.55762243270874,
      "learning_rate": 3.9693738191345495e-05,
      "loss": 2.566722869873047,
      "memory(GiB)": 77.56,
      "step": 66065,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.437423
    },
    {
      "epoch": 2.8306413606957714,
      "grad_norm": 5.159157752990723,
      "learning_rate": 3.968715302938317e-05,
      "loss": 2.320499229431152,
      "memory(GiB)": 77.56,
      "step": 66070,
      "token_acc": 0.4743935309973046,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 2.8308555760250202,
      "grad_norm": 6.623139381408691,
      "learning_rate": 3.9680568054247744e-05,
      "loss": 2.432745361328125,
      "memory(GiB)": 77.56,
      "step": 66075,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.437466
    },
    {
      "epoch": 2.8310697913542695,
      "grad_norm": 3.9942383766174316,
      "learning_rate": 3.9673983266058504e-05,
      "loss": 2.4070995330810545,
      "memory(GiB)": 77.56,
      "step": 66080,
      "token_acc": 0.5015974440894568,
      "train_speed(iter/s)": 1.437464
    },
    {
      "epoch": 2.8312840066835183,
      "grad_norm": 5.720208644866943,
      "learning_rate": 3.9667398664934735e-05,
      "loss": 2.1331079483032225,
      "memory(GiB)": 77.56,
      "step": 66085,
      "token_acc": 0.5077881619937694,
      "train_speed(iter/s)": 1.437483
    },
    {
      "epoch": 2.831498222012767,
      "grad_norm": 3.524899959564209,
      "learning_rate": 3.966081425099575e-05,
      "loss": 1.8924365997314454,
      "memory(GiB)": 77.56,
      "step": 66090,
      "token_acc": 0.570281124497992,
      "train_speed(iter/s)": 1.437502
    },
    {
      "epoch": 2.8317124373420164,
      "grad_norm": 5.937251567840576,
      "learning_rate": 3.96542300243608e-05,
      "loss": 2.2823408126831053,
      "memory(GiB)": 77.56,
      "step": 66095,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.437511
    },
    {
      "epoch": 2.831926652671265,
      "grad_norm": 5.689079761505127,
      "learning_rate": 3.9647645985149184e-05,
      "loss": 2.394300842285156,
      "memory(GiB)": 77.56,
      "step": 66100,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.437524
    },
    {
      "epoch": 2.832140868000514,
      "grad_norm": 5.338691711425781,
      "learning_rate": 3.964106213348017e-05,
      "loss": 2.508581352233887,
      "memory(GiB)": 77.56,
      "step": 66105,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.437492
    },
    {
      "epoch": 2.8323550833297633,
      "grad_norm": 5.338334083557129,
      "learning_rate": 3.963447846947304e-05,
      "loss": 2.659164047241211,
      "memory(GiB)": 77.56,
      "step": 66110,
      "token_acc": 0.47547169811320755,
      "train_speed(iter/s)": 1.437514
    },
    {
      "epoch": 2.832569298659012,
      "grad_norm": 3.856503486633301,
      "learning_rate": 3.962789499324703e-05,
      "loss": 2.4176567077636717,
      "memory(GiB)": 77.56,
      "step": 66115,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.437517
    },
    {
      "epoch": 2.832783513988261,
      "grad_norm": 4.626069068908691,
      "learning_rate": 3.962131170492145e-05,
      "loss": 2.5557384490966797,
      "memory(GiB)": 77.56,
      "step": 66120,
      "token_acc": 0.4662576687116564,
      "train_speed(iter/s)": 1.437537
    },
    {
      "epoch": 2.83299772931751,
      "grad_norm": 6.309337615966797,
      "learning_rate": 3.961472860461555e-05,
      "loss": 2.7593154907226562,
      "memory(GiB)": 77.56,
      "step": 66125,
      "token_acc": 0.4263565891472868,
      "train_speed(iter/s)": 1.437559
    },
    {
      "epoch": 2.833211944646759,
      "grad_norm": 5.2441582679748535,
      "learning_rate": 3.9608145692448575e-05,
      "loss": 2.5551788330078127,
      "memory(GiB)": 77.56,
      "step": 66130,
      "token_acc": 0.5015974440894568,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 2.8334261599760078,
      "grad_norm": 5.7819318771362305,
      "learning_rate": 3.9601562968539796e-05,
      "loss": 2.4453847885131834,
      "memory(GiB)": 77.56,
      "step": 66135,
      "token_acc": 0.4969512195121951,
      "train_speed(iter/s)": 1.437498
    },
    {
      "epoch": 2.833640375305257,
      "grad_norm": 6.63046932220459,
      "learning_rate": 3.959498043300846e-05,
      "loss": 2.020421600341797,
      "memory(GiB)": 77.56,
      "step": 66140,
      "token_acc": 0.5795918367346938,
      "train_speed(iter/s)": 1.437483
    },
    {
      "epoch": 2.833854590634506,
      "grad_norm": 5.306735515594482,
      "learning_rate": 3.958839808597381e-05,
      "loss": 2.56314640045166,
      "memory(GiB)": 77.56,
      "step": 66145,
      "token_acc": 0.46788990825688076,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 2.8340688059637547,
      "grad_norm": 5.529629230499268,
      "learning_rate": 3.95818159275551e-05,
      "loss": 2.550008773803711,
      "memory(GiB)": 77.56,
      "step": 66150,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.43749
    },
    {
      "epoch": 2.834283021293004,
      "grad_norm": 4.78882360458374,
      "learning_rate": 3.957523395787156e-05,
      "loss": 2.6957208633422853,
      "memory(GiB)": 77.56,
      "step": 66155,
      "token_acc": 0.46440677966101696,
      "train_speed(iter/s)": 1.437475
    },
    {
      "epoch": 2.8344972366222527,
      "grad_norm": 6.711662769317627,
      "learning_rate": 3.956865217704244e-05,
      "loss": 2.79765625,
      "memory(GiB)": 77.56,
      "step": 66160,
      "token_acc": 0.44061302681992337,
      "train_speed(iter/s)": 1.437496
    },
    {
      "epoch": 2.8347114519515015,
      "grad_norm": 11.294705390930176,
      "learning_rate": 3.956207058518697e-05,
      "loss": 2.3666833877563476,
      "memory(GiB)": 77.56,
      "step": 66165,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 2.834925667280751,
      "grad_norm": 5.45749044418335,
      "learning_rate": 3.955548918242438e-05,
      "loss": 2.3976718902587892,
      "memory(GiB)": 77.56,
      "step": 66170,
      "token_acc": 0.4927007299270073,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 2.8351398826099996,
      "grad_norm": 5.315276622772217,
      "learning_rate": 3.954890796887391e-05,
      "loss": 2.3737110137939452,
      "memory(GiB)": 77.56,
      "step": 66175,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 2.8353540979392484,
      "grad_norm": 5.749845504760742,
      "learning_rate": 3.9542326944654775e-05,
      "loss": 2.370409393310547,
      "memory(GiB)": 77.56,
      "step": 66180,
      "token_acc": 0.4609053497942387,
      "train_speed(iter/s)": 1.437524
    },
    {
      "epoch": 2.8355683132684977,
      "grad_norm": 4.790119171142578,
      "learning_rate": 3.953574610988619e-05,
      "loss": 2.2135387420654298,
      "memory(GiB)": 77.56,
      "step": 66185,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 2.8357825285977465,
      "grad_norm": 5.460172176361084,
      "learning_rate": 3.952916546468737e-05,
      "loss": 2.429324722290039,
      "memory(GiB)": 77.56,
      "step": 66190,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 2.8359967439269953,
      "grad_norm": 3.906254529953003,
      "learning_rate": 3.9522585009177554e-05,
      "loss": 2.353445053100586,
      "memory(GiB)": 77.56,
      "step": 66195,
      "token_acc": 0.5070422535211268,
      "train_speed(iter/s)": 1.437542
    },
    {
      "epoch": 2.8362109592562446,
      "grad_norm": 5.391211032867432,
      "learning_rate": 3.951600474347594e-05,
      "loss": 2.3955163955688477,
      "memory(GiB)": 77.56,
      "step": 66200,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.437552
    },
    {
      "epoch": 2.8364251745854934,
      "grad_norm": 4.388646125793457,
      "learning_rate": 3.950942466770173e-05,
      "loss": 2.410428047180176,
      "memory(GiB)": 77.56,
      "step": 66205,
      "token_acc": 0.46835443037974683,
      "train_speed(iter/s)": 1.437552
    },
    {
      "epoch": 2.836639389914742,
      "grad_norm": 5.6095380783081055,
      "learning_rate": 3.950284478197414e-05,
      "loss": 2.4293533325195313,
      "memory(GiB)": 77.56,
      "step": 66210,
      "token_acc": 0.4740740740740741,
      "train_speed(iter/s)": 1.43756
    },
    {
      "epoch": 2.8368536052439914,
      "grad_norm": 5.41325569152832,
      "learning_rate": 3.9496265086412364e-05,
      "loss": 2.242751884460449,
      "memory(GiB)": 77.56,
      "step": 66215,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 2.8370678205732403,
      "grad_norm": 4.969966411590576,
      "learning_rate": 3.948968558113559e-05,
      "loss": 2.41613826751709,
      "memory(GiB)": 77.56,
      "step": 66220,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.437587
    },
    {
      "epoch": 2.837282035902489,
      "grad_norm": 12.915497779846191,
      "learning_rate": 3.9483106266263036e-05,
      "loss": 2.6022666931152343,
      "memory(GiB)": 77.56,
      "step": 66225,
      "token_acc": 0.4584450402144772,
      "train_speed(iter/s)": 1.437575
    },
    {
      "epoch": 2.8374962512317383,
      "grad_norm": 5.276846885681152,
      "learning_rate": 3.9476527141913866e-05,
      "loss": 2.3101612091064454,
      "memory(GiB)": 77.56,
      "step": 66230,
      "token_acc": 0.5136186770428015,
      "train_speed(iter/s)": 1.437552
    },
    {
      "epoch": 2.837710466560987,
      "grad_norm": 4.95725154876709,
      "learning_rate": 3.946994820820728e-05,
      "loss": 2.733891487121582,
      "memory(GiB)": 77.56,
      "step": 66235,
      "token_acc": 0.468503937007874,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 2.837924681890236,
      "grad_norm": 4.425442695617676,
      "learning_rate": 3.9463369465262466e-05,
      "loss": 2.391596221923828,
      "memory(GiB)": 77.56,
      "step": 66240,
      "token_acc": 0.509375,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 2.838138897219485,
      "grad_norm": 5.859916687011719,
      "learning_rate": 3.945679091319859e-05,
      "loss": 2.340145492553711,
      "memory(GiB)": 77.56,
      "step": 66245,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 2.838353112548734,
      "grad_norm": 5.283370018005371,
      "learning_rate": 3.9450212552134845e-05,
      "loss": 2.3300167083740235,
      "memory(GiB)": 77.56,
      "step": 66250,
      "token_acc": 0.4981684981684982,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 2.838567327877983,
      "grad_norm": 5.853363990783691,
      "learning_rate": 3.9443634382190396e-05,
      "loss": 2.3598812103271483,
      "memory(GiB)": 77.56,
      "step": 66255,
      "token_acc": 0.463768115942029,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 2.838781543207232,
      "grad_norm": 6.4891037940979,
      "learning_rate": 3.9437056403484404e-05,
      "loss": 2.3071685791015626,
      "memory(GiB)": 77.56,
      "step": 66260,
      "token_acc": 0.5064516129032258,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 2.838995758536481,
      "grad_norm": 4.6603007316589355,
      "learning_rate": 3.9430478616136036e-05,
      "loss": 2.272079277038574,
      "memory(GiB)": 77.56,
      "step": 66265,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.437591
    },
    {
      "epoch": 2.8392099738657297,
      "grad_norm": 7.033201694488525,
      "learning_rate": 3.9423901020264474e-05,
      "loss": 2.3999202728271483,
      "memory(GiB)": 77.56,
      "step": 66270,
      "token_acc": 0.4628975265017668,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 2.839424189194979,
      "grad_norm": 6.5980048179626465,
      "learning_rate": 3.9417323615988864e-05,
      "loss": 2.3603357315063476,
      "memory(GiB)": 77.56,
      "step": 66275,
      "token_acc": 0.4844290657439446,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 2.8396384045242278,
      "grad_norm": 5.733865737915039,
      "learning_rate": 3.941074640342838e-05,
      "loss": 2.4446090698242187,
      "memory(GiB)": 77.56,
      "step": 66280,
      "token_acc": 0.48417721518987344,
      "train_speed(iter/s)": 1.437585
    },
    {
      "epoch": 2.8398526198534766,
      "grad_norm": 5.516995429992676,
      "learning_rate": 3.940416938270215e-05,
      "loss": 2.331661605834961,
      "memory(GiB)": 77.56,
      "step": 66285,
      "token_acc": 0.46987951807228917,
      "train_speed(iter/s)": 1.437614
    },
    {
      "epoch": 2.840066835182726,
      "grad_norm": 6.06817626953125,
      "learning_rate": 3.939759255392932e-05,
      "loss": 2.46845703125,
      "memory(GiB)": 77.56,
      "step": 66290,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.437623
    },
    {
      "epoch": 2.8402810505119747,
      "grad_norm": 6.026614665985107,
      "learning_rate": 3.939101591722906e-05,
      "loss": 2.1548831939697264,
      "memory(GiB)": 77.56,
      "step": 66295,
      "token_acc": 0.5020242914979757,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 2.8404952658412235,
      "grad_norm": 5.751220703125,
      "learning_rate": 3.93844394727205e-05,
      "loss": 2.576325225830078,
      "memory(GiB)": 77.56,
      "step": 66300,
      "token_acc": 0.4597014925373134,
      "train_speed(iter/s)": 1.437662
    },
    {
      "epoch": 2.8407094811704727,
      "grad_norm": 6.084564685821533,
      "learning_rate": 3.937786322052276e-05,
      "loss": 2.3021015167236327,
      "memory(GiB)": 77.56,
      "step": 66305,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.43769
    },
    {
      "epoch": 2.8409236964997215,
      "grad_norm": 4.717773914337158,
      "learning_rate": 3.937128716075501e-05,
      "loss": 2.096889877319336,
      "memory(GiB)": 77.56,
      "step": 66310,
      "token_acc": 0.563573883161512,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.8411379118289704,
      "grad_norm": 5.652832984924316,
      "learning_rate": 3.936471129353635e-05,
      "loss": 2.3126296997070312,
      "memory(GiB)": 77.56,
      "step": 66315,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 2.8413521271582196,
      "grad_norm": 5.348511695861816,
      "learning_rate": 3.935813561898593e-05,
      "loss": 2.539601707458496,
      "memory(GiB)": 77.56,
      "step": 66320,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437701
    },
    {
      "epoch": 2.8415663424874684,
      "grad_norm": 4.41096305847168,
      "learning_rate": 3.935156013722287e-05,
      "loss": 2.1715204238891603,
      "memory(GiB)": 77.56,
      "step": 66325,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 2.8417805578167172,
      "grad_norm": 4.9178595542907715,
      "learning_rate": 3.934498484836627e-05,
      "loss": 2.5662769317626952,
      "memory(GiB)": 77.56,
      "step": 66330,
      "token_acc": 0.46387832699619774,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 2.8419947731459665,
      "grad_norm": 5.927366733551025,
      "learning_rate": 3.933840975253527e-05,
      "loss": 2.3385799407958983,
      "memory(GiB)": 77.56,
      "step": 66335,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.437738
    },
    {
      "epoch": 2.8422089884752153,
      "grad_norm": 5.280552864074707,
      "learning_rate": 3.933183484984898e-05,
      "loss": 2.250534439086914,
      "memory(GiB)": 77.56,
      "step": 66340,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.842423203804464,
      "grad_norm": 4.954792499542236,
      "learning_rate": 3.932526014042652e-05,
      "loss": 2.6814754486083983,
      "memory(GiB)": 77.56,
      "step": 66345,
      "token_acc": 0.4773413897280967,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.8426374191337134,
      "grad_norm": 4.5819902420043945,
      "learning_rate": 3.9318685624386975e-05,
      "loss": 2.4675159454345703,
      "memory(GiB)": 77.56,
      "step": 66350,
      "token_acc": 0.45348837209302323,
      "train_speed(iter/s)": 1.437766
    },
    {
      "epoch": 2.842851634462962,
      "grad_norm": 4.561880111694336,
      "learning_rate": 3.931211130184947e-05,
      "loss": 2.1820587158203124,
      "memory(GiB)": 77.56,
      "step": 66355,
      "token_acc": 0.5171339563862928,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 2.843065849792211,
      "grad_norm": 4.890132427215576,
      "learning_rate": 3.9305537172933085e-05,
      "loss": 2.551707077026367,
      "memory(GiB)": 77.56,
      "step": 66360,
      "token_acc": 0.4554140127388535,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 2.8432800651214603,
      "grad_norm": 7.599239349365234,
      "learning_rate": 3.9298963237756934e-05,
      "loss": 2.608646202087402,
      "memory(GiB)": 77.56,
      "step": 66365,
      "token_acc": 0.46613545816733065,
      "train_speed(iter/s)": 1.437814
    },
    {
      "epoch": 2.843494280450709,
      "grad_norm": 5.214544773101807,
      "learning_rate": 3.92923894964401e-05,
      "loss": 2.417283630371094,
      "memory(GiB)": 77.56,
      "step": 66370,
      "token_acc": 0.55,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 2.843708495779958,
      "grad_norm": 5.644783020019531,
      "learning_rate": 3.9285815949101675e-05,
      "loss": 2.54773063659668,
      "memory(GiB)": 77.56,
      "step": 66375,
      "token_acc": 0.47774480712166173,
      "train_speed(iter/s)": 1.437844
    },
    {
      "epoch": 2.843922711109207,
      "grad_norm": 4.35864782333374,
      "learning_rate": 3.9279242595860746e-05,
      "loss": 2.6800617218017577,
      "memory(GiB)": 77.56,
      "step": 66380,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 2.844136926438456,
      "grad_norm": 5.798492431640625,
      "learning_rate": 3.9272669436836395e-05,
      "loss": 2.5936920166015627,
      "memory(GiB)": 77.56,
      "step": 66385,
      "token_acc": 0.4790996784565916,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 2.8443511417677048,
      "grad_norm": 6.025155544281006,
      "learning_rate": 3.9266096472147694e-05,
      "loss": 2.4738348007202147,
      "memory(GiB)": 77.56,
      "step": 66390,
      "token_acc": 0.5219941348973607,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 2.844565357096954,
      "grad_norm": 4.408134937286377,
      "learning_rate": 3.925952370191373e-05,
      "loss": 2.3777000427246096,
      "memory(GiB)": 77.56,
      "step": 66395,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.437849
    },
    {
      "epoch": 2.844779572426203,
      "grad_norm": 6.43510103225708,
      "learning_rate": 3.9252951126253565e-05,
      "loss": 2.463430404663086,
      "memory(GiB)": 77.56,
      "step": 66400,
      "token_acc": 0.4907749077490775,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 2.8449937877554516,
      "grad_norm": 4.5750732421875,
      "learning_rate": 3.9246378745286276e-05,
      "loss": 2.2757320404052734,
      "memory(GiB)": 77.56,
      "step": 66405,
      "token_acc": 0.5211726384364821,
      "train_speed(iter/s)": 1.437883
    },
    {
      "epoch": 2.845208003084701,
      "grad_norm": 5.788753509521484,
      "learning_rate": 3.923980655913091e-05,
      "loss": 2.260826301574707,
      "memory(GiB)": 77.56,
      "step": 66410,
      "token_acc": 0.4948805460750853,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 2.8454222184139497,
      "grad_norm": 5.856846332550049,
      "learning_rate": 3.923323456790656e-05,
      "loss": 2.002977752685547,
      "memory(GiB)": 77.56,
      "step": 66415,
      "token_acc": 0.5163636363636364,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 2.8456364337431985,
      "grad_norm": 5.099117279052734,
      "learning_rate": 3.9226662771732244e-05,
      "loss": 2.43624324798584,
      "memory(GiB)": 77.56,
      "step": 66420,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 2.845850649072448,
      "grad_norm": 4.795126438140869,
      "learning_rate": 3.922009117072706e-05,
      "loss": 2.4289674758911133,
      "memory(GiB)": 77.56,
      "step": 66425,
      "token_acc": 0.4752475247524752,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 2.8460648644016966,
      "grad_norm": 5.6901631355285645,
      "learning_rate": 3.921351976501004e-05,
      "loss": 2.2836795806884767,
      "memory(GiB)": 77.56,
      "step": 66430,
      "token_acc": 0.5283018867924528,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 2.8462790797309454,
      "grad_norm": 4.898319244384766,
      "learning_rate": 3.920694855470021e-05,
      "loss": 2.6279300689697265,
      "memory(GiB)": 77.56,
      "step": 66435,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.437861
    },
    {
      "epoch": 2.8464932950601947,
      "grad_norm": 6.0314836502075195,
      "learning_rate": 3.920037753991665e-05,
      "loss": 2.5557987213134767,
      "memory(GiB)": 77.56,
      "step": 66440,
      "token_acc": 0.45787545787545786,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 2.8467075103894435,
      "grad_norm": 4.333067417144775,
      "learning_rate": 3.919380672077838e-05,
      "loss": 2.7301849365234374,
      "memory(GiB)": 77.56,
      "step": 66445,
      "token_acc": 0.4369230769230769,
      "train_speed(iter/s)": 1.437892
    },
    {
      "epoch": 2.8469217257186923,
      "grad_norm": 6.202364444732666,
      "learning_rate": 3.918723609740442e-05,
      "loss": 2.3933507919311525,
      "memory(GiB)": 77.56,
      "step": 66450,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 2.8471359410479415,
      "grad_norm": 5.349114418029785,
      "learning_rate": 3.918066566991384e-05,
      "loss": 2.7145721435546877,
      "memory(GiB)": 77.56,
      "step": 66455,
      "token_acc": 0.44816053511705684,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 2.8473501563771904,
      "grad_norm": 4.3693718910217285,
      "learning_rate": 3.917409543842564e-05,
      "loss": 2.294993591308594,
      "memory(GiB)": 77.56,
      "step": 66460,
      "token_acc": 0.48427672955974843,
      "train_speed(iter/s)": 1.437892
    },
    {
      "epoch": 2.847564371706439,
      "grad_norm": 4.617372512817383,
      "learning_rate": 3.9167525403058855e-05,
      "loss": 2.529158020019531,
      "memory(GiB)": 77.56,
      "step": 66465,
      "token_acc": 0.44904458598726116,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 2.8477785870356884,
      "grad_norm": 5.825178623199463,
      "learning_rate": 3.9160955563932524e-05,
      "loss": 2.4784067153930662,
      "memory(GiB)": 77.56,
      "step": 66470,
      "token_acc": 0.4523809523809524,
      "train_speed(iter/s)": 1.437882
    },
    {
      "epoch": 2.8479928023649372,
      "grad_norm": 5.506708145141602,
      "learning_rate": 3.9154385921165635e-05,
      "loss": 2.5322898864746093,
      "memory(GiB)": 77.56,
      "step": 66475,
      "token_acc": 0.44983818770226536,
      "train_speed(iter/s)": 1.437898
    },
    {
      "epoch": 2.848207017694186,
      "grad_norm": 4.318180084228516,
      "learning_rate": 3.9147816474877226e-05,
      "loss": 2.268649673461914,
      "memory(GiB)": 77.56,
      "step": 66480,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 2.8484212330234353,
      "grad_norm": 5.781953811645508,
      "learning_rate": 3.914124722518628e-05,
      "loss": 2.2700672149658203,
      "memory(GiB)": 77.56,
      "step": 66485,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.437912
    },
    {
      "epoch": 2.848635448352684,
      "grad_norm": 6.998129367828369,
      "learning_rate": 3.913467817221185e-05,
      "loss": 2.3082517623901366,
      "memory(GiB)": 77.56,
      "step": 66490,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 2.848849663681933,
      "grad_norm": 6.134349822998047,
      "learning_rate": 3.912810931607291e-05,
      "loss": 2.4138042449951174,
      "memory(GiB)": 77.56,
      "step": 66495,
      "token_acc": 0.46360153256704983,
      "train_speed(iter/s)": 1.437932
    },
    {
      "epoch": 2.849063879011182,
      "grad_norm": 6.053704738616943,
      "learning_rate": 3.9121540656888475e-05,
      "loss": 2.2307310104370117,
      "memory(GiB)": 77.56,
      "step": 66500,
      "token_acc": 0.5486381322957199,
      "train_speed(iter/s)": 1.43793
    },
    {
      "epoch": 2.849063879011182,
      "eval_loss": 2.191592216491699,
      "eval_runtime": 14.3462,
      "eval_samples_per_second": 6.97,
      "eval_steps_per_second": 6.97,
      "eval_token_acc": 0.4857142857142857,
      "step": 66500
    },
    {
      "epoch": 2.849278094340431,
      "grad_norm": 4.378180503845215,
      "learning_rate": 3.9114972194777534e-05,
      "loss": 2.838070106506348,
      "memory(GiB)": 77.56,
      "step": 66505,
      "token_acc": 0.47110675808031344,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 2.84949230966968,
      "grad_norm": 5.601437568664551,
      "learning_rate": 3.910840392985908e-05,
      "loss": 2.3315408706665037,
      "memory(GiB)": 77.56,
      "step": 66510,
      "token_acc": 0.5371621621621622,
      "train_speed(iter/s)": 1.437448
    },
    {
      "epoch": 2.849706524998929,
      "grad_norm": 7.7358269691467285,
      "learning_rate": 3.910183586225212e-05,
      "loss": 2.4238819122314452,
      "memory(GiB)": 77.56,
      "step": 66515,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.437461
    },
    {
      "epoch": 2.849920740328178,
      "grad_norm": 4.937651634216309,
      "learning_rate": 3.9095267992075604e-05,
      "loss": 2.492909049987793,
      "memory(GiB)": 77.56,
      "step": 66520,
      "token_acc": 0.5056603773584906,
      "train_speed(iter/s)": 1.437491
    },
    {
      "epoch": 2.8501349556574267,
      "grad_norm": 7.730500221252441,
      "learning_rate": 3.908870031944856e-05,
      "loss": 2.6884361267089845,
      "memory(GiB)": 77.56,
      "step": 66525,
      "token_acc": 0.4397163120567376,
      "train_speed(iter/s)": 1.437493
    },
    {
      "epoch": 2.850349170986676,
      "grad_norm": 6.503012180328369,
      "learning_rate": 3.9082132844489935e-05,
      "loss": 2.5664562225341796,
      "memory(GiB)": 77.56,
      "step": 66530,
      "token_acc": 0.4934640522875817,
      "train_speed(iter/s)": 1.437489
    },
    {
      "epoch": 2.8505633863159248,
      "grad_norm": 7.859287738800049,
      "learning_rate": 3.907556556731871e-05,
      "loss": 2.737318992614746,
      "memory(GiB)": 77.56,
      "step": 66535,
      "token_acc": 0.5057034220532319,
      "train_speed(iter/s)": 1.437503
    },
    {
      "epoch": 2.8507776016451736,
      "grad_norm": 5.697571754455566,
      "learning_rate": 3.906899848805387e-05,
      "loss": 2.72196102142334,
      "memory(GiB)": 77.56,
      "step": 66540,
      "token_acc": 0.444794952681388,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 2.850991816974423,
      "grad_norm": 5.560765743255615,
      "learning_rate": 3.906243160681436e-05,
      "loss": 2.4972421646118166,
      "memory(GiB)": 77.56,
      "step": 66545,
      "token_acc": 0.4984984984984985,
      "train_speed(iter/s)": 1.437432
    },
    {
      "epoch": 2.8512060323036716,
      "grad_norm": 8.9849271774292,
      "learning_rate": 3.9055864923719185e-05,
      "loss": 2.608599853515625,
      "memory(GiB)": 77.56,
      "step": 66550,
      "token_acc": 0.43462897526501765,
      "train_speed(iter/s)": 1.437435
    },
    {
      "epoch": 2.8514202476329205,
      "grad_norm": 7.900022983551025,
      "learning_rate": 3.9049298438887276e-05,
      "loss": 2.554580307006836,
      "memory(GiB)": 77.56,
      "step": 66555,
      "token_acc": 0.5043731778425656,
      "train_speed(iter/s)": 1.437449
    },
    {
      "epoch": 2.8516344629621697,
      "grad_norm": 5.279074668884277,
      "learning_rate": 3.904273215243758e-05,
      "loss": 2.5848312377929688,
      "memory(GiB)": 77.56,
      "step": 66560,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.437472
    },
    {
      "epoch": 2.8518486782914185,
      "grad_norm": 5.624842643737793,
      "learning_rate": 3.903616606448908e-05,
      "loss": 2.5240001678466797,
      "memory(GiB)": 77.56,
      "step": 66565,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437496
    },
    {
      "epoch": 2.8520628936206673,
      "grad_norm": 5.991982460021973,
      "learning_rate": 3.902960017516073e-05,
      "loss": 2.513314628601074,
      "memory(GiB)": 77.56,
      "step": 66570,
      "token_acc": 0.4673202614379085,
      "train_speed(iter/s)": 1.437492
    },
    {
      "epoch": 2.8522771089499166,
      "grad_norm": 7.103693008422852,
      "learning_rate": 3.902303448457146e-05,
      "loss": 2.1770584106445314,
      "memory(GiB)": 77.56,
      "step": 66575,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.437469
    },
    {
      "epoch": 2.8524913242791654,
      "grad_norm": 5.116634368896484,
      "learning_rate": 3.9016468992840203e-05,
      "loss": 2.6007291793823244,
      "memory(GiB)": 77.56,
      "step": 66580,
      "token_acc": 0.46130952380952384,
      "train_speed(iter/s)": 1.437473
    },
    {
      "epoch": 2.852705539608414,
      "grad_norm": 4.407067775726318,
      "learning_rate": 3.900990370008593e-05,
      "loss": 2.4688270568847654,
      "memory(GiB)": 77.56,
      "step": 66585,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 2.8529197549376635,
      "grad_norm": 4.946870803833008,
      "learning_rate": 3.9003338606427564e-05,
      "loss": 2.391340637207031,
      "memory(GiB)": 77.56,
      "step": 66590,
      "token_acc": 0.5087719298245614,
      "train_speed(iter/s)": 1.437494
    },
    {
      "epoch": 2.8531339702669123,
      "grad_norm": 4.987757682800293,
      "learning_rate": 3.899677371198402e-05,
      "loss": 2.258224868774414,
      "memory(GiB)": 77.56,
      "step": 66595,
      "token_acc": 0.541958041958042,
      "train_speed(iter/s)": 1.437503
    },
    {
      "epoch": 2.853348185596161,
      "grad_norm": 5.315586566925049,
      "learning_rate": 3.8990209016874254e-05,
      "loss": 2.4487504959106445,
      "memory(GiB)": 77.56,
      "step": 66600,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.43749
    },
    {
      "epoch": 2.8535624009254104,
      "grad_norm": 6.663456916809082,
      "learning_rate": 3.898364452121717e-05,
      "loss": 2.3310081481933596,
      "memory(GiB)": 77.56,
      "step": 66605,
      "token_acc": 0.49606299212598426,
      "train_speed(iter/s)": 1.43751
    },
    {
      "epoch": 2.853776616254659,
      "grad_norm": 6.007467269897461,
      "learning_rate": 3.89770802251317e-05,
      "loss": 2.4017011642456056,
      "memory(GiB)": 77.56,
      "step": 66610,
      "token_acc": 0.4980544747081712,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 2.853990831583908,
      "grad_norm": 5.462390899658203,
      "learning_rate": 3.897051612873677e-05,
      "loss": 2.4776123046875,
      "memory(GiB)": 77.56,
      "step": 66615,
      "token_acc": 0.47038327526132406,
      "train_speed(iter/s)": 1.437523
    },
    {
      "epoch": 2.8542050469131572,
      "grad_norm": 4.1070637702941895,
      "learning_rate": 3.896395223215128e-05,
      "loss": 2.6370866775512694,
      "memory(GiB)": 77.56,
      "step": 66620,
      "token_acc": 0.4161073825503356,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 2.854419262242406,
      "grad_norm": 6.080334663391113,
      "learning_rate": 3.8957388535494156e-05,
      "loss": 2.1869640350341797,
      "memory(GiB)": 77.56,
      "step": 66625,
      "token_acc": 0.5113636363636364,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 2.854633477571655,
      "grad_norm": 4.292595386505127,
      "learning_rate": 3.8950825038884296e-05,
      "loss": 2.261898231506348,
      "memory(GiB)": 77.56,
      "step": 66630,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.437494
    },
    {
      "epoch": 2.854847692900904,
      "grad_norm": 5.835457801818848,
      "learning_rate": 3.894426174244058e-05,
      "loss": 2.2565582275390623,
      "memory(GiB)": 77.56,
      "step": 66635,
      "token_acc": 0.5518672199170125,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 2.855061908230153,
      "grad_norm": 5.705355644226074,
      "learning_rate": 3.893769864628195e-05,
      "loss": 2.4744895935058593,
      "memory(GiB)": 77.56,
      "step": 66640,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437528
    },
    {
      "epoch": 2.8552761235594017,
      "grad_norm": 5.244635581970215,
      "learning_rate": 3.89311357505273e-05,
      "loss": 2.405169868469238,
      "memory(GiB)": 77.56,
      "step": 66645,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.437563
    },
    {
      "epoch": 2.855490338888651,
      "grad_norm": 5.433347702026367,
      "learning_rate": 3.892457305529549e-05,
      "loss": 2.1917551040649412,
      "memory(GiB)": 77.56,
      "step": 66650,
      "token_acc": 0.5693430656934306,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 2.8557045542179,
      "grad_norm": 4.850085258483887,
      "learning_rate": 3.891801056070545e-05,
      "loss": 2.1592260360717774,
      "memory(GiB)": 77.56,
      "step": 66655,
      "token_acc": 0.5382165605095541,
      "train_speed(iter/s)": 1.437559
    },
    {
      "epoch": 2.8559187695471486,
      "grad_norm": 6.711618423461914,
      "learning_rate": 3.891144826687603e-05,
      "loss": 2.4707393646240234,
      "memory(GiB)": 77.56,
      "step": 66660,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437586
    },
    {
      "epoch": 2.856132984876398,
      "grad_norm": 7.5741286277771,
      "learning_rate": 3.890488617392613e-05,
      "loss": 2.37469482421875,
      "memory(GiB)": 77.56,
      "step": 66665,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.437599
    },
    {
      "epoch": 2.8563472002056467,
      "grad_norm": 6.927433013916016,
      "learning_rate": 3.889832428197463e-05,
      "loss": 2.4204429626464843,
      "memory(GiB)": 77.56,
      "step": 66670,
      "token_acc": 0.5066225165562914,
      "train_speed(iter/s)": 1.437614
    },
    {
      "epoch": 2.8565614155348955,
      "grad_norm": 5.925686836242676,
      "learning_rate": 3.88917625911404e-05,
      "loss": 2.3027008056640623,
      "memory(GiB)": 77.56,
      "step": 66675,
      "token_acc": 0.5186567164179104,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 2.8567756308641448,
      "grad_norm": 6.618916034698486,
      "learning_rate": 3.8885201101542304e-05,
      "loss": 2.3661354064941404,
      "memory(GiB)": 77.56,
      "step": 66680,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.437649
    },
    {
      "epoch": 2.8569898461933936,
      "grad_norm": 4.6011552810668945,
      "learning_rate": 3.8878639813299214e-05,
      "loss": 2.453974723815918,
      "memory(GiB)": 77.56,
      "step": 66685,
      "token_acc": 0.4980392156862745,
      "train_speed(iter/s)": 1.437653
    },
    {
      "epoch": 2.8572040615226424,
      "grad_norm": 4.523308277130127,
      "learning_rate": 3.887207872653e-05,
      "loss": 2.26114501953125,
      "memory(GiB)": 77.56,
      "step": 66690,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 2.8574182768518916,
      "grad_norm": 7.806247234344482,
      "learning_rate": 3.886551784135352e-05,
      "loss": 2.554484176635742,
      "memory(GiB)": 77.56,
      "step": 66695,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.437669
    },
    {
      "epoch": 2.8576324921811405,
      "grad_norm": 4.85680627822876,
      "learning_rate": 3.885895715788864e-05,
      "loss": 2.5340404510498047,
      "memory(GiB)": 77.56,
      "step": 66700,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.437675
    },
    {
      "epoch": 2.8578467075103893,
      "grad_norm": 5.873867034912109,
      "learning_rate": 3.8852396676254175e-05,
      "loss": 2.2370100021362305,
      "memory(GiB)": 77.56,
      "step": 66705,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.437697
    },
    {
      "epoch": 2.8580609228396385,
      "grad_norm": 5.588469982147217,
      "learning_rate": 3.8845836396569007e-05,
      "loss": 2.354682731628418,
      "memory(GiB)": 77.56,
      "step": 66710,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.437655
    },
    {
      "epoch": 2.8582751381688873,
      "grad_norm": 4.8850274085998535,
      "learning_rate": 3.8839276318952e-05,
      "loss": 2.4714897155761717,
      "memory(GiB)": 77.56,
      "step": 66715,
      "token_acc": 0.4484848484848485,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 2.858489353498136,
      "grad_norm": 6.62422513961792,
      "learning_rate": 3.883271644352196e-05,
      "loss": 2.377308464050293,
      "memory(GiB)": 77.56,
      "step": 66720,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.43761
    },
    {
      "epoch": 2.8587035688273854,
      "grad_norm": 5.528073787689209,
      "learning_rate": 3.882615677039773e-05,
      "loss": 2.336688995361328,
      "memory(GiB)": 77.56,
      "step": 66725,
      "token_acc": 0.5177514792899408,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 2.8589177841566342,
      "grad_norm": 8.789603233337402,
      "learning_rate": 3.881959729969816e-05,
      "loss": 2.3863702774047852,
      "memory(GiB)": 77.56,
      "step": 66730,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.437617
    },
    {
      "epoch": 2.859131999485883,
      "grad_norm": 5.422511100769043,
      "learning_rate": 3.881303803154208e-05,
      "loss": 2.707002639770508,
      "memory(GiB)": 77.56,
      "step": 66735,
      "token_acc": 0.4551282051282051,
      "train_speed(iter/s)": 1.437619
    },
    {
      "epoch": 2.8593462148151323,
      "grad_norm": 4.357863426208496,
      "learning_rate": 3.8806478966048295e-05,
      "loss": 2.2606287002563477,
      "memory(GiB)": 77.56,
      "step": 66740,
      "token_acc": 0.5464684014869888,
      "train_speed(iter/s)": 1.437613
    },
    {
      "epoch": 2.859560430144381,
      "grad_norm": 5.374544143676758,
      "learning_rate": 3.879992010333565e-05,
      "loss": 2.470383071899414,
      "memory(GiB)": 77.56,
      "step": 66745,
      "token_acc": 0.476878612716763,
      "train_speed(iter/s)": 1.437592
    },
    {
      "epoch": 2.85977464547363,
      "grad_norm": 7.355118274688721,
      "learning_rate": 3.879336144352296e-05,
      "loss": 2.633053016662598,
      "memory(GiB)": 77.56,
      "step": 66750,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 2.859988860802879,
      "grad_norm": 5.335455417633057,
      "learning_rate": 3.878680298672904e-05,
      "loss": 2.155171203613281,
      "memory(GiB)": 77.56,
      "step": 66755,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.437545
    },
    {
      "epoch": 2.860203076132128,
      "grad_norm": 5.805242538452148,
      "learning_rate": 3.8780244733072706e-05,
      "loss": 2.1418052673339845,
      "memory(GiB)": 77.56,
      "step": 66760,
      "token_acc": 0.4981684981684982,
      "train_speed(iter/s)": 1.43754
    },
    {
      "epoch": 2.860417291461377,
      "grad_norm": 7.094827651977539,
      "learning_rate": 3.8773686682672756e-05,
      "loss": 2.9565858840942383,
      "memory(GiB)": 77.56,
      "step": 66765,
      "token_acc": 0.43023255813953487,
      "train_speed(iter/s)": 1.437528
    },
    {
      "epoch": 2.860631506790626,
      "grad_norm": 5.819697380065918,
      "learning_rate": 3.876712883564801e-05,
      "loss": 2.5334810256958007,
      "memory(GiB)": 77.56,
      "step": 66770,
      "token_acc": 0.4652014652014652,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 2.860845722119875,
      "grad_norm": 7.383658409118652,
      "learning_rate": 3.876057119211727e-05,
      "loss": 2.4060733795166014,
      "memory(GiB)": 77.56,
      "step": 66775,
      "token_acc": 0.46283783783783783,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 2.8610599374491237,
      "grad_norm": 5.543306350708008,
      "learning_rate": 3.875401375219929e-05,
      "loss": 2.1814397811889648,
      "memory(GiB)": 77.56,
      "step": 66780,
      "token_acc": 0.5381944444444444,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 2.861274152778373,
      "grad_norm": 5.551831245422363,
      "learning_rate": 3.8747456516012926e-05,
      "loss": 2.328445053100586,
      "memory(GiB)": 77.56,
      "step": 66785,
      "token_acc": 0.47513812154696133,
      "train_speed(iter/s)": 1.43759
    },
    {
      "epoch": 2.8614883681076217,
      "grad_norm": 4.948244571685791,
      "learning_rate": 3.8740899483676946e-05,
      "loss": 2.114749717712402,
      "memory(GiB)": 77.56,
      "step": 66790,
      "token_acc": 0.532608695652174,
      "train_speed(iter/s)": 1.437542
    },
    {
      "epoch": 2.8617025834368706,
      "grad_norm": 4.717979907989502,
      "learning_rate": 3.8734342655310124e-05,
      "loss": 2.2499460220336913,
      "memory(GiB)": 77.56,
      "step": 66795,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.437539
    },
    {
      "epoch": 2.86191679876612,
      "grad_norm": 5.059933185577393,
      "learning_rate": 3.872778603103126e-05,
      "loss": 2.027072715759277,
      "memory(GiB)": 77.56,
      "step": 66800,
      "token_acc": 0.5748031496062992,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 2.8621310140953686,
      "grad_norm": 6.091126441955566,
      "learning_rate": 3.8721229610959136e-05,
      "loss": 2.4446929931640624,
      "memory(GiB)": 77.56,
      "step": 66805,
      "token_acc": 0.455026455026455,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 2.8623452294246174,
      "grad_norm": 9.122893333435059,
      "learning_rate": 3.87146733952125e-05,
      "loss": 2.501746940612793,
      "memory(GiB)": 77.56,
      "step": 66810,
      "token_acc": 0.4522058823529412,
      "train_speed(iter/s)": 1.437577
    },
    {
      "epoch": 2.8625594447538667,
      "grad_norm": 5.040594577789307,
      "learning_rate": 3.870811738391015e-05,
      "loss": 2.5487319946289064,
      "memory(GiB)": 77.56,
      "step": 66815,
      "token_acc": 0.4479495268138801,
      "train_speed(iter/s)": 1.437577
    },
    {
      "epoch": 2.8627736600831155,
      "grad_norm": 6.156131744384766,
      "learning_rate": 3.8701561577170855e-05,
      "loss": 2.3668354034423826,
      "memory(GiB)": 77.56,
      "step": 66820,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.437587
    },
    {
      "epoch": 2.8629878754123643,
      "grad_norm": 4.535671710968018,
      "learning_rate": 3.8695005975113357e-05,
      "loss": 2.6816427230834963,
      "memory(GiB)": 77.56,
      "step": 66825,
      "token_acc": 0.43558282208588955,
      "train_speed(iter/s)": 1.437606
    },
    {
      "epoch": 2.8632020907416136,
      "grad_norm": 5.114601135253906,
      "learning_rate": 3.868845057785644e-05,
      "loss": 2.303553009033203,
      "memory(GiB)": 77.56,
      "step": 66830,
      "token_acc": 0.5050847457627119,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 2.8634163060708624,
      "grad_norm": 5.828295707702637,
      "learning_rate": 3.868189538551884e-05,
      "loss": 2.3304513931274413,
      "memory(GiB)": 77.56,
      "step": 66835,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.437591
    },
    {
      "epoch": 2.863630521400111,
      "grad_norm": 6.106845378875732,
      "learning_rate": 3.8675340398219335e-05,
      "loss": 2.491108703613281,
      "memory(GiB)": 77.56,
      "step": 66840,
      "token_acc": 0.5224358974358975,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 2.8638447367293605,
      "grad_norm": 5.234866142272949,
      "learning_rate": 3.8668785616076665e-05,
      "loss": 2.6317298889160154,
      "memory(GiB)": 77.56,
      "step": 66845,
      "token_acc": 0.4591549295774648,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 2.8640589520586093,
      "grad_norm": 5.1936211585998535,
      "learning_rate": 3.866223103920956e-05,
      "loss": 2.3543069839477537,
      "memory(GiB)": 77.56,
      "step": 66850,
      "token_acc": 0.5280898876404494,
      "train_speed(iter/s)": 1.437589
    },
    {
      "epoch": 2.864273167387858,
      "grad_norm": 5.63826322555542,
      "learning_rate": 3.865567666773676e-05,
      "loss": 2.451177215576172,
      "memory(GiB)": 77.56,
      "step": 66855,
      "token_acc": 0.5192307692307693,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 2.8644873827171073,
      "grad_norm": 5.010950565338135,
      "learning_rate": 3.864912250177704e-05,
      "loss": 2.4862348556518556,
      "memory(GiB)": 77.56,
      "step": 66860,
      "token_acc": 0.4872611464968153,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 2.864701598046356,
      "grad_norm": 3.9060218334198,
      "learning_rate": 3.864256854144912e-05,
      "loss": 2.618460464477539,
      "memory(GiB)": 77.56,
      "step": 66865,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.437629
    },
    {
      "epoch": 2.864915813375605,
      "grad_norm": 4.667769432067871,
      "learning_rate": 3.863601478687171e-05,
      "loss": 2.0842124938964846,
      "memory(GiB)": 77.56,
      "step": 66870,
      "token_acc": 0.5397350993377483,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 2.8651300287048542,
      "grad_norm": 5.057943344116211,
      "learning_rate": 3.862946123816357e-05,
      "loss": 2.2692262649536135,
      "memory(GiB)": 77.56,
      "step": 66875,
      "token_acc": 0.4970414201183432,
      "train_speed(iter/s)": 1.43761
    },
    {
      "epoch": 2.865344244034103,
      "grad_norm": 7.909023284912109,
      "learning_rate": 3.86229078954434e-05,
      "loss": 2.307342529296875,
      "memory(GiB)": 77.56,
      "step": 66880,
      "token_acc": 0.4915254237288136,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 2.865558459363352,
      "grad_norm": 4.595394134521484,
      "learning_rate": 3.8616354758829915e-05,
      "loss": 2.4317653656005858,
      "memory(GiB)": 77.56,
      "step": 66885,
      "token_acc": 0.4306569343065693,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 2.865772674692601,
      "grad_norm": 5.636876583099365,
      "learning_rate": 3.8609801828441855e-05,
      "loss": 2.557910919189453,
      "memory(GiB)": 77.56,
      "step": 66890,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.437609
    },
    {
      "epoch": 2.86598689002185,
      "grad_norm": 5.9952521324157715,
      "learning_rate": 3.860324910439791e-05,
      "loss": 2.1720706939697267,
      "memory(GiB)": 77.56,
      "step": 66895,
      "token_acc": 0.484149855907781,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 2.8662011053510987,
      "grad_norm": 4.391422748565674,
      "learning_rate": 3.859669658681681e-05,
      "loss": 2.628008270263672,
      "memory(GiB)": 77.56,
      "step": 66900,
      "token_acc": 0.4713375796178344,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 2.866415320680348,
      "grad_norm": 6.580483436584473,
      "learning_rate": 3.859014427581724e-05,
      "loss": 2.326628303527832,
      "memory(GiB)": 77.56,
      "step": 66905,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 2.866629536009597,
      "grad_norm": 5.895464897155762,
      "learning_rate": 3.858359217151791e-05,
      "loss": 2.2959293365478515,
      "memory(GiB)": 77.56,
      "step": 66910,
      "token_acc": 0.5167785234899329,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 2.8668437513388456,
      "grad_norm": 3.5169382095336914,
      "learning_rate": 3.857704027403752e-05,
      "loss": 2.210647201538086,
      "memory(GiB)": 77.56,
      "step": 66915,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 2.867057966668095,
      "grad_norm": 6.579715728759766,
      "learning_rate": 3.8570488583494766e-05,
      "loss": 2.456513214111328,
      "memory(GiB)": 77.56,
      "step": 66920,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.437679
    },
    {
      "epoch": 2.8672721819973437,
      "grad_norm": 6.019247531890869,
      "learning_rate": 3.856393710000832e-05,
      "loss": 2.227117919921875,
      "memory(GiB)": 77.56,
      "step": 66925,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.437701
    },
    {
      "epoch": 2.8674863973265925,
      "grad_norm": 9.02134895324707,
      "learning_rate": 3.8557385823696875e-05,
      "loss": 2.2834991455078124,
      "memory(GiB)": 77.56,
      "step": 66930,
      "token_acc": 0.5241635687732342,
      "train_speed(iter/s)": 1.43771
    },
    {
      "epoch": 2.8677006126558418,
      "grad_norm": 6.506422996520996,
      "learning_rate": 3.855083475467913e-05,
      "loss": 2.6903707504272463,
      "memory(GiB)": 77.56,
      "step": 66935,
      "token_acc": 0.4370860927152318,
      "train_speed(iter/s)": 1.437716
    },
    {
      "epoch": 2.8679148279850906,
      "grad_norm": 5.70438814163208,
      "learning_rate": 3.854428389307376e-05,
      "loss": 2.3973888397216796,
      "memory(GiB)": 77.56,
      "step": 66940,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 2.8681290433143394,
      "grad_norm": 6.45711612701416,
      "learning_rate": 3.853773323899943e-05,
      "loss": 2.4502588272094727,
      "memory(GiB)": 77.56,
      "step": 66945,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 2.8683432586435886,
      "grad_norm": 6.896414756774902,
      "learning_rate": 3.853118279257482e-05,
      "loss": 2.211759567260742,
      "memory(GiB)": 77.56,
      "step": 66950,
      "token_acc": 0.5062761506276151,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 2.8685574739728374,
      "grad_norm": 4.425848484039307,
      "learning_rate": 3.852463255391858e-05,
      "loss": 2.283206558227539,
      "memory(GiB)": 77.56,
      "step": 66955,
      "token_acc": 0.5088757396449705,
      "train_speed(iter/s)": 1.437766
    },
    {
      "epoch": 2.8687716893020863,
      "grad_norm": 5.747751235961914,
      "learning_rate": 3.85180825231494e-05,
      "loss": 2.165032958984375,
      "memory(GiB)": 77.56,
      "step": 66960,
      "token_acc": 0.5570934256055363,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 2.8689859046313355,
      "grad_norm": 6.781982898712158,
      "learning_rate": 3.851153270038592e-05,
      "loss": 2.479697036743164,
      "memory(GiB)": 77.56,
      "step": 66965,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.437802
    },
    {
      "epoch": 2.8692001199605843,
      "grad_norm": 7.020091533660889,
      "learning_rate": 3.85049830857468e-05,
      "loss": 2.521246910095215,
      "memory(GiB)": 77.56,
      "step": 66970,
      "token_acc": 0.532608695652174,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 2.869414335289833,
      "grad_norm": 5.100048542022705,
      "learning_rate": 3.84984336793507e-05,
      "loss": 2.2837223052978515,
      "memory(GiB)": 77.56,
      "step": 66975,
      "token_acc": 0.5289855072463768,
      "train_speed(iter/s)": 1.437855
    },
    {
      "epoch": 2.8696285506190824,
      "grad_norm": 6.363317966461182,
      "learning_rate": 3.849188448131625e-05,
      "loss": 2.108646011352539,
      "memory(GiB)": 77.56,
      "step": 66980,
      "token_acc": 0.5613382899628253,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 2.869842765948331,
      "grad_norm": 5.7891645431518555,
      "learning_rate": 3.8485335491762126e-05,
      "loss": 2.3036727905273438,
      "memory(GiB)": 77.56,
      "step": 66985,
      "token_acc": 0.5379061371841155,
      "train_speed(iter/s)": 1.437849
    },
    {
      "epoch": 2.87005698127758,
      "grad_norm": 5.255712509155273,
      "learning_rate": 3.8478786710806946e-05,
      "loss": 2.4827831268310545,
      "memory(GiB)": 77.56,
      "step": 66990,
      "token_acc": 0.46886446886446886,
      "train_speed(iter/s)": 1.437855
    },
    {
      "epoch": 2.8702711966068293,
      "grad_norm": 5.912222862243652,
      "learning_rate": 3.847223813856934e-05,
      "loss": 2.238930892944336,
      "memory(GiB)": 77.56,
      "step": 66995,
      "token_acc": 0.5432098765432098,
      "train_speed(iter/s)": 1.437814
    },
    {
      "epoch": 2.870485411936078,
      "grad_norm": 6.585913181304932,
      "learning_rate": 3.846568977516797e-05,
      "loss": 2.569080352783203,
      "memory(GiB)": 77.56,
      "step": 67000,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 2.870485411936078,
      "eval_loss": 2.285208225250244,
      "eval_runtime": 14.177,
      "eval_samples_per_second": 7.054,
      "eval_steps_per_second": 7.054,
      "eval_token_acc": 0.4603616133518776,
      "step": 67000
    },
    {
      "epoch": 2.870699627265327,
      "grad_norm": 5.607452392578125,
      "learning_rate": 3.845914162072143e-05,
      "loss": 2.2495880126953125,
      "memory(GiB)": 77.56,
      "step": 67005,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.437342
    },
    {
      "epoch": 2.870913842594576,
      "grad_norm": 5.4821319580078125,
      "learning_rate": 3.845259367534838e-05,
      "loss": 2.571170425415039,
      "memory(GiB)": 77.56,
      "step": 67010,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.43737
    },
    {
      "epoch": 2.871128057923825,
      "grad_norm": 4.354070663452148,
      "learning_rate": 3.8446045939167425e-05,
      "loss": 2.3739681243896484,
      "memory(GiB)": 77.56,
      "step": 67015,
      "token_acc": 0.52046783625731,
      "train_speed(iter/s)": 1.437403
    },
    {
      "epoch": 2.871342273253074,
      "grad_norm": 5.79425048828125,
      "learning_rate": 3.8439498412297195e-05,
      "loss": 2.485355758666992,
      "memory(GiB)": 77.56,
      "step": 67020,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.43739
    },
    {
      "epoch": 2.871556488582323,
      "grad_norm": 5.1980743408203125,
      "learning_rate": 3.843295109485629e-05,
      "loss": 2.444027328491211,
      "memory(GiB)": 77.56,
      "step": 67025,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 2.871770703911572,
      "grad_norm": 7.363042831420898,
      "learning_rate": 3.842640398696332e-05,
      "loss": 2.584497833251953,
      "memory(GiB)": 77.56,
      "step": 67030,
      "token_acc": 0.4621212121212121,
      "train_speed(iter/s)": 1.437383
    },
    {
      "epoch": 2.8719849192408207,
      "grad_norm": 5.0718536376953125,
      "learning_rate": 3.841985708873691e-05,
      "loss": 2.4830461502075196,
      "memory(GiB)": 77.56,
      "step": 67035,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.437387
    },
    {
      "epoch": 2.87219913457007,
      "grad_norm": 5.529065132141113,
      "learning_rate": 3.8413310400295636e-05,
      "loss": 2.332920455932617,
      "memory(GiB)": 77.56,
      "step": 67040,
      "token_acc": 0.4783950617283951,
      "train_speed(iter/s)": 1.437412
    },
    {
      "epoch": 2.8724133498993187,
      "grad_norm": 8.685287475585938,
      "learning_rate": 3.8406763921758124e-05,
      "loss": 2.4722246170043944,
      "memory(GiB)": 77.56,
      "step": 67045,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 2.8726275652285675,
      "grad_norm": 6.294495582580566,
      "learning_rate": 3.840021765324297e-05,
      "loss": 2.3772228240966795,
      "memory(GiB)": 77.56,
      "step": 67050,
      "token_acc": 0.5205992509363296,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 2.872841780557817,
      "grad_norm": 5.3892388343811035,
      "learning_rate": 3.839367159486874e-05,
      "loss": 2.1171764373779296,
      "memory(GiB)": 77.56,
      "step": 67055,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 2.8730559958870656,
      "grad_norm": 5.201147556304932,
      "learning_rate": 3.838712574675404e-05,
      "loss": 2.3917259216308593,
      "memory(GiB)": 77.56,
      "step": 67060,
      "token_acc": 0.5236593059936908,
      "train_speed(iter/s)": 1.437456
    },
    {
      "epoch": 2.8732702112163144,
      "grad_norm": 7.4396562576293945,
      "learning_rate": 3.838058010901746e-05,
      "loss": 2.3074174880981446,
      "memory(GiB)": 77.56,
      "step": 67065,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437475
    },
    {
      "epoch": 2.8734844265455637,
      "grad_norm": 6.192290306091309,
      "learning_rate": 3.8374034681777546e-05,
      "loss": 2.577297019958496,
      "memory(GiB)": 77.56,
      "step": 67070,
      "token_acc": 0.4744744744744745,
      "train_speed(iter/s)": 1.437502
    },
    {
      "epoch": 2.8736986418748125,
      "grad_norm": 4.847055435180664,
      "learning_rate": 3.8367489465152905e-05,
      "loss": 2.4470558166503906,
      "memory(GiB)": 77.56,
      "step": 67075,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.437531
    },
    {
      "epoch": 2.8739128572040613,
      "grad_norm": 4.06255578994751,
      "learning_rate": 3.836094445926212e-05,
      "loss": 2.683848762512207,
      "memory(GiB)": 77.56,
      "step": 67080,
      "token_acc": 0.4604904632152589,
      "train_speed(iter/s)": 1.437554
    },
    {
      "epoch": 2.8741270725333106,
      "grad_norm": 5.534252643585205,
      "learning_rate": 3.835439966422373e-05,
      "loss": 2.389699172973633,
      "memory(GiB)": 77.56,
      "step": 67085,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.437554
    },
    {
      "epoch": 2.8743412878625594,
      "grad_norm": 4.587975978851318,
      "learning_rate": 3.834785508015633e-05,
      "loss": 2.599444580078125,
      "memory(GiB)": 77.56,
      "step": 67090,
      "token_acc": 0.4358108108108108,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 2.874555503191808,
      "grad_norm": 6.065348148345947,
      "learning_rate": 3.834131070717847e-05,
      "loss": 2.6929557800292967,
      "memory(GiB)": 77.56,
      "step": 67095,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.437598
    },
    {
      "epoch": 2.8747697185210574,
      "grad_norm": 5.553237438201904,
      "learning_rate": 3.833476654540869e-05,
      "loss": 2.803675079345703,
      "memory(GiB)": 77.56,
      "step": 67100,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.4376
    },
    {
      "epoch": 2.8749839338503063,
      "grad_norm": 5.479255199432373,
      "learning_rate": 3.8328222594965555e-05,
      "loss": 2.485291862487793,
      "memory(GiB)": 77.56,
      "step": 67105,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.437585
    },
    {
      "epoch": 2.875198149179555,
      "grad_norm": 5.464146137237549,
      "learning_rate": 3.832167885596763e-05,
      "loss": 2.490620422363281,
      "memory(GiB)": 77.56,
      "step": 67110,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 2.8754123645088043,
      "grad_norm": 5.560009479522705,
      "learning_rate": 3.8315135328533435e-05,
      "loss": 2.253011131286621,
      "memory(GiB)": 77.56,
      "step": 67115,
      "token_acc": 0.5252100840336135,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 2.875626579838053,
      "grad_norm": 4.747530460357666,
      "learning_rate": 3.830859201278153e-05,
      "loss": 2.211527633666992,
      "memory(GiB)": 77.56,
      "step": 67120,
      "token_acc": 0.528169014084507,
      "train_speed(iter/s)": 1.437558
    },
    {
      "epoch": 2.875840795167302,
      "grad_norm": 4.919352054595947,
      "learning_rate": 3.830204890883044e-05,
      "loss": 2.3209238052368164,
      "memory(GiB)": 77.56,
      "step": 67125,
      "token_acc": 0.4820717131474104,
      "train_speed(iter/s)": 1.437588
    },
    {
      "epoch": 2.876055010496551,
      "grad_norm": 4.848925590515137,
      "learning_rate": 3.829550601679872e-05,
      "loss": 2.360421371459961,
      "memory(GiB)": 77.56,
      "step": 67130,
      "token_acc": 0.5360824742268041,
      "train_speed(iter/s)": 1.437613
    },
    {
      "epoch": 2.8762692258258,
      "grad_norm": 4.4091644287109375,
      "learning_rate": 3.828896333680489e-05,
      "loss": 2.502337837219238,
      "memory(GiB)": 77.56,
      "step": 67135,
      "token_acc": 0.48089171974522293,
      "train_speed(iter/s)": 1.437604
    },
    {
      "epoch": 2.876483441155049,
      "grad_norm": 6.011263847351074,
      "learning_rate": 3.828242086896746e-05,
      "loss": 2.267398452758789,
      "memory(GiB)": 77.56,
      "step": 67140,
      "token_acc": 0.521311475409836,
      "train_speed(iter/s)": 1.437606
    },
    {
      "epoch": 2.876697656484298,
      "grad_norm": 6.308224201202393,
      "learning_rate": 3.827587861340498e-05,
      "loss": 2.236736297607422,
      "memory(GiB)": 77.56,
      "step": 67145,
      "token_acc": 0.5346534653465347,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 2.876911871813547,
      "grad_norm": 5.017220497131348,
      "learning_rate": 3.826933657023594e-05,
      "loss": 2.130982780456543,
      "memory(GiB)": 77.56,
      "step": 67150,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.437599
    },
    {
      "epoch": 2.8771260871427957,
      "grad_norm": 9.085668563842773,
      "learning_rate": 3.826279473957889e-05,
      "loss": 2.5644020080566405,
      "memory(GiB)": 77.56,
      "step": 67155,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.437605
    },
    {
      "epoch": 2.877340302472045,
      "grad_norm": 8.26830768585205,
      "learning_rate": 3.825625312155231e-05,
      "loss": 2.3862737655639648,
      "memory(GiB)": 77.56,
      "step": 67160,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 2.877554517801294,
      "grad_norm": 5.186570644378662,
      "learning_rate": 3.824971171627474e-05,
      "loss": 2.638486862182617,
      "memory(GiB)": 77.56,
      "step": 67165,
      "token_acc": 0.45110410094637227,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 2.8777687331305426,
      "grad_norm": 6.128634929656982,
      "learning_rate": 3.8243170523864655e-05,
      "loss": 2.4484525680541993,
      "memory(GiB)": 77.56,
      "step": 67170,
      "token_acc": 0.5175097276264592,
      "train_speed(iter/s)": 1.437654
    },
    {
      "epoch": 2.877982948459792,
      "grad_norm": 4.915328025817871,
      "learning_rate": 3.823662954444056e-05,
      "loss": 2.40826416015625,
      "memory(GiB)": 77.56,
      "step": 67175,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.437664
    },
    {
      "epoch": 2.8781971637890407,
      "grad_norm": 4.873974800109863,
      "learning_rate": 3.823008877812097e-05,
      "loss": 2.310495948791504,
      "memory(GiB)": 77.56,
      "step": 67180,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.8784113791182895,
      "grad_norm": 5.154747009277344,
      "learning_rate": 3.822354822502434e-05,
      "loss": 2.3504953384399414,
      "memory(GiB)": 77.56,
      "step": 67185,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.437706
    },
    {
      "epoch": 2.8786255944475387,
      "grad_norm": 5.984213352203369,
      "learning_rate": 3.821700788526921e-05,
      "loss": 2.4143396377563477,
      "memory(GiB)": 77.56,
      "step": 67190,
      "token_acc": 0.44984802431610943,
      "train_speed(iter/s)": 1.437687
    },
    {
      "epoch": 2.8788398097767876,
      "grad_norm": 5.142282962799072,
      "learning_rate": 3.8210467758974026e-05,
      "loss": 2.1933261871337892,
      "memory(GiB)": 77.56,
      "step": 67195,
      "token_acc": 0.5270758122743683,
      "train_speed(iter/s)": 1.437662
    },
    {
      "epoch": 2.8790540251060364,
      "grad_norm": 4.387335777282715,
      "learning_rate": 3.820392784625726e-05,
      "loss": 2.368913459777832,
      "memory(GiB)": 77.56,
      "step": 67200,
      "token_acc": 0.5092936802973977,
      "train_speed(iter/s)": 1.437664
    },
    {
      "epoch": 2.8792682404352856,
      "grad_norm": 6.414888381958008,
      "learning_rate": 3.819738814723742e-05,
      "loss": 2.5073263168334963,
      "memory(GiB)": 77.56,
      "step": 67205,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.437657
    },
    {
      "epoch": 2.8794824557645344,
      "grad_norm": 6.047554969787598,
      "learning_rate": 3.8190848662032974e-05,
      "loss": 2.3778564453125,
      "memory(GiB)": 77.56,
      "step": 67210,
      "token_acc": 0.4786885245901639,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 2.8796966710937832,
      "grad_norm": 7.756847381591797,
      "learning_rate": 3.8184309390762364e-05,
      "loss": 2.697977828979492,
      "memory(GiB)": 77.56,
      "step": 67215,
      "token_acc": 0.4280821917808219,
      "train_speed(iter/s)": 1.437693
    },
    {
      "epoch": 2.8799108864230325,
      "grad_norm": 7.670078754425049,
      "learning_rate": 3.817777033354409e-05,
      "loss": 2.2733713150024415,
      "memory(GiB)": 77.56,
      "step": 67220,
      "token_acc": 0.4860557768924303,
      "train_speed(iter/s)": 1.437706
    },
    {
      "epoch": 2.8801251017522813,
      "grad_norm": 8.820830345153809,
      "learning_rate": 3.817123149049657e-05,
      "loss": 2.411340522766113,
      "memory(GiB)": 77.56,
      "step": 67225,
      "token_acc": 0.5068027210884354,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 2.88033931708153,
      "grad_norm": 4.12614107131958,
      "learning_rate": 3.81646928617383e-05,
      "loss": 2.4285394668579103,
      "memory(GiB)": 77.56,
      "step": 67230,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.437751
    },
    {
      "epoch": 2.8805535324107794,
      "grad_norm": 6.352338790893555,
      "learning_rate": 3.8158154447387737e-05,
      "loss": 2.4290489196777343,
      "memory(GiB)": 77.56,
      "step": 67235,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 2.880767747740028,
      "grad_norm": 4.837244987487793,
      "learning_rate": 3.81516162475633e-05,
      "loss": 2.3741870880126954,
      "memory(GiB)": 77.56,
      "step": 67240,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 2.880981963069277,
      "grad_norm": 6.899613857269287,
      "learning_rate": 3.814507826238345e-05,
      "loss": 2.3662992477416993,
      "memory(GiB)": 77.56,
      "step": 67245,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 2.8811961783985263,
      "grad_norm": 4.979419708251953,
      "learning_rate": 3.813854049196664e-05,
      "loss": 2.2906234741210936,
      "memory(GiB)": 77.56,
      "step": 67250,
      "token_acc": 0.46785714285714286,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 2.881410393727775,
      "grad_norm": 4.446743011474609,
      "learning_rate": 3.813200293643129e-05,
      "loss": 2.6041790008544923,
      "memory(GiB)": 77.56,
      "step": 67255,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.881624609057024,
      "grad_norm": 6.072567939758301,
      "learning_rate": 3.8125465595895834e-05,
      "loss": 2.1654294967651366,
      "memory(GiB)": 77.56,
      "step": 67260,
      "token_acc": 0.5492957746478874,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.881838824386273,
      "grad_norm": 11.573997497558594,
      "learning_rate": 3.811892847047872e-05,
      "loss": 2.752538299560547,
      "memory(GiB)": 77.56,
      "step": 67265,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 2.882053039715522,
      "grad_norm": 4.243500709533691,
      "learning_rate": 3.811239156029835e-05,
      "loss": 2.3777835845947264,
      "memory(GiB)": 77.56,
      "step": 67270,
      "token_acc": 0.5058823529411764,
      "train_speed(iter/s)": 1.437737
    },
    {
      "epoch": 2.8822672550447708,
      "grad_norm": 6.113096714019775,
      "learning_rate": 3.8105854865473164e-05,
      "loss": 2.6341041564941405,
      "memory(GiB)": 77.56,
      "step": 67275,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 2.88248147037402,
      "grad_norm": 5.571936130523682,
      "learning_rate": 3.809931838612159e-05,
      "loss": 2.5526802062988283,
      "memory(GiB)": 77.56,
      "step": 67280,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437787
    },
    {
      "epoch": 2.882695685703269,
      "grad_norm": 4.928206443786621,
      "learning_rate": 3.8092782122362e-05,
      "loss": 1.855202865600586,
      "memory(GiB)": 77.56,
      "step": 67285,
      "token_acc": 0.5729537366548043,
      "train_speed(iter/s)": 1.437784
    },
    {
      "epoch": 2.8829099010325177,
      "grad_norm": 7.276062965393066,
      "learning_rate": 3.808624607431285e-05,
      "loss": 2.2683895111083983,
      "memory(GiB)": 77.56,
      "step": 67290,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.437797
    },
    {
      "epoch": 2.883124116361767,
      "grad_norm": 4.760739803314209,
      "learning_rate": 3.807971024209252e-05,
      "loss": 2.5077787399291993,
      "memory(GiB)": 77.56,
      "step": 67295,
      "token_acc": 0.5259259259259259,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 2.8833383316910157,
      "grad_norm": 5.484633922576904,
      "learning_rate": 3.8073174625819406e-05,
      "loss": 2.411721420288086,
      "memory(GiB)": 77.56,
      "step": 67300,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 2.8835525470202645,
      "grad_norm": 5.4311418533325195,
      "learning_rate": 3.8066639225611934e-05,
      "loss": 2.6528240203857423,
      "memory(GiB)": 77.56,
      "step": 67305,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 2.883766762349514,
      "grad_norm": 4.820439338684082,
      "learning_rate": 3.8060104041588494e-05,
      "loss": 2.3418437957763674,
      "memory(GiB)": 77.56,
      "step": 67310,
      "token_acc": 0.49828178694158076,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 2.8839809776787626,
      "grad_norm": 5.935538291931152,
      "learning_rate": 3.805356907386747e-05,
      "loss": 2.3616310119628907,
      "memory(GiB)": 77.56,
      "step": 67315,
      "token_acc": 0.46864686468646866,
      "train_speed(iter/s)": 1.437836
    },
    {
      "epoch": 2.8841951930080114,
      "grad_norm": 5.296969413757324,
      "learning_rate": 3.8047034322567246e-05,
      "loss": 2.4964599609375,
      "memory(GiB)": 77.56,
      "step": 67320,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437828
    },
    {
      "epoch": 2.8844094083372607,
      "grad_norm": 5.347250461578369,
      "learning_rate": 3.804049978780621e-05,
      "loss": 2.402444267272949,
      "memory(GiB)": 77.56,
      "step": 67325,
      "token_acc": 0.4751552795031056,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 2.8846236236665095,
      "grad_norm": 5.347058296203613,
      "learning_rate": 3.803396546970273e-05,
      "loss": 2.3301054000854493,
      "memory(GiB)": 77.56,
      "step": 67330,
      "token_acc": 0.4952978056426332,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 2.8848378389957583,
      "grad_norm": 6.328917503356934,
      "learning_rate": 3.80274313683752e-05,
      "loss": 2.4490352630615235,
      "memory(GiB)": 77.56,
      "step": 67335,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 2.8850520543250076,
      "grad_norm": 5.516933917999268,
      "learning_rate": 3.802089748394199e-05,
      "loss": 2.2419479370117186,
      "memory(GiB)": 77.56,
      "step": 67340,
      "token_acc": 0.5483870967741935,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 2.8852662696542564,
      "grad_norm": 4.902294158935547,
      "learning_rate": 3.801436381652144e-05,
      "loss": 2.5699737548828123,
      "memory(GiB)": 77.56,
      "step": 67345,
      "token_acc": 0.486404833836858,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 2.885480484983505,
      "grad_norm": 4.538571357727051,
      "learning_rate": 3.800783036623194e-05,
      "loss": 2.6129196166992186,
      "memory(GiB)": 77.56,
      "step": 67350,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.437867
    },
    {
      "epoch": 2.8856947003127544,
      "grad_norm": 5.637793064117432,
      "learning_rate": 3.8001297133191855e-05,
      "loss": 2.636853790283203,
      "memory(GiB)": 77.56,
      "step": 67355,
      "token_acc": 0.5162337662337663,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 2.8859089156420032,
      "grad_norm": 5.997646331787109,
      "learning_rate": 3.799476411751951e-05,
      "loss": 2.452348327636719,
      "memory(GiB)": 77.56,
      "step": 67360,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.437903
    },
    {
      "epoch": 2.886123130971252,
      "grad_norm": 5.581997871398926,
      "learning_rate": 3.798823131933329e-05,
      "loss": 2.5143415451049806,
      "memory(GiB)": 77.56,
      "step": 67365,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.437915
    },
    {
      "epoch": 2.8863373463005013,
      "grad_norm": 5.182910442352295,
      "learning_rate": 3.79816987387515e-05,
      "loss": 2.539763069152832,
      "memory(GiB)": 77.56,
      "step": 67370,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.437924
    },
    {
      "epoch": 2.88655156162975,
      "grad_norm": 4.865305423736572,
      "learning_rate": 3.797516637589252e-05,
      "loss": 2.399571990966797,
      "memory(GiB)": 77.56,
      "step": 67375,
      "token_acc": 0.48398576512455516,
      "train_speed(iter/s)": 1.43795
    },
    {
      "epoch": 2.886765776958999,
      "grad_norm": 6.325377941131592,
      "learning_rate": 3.79686342308747e-05,
      "loss": 2.6419435501098634,
      "memory(GiB)": 77.56,
      "step": 67380,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 2.886979992288248,
      "grad_norm": 4.160820484161377,
      "learning_rate": 3.796210230381635e-05,
      "loss": 2.205731010437012,
      "memory(GiB)": 77.56,
      "step": 67385,
      "token_acc": 0.5325077399380805,
      "train_speed(iter/s)": 1.437932
    },
    {
      "epoch": 2.887194207617497,
      "grad_norm": 6.333376884460449,
      "learning_rate": 3.795557059483581e-05,
      "loss": 2.538529396057129,
      "memory(GiB)": 77.56,
      "step": 67390,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 2.887408422946746,
      "grad_norm": 5.403401851654053,
      "learning_rate": 3.794903910405141e-05,
      "loss": 2.198395538330078,
      "memory(GiB)": 77.56,
      "step": 67395,
      "token_acc": 0.5397489539748954,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 2.887622638275995,
      "grad_norm": 6.6402130126953125,
      "learning_rate": 3.794250783158148e-05,
      "loss": 2.176066780090332,
      "memory(GiB)": 77.56,
      "step": 67400,
      "token_acc": 0.4962962962962963,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 2.887836853605244,
      "grad_norm": 6.510558128356934,
      "learning_rate": 3.793597677754431e-05,
      "loss": 2.2711370468139647,
      "memory(GiB)": 77.56,
      "step": 67405,
      "token_acc": 0.5044776119402985,
      "train_speed(iter/s)": 1.437964
    },
    {
      "epoch": 2.8880510689344927,
      "grad_norm": 5.322099685668945,
      "learning_rate": 3.792944594205825e-05,
      "loss": 2.542129707336426,
      "memory(GiB)": 77.56,
      "step": 67410,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.43795
    },
    {
      "epoch": 2.888265284263742,
      "grad_norm": 5.281303405761719,
      "learning_rate": 3.792291532524159e-05,
      "loss": 2.512421417236328,
      "memory(GiB)": 77.56,
      "step": 67415,
      "token_acc": 0.49107142857142855,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 2.8884794995929908,
      "grad_norm": 6.071463108062744,
      "learning_rate": 3.791638492721266e-05,
      "loss": 2.574600028991699,
      "memory(GiB)": 77.56,
      "step": 67420,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 2.8886937149222396,
      "grad_norm": 5.547637939453125,
      "learning_rate": 3.790985474808975e-05,
      "loss": 2.6919523239135743,
      "memory(GiB)": 77.56,
      "step": 67425,
      "token_acc": 0.47183098591549294,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 2.888907930251489,
      "grad_norm": 5.751314640045166,
      "learning_rate": 3.790332478799115e-05,
      "loss": 2.456285095214844,
      "memory(GiB)": 77.56,
      "step": 67430,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 2.8891221455807377,
      "grad_norm": 5.877937316894531,
      "learning_rate": 3.789679504703519e-05,
      "loss": 2.383719253540039,
      "memory(GiB)": 77.56,
      "step": 67435,
      "token_acc": 0.5034965034965035,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 2.8893363609099865,
      "grad_norm": 5.98226261138916,
      "learning_rate": 3.789026552534013e-05,
      "loss": 2.4926692962646486,
      "memory(GiB)": 77.56,
      "step": 67440,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.438019
    },
    {
      "epoch": 2.8895505762392357,
      "grad_norm": 4.849386215209961,
      "learning_rate": 3.788373622302426e-05,
      "loss": 2.1820966720581056,
      "memory(GiB)": 77.56,
      "step": 67445,
      "token_acc": 0.4935897435897436,
      "train_speed(iter/s)": 1.438034
    },
    {
      "epoch": 2.8897647915684845,
      "grad_norm": 4.841516971588135,
      "learning_rate": 3.787720714020587e-05,
      "loss": 2.5233510971069335,
      "memory(GiB)": 77.56,
      "step": 67450,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.438032
    },
    {
      "epoch": 2.8899790068977333,
      "grad_norm": 7.413880348205566,
      "learning_rate": 3.787067827700327e-05,
      "loss": 2.2927804946899415,
      "memory(GiB)": 77.56,
      "step": 67455,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 2.8901932222269826,
      "grad_norm": 5.3944411277771,
      "learning_rate": 3.7864149633534704e-05,
      "loss": 2.2840316772460936,
      "memory(GiB)": 77.56,
      "step": 67460,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.438031
    },
    {
      "epoch": 2.8904074375562314,
      "grad_norm": 6.688461780548096,
      "learning_rate": 3.785762120991844e-05,
      "loss": 2.2583633422851563,
      "memory(GiB)": 77.56,
      "step": 67465,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 2.8906216528854802,
      "grad_norm": 4.892764091491699,
      "learning_rate": 3.7851093006272775e-05,
      "loss": 2.6320043563842774,
      "memory(GiB)": 77.56,
      "step": 67470,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 2.8908358682147295,
      "grad_norm": 6.65464448928833,
      "learning_rate": 3.784456502271594e-05,
      "loss": 2.7293590545654296,
      "memory(GiB)": 77.56,
      "step": 67475,
      "token_acc": 0.45483870967741935,
      "train_speed(iter/s)": 1.438061
    },
    {
      "epoch": 2.8910500835439783,
      "grad_norm": 5.529016017913818,
      "learning_rate": 3.783803725936622e-05,
      "loss": 2.436013031005859,
      "memory(GiB)": 77.56,
      "step": 67480,
      "token_acc": 0.4560669456066946,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 2.891264298873227,
      "grad_norm": 6.098552703857422,
      "learning_rate": 3.783150971634187e-05,
      "loss": 2.463293266296387,
      "memory(GiB)": 77.56,
      "step": 67485,
      "token_acc": 0.4831804281345566,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 2.8914785142024764,
      "grad_norm": 5.520616054534912,
      "learning_rate": 3.782498239376113e-05,
      "loss": 2.5195693969726562,
      "memory(GiB)": 77.56,
      "step": 67490,
      "token_acc": 0.4448818897637795,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 2.891692729531725,
      "grad_norm": 6.759217262268066,
      "learning_rate": 3.781845529174226e-05,
      "loss": 2.475371742248535,
      "memory(GiB)": 77.56,
      "step": 67495,
      "token_acc": 0.4470588235294118,
      "train_speed(iter/s)": 1.438082
    },
    {
      "epoch": 2.891906944860974,
      "grad_norm": 5.455395221710205,
      "learning_rate": 3.7811928410403495e-05,
      "loss": 2.7228761672973634,
      "memory(GiB)": 77.56,
      "step": 67500,
      "token_acc": 0.46547314578005117,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 2.891906944860974,
      "eval_loss": 2.264267683029175,
      "eval_runtime": 14.619,
      "eval_samples_per_second": 6.84,
      "eval_steps_per_second": 6.84,
      "eval_token_acc": 0.458656330749354,
      "step": 67500
    },
    {
      "epoch": 2.8921211601902233,
      "grad_norm": 6.082050800323486,
      "learning_rate": 3.780540174986308e-05,
      "loss": 2.2605606079101563,
      "memory(GiB)": 77.56,
      "step": 67505,
      "token_acc": 0.4785992217898833,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 2.892335375519472,
      "grad_norm": 5.861954689025879,
      "learning_rate": 3.7798875310239254e-05,
      "loss": 2.2912506103515624,
      "memory(GiB)": 77.56,
      "step": 67510,
      "token_acc": 0.5019455252918288,
      "train_speed(iter/s)": 1.437649
    },
    {
      "epoch": 2.892549590848721,
      "grad_norm": 5.445698261260986,
      "learning_rate": 3.779234909165024e-05,
      "loss": 2.5807964324951174,
      "memory(GiB)": 77.56,
      "step": 67515,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 2.89276380617797,
      "grad_norm": 6.062221050262451,
      "learning_rate": 3.7785823094214265e-05,
      "loss": 2.5411703109741213,
      "memory(GiB)": 77.56,
      "step": 67520,
      "token_acc": 0.4674329501915709,
      "train_speed(iter/s)": 1.437661
    },
    {
      "epoch": 2.892978021507219,
      "grad_norm": 4.514892101287842,
      "learning_rate": 3.777929731804958e-05,
      "loss": 2.2641124725341797,
      "memory(GiB)": 77.56,
      "step": 67525,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.437659
    },
    {
      "epoch": 2.8931922368364678,
      "grad_norm": 5.516725063323975,
      "learning_rate": 3.777277176327439e-05,
      "loss": 2.4197708129882813,
      "memory(GiB)": 77.56,
      "step": 67530,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437656
    },
    {
      "epoch": 2.893406452165717,
      "grad_norm": 4.598053455352783,
      "learning_rate": 3.7766246430006904e-05,
      "loss": 2.4598995208740235,
      "memory(GiB)": 77.56,
      "step": 67535,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.437665
    },
    {
      "epoch": 2.893620667494966,
      "grad_norm": 4.395597457885742,
      "learning_rate": 3.7759721318365344e-05,
      "loss": 2.346821403503418,
      "memory(GiB)": 77.56,
      "step": 67540,
      "token_acc": 0.5094850948509485,
      "train_speed(iter/s)": 1.437669
    },
    {
      "epoch": 2.8938348828242146,
      "grad_norm": 6.470637798309326,
      "learning_rate": 3.7753196428467915e-05,
      "loss": 2.3654699325561523,
      "memory(GiB)": 77.56,
      "step": 67545,
      "token_acc": 0.5099337748344371,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 2.894049098153464,
      "grad_norm": 4.622028827667236,
      "learning_rate": 3.774667176043281e-05,
      "loss": 2.2633066177368164,
      "memory(GiB)": 77.56,
      "step": 67550,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 2.8942633134827127,
      "grad_norm": 4.285539150238037,
      "learning_rate": 3.7740147314378254e-05,
      "loss": 2.409061241149902,
      "memory(GiB)": 77.56,
      "step": 67555,
      "token_acc": 0.5313653136531366,
      "train_speed(iter/s)": 1.437646
    },
    {
      "epoch": 2.8944775288119615,
      "grad_norm": 5.370772838592529,
      "learning_rate": 3.7733623090422424e-05,
      "loss": 2.349081802368164,
      "memory(GiB)": 77.56,
      "step": 67560,
      "token_acc": 0.44666666666666666,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 2.8946917441412108,
      "grad_norm": 6.547722816467285,
      "learning_rate": 3.772709908868351e-05,
      "loss": 2.361754608154297,
      "memory(GiB)": 77.56,
      "step": 67565,
      "token_acc": 0.4859437751004016,
      "train_speed(iter/s)": 1.437616
    },
    {
      "epoch": 2.8949059594704596,
      "grad_norm": 5.333826541900635,
      "learning_rate": 3.772057530927973e-05,
      "loss": 2.525577163696289,
      "memory(GiB)": 77.56,
      "step": 67570,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 2.8951201747997084,
      "grad_norm": 6.340412139892578,
      "learning_rate": 3.771405175232923e-05,
      "loss": 2.3988569259643553,
      "memory(GiB)": 77.56,
      "step": 67575,
      "token_acc": 0.5068493150684932,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 2.8953343901289577,
      "grad_norm": 7.052786827087402,
      "learning_rate": 3.7707528417950215e-05,
      "loss": 2.3821653366088866,
      "memory(GiB)": 77.56,
      "step": 67580,
      "token_acc": 0.522633744855967,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 2.8955486054582065,
      "grad_norm": 4.3135986328125,
      "learning_rate": 3.770100530626086e-05,
      "loss": 2.5312820434570313,
      "memory(GiB)": 77.56,
      "step": 67585,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.43768
    },
    {
      "epoch": 2.8957628207874553,
      "grad_norm": 10.569611549377441,
      "learning_rate": 3.769448241737932e-05,
      "loss": 2.485857391357422,
      "memory(GiB)": 77.56,
      "step": 67590,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.8959770361167045,
      "grad_norm": 5.541058540344238,
      "learning_rate": 3.7687959751423755e-05,
      "loss": 2.4066404342651366,
      "memory(GiB)": 77.56,
      "step": 67595,
      "token_acc": 0.46984126984126984,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 2.8961912514459534,
      "grad_norm": 5.763273239135742,
      "learning_rate": 3.768143730851238e-05,
      "loss": 2.345656394958496,
      "memory(GiB)": 77.56,
      "step": 67600,
      "token_acc": 0.49372384937238495,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.896405466775202,
      "grad_norm": 9.806373596191406,
      "learning_rate": 3.767491508876331e-05,
      "loss": 2.5688247680664062,
      "memory(GiB)": 77.56,
      "step": 67605,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.437709
    },
    {
      "epoch": 2.8966196821044514,
      "grad_norm": 7.553590297698975,
      "learning_rate": 3.7668393092294723e-05,
      "loss": 2.611219024658203,
      "memory(GiB)": 77.56,
      "step": 67610,
      "token_acc": 0.4470198675496689,
      "train_speed(iter/s)": 1.437687
    },
    {
      "epoch": 2.8968338974337002,
      "grad_norm": 4.764229774475098,
      "learning_rate": 3.766187131922476e-05,
      "loss": 2.3265857696533203,
      "memory(GiB)": 77.56,
      "step": 67615,
      "token_acc": 0.49693251533742333,
      "train_speed(iter/s)": 1.437692
    },
    {
      "epoch": 2.897048112762949,
      "grad_norm": 5.240024089813232,
      "learning_rate": 3.765534976967157e-05,
      "loss": 2.3967214584350587,
      "memory(GiB)": 77.56,
      "step": 67620,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.437699
    },
    {
      "epoch": 2.8972623280921983,
      "grad_norm": 6.137687683105469,
      "learning_rate": 3.7648828443753305e-05,
      "loss": 2.387478256225586,
      "memory(GiB)": 77.56,
      "step": 67625,
      "token_acc": 0.46311475409836067,
      "train_speed(iter/s)": 1.43773
    },
    {
      "epoch": 2.897476543421447,
      "grad_norm": 5.4151105880737305,
      "learning_rate": 3.76423073415881e-05,
      "loss": 2.5631446838378906,
      "memory(GiB)": 77.56,
      "step": 67630,
      "token_acc": 0.46332046332046334,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 2.897690758750696,
      "grad_norm": 7.524933815002441,
      "learning_rate": 3.763578646329408e-05,
      "loss": 2.2344633102416993,
      "memory(GiB)": 77.56,
      "step": 67635,
      "token_acc": 0.5145985401459854,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 2.897904974079945,
      "grad_norm": 5.050754070281982,
      "learning_rate": 3.76292658089894e-05,
      "loss": 2.4366222381591798,
      "memory(GiB)": 77.56,
      "step": 67640,
      "token_acc": 0.43253968253968256,
      "train_speed(iter/s)": 1.437753
    },
    {
      "epoch": 2.898119189409194,
      "grad_norm": 4.685484886169434,
      "learning_rate": 3.7622745378792156e-05,
      "loss": 2.6455270767211916,
      "memory(GiB)": 77.56,
      "step": 67645,
      "token_acc": 0.4894894894894895,
      "train_speed(iter/s)": 1.437753
    },
    {
      "epoch": 2.898333404738443,
      "grad_norm": 5.1919403076171875,
      "learning_rate": 3.761622517282051e-05,
      "loss": 2.156467819213867,
      "memory(GiB)": 77.56,
      "step": 67650,
      "token_acc": 0.5670498084291188,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 2.898547620067692,
      "grad_norm": 4.4663777351379395,
      "learning_rate": 3.7609705191192555e-05,
      "loss": 2.167580223083496,
      "memory(GiB)": 77.56,
      "step": 67655,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437744
    },
    {
      "epoch": 2.898761835396941,
      "grad_norm": 6.099833011627197,
      "learning_rate": 3.760318543402641e-05,
      "loss": 2.646506500244141,
      "memory(GiB)": 77.56,
      "step": 67660,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.8989760507261897,
      "grad_norm": 6.132604598999023,
      "learning_rate": 3.75966659014402e-05,
      "loss": 2.4978260040283202,
      "memory(GiB)": 77.56,
      "step": 67665,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 2.899190266055439,
      "grad_norm": 5.117307662963867,
      "learning_rate": 3.759014659355199e-05,
      "loss": 2.4953018188476563,
      "memory(GiB)": 77.56,
      "step": 67670,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 2.8994044813846878,
      "grad_norm": 5.843194484710693,
      "learning_rate": 3.758362751047995e-05,
      "loss": 2.368437576293945,
      "memory(GiB)": 77.56,
      "step": 67675,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 2.8996186967139366,
      "grad_norm": 4.256525993347168,
      "learning_rate": 3.7577108652342124e-05,
      "loss": 2.5725986480712892,
      "memory(GiB)": 77.56,
      "step": 67680,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 2.899832912043186,
      "grad_norm": 5.475727558135986,
      "learning_rate": 3.7570590019256645e-05,
      "loss": 2.2663818359375,
      "memory(GiB)": 77.56,
      "step": 67685,
      "token_acc": 0.5189873417721519,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 2.9000471273724346,
      "grad_norm": 5.523690700531006,
      "learning_rate": 3.756407161134159e-05,
      "loss": 2.3974020004272463,
      "memory(GiB)": 77.56,
      "step": 67690,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.437746
    },
    {
      "epoch": 2.900261342701684,
      "grad_norm": 4.970705509185791,
      "learning_rate": 3.7557553428715026e-05,
      "loss": 2.6790163040161135,
      "memory(GiB)": 77.56,
      "step": 67695,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 2.9004755580309327,
      "grad_norm": 4.712221145629883,
      "learning_rate": 3.7551035471495065e-05,
      "loss": 2.28826961517334,
      "memory(GiB)": 77.56,
      "step": 67700,
      "token_acc": 0.49640287769784175,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 2.9006897733601815,
      "grad_norm": 6.098865509033203,
      "learning_rate": 3.7544517739799775e-05,
      "loss": 2.6099933624267577,
      "memory(GiB)": 77.56,
      "step": 67705,
      "token_acc": 0.4591439688715953,
      "train_speed(iter/s)": 1.43768
    },
    {
      "epoch": 2.900903988689431,
      "grad_norm": 6.39176082611084,
      "learning_rate": 3.753800023374723e-05,
      "loss": 2.5080766677856445,
      "memory(GiB)": 77.56,
      "step": 67710,
      "token_acc": 0.4797687861271676,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 2.9011182040186796,
      "grad_norm": 4.354471683502197,
      "learning_rate": 3.7531482953455504e-05,
      "loss": 2.0768817901611327,
      "memory(GiB)": 77.56,
      "step": 67715,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.437654
    },
    {
      "epoch": 2.9013324193479284,
      "grad_norm": 4.666151523590088,
      "learning_rate": 3.752496589904265e-05,
      "loss": 2.4172800064086912,
      "memory(GiB)": 77.56,
      "step": 67720,
      "token_acc": 0.4774774774774775,
      "train_speed(iter/s)": 1.437677
    },
    {
      "epoch": 2.9015466346771777,
      "grad_norm": 4.78005313873291,
      "learning_rate": 3.751844907062675e-05,
      "loss": 2.4841840744018553,
      "memory(GiB)": 77.56,
      "step": 67725,
      "token_acc": 0.44107744107744107,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 2.9017608500064265,
      "grad_norm": 4.88601541519165,
      "learning_rate": 3.7511932468325855e-05,
      "loss": 2.3090999603271483,
      "memory(GiB)": 77.56,
      "step": 67730,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.437674
    },
    {
      "epoch": 2.9019750653356753,
      "grad_norm": 6.03407621383667,
      "learning_rate": 3.750541609225801e-05,
      "loss": 2.605454444885254,
      "memory(GiB)": 77.56,
      "step": 67735,
      "token_acc": 0.46325878594249204,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 2.9021892806649245,
      "grad_norm": 4.889484405517578,
      "learning_rate": 3.749889994254126e-05,
      "loss": 2.414371681213379,
      "memory(GiB)": 77.56,
      "step": 67740,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.437673
    },
    {
      "epoch": 2.9024034959941734,
      "grad_norm": 7.421288967132568,
      "learning_rate": 3.7492384019293686e-05,
      "loss": 2.4009025573730467,
      "memory(GiB)": 77.56,
      "step": 67745,
      "token_acc": 0.5204081632653061,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 2.902617711323422,
      "grad_norm": 6.974843502044678,
      "learning_rate": 3.74858683226333e-05,
      "loss": 2.2267650604248046,
      "memory(GiB)": 77.56,
      "step": 67750,
      "token_acc": 0.5058365758754864,
      "train_speed(iter/s)": 1.437635
    },
    {
      "epoch": 2.9028319266526714,
      "grad_norm": 4.531940937042236,
      "learning_rate": 3.747935285267816e-05,
      "loss": 2.5790096282958985,
      "memory(GiB)": 77.56,
      "step": 67755,
      "token_acc": 0.45686900958466453,
      "train_speed(iter/s)": 1.437636
    },
    {
      "epoch": 2.9030461419819202,
      "grad_norm": 5.8044753074646,
      "learning_rate": 3.747283760954629e-05,
      "loss": 2.391276741027832,
      "memory(GiB)": 77.56,
      "step": 67760,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.437651
    },
    {
      "epoch": 2.903260357311169,
      "grad_norm": 6.199559211730957,
      "learning_rate": 3.74663225933557e-05,
      "loss": 2.37835578918457,
      "memory(GiB)": 77.56,
      "step": 67765,
      "token_acc": 0.4825174825174825,
      "train_speed(iter/s)": 1.437656
    },
    {
      "epoch": 2.9034745726404183,
      "grad_norm": 5.674834728240967,
      "learning_rate": 3.745980780422445e-05,
      "loss": 2.1458560943603517,
      "memory(GiB)": 77.56,
      "step": 67770,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.437676
    },
    {
      "epoch": 2.903688787969667,
      "grad_norm": 5.815678119659424,
      "learning_rate": 3.7453293242270536e-05,
      "loss": 2.400010871887207,
      "memory(GiB)": 77.56,
      "step": 67775,
      "token_acc": 0.47876447876447875,
      "train_speed(iter/s)": 1.43769
    },
    {
      "epoch": 2.903903003298916,
      "grad_norm": 4.094466209411621,
      "learning_rate": 3.744677890761198e-05,
      "loss": 2.449346160888672,
      "memory(GiB)": 77.56,
      "step": 67780,
      "token_acc": 0.47278911564625853,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 2.904117218628165,
      "grad_norm": 8.096376419067383,
      "learning_rate": 3.7440264800366815e-05,
      "loss": 2.468935966491699,
      "memory(GiB)": 77.56,
      "step": 67785,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.904331433957414,
      "grad_norm": 6.131056785583496,
      "learning_rate": 3.743375092065302e-05,
      "loss": 2.5726037979125977,
      "memory(GiB)": 77.56,
      "step": 67790,
      "token_acc": 0.5145985401459854,
      "train_speed(iter/s)": 1.437702
    },
    {
      "epoch": 2.904545649286663,
      "grad_norm": 4.5523881912231445,
      "learning_rate": 3.7427237268588626e-05,
      "loss": 2.3564420700073243,
      "memory(GiB)": 77.56,
      "step": 67795,
      "token_acc": 0.5335570469798657,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 2.904759864615912,
      "grad_norm": 4.512152194976807,
      "learning_rate": 3.7420723844291625e-05,
      "loss": 2.4989864349365236,
      "memory(GiB)": 77.56,
      "step": 67800,
      "token_acc": 0.45187165775401067,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 2.904974079945161,
      "grad_norm": 5.770841121673584,
      "learning_rate": 3.7414210647879996e-05,
      "loss": 3.049637794494629,
      "memory(GiB)": 77.56,
      "step": 67805,
      "token_acc": 0.4057507987220447,
      "train_speed(iter/s)": 1.437717
    },
    {
      "epoch": 2.9051882952744097,
      "grad_norm": 6.343110084533691,
      "learning_rate": 3.740769767947177e-05,
      "loss": 2.3117307662963866,
      "memory(GiB)": 77.56,
      "step": 67810,
      "token_acc": 0.4894366197183099,
      "train_speed(iter/s)": 1.43772
    },
    {
      "epoch": 2.905402510603659,
      "grad_norm": 6.3243207931518555,
      "learning_rate": 3.740118493918489e-05,
      "loss": 2.2327476501464845,
      "memory(GiB)": 77.56,
      "step": 67815,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 2.9056167259329078,
      "grad_norm": 4.467010021209717,
      "learning_rate": 3.739467242713737e-05,
      "loss": 2.192717742919922,
      "memory(GiB)": 77.56,
      "step": 67820,
      "token_acc": 0.5415162454873647,
      "train_speed(iter/s)": 1.437771
    },
    {
      "epoch": 2.9058309412621566,
      "grad_norm": 9.37385082244873,
      "learning_rate": 3.7388160143447184e-05,
      "loss": 2.5006927490234374,
      "memory(GiB)": 77.56,
      "step": 67825,
      "token_acc": 0.46887966804979253,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 2.906045156591406,
      "grad_norm": 8.409736633300781,
      "learning_rate": 3.738164808823231e-05,
      "loss": 2.2262020111083984,
      "memory(GiB)": 77.56,
      "step": 67830,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 2.9062593719206546,
      "grad_norm": 6.2491841316223145,
      "learning_rate": 3.737513626161073e-05,
      "loss": 2.3818056106567385,
      "memory(GiB)": 77.56,
      "step": 67835,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437796
    },
    {
      "epoch": 2.9064735872499035,
      "grad_norm": 5.542945861816406,
      "learning_rate": 3.736862466370038e-05,
      "loss": 2.5873130798339843,
      "memory(GiB)": 77.56,
      "step": 67840,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.437803
    },
    {
      "epoch": 2.9066878025791527,
      "grad_norm": 5.3912458419799805,
      "learning_rate": 3.7362113294619264e-05,
      "loss": 2.364415740966797,
      "memory(GiB)": 77.56,
      "step": 67845,
      "token_acc": 0.503030303030303,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 2.9069020179084015,
      "grad_norm": 4.695046424865723,
      "learning_rate": 3.73556021544853e-05,
      "loss": 2.3008392333984373,
      "memory(GiB)": 77.56,
      "step": 67850,
      "token_acc": 0.5150501672240803,
      "train_speed(iter/s)": 1.437787
    },
    {
      "epoch": 2.9071162332376503,
      "grad_norm": 6.187320232391357,
      "learning_rate": 3.734909124341648e-05,
      "loss": 2.3105842590332033,
      "memory(GiB)": 77.56,
      "step": 67855,
      "token_acc": 0.508695652173913,
      "train_speed(iter/s)": 1.437771
    },
    {
      "epoch": 2.9073304485668996,
      "grad_norm": 6.184545993804932,
      "learning_rate": 3.734258056153074e-05,
      "loss": 2.7187423706054688,
      "memory(GiB)": 77.56,
      "step": 67860,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.437773
    },
    {
      "epoch": 2.9075446638961484,
      "grad_norm": 5.685884475708008,
      "learning_rate": 3.7336070108946016e-05,
      "loss": 2.7080142974853514,
      "memory(GiB)": 77.56,
      "step": 67865,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 2.907758879225397,
      "grad_norm": 4.762025356292725,
      "learning_rate": 3.732955988578027e-05,
      "loss": 2.336121368408203,
      "memory(GiB)": 77.56,
      "step": 67870,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 2.9079730945546465,
      "grad_norm": 3.8940138816833496,
      "learning_rate": 3.732304989215143e-05,
      "loss": 2.3883819580078125,
      "memory(GiB)": 77.56,
      "step": 67875,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.437786
    },
    {
      "epoch": 2.9081873098838953,
      "grad_norm": 6.916037082672119,
      "learning_rate": 3.731654012817743e-05,
      "loss": 2.383189582824707,
      "memory(GiB)": 77.56,
      "step": 67880,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 2.908401525213144,
      "grad_norm": 7.057248115539551,
      "learning_rate": 3.7310030593976205e-05,
      "loss": 2.4842044830322267,
      "memory(GiB)": 77.56,
      "step": 67885,
      "token_acc": 0.4379310344827586,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 2.9086157405423934,
      "grad_norm": 5.1528143882751465,
      "learning_rate": 3.730352128966567e-05,
      "loss": 2.4627311706542967,
      "memory(GiB)": 77.56,
      "step": 67890,
      "token_acc": 0.5043478260869565,
      "train_speed(iter/s)": 1.437779
    },
    {
      "epoch": 2.908829955871642,
      "grad_norm": 5.144994258880615,
      "learning_rate": 3.7297012215363756e-05,
      "loss": 2.4363882064819338,
      "memory(GiB)": 77.56,
      "step": 67895,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 2.909044171200891,
      "grad_norm": 4.467019557952881,
      "learning_rate": 3.7290503371188404e-05,
      "loss": 2.1492786407470703,
      "memory(GiB)": 77.56,
      "step": 67900,
      "token_acc": 0.5348837209302325,
      "train_speed(iter/s)": 1.437796
    },
    {
      "epoch": 2.9092583865301402,
      "grad_norm": 6.21030330657959,
      "learning_rate": 3.72839947572575e-05,
      "loss": 2.627834701538086,
      "memory(GiB)": 77.56,
      "step": 67905,
      "token_acc": 0.46439628482972134,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 2.909472601859389,
      "grad_norm": 4.537671089172363,
      "learning_rate": 3.727748637368895e-05,
      "loss": 2.5181861877441407,
      "memory(GiB)": 77.56,
      "step": 67910,
      "token_acc": 0.45609065155807366,
      "train_speed(iter/s)": 1.437817
    },
    {
      "epoch": 2.909686817188638,
      "grad_norm": 4.239520072937012,
      "learning_rate": 3.727097822060068e-05,
      "loss": 2.6376447677612305,
      "memory(GiB)": 77.56,
      "step": 67915,
      "token_acc": 0.47678018575851394,
      "train_speed(iter/s)": 1.437789
    },
    {
      "epoch": 2.909901032517887,
      "grad_norm": 4.422943592071533,
      "learning_rate": 3.7264470298110576e-05,
      "loss": 2.459255599975586,
      "memory(GiB)": 77.56,
      "step": 67920,
      "token_acc": 0.5015673981191222,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 2.910115247847136,
      "grad_norm": 6.4508957862854,
      "learning_rate": 3.725796260633653e-05,
      "loss": 2.1374263763427734,
      "memory(GiB)": 77.56,
      "step": 67925,
      "token_acc": 0.541095890410959,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 2.9103294631763847,
      "grad_norm": 7.026559352874756,
      "learning_rate": 3.725145514539646e-05,
      "loss": 2.7312219619750975,
      "memory(GiB)": 77.56,
      "step": 67930,
      "token_acc": 0.45733788395904434,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 2.910543678505634,
      "grad_norm": 7.377779960632324,
      "learning_rate": 3.724494791540823e-05,
      "loss": 2.2984180450439453,
      "memory(GiB)": 77.56,
      "step": 67935,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 2.910757893834883,
      "grad_norm": 5.123602390289307,
      "learning_rate": 3.723844091648974e-05,
      "loss": 2.4122140884399412,
      "memory(GiB)": 77.56,
      "step": 67940,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.437805
    },
    {
      "epoch": 2.9109721091641316,
      "grad_norm": 5.724231719970703,
      "learning_rate": 3.7231934148758865e-05,
      "loss": 2.4591293334960938,
      "memory(GiB)": 77.56,
      "step": 67945,
      "token_acc": 0.49387755102040815,
      "train_speed(iter/s)": 1.4378
    },
    {
      "epoch": 2.911186324493381,
      "grad_norm": 4.906028747558594,
      "learning_rate": 3.7225427612333465e-05,
      "loss": 2.4353408813476562,
      "memory(GiB)": 77.56,
      "step": 67950,
      "token_acc": 0.45390070921985815,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 2.9114005398226297,
      "grad_norm": 5.741007328033447,
      "learning_rate": 3.7218921307331446e-05,
      "loss": 2.701954650878906,
      "memory(GiB)": 77.56,
      "step": 67955,
      "token_acc": 0.46835443037974683,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 2.9116147551518785,
      "grad_norm": 6.2724432945251465,
      "learning_rate": 3.7212415233870646e-05,
      "loss": 2.256239318847656,
      "memory(GiB)": 77.56,
      "step": 67960,
      "token_acc": 0.4708333333333333,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 2.9118289704811278,
      "grad_norm": 4.073914527893066,
      "learning_rate": 3.7205909392068926e-05,
      "loss": 2.565846061706543,
      "memory(GiB)": 77.56,
      "step": 67965,
      "token_acc": 0.478125,
      "train_speed(iter/s)": 1.437815
    },
    {
      "epoch": 2.9120431858103766,
      "grad_norm": 7.0176262855529785,
      "learning_rate": 3.719940378204417e-05,
      "loss": 2.505396270751953,
      "memory(GiB)": 77.56,
      "step": 67970,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.43784
    },
    {
      "epoch": 2.9122574011396254,
      "grad_norm": 4.375279903411865,
      "learning_rate": 3.7192898403914236e-05,
      "loss": 2.441196250915527,
      "memory(GiB)": 77.56,
      "step": 67975,
      "token_acc": 0.44884488448844884,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 2.9124716164688746,
      "grad_norm": 4.9972333908081055,
      "learning_rate": 3.718639325779696e-05,
      "loss": 2.30728702545166,
      "memory(GiB)": 77.56,
      "step": 67980,
      "token_acc": 0.5390070921985816,
      "train_speed(iter/s)": 1.437855
    },
    {
      "epoch": 2.9126858317981235,
      "grad_norm": 5.396003723144531,
      "learning_rate": 3.717988834381017e-05,
      "loss": 2.4788684844970703,
      "memory(GiB)": 77.56,
      "step": 67985,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.437844
    },
    {
      "epoch": 2.9129000471273723,
      "grad_norm": 6.3774189949035645,
      "learning_rate": 3.717338366207175e-05,
      "loss": 2.435546112060547,
      "memory(GiB)": 77.56,
      "step": 67990,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 2.9131142624566215,
      "grad_norm": 4.868289470672607,
      "learning_rate": 3.71668792126995e-05,
      "loss": 2.606495666503906,
      "memory(GiB)": 77.56,
      "step": 67995,
      "token_acc": 0.44904458598726116,
      "train_speed(iter/s)": 1.437854
    },
    {
      "epoch": 2.9133284777858703,
      "grad_norm": 4.707134246826172,
      "learning_rate": 3.716037499581129e-05,
      "loss": 2.5440460205078126,
      "memory(GiB)": 77.56,
      "step": 68000,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.437863
    },
    {
      "epoch": 2.9133284777858703,
      "eval_loss": 2.0957367420196533,
      "eval_runtime": 14.1304,
      "eval_samples_per_second": 7.077,
      "eval_steps_per_second": 7.077,
      "eval_token_acc": 0.49319213313161875,
      "step": 68000
    },
    {
      "epoch": 2.913542693115119,
      "grad_norm": 5.822622776031494,
      "learning_rate": 3.7153871011524924e-05,
      "loss": 2.9383464813232423,
      "memory(GiB)": 77.56,
      "step": 68005,
      "token_acc": 0.4668737060041408,
      "train_speed(iter/s)": 1.437398
    },
    {
      "epoch": 2.9137569084443684,
      "grad_norm": 4.303014755249023,
      "learning_rate": 3.714736725995823e-05,
      "loss": 2.4112377166748047,
      "memory(GiB)": 77.56,
      "step": 68010,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.437415
    },
    {
      "epoch": 2.9139711237736172,
      "grad_norm": 6.689478397369385,
      "learning_rate": 3.714086374122904e-05,
      "loss": 2.498320770263672,
      "memory(GiB)": 77.56,
      "step": 68015,
      "token_acc": 0.4775510204081633,
      "train_speed(iter/s)": 1.437439
    },
    {
      "epoch": 2.914185339102866,
      "grad_norm": 5.524967670440674,
      "learning_rate": 3.713436045545518e-05,
      "loss": 2.1342063903808595,
      "memory(GiB)": 77.56,
      "step": 68020,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 2.9143995544321153,
      "grad_norm": 5.780593395233154,
      "learning_rate": 3.712785740275442e-05,
      "loss": 2.1418651580810546,
      "memory(GiB)": 77.56,
      "step": 68025,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 2.914613769761364,
      "grad_norm": 4.24763298034668,
      "learning_rate": 3.7121354583244615e-05,
      "loss": 2.6473567962646483,
      "memory(GiB)": 77.56,
      "step": 68030,
      "token_acc": 0.49038461538461536,
      "train_speed(iter/s)": 1.437443
    },
    {
      "epoch": 2.9148279850906134,
      "grad_norm": 6.561647415161133,
      "learning_rate": 3.711485199704353e-05,
      "loss": 2.3959327697753907,
      "memory(GiB)": 77.56,
      "step": 68035,
      "token_acc": 0.472,
      "train_speed(iter/s)": 1.43745
    },
    {
      "epoch": 2.915042200419862,
      "grad_norm": 5.502318382263184,
      "learning_rate": 3.710834964426901e-05,
      "loss": 2.190201187133789,
      "memory(GiB)": 77.56,
      "step": 68040,
      "token_acc": 0.5432525951557093,
      "train_speed(iter/s)": 1.437458
    },
    {
      "epoch": 2.915256415749111,
      "grad_norm": 5.741339206695557,
      "learning_rate": 3.710184752503883e-05,
      "loss": 2.3484983444213867,
      "memory(GiB)": 77.56,
      "step": 68045,
      "token_acc": 0.5104602510460251,
      "train_speed(iter/s)": 1.437478
    },
    {
      "epoch": 2.9154706310783602,
      "grad_norm": 5.461989402770996,
      "learning_rate": 3.709534563947078e-05,
      "loss": 2.5413816452026365,
      "memory(GiB)": 77.56,
      "step": 68050,
      "token_acc": 0.4559748427672956,
      "train_speed(iter/s)": 1.437497
    },
    {
      "epoch": 2.915684846407609,
      "grad_norm": 4.681112766265869,
      "learning_rate": 3.708884398768265e-05,
      "loss": 2.6381193161010743,
      "memory(GiB)": 77.56,
      "step": 68055,
      "token_acc": 0.4560810810810811,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 2.915899061736858,
      "grad_norm": 4.501175403594971,
      "learning_rate": 3.708234256979223e-05,
      "loss": 2.4485021591186524,
      "memory(GiB)": 77.56,
      "step": 68060,
      "token_acc": 0.4968944099378882,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 2.916113277066107,
      "grad_norm": 6.423325061798096,
      "learning_rate": 3.7075841385917285e-05,
      "loss": 2.4590570449829103,
      "memory(GiB)": 77.56,
      "step": 68065,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.43754
    },
    {
      "epoch": 2.916327492395356,
      "grad_norm": 5.460845470428467,
      "learning_rate": 3.706934043617559e-05,
      "loss": 2.239404487609863,
      "memory(GiB)": 77.56,
      "step": 68070,
      "token_acc": 0.511400651465798,
      "train_speed(iter/s)": 1.43755
    },
    {
      "epoch": 2.9165417077246047,
      "grad_norm": 4.988187789916992,
      "learning_rate": 3.7062839720684926e-05,
      "loss": 2.6336437225341798,
      "memory(GiB)": 77.56,
      "step": 68075,
      "token_acc": 0.47147147147147145,
      "train_speed(iter/s)": 1.437523
    },
    {
      "epoch": 2.916755923053854,
      "grad_norm": 5.610249996185303,
      "learning_rate": 3.705633923956306e-05,
      "loss": 2.63223876953125,
      "memory(GiB)": 77.56,
      "step": 68080,
      "token_acc": 0.4463667820069204,
      "train_speed(iter/s)": 1.437539
    },
    {
      "epoch": 2.916970138383103,
      "grad_norm": 5.455048561096191,
      "learning_rate": 3.7049838992927755e-05,
      "loss": 2.0385772705078127,
      "memory(GiB)": 77.56,
      "step": 68085,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 2.9171843537123516,
      "grad_norm": 4.9343061447143555,
      "learning_rate": 3.704333898089676e-05,
      "loss": 2.596497917175293,
      "memory(GiB)": 77.56,
      "step": 68090,
      "token_acc": 0.4679245283018868,
      "train_speed(iter/s)": 1.437547
    },
    {
      "epoch": 2.917398569041601,
      "grad_norm": 6.102274417877197,
      "learning_rate": 3.703683920358783e-05,
      "loss": 2.6200613021850585,
      "memory(GiB)": 77.56,
      "step": 68095,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.437527
    },
    {
      "epoch": 2.9176127843708497,
      "grad_norm": 8.687932968139648,
      "learning_rate": 3.7030339661118726e-05,
      "loss": 2.26722412109375,
      "memory(GiB)": 77.56,
      "step": 68100,
      "token_acc": 0.5291828793774319,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 2.9178269997000985,
      "grad_norm": 5.289733409881592,
      "learning_rate": 3.7023840353607184e-05,
      "loss": 2.4978969573974608,
      "memory(GiB)": 77.56,
      "step": 68105,
      "token_acc": 0.43630573248407645,
      "train_speed(iter/s)": 1.437495
    },
    {
      "epoch": 2.9180412150293478,
      "grad_norm": 4.299315929412842,
      "learning_rate": 3.701734128117093e-05,
      "loss": 2.5657699584960936,
      "memory(GiB)": 77.56,
      "step": 68110,
      "token_acc": 0.45674740484429066,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 2.9182554303585966,
      "grad_norm": 4.942446231842041,
      "learning_rate": 3.701084244392773e-05,
      "loss": 2.413589286804199,
      "memory(GiB)": 77.56,
      "step": 68115,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 2.9184696456878454,
      "grad_norm": 5.2760419845581055,
      "learning_rate": 3.70043438419953e-05,
      "loss": 2.2743240356445313,
      "memory(GiB)": 77.56,
      "step": 68120,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.437519
    },
    {
      "epoch": 2.9186838610170946,
      "grad_norm": 6.663489818572998,
      "learning_rate": 3.699784547549139e-05,
      "loss": 2.316331672668457,
      "memory(GiB)": 77.56,
      "step": 68125,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.437535
    },
    {
      "epoch": 2.9188980763463435,
      "grad_norm": 5.129056453704834,
      "learning_rate": 3.699134734453369e-05,
      "loss": 2.6504919052124025,
      "memory(GiB)": 77.56,
      "step": 68130,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.437516
    },
    {
      "epoch": 2.9191122916755923,
      "grad_norm": 4.9390740394592285,
      "learning_rate": 3.698484944923994e-05,
      "loss": 2.1255060195922852,
      "memory(GiB)": 77.56,
      "step": 68135,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.437522
    },
    {
      "epoch": 2.9193265070048415,
      "grad_norm": 6.614055156707764,
      "learning_rate": 3.697835178972785e-05,
      "loss": 2.524114990234375,
      "memory(GiB)": 77.56,
      "step": 68140,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.437514
    },
    {
      "epoch": 2.9195407223340903,
      "grad_norm": 4.805367946624756,
      "learning_rate": 3.6971854366115136e-05,
      "loss": 2.5792205810546873,
      "memory(GiB)": 77.56,
      "step": 68145,
      "token_acc": 0.4520123839009288,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 2.919754937663339,
      "grad_norm": 6.10288143157959,
      "learning_rate": 3.69653571785195e-05,
      "loss": 2.163959503173828,
      "memory(GiB)": 77.56,
      "step": 68150,
      "token_acc": 0.5178571428571429,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 2.9199691529925884,
      "grad_norm": 6.391965866088867,
      "learning_rate": 3.695886022705865e-05,
      "loss": 2.1923961639404297,
      "memory(GiB)": 77.56,
      "step": 68155,
      "token_acc": 0.5447154471544715,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 2.9201833683218372,
      "grad_norm": 5.049295425415039,
      "learning_rate": 3.6952363511850274e-05,
      "loss": 2.498997688293457,
      "memory(GiB)": 77.56,
      "step": 68160,
      "token_acc": 0.44664031620553357,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 2.920397583651086,
      "grad_norm": 5.195949077606201,
      "learning_rate": 3.6945867033012084e-05,
      "loss": 2.214436149597168,
      "memory(GiB)": 77.56,
      "step": 68165,
      "token_acc": 0.5353846153846153,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 2.9206117989803353,
      "grad_norm": 7.036581993103027,
      "learning_rate": 3.6939370790661745e-05,
      "loss": 2.429710578918457,
      "memory(GiB)": 77.56,
      "step": 68170,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.437547
    },
    {
      "epoch": 2.920826014309584,
      "grad_norm": 6.2165446281433105,
      "learning_rate": 3.693287478491696e-05,
      "loss": 2.4742910385131838,
      "memory(GiB)": 77.56,
      "step": 68175,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.437565
    },
    {
      "epoch": 2.921040229638833,
      "grad_norm": 5.181307792663574,
      "learning_rate": 3.6926379015895395e-05,
      "loss": 2.344084548950195,
      "memory(GiB)": 77.56,
      "step": 68180,
      "token_acc": 0.5029585798816568,
      "train_speed(iter/s)": 1.437584
    },
    {
      "epoch": 2.921254444968082,
      "grad_norm": 6.763940811157227,
      "learning_rate": 3.691988348371473e-05,
      "loss": 2.3910369873046875,
      "memory(GiB)": 77.56,
      "step": 68185,
      "token_acc": 0.46122448979591835,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 2.921468660297331,
      "grad_norm": 4.198749542236328,
      "learning_rate": 3.691338818849268e-05,
      "loss": 2.419862174987793,
      "memory(GiB)": 77.56,
      "step": 68190,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.437558
    },
    {
      "epoch": 2.92168287562658,
      "grad_norm": 8.38559341430664,
      "learning_rate": 3.690689313034686e-05,
      "loss": 2.664231872558594,
      "memory(GiB)": 77.56,
      "step": 68195,
      "token_acc": 0.4774193548387097,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 2.921897090955829,
      "grad_norm": 5.795619487762451,
      "learning_rate": 3.690039830939496e-05,
      "loss": 2.1378347396850588,
      "memory(GiB)": 77.56,
      "step": 68200,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.437609
    },
    {
      "epoch": 2.922111306285078,
      "grad_norm": 4.589444160461426,
      "learning_rate": 3.689390372575463e-05,
      "loss": 2.359437370300293,
      "memory(GiB)": 77.56,
      "step": 68205,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.437596
    },
    {
      "epoch": 2.9223255216143267,
      "grad_norm": 7.643991470336914,
      "learning_rate": 3.688740937954354e-05,
      "loss": 2.4506202697753907,
      "memory(GiB)": 77.56,
      "step": 68210,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.437585
    },
    {
      "epoch": 2.922539736943576,
      "grad_norm": 5.98811149597168,
      "learning_rate": 3.6880915270879316e-05,
      "loss": 2.3916858673095702,
      "memory(GiB)": 77.56,
      "step": 68215,
      "token_acc": 0.49514563106796117,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 2.9227539522728248,
      "grad_norm": 4.503275394439697,
      "learning_rate": 3.6874421399879624e-05,
      "loss": 2.5401371002197264,
      "memory(GiB)": 77.56,
      "step": 68220,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 2.9229681676020736,
      "grad_norm": 6.436376571655273,
      "learning_rate": 3.686792776666209e-05,
      "loss": 2.7383283615112304,
      "memory(GiB)": 77.56,
      "step": 68225,
      "token_acc": 0.4463768115942029,
      "train_speed(iter/s)": 1.437628
    },
    {
      "epoch": 2.923182382931323,
      "grad_norm": 4.667707920074463,
      "learning_rate": 3.686143437134438e-05,
      "loss": 2.2121482849121095,
      "memory(GiB)": 77.56,
      "step": 68230,
      "token_acc": 0.504885993485342,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 2.9233965982605716,
      "grad_norm": 4.4454665184021,
      "learning_rate": 3.6854941214044105e-05,
      "loss": 2.272251510620117,
      "memory(GiB)": 77.56,
      "step": 68235,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 2.9236108135898204,
      "grad_norm": 6.854433536529541,
      "learning_rate": 3.68484482948789e-05,
      "loss": 2.2384340286254885,
      "memory(GiB)": 77.56,
      "step": 68240,
      "token_acc": 0.5551470588235294,
      "train_speed(iter/s)": 1.437628
    },
    {
      "epoch": 2.9238250289190697,
      "grad_norm": 5.541005611419678,
      "learning_rate": 3.684195561396639e-05,
      "loss": 2.3282079696655273,
      "memory(GiB)": 77.56,
      "step": 68245,
      "token_acc": 0.5190311418685121,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 2.9240392442483185,
      "grad_norm": 6.127822399139404,
      "learning_rate": 3.6835463171424204e-05,
      "loss": 2.191803550720215,
      "memory(GiB)": 77.56,
      "step": 68250,
      "token_acc": 0.5301204819277109,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 2.9242534595775673,
      "grad_norm": 6.204081058502197,
      "learning_rate": 3.682897096736994e-05,
      "loss": 2.765501022338867,
      "memory(GiB)": 77.56,
      "step": 68255,
      "token_acc": 0.4470899470899471,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 2.9244676749068166,
      "grad_norm": 7.104529857635498,
      "learning_rate": 3.682247900192121e-05,
      "loss": 2.4396389007568358,
      "memory(GiB)": 77.56,
      "step": 68260,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.437617
    },
    {
      "epoch": 2.9246818902360654,
      "grad_norm": 5.439451217651367,
      "learning_rate": 3.6815987275195664e-05,
      "loss": 2.4252492904663088,
      "memory(GiB)": 77.56,
      "step": 68265,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 2.924896105565314,
      "grad_norm": 6.014585018157959,
      "learning_rate": 3.680949578731086e-05,
      "loss": 2.6215152740478516,
      "memory(GiB)": 77.56,
      "step": 68270,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.437616
    },
    {
      "epoch": 2.9251103208945635,
      "grad_norm": 4.763463497161865,
      "learning_rate": 3.6803004538384425e-05,
      "loss": 2.1167179107666017,
      "memory(GiB)": 77.56,
      "step": 68275,
      "token_acc": 0.555956678700361,
      "train_speed(iter/s)": 1.437624
    },
    {
      "epoch": 2.9253245362238123,
      "grad_norm": 5.733205318450928,
      "learning_rate": 3.679651352853394e-05,
      "loss": 2.578509521484375,
      "memory(GiB)": 77.56,
      "step": 68280,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 2.925538751553061,
      "grad_norm": 5.081930160522461,
      "learning_rate": 3.679002275787698e-05,
      "loss": 2.3902057647705077,
      "memory(GiB)": 77.56,
      "step": 68285,
      "token_acc": 0.48055555555555557,
      "train_speed(iter/s)": 1.437637
    },
    {
      "epoch": 2.9257529668823103,
      "grad_norm": 6.3652191162109375,
      "learning_rate": 3.678353222653117e-05,
      "loss": 2.328163719177246,
      "memory(GiB)": 77.56,
      "step": 68290,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 2.925967182211559,
      "grad_norm": 5.1677751541137695,
      "learning_rate": 3.6777041934614076e-05,
      "loss": 2.733831596374512,
      "memory(GiB)": 77.56,
      "step": 68295,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 2.926181397540808,
      "grad_norm": 4.527955055236816,
      "learning_rate": 3.6770551882243256e-05,
      "loss": 2.3957286834716798,
      "memory(GiB)": 77.56,
      "step": 68300,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 2.9263956128700572,
      "grad_norm": 8.41672420501709,
      "learning_rate": 3.676406206953631e-05,
      "loss": 2.4217927932739256,
      "memory(GiB)": 77.56,
      "step": 68305,
      "token_acc": 0.48125,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 2.926609828199306,
      "grad_norm": 5.73418664932251,
      "learning_rate": 3.6757572496610794e-05,
      "loss": 2.4105131149291994,
      "memory(GiB)": 77.56,
      "step": 68310,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 2.926824043528555,
      "grad_norm": 5.170902729034424,
      "learning_rate": 3.675108316358426e-05,
      "loss": 2.554925537109375,
      "memory(GiB)": 77.56,
      "step": 68315,
      "token_acc": 0.43944636678200694,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 2.927038258857804,
      "grad_norm": 6.31304931640625,
      "learning_rate": 3.67445940705743e-05,
      "loss": 2.6427539825439452,
      "memory(GiB)": 77.56,
      "step": 68320,
      "token_acc": 0.46564885496183206,
      "train_speed(iter/s)": 1.437653
    },
    {
      "epoch": 2.927252474187053,
      "grad_norm": 6.32892370223999,
      "learning_rate": 3.673810521769844e-05,
      "loss": 2.6022083282470705,
      "memory(GiB)": 77.56,
      "step": 68325,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.437644
    },
    {
      "epoch": 2.9274666895163017,
      "grad_norm": 6.0275492668151855,
      "learning_rate": 3.673161660507426e-05,
      "loss": 2.533201217651367,
      "memory(GiB)": 77.56,
      "step": 68330,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.437658
    },
    {
      "epoch": 2.927680904845551,
      "grad_norm": 5.365231513977051,
      "learning_rate": 3.672512823281926e-05,
      "loss": 2.4699285507202147,
      "memory(GiB)": 77.56,
      "step": 68335,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.43768
    },
    {
      "epoch": 2.9278951201748,
      "grad_norm": 6.795527935028076,
      "learning_rate": 3.671864010105105e-05,
      "loss": 2.606638717651367,
      "memory(GiB)": 77.56,
      "step": 68340,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.43769
    },
    {
      "epoch": 2.9281093355040486,
      "grad_norm": 11.144953727722168,
      "learning_rate": 3.671215220988711e-05,
      "loss": 2.7022300720214845,
      "memory(GiB)": 77.56,
      "step": 68345,
      "token_acc": 0.4610169491525424,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 2.928323550833298,
      "grad_norm": 5.704421520233154,
      "learning_rate": 3.670566455944502e-05,
      "loss": 2.378597640991211,
      "memory(GiB)": 77.56,
      "step": 68350,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 2.9285377661625467,
      "grad_norm": 5.122502326965332,
      "learning_rate": 3.6699177149842276e-05,
      "loss": 2.5273069381713866,
      "memory(GiB)": 77.56,
      "step": 68355,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.437698
    },
    {
      "epoch": 2.9287519814917955,
      "grad_norm": 5.294159412384033,
      "learning_rate": 3.669268998119642e-05,
      "loss": 2.3849491119384765,
      "memory(GiB)": 77.56,
      "step": 68360,
      "token_acc": 0.4830188679245283,
      "train_speed(iter/s)": 1.437716
    },
    {
      "epoch": 2.9289661968210448,
      "grad_norm": 7.5218634605407715,
      "learning_rate": 3.668620305362496e-05,
      "loss": 2.4301084518432616,
      "memory(GiB)": 77.56,
      "step": 68365,
      "token_acc": 0.4981684981684982,
      "train_speed(iter/s)": 1.437704
    },
    {
      "epoch": 2.9291804121502936,
      "grad_norm": 5.494652271270752,
      "learning_rate": 3.667971636724542e-05,
      "loss": 2.234860420227051,
      "memory(GiB)": 77.56,
      "step": 68370,
      "token_acc": 0.5347985347985348,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 2.9293946274795424,
      "grad_norm": 5.734588623046875,
      "learning_rate": 3.667322992217532e-05,
      "loss": 2.5520034790039063,
      "memory(GiB)": 77.56,
      "step": 68375,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 2.9296088428087916,
      "grad_norm": 5.948920249938965,
      "learning_rate": 3.666674371853217e-05,
      "loss": 2.188946533203125,
      "memory(GiB)": 77.56,
      "step": 68380,
      "token_acc": 0.5098814229249012,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 2.9298230581380404,
      "grad_norm": 6.205685615539551,
      "learning_rate": 3.6660257756433444e-05,
      "loss": 2.5464221954345705,
      "memory(GiB)": 77.56,
      "step": 68385,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 2.9300372734672893,
      "grad_norm": 4.757958889007568,
      "learning_rate": 3.665377203599668e-05,
      "loss": 2.375376892089844,
      "memory(GiB)": 77.56,
      "step": 68390,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 2.9302514887965385,
      "grad_norm": 5.090051174163818,
      "learning_rate": 3.664728655733936e-05,
      "loss": 2.354806900024414,
      "memory(GiB)": 77.56,
      "step": 68395,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 2.9304657041257873,
      "grad_norm": 5.366011619567871,
      "learning_rate": 3.664080132057896e-05,
      "loss": 2.7003589630126954,
      "memory(GiB)": 77.56,
      "step": 68400,
      "token_acc": 0.40942028985507245,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 2.930679919455036,
      "grad_norm": 5.187288761138916,
      "learning_rate": 3.6634316325832954e-05,
      "loss": 2.4137845993041993,
      "memory(GiB)": 77.56,
      "step": 68405,
      "token_acc": 0.48265895953757226,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 2.9308941347842854,
      "grad_norm": 4.5555739402771,
      "learning_rate": 3.662783157321887e-05,
      "loss": 2.3343963623046875,
      "memory(GiB)": 77.56,
      "step": 68410,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 2.931108350113534,
      "grad_norm": 4.508800506591797,
      "learning_rate": 3.6621347062854164e-05,
      "loss": 2.3369203567504884,
      "memory(GiB)": 77.56,
      "step": 68415,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.437671
    },
    {
      "epoch": 2.931322565442783,
      "grad_norm": 5.752854824066162,
      "learning_rate": 3.661486279485629e-05,
      "loss": 2.4629842758178713,
      "memory(GiB)": 77.56,
      "step": 68420,
      "token_acc": 0.45555555555555555,
      "train_speed(iter/s)": 1.437704
    },
    {
      "epoch": 2.9315367807720323,
      "grad_norm": 4.406851291656494,
      "learning_rate": 3.6608378769342746e-05,
      "loss": 2.525054168701172,
      "memory(GiB)": 77.56,
      "step": 68425,
      "token_acc": 0.4732142857142857,
      "train_speed(iter/s)": 1.437679
    },
    {
      "epoch": 2.931750996101281,
      "grad_norm": 5.786430835723877,
      "learning_rate": 3.660189498643097e-05,
      "loss": 2.5697994232177734,
      "memory(GiB)": 77.56,
      "step": 68430,
      "token_acc": 0.5028901734104047,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 2.93196521143053,
      "grad_norm": 4.407654285430908,
      "learning_rate": 3.659541144623846e-05,
      "loss": 2.5435739517211915,
      "memory(GiB)": 77.56,
      "step": 68435,
      "token_acc": 0.46688741721854304,
      "train_speed(iter/s)": 1.437682
    },
    {
      "epoch": 2.932179426759779,
      "grad_norm": 6.907027721405029,
      "learning_rate": 3.658892814888264e-05,
      "loss": 2.5674554824829103,
      "memory(GiB)": 77.56,
      "step": 68440,
      "token_acc": 0.423841059602649,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 2.932393642089028,
      "grad_norm": 5.145559787750244,
      "learning_rate": 3.658244509448094e-05,
      "loss": 2.2846824645996096,
      "memory(GiB)": 77.56,
      "step": 68445,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 2.932607857418277,
      "grad_norm": 4.6332926750183105,
      "learning_rate": 3.657596228315086e-05,
      "loss": 2.3373672485351564,
      "memory(GiB)": 77.56,
      "step": 68450,
      "token_acc": 0.5096774193548387,
      "train_speed(iter/s)": 1.437669
    },
    {
      "epoch": 2.932822072747526,
      "grad_norm": 4.5164361000061035,
      "learning_rate": 3.656947971500982e-05,
      "loss": 2.3477684020996095,
      "memory(GiB)": 77.56,
      "step": 68455,
      "token_acc": 0.49429657794676807,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 2.933036288076775,
      "grad_norm": 6.285327434539795,
      "learning_rate": 3.656299739017523e-05,
      "loss": 2.2973634719848635,
      "memory(GiB)": 77.56,
      "step": 68460,
      "token_acc": 0.5175718849840255,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 2.9332505034060237,
      "grad_norm": 5.199221134185791,
      "learning_rate": 3.655651530876456e-05,
      "loss": 2.5503076553344726,
      "memory(GiB)": 77.56,
      "step": 68465,
      "token_acc": 0.44483985765124556,
      "train_speed(iter/s)": 1.437687
    },
    {
      "epoch": 2.933464718735273,
      "grad_norm": 8.812651634216309,
      "learning_rate": 3.6550033470895225e-05,
      "loss": 2.6162412643432615,
      "memory(GiB)": 77.56,
      "step": 68470,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 2.9336789340645217,
      "grad_norm": 4.474532127380371,
      "learning_rate": 3.6543551876684656e-05,
      "loss": 2.5346126556396484,
      "memory(GiB)": 77.56,
      "step": 68475,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 2.9338931493937705,
      "grad_norm": 5.679563999176025,
      "learning_rate": 3.6537070526250244e-05,
      "loss": 2.268147850036621,
      "memory(GiB)": 77.56,
      "step": 68480,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.437737
    },
    {
      "epoch": 2.93410736472302,
      "grad_norm": 6.245189189910889,
      "learning_rate": 3.653058941970945e-05,
      "loss": 2.347981262207031,
      "memory(GiB)": 77.56,
      "step": 68485,
      "token_acc": 0.49224806201550386,
      "train_speed(iter/s)": 1.437738
    },
    {
      "epoch": 2.9343215800522686,
      "grad_norm": 5.599194526672363,
      "learning_rate": 3.652410855717966e-05,
      "loss": 2.276208686828613,
      "memory(GiB)": 77.56,
      "step": 68490,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 2.9345357953815174,
      "grad_norm": 4.174640655517578,
      "learning_rate": 3.651762793877829e-05,
      "loss": 2.4413013458251953,
      "memory(GiB)": 77.56,
      "step": 68495,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 2.9347500107107667,
      "grad_norm": 7.646904468536377,
      "learning_rate": 3.6511147564622736e-05,
      "loss": 2.4015220642089843,
      "memory(GiB)": 77.56,
      "step": 68500,
      "token_acc": 0.5062111801242236,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 2.9347500107107667,
      "eval_loss": 2.31423282623291,
      "eval_runtime": 13.8976,
      "eval_samples_per_second": 7.196,
      "eval_steps_per_second": 7.196,
      "eval_token_acc": 0.43351800554016623,
      "step": 68500
    },
    {
      "epoch": 2.9349642260400155,
      "grad_norm": 4.358438968658447,
      "learning_rate": 3.6504667434830394e-05,
      "loss": 2.3032392501831054,
      "memory(GiB)": 77.56,
      "step": 68505,
      "token_acc": 0.44831013916500995,
      "train_speed(iter/s)": 1.437311
    },
    {
      "epoch": 2.9351784413692643,
      "grad_norm": 8.29546070098877,
      "learning_rate": 3.649818754951866e-05,
      "loss": 2.151364326477051,
      "memory(GiB)": 77.56,
      "step": 68510,
      "token_acc": 0.5331325301204819,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 2.9353926566985136,
      "grad_norm": 5.173444747924805,
      "learning_rate": 3.6491707908804926e-05,
      "loss": 2.5007471084594726,
      "memory(GiB)": 77.56,
      "step": 68515,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 2.9356068720277624,
      "grad_norm": 6.332651615142822,
      "learning_rate": 3.648522851280658e-05,
      "loss": 2.943612289428711,
      "memory(GiB)": 77.56,
      "step": 68520,
      "token_acc": 0.3857677902621723,
      "train_speed(iter/s)": 1.437325
    },
    {
      "epoch": 2.935821087357011,
      "grad_norm": 5.154989719390869,
      "learning_rate": 3.6478749361640996e-05,
      "loss": 2.495052719116211,
      "memory(GiB)": 77.56,
      "step": 68525,
      "token_acc": 0.4766355140186916,
      "train_speed(iter/s)": 1.437322
    },
    {
      "epoch": 2.9360353026862605,
      "grad_norm": 4.607423305511475,
      "learning_rate": 3.647227045542554e-05,
      "loss": 2.5632247924804688,
      "memory(GiB)": 77.56,
      "step": 68530,
      "token_acc": 0.44947735191637633,
      "train_speed(iter/s)": 1.437314
    },
    {
      "epoch": 2.9362495180155093,
      "grad_norm": 5.022364139556885,
      "learning_rate": 3.646579179427761e-05,
      "loss": 1.8762619018554687,
      "memory(GiB)": 77.56,
      "step": 68535,
      "token_acc": 0.5580357142857143,
      "train_speed(iter/s)": 1.437327
    },
    {
      "epoch": 2.936463733344758,
      "grad_norm": 6.142657279968262,
      "learning_rate": 3.6459313378314554e-05,
      "loss": 2.215517044067383,
      "memory(GiB)": 77.56,
      "step": 68540,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.437319
    },
    {
      "epoch": 2.9366779486740073,
      "grad_norm": 4.5446577072143555,
      "learning_rate": 3.645283520765373e-05,
      "loss": 2.1593366622924806,
      "memory(GiB)": 77.56,
      "step": 68545,
      "token_acc": 0.5222929936305732,
      "train_speed(iter/s)": 1.437321
    },
    {
      "epoch": 2.936892164003256,
      "grad_norm": 4.72842264175415,
      "learning_rate": 3.6446357282412514e-05,
      "loss": 2.3668373107910154,
      "memory(GiB)": 77.56,
      "step": 68550,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.437326
    },
    {
      "epoch": 2.937106379332505,
      "grad_norm": 6.1560468673706055,
      "learning_rate": 3.6439879602708224e-05,
      "loss": 2.593875503540039,
      "memory(GiB)": 77.56,
      "step": 68555,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.437323
    },
    {
      "epoch": 2.937320594661754,
      "grad_norm": 4.385132789611816,
      "learning_rate": 3.643340216865827e-05,
      "loss": 2.6797506332397463,
      "memory(GiB)": 77.56,
      "step": 68560,
      "token_acc": 0.45217391304347826,
      "train_speed(iter/s)": 1.437295
    },
    {
      "epoch": 2.937534809991003,
      "grad_norm": 5.3138532638549805,
      "learning_rate": 3.642692498037994e-05,
      "loss": 2.421952819824219,
      "memory(GiB)": 77.56,
      "step": 68565,
      "token_acc": 0.4765625,
      "train_speed(iter/s)": 1.437323
    },
    {
      "epoch": 2.937749025320252,
      "grad_norm": 4.772027492523193,
      "learning_rate": 3.642044803799061e-05,
      "loss": 1.9963918685913087,
      "memory(GiB)": 77.56,
      "step": 68570,
      "token_acc": 0.5772357723577236,
      "train_speed(iter/s)": 1.437291
    },
    {
      "epoch": 2.937963240649501,
      "grad_norm": 6.275188446044922,
      "learning_rate": 3.6413971341607596e-05,
      "loss": 2.5237409591674806,
      "memory(GiB)": 77.56,
      "step": 68575,
      "token_acc": 0.4620253164556962,
      "train_speed(iter/s)": 1.437274
    },
    {
      "epoch": 2.93817745597875,
      "grad_norm": 5.677504062652588,
      "learning_rate": 3.6407494891348244e-05,
      "loss": 2.542437934875488,
      "memory(GiB)": 77.56,
      "step": 68580,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.437288
    },
    {
      "epoch": 2.9383916713079987,
      "grad_norm": 11.629992485046387,
      "learning_rate": 3.640101868732987e-05,
      "loss": 2.2602775573730467,
      "memory(GiB)": 77.56,
      "step": 68585,
      "token_acc": 0.48338368580060426,
      "train_speed(iter/s)": 1.437279
    },
    {
      "epoch": 2.938605886637248,
      "grad_norm": 5.792952060699463,
      "learning_rate": 3.639454272966979e-05,
      "loss": 2.2910930633544924,
      "memory(GiB)": 77.56,
      "step": 68590,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.437276
    },
    {
      "epoch": 2.938820101966497,
      "grad_norm": 7.0362229347229,
      "learning_rate": 3.6388067018485337e-05,
      "loss": 2.2122020721435547,
      "memory(GiB)": 77.56,
      "step": 68595,
      "token_acc": 0.5445205479452054,
      "train_speed(iter/s)": 1.437301
    },
    {
      "epoch": 2.9390343172957456,
      "grad_norm": 4.994625091552734,
      "learning_rate": 3.638159155389383e-05,
      "loss": 2.540998077392578,
      "memory(GiB)": 77.56,
      "step": 68600,
      "token_acc": 0.449468085106383,
      "train_speed(iter/s)": 1.437284
    },
    {
      "epoch": 2.939248532624995,
      "grad_norm": 6.120875835418701,
      "learning_rate": 3.637511633601255e-05,
      "loss": 2.559706687927246,
      "memory(GiB)": 77.56,
      "step": 68605,
      "token_acc": 0.4485294117647059,
      "train_speed(iter/s)": 1.437288
    },
    {
      "epoch": 2.9394627479542437,
      "grad_norm": 4.160299301147461,
      "learning_rate": 3.636864136495883e-05,
      "loss": 2.440188980102539,
      "memory(GiB)": 77.56,
      "step": 68610,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437307
    },
    {
      "epoch": 2.9396769632834925,
      "grad_norm": 5.6210479736328125,
      "learning_rate": 3.636216664084995e-05,
      "loss": 1.8378814697265624,
      "memory(GiB)": 77.56,
      "step": 68615,
      "token_acc": 0.5829787234042553,
      "train_speed(iter/s)": 1.437287
    },
    {
      "epoch": 2.9398911786127417,
      "grad_norm": 7.1883955001831055,
      "learning_rate": 3.635569216380322e-05,
      "loss": 2.498285675048828,
      "memory(GiB)": 77.56,
      "step": 68620,
      "token_acc": 0.45733788395904434,
      "train_speed(iter/s)": 1.437311
    },
    {
      "epoch": 2.9401053939419906,
      "grad_norm": 8.345105171203613,
      "learning_rate": 3.634921793393593e-05,
      "loss": 2.493545913696289,
      "memory(GiB)": 77.56,
      "step": 68625,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.437331
    },
    {
      "epoch": 2.9403196092712394,
      "grad_norm": 5.7748122215271,
      "learning_rate": 3.634274395136534e-05,
      "loss": 2.359173583984375,
      "memory(GiB)": 77.56,
      "step": 68630,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.437338
    },
    {
      "epoch": 2.9405338246004886,
      "grad_norm": 6.156806945800781,
      "learning_rate": 3.633627021620877e-05,
      "loss": 2.3151174545288087,
      "memory(GiB)": 77.56,
      "step": 68635,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.437343
    },
    {
      "epoch": 2.9407480399297374,
      "grad_norm": 5.841317176818848,
      "learning_rate": 3.632979672858349e-05,
      "loss": 2.6829769134521486,
      "memory(GiB)": 77.56,
      "step": 68640,
      "token_acc": 0.42066420664206644,
      "train_speed(iter/s)": 1.437343
    },
    {
      "epoch": 2.9409622552589862,
      "grad_norm": 4.903331279754639,
      "learning_rate": 3.632332348860676e-05,
      "loss": 2.2645505905151366,
      "memory(GiB)": 77.56,
      "step": 68645,
      "token_acc": 0.5377358490566038,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 2.9411764705882355,
      "grad_norm": 5.157782077789307,
      "learning_rate": 3.631685049639586e-05,
      "loss": 2.596482849121094,
      "memory(GiB)": 77.56,
      "step": 68650,
      "token_acc": 0.4559386973180077,
      "train_speed(iter/s)": 1.437354
    },
    {
      "epoch": 2.9413906859174843,
      "grad_norm": 6.980854511260986,
      "learning_rate": 3.6310377752068046e-05,
      "loss": 2.601549530029297,
      "memory(GiB)": 77.56,
      "step": 68655,
      "token_acc": 0.5052631578947369,
      "train_speed(iter/s)": 1.437344
    },
    {
      "epoch": 2.941604901246733,
      "grad_norm": 4.489633083343506,
      "learning_rate": 3.6303905255740575e-05,
      "loss": 2.5341041564941404,
      "memory(GiB)": 77.56,
      "step": 68660,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.437351
    },
    {
      "epoch": 2.9418191165759824,
      "grad_norm": 6.645514011383057,
      "learning_rate": 3.629743300753072e-05,
      "loss": 2.7830312728881834,
      "memory(GiB)": 77.56,
      "step": 68665,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 2.942033331905231,
      "grad_norm": 6.095151424407959,
      "learning_rate": 3.629096100755571e-05,
      "loss": 2.603617858886719,
      "memory(GiB)": 77.56,
      "step": 68670,
      "token_acc": 0.4490861618798956,
      "train_speed(iter/s)": 1.437301
    },
    {
      "epoch": 2.94224754723448,
      "grad_norm": 5.025662422180176,
      "learning_rate": 3.6284489255932805e-05,
      "loss": 2.5306350708007814,
      "memory(GiB)": 77.56,
      "step": 68675,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437272
    },
    {
      "epoch": 2.9424617625637293,
      "grad_norm": 7.614643573760986,
      "learning_rate": 3.627801775277925e-05,
      "loss": 2.505751037597656,
      "memory(GiB)": 77.56,
      "step": 68680,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.437283
    },
    {
      "epoch": 2.942675977892978,
      "grad_norm": 6.706848621368408,
      "learning_rate": 3.627154649821227e-05,
      "loss": 2.1313541412353514,
      "memory(GiB)": 77.56,
      "step": 68685,
      "token_acc": 0.5437262357414449,
      "train_speed(iter/s)": 1.43727
    },
    {
      "epoch": 2.942890193222227,
      "grad_norm": 6.144601821899414,
      "learning_rate": 3.62650754923491e-05,
      "loss": 2.2737705230712892,
      "memory(GiB)": 77.56,
      "step": 68690,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.437282
    },
    {
      "epoch": 2.943104408551476,
      "grad_norm": 6.145516395568848,
      "learning_rate": 3.625860473530698e-05,
      "loss": 2.3433048248291017,
      "memory(GiB)": 77.56,
      "step": 68695,
      "token_acc": 0.5111940298507462,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 2.943318623880725,
      "grad_norm": 6.162186622619629,
      "learning_rate": 3.62521342272031e-05,
      "loss": 2.4569093704223635,
      "memory(GiB)": 77.56,
      "step": 68700,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.437326
    },
    {
      "epoch": 2.9435328392099738,
      "grad_norm": 7.319283962249756,
      "learning_rate": 3.624566396815473e-05,
      "loss": 2.6335189819335936,
      "memory(GiB)": 77.56,
      "step": 68705,
      "token_acc": 0.4266666666666667,
      "train_speed(iter/s)": 1.437338
    },
    {
      "epoch": 2.943747054539223,
      "grad_norm": 4.756680488586426,
      "learning_rate": 3.623919395827905e-05,
      "loss": 2.4158279418945314,
      "memory(GiB)": 77.56,
      "step": 68710,
      "token_acc": 0.4981132075471698,
      "train_speed(iter/s)": 1.437301
    },
    {
      "epoch": 2.943961269868472,
      "grad_norm": 6.41449499130249,
      "learning_rate": 3.623272419769329e-05,
      "loss": 2.5319637298583983,
      "memory(GiB)": 77.56,
      "step": 68715,
      "token_acc": 0.47509578544061304,
      "train_speed(iter/s)": 1.437305
    },
    {
      "epoch": 2.9441754851977207,
      "grad_norm": 5.517734050750732,
      "learning_rate": 3.622625468651463e-05,
      "loss": 2.489777755737305,
      "memory(GiB)": 77.56,
      "step": 68720,
      "token_acc": 0.46956521739130436,
      "train_speed(iter/s)": 1.43732
    },
    {
      "epoch": 2.94438970052697,
      "grad_norm": 6.091075420379639,
      "learning_rate": 3.62197854248603e-05,
      "loss": 2.700113868713379,
      "memory(GiB)": 77.56,
      "step": 68725,
      "token_acc": 0.468503937007874,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 2.9446039158562187,
      "grad_norm": 5.145681381225586,
      "learning_rate": 3.621331641284749e-05,
      "loss": 2.181728172302246,
      "memory(GiB)": 77.56,
      "step": 68730,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.437306
    },
    {
      "epoch": 2.9448181311854675,
      "grad_norm": 4.653256893157959,
      "learning_rate": 3.620684765059337e-05,
      "loss": 2.333662223815918,
      "memory(GiB)": 77.56,
      "step": 68735,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.437334
    },
    {
      "epoch": 2.945032346514717,
      "grad_norm": 4.413647174835205,
      "learning_rate": 3.620037913821516e-05,
      "loss": 2.6249488830566405,
      "memory(GiB)": 77.56,
      "step": 68740,
      "token_acc": 0.43086816720257237,
      "train_speed(iter/s)": 1.437344
    },
    {
      "epoch": 2.9452465618439656,
      "grad_norm": 6.1817522048950195,
      "learning_rate": 3.619391087583002e-05,
      "loss": 2.759774398803711,
      "memory(GiB)": 77.56,
      "step": 68745,
      "token_acc": 0.4582043343653251,
      "train_speed(iter/s)": 1.437354
    },
    {
      "epoch": 2.9454607771732144,
      "grad_norm": 5.204505443572998,
      "learning_rate": 3.618744286355513e-05,
      "loss": 2.450588607788086,
      "memory(GiB)": 77.56,
      "step": 68750,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.437361
    },
    {
      "epoch": 2.9456749925024637,
      "grad_norm": 5.7504191398620605,
      "learning_rate": 3.618097510150768e-05,
      "loss": 2.284286880493164,
      "memory(GiB)": 77.56,
      "step": 68755,
      "token_acc": 0.5269709543568465,
      "train_speed(iter/s)": 1.437371
    },
    {
      "epoch": 2.9458892078317125,
      "grad_norm": 5.748534202575684,
      "learning_rate": 3.617450758980482e-05,
      "loss": 2.293759346008301,
      "memory(GiB)": 77.56,
      "step": 68760,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 2.9461034231609613,
      "grad_norm": 4.761528968811035,
      "learning_rate": 3.616804032856373e-05,
      "loss": 2.5627391815185545,
      "memory(GiB)": 77.56,
      "step": 68765,
      "token_acc": 0.4555160142348754,
      "train_speed(iter/s)": 1.437332
    },
    {
      "epoch": 2.9463176384902106,
      "grad_norm": 5.485678672790527,
      "learning_rate": 3.6161573317901564e-05,
      "loss": 2.6419851303100588,
      "memory(GiB)": 77.56,
      "step": 68770,
      "token_acc": 0.4701492537313433,
      "train_speed(iter/s)": 1.437322
    },
    {
      "epoch": 2.9465318538194594,
      "grad_norm": 5.006595134735107,
      "learning_rate": 3.6155106557935464e-05,
      "loss": 2.4364442825317383,
      "memory(GiB)": 77.56,
      "step": 68775,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.43733
    },
    {
      "epoch": 2.946746069148708,
      "grad_norm": 4.9011921882629395,
      "learning_rate": 3.6148640048782604e-05,
      "loss": 2.335931587219238,
      "memory(GiB)": 77.56,
      "step": 68780,
      "token_acc": 0.5511111111111111,
      "train_speed(iter/s)": 1.437357
    },
    {
      "epoch": 2.9469602844779574,
      "grad_norm": 6.965666770935059,
      "learning_rate": 3.614217379056013e-05,
      "loss": 2.2329822540283204,
      "memory(GiB)": 77.56,
      "step": 68785,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 2.9471744998072062,
      "grad_norm": 4.601630210876465,
      "learning_rate": 3.6135707783385183e-05,
      "loss": 2.626714324951172,
      "memory(GiB)": 77.56,
      "step": 68790,
      "token_acc": 0.4502923976608187,
      "train_speed(iter/s)": 1.43738
    },
    {
      "epoch": 2.947388715136455,
      "grad_norm": 4.9763264656066895,
      "learning_rate": 3.6129242027374886e-05,
      "loss": 2.3513513565063477,
      "memory(GiB)": 77.56,
      "step": 68795,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.437387
    },
    {
      "epoch": 2.9476029304657043,
      "grad_norm": 4.851491451263428,
      "learning_rate": 3.6122776522646386e-05,
      "loss": 2.5669939041137697,
      "memory(GiB)": 77.56,
      "step": 68800,
      "token_acc": 0.4562334217506631,
      "train_speed(iter/s)": 1.437387
    },
    {
      "epoch": 2.947817145794953,
      "grad_norm": 4.193472385406494,
      "learning_rate": 3.6116311269316804e-05,
      "loss": 2.3566879272460937,
      "memory(GiB)": 77.56,
      "step": 68805,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437362
    },
    {
      "epoch": 2.948031361124202,
      "grad_norm": 5.2125678062438965,
      "learning_rate": 3.610984626750327e-05,
      "loss": 2.473443031311035,
      "memory(GiB)": 77.56,
      "step": 68810,
      "token_acc": 0.4684385382059801,
      "train_speed(iter/s)": 1.437374
    },
    {
      "epoch": 2.948245576453451,
      "grad_norm": 4.731603145599365,
      "learning_rate": 3.6103381517322905e-05,
      "loss": 2.3626129150390627,
      "memory(GiB)": 77.56,
      "step": 68815,
      "token_acc": 0.516245487364621,
      "train_speed(iter/s)": 1.437373
    },
    {
      "epoch": 2.9484597917827,
      "grad_norm": 5.903083801269531,
      "learning_rate": 3.609691701889281e-05,
      "loss": 2.4182340621948244,
      "memory(GiB)": 77.56,
      "step": 68820,
      "token_acc": 0.5033333333333333,
      "train_speed(iter/s)": 1.437374
    },
    {
      "epoch": 2.948674007111949,
      "grad_norm": 6.340105056762695,
      "learning_rate": 3.6090452772330115e-05,
      "loss": 2.312301254272461,
      "memory(GiB)": 77.56,
      "step": 68825,
      "token_acc": 0.4740061162079511,
      "train_speed(iter/s)": 1.437383
    },
    {
      "epoch": 2.948888222441198,
      "grad_norm": 7.80027961730957,
      "learning_rate": 3.6083988777751916e-05,
      "loss": 2.6433765411376955,
      "memory(GiB)": 77.56,
      "step": 68830,
      "token_acc": 0.4303030303030303,
      "train_speed(iter/s)": 1.437408
    },
    {
      "epoch": 2.949102437770447,
      "grad_norm": 5.485929012298584,
      "learning_rate": 3.60775250352753e-05,
      "loss": 2.530671310424805,
      "memory(GiB)": 77.56,
      "step": 68835,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 2.9493166530996957,
      "grad_norm": 4.73348331451416,
      "learning_rate": 3.6071061545017395e-05,
      "loss": 2.529583549499512,
      "memory(GiB)": 77.56,
      "step": 68840,
      "token_acc": 0.4797507788161994,
      "train_speed(iter/s)": 1.437418
    },
    {
      "epoch": 2.949530868428945,
      "grad_norm": 4.363133907318115,
      "learning_rate": 3.6064598307095266e-05,
      "loss": 2.4248477935791017,
      "memory(GiB)": 77.56,
      "step": 68845,
      "token_acc": 0.4740740740740741,
      "train_speed(iter/s)": 1.437426
    },
    {
      "epoch": 2.9497450837581938,
      "grad_norm": 5.332622528076172,
      "learning_rate": 3.6058135321625995e-05,
      "loss": 2.3236104965209963,
      "memory(GiB)": 77.56,
      "step": 68850,
      "token_acc": 0.535483870967742,
      "train_speed(iter/s)": 1.437437
    },
    {
      "epoch": 2.9499592990874426,
      "grad_norm": 6.431833267211914,
      "learning_rate": 3.60516725887267e-05,
      "loss": 2.568436622619629,
      "memory(GiB)": 77.56,
      "step": 68855,
      "token_acc": 0.48132780082987553,
      "train_speed(iter/s)": 1.437467
    },
    {
      "epoch": 2.950173514416692,
      "grad_norm": 4.678027153015137,
      "learning_rate": 3.604521010851445e-05,
      "loss": 2.5360055923461915,
      "memory(GiB)": 77.56,
      "step": 68860,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.437464
    },
    {
      "epoch": 2.9503877297459407,
      "grad_norm": 5.996098518371582,
      "learning_rate": 3.60387478811063e-05,
      "loss": 2.6224271774291994,
      "memory(GiB)": 77.56,
      "step": 68865,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 2.9506019450751895,
      "grad_norm": 4.644439220428467,
      "learning_rate": 3.603228590661933e-05,
      "loss": 2.233340835571289,
      "memory(GiB)": 77.56,
      "step": 68870,
      "token_acc": 0.4781021897810219,
      "train_speed(iter/s)": 1.437488
    },
    {
      "epoch": 2.9508161604044387,
      "grad_norm": 6.249563694000244,
      "learning_rate": 3.602582418517061e-05,
      "loss": 2.5900638580322264,
      "memory(GiB)": 77.56,
      "step": 68875,
      "token_acc": 0.4542372881355932,
      "train_speed(iter/s)": 1.437449
    },
    {
      "epoch": 2.9510303757336875,
      "grad_norm": 4.71757173538208,
      "learning_rate": 3.601936271687718e-05,
      "loss": 2.4595561981201173,
      "memory(GiB)": 77.56,
      "step": 68880,
      "token_acc": 0.4738562091503268,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 2.9512445910629364,
      "grad_norm": 6.177444934844971,
      "learning_rate": 3.601290150185612e-05,
      "loss": 2.3477642059326174,
      "memory(GiB)": 77.56,
      "step": 68885,
      "token_acc": 0.5405405405405406,
      "train_speed(iter/s)": 1.437483
    },
    {
      "epoch": 2.9514588063921856,
      "grad_norm": 4.402703285217285,
      "learning_rate": 3.6006440540224465e-05,
      "loss": 2.4873001098632814,
      "memory(GiB)": 77.56,
      "step": 68890,
      "token_acc": 0.46174142480211083,
      "train_speed(iter/s)": 1.437488
    },
    {
      "epoch": 2.9516730217214344,
      "grad_norm": 5.3722429275512695,
      "learning_rate": 3.599997983209927e-05,
      "loss": 2.5535905838012694,
      "memory(GiB)": 77.56,
      "step": 68895,
      "token_acc": 0.44970414201183434,
      "train_speed(iter/s)": 1.437492
    },
    {
      "epoch": 2.9518872370506832,
      "grad_norm": 5.824199676513672,
      "learning_rate": 3.5993519377597576e-05,
      "loss": 2.6637346267700197,
      "memory(GiB)": 77.56,
      "step": 68900,
      "token_acc": 0.46440677966101696,
      "train_speed(iter/s)": 1.437501
    },
    {
      "epoch": 2.9521014523799325,
      "grad_norm": 6.875514984130859,
      "learning_rate": 3.5987059176836404e-05,
      "loss": 2.3171222686767576,
      "memory(GiB)": 77.56,
      "step": 68905,
      "token_acc": 0.45075757575757575,
      "train_speed(iter/s)": 1.437527
    },
    {
      "epoch": 2.9523156677091813,
      "grad_norm": 4.838869571685791,
      "learning_rate": 3.598059922993282e-05,
      "loss": 2.1232067108154298,
      "memory(GiB)": 77.56,
      "step": 68910,
      "token_acc": 0.5259515570934256,
      "train_speed(iter/s)": 1.437535
    },
    {
      "epoch": 2.95252988303843,
      "grad_norm": 5.742547988891602,
      "learning_rate": 3.597413953700382e-05,
      "loss": 2.31732234954834,
      "memory(GiB)": 77.56,
      "step": 68915,
      "token_acc": 0.5160142348754448,
      "train_speed(iter/s)": 1.437535
    },
    {
      "epoch": 2.9527440983676794,
      "grad_norm": 6.010509967803955,
      "learning_rate": 3.596768009816644e-05,
      "loss": 2.3423095703125,
      "memory(GiB)": 77.56,
      "step": 68920,
      "token_acc": 0.5036231884057971,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 2.952958313696928,
      "grad_norm": 5.6316728591918945,
      "learning_rate": 3.5961220913537683e-05,
      "loss": 2.213157844543457,
      "memory(GiB)": 77.56,
      "step": 68925,
      "token_acc": 0.5387596899224806,
      "train_speed(iter/s)": 1.437545
    },
    {
      "epoch": 2.953172529026177,
      "grad_norm": 4.8132734298706055,
      "learning_rate": 3.5954761983234595e-05,
      "loss": 2.708344268798828,
      "memory(GiB)": 77.56,
      "step": 68930,
      "token_acc": 0.48623853211009177,
      "train_speed(iter/s)": 1.437547
    },
    {
      "epoch": 2.9533867443554263,
      "grad_norm": 5.64206600189209,
      "learning_rate": 3.594830330737417e-05,
      "loss": 2.413673210144043,
      "memory(GiB)": 77.56,
      "step": 68935,
      "token_acc": 0.4651898734177215,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 2.953600959684675,
      "grad_norm": 4.685112476348877,
      "learning_rate": 3.5941844886073416e-05,
      "loss": 2.428424263000488,
      "memory(GiB)": 77.56,
      "step": 68940,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 2.953815175013924,
      "grad_norm": 4.130372047424316,
      "learning_rate": 3.5935386719449324e-05,
      "loss": 2.4846450805664064,
      "memory(GiB)": 77.56,
      "step": 68945,
      "token_acc": 0.5058365758754864,
      "train_speed(iter/s)": 1.4376
    },
    {
      "epoch": 2.954029390343173,
      "grad_norm": 4.273837089538574,
      "learning_rate": 3.5928928807618896e-05,
      "loss": 2.602483367919922,
      "memory(GiB)": 77.56,
      "step": 68950,
      "token_acc": 0.472636815920398,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 2.954243605672422,
      "grad_norm": 4.532648086547852,
      "learning_rate": 3.592247115069913e-05,
      "loss": 2.5536243438720705,
      "memory(GiB)": 77.56,
      "step": 68955,
      "token_acc": 0.46503496503496505,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 2.9544578210016708,
      "grad_norm": 5.601273536682129,
      "learning_rate": 3.5916013748807e-05,
      "loss": 2.231674575805664,
      "memory(GiB)": 77.56,
      "step": 68960,
      "token_acc": 0.49523809523809526,
      "train_speed(iter/s)": 1.437674
    },
    {
      "epoch": 2.95467203633092,
      "grad_norm": 7.24081563949585,
      "learning_rate": 3.590955660205948e-05,
      "loss": 2.558842086791992,
      "memory(GiB)": 77.56,
      "step": 68965,
      "token_acc": 0.46938775510204084,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 2.954886251660169,
      "grad_norm": 4.85072660446167,
      "learning_rate": 3.590309971057358e-05,
      "loss": 2.4630973815917967,
      "memory(GiB)": 77.56,
      "step": 68970,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 2.9551004669894176,
      "grad_norm": 6.246769905090332,
      "learning_rate": 3.5896643074466246e-05,
      "loss": 2.334185028076172,
      "memory(GiB)": 77.56,
      "step": 68975,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437679
    },
    {
      "epoch": 2.955314682318667,
      "grad_norm": 6.290075302124023,
      "learning_rate": 3.5890186693854444e-05,
      "loss": 2.1810699462890626,
      "memory(GiB)": 77.56,
      "step": 68980,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437682
    },
    {
      "epoch": 2.9555288976479157,
      "grad_norm": 4.407094955444336,
      "learning_rate": 3.5883730568855156e-05,
      "loss": 2.5275590896606444,
      "memory(GiB)": 77.56,
      "step": 68985,
      "token_acc": 0.44011976047904194,
      "train_speed(iter/s)": 1.437706
    },
    {
      "epoch": 2.9557431129771645,
      "grad_norm": 6.0976433753967285,
      "learning_rate": 3.587727469958532e-05,
      "loss": 2.6480321884155273,
      "memory(GiB)": 77.56,
      "step": 68990,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 2.955957328306414,
      "grad_norm": 4.6575517654418945,
      "learning_rate": 3.5870819086161913e-05,
      "loss": 2.3110029220581056,
      "memory(GiB)": 77.56,
      "step": 68995,
      "token_acc": 0.5490196078431373,
      "train_speed(iter/s)": 1.43774
    },
    {
      "epoch": 2.9561715436356626,
      "grad_norm": 6.478888988494873,
      "learning_rate": 3.5864363728701855e-05,
      "loss": 1.9367637634277344,
      "memory(GiB)": 77.56,
      "step": 69000,
      "token_acc": 0.5495867768595041,
      "train_speed(iter/s)": 1.437763
    },
    {
      "epoch": 2.9561715436356626,
      "eval_loss": 2.1820437908172607,
      "eval_runtime": 14.9369,
      "eval_samples_per_second": 6.695,
      "eval_steps_per_second": 6.695,
      "eval_token_acc": 0.4807436918990704,
      "step": 69000
    },
    {
      "epoch": 2.9563857589649114,
      "grad_norm": 4.883810520172119,
      "learning_rate": 3.5857908627322124e-05,
      "loss": 2.6500858306884765,
      "memory(GiB)": 77.56,
      "step": 69005,
      "token_acc": 0.47880539499036606,
      "train_speed(iter/s)": 1.437289
    },
    {
      "epoch": 2.9565999742941607,
      "grad_norm": 4.120144367218018,
      "learning_rate": 3.585145378213963e-05,
      "loss": 2.456507110595703,
      "memory(GiB)": 77.56,
      "step": 69010,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.43728
    },
    {
      "epoch": 2.9568141896234095,
      "grad_norm": 4.422980308532715,
      "learning_rate": 3.584499919327135e-05,
      "loss": 2.5950057983398436,
      "memory(GiB)": 77.56,
      "step": 69015,
      "token_acc": 0.49480968858131485,
      "train_speed(iter/s)": 1.437275
    },
    {
      "epoch": 2.9570284049526583,
      "grad_norm": 5.577856063842773,
      "learning_rate": 3.583854486083417e-05,
      "loss": 2.1856239318847654,
      "memory(GiB)": 77.56,
      "step": 69020,
      "token_acc": 0.4858156028368794,
      "train_speed(iter/s)": 1.437311
    },
    {
      "epoch": 2.9572426202819075,
      "grad_norm": 4.3809943199157715,
      "learning_rate": 3.583209078494503e-05,
      "loss": 2.690225028991699,
      "memory(GiB)": 77.56,
      "step": 69025,
      "token_acc": 0.42028985507246375,
      "train_speed(iter/s)": 1.437325
    },
    {
      "epoch": 2.9574568356111564,
      "grad_norm": 5.803936958312988,
      "learning_rate": 3.582563696572087e-05,
      "loss": 2.4410112380981444,
      "memory(GiB)": 77.56,
      "step": 69030,
      "token_acc": 0.5354330708661418,
      "train_speed(iter/s)": 1.437307
    },
    {
      "epoch": 2.957671050940405,
      "grad_norm": 5.099001884460449,
      "learning_rate": 3.581918340327858e-05,
      "loss": 2.3389801025390624,
      "memory(GiB)": 77.56,
      "step": 69035,
      "token_acc": 0.4620938628158845,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 2.9578852662696544,
      "grad_norm": 6.368360996246338,
      "learning_rate": 3.581273009773509e-05,
      "loss": 2.463871955871582,
      "memory(GiB)": 77.56,
      "step": 69040,
      "token_acc": 0.47808764940239046,
      "train_speed(iter/s)": 1.437298
    },
    {
      "epoch": 2.9580994815989032,
      "grad_norm": 6.084715843200684,
      "learning_rate": 3.5806277049207315e-05,
      "loss": 2.5276683807373046,
      "memory(GiB)": 77.56,
      "step": 69045,
      "token_acc": 0.4539249146757679,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 2.958313696928152,
      "grad_norm": 6.471798419952393,
      "learning_rate": 3.579982425781213e-05,
      "loss": 2.6048458099365233,
      "memory(GiB)": 77.56,
      "step": 69050,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 2.9585279122574013,
      "grad_norm": 6.593715667724609,
      "learning_rate": 3.579337172366646e-05,
      "loss": 2.7753677368164062,
      "memory(GiB)": 77.56,
      "step": 69055,
      "token_acc": 0.4204946996466431,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 2.95874212758665,
      "grad_norm": 5.319879531860352,
      "learning_rate": 3.578691944688719e-05,
      "loss": 2.4280622482299803,
      "memory(GiB)": 77.56,
      "step": 69060,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.437316
    },
    {
      "epoch": 2.958956342915899,
      "grad_norm": 4.729599952697754,
      "learning_rate": 3.5780467427591194e-05,
      "loss": 2.5661991119384764,
      "memory(GiB)": 77.56,
      "step": 69065,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.437324
    },
    {
      "epoch": 2.959170558245148,
      "grad_norm": 4.614162445068359,
      "learning_rate": 3.577401566589535e-05,
      "loss": 2.346187782287598,
      "memory(GiB)": 77.56,
      "step": 69070,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.437337
    },
    {
      "epoch": 2.959384773574397,
      "grad_norm": 6.547674179077148,
      "learning_rate": 3.576756416191659e-05,
      "loss": 2.5412033081054686,
      "memory(GiB)": 77.56,
      "step": 69075,
      "token_acc": 0.46111111111111114,
      "train_speed(iter/s)": 1.437372
    },
    {
      "epoch": 2.959598988903646,
      "grad_norm": 5.097473621368408,
      "learning_rate": 3.5761112915771756e-05,
      "loss": 2.3003753662109374,
      "memory(GiB)": 77.56,
      "step": 69080,
      "token_acc": 0.5251798561151079,
      "train_speed(iter/s)": 1.437393
    },
    {
      "epoch": 2.959813204232895,
      "grad_norm": 6.277078151702881,
      "learning_rate": 3.57546619275777e-05,
      "loss": 2.357630157470703,
      "memory(GiB)": 77.56,
      "step": 69085,
      "token_acc": 0.5335820895522388,
      "train_speed(iter/s)": 1.43741
    },
    {
      "epoch": 2.960027419562144,
      "grad_norm": 5.296885967254639,
      "learning_rate": 3.574821119745133e-05,
      "loss": 2.576180076599121,
      "memory(GiB)": 77.56,
      "step": 69090,
      "token_acc": 0.4463087248322148,
      "train_speed(iter/s)": 1.437416
    },
    {
      "epoch": 2.9602416348913927,
      "grad_norm": 5.158799648284912,
      "learning_rate": 3.5741760725509464e-05,
      "loss": 2.739981842041016,
      "memory(GiB)": 77.56,
      "step": 69095,
      "token_acc": 0.4329608938547486,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 2.960455850220642,
      "grad_norm": 4.835994720458984,
      "learning_rate": 3.5735310511868994e-05,
      "loss": 2.711231994628906,
      "memory(GiB)": 77.56,
      "step": 69100,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.437445
    },
    {
      "epoch": 2.9606700655498908,
      "grad_norm": 4.914797306060791,
      "learning_rate": 3.572886055664675e-05,
      "loss": 2.5512765884399413,
      "memory(GiB)": 77.56,
      "step": 69105,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.437464
    },
    {
      "epoch": 2.9608842808791396,
      "grad_norm": 5.366225242614746,
      "learning_rate": 3.5722410859959574e-05,
      "loss": 2.6116432189941405,
      "memory(GiB)": 77.56,
      "step": 69110,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.437489
    },
    {
      "epoch": 2.961098496208389,
      "grad_norm": 4.803067684173584,
      "learning_rate": 3.571596142192433e-05,
      "loss": 2.1961696624755858,
      "memory(GiB)": 77.56,
      "step": 69115,
      "token_acc": 0.5475409836065573,
      "train_speed(iter/s)": 1.437489
    },
    {
      "epoch": 2.9613127115376376,
      "grad_norm": 6.650675296783447,
      "learning_rate": 3.570951224265785e-05,
      "loss": 2.662397766113281,
      "memory(GiB)": 77.56,
      "step": 69120,
      "token_acc": 0.45514950166112955,
      "train_speed(iter/s)": 1.437495
    },
    {
      "epoch": 2.9615269268668865,
      "grad_norm": 5.322486877441406,
      "learning_rate": 3.570306332227694e-05,
      "loss": 2.2299646377563476,
      "memory(GiB)": 77.56,
      "step": 69125,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.437505
    },
    {
      "epoch": 2.9617411421961357,
      "grad_norm": 7.098770618438721,
      "learning_rate": 3.5696614660898465e-05,
      "loss": 2.410774993896484,
      "memory(GiB)": 77.56,
      "step": 69130,
      "token_acc": 0.5225225225225225,
      "train_speed(iter/s)": 1.437519
    },
    {
      "epoch": 2.9619553575253845,
      "grad_norm": 4.78179407119751,
      "learning_rate": 3.5690166258639226e-05,
      "loss": 2.2891971588134767,
      "memory(GiB)": 77.56,
      "step": 69135,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.437533
    },
    {
      "epoch": 2.9621695728546333,
      "grad_norm": 5.066426753997803,
      "learning_rate": 3.568371811561606e-05,
      "loss": 2.3036334991455076,
      "memory(GiB)": 77.56,
      "step": 69140,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437534
    },
    {
      "epoch": 2.9623837881838826,
      "grad_norm": 5.383299350738525,
      "learning_rate": 3.5677270231945745e-05,
      "loss": 2.3096208572387695,
      "memory(GiB)": 77.56,
      "step": 69145,
      "token_acc": 0.5058823529411764,
      "train_speed(iter/s)": 1.437543
    },
    {
      "epoch": 2.9625980035131314,
      "grad_norm": 5.121975421905518,
      "learning_rate": 3.5670822607745134e-05,
      "loss": 2.5336372375488283,
      "memory(GiB)": 77.56,
      "step": 69150,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.437512
    },
    {
      "epoch": 2.96281221884238,
      "grad_norm": 4.660567760467529,
      "learning_rate": 3.566437524313101e-05,
      "loss": 2.37908935546875,
      "memory(GiB)": 77.56,
      "step": 69155,
      "token_acc": 0.49809885931558934,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 2.9630264341716295,
      "grad_norm": 5.889187335968018,
      "learning_rate": 3.5657928138220184e-05,
      "loss": 2.4170663833618162,
      "memory(GiB)": 77.56,
      "step": 69160,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.437546
    },
    {
      "epoch": 2.9632406495008783,
      "grad_norm": 8.472227096557617,
      "learning_rate": 3.565148129312944e-05,
      "loss": 2.180571746826172,
      "memory(GiB)": 77.56,
      "step": 69165,
      "token_acc": 0.5283842794759825,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 2.963454864830127,
      "grad_norm": 5.193587779998779,
      "learning_rate": 3.564503470797556e-05,
      "loss": 2.653279113769531,
      "memory(GiB)": 77.56,
      "step": 69170,
      "token_acc": 0.45634920634920634,
      "train_speed(iter/s)": 1.437564
    },
    {
      "epoch": 2.9636690801593764,
      "grad_norm": 4.036701202392578,
      "learning_rate": 3.563858838287536e-05,
      "loss": 2.2770002365112303,
      "memory(GiB)": 77.56,
      "step": 69175,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.437579
    },
    {
      "epoch": 2.963883295488625,
      "grad_norm": 5.415582656860352,
      "learning_rate": 3.5632142317945596e-05,
      "loss": 2.3588560104370115,
      "memory(GiB)": 77.56,
      "step": 69180,
      "token_acc": 0.5321428571428571,
      "train_speed(iter/s)": 1.437586
    },
    {
      "epoch": 2.964097510817874,
      "grad_norm": 6.111235618591309,
      "learning_rate": 3.562569651330305e-05,
      "loss": 2.3530622482299806,
      "memory(GiB)": 77.56,
      "step": 69185,
      "token_acc": 0.4617737003058104,
      "train_speed(iter/s)": 1.437585
    },
    {
      "epoch": 2.9643117261471232,
      "grad_norm": 6.13635778427124,
      "learning_rate": 3.561925096906451e-05,
      "loss": 2.367791938781738,
      "memory(GiB)": 77.56,
      "step": 69190,
      "token_acc": 0.466403162055336,
      "train_speed(iter/s)": 1.437605
    },
    {
      "epoch": 2.964525941476372,
      "grad_norm": 5.716488361358643,
      "learning_rate": 3.56128056853467e-05,
      "loss": 2.65100212097168,
      "memory(GiB)": 77.56,
      "step": 69195,
      "token_acc": 0.3944636678200692,
      "train_speed(iter/s)": 1.437635
    },
    {
      "epoch": 2.964740156805621,
      "grad_norm": 7.359189987182617,
      "learning_rate": 3.560636066226644e-05,
      "loss": 2.4877765655517576,
      "memory(GiB)": 77.56,
      "step": 69200,
      "token_acc": 0.45493562231759654,
      "train_speed(iter/s)": 1.43766
    },
    {
      "epoch": 2.96495437213487,
      "grad_norm": 6.101253509521484,
      "learning_rate": 3.5599915899940454e-05,
      "loss": 2.5385768890380858,
      "memory(GiB)": 77.56,
      "step": 69205,
      "token_acc": 0.461038961038961,
      "train_speed(iter/s)": 1.437658
    },
    {
      "epoch": 2.965168587464119,
      "grad_norm": 4.345081806182861,
      "learning_rate": 3.559347139848549e-05,
      "loss": 2.2688720703125,
      "memory(GiB)": 77.56,
      "step": 69210,
      "token_acc": 0.5080906148867314,
      "train_speed(iter/s)": 1.437669
    },
    {
      "epoch": 2.9653828027933677,
      "grad_norm": 4.58785343170166,
      "learning_rate": 3.558702715801832e-05,
      "loss": 2.595186233520508,
      "memory(GiB)": 77.56,
      "step": 69215,
      "token_acc": 0.48135593220338985,
      "train_speed(iter/s)": 1.437676
    },
    {
      "epoch": 2.965597018122617,
      "grad_norm": 5.516713619232178,
      "learning_rate": 3.5580583178655637e-05,
      "loss": 2.162289047241211,
      "memory(GiB)": 77.56,
      "step": 69220,
      "token_acc": 0.528169014084507,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 2.965811233451866,
      "grad_norm": 5.593578338623047,
      "learning_rate": 3.557413946051425e-05,
      "loss": 2.386058044433594,
      "memory(GiB)": 77.56,
      "step": 69225,
      "token_acc": 0.4979253112033195,
      "train_speed(iter/s)": 1.437693
    },
    {
      "epoch": 2.9660254487811146,
      "grad_norm": 6.148902416229248,
      "learning_rate": 3.556769600371084e-05,
      "loss": 2.6028745651245115,
      "memory(GiB)": 77.56,
      "step": 69230,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 2.966239664110364,
      "grad_norm": 8.898268699645996,
      "learning_rate": 3.5561252808362176e-05,
      "loss": 2.7592483520507813,
      "memory(GiB)": 77.56,
      "step": 69235,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437696
    },
    {
      "epoch": 2.9664538794396127,
      "grad_norm": 6.478043079376221,
      "learning_rate": 3.555480987458495e-05,
      "loss": 2.4878005981445312,
      "memory(GiB)": 77.56,
      "step": 69240,
      "token_acc": 0.468,
      "train_speed(iter/s)": 1.437723
    },
    {
      "epoch": 2.9666680947688615,
      "grad_norm": 5.717689514160156,
      "learning_rate": 3.5548367202495894e-05,
      "loss": 2.4582868576049806,
      "memory(GiB)": 77.56,
      "step": 69245,
      "token_acc": 0.4795539033457249,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 2.9668823100981108,
      "grad_norm": 4.743619918823242,
      "learning_rate": 3.554192479221173e-05,
      "loss": 2.4761852264404296,
      "memory(GiB)": 77.56,
      "step": 69250,
      "token_acc": 0.47043010752688175,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 2.9670965254273596,
      "grad_norm": 5.507030963897705,
      "learning_rate": 3.5535482643849153e-05,
      "loss": 2.4245632171630858,
      "memory(GiB)": 77.56,
      "step": 69255,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.437716
    },
    {
      "epoch": 2.9673107407566084,
      "grad_norm": 6.4472455978393555,
      "learning_rate": 3.552904075752489e-05,
      "loss": 2.163592529296875,
      "memory(GiB)": 77.56,
      "step": 69260,
      "token_acc": 0.519650655021834,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 2.9675249560858576,
      "grad_norm": 4.059402942657471,
      "learning_rate": 3.552259913335562e-05,
      "loss": 2.487284469604492,
      "memory(GiB)": 77.56,
      "step": 69265,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 2.9677391714151065,
      "grad_norm": 4.72160005569458,
      "learning_rate": 3.5516157771458045e-05,
      "loss": 2.558095169067383,
      "memory(GiB)": 77.56,
      "step": 69270,
      "token_acc": 0.46075085324232085,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 2.9679533867443553,
      "grad_norm": 4.550678730010986,
      "learning_rate": 3.550971667194886e-05,
      "loss": 2.3446065902709963,
      "memory(GiB)": 77.56,
      "step": 69275,
      "token_acc": 0.5036231884057971,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 2.9681676020736045,
      "grad_norm": 5.5712480545043945,
      "learning_rate": 3.550327583494475e-05,
      "loss": 2.5355155944824217,
      "memory(GiB)": 77.56,
      "step": 69280,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 2.9683818174028533,
      "grad_norm": 5.873742580413818,
      "learning_rate": 3.54968352605624e-05,
      "loss": 2.7183086395263674,
      "memory(GiB)": 77.56,
      "step": 69285,
      "token_acc": 0.4965034965034965,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 2.968596032732102,
      "grad_norm": 6.629122734069824,
      "learning_rate": 3.549039494891849e-05,
      "loss": 2.4706657409667967,
      "memory(GiB)": 77.56,
      "step": 69290,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.437753
    },
    {
      "epoch": 2.9688102480613514,
      "grad_norm": 4.271261692047119,
      "learning_rate": 3.548395490012966e-05,
      "loss": 2.3545095443725588,
      "memory(GiB)": 77.56,
      "step": 69295,
      "token_acc": 0.5202952029520295,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 2.9690244633906,
      "grad_norm": 6.33363151550293,
      "learning_rate": 3.547751511431262e-05,
      "loss": 2.7936716079711914,
      "memory(GiB)": 77.56,
      "step": 69300,
      "token_acc": 0.4275618374558304,
      "train_speed(iter/s)": 1.437751
    },
    {
      "epoch": 2.969238678719849,
      "grad_norm": 5.481980323791504,
      "learning_rate": 3.5471075591584024e-05,
      "loss": 2.181114387512207,
      "memory(GiB)": 77.56,
      "step": 69305,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.437737
    },
    {
      "epoch": 2.9694528940490983,
      "grad_norm": 5.386048316955566,
      "learning_rate": 3.546463633206052e-05,
      "loss": 2.0404840469360352,
      "memory(GiB)": 77.56,
      "step": 69310,
      "token_acc": 0.5625,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 2.969667109378347,
      "grad_norm": 5.444461822509766,
      "learning_rate": 3.545819733585876e-05,
      "loss": 2.879924201965332,
      "memory(GiB)": 77.56,
      "step": 69315,
      "token_acc": 0.4359673024523161,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 2.969881324707596,
      "grad_norm": 6.090682506561279,
      "learning_rate": 3.5451758603095404e-05,
      "loss": 2.5262264251708983,
      "memory(GiB)": 77.56,
      "step": 69320,
      "token_acc": 0.458955223880597,
      "train_speed(iter/s)": 1.437742
    },
    {
      "epoch": 2.970095540036845,
      "grad_norm": 4.950562000274658,
      "learning_rate": 3.5445320133887075e-05,
      "loss": 2.6107624053955076,
      "memory(GiB)": 77.56,
      "step": 69325,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 2.970309755366094,
      "grad_norm": 4.2823686599731445,
      "learning_rate": 3.543888192835044e-05,
      "loss": 2.3906349182128905,
      "memory(GiB)": 77.56,
      "step": 69330,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 2.970523970695343,
      "grad_norm": 5.902844429016113,
      "learning_rate": 3.543244398660212e-05,
      "loss": 2.576519584655762,
      "memory(GiB)": 77.56,
      "step": 69335,
      "token_acc": 0.46105919003115264,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 2.970738186024592,
      "grad_norm": 6.1937642097473145,
      "learning_rate": 3.542600630875873e-05,
      "loss": 2.6070356369018555,
      "memory(GiB)": 77.56,
      "step": 69340,
      "token_acc": 0.4714828897338403,
      "train_speed(iter/s)": 1.437802
    },
    {
      "epoch": 2.970952401353841,
      "grad_norm": 6.378774642944336,
      "learning_rate": 3.541956889493692e-05,
      "loss": 2.4436540603637695,
      "memory(GiB)": 77.56,
      "step": 69345,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.437807
    },
    {
      "epoch": 2.9711666166830897,
      "grad_norm": 6.986944675445557,
      "learning_rate": 3.541313174525329e-05,
      "loss": 2.5839462280273438,
      "memory(GiB)": 77.56,
      "step": 69350,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 2.971380832012339,
      "grad_norm": 6.754354476928711,
      "learning_rate": 3.540669485982445e-05,
      "loss": 2.0642425537109377,
      "memory(GiB)": 77.56,
      "step": 69355,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 2.9715950473415877,
      "grad_norm": 6.530547142028809,
      "learning_rate": 3.540025823876704e-05,
      "loss": 2.606150436401367,
      "memory(GiB)": 77.56,
      "step": 69360,
      "token_acc": 0.45112781954887216,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 2.9718092626708366,
      "grad_norm": 4.96040153503418,
      "learning_rate": 3.539382188219764e-05,
      "loss": 2.6040573120117188,
      "memory(GiB)": 77.56,
      "step": 69365,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.437857
    },
    {
      "epoch": 2.972023478000086,
      "grad_norm": 6.430371284484863,
      "learning_rate": 3.5387385790232854e-05,
      "loss": 2.4298574447631838,
      "memory(GiB)": 77.56,
      "step": 69370,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 2.9722376933293346,
      "grad_norm": 7.155005931854248,
      "learning_rate": 3.538094996298928e-05,
      "loss": 2.4761009216308594,
      "memory(GiB)": 77.56,
      "step": 69375,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 2.9724519086585834,
      "grad_norm": 5.11592960357666,
      "learning_rate": 3.537451440058353e-05,
      "loss": 2.630797576904297,
      "memory(GiB)": 77.56,
      "step": 69380,
      "token_acc": 0.44594594594594594,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 2.9726661239878327,
      "grad_norm": 7.512523651123047,
      "learning_rate": 3.5368079103132143e-05,
      "loss": 2.3299076080322267,
      "memory(GiB)": 77.56,
      "step": 69385,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.437838
    },
    {
      "epoch": 2.9728803393170815,
      "grad_norm": 5.576261043548584,
      "learning_rate": 3.536164407075175e-05,
      "loss": 2.2350683212280273,
      "memory(GiB)": 77.56,
      "step": 69390,
      "token_acc": 0.5487364620938628,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 2.9730945546463303,
      "grad_norm": 5.0224127769470215,
      "learning_rate": 3.535520930355891e-05,
      "loss": 2.3432601928710937,
      "memory(GiB)": 77.56,
      "step": 69395,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 2.9733087699755796,
      "grad_norm": 6.845538139343262,
      "learning_rate": 3.5348774801670184e-05,
      "loss": 2.373804473876953,
      "memory(GiB)": 77.56,
      "step": 69400,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 2.9735229853048284,
      "grad_norm": 5.384983539581299,
      "learning_rate": 3.5342340565202146e-05,
      "loss": 2.216988372802734,
      "memory(GiB)": 77.56,
      "step": 69405,
      "token_acc": 0.532608695652174,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 2.973737200634077,
      "grad_norm": 6.58991003036499,
      "learning_rate": 3.533590659427137e-05,
      "loss": 2.7947261810302733,
      "memory(GiB)": 77.56,
      "step": 69410,
      "token_acc": 0.4205298013245033,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 2.9739514159633265,
      "grad_norm": 5.856507301330566,
      "learning_rate": 3.532947288899439e-05,
      "loss": 2.368599700927734,
      "memory(GiB)": 77.56,
      "step": 69415,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.437921
    },
    {
      "epoch": 2.9741656312925753,
      "grad_norm": 5.09470796585083,
      "learning_rate": 3.532303944948777e-05,
      "loss": 2.3902076721191405,
      "memory(GiB)": 77.56,
      "step": 69420,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 2.974379846621824,
      "grad_norm": 8.592560768127441,
      "learning_rate": 3.5316606275868056e-05,
      "loss": 2.418587875366211,
      "memory(GiB)": 77.56,
      "step": 69425,
      "token_acc": 0.46742209631728043,
      "train_speed(iter/s)": 1.437922
    },
    {
      "epoch": 2.9745940619510733,
      "grad_norm": 5.40716552734375,
      "learning_rate": 3.5310173368251794e-05,
      "loss": 2.3445207595825197,
      "memory(GiB)": 77.56,
      "step": 69430,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437902
    },
    {
      "epoch": 2.974808277280322,
      "grad_norm": 4.981843948364258,
      "learning_rate": 3.5303740726755544e-05,
      "loss": 2.318739891052246,
      "memory(GiB)": 77.56,
      "step": 69435,
      "token_acc": 0.488135593220339,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 2.975022492609571,
      "grad_norm": 7.428905963897705,
      "learning_rate": 3.529730835149577e-05,
      "loss": 2.4202529907226564,
      "memory(GiB)": 77.56,
      "step": 69440,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 2.9752367079388202,
      "grad_norm": 6.593773365020752,
      "learning_rate": 3.5290876242589076e-05,
      "loss": 2.227604866027832,
      "memory(GiB)": 77.56,
      "step": 69445,
      "token_acc": 0.5376712328767124,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 2.975450923268069,
      "grad_norm": 4.56346321105957,
      "learning_rate": 3.528444440015196e-05,
      "loss": 2.352486801147461,
      "memory(GiB)": 77.56,
      "step": 69450,
      "token_acc": 0.46229508196721314,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 2.975665138597318,
      "grad_norm": 7.1637420654296875,
      "learning_rate": 3.5278012824300944e-05,
      "loss": 2.3603553771972656,
      "memory(GiB)": 77.56,
      "step": 69455,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 2.975879353926567,
      "grad_norm": 5.515501022338867,
      "learning_rate": 3.527158151515252e-05,
      "loss": 2.4818950653076173,
      "memory(GiB)": 77.56,
      "step": 69460,
      "token_acc": 0.5097402597402597,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 2.976093569255816,
      "grad_norm": 6.24483060836792,
      "learning_rate": 3.526515047282323e-05,
      "loss": 2.440196228027344,
      "memory(GiB)": 77.56,
      "step": 69465,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 2.9763077845850647,
      "grad_norm": 4.1129350662231445,
      "learning_rate": 3.525871969742954e-05,
      "loss": 2.4850826263427734,
      "memory(GiB)": 77.56,
      "step": 69470,
      "token_acc": 0.5129682997118156,
      "train_speed(iter/s)": 1.437898
    },
    {
      "epoch": 2.976521999914314,
      "grad_norm": 5.955257892608643,
      "learning_rate": 3.525228918908799e-05,
      "loss": 2.193038558959961,
      "memory(GiB)": 77.56,
      "step": 69475,
      "token_acc": 0.5232558139534884,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 2.976736215243563,
      "grad_norm": 5.333489418029785,
      "learning_rate": 3.524585894791506e-05,
      "loss": 2.424423408508301,
      "memory(GiB)": 77.56,
      "step": 69480,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.437874
    },
    {
      "epoch": 2.9769504305728116,
      "grad_norm": 5.985816955566406,
      "learning_rate": 3.523942897402721e-05,
      "loss": 2.375263214111328,
      "memory(GiB)": 77.56,
      "step": 69485,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 2.977164645902061,
      "grad_norm": 6.023412227630615,
      "learning_rate": 3.5232999267540964e-05,
      "loss": 2.254396438598633,
      "memory(GiB)": 77.56,
      "step": 69490,
      "token_acc": 0.4979757085020243,
      "train_speed(iter/s)": 1.437904
    },
    {
      "epoch": 2.9773788612313097,
      "grad_norm": 5.949342250823975,
      "learning_rate": 3.52265698285728e-05,
      "loss": 2.3097095489501953,
      "memory(GiB)": 77.56,
      "step": 69495,
      "token_acc": 0.5226480836236934,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 2.9775930765605585,
      "grad_norm": 6.291986465454102,
      "learning_rate": 3.5220140657239164e-05,
      "loss": 2.600223922729492,
      "memory(GiB)": 77.56,
      "step": 69500,
      "token_acc": 0.43521594684385384,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 2.9775930765605585,
      "eval_loss": 2.2632994651794434,
      "eval_runtime": 13.8832,
      "eval_samples_per_second": 7.203,
      "eval_steps_per_second": 7.203,
      "eval_token_acc": 0.48267008985879334,
      "step": 69500
    },
    {
      "epoch": 2.9778072918898078,
      "grad_norm": 4.687198638916016,
      "learning_rate": 3.5213711753656565e-05,
      "loss": 2.4443904876708986,
      "memory(GiB)": 77.56,
      "step": 69505,
      "token_acc": 0.4800389483933788,
      "train_speed(iter/s)": 1.437482
    },
    {
      "epoch": 2.9780215072190566,
      "grad_norm": 6.376140117645264,
      "learning_rate": 3.520728311794143e-05,
      "loss": 2.7297183990478517,
      "memory(GiB)": 77.56,
      "step": 69510,
      "token_acc": 0.4629080118694362,
      "train_speed(iter/s)": 1.43746
    },
    {
      "epoch": 2.9782357225483054,
      "grad_norm": 5.8469319343566895,
      "learning_rate": 3.5200854750210235e-05,
      "loss": 2.2844945907592775,
      "memory(GiB)": 77.56,
      "step": 69515,
      "token_acc": 0.4901315789473684,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 2.9784499378775546,
      "grad_norm": 6.49983024597168,
      "learning_rate": 3.5194426650579445e-05,
      "loss": 2.2972564697265625,
      "memory(GiB)": 77.56,
      "step": 69520,
      "token_acc": 0.4731182795698925,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 2.9786641532068034,
      "grad_norm": 5.930006980895996,
      "learning_rate": 3.518799881916551e-05,
      "loss": 2.3452810287475585,
      "memory(GiB)": 77.56,
      "step": 69525,
      "token_acc": 0.46396396396396394,
      "train_speed(iter/s)": 1.437517
    },
    {
      "epoch": 2.9788783685360523,
      "grad_norm": 6.153254985809326,
      "learning_rate": 3.518157125608487e-05,
      "loss": 2.435942268371582,
      "memory(GiB)": 77.56,
      "step": 69530,
      "token_acc": 0.47147147147147145,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 2.9790925838653015,
      "grad_norm": 5.583297252655029,
      "learning_rate": 3.5175143961453965e-05,
      "loss": 2.1783119201660157,
      "memory(GiB)": 77.56,
      "step": 69535,
      "token_acc": 0.5244755244755245,
      "train_speed(iter/s)": 1.437512
    },
    {
      "epoch": 2.9793067991945503,
      "grad_norm": 7.191718578338623,
      "learning_rate": 3.516871693538924e-05,
      "loss": 2.603336715698242,
      "memory(GiB)": 77.56,
      "step": 69540,
      "token_acc": 0.44947735191637633,
      "train_speed(iter/s)": 1.437502
    },
    {
      "epoch": 2.979521014523799,
      "grad_norm": 5.396605968475342,
      "learning_rate": 3.516229017800711e-05,
      "loss": 2.48966064453125,
      "memory(GiB)": 77.56,
      "step": 69545,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 2.9797352298530484,
      "grad_norm": 5.894193649291992,
      "learning_rate": 3.515586368942402e-05,
      "loss": 2.366547393798828,
      "memory(GiB)": 77.56,
      "step": 69550,
      "token_acc": 0.5318352059925093,
      "train_speed(iter/s)": 1.43753
    },
    {
      "epoch": 2.979949445182297,
      "grad_norm": 5.607132911682129,
      "learning_rate": 3.514943746975639e-05,
      "loss": 2.6504955291748047,
      "memory(GiB)": 77.56,
      "step": 69555,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437543
    },
    {
      "epoch": 2.980163660511546,
      "grad_norm": 5.978559494018555,
      "learning_rate": 3.514301151912062e-05,
      "loss": 2.3608266830444338,
      "memory(GiB)": 77.56,
      "step": 69560,
      "token_acc": 0.48659003831417624,
      "train_speed(iter/s)": 1.43756
    },
    {
      "epoch": 2.9803778758407953,
      "grad_norm": 6.648810863494873,
      "learning_rate": 3.513658583763314e-05,
      "loss": 2.18863468170166,
      "memory(GiB)": 77.56,
      "step": 69565,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.437576
    },
    {
      "epoch": 2.980592091170044,
      "grad_norm": 5.128035068511963,
      "learning_rate": 3.513016042541034e-05,
      "loss": 2.5890134811401366,
      "memory(GiB)": 77.56,
      "step": 69570,
      "token_acc": 0.43154761904761907,
      "train_speed(iter/s)": 1.437587
    },
    {
      "epoch": 2.980806306499293,
      "grad_norm": 8.38038158416748,
      "learning_rate": 3.5123735282568646e-05,
      "loss": 2.654056930541992,
      "memory(GiB)": 77.56,
      "step": 69575,
      "token_acc": 0.42990654205607476,
      "train_speed(iter/s)": 1.437602
    },
    {
      "epoch": 2.981020521828542,
      "grad_norm": 5.498917102813721,
      "learning_rate": 3.5117310409224443e-05,
      "loss": 2.470449447631836,
      "memory(GiB)": 77.56,
      "step": 69580,
      "token_acc": 0.4647058823529412,
      "train_speed(iter/s)": 1.437623
    },
    {
      "epoch": 2.981234737157791,
      "grad_norm": 8.832070350646973,
      "learning_rate": 3.5110885805494115e-05,
      "loss": 2.570925712585449,
      "memory(GiB)": 77.56,
      "step": 69585,
      "token_acc": 0.476,
      "train_speed(iter/s)": 1.437628
    },
    {
      "epoch": 2.98144895248704,
      "grad_norm": 6.388439178466797,
      "learning_rate": 3.510446147149404e-05,
      "loss": 2.279448318481445,
      "memory(GiB)": 77.56,
      "step": 69590,
      "token_acc": 0.5352112676056338,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 2.981663167816289,
      "grad_norm": 4.355556488037109,
      "learning_rate": 3.509803740734065e-05,
      "loss": 2.0512977600097657,
      "memory(GiB)": 77.56,
      "step": 69595,
      "token_acc": 0.5178571428571429,
      "train_speed(iter/s)": 1.437632
    },
    {
      "epoch": 2.981877383145538,
      "grad_norm": 5.755629062652588,
      "learning_rate": 3.509161361315028e-05,
      "loss": 2.5466222763061523,
      "memory(GiB)": 77.56,
      "step": 69600,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 2.9820915984747867,
      "grad_norm": 7.499678611755371,
      "learning_rate": 3.508519008903931e-05,
      "loss": 2.554818534851074,
      "memory(GiB)": 77.56,
      "step": 69605,
      "token_acc": 0.46757679180887374,
      "train_speed(iter/s)": 1.437616
    },
    {
      "epoch": 2.982305813804036,
      "grad_norm": 4.489687919616699,
      "learning_rate": 3.507876683512412e-05,
      "loss": 2.1566585540771483,
      "memory(GiB)": 77.56,
      "step": 69610,
      "token_acc": 0.5144694533762058,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 2.9825200291332847,
      "grad_norm": 4.953376293182373,
      "learning_rate": 3.507234385152106e-05,
      "loss": 2.4046310424804687,
      "memory(GiB)": 77.56,
      "step": 69615,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.437604
    },
    {
      "epoch": 2.9827342444625335,
      "grad_norm": 5.839040279388428,
      "learning_rate": 3.5065921138346504e-05,
      "loss": 2.527578926086426,
      "memory(GiB)": 77.56,
      "step": 69620,
      "token_acc": 0.5036231884057971,
      "train_speed(iter/s)": 1.437617
    },
    {
      "epoch": 2.982948459791783,
      "grad_norm": 9.756697654724121,
      "learning_rate": 3.505949869571679e-05,
      "loss": 2.232336235046387,
      "memory(GiB)": 77.56,
      "step": 69625,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437612
    },
    {
      "epoch": 2.9831626751210316,
      "grad_norm": 5.325991153717041,
      "learning_rate": 3.505307652374827e-05,
      "loss": 2.2279457092285155,
      "memory(GiB)": 77.56,
      "step": 69630,
      "token_acc": 0.5628930817610063,
      "train_speed(iter/s)": 1.437608
    },
    {
      "epoch": 2.9833768904502804,
      "grad_norm": 4.278589248657227,
      "learning_rate": 3.5046654622557295e-05,
      "loss": 2.436077690124512,
      "memory(GiB)": 77.56,
      "step": 69635,
      "token_acc": 0.5097402597402597,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 2.9835911057795297,
      "grad_norm": 3.8859148025512695,
      "learning_rate": 3.50402329922602e-05,
      "loss": 2.3893901824951174,
      "memory(GiB)": 77.56,
      "step": 69640,
      "token_acc": 0.4891640866873065,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 2.9838053211087785,
      "grad_norm": 8.65683364868164,
      "learning_rate": 3.5033811632973315e-05,
      "loss": 2.6062952041625977,
      "memory(GiB)": 77.56,
      "step": 69645,
      "token_acc": 0.436046511627907,
      "train_speed(iter/s)": 1.437611
    },
    {
      "epoch": 2.9840195364380273,
      "grad_norm": 6.522454738616943,
      "learning_rate": 3.502739054481297e-05,
      "loss": 2.6266656875610352,
      "memory(GiB)": 77.56,
      "step": 69650,
      "token_acc": 0.4329268292682927,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 2.9842337517672766,
      "grad_norm": 5.323433876037598,
      "learning_rate": 3.5020969727895484e-05,
      "loss": 2.34755859375,
      "memory(GiB)": 77.56,
      "step": 69655,
      "token_acc": 0.48615384615384616,
      "train_speed(iter/s)": 1.437639
    },
    {
      "epoch": 2.9844479670965254,
      "grad_norm": 7.181215763092041,
      "learning_rate": 3.50145491823372e-05,
      "loss": 2.919140625,
      "memory(GiB)": 77.56,
      "step": 69660,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.437652
    },
    {
      "epoch": 2.984662182425774,
      "grad_norm": 6.406488418579102,
      "learning_rate": 3.500812890825439e-05,
      "loss": 2.5161109924316407,
      "memory(GiB)": 77.56,
      "step": 69665,
      "token_acc": 0.4936061381074169,
      "train_speed(iter/s)": 1.437669
    },
    {
      "epoch": 2.9848763977550234,
      "grad_norm": 5.650805950164795,
      "learning_rate": 3.50017089057634e-05,
      "loss": 2.3362913131713867,
      "memory(GiB)": 77.56,
      "step": 69670,
      "token_acc": 0.5049180327868853,
      "train_speed(iter/s)": 1.437687
    },
    {
      "epoch": 2.9850906130842723,
      "grad_norm": 6.8401713371276855,
      "learning_rate": 3.499528917498053e-05,
      "loss": 2.4290498733520507,
      "memory(GiB)": 77.56,
      "step": 69675,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.437671
    },
    {
      "epoch": 2.985304828413521,
      "grad_norm": 6.590735912322998,
      "learning_rate": 3.4988869716022065e-05,
      "loss": 2.361783027648926,
      "memory(GiB)": 77.56,
      "step": 69680,
      "token_acc": 0.5316455696202531,
      "train_speed(iter/s)": 1.437708
    },
    {
      "epoch": 2.9855190437427703,
      "grad_norm": 7.112832069396973,
      "learning_rate": 3.498245052900432e-05,
      "loss": 2.619435691833496,
      "memory(GiB)": 77.56,
      "step": 69685,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.437726
    },
    {
      "epoch": 2.985733259072019,
      "grad_norm": 5.147935390472412,
      "learning_rate": 3.4976031614043555e-05,
      "loss": 2.3561704635620115,
      "memory(GiB)": 77.56,
      "step": 69690,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 2.985947474401268,
      "grad_norm": 3.965451240539551,
      "learning_rate": 3.496961297125608e-05,
      "loss": 2.8114501953125,
      "memory(GiB)": 77.56,
      "step": 69695,
      "token_acc": 0.46407185628742514,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 2.986161689730517,
      "grad_norm": 5.191863536834717,
      "learning_rate": 3.4963194600758166e-05,
      "loss": 2.213589668273926,
      "memory(GiB)": 77.56,
      "step": 69700,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.437692
    },
    {
      "epoch": 2.986375905059766,
      "grad_norm": 10.08134937286377,
      "learning_rate": 3.4956776502666076e-05,
      "loss": 2.4462020874023436,
      "memory(GiB)": 77.56,
      "step": 69705,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 2.986590120389015,
      "grad_norm": 5.610540390014648,
      "learning_rate": 3.4950358677096103e-05,
      "loss": 2.6670482635498045,
      "memory(GiB)": 77.56,
      "step": 69710,
      "token_acc": 0.43,
      "train_speed(iter/s)": 1.437709
    },
    {
      "epoch": 2.986804335718264,
      "grad_norm": 5.5300445556640625,
      "learning_rate": 3.4943941124164494e-05,
      "loss": 2.5622278213500977,
      "memory(GiB)": 77.56,
      "step": 69715,
      "token_acc": 0.47648902821316613,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 2.987018551047513,
      "grad_norm": 6.6703572273254395,
      "learning_rate": 3.493752384398753e-05,
      "loss": 2.4519445419311525,
      "memory(GiB)": 77.56,
      "step": 69720,
      "token_acc": 0.45564516129032256,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 2.9872327663767617,
      "grad_norm": 7.126587390899658,
      "learning_rate": 3.493110683668144e-05,
      "loss": 2.4102910995483398,
      "memory(GiB)": 77.56,
      "step": 69725,
      "token_acc": 0.47491638795986624,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 2.987446981706011,
      "grad_norm": 5.214530944824219,
      "learning_rate": 3.4924690102362475e-05,
      "loss": 2.3161338806152343,
      "memory(GiB)": 77.56,
      "step": 69730,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 2.98766119703526,
      "grad_norm": 6.560204029083252,
      "learning_rate": 3.491827364114689e-05,
      "loss": 2.3912942886352537,
      "memory(GiB)": 77.56,
      "step": 69735,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 2.9878754123645086,
      "grad_norm": 4.339321136474609,
      "learning_rate": 3.491185745315094e-05,
      "loss": 2.397929382324219,
      "memory(GiB)": 77.56,
      "step": 69740,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 2.988089627693758,
      "grad_norm": 8.264789581298828,
      "learning_rate": 3.490544153849085e-05,
      "loss": 2.492336463928223,
      "memory(GiB)": 77.56,
      "step": 69745,
      "token_acc": 0.4778761061946903,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 2.9883038430230067,
      "grad_norm": 7.304971218109131,
      "learning_rate": 3.489902589728283e-05,
      "loss": 2.107792854309082,
      "memory(GiB)": 77.56,
      "step": 69750,
      "token_acc": 0.5473684210526316,
      "train_speed(iter/s)": 1.43773
    },
    {
      "epoch": 2.9885180583522555,
      "grad_norm": 5.320586681365967,
      "learning_rate": 3.4892610529643135e-05,
      "loss": 2.702079010009766,
      "memory(GiB)": 77.56,
      "step": 69755,
      "token_acc": 0.4426229508196721,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 2.9887322736815047,
      "grad_norm": 5.018779754638672,
      "learning_rate": 3.488619543568796e-05,
      "loss": 2.6156057357788085,
      "memory(GiB)": 77.56,
      "step": 69760,
      "token_acc": 0.4836795252225519,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 2.9889464890107535,
      "grad_norm": 5.850031852722168,
      "learning_rate": 3.487978061553355e-05,
      "loss": 2.377509880065918,
      "memory(GiB)": 77.56,
      "step": 69765,
      "token_acc": 0.48639455782312924,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 2.9891607043400024,
      "grad_norm": 5.953535556793213,
      "learning_rate": 3.4873366069296095e-05,
      "loss": 2.351734924316406,
      "memory(GiB)": 77.56,
      "step": 69770,
      "token_acc": 0.48046875,
      "train_speed(iter/s)": 1.437764
    },
    {
      "epoch": 2.9893749196692516,
      "grad_norm": 5.522590637207031,
      "learning_rate": 3.4866951797091786e-05,
      "loss": 2.4259239196777345,
      "memory(GiB)": 77.56,
      "step": 69775,
      "token_acc": 0.49337748344370863,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 2.9895891349985004,
      "grad_norm": 4.952304840087891,
      "learning_rate": 3.486053779903686e-05,
      "loss": 2.4415794372558595,
      "memory(GiB)": 77.56,
      "step": 69780,
      "token_acc": 0.47330960854092524,
      "train_speed(iter/s)": 1.43774
    },
    {
      "epoch": 2.9898033503277492,
      "grad_norm": 5.014679908752441,
      "learning_rate": 3.485412407524749e-05,
      "loss": 2.6219457626342773,
      "memory(GiB)": 77.56,
      "step": 69785,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 2.9900175656569985,
      "grad_norm": 4.547702312469482,
      "learning_rate": 3.484771062583986e-05,
      "loss": 2.3821346282958986,
      "memory(GiB)": 77.56,
      "step": 69790,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 2.9902317809862473,
      "grad_norm": 5.8471455574035645,
      "learning_rate": 3.484129745093018e-05,
      "loss": 2.6142168045043945,
      "memory(GiB)": 77.56,
      "step": 69795,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.437761
    },
    {
      "epoch": 2.990445996315496,
      "grad_norm": 5.91321325302124,
      "learning_rate": 3.48348845506346e-05,
      "loss": 2.4356136322021484,
      "memory(GiB)": 77.56,
      "step": 69800,
      "token_acc": 0.4492307692307692,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 2.9906602116447454,
      "grad_norm": 4.776285648345947,
      "learning_rate": 3.482847192506933e-05,
      "loss": 2.455434799194336,
      "memory(GiB)": 77.56,
      "step": 69805,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.437746
    },
    {
      "epoch": 2.990874426973994,
      "grad_norm": 4.8118085861206055,
      "learning_rate": 3.48220595743505e-05,
      "loss": 2.3861270904541017,
      "memory(GiB)": 77.56,
      "step": 69810,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 2.991088642303243,
      "grad_norm": 6.598339557647705,
      "learning_rate": 3.481564749859431e-05,
      "loss": 2.598998260498047,
      "memory(GiB)": 77.56,
      "step": 69815,
      "token_acc": 0.4766355140186916,
      "train_speed(iter/s)": 1.43774
    },
    {
      "epoch": 2.9913028576324923,
      "grad_norm": 5.918899059295654,
      "learning_rate": 3.480923569791691e-05,
      "loss": 2.6179405212402345,
      "memory(GiB)": 77.56,
      "step": 69820,
      "token_acc": 0.4152046783625731,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 2.991517072961741,
      "grad_norm": 6.022837162017822,
      "learning_rate": 3.480282417243446e-05,
      "loss": 2.21358585357666,
      "memory(GiB)": 77.56,
      "step": 69825,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 2.99173128829099,
      "grad_norm": 5.195446014404297,
      "learning_rate": 3.479641292226311e-05,
      "loss": 2.460297393798828,
      "memory(GiB)": 77.56,
      "step": 69830,
      "token_acc": 0.4649122807017544,
      "train_speed(iter/s)": 1.437764
    },
    {
      "epoch": 2.991945503620239,
      "grad_norm": 5.995636940002441,
      "learning_rate": 3.479000194751899e-05,
      "loss": 2.3766172409057615,
      "memory(GiB)": 77.56,
      "step": 69835,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 2.992159718949488,
      "grad_norm": 4.970757484436035,
      "learning_rate": 3.478359124831827e-05,
      "loss": 2.352031135559082,
      "memory(GiB)": 77.56,
      "step": 69840,
      "token_acc": 0.5418060200668896,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 2.9923739342787368,
      "grad_norm": 4.864602565765381,
      "learning_rate": 3.4777180824777057e-05,
      "loss": 2.4529241561889648,
      "memory(GiB)": 77.56,
      "step": 69845,
      "token_acc": 0.46449704142011833,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 2.992588149607986,
      "grad_norm": 6.022401809692383,
      "learning_rate": 3.477077067701149e-05,
      "loss": 2.3812414169311524,
      "memory(GiB)": 77.56,
      "step": 69850,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 2.992802364937235,
      "grad_norm": 4.516592502593994,
      "learning_rate": 3.476436080513771e-05,
      "loss": 2.2282077789306642,
      "memory(GiB)": 77.56,
      "step": 69855,
      "token_acc": 0.5466666666666666,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 2.9930165802664837,
      "grad_norm": 7.423250675201416,
      "learning_rate": 3.475795120927181e-05,
      "loss": 2.2731250762939452,
      "memory(GiB)": 77.56,
      "step": 69860,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 2.993230795595733,
      "grad_norm": 5.776782989501953,
      "learning_rate": 3.475154188952994e-05,
      "loss": 2.3110647201538086,
      "memory(GiB)": 77.56,
      "step": 69865,
      "token_acc": 0.501628664495114,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 2.9934450109249817,
      "grad_norm": 7.36826229095459,
      "learning_rate": 3.4745132846028185e-05,
      "loss": 2.6141178131103517,
      "memory(GiB)": 77.56,
      "step": 69870,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 2.9936592262542305,
      "grad_norm": 4.841373443603516,
      "learning_rate": 3.473872407888266e-05,
      "loss": 2.1498823165893555,
      "memory(GiB)": 77.56,
      "step": 69875,
      "token_acc": 0.5434782608695652,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 2.99387344158348,
      "grad_norm": 6.328704833984375,
      "learning_rate": 3.473231558820946e-05,
      "loss": 2.7730695724487306,
      "memory(GiB)": 77.56,
      "step": 69880,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 2.9940876569127286,
      "grad_norm": 4.848683834075928,
      "learning_rate": 3.472590737412467e-05,
      "loss": 2.423569107055664,
      "memory(GiB)": 77.56,
      "step": 69885,
      "token_acc": 0.5165562913907285,
      "train_speed(iter/s)": 1.437802
    },
    {
      "epoch": 2.9943018722419774,
      "grad_norm": 5.444921016693115,
      "learning_rate": 3.471949943674442e-05,
      "loss": 2.6714550018310548,
      "memory(GiB)": 77.56,
      "step": 69890,
      "token_acc": 0.47477744807121663,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 2.9945160875712267,
      "grad_norm": 5.013216018676758,
      "learning_rate": 3.471309177618476e-05,
      "loss": 2.217668151855469,
      "memory(GiB)": 77.56,
      "step": 69895,
      "token_acc": 0.4980694980694981,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 2.9947303029004755,
      "grad_norm": 5.5079665184021,
      "learning_rate": 3.47066843925618e-05,
      "loss": 2.409965705871582,
      "memory(GiB)": 77.56,
      "step": 69900,
      "token_acc": 0.549407114624506,
      "train_speed(iter/s)": 1.437764
    },
    {
      "epoch": 2.9949445182297243,
      "grad_norm": 6.3039703369140625,
      "learning_rate": 3.4700277285991575e-05,
      "loss": 2.1393192291259764,
      "memory(GiB)": 77.56,
      "step": 69905,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 2.9951587335589736,
      "grad_norm": 4.885385036468506,
      "learning_rate": 3.469387045659019e-05,
      "loss": 2.515911865234375,
      "memory(GiB)": 77.56,
      "step": 69910,
      "token_acc": 0.5050505050505051,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 2.9953729488882224,
      "grad_norm": 4.702334880828857,
      "learning_rate": 3.4687463904473716e-05,
      "loss": 2.3052322387695314,
      "memory(GiB)": 77.56,
      "step": 69915,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 2.995587164217471,
      "grad_norm": 5.00528621673584,
      "learning_rate": 3.468105762975817e-05,
      "loss": 2.4896568298339843,
      "memory(GiB)": 77.56,
      "step": 69920,
      "token_acc": 0.4748427672955975,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 2.9958013795467204,
      "grad_norm": 6.449862480163574,
      "learning_rate": 3.467465163255966e-05,
      "loss": 2.3143749237060547,
      "memory(GiB)": 77.56,
      "step": 69925,
      "token_acc": 0.5207547169811321,
      "train_speed(iter/s)": 1.437755
    },
    {
      "epoch": 2.9960155948759692,
      "grad_norm": 4.4548492431640625,
      "learning_rate": 3.46682459129942e-05,
      "loss": 2.471028518676758,
      "memory(GiB)": 77.56,
      "step": 69930,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437788
    },
    {
      "epoch": 2.996229810205218,
      "grad_norm": 5.446771621704102,
      "learning_rate": 3.466184047117784e-05,
      "loss": 2.3469188690185545,
      "memory(GiB)": 77.56,
      "step": 69935,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.437796
    },
    {
      "epoch": 2.9964440255344673,
      "grad_norm": 5.143713474273682,
      "learning_rate": 3.4655435307226645e-05,
      "loss": 2.495412826538086,
      "memory(GiB)": 77.56,
      "step": 69940,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 2.996658240863716,
      "grad_norm": 8.154741287231445,
      "learning_rate": 3.4649030421256625e-05,
      "loss": 2.307531547546387,
      "memory(GiB)": 77.56,
      "step": 69945,
      "token_acc": 0.5027932960893855,
      "train_speed(iter/s)": 1.437782
    },
    {
      "epoch": 2.996872456192965,
      "grad_norm": 4.66604471206665,
      "learning_rate": 3.464262581338382e-05,
      "loss": 2.464568519592285,
      "memory(GiB)": 77.56,
      "step": 69950,
      "token_acc": 0.48297213622291024,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 2.997086671522214,
      "grad_norm": 5.159629821777344,
      "learning_rate": 3.463622148372426e-05,
      "loss": 2.203643798828125,
      "memory(GiB)": 77.56,
      "step": 69955,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 2.997300886851463,
      "grad_norm": 5.51721715927124,
      "learning_rate": 3.462981743239394e-05,
      "loss": 2.3900142669677735,
      "memory(GiB)": 77.56,
      "step": 69960,
      "token_acc": 0.47333333333333333,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 2.997515102180712,
      "grad_norm": 8.074501991271973,
      "learning_rate": 3.46234136595089e-05,
      "loss": 2.333910369873047,
      "memory(GiB)": 77.56,
      "step": 69965,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.437786
    },
    {
      "epoch": 2.997729317509961,
      "grad_norm": 7.173497676849365,
      "learning_rate": 3.461701016518516e-05,
      "loss": 2.3989498138427736,
      "memory(GiB)": 77.56,
      "step": 69970,
      "token_acc": 0.4730290456431535,
      "train_speed(iter/s)": 1.43774
    },
    {
      "epoch": 2.99794353283921,
      "grad_norm": 7.386610984802246,
      "learning_rate": 3.461060694953871e-05,
      "loss": 2.425118255615234,
      "memory(GiB)": 77.56,
      "step": 69975,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.437763
    },
    {
      "epoch": 2.9981577481684587,
      "grad_norm": 5.4531426429748535,
      "learning_rate": 3.4604204012685546e-05,
      "loss": 2.3320119857788084,
      "memory(GiB)": 77.56,
      "step": 69980,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.437765
    },
    {
      "epoch": 2.998371963497708,
      "grad_norm": 5.652688980102539,
      "learning_rate": 3.459780135474168e-05,
      "loss": 2.392618179321289,
      "memory(GiB)": 77.56,
      "step": 69985,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 2.9985861788269568,
      "grad_norm": 5.391223430633545,
      "learning_rate": 3.4591398975823084e-05,
      "loss": 2.3509243011474608,
      "memory(GiB)": 77.56,
      "step": 69990,
      "token_acc": 0.47878787878787876,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 2.9988003941562056,
      "grad_norm": 7.125851631164551,
      "learning_rate": 3.458499687604575e-05,
      "loss": 2.5453536987304686,
      "memory(GiB)": 77.56,
      "step": 69995,
      "token_acc": 0.4733893557422969,
      "train_speed(iter/s)": 1.437775
    },
    {
      "epoch": 2.999014609485455,
      "grad_norm": 6.092185974121094,
      "learning_rate": 3.457859505552565e-05,
      "loss": 2.4291168212890626,
      "memory(GiB)": 77.56,
      "step": 70000,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 2.999014609485455,
      "eval_loss": 2.0748212337493896,
      "eval_runtime": 14.4906,
      "eval_samples_per_second": 6.901,
      "eval_steps_per_second": 6.901,
      "eval_token_acc": 0.481029810298103,
      "step": 70000
    },
    {
      "epoch": 2.9992288248147037,
      "grad_norm": 4.501657485961914,
      "learning_rate": 3.457219351437877e-05,
      "loss": 2.2379199981689455,
      "memory(GiB)": 77.56,
      "step": 70005,
      "token_acc": 0.47448015122873344,
      "train_speed(iter/s)": 1.437305
    },
    {
      "epoch": 2.9994430401439525,
      "grad_norm": 5.824315071105957,
      "learning_rate": 3.456579225272109e-05,
      "loss": 2.425770378112793,
      "memory(GiB)": 77.56,
      "step": 70010,
      "token_acc": 0.5220588235294118,
      "train_speed(iter/s)": 1.437324
    },
    {
      "epoch": 2.9996572554732017,
      "grad_norm": 4.7570414543151855,
      "learning_rate": 3.4559391270668554e-05,
      "loss": 2.3508453369140625,
      "memory(GiB)": 77.56,
      "step": 70015,
      "token_acc": 0.4876325088339223,
      "train_speed(iter/s)": 1.437341
    },
    {
      "epoch": 2.9998714708024505,
      "grad_norm": 5.920152187347412,
      "learning_rate": 3.4552990568337124e-05,
      "loss": 2.4949527740478517,
      "memory(GiB)": 77.56,
      "step": 70020,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437345
    },
    {
      "epoch": 3.0000856861317,
      "grad_norm": 5.641233921051025,
      "learning_rate": 3.454659014584277e-05,
      "loss": 2.28286075592041,
      "memory(GiB)": 77.56,
      "step": 70025,
      "token_acc": 0.5175097276264592,
      "train_speed(iter/s)": 1.437358
    },
    {
      "epoch": 3.0002999014609486,
      "grad_norm": 4.419128894805908,
      "learning_rate": 3.4540190003301434e-05,
      "loss": 2.1704872131347654,
      "memory(GiB)": 77.56,
      "step": 70030,
      "token_acc": 0.541958041958042,
      "train_speed(iter/s)": 1.437362
    },
    {
      "epoch": 3.0005141167901974,
      "grad_norm": 6.164981365203857,
      "learning_rate": 3.453379014082902e-05,
      "loss": 2.470066452026367,
      "memory(GiB)": 77.56,
      "step": 70035,
      "token_acc": 0.51953125,
      "train_speed(iter/s)": 1.437354
    },
    {
      "epoch": 3.0007283321194467,
      "grad_norm": 4.922651767730713,
      "learning_rate": 3.452739055854154e-05,
      "loss": 2.340447998046875,
      "memory(GiB)": 77.56,
      "step": 70040,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.437387
    },
    {
      "epoch": 3.0009425474486955,
      "grad_norm": 5.426289081573486,
      "learning_rate": 3.4520991256554895e-05,
      "loss": 1.9927356719970704,
      "memory(GiB)": 77.56,
      "step": 70045,
      "token_acc": 0.588,
      "train_speed(iter/s)": 1.437397
    },
    {
      "epoch": 3.0011567627779443,
      "grad_norm": 7.255398750305176,
      "learning_rate": 3.4514592234985006e-05,
      "loss": 2.410924530029297,
      "memory(GiB)": 77.56,
      "step": 70050,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 3.0013709781071936,
      "grad_norm": 5.883856296539307,
      "learning_rate": 3.4508193493947816e-05,
      "loss": 2.414633369445801,
      "memory(GiB)": 77.56,
      "step": 70055,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.437379
    },
    {
      "epoch": 3.0015851934364424,
      "grad_norm": 5.694874286651611,
      "learning_rate": 3.4501795033559224e-05,
      "loss": 2.5532037734985353,
      "memory(GiB)": 77.56,
      "step": 70060,
      "token_acc": 0.4708029197080292,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 3.001799408765691,
      "grad_norm": 4.599838733673096,
      "learning_rate": 3.449539685393516e-05,
      "loss": 2.2517772674560548,
      "memory(GiB)": 77.56,
      "step": 70065,
      "token_acc": 0.5348837209302325,
      "train_speed(iter/s)": 1.437393
    },
    {
      "epoch": 3.0020136240949404,
      "grad_norm": 4.233546733856201,
      "learning_rate": 3.448899895519152e-05,
      "loss": 2.3383678436279296,
      "memory(GiB)": 77.56,
      "step": 70070,
      "token_acc": 0.49866666666666665,
      "train_speed(iter/s)": 1.437388
    },
    {
      "epoch": 3.0022278394241892,
      "grad_norm": 5.9302659034729,
      "learning_rate": 3.448260133744422e-05,
      "loss": 2.395679473876953,
      "memory(GiB)": 77.56,
      "step": 70075,
      "token_acc": 0.4588607594936709,
      "train_speed(iter/s)": 1.437397
    },
    {
      "epoch": 3.002442054753438,
      "grad_norm": 5.467451572418213,
      "learning_rate": 3.4476204000809145e-05,
      "loss": 2.3761524200439452,
      "memory(GiB)": 77.56,
      "step": 70080,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.437381
    },
    {
      "epoch": 3.0026562700826873,
      "grad_norm": 4.680141925811768,
      "learning_rate": 3.446980694540221e-05,
      "loss": 2.249062156677246,
      "memory(GiB)": 77.56,
      "step": 70085,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.437393
    },
    {
      "epoch": 3.002870485411936,
      "grad_norm": 6.755695819854736,
      "learning_rate": 3.4463410171339275e-05,
      "loss": 2.548707389831543,
      "memory(GiB)": 77.56,
      "step": 70090,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 3.003084700741185,
      "grad_norm": 5.0909600257873535,
      "learning_rate": 3.445701367873625e-05,
      "loss": 2.396769332885742,
      "memory(GiB)": 77.56,
      "step": 70095,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.437385
    },
    {
      "epoch": 3.003298916070434,
      "grad_norm": 6.153361797332764,
      "learning_rate": 3.445061746770901e-05,
      "loss": 1.972971534729004,
      "memory(GiB)": 77.56,
      "step": 70100,
      "token_acc": 0.5528169014084507,
      "train_speed(iter/s)": 1.437398
    },
    {
      "epoch": 3.003513131399683,
      "grad_norm": 6.290983200073242,
      "learning_rate": 3.444422153837339e-05,
      "loss": 2.1196720123291017,
      "memory(GiB)": 77.56,
      "step": 70105,
      "token_acc": 0.5328947368421053,
      "train_speed(iter/s)": 1.437405
    },
    {
      "epoch": 3.003727346728932,
      "grad_norm": 6.0566301345825195,
      "learning_rate": 3.443782589084531e-05,
      "loss": 2.595070648193359,
      "memory(GiB)": 77.56,
      "step": 70110,
      "token_acc": 0.45396825396825397,
      "train_speed(iter/s)": 1.437404
    },
    {
      "epoch": 3.003941562058181,
      "grad_norm": 4.441219329833984,
      "learning_rate": 3.443143052524062e-05,
      "loss": 2.3609437942504883,
      "memory(GiB)": 77.56,
      "step": 70115,
      "token_acc": 0.47413793103448276,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 3.00415577738743,
      "grad_norm": 4.946651458740234,
      "learning_rate": 3.4425035441675165e-05,
      "loss": 2.354739952087402,
      "memory(GiB)": 77.56,
      "step": 70120,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.43737
    },
    {
      "epoch": 3.0043699927166787,
      "grad_norm": 5.974501609802246,
      "learning_rate": 3.441864064026479e-05,
      "loss": 2.1505544662475584,
      "memory(GiB)": 77.56,
      "step": 70125,
      "token_acc": 0.512,
      "train_speed(iter/s)": 1.437372
    },
    {
      "epoch": 3.004584208045928,
      "grad_norm": 7.834686279296875,
      "learning_rate": 3.441224612112538e-05,
      "loss": 2.222279930114746,
      "memory(GiB)": 77.56,
      "step": 70130,
      "token_acc": 0.4952681388012618,
      "train_speed(iter/s)": 1.437379
    },
    {
      "epoch": 3.0047984233751768,
      "grad_norm": 5.017378807067871,
      "learning_rate": 3.440585188437273e-05,
      "loss": 2.4043447494506838,
      "memory(GiB)": 77.56,
      "step": 70135,
      "token_acc": 0.5034722222222222,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 3.0050126387044256,
      "grad_norm": 5.082510948181152,
      "learning_rate": 3.439945793012272e-05,
      "loss": 2.595010757446289,
      "memory(GiB)": 77.56,
      "step": 70140,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437385
    },
    {
      "epoch": 3.005226854033675,
      "grad_norm": 5.418440341949463,
      "learning_rate": 3.439306425849116e-05,
      "loss": 2.23193244934082,
      "memory(GiB)": 77.56,
      "step": 70145,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.437377
    },
    {
      "epoch": 3.0054410693629237,
      "grad_norm": 6.121163368225098,
      "learning_rate": 3.438667086959388e-05,
      "loss": 2.2608058929443358,
      "memory(GiB)": 77.56,
      "step": 70150,
      "token_acc": 0.5261194029850746,
      "train_speed(iter/s)": 1.437352
    },
    {
      "epoch": 3.0056552846921725,
      "grad_norm": 5.88928747177124,
      "learning_rate": 3.438027776354671e-05,
      "loss": 3.0113218307495115,
      "memory(GiB)": 77.56,
      "step": 70155,
      "token_acc": 0.4492307692307692,
      "train_speed(iter/s)": 1.437336
    },
    {
      "epoch": 3.0058695000214217,
      "grad_norm": 7.035855293273926,
      "learning_rate": 3.437388494046545e-05,
      "loss": 2.4730562210083007,
      "memory(GiB)": 77.56,
      "step": 70160,
      "token_acc": 0.45985401459854014,
      "train_speed(iter/s)": 1.437342
    },
    {
      "epoch": 3.0060837153506705,
      "grad_norm": 4.201283931732178,
      "learning_rate": 3.4367492400465926e-05,
      "loss": 2.0827741622924805,
      "memory(GiB)": 77.56,
      "step": 70165,
      "token_acc": 0.5680272108843537,
      "train_speed(iter/s)": 1.43735
    },
    {
      "epoch": 3.0062979306799194,
      "grad_norm": 5.9304633140563965,
      "learning_rate": 3.436110014366395e-05,
      "loss": 2.1160346984863283,
      "memory(GiB)": 77.56,
      "step": 70170,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.437367
    },
    {
      "epoch": 3.0065121460091686,
      "grad_norm": 3.9505488872528076,
      "learning_rate": 3.4354708170175296e-05,
      "loss": 2.3902923583984377,
      "memory(GiB)": 77.56,
      "step": 70175,
      "token_acc": 0.5222222222222223,
      "train_speed(iter/s)": 1.437393
    },
    {
      "epoch": 3.0067263613384174,
      "grad_norm": 6.202552318572998,
      "learning_rate": 3.4348316480115786e-05,
      "loss": 2.3049552917480467,
      "memory(GiB)": 77.56,
      "step": 70180,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.437374
    },
    {
      "epoch": 3.0069405766676662,
      "grad_norm": 5.326395511627197,
      "learning_rate": 3.4341925073601206e-05,
      "loss": 2.3128583908081053,
      "memory(GiB)": 77.56,
      "step": 70185,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 3.0071547919969155,
      "grad_norm": 6.259341716766357,
      "learning_rate": 3.433553395074735e-05,
      "loss": 2.373114585876465,
      "memory(GiB)": 77.56,
      "step": 70190,
      "token_acc": 0.5065359477124183,
      "train_speed(iter/s)": 1.437388
    },
    {
      "epoch": 3.0073690073261643,
      "grad_norm": 7.973504066467285,
      "learning_rate": 3.432914311166998e-05,
      "loss": 2.333552360534668,
      "memory(GiB)": 77.56,
      "step": 70195,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.437408
    },
    {
      "epoch": 3.007583222655413,
      "grad_norm": 6.080653667449951,
      "learning_rate": 3.4322752556484896e-05,
      "loss": 2.4298679351806642,
      "memory(GiB)": 77.56,
      "step": 70200,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 3.0077974379846624,
      "grad_norm": 7.564701080322266,
      "learning_rate": 3.4316362285307854e-05,
      "loss": 2.389156150817871,
      "memory(GiB)": 77.56,
      "step": 70205,
      "token_acc": 0.47572815533980584,
      "train_speed(iter/s)": 1.437427
    },
    {
      "epoch": 3.008011653313911,
      "grad_norm": 5.251187801361084,
      "learning_rate": 3.430997229825462e-05,
      "loss": 2.404718017578125,
      "memory(GiB)": 77.56,
      "step": 70210,
      "token_acc": 0.4786885245901639,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 3.00822586864316,
      "grad_norm": 5.972806453704834,
      "learning_rate": 3.430358259544095e-05,
      "loss": 2.184909439086914,
      "memory(GiB)": 77.56,
      "step": 70215,
      "token_acc": 0.531986531986532,
      "train_speed(iter/s)": 1.43741
    },
    {
      "epoch": 3.0084400839724093,
      "grad_norm": 5.10725736618042,
      "learning_rate": 3.429719317698262e-05,
      "loss": 2.4998268127441405,
      "memory(GiB)": 77.56,
      "step": 70220,
      "token_acc": 0.48578199052132703,
      "train_speed(iter/s)": 1.437406
    },
    {
      "epoch": 3.008654299301658,
      "grad_norm": 4.5519514083862305,
      "learning_rate": 3.429080404299535e-05,
      "loss": 2.054467964172363,
      "memory(GiB)": 77.56,
      "step": 70225,
      "token_acc": 0.48466257668711654,
      "train_speed(iter/s)": 1.437403
    },
    {
      "epoch": 3.008868514630907,
      "grad_norm": 5.849546909332275,
      "learning_rate": 3.428441519359491e-05,
      "loss": 2.0395259857177734,
      "memory(GiB)": 77.56,
      "step": 70230,
      "token_acc": 0.5506072874493927,
      "train_speed(iter/s)": 1.43741
    },
    {
      "epoch": 3.009082729960156,
      "grad_norm": 6.683048248291016,
      "learning_rate": 3.4278026628897024e-05,
      "loss": 2.5393098831176757,
      "memory(GiB)": 77.56,
      "step": 70235,
      "token_acc": 0.5014925373134328,
      "train_speed(iter/s)": 1.437416
    },
    {
      "epoch": 3.009296945289405,
      "grad_norm": 6.5449018478393555,
      "learning_rate": 3.427163834901744e-05,
      "loss": 2.19445915222168,
      "memory(GiB)": 77.56,
      "step": 70240,
      "token_acc": 0.5517241379310345,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 3.0095111606186538,
      "grad_norm": 6.474233150482178,
      "learning_rate": 3.426525035407189e-05,
      "loss": 2.316154670715332,
      "memory(GiB)": 77.56,
      "step": 70245,
      "token_acc": 0.5054151624548736,
      "train_speed(iter/s)": 1.437451
    },
    {
      "epoch": 3.009725375947903,
      "grad_norm": 8.122316360473633,
      "learning_rate": 3.425886264417607e-05,
      "loss": 2.3829437255859376,
      "memory(GiB)": 77.56,
      "step": 70250,
      "token_acc": 0.4591194968553459,
      "train_speed(iter/s)": 1.437456
    },
    {
      "epoch": 3.009939591277152,
      "grad_norm": 7.372654438018799,
      "learning_rate": 3.425247521944572e-05,
      "loss": 2.374039649963379,
      "memory(GiB)": 77.56,
      "step": 70255,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437489
    },
    {
      "epoch": 3.0101538066064006,
      "grad_norm": 13.866686820983887,
      "learning_rate": 3.424608807999656e-05,
      "loss": 2.520269203186035,
      "memory(GiB)": 77.56,
      "step": 70260,
      "token_acc": 0.5034722222222222,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 3.01036802193565,
      "grad_norm": 8.366415977478027,
      "learning_rate": 3.423970122594431e-05,
      "loss": 2.223109245300293,
      "memory(GiB)": 77.56,
      "step": 70265,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.437485
    },
    {
      "epoch": 3.0105822372648987,
      "grad_norm": 4.910088062286377,
      "learning_rate": 3.423331465740463e-05,
      "loss": 2.357571029663086,
      "memory(GiB)": 77.56,
      "step": 70270,
      "token_acc": 0.5090361445783133,
      "train_speed(iter/s)": 1.437482
    },
    {
      "epoch": 3.0107964525941475,
      "grad_norm": 4.795945167541504,
      "learning_rate": 3.4226928374493275e-05,
      "loss": 2.013970947265625,
      "memory(GiB)": 77.56,
      "step": 70275,
      "token_acc": 0.5360501567398119,
      "train_speed(iter/s)": 1.437463
    },
    {
      "epoch": 3.0110106679233968,
      "grad_norm": 6.291277885437012,
      "learning_rate": 3.422054237732588e-05,
      "loss": 2.191616249084473,
      "memory(GiB)": 77.56,
      "step": 70280,
      "token_acc": 0.5188284518828452,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 3.0112248832526456,
      "grad_norm": 7.073383808135986,
      "learning_rate": 3.421415666601818e-05,
      "loss": 2.198455810546875,
      "memory(GiB)": 77.56,
      "step": 70285,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.4375
    },
    {
      "epoch": 3.0114390985818944,
      "grad_norm": 5.46431827545166,
      "learning_rate": 3.420777124068584e-05,
      "loss": 2.4840248107910154,
      "memory(GiB)": 77.56,
      "step": 70290,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.437524
    },
    {
      "epoch": 3.0116533139111437,
      "grad_norm": 6.905582904815674,
      "learning_rate": 3.4201386101444524e-05,
      "loss": 2.1254100799560547,
      "memory(GiB)": 77.56,
      "step": 70295,
      "token_acc": 0.5377049180327869,
      "train_speed(iter/s)": 1.437524
    },
    {
      "epoch": 3.0118675292403925,
      "grad_norm": 6.294972896575928,
      "learning_rate": 3.419500124840994e-05,
      "loss": 2.311610221862793,
      "memory(GiB)": 77.56,
      "step": 70300,
      "token_acc": 0.5183098591549296,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 3.0120817445696413,
      "grad_norm": 6.419466018676758,
      "learning_rate": 3.418861668169773e-05,
      "loss": 2.239666748046875,
      "memory(GiB)": 77.56,
      "step": 70305,
      "token_acc": 0.5057034220532319,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 3.0122959598988905,
      "grad_norm": 5.778848171234131,
      "learning_rate": 3.4182232401423544e-05,
      "loss": 2.308125686645508,
      "memory(GiB)": 77.56,
      "step": 70310,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.437482
    },
    {
      "epoch": 3.0125101752281394,
      "grad_norm": 8.681206703186035,
      "learning_rate": 3.417584840770307e-05,
      "loss": 2.50396842956543,
      "memory(GiB)": 77.56,
      "step": 70315,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.437499
    },
    {
      "epoch": 3.012724390557388,
      "grad_norm": 6.515965461730957,
      "learning_rate": 3.416946470065193e-05,
      "loss": 2.5161258697509767,
      "memory(GiB)": 77.56,
      "step": 70320,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437507
    },
    {
      "epoch": 3.0129386058866374,
      "grad_norm": 5.736940860748291,
      "learning_rate": 3.416308128038581e-05,
      "loss": 2.4026493072509765,
      "memory(GiB)": 77.56,
      "step": 70325,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437519
    },
    {
      "epoch": 3.0131528212158862,
      "grad_norm": 5.693610191345215,
      "learning_rate": 3.415669814702031e-05,
      "loss": 2.4020872116088867,
      "memory(GiB)": 77.56,
      "step": 70330,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 3.013367036545135,
      "grad_norm": 6.333526134490967,
      "learning_rate": 3.415031530067109e-05,
      "loss": 2.2822597503662108,
      "memory(GiB)": 77.56,
      "step": 70335,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437535
    },
    {
      "epoch": 3.0135812518743843,
      "grad_norm": 7.3658857345581055,
      "learning_rate": 3.414393274145378e-05,
      "loss": 2.291658401489258,
      "memory(GiB)": 77.56,
      "step": 70340,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.437544
    },
    {
      "epoch": 3.013795467203633,
      "grad_norm": 5.9387102127075195,
      "learning_rate": 3.4137550469484e-05,
      "loss": 2.5541927337646486,
      "memory(GiB)": 77.56,
      "step": 70345,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.4375
    },
    {
      "epoch": 3.014009682532882,
      "grad_norm": 6.346806049346924,
      "learning_rate": 3.4131168484877385e-05,
      "loss": 2.3148178100585937,
      "memory(GiB)": 77.56,
      "step": 70350,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437517
    },
    {
      "epoch": 3.014223897862131,
      "grad_norm": 5.464722156524658,
      "learning_rate": 3.412478678774952e-05,
      "loss": 2.59960823059082,
      "memory(GiB)": 77.56,
      "step": 70355,
      "token_acc": 0.42727272727272725,
      "train_speed(iter/s)": 1.437539
    },
    {
      "epoch": 3.01443811319138,
      "grad_norm": 5.355817794799805,
      "learning_rate": 3.4118405378216056e-05,
      "loss": 2.1365318298339844,
      "memory(GiB)": 77.56,
      "step": 70360,
      "token_acc": 0.5390070921985816,
      "train_speed(iter/s)": 1.437528
    },
    {
      "epoch": 3.014652328520629,
      "grad_norm": 6.916172504425049,
      "learning_rate": 3.411202425639257e-05,
      "loss": 1.8687414169311523,
      "memory(GiB)": 77.56,
      "step": 70365,
      "token_acc": 0.55078125,
      "train_speed(iter/s)": 1.437526
    },
    {
      "epoch": 3.014866543849878,
      "grad_norm": 5.728539943695068,
      "learning_rate": 3.410564342239466e-05,
      "loss": 2.389829635620117,
      "memory(GiB)": 77.56,
      "step": 70370,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.437542
    },
    {
      "epoch": 3.015080759179127,
      "grad_norm": 9.179864883422852,
      "learning_rate": 3.409926287633793e-05,
      "loss": 2.2522197723388673,
      "memory(GiB)": 77.56,
      "step": 70375,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.437567
    },
    {
      "epoch": 3.0152949745083757,
      "grad_norm": 5.689191818237305,
      "learning_rate": 3.409288261833797e-05,
      "loss": 2.503635597229004,
      "memory(GiB)": 77.56,
      "step": 70380,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 3.015509189837625,
      "grad_norm": 6.755014896392822,
      "learning_rate": 3.408650264851038e-05,
      "loss": 2.354074478149414,
      "memory(GiB)": 77.56,
      "step": 70385,
      "token_acc": 0.5396825396825397,
      "train_speed(iter/s)": 1.437597
    },
    {
      "epoch": 3.0157234051668738,
      "grad_norm": 5.1933698654174805,
      "learning_rate": 3.4080122966970705e-05,
      "loss": 2.2398874282836916,
      "memory(GiB)": 77.56,
      "step": 70390,
      "token_acc": 0.5103448275862069,
      "train_speed(iter/s)": 1.437562
    },
    {
      "epoch": 3.0159376204961226,
      "grad_norm": 4.797146797180176,
      "learning_rate": 3.4073743573834535e-05,
      "loss": 2.3041778564453126,
      "memory(GiB)": 77.56,
      "step": 70395,
      "token_acc": 0.4651898734177215,
      "train_speed(iter/s)": 1.437579
    },
    {
      "epoch": 3.016151835825372,
      "grad_norm": 6.0858588218688965,
      "learning_rate": 3.406736446921743e-05,
      "loss": 2.4038543701171875,
      "memory(GiB)": 77.56,
      "step": 70400,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.437605
    },
    {
      "epoch": 3.0163660511546206,
      "grad_norm": 4.796118259429932,
      "learning_rate": 3.406098565323499e-05,
      "loss": 2.3136032104492186,
      "memory(GiB)": 77.56,
      "step": 70405,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.437604
    },
    {
      "epoch": 3.0165802664838695,
      "grad_norm": 5.765377521514893,
      "learning_rate": 3.4054607126002734e-05,
      "loss": 2.2307804107666014,
      "memory(GiB)": 77.56,
      "step": 70410,
      "token_acc": 0.5103857566765578,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 3.0167944818131187,
      "grad_norm": 6.051815509796143,
      "learning_rate": 3.404822888763623e-05,
      "loss": 2.281178855895996,
      "memory(GiB)": 77.56,
      "step": 70415,
      "token_acc": 0.5376712328767124,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 3.0170086971423675,
      "grad_norm": 7.341418266296387,
      "learning_rate": 3.404185093825103e-05,
      "loss": 2.264250373840332,
      "memory(GiB)": 77.56,
      "step": 70420,
      "token_acc": 0.4676258992805755,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 3.0172229124716163,
      "grad_norm": 7.620987415313721,
      "learning_rate": 3.403547327796266e-05,
      "loss": 2.4151514053344725,
      "memory(GiB)": 77.56,
      "step": 70425,
      "token_acc": 0.5174825174825175,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 3.0174371278008656,
      "grad_norm": 6.364876747131348,
      "learning_rate": 3.4029095906886676e-05,
      "loss": 2.0931800842285155,
      "memory(GiB)": 77.56,
      "step": 70430,
      "token_acc": 0.583916083916084,
      "train_speed(iter/s)": 1.43763
    },
    {
      "epoch": 3.0176513431301144,
      "grad_norm": 5.987843036651611,
      "learning_rate": 3.40227188251386e-05,
      "loss": 2.36295108795166,
      "memory(GiB)": 77.56,
      "step": 70435,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.437652
    },
    {
      "epoch": 3.017865558459363,
      "grad_norm": 5.084464073181152,
      "learning_rate": 3.401634203283395e-05,
      "loss": 2.233847999572754,
      "memory(GiB)": 77.56,
      "step": 70440,
      "token_acc": 0.5326797385620915,
      "train_speed(iter/s)": 1.437659
    },
    {
      "epoch": 3.0180797737886125,
      "grad_norm": 4.420731544494629,
      "learning_rate": 3.4009965530088274e-05,
      "loss": 2.244358253479004,
      "memory(GiB)": 77.56,
      "step": 70445,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.437687
    },
    {
      "epoch": 3.0182939891178613,
      "grad_norm": 6.721184730529785,
      "learning_rate": 3.4003589317017074e-05,
      "loss": 2.3993999481201174,
      "memory(GiB)": 77.56,
      "step": 70450,
      "token_acc": 0.43703703703703706,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.01850820444711,
      "grad_norm": 5.2902069091796875,
      "learning_rate": 3.399721339373584e-05,
      "loss": 1.9199289321899413,
      "memory(GiB)": 77.56,
      "step": 70455,
      "token_acc": 0.5253164556962026,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 3.0187224197763594,
      "grad_norm": 5.390171051025391,
      "learning_rate": 3.3990837760360106e-05,
      "loss": 2.3308927536010744,
      "memory(GiB)": 77.56,
      "step": 70460,
      "token_acc": 0.5303867403314917,
      "train_speed(iter/s)": 1.437681
    },
    {
      "epoch": 3.018936635105608,
      "grad_norm": 5.8123698234558105,
      "learning_rate": 3.398446241700536e-05,
      "loss": 2.1574642181396486,
      "memory(GiB)": 77.56,
      "step": 70465,
      "token_acc": 0.5340136054421769,
      "train_speed(iter/s)": 1.437674
    },
    {
      "epoch": 3.019150850434857,
      "grad_norm": 5.640305042266846,
      "learning_rate": 3.397808736378711e-05,
      "loss": 2.6134563446044923,
      "memory(GiB)": 77.56,
      "step": 70470,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 3.0193650657641062,
      "grad_norm": 6.796550750732422,
      "learning_rate": 3.397171260082082e-05,
      "loss": 2.414898681640625,
      "memory(GiB)": 77.56,
      "step": 70475,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.437662
    },
    {
      "epoch": 3.019579281093355,
      "grad_norm": 5.607170581817627,
      "learning_rate": 3.3965338128222e-05,
      "loss": 2.606097412109375,
      "memory(GiB)": 77.56,
      "step": 70480,
      "token_acc": 0.4851190476190476,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 3.019793496422604,
      "grad_norm": 6.351754665374756,
      "learning_rate": 3.3958963946106126e-05,
      "loss": 2.2404232025146484,
      "memory(GiB)": 77.56,
      "step": 70485,
      "token_acc": 0.4780701754385965,
      "train_speed(iter/s)": 1.437673
    },
    {
      "epoch": 3.020007711751853,
      "grad_norm": 5.325263500213623,
      "learning_rate": 3.3952590054588675e-05,
      "loss": 2.297855567932129,
      "memory(GiB)": 77.56,
      "step": 70490,
      "token_acc": 0.5342465753424658,
      "train_speed(iter/s)": 1.437675
    },
    {
      "epoch": 3.020221927081102,
      "grad_norm": 6.447782039642334,
      "learning_rate": 3.394621645378511e-05,
      "loss": 2.354689598083496,
      "memory(GiB)": 77.56,
      "step": 70495,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.437672
    },
    {
      "epoch": 3.0204361424103507,
      "grad_norm": 4.932530879974365,
      "learning_rate": 3.3939843143810874e-05,
      "loss": 2.6362049102783205,
      "memory(GiB)": 77.56,
      "step": 70500,
      "token_acc": 0.528052805280528,
      "train_speed(iter/s)": 1.437681
    },
    {
      "epoch": 3.0204361424103507,
      "eval_loss": 2.2287631034851074,
      "eval_runtime": 15.0191,
      "eval_samples_per_second": 6.658,
      "eval_steps_per_second": 6.658,
      "eval_token_acc": 0.4685408299866131,
      "step": 70500
    },
    {
      "epoch": 3.0206503577396,
      "grad_norm": 5.360347270965576,
      "learning_rate": 3.3933470124781466e-05,
      "loss": 2.4924240112304688,
      "memory(GiB)": 77.56,
      "step": 70505,
      "token_acc": 0.47696737044145876,
      "train_speed(iter/s)": 1.43721
    },
    {
      "epoch": 3.020864573068849,
      "grad_norm": 5.935469627380371,
      "learning_rate": 3.3927097396812304e-05,
      "loss": 1.7939041137695313,
      "memory(GiB)": 77.56,
      "step": 70510,
      "token_acc": 0.576271186440678,
      "train_speed(iter/s)": 1.437218
    },
    {
      "epoch": 3.0210787883980976,
      "grad_norm": 6.828772068023682,
      "learning_rate": 3.3920724960018856e-05,
      "loss": 2.6675512313842775,
      "memory(GiB)": 77.56,
      "step": 70515,
      "token_acc": 0.48828125,
      "train_speed(iter/s)": 1.437235
    },
    {
      "epoch": 3.021293003727347,
      "grad_norm": 5.44118070602417,
      "learning_rate": 3.3914352814516556e-05,
      "loss": 2.379140090942383,
      "memory(GiB)": 77.56,
      "step": 70520,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.437224
    },
    {
      "epoch": 3.0215072190565957,
      "grad_norm": 5.2825751304626465,
      "learning_rate": 3.390798096042085e-05,
      "loss": 2.429238510131836,
      "memory(GiB)": 77.56,
      "step": 70525,
      "token_acc": 0.4844290657439446,
      "train_speed(iter/s)": 1.437237
    },
    {
      "epoch": 3.0217214343858445,
      "grad_norm": 6.980903148651123,
      "learning_rate": 3.390160939784716e-05,
      "loss": 2.152998352050781,
      "memory(GiB)": 77.56,
      "step": 70530,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.437261
    },
    {
      "epoch": 3.0219356497150938,
      "grad_norm": 5.881457328796387,
      "learning_rate": 3.3895238126910924e-05,
      "loss": 2.1978784561157227,
      "memory(GiB)": 77.56,
      "step": 70535,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.437259
    },
    {
      "epoch": 3.0221498650443426,
      "grad_norm": 4.331933975219727,
      "learning_rate": 3.388886714772754e-05,
      "loss": 2.080428123474121,
      "memory(GiB)": 77.56,
      "step": 70540,
      "token_acc": 0.5124223602484472,
      "train_speed(iter/s)": 1.437276
    },
    {
      "epoch": 3.0223640803735914,
      "grad_norm": 4.232517242431641,
      "learning_rate": 3.388249646041246e-05,
      "loss": 2.3634016036987306,
      "memory(GiB)": 77.56,
      "step": 70545,
      "token_acc": 0.5292096219931272,
      "train_speed(iter/s)": 1.437273
    },
    {
      "epoch": 3.0225782957028406,
      "grad_norm": 6.706959247589111,
      "learning_rate": 3.387612606508105e-05,
      "loss": 2.637325668334961,
      "memory(GiB)": 77.56,
      "step": 70550,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.437265
    },
    {
      "epoch": 3.0227925110320895,
      "grad_norm": 6.284237861633301,
      "learning_rate": 3.386975596184877e-05,
      "loss": 2.4386037826538085,
      "memory(GiB)": 77.56,
      "step": 70555,
      "token_acc": 0.49185667752442996,
      "train_speed(iter/s)": 1.43726
    },
    {
      "epoch": 3.0230067263613383,
      "grad_norm": 5.004030227661133,
      "learning_rate": 3.3863386150830974e-05,
      "loss": 2.47908992767334,
      "memory(GiB)": 77.56,
      "step": 70560,
      "token_acc": 0.5145228215767634,
      "train_speed(iter/s)": 1.437269
    },
    {
      "epoch": 3.0232209416905875,
      "grad_norm": 5.973959445953369,
      "learning_rate": 3.385701663214309e-05,
      "loss": 2.30418701171875,
      "memory(GiB)": 77.56,
      "step": 70565,
      "token_acc": 0.5257731958762887,
      "train_speed(iter/s)": 1.437271
    },
    {
      "epoch": 3.0234351570198363,
      "grad_norm": 10.449799537658691,
      "learning_rate": 3.385064740590048e-05,
      "loss": 2.459584617614746,
      "memory(GiB)": 77.56,
      "step": 70570,
      "token_acc": 0.4819672131147541,
      "train_speed(iter/s)": 1.437288
    },
    {
      "epoch": 3.023649372349085,
      "grad_norm": 5.968154430389404,
      "learning_rate": 3.384427847221855e-05,
      "loss": 2.3906455993652345,
      "memory(GiB)": 77.56,
      "step": 70575,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.437293
    },
    {
      "epoch": 3.0238635876783344,
      "grad_norm": 5.551918029785156,
      "learning_rate": 3.383790983121267e-05,
      "loss": 2.3442138671875,
      "memory(GiB)": 77.56,
      "step": 70580,
      "token_acc": 0.4956521739130435,
      "train_speed(iter/s)": 1.437322
    },
    {
      "epoch": 3.024077803007583,
      "grad_norm": 6.067318439483643,
      "learning_rate": 3.3831541482998205e-05,
      "loss": 1.964788055419922,
      "memory(GiB)": 77.56,
      "step": 70585,
      "token_acc": 0.558303886925795,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 3.024292018336832,
      "grad_norm": 5.7133283615112305,
      "learning_rate": 3.3825173427690546e-05,
      "loss": 2.325041580200195,
      "memory(GiB)": 77.56,
      "step": 70590,
      "token_acc": 0.540625,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 3.0245062336660813,
      "grad_norm": 5.407682418823242,
      "learning_rate": 3.381880566540505e-05,
      "loss": 2.341222381591797,
      "memory(GiB)": 77.56,
      "step": 70595,
      "token_acc": 0.5248447204968945,
      "train_speed(iter/s)": 1.437363
    },
    {
      "epoch": 3.02472044899533,
      "grad_norm": 7.491780757904053,
      "learning_rate": 3.381243819625705e-05,
      "loss": 2.2960657119750976,
      "memory(GiB)": 77.56,
      "step": 70600,
      "token_acc": 0.5314685314685315,
      "train_speed(iter/s)": 1.437363
    },
    {
      "epoch": 3.024934664324579,
      "grad_norm": 4.596776485443115,
      "learning_rate": 3.380607102036193e-05,
      "loss": 2.2798250198364256,
      "memory(GiB)": 77.56,
      "step": 70605,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.437356
    },
    {
      "epoch": 3.025148879653828,
      "grad_norm": 6.106468200683594,
      "learning_rate": 3.3799704137835014e-05,
      "loss": 2.1350807189941405,
      "memory(GiB)": 77.56,
      "step": 70610,
      "token_acc": 0.5119453924914675,
      "train_speed(iter/s)": 1.43732
    },
    {
      "epoch": 3.025363094983077,
      "grad_norm": 5.095737457275391,
      "learning_rate": 3.3793337548791666e-05,
      "loss": 2.242348289489746,
      "memory(GiB)": 77.56,
      "step": 70615,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.437313
    },
    {
      "epoch": 3.025577310312326,
      "grad_norm": 5.405594348907471,
      "learning_rate": 3.378697125334721e-05,
      "loss": 2.3909561157226564,
      "memory(GiB)": 77.56,
      "step": 70620,
      "token_acc": 0.4837662337662338,
      "train_speed(iter/s)": 1.437328
    },
    {
      "epoch": 3.025791525641575,
      "grad_norm": 6.876720428466797,
      "learning_rate": 3.378060525161696e-05,
      "loss": 2.1319761276245117,
      "memory(GiB)": 77.56,
      "step": 70625,
      "token_acc": 0.5096525096525096,
      "train_speed(iter/s)": 1.437351
    },
    {
      "epoch": 3.026005740970824,
      "grad_norm": 5.188251495361328,
      "learning_rate": 3.377423954371627e-05,
      "loss": 2.487404632568359,
      "memory(GiB)": 77.56,
      "step": 70630,
      "token_acc": 0.47491638795986624,
      "train_speed(iter/s)": 1.437344
    },
    {
      "epoch": 3.0262199563000727,
      "grad_norm": 4.916665077209473,
      "learning_rate": 3.3767874129760455e-05,
      "loss": 2.3435245513916017,
      "memory(GiB)": 77.56,
      "step": 70635,
      "token_acc": 0.5377049180327869,
      "train_speed(iter/s)": 1.437344
    },
    {
      "epoch": 3.026434171629322,
      "grad_norm": 5.324584484100342,
      "learning_rate": 3.3761509009864834e-05,
      "loss": 2.0037839889526365,
      "memory(GiB)": 77.56,
      "step": 70640,
      "token_acc": 0.5693215339233039,
      "train_speed(iter/s)": 1.43736
    },
    {
      "epoch": 3.0266483869585707,
      "grad_norm": 5.0959930419921875,
      "learning_rate": 3.3755144184144704e-05,
      "loss": 2.244406890869141,
      "memory(GiB)": 77.56,
      "step": 70645,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.437368
    },
    {
      "epoch": 3.0268626022878196,
      "grad_norm": 5.386826515197754,
      "learning_rate": 3.374877965271538e-05,
      "loss": 2.1589298248291016,
      "memory(GiB)": 77.56,
      "step": 70650,
      "token_acc": 0.5177304964539007,
      "train_speed(iter/s)": 1.437367
    },
    {
      "epoch": 3.027076817617069,
      "grad_norm": 5.784743309020996,
      "learning_rate": 3.374241541569216e-05,
      "loss": 2.434076690673828,
      "memory(GiB)": 77.56,
      "step": 70655,
      "token_acc": 0.48753462603878117,
      "train_speed(iter/s)": 1.43736
    },
    {
      "epoch": 3.0272910329463176,
      "grad_norm": 7.675990104675293,
      "learning_rate": 3.3736051473190324e-05,
      "loss": 2.4441043853759767,
      "memory(GiB)": 77.56,
      "step": 70660,
      "token_acc": 0.48828125,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.0275052482755664,
      "grad_norm": 6.093672275543213,
      "learning_rate": 3.372968782532519e-05,
      "loss": 2.062820816040039,
      "memory(GiB)": 77.56,
      "step": 70665,
      "token_acc": 0.5138461538461538,
      "train_speed(iter/s)": 1.437381
    },
    {
      "epoch": 3.0277194636048157,
      "grad_norm": 7.627948760986328,
      "learning_rate": 3.3723324472212e-05,
      "loss": 2.190084457397461,
      "memory(GiB)": 77.56,
      "step": 70670,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.437359
    },
    {
      "epoch": 3.0279336789340645,
      "grad_norm": 5.10483980178833,
      "learning_rate": 3.371696141396608e-05,
      "loss": 2.5277629852294923,
      "memory(GiB)": 77.56,
      "step": 70675,
      "token_acc": 0.4560260586319218,
      "train_speed(iter/s)": 1.43737
    },
    {
      "epoch": 3.0281478942633133,
      "grad_norm": 5.51690673828125,
      "learning_rate": 3.371059865070266e-05,
      "loss": 2.3362171173095705,
      "memory(GiB)": 77.56,
      "step": 70680,
      "token_acc": 0.5059288537549407,
      "train_speed(iter/s)": 1.437388
    },
    {
      "epoch": 3.0283621095925626,
      "grad_norm": 6.007562160491943,
      "learning_rate": 3.370423618253703e-05,
      "loss": 2.424604606628418,
      "memory(GiB)": 77.56,
      "step": 70685,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.437393
    },
    {
      "epoch": 3.0285763249218114,
      "grad_norm": 11.194559097290039,
      "learning_rate": 3.369787400958446e-05,
      "loss": 2.4155141830444338,
      "memory(GiB)": 77.56,
      "step": 70690,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.437376
    },
    {
      "epoch": 3.02879054025106,
      "grad_norm": 7.8410468101501465,
      "learning_rate": 3.369151213196019e-05,
      "loss": 2.267900085449219,
      "memory(GiB)": 77.56,
      "step": 70695,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.437389
    },
    {
      "epoch": 3.0290047555803095,
      "grad_norm": 5.812904357910156,
      "learning_rate": 3.368515054977945e-05,
      "loss": 2.752921295166016,
      "memory(GiB)": 77.56,
      "step": 70700,
      "token_acc": 0.4622356495468278,
      "train_speed(iter/s)": 1.43738
    },
    {
      "epoch": 3.0292189709095583,
      "grad_norm": 6.987752437591553,
      "learning_rate": 3.3678789263157534e-05,
      "loss": 2.320968246459961,
      "memory(GiB)": 77.56,
      "step": 70705,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.437354
    },
    {
      "epoch": 3.029433186238807,
      "grad_norm": 7.077080249786377,
      "learning_rate": 3.367242827220967e-05,
      "loss": 2.3458202362060545,
      "memory(GiB)": 77.56,
      "step": 70710,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.437356
    },
    {
      "epoch": 3.0296474015680563,
      "grad_norm": 7.908858776092529,
      "learning_rate": 3.3666067577051064e-05,
      "loss": 2.483523368835449,
      "memory(GiB)": 77.56,
      "step": 70715,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.437372
    },
    {
      "epoch": 3.029861616897305,
      "grad_norm": 6.426355361938477,
      "learning_rate": 3.3659707177796984e-05,
      "loss": 2.22983341217041,
      "memory(GiB)": 77.56,
      "step": 70720,
      "token_acc": 0.5234657039711191,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 3.030075832226554,
      "grad_norm": 5.172299861907959,
      "learning_rate": 3.365334707456264e-05,
      "loss": 2.438453483581543,
      "memory(GiB)": 77.56,
      "step": 70725,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.437372
    },
    {
      "epoch": 3.0302900475558032,
      "grad_norm": 5.077765464782715,
      "learning_rate": 3.364698726746323e-05,
      "loss": 2.3051753997802735,
      "memory(GiB)": 77.56,
      "step": 70730,
      "token_acc": 0.5298245614035088,
      "train_speed(iter/s)": 1.437368
    },
    {
      "epoch": 3.030504262885052,
      "grad_norm": 5.132670879364014,
      "learning_rate": 3.364062775661401e-05,
      "loss": 2.370410346984863,
      "memory(GiB)": 77.56,
      "step": 70735,
      "token_acc": 0.5311475409836065,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 3.030718478214301,
      "grad_norm": 6.580378532409668,
      "learning_rate": 3.3634268542130146e-05,
      "loss": 2.4990028381347655,
      "memory(GiB)": 77.56,
      "step": 70740,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.437329
    },
    {
      "epoch": 3.03093269354355,
      "grad_norm": 8.148248672485352,
      "learning_rate": 3.3627909624126864e-05,
      "loss": 2.4106117248535157,
      "memory(GiB)": 77.56,
      "step": 70745,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.437338
    },
    {
      "epoch": 3.031146908872799,
      "grad_norm": 6.455301761627197,
      "learning_rate": 3.362155100271936e-05,
      "loss": 2.394194793701172,
      "memory(GiB)": 77.56,
      "step": 70750,
      "token_acc": 0.47041420118343197,
      "train_speed(iter/s)": 1.437336
    },
    {
      "epoch": 3.0313611242020477,
      "grad_norm": 5.465808391571045,
      "learning_rate": 3.361519267802281e-05,
      "loss": 2.0059978485107424,
      "memory(GiB)": 77.56,
      "step": 70755,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.437345
    },
    {
      "epoch": 3.031575339531297,
      "grad_norm": 6.298335552215576,
      "learning_rate": 3.360883465015243e-05,
      "loss": 2.4253538131713865,
      "memory(GiB)": 77.56,
      "step": 70760,
      "token_acc": 0.522633744855967,
      "train_speed(iter/s)": 1.437346
    },
    {
      "epoch": 3.031789554860546,
      "grad_norm": 5.976151943206787,
      "learning_rate": 3.3602476919223393e-05,
      "loss": 2.414750671386719,
      "memory(GiB)": 77.56,
      "step": 70765,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.0320037701897946,
      "grad_norm": 6.39660120010376,
      "learning_rate": 3.359611948535084e-05,
      "loss": 2.361659049987793,
      "memory(GiB)": 77.56,
      "step": 70770,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.437389
    },
    {
      "epoch": 3.032217985519044,
      "grad_norm": 5.713761806488037,
      "learning_rate": 3.358976234864999e-05,
      "loss": 2.5065639495849608,
      "memory(GiB)": 77.56,
      "step": 70775,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437372
    },
    {
      "epoch": 3.0324322008482927,
      "grad_norm": 6.726590156555176,
      "learning_rate": 3.3583405509235986e-05,
      "loss": 2.1831663131713865,
      "memory(GiB)": 77.56,
      "step": 70780,
      "token_acc": 0.5177865612648221,
      "train_speed(iter/s)": 1.437341
    },
    {
      "epoch": 3.0326464161775415,
      "grad_norm": 4.485472202301025,
      "learning_rate": 3.3577048967224e-05,
      "loss": 2.03262939453125,
      "memory(GiB)": 77.56,
      "step": 70785,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.437346
    },
    {
      "epoch": 3.0328606315067907,
      "grad_norm": 5.325892925262451,
      "learning_rate": 3.3570692722729167e-05,
      "loss": 2.202468490600586,
      "memory(GiB)": 77.56,
      "step": 70790,
      "token_acc": 0.5265017667844523,
      "train_speed(iter/s)": 1.437347
    },
    {
      "epoch": 3.0330748468360396,
      "grad_norm": 7.766354084014893,
      "learning_rate": 3.3564336775866654e-05,
      "loss": 2.377099800109863,
      "memory(GiB)": 77.56,
      "step": 70795,
      "token_acc": 0.47161572052401746,
      "train_speed(iter/s)": 1.437356
    },
    {
      "epoch": 3.0332890621652884,
      "grad_norm": 4.616634845733643,
      "learning_rate": 3.35579811267516e-05,
      "loss": 2.4821651458740233,
      "memory(GiB)": 77.56,
      "step": 70800,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 3.0335032774945376,
      "grad_norm": 4.715622901916504,
      "learning_rate": 3.355162577549913e-05,
      "loss": 1.9905763626098634,
      "memory(GiB)": 77.56,
      "step": 70805,
      "token_acc": 0.5328185328185329,
      "train_speed(iter/s)": 1.43738
    },
    {
      "epoch": 3.0337174928237864,
      "grad_norm": 4.912971496582031,
      "learning_rate": 3.35452707222244e-05,
      "loss": 2.2994430541992186,
      "memory(GiB)": 77.56,
      "step": 70810,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 3.0339317081530353,
      "grad_norm": 4.222915172576904,
      "learning_rate": 3.353891596704252e-05,
      "loss": 2.3573863983154295,
      "memory(GiB)": 77.56,
      "step": 70815,
      "token_acc": 0.5258064516129032,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 3.0341459234822845,
      "grad_norm": 4.757033824920654,
      "learning_rate": 3.353256151006863e-05,
      "loss": 2.629930877685547,
      "memory(GiB)": 77.56,
      "step": 70820,
      "token_acc": 0.44518272425249167,
      "train_speed(iter/s)": 1.437393
    },
    {
      "epoch": 3.0343601388115333,
      "grad_norm": 5.285696506500244,
      "learning_rate": 3.3526207351417825e-05,
      "loss": 2.3617713928222654,
      "memory(GiB)": 77.56,
      "step": 70825,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.437388
    },
    {
      "epoch": 3.034574354140782,
      "grad_norm": 6.275881767272949,
      "learning_rate": 3.351985349120523e-05,
      "loss": 2.3016841888427733,
      "memory(GiB)": 77.56,
      "step": 70830,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.437381
    },
    {
      "epoch": 3.0347885694700314,
      "grad_norm": 6.186609268188477,
      "learning_rate": 3.351349992954595e-05,
      "loss": 2.4427631378173826,
      "memory(GiB)": 77.56,
      "step": 70835,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.437345
    },
    {
      "epoch": 3.03500278479928,
      "grad_norm": 4.848016738891602,
      "learning_rate": 3.350714666655509e-05,
      "loss": 2.1069488525390625,
      "memory(GiB)": 77.56,
      "step": 70840,
      "token_acc": 0.5821917808219178,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 3.035217000128529,
      "grad_norm": 6.821707248687744,
      "learning_rate": 3.3500793702347716e-05,
      "loss": 2.554111862182617,
      "memory(GiB)": 77.56,
      "step": 70845,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.437352
    },
    {
      "epoch": 3.0354312154577783,
      "grad_norm": 5.950802326202393,
      "learning_rate": 3.349444103703895e-05,
      "loss": 2.418069076538086,
      "memory(GiB)": 77.56,
      "step": 70850,
      "token_acc": 0.5179856115107914,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 3.035645430787027,
      "grad_norm": 5.721039295196533,
      "learning_rate": 3.3488088670743886e-05,
      "loss": 2.1435009002685548,
      "memory(GiB)": 77.56,
      "step": 70855,
      "token_acc": 0.5095541401273885,
      "train_speed(iter/s)": 1.43737
    },
    {
      "epoch": 3.035859646116276,
      "grad_norm": 5.344554424285889,
      "learning_rate": 3.3481736603577565e-05,
      "loss": 2.1546920776367187,
      "memory(GiB)": 77.56,
      "step": 70860,
      "token_acc": 0.555984555984556,
      "train_speed(iter/s)": 1.437373
    },
    {
      "epoch": 3.036073861445525,
      "grad_norm": 6.1356353759765625,
      "learning_rate": 3.34753848356551e-05,
      "loss": 2.4321304321289063,
      "memory(GiB)": 77.56,
      "step": 70865,
      "token_acc": 0.4784172661870504,
      "train_speed(iter/s)": 1.437384
    },
    {
      "epoch": 3.036288076774774,
      "grad_norm": 5.542874336242676,
      "learning_rate": 3.3469033367091534e-05,
      "loss": 2.246772575378418,
      "memory(GiB)": 77.56,
      "step": 70870,
      "token_acc": 0.5303643724696356,
      "train_speed(iter/s)": 1.437405
    },
    {
      "epoch": 3.036502292104023,
      "grad_norm": 5.28950309753418,
      "learning_rate": 3.3462682198001925e-05,
      "loss": 2.7895410537719725,
      "memory(GiB)": 77.56,
      "step": 70875,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.437412
    },
    {
      "epoch": 3.036716507433272,
      "grad_norm": 8.909818649291992,
      "learning_rate": 3.345633132850136e-05,
      "loss": 2.134360122680664,
      "memory(GiB)": 77.56,
      "step": 70880,
      "token_acc": 0.524822695035461,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 3.036930722762521,
      "grad_norm": 6.036808490753174,
      "learning_rate": 3.344998075870487e-05,
      "loss": 2.3594997406005858,
      "memory(GiB)": 77.56,
      "step": 70885,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.437417
    },
    {
      "epoch": 3.0371449380917697,
      "grad_norm": 7.076190948486328,
      "learning_rate": 3.344363048872748e-05,
      "loss": 2.3516359329223633,
      "memory(GiB)": 77.56,
      "step": 70890,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.437441
    },
    {
      "epoch": 3.037359153421019,
      "grad_norm": 9.171173095703125,
      "learning_rate": 3.3437280518684275e-05,
      "loss": 2.3217939376831054,
      "memory(GiB)": 77.56,
      "step": 70895,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.437441
    },
    {
      "epoch": 3.0375733687502677,
      "grad_norm": 4.8570685386657715,
      "learning_rate": 3.3430930848690266e-05,
      "loss": 2.5207239151000977,
      "memory(GiB)": 77.56,
      "step": 70900,
      "token_acc": 0.47493403693931396,
      "train_speed(iter/s)": 1.437459
    },
    {
      "epoch": 3.0377875840795165,
      "grad_norm": 6.4781317710876465,
      "learning_rate": 3.3424581478860495e-05,
      "loss": 2.1019229888916016,
      "memory(GiB)": 77.56,
      "step": 70905,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.038001799408766,
      "grad_norm": 5.527780055999756,
      "learning_rate": 3.3418232409309975e-05,
      "loss": 2.089214897155762,
      "memory(GiB)": 77.56,
      "step": 70910,
      "token_acc": 0.5444839857651246,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.0382160147380146,
      "grad_norm": 6.387155055999756,
      "learning_rate": 3.341188364015372e-05,
      "loss": 2.4615354537963867,
      "memory(GiB)": 77.56,
      "step": 70915,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.437498
    },
    {
      "epoch": 3.0384302300672634,
      "grad_norm": 6.141441345214844,
      "learning_rate": 3.340553517150674e-05,
      "loss": 2.3892189025878907,
      "memory(GiB)": 77.56,
      "step": 70920,
      "token_acc": 0.4533898305084746,
      "train_speed(iter/s)": 1.43751
    },
    {
      "epoch": 3.0386444453965127,
      "grad_norm": 6.831070899963379,
      "learning_rate": 3.339918700348409e-05,
      "loss": 2.385594367980957,
      "memory(GiB)": 77.56,
      "step": 70925,
      "token_acc": 0.478125,
      "train_speed(iter/s)": 1.437534
    },
    {
      "epoch": 3.0388586607257615,
      "grad_norm": 5.203241348266602,
      "learning_rate": 3.339283913620073e-05,
      "loss": 2.3997777938842773,
      "memory(GiB)": 77.56,
      "step": 70930,
      "token_acc": 0.5078369905956113,
      "train_speed(iter/s)": 1.437515
    },
    {
      "epoch": 3.0390728760550103,
      "grad_norm": 6.753215312957764,
      "learning_rate": 3.338649156977166e-05,
      "loss": 2.4418935775756836,
      "memory(GiB)": 77.56,
      "step": 70935,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.437507
    },
    {
      "epoch": 3.0392870913842596,
      "grad_norm": 6.383613109588623,
      "learning_rate": 3.338014430431189e-05,
      "loss": 2.1237998962402345,
      "memory(GiB)": 77.56,
      "step": 70940,
      "token_acc": 0.5335968379446641,
      "train_speed(iter/s)": 1.43749
    },
    {
      "epoch": 3.0395013067135084,
      "grad_norm": 6.2430500984191895,
      "learning_rate": 3.33737973399364e-05,
      "loss": 2.531260108947754,
      "memory(GiB)": 77.56,
      "step": 70945,
      "token_acc": 0.48554913294797686,
      "train_speed(iter/s)": 1.437493
    },
    {
      "epoch": 3.039715522042757,
      "grad_norm": 6.0232625007629395,
      "learning_rate": 3.336745067676015e-05,
      "loss": 2.4142019271850588,
      "memory(GiB)": 77.56,
      "step": 70950,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.437496
    },
    {
      "epoch": 3.0399297373720064,
      "grad_norm": 5.738032817840576,
      "learning_rate": 3.336110431489815e-05,
      "loss": 2.1952428817749023,
      "memory(GiB)": 77.56,
      "step": 70955,
      "token_acc": 0.4899598393574297,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 3.0401439527012553,
      "grad_norm": 5.492711544036865,
      "learning_rate": 3.3354758254465346e-05,
      "loss": 2.259307861328125,
      "memory(GiB)": 77.56,
      "step": 70960,
      "token_acc": 0.548951048951049,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 3.040358168030504,
      "grad_norm": 5.665395736694336,
      "learning_rate": 3.334841249557672e-05,
      "loss": 2.469676208496094,
      "memory(GiB)": 77.56,
      "step": 70965,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.437549
    },
    {
      "epoch": 3.0405723833597533,
      "grad_norm": 5.712028980255127,
      "learning_rate": 3.334206703834721e-05,
      "loss": 2.2087820053100584,
      "memory(GiB)": 77.56,
      "step": 70970,
      "token_acc": 0.5034722222222222,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 3.040786598689002,
      "grad_norm": 5.499817371368408,
      "learning_rate": 3.333572188289179e-05,
      "loss": 1.9762527465820312,
      "memory(GiB)": 77.56,
      "step": 70975,
      "token_acc": 0.5608856088560885,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 3.041000814018251,
      "grad_norm": 4.6744818687438965,
      "learning_rate": 3.3329377029325396e-05,
      "loss": 2.1812816619873048,
      "memory(GiB)": 77.56,
      "step": 70980,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.437603
    },
    {
      "epoch": 3.0412150293475,
      "grad_norm": 6.415157318115234,
      "learning_rate": 3.3323032477762984e-05,
      "loss": 2.1204456329345702,
      "memory(GiB)": 77.56,
      "step": 70985,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.437611
    },
    {
      "epoch": 3.041429244676749,
      "grad_norm": 5.196455955505371,
      "learning_rate": 3.331668822831947e-05,
      "loss": 2.221500587463379,
      "memory(GiB)": 77.56,
      "step": 70990,
      "token_acc": 0.5056603773584906,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 3.041643460005998,
      "grad_norm": 4.392649173736572,
      "learning_rate": 3.331034428110979e-05,
      "loss": 2.1638776779174806,
      "memory(GiB)": 77.56,
      "step": 70995,
      "token_acc": 0.5234375,
      "train_speed(iter/s)": 1.437644
    },
    {
      "epoch": 3.041857675335247,
      "grad_norm": 9.233529090881348,
      "learning_rate": 3.33040006362489e-05,
      "loss": 2.483344078063965,
      "memory(GiB)": 77.56,
      "step": 71000,
      "token_acc": 0.49624060150375937,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 3.041857675335247,
      "eval_loss": 2.188765525817871,
      "eval_runtime": 14.8012,
      "eval_samples_per_second": 6.756,
      "eval_steps_per_second": 6.756,
      "eval_token_acc": 0.49670619235836627,
      "step": 71000
    },
    {
      "epoch": 3.042071890664496,
      "grad_norm": 6.377941131591797,
      "learning_rate": 3.329765729385169e-05,
      "loss": 2.4590631484985352,
      "memory(GiB)": 77.56,
      "step": 71005,
      "token_acc": 0.491412213740458,
      "train_speed(iter/s)": 1.437235
    },
    {
      "epoch": 3.0422861059937447,
      "grad_norm": 5.215327739715576,
      "learning_rate": 3.329131425403309e-05,
      "loss": 2.3740257263183593,
      "memory(GiB)": 77.56,
      "step": 71010,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.437246
    },
    {
      "epoch": 3.042500321322994,
      "grad_norm": 5.504948616027832,
      "learning_rate": 3.3284971516908014e-05,
      "loss": 2.3258388519287108,
      "memory(GiB)": 77.56,
      "step": 71015,
      "token_acc": 0.49828178694158076,
      "train_speed(iter/s)": 1.437224
    },
    {
      "epoch": 3.042714536652243,
      "grad_norm": 5.201958179473877,
      "learning_rate": 3.327862908259135e-05,
      "loss": 2.282832145690918,
      "memory(GiB)": 77.56,
      "step": 71020,
      "token_acc": 0.5343511450381679,
      "train_speed(iter/s)": 1.4372
    },
    {
      "epoch": 3.0429287519814916,
      "grad_norm": 4.988426208496094,
      "learning_rate": 3.327228695119801e-05,
      "loss": 2.3858795166015625,
      "memory(GiB)": 77.56,
      "step": 71025,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.43719
    },
    {
      "epoch": 3.043142967310741,
      "grad_norm": 6.245116710662842,
      "learning_rate": 3.326594512284289e-05,
      "loss": 2.2175689697265626,
      "memory(GiB)": 77.56,
      "step": 71030,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.437201
    },
    {
      "epoch": 3.0433571826399897,
      "grad_norm": 6.072200775146484,
      "learning_rate": 3.325960359764085e-05,
      "loss": 2.4529499053955077,
      "memory(GiB)": 77.56,
      "step": 71035,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437188
    },
    {
      "epoch": 3.0435713979692385,
      "grad_norm": 6.2255964279174805,
      "learning_rate": 3.3253262375706815e-05,
      "loss": 2.3310726165771483,
      "memory(GiB)": 77.56,
      "step": 71040,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.437182
    },
    {
      "epoch": 3.0437856132984877,
      "grad_norm": 4.321385383605957,
      "learning_rate": 3.3246921457155634e-05,
      "loss": 2.379546356201172,
      "memory(GiB)": 77.56,
      "step": 71045,
      "token_acc": 0.4759036144578313,
      "train_speed(iter/s)": 1.437176
    },
    {
      "epoch": 3.0439998286277365,
      "grad_norm": 6.137400150299072,
      "learning_rate": 3.3240580842102186e-05,
      "loss": 2.3142412185668944,
      "memory(GiB)": 77.56,
      "step": 71050,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.437181
    },
    {
      "epoch": 3.0442140439569854,
      "grad_norm": 5.424540042877197,
      "learning_rate": 3.323424053066135e-05,
      "loss": 2.321827697753906,
      "memory(GiB)": 77.56,
      "step": 71055,
      "token_acc": 0.5061349693251533,
      "train_speed(iter/s)": 1.437196
    },
    {
      "epoch": 3.0444282592862346,
      "grad_norm": 5.927058219909668,
      "learning_rate": 3.322790052294796e-05,
      "loss": 2.2751163482666015,
      "memory(GiB)": 77.56,
      "step": 71060,
      "token_acc": 0.5113636363636364,
      "train_speed(iter/s)": 1.437217
    },
    {
      "epoch": 3.0446424746154834,
      "grad_norm": 5.897186756134033,
      "learning_rate": 3.3221560819076894e-05,
      "loss": 2.4224424362182617,
      "memory(GiB)": 77.56,
      "step": 71065,
      "token_acc": 0.5214285714285715,
      "train_speed(iter/s)": 1.437222
    },
    {
      "epoch": 3.0448566899447322,
      "grad_norm": 4.345804691314697,
      "learning_rate": 3.3215221419162966e-05,
      "loss": 2.2936595916748046,
      "memory(GiB)": 77.56,
      "step": 71070,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.437234
    },
    {
      "epoch": 3.0450709052739815,
      "grad_norm": 5.58962345123291,
      "learning_rate": 3.320888232332108e-05,
      "loss": 2.4791725158691404,
      "memory(GiB)": 77.56,
      "step": 71075,
      "token_acc": 0.5016835016835017,
      "train_speed(iter/s)": 1.437223
    },
    {
      "epoch": 3.0452851206032303,
      "grad_norm": 7.8404717445373535,
      "learning_rate": 3.320254353166602e-05,
      "loss": 2.495720100402832,
      "memory(GiB)": 77.56,
      "step": 71080,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437235
    },
    {
      "epoch": 3.045499335932479,
      "grad_norm": 5.145702362060547,
      "learning_rate": 3.3196205044312664e-05,
      "loss": 2.3742069244384765,
      "memory(GiB)": 77.56,
      "step": 71085,
      "token_acc": 0.5045317220543807,
      "train_speed(iter/s)": 1.437247
    },
    {
      "epoch": 3.0457135512617284,
      "grad_norm": 8.874959945678711,
      "learning_rate": 3.318986686137581e-05,
      "loss": 2.489121437072754,
      "memory(GiB)": 77.56,
      "step": 71090,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437264
    },
    {
      "epoch": 3.045927766590977,
      "grad_norm": 11.734989166259766,
      "learning_rate": 3.318352898297027e-05,
      "loss": 2.620874214172363,
      "memory(GiB)": 77.56,
      "step": 71095,
      "token_acc": 0.4557522123893805,
      "train_speed(iter/s)": 1.437257
    },
    {
      "epoch": 3.046141981920226,
      "grad_norm": 6.317271709442139,
      "learning_rate": 3.31771914092109e-05,
      "loss": 2.4212602615356444,
      "memory(GiB)": 77.56,
      "step": 71100,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.437278
    },
    {
      "epoch": 3.0463561972494753,
      "grad_norm": 5.987992763519287,
      "learning_rate": 3.317085414021247e-05,
      "loss": 2.331417274475098,
      "memory(GiB)": 77.56,
      "step": 71105,
      "token_acc": 0.5244299674267101,
      "train_speed(iter/s)": 1.437277
    },
    {
      "epoch": 3.046570412578724,
      "grad_norm": 4.15399694442749,
      "learning_rate": 3.316451717608982e-05,
      "loss": 2.127524566650391,
      "memory(GiB)": 77.56,
      "step": 71110,
      "token_acc": 0.558303886925795,
      "train_speed(iter/s)": 1.437307
    },
    {
      "epoch": 3.046784627907973,
      "grad_norm": 5.694718360900879,
      "learning_rate": 3.3158180516957736e-05,
      "loss": 2.223400115966797,
      "memory(GiB)": 77.56,
      "step": 71115,
      "token_acc": 0.4984520123839009,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 3.046998843237222,
      "grad_norm": 7.442710876464844,
      "learning_rate": 3.315184416293099e-05,
      "loss": 2.4708282470703127,
      "memory(GiB)": 77.56,
      "step": 71120,
      "token_acc": 0.45318352059925093,
      "train_speed(iter/s)": 1.437296
    },
    {
      "epoch": 3.047213058566471,
      "grad_norm": 4.6826558113098145,
      "learning_rate": 3.314550811412441e-05,
      "loss": 2.183447265625,
      "memory(GiB)": 77.56,
      "step": 71125,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.437313
    },
    {
      "epoch": 3.0474272738957198,
      "grad_norm": 7.514987945556641,
      "learning_rate": 3.313917237065277e-05,
      "loss": 2.3281620025634764,
      "memory(GiB)": 77.56,
      "step": 71130,
      "token_acc": 0.5201342281879194,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 3.047641489224969,
      "grad_norm": 6.463863849639893,
      "learning_rate": 3.313283693263082e-05,
      "loss": 2.2465572357177734,
      "memory(GiB)": 77.56,
      "step": 71135,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.437328
    },
    {
      "epoch": 3.047855704554218,
      "grad_norm": 6.358994007110596,
      "learning_rate": 3.3126501800173334e-05,
      "loss": 2.250709342956543,
      "memory(GiB)": 77.56,
      "step": 71140,
      "token_acc": 0.5029239766081871,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 3.0480699198834666,
      "grad_norm": 6.5687336921691895,
      "learning_rate": 3.312016697339513e-05,
      "loss": 2.2080333709716795,
      "memory(GiB)": 77.56,
      "step": 71145,
      "token_acc": 0.5435540069686411,
      "train_speed(iter/s)": 1.437377
    },
    {
      "epoch": 3.048284135212716,
      "grad_norm": 6.290611267089844,
      "learning_rate": 3.311383245241092e-05,
      "loss": 2.416961097717285,
      "memory(GiB)": 77.56,
      "step": 71150,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.437392
    },
    {
      "epoch": 3.0484983505419647,
      "grad_norm": 5.255943775177002,
      "learning_rate": 3.3107498237335476e-05,
      "loss": 2.1069141387939454,
      "memory(GiB)": 77.56,
      "step": 71155,
      "token_acc": 0.5212121212121212,
      "train_speed(iter/s)": 1.437377
    },
    {
      "epoch": 3.0487125658712135,
      "grad_norm": 5.697054862976074,
      "learning_rate": 3.3101164328283566e-05,
      "loss": 2.2210378646850586,
      "memory(GiB)": 77.56,
      "step": 71160,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.048926781200463,
      "grad_norm": 5.799210071563721,
      "learning_rate": 3.3094830725369894e-05,
      "loss": 2.4000701904296875,
      "memory(GiB)": 77.56,
      "step": 71165,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.437381
    },
    {
      "epoch": 3.0491409965297116,
      "grad_norm": 6.0237836837768555,
      "learning_rate": 3.308849742870923e-05,
      "loss": 2.0678529739379883,
      "memory(GiB)": 77.56,
      "step": 71170,
      "token_acc": 0.5566666666666666,
      "train_speed(iter/s)": 1.437354
    },
    {
      "epoch": 3.0493552118589604,
      "grad_norm": 6.069119930267334,
      "learning_rate": 3.3082164438416306e-05,
      "loss": 2.3057491302490236,
      "memory(GiB)": 77.56,
      "step": 71175,
      "token_acc": 0.5184049079754601,
      "train_speed(iter/s)": 1.437367
    },
    {
      "epoch": 3.0495694271882097,
      "grad_norm": 7.07837438583374,
      "learning_rate": 3.3075831754605834e-05,
      "loss": 2.2990318298339845,
      "memory(GiB)": 77.56,
      "step": 71180,
      "token_acc": 0.5418060200668896,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.0497836425174585,
      "grad_norm": 7.270720481872559,
      "learning_rate": 3.306949937739255e-05,
      "loss": 2.0276330947875976,
      "memory(GiB)": 77.56,
      "step": 71185,
      "token_acc": 0.5350877192982456,
      "train_speed(iter/s)": 1.437378
    },
    {
      "epoch": 3.0499978578467073,
      "grad_norm": 5.985006809234619,
      "learning_rate": 3.3063167306891165e-05,
      "loss": 2.456471252441406,
      "memory(GiB)": 77.56,
      "step": 71190,
      "token_acc": 0.4768211920529801,
      "train_speed(iter/s)": 1.437391
    },
    {
      "epoch": 3.0502120731759566,
      "grad_norm": 7.572167873382568,
      "learning_rate": 3.30568355432164e-05,
      "loss": 2.4999338150024415,
      "memory(GiB)": 77.56,
      "step": 71195,
      "token_acc": 0.5390334572490706,
      "train_speed(iter/s)": 1.437416
    },
    {
      "epoch": 3.0504262885052054,
      "grad_norm": 6.423421859741211,
      "learning_rate": 3.305050408648295e-05,
      "loss": 2.3983505249023436,
      "memory(GiB)": 77.56,
      "step": 71200,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.437418
    },
    {
      "epoch": 3.050640503834454,
      "grad_norm": 5.355645179748535,
      "learning_rate": 3.3044172936805505e-05,
      "loss": 2.467995834350586,
      "memory(GiB)": 77.56,
      "step": 71205,
      "token_acc": 0.4703703703703704,
      "train_speed(iter/s)": 1.437441
    },
    {
      "epoch": 3.0508547191637034,
      "grad_norm": 5.814355373382568,
      "learning_rate": 3.3037842094298795e-05,
      "loss": 2.3652368545532227,
      "memory(GiB)": 77.56,
      "step": 71210,
      "token_acc": 0.5646551724137931,
      "train_speed(iter/s)": 1.437455
    },
    {
      "epoch": 3.0510689344929522,
      "grad_norm": 5.717180252075195,
      "learning_rate": 3.303151155907746e-05,
      "loss": 2.660415840148926,
      "memory(GiB)": 77.56,
      "step": 71215,
      "token_acc": 0.43312101910828027,
      "train_speed(iter/s)": 1.437443
    },
    {
      "epoch": 3.0512831498222015,
      "grad_norm": 5.1082024574279785,
      "learning_rate": 3.302518133125623e-05,
      "loss": 2.3493471145629883,
      "memory(GiB)": 77.56,
      "step": 71220,
      "token_acc": 0.5241157556270096,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 3.0514973651514503,
      "grad_norm": 5.413752555847168,
      "learning_rate": 3.301885141094975e-05,
      "loss": 2.399314308166504,
      "memory(GiB)": 77.56,
      "step": 71225,
      "token_acc": 0.47878787878787876,
      "train_speed(iter/s)": 1.437466
    },
    {
      "epoch": 3.051711580480699,
      "grad_norm": 5.756601333618164,
      "learning_rate": 3.3012521798272723e-05,
      "loss": 2.715800476074219,
      "memory(GiB)": 77.56,
      "step": 71230,
      "token_acc": 0.4437299035369775,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.0519257958099484,
      "grad_norm": 7.276020526885986,
      "learning_rate": 3.3006192493339785e-05,
      "loss": 1.7715030670166017,
      "memory(GiB)": 77.56,
      "step": 71235,
      "token_acc": 0.5675675675675675,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 3.052140011139197,
      "grad_norm": 5.446733474731445,
      "learning_rate": 3.299986349626563e-05,
      "loss": 2.161296844482422,
      "memory(GiB)": 77.56,
      "step": 71240,
      "token_acc": 0.5048231511254019,
      "train_speed(iter/s)": 1.437524
    },
    {
      "epoch": 3.052354226468446,
      "grad_norm": 5.676682472229004,
      "learning_rate": 3.299353480716488e-05,
      "loss": 2.6087387084960936,
      "memory(GiB)": 77.56,
      "step": 71245,
      "token_acc": 0.4586206896551724,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 3.0525684417976953,
      "grad_norm": 4.904717922210693,
      "learning_rate": 3.298720642615221e-05,
      "loss": 1.9600635528564454,
      "memory(GiB)": 77.56,
      "step": 71250,
      "token_acc": 0.5377777777777778,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 3.052782657126944,
      "grad_norm": 6.122511863708496,
      "learning_rate": 3.298087835334225e-05,
      "loss": 2.3314563751220705,
      "memory(GiB)": 77.56,
      "step": 71255,
      "token_acc": 0.5054945054945055,
      "train_speed(iter/s)": 1.437549
    },
    {
      "epoch": 3.052996872456193,
      "grad_norm": 5.856444835662842,
      "learning_rate": 3.2974550588849645e-05,
      "loss": 2.398816680908203,
      "memory(GiB)": 77.56,
      "step": 71260,
      "token_acc": 0.4753521126760563,
      "train_speed(iter/s)": 1.437558
    },
    {
      "epoch": 3.053211087785442,
      "grad_norm": 5.47257137298584,
      "learning_rate": 3.296822313278902e-05,
      "loss": 2.5130504608154296,
      "memory(GiB)": 77.56,
      "step": 71265,
      "token_acc": 0.47093023255813954,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 3.053425303114691,
      "grad_norm": 6.759922027587891,
      "learning_rate": 3.296189598527502e-05,
      "loss": 2.0901737213134766,
      "memory(GiB)": 77.56,
      "step": 71270,
      "token_acc": 0.5407407407407407,
      "train_speed(iter/s)": 1.437544
    },
    {
      "epoch": 3.0536395184439398,
      "grad_norm": 4.901780605316162,
      "learning_rate": 3.295556914642226e-05,
      "loss": 2.340744209289551,
      "memory(GiB)": 77.56,
      "step": 71275,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.437576
    },
    {
      "epoch": 3.053853733773189,
      "grad_norm": 6.798420429229736,
      "learning_rate": 3.294924261634534e-05,
      "loss": 2.2466558456420898,
      "memory(GiB)": 77.56,
      "step": 71280,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 3.054067949102438,
      "grad_norm": 5.200614929199219,
      "learning_rate": 3.29429163951589e-05,
      "loss": 2.4750911712646486,
      "memory(GiB)": 77.56,
      "step": 71285,
      "token_acc": 0.4980544747081712,
      "train_speed(iter/s)": 1.437614
    },
    {
      "epoch": 3.0542821644316867,
      "grad_norm": 9.178266525268555,
      "learning_rate": 3.29365904829775e-05,
      "loss": 2.514382553100586,
      "memory(GiB)": 77.56,
      "step": 71290,
      "token_acc": 0.4720496894409938,
      "train_speed(iter/s)": 1.43763
    },
    {
      "epoch": 3.054496379760936,
      "grad_norm": 4.766347885131836,
      "learning_rate": 3.293026487991579e-05,
      "loss": 2.4030633926391602,
      "memory(GiB)": 77.56,
      "step": 71295,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.437649
    },
    {
      "epoch": 3.0547105950901847,
      "grad_norm": 5.82194185256958,
      "learning_rate": 3.2923939586088346e-05,
      "loss": 2.2792917251586915,
      "memory(GiB)": 77.56,
      "step": 71300,
      "token_acc": 0.5111821086261981,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 3.0549248104194335,
      "grad_norm": 6.4278974533081055,
      "learning_rate": 3.291761460160976e-05,
      "loss": 2.531342124938965,
      "memory(GiB)": 77.56,
      "step": 71305,
      "token_acc": 0.44404332129963897,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 3.055139025748683,
      "grad_norm": 7.563528060913086,
      "learning_rate": 3.2911289926594604e-05,
      "loss": 2.0771963119506838,
      "memory(GiB)": 77.56,
      "step": 71310,
      "token_acc": 0.5430711610486891,
      "train_speed(iter/s)": 1.437646
    },
    {
      "epoch": 3.0553532410779316,
      "grad_norm": 4.607213020324707,
      "learning_rate": 3.2904965561157466e-05,
      "loss": 2.1626506805419923,
      "memory(GiB)": 77.56,
      "step": 71315,
      "token_acc": 0.49523809523809526,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.0555674564071804,
      "grad_norm": 7.028500080108643,
      "learning_rate": 3.289864150541292e-05,
      "loss": 2.5822845458984376,
      "memory(GiB)": 77.56,
      "step": 71320,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437698
    },
    {
      "epoch": 3.0557816717364297,
      "grad_norm": 6.545118808746338,
      "learning_rate": 3.289231775947551e-05,
      "loss": 2.5390146255493162,
      "memory(GiB)": 77.56,
      "step": 71325,
      "token_acc": 0.46446700507614214,
      "train_speed(iter/s)": 1.437717
    },
    {
      "epoch": 3.0559958870656785,
      "grad_norm": 5.889909267425537,
      "learning_rate": 3.2885994323459826e-05,
      "loss": 2.555874824523926,
      "memory(GiB)": 77.56,
      "step": 71330,
      "token_acc": 0.4633431085043988,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 3.0562101023949273,
      "grad_norm": 5.655017375946045,
      "learning_rate": 3.287967119748039e-05,
      "loss": 2.270869827270508,
      "memory(GiB)": 77.56,
      "step": 71335,
      "token_acc": 0.4904214559386973,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 3.0564243177241766,
      "grad_norm": 6.0177106857299805,
      "learning_rate": 3.287334838165179e-05,
      "loss": 2.3454551696777344,
      "memory(GiB)": 77.56,
      "step": 71340,
      "token_acc": 0.5033783783783784,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 3.0566385330534254,
      "grad_norm": 6.671818256378174,
      "learning_rate": 3.286702587608854e-05,
      "loss": 2.1206298828125,
      "memory(GiB)": 77.56,
      "step": 71345,
      "token_acc": 0.4866920152091255,
      "train_speed(iter/s)": 1.437738
    },
    {
      "epoch": 3.056852748382674,
      "grad_norm": 5.391358852386475,
      "learning_rate": 3.286070368090519e-05,
      "loss": 1.9761165618896483,
      "memory(GiB)": 77.56,
      "step": 71350,
      "token_acc": 0.5505226480836237,
      "train_speed(iter/s)": 1.437742
    },
    {
      "epoch": 3.0570669637119234,
      "grad_norm": 5.055149555206299,
      "learning_rate": 3.285438179621627e-05,
      "loss": 2.3021852493286135,
      "memory(GiB)": 77.56,
      "step": 71355,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.43775
    },
    {
      "epoch": 3.0572811790411722,
      "grad_norm": 7.399731159210205,
      "learning_rate": 3.284806022213632e-05,
      "loss": 2.6837385177612303,
      "memory(GiB)": 77.56,
      "step": 71360,
      "token_acc": 0.44554455445544555,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 3.057495394370421,
      "grad_norm": 5.64953088760376,
      "learning_rate": 3.284173895877982e-05,
      "loss": 2.4469409942626954,
      "memory(GiB)": 77.56,
      "step": 71365,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 3.0577096096996703,
      "grad_norm": 4.701729774475098,
      "learning_rate": 3.283541800626132e-05,
      "loss": 2.144628143310547,
      "memory(GiB)": 77.56,
      "step": 71370,
      "token_acc": 0.525691699604743,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 3.057923825028919,
      "grad_norm": 6.855034351348877,
      "learning_rate": 3.282909736469535e-05,
      "loss": 2.5289798736572267,
      "memory(GiB)": 77.56,
      "step": 71375,
      "token_acc": 0.4766355140186916,
      "train_speed(iter/s)": 1.437699
    },
    {
      "epoch": 3.058138040358168,
      "grad_norm": 8.932657241821289,
      "learning_rate": 3.2822777034196364e-05,
      "loss": 2.2341377258300783,
      "memory(GiB)": 77.56,
      "step": 71380,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 3.058352255687417,
      "grad_norm": 6.1834635734558105,
      "learning_rate": 3.28164570148789e-05,
      "loss": 2.301803779602051,
      "memory(GiB)": 77.56,
      "step": 71385,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.437716
    },
    {
      "epoch": 3.058566471016666,
      "grad_norm": 6.1221394538879395,
      "learning_rate": 3.281013730685744e-05,
      "loss": 2.253335189819336,
      "memory(GiB)": 77.56,
      "step": 71390,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437707
    },
    {
      "epoch": 3.058780686345915,
      "grad_norm": 6.424152851104736,
      "learning_rate": 3.280381791024646e-05,
      "loss": 2.4969079971313475,
      "memory(GiB)": 77.56,
      "step": 71395,
      "token_acc": 0.4657534246575342,
      "train_speed(iter/s)": 1.437701
    },
    {
      "epoch": 3.058994901675164,
      "grad_norm": 5.603373050689697,
      "learning_rate": 3.2797498825160456e-05,
      "loss": 2.539145088195801,
      "memory(GiB)": 77.56,
      "step": 71400,
      "token_acc": 0.4753246753246753,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 3.059209117004413,
      "grad_norm": 6.993089199066162,
      "learning_rate": 3.279118005171389e-05,
      "loss": 2.6204206466674806,
      "memory(GiB)": 77.56,
      "step": 71405,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 3.0594233323336617,
      "grad_norm": 4.646661758422852,
      "learning_rate": 3.278486159002124e-05,
      "loss": 2.394352912902832,
      "memory(GiB)": 77.56,
      "step": 71410,
      "token_acc": 0.5110294117647058,
      "train_speed(iter/s)": 1.437686
    },
    {
      "epoch": 3.059637547662911,
      "grad_norm": 6.484498023986816,
      "learning_rate": 3.277854344019699e-05,
      "loss": 2.381487274169922,
      "memory(GiB)": 77.56,
      "step": 71415,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.437679
    },
    {
      "epoch": 3.0598517629921598,
      "grad_norm": 5.069632530212402,
      "learning_rate": 3.2772225602355554e-05,
      "loss": 2.4179157257080077,
      "memory(GiB)": 77.56,
      "step": 71420,
      "token_acc": 0.5048231511254019,
      "train_speed(iter/s)": 1.437689
    },
    {
      "epoch": 3.0600659783214086,
      "grad_norm": 6.545305252075195,
      "learning_rate": 3.2765908076611426e-05,
      "loss": 2.190912628173828,
      "memory(GiB)": 77.56,
      "step": 71425,
      "token_acc": 0.4641350210970464,
      "train_speed(iter/s)": 1.437677
    },
    {
      "epoch": 3.060280193650658,
      "grad_norm": 6.894156455993652,
      "learning_rate": 3.275959086307905e-05,
      "loss": 2.5124589920043947,
      "memory(GiB)": 77.56,
      "step": 71430,
      "token_acc": 0.4553846153846154,
      "train_speed(iter/s)": 1.437692
    },
    {
      "epoch": 3.0604944089799067,
      "grad_norm": 8.655752182006836,
      "learning_rate": 3.2753273961872824e-05,
      "loss": 2.5310384750366213,
      "memory(GiB)": 77.56,
      "step": 71435,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.437707
    },
    {
      "epoch": 3.0607086243091555,
      "grad_norm": 5.756744861602783,
      "learning_rate": 3.274695737310723e-05,
      "loss": 2.5682937622070314,
      "memory(GiB)": 77.56,
      "step": 71440,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.437708
    },
    {
      "epoch": 3.0609228396384047,
      "grad_norm": 6.639272689819336,
      "learning_rate": 3.2740641096896715e-05,
      "loss": 2.3876296997070314,
      "memory(GiB)": 77.56,
      "step": 71445,
      "token_acc": 0.5389408099688473,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 3.0611370549676535,
      "grad_norm": 5.637764930725098,
      "learning_rate": 3.273432513335566e-05,
      "loss": 2.158272933959961,
      "memory(GiB)": 77.56,
      "step": 71450,
      "token_acc": 0.5692307692307692,
      "train_speed(iter/s)": 1.437744
    },
    {
      "epoch": 3.0613512702969023,
      "grad_norm": 5.608621597290039,
      "learning_rate": 3.27280094825985e-05,
      "loss": 2.142215347290039,
      "memory(GiB)": 77.56,
      "step": 71455,
      "token_acc": 0.53515625,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 3.0615654856261516,
      "grad_norm": 5.965383529663086,
      "learning_rate": 3.2721694144739654e-05,
      "loss": 2.2114877700805664,
      "memory(GiB)": 77.56,
      "step": 71460,
      "token_acc": 0.5206896551724138,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 3.0617797009554004,
      "grad_norm": 8.144937515258789,
      "learning_rate": 3.2715379119893534e-05,
      "loss": 2.2537193298339844,
      "memory(GiB)": 77.56,
      "step": 71465,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 3.0619939162846492,
      "grad_norm": 6.5725016593933105,
      "learning_rate": 3.270906440817453e-05,
      "loss": 2.2744071960449217,
      "memory(GiB)": 77.56,
      "step": 71470,
      "token_acc": 0.5648854961832062,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.0622081316138985,
      "grad_norm": 5.586136817932129,
      "learning_rate": 3.270275000969704e-05,
      "loss": 2.4055850982666014,
      "memory(GiB)": 77.56,
      "step": 71475,
      "token_acc": 0.4553191489361702,
      "train_speed(iter/s)": 1.437778
    },
    {
      "epoch": 3.0624223469431473,
      "grad_norm": 6.206053256988525,
      "learning_rate": 3.2696435924575456e-05,
      "loss": 2.4393325805664063,
      "memory(GiB)": 77.56,
      "step": 71480,
      "token_acc": 0.5288135593220339,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.062636562272396,
      "grad_norm": 5.872185230255127,
      "learning_rate": 3.269012215292418e-05,
      "loss": 2.1251575469970705,
      "memory(GiB)": 77.56,
      "step": 71485,
      "token_acc": 0.5465587044534413,
      "train_speed(iter/s)": 1.437817
    },
    {
      "epoch": 3.0628507776016454,
      "grad_norm": 5.686715126037598,
      "learning_rate": 3.268380869485758e-05,
      "loss": 2.4605125427246093,
      "memory(GiB)": 77.56,
      "step": 71490,
      "token_acc": 0.49032258064516127,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 3.063064992930894,
      "grad_norm": 10.681577682495117,
      "learning_rate": 3.267749555049002e-05,
      "loss": 2.450027656555176,
      "memory(GiB)": 77.56,
      "step": 71495,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 3.063279208260143,
      "grad_norm": 5.863169193267822,
      "learning_rate": 3.2671182719935874e-05,
      "loss": 2.264522933959961,
      "memory(GiB)": 77.56,
      "step": 71500,
      "token_acc": 0.5485714285714286,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.063279208260143,
      "eval_loss": 2.1684317588806152,
      "eval_runtime": 14.4922,
      "eval_samples_per_second": 6.9,
      "eval_steps_per_second": 6.9,
      "eval_token_acc": 0.479328165374677,
      "step": 71500
    },
    {
      "epoch": 3.0634934235893923,
      "grad_norm": 6.721035957336426,
      "learning_rate": 3.266487020330953e-05,
      "loss": 2.5304203033447266,
      "memory(GiB)": 77.56,
      "step": 71505,
      "token_acc": 0.48123195380173245,
      "train_speed(iter/s)": 1.437387
    },
    {
      "epoch": 3.063707638918641,
      "grad_norm": 4.567585468292236,
      "learning_rate": 3.2658558000725295e-05,
      "loss": 2.3212718963623047,
      "memory(GiB)": 77.56,
      "step": 71510,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.437392
    },
    {
      "epoch": 3.06392185424789,
      "grad_norm": 6.7867865562438965,
      "learning_rate": 3.265224611229755e-05,
      "loss": 2.414967346191406,
      "memory(GiB)": 77.56,
      "step": 71515,
      "token_acc": 0.47701149425287354,
      "train_speed(iter/s)": 1.437377
    },
    {
      "epoch": 3.064136069577139,
      "grad_norm": 3.9770381450653076,
      "learning_rate": 3.264593453814066e-05,
      "loss": 2.435228729248047,
      "memory(GiB)": 77.56,
      "step": 71520,
      "token_acc": 0.4818181818181818,
      "train_speed(iter/s)": 1.437386
    },
    {
      "epoch": 3.064350284906388,
      "grad_norm": 6.722285747528076,
      "learning_rate": 3.2639623278368936e-05,
      "loss": 2.3312519073486326,
      "memory(GiB)": 77.56,
      "step": 71525,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.437408
    },
    {
      "epoch": 3.0645645002356368,
      "grad_norm": 5.002237796783447,
      "learning_rate": 3.263331233309673e-05,
      "loss": 2.897745132446289,
      "memory(GiB)": 77.56,
      "step": 71530,
      "token_acc": 0.44146341463414634,
      "train_speed(iter/s)": 1.437412
    },
    {
      "epoch": 3.064778715564886,
      "grad_norm": 5.541676998138428,
      "learning_rate": 3.262700170243836e-05,
      "loss": 2.232503890991211,
      "memory(GiB)": 77.56,
      "step": 71535,
      "token_acc": 0.5178571428571429,
      "train_speed(iter/s)": 1.437439
    },
    {
      "epoch": 3.064992930894135,
      "grad_norm": 5.969266891479492,
      "learning_rate": 3.262069138650814e-05,
      "loss": 2.3966352462768556,
      "memory(GiB)": 77.56,
      "step": 71540,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.437445
    },
    {
      "epoch": 3.0652071462233836,
      "grad_norm": 5.5641069412231445,
      "learning_rate": 3.261438138542041e-05,
      "loss": 2.4115156173706054,
      "memory(GiB)": 77.56,
      "step": 71545,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.437456
    },
    {
      "epoch": 3.065421361552633,
      "grad_norm": 6.443191051483154,
      "learning_rate": 3.260807169928948e-05,
      "loss": 2.166291618347168,
      "memory(GiB)": 77.56,
      "step": 71550,
      "token_acc": 0.532258064516129,
      "train_speed(iter/s)": 1.437456
    },
    {
      "epoch": 3.0656355768818817,
      "grad_norm": 5.40637731552124,
      "learning_rate": 3.2601762328229624e-05,
      "loss": 2.4704456329345703,
      "memory(GiB)": 77.56,
      "step": 71555,
      "token_acc": 0.4746268656716418,
      "train_speed(iter/s)": 1.437439
    },
    {
      "epoch": 3.0658497922111305,
      "grad_norm": 6.787854194641113,
      "learning_rate": 3.259545327235518e-05,
      "loss": 2.2822240829467773,
      "memory(GiB)": 77.56,
      "step": 71560,
      "token_acc": 0.5045871559633027,
      "train_speed(iter/s)": 1.437453
    },
    {
      "epoch": 3.0660640075403798,
      "grad_norm": 4.851698875427246,
      "learning_rate": 3.258914453178041e-05,
      "loss": 2.2787031173706054,
      "memory(GiB)": 77.56,
      "step": 71565,
      "token_acc": 0.5317725752508361,
      "train_speed(iter/s)": 1.437437
    },
    {
      "epoch": 3.0662782228696286,
      "grad_norm": 7.534502029418945,
      "learning_rate": 3.2582836106619636e-05,
      "loss": 2.356695365905762,
      "memory(GiB)": 77.56,
      "step": 71570,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.437449
    },
    {
      "epoch": 3.0664924381988774,
      "grad_norm": 6.021844387054443,
      "learning_rate": 3.257652799698712e-05,
      "loss": 2.5388763427734373,
      "memory(GiB)": 77.56,
      "step": 71575,
      "token_acc": 0.4676470588235294,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 3.0667066535281267,
      "grad_norm": 7.734993934631348,
      "learning_rate": 3.2570220202997134e-05,
      "loss": 2.0558521270751955,
      "memory(GiB)": 77.56,
      "step": 71580,
      "token_acc": 0.5325670498084292,
      "train_speed(iter/s)": 1.437478
    },
    {
      "epoch": 3.0669208688573755,
      "grad_norm": 5.977901935577393,
      "learning_rate": 3.256391272476396e-05,
      "loss": 2.504092788696289,
      "memory(GiB)": 77.56,
      "step": 71585,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 3.0671350841866243,
      "grad_norm": 5.679624557495117,
      "learning_rate": 3.255760556240187e-05,
      "loss": 2.3099443435668947,
      "memory(GiB)": 77.56,
      "step": 71590,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.437531
    },
    {
      "epoch": 3.0673492995158735,
      "grad_norm": 5.730276584625244,
      "learning_rate": 3.255129871602512e-05,
      "loss": 2.5747289657592773,
      "memory(GiB)": 77.56,
      "step": 71595,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.437558
    },
    {
      "epoch": 3.0675635148451224,
      "grad_norm": 4.948228359222412,
      "learning_rate": 3.2544992185747955e-05,
      "loss": 2.3366249084472654,
      "memory(GiB)": 77.56,
      "step": 71600,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 3.067777730174371,
      "grad_norm": 4.9751482009887695,
      "learning_rate": 3.2538685971684646e-05,
      "loss": 2.1585954666137694,
      "memory(GiB)": 77.56,
      "step": 71605,
      "token_acc": 0.5398550724637681,
      "train_speed(iter/s)": 1.437563
    },
    {
      "epoch": 3.0679919455036204,
      "grad_norm": 5.517768383026123,
      "learning_rate": 3.253238007394942e-05,
      "loss": 2.2570945739746096,
      "memory(GiB)": 77.56,
      "step": 71610,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 3.0682061608328692,
      "grad_norm": 5.870797634124756,
      "learning_rate": 3.2526074492656496e-05,
      "loss": 2.14916877746582,
      "memory(GiB)": 77.56,
      "step": 71615,
      "token_acc": 0.5341880341880342,
      "train_speed(iter/s)": 1.437562
    },
    {
      "epoch": 3.068420376162118,
      "grad_norm": 5.81953763961792,
      "learning_rate": 3.2519769227920135e-05,
      "loss": 2.2913625717163084,
      "memory(GiB)": 77.56,
      "step": 71620,
      "token_acc": 0.5067114093959731,
      "train_speed(iter/s)": 1.43757
    },
    {
      "epoch": 3.0686345914913673,
      "grad_norm": 6.246188640594482,
      "learning_rate": 3.251346427985455e-05,
      "loss": 2.4065980911254883,
      "memory(GiB)": 77.56,
      "step": 71625,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.437576
    },
    {
      "epoch": 3.068848806820616,
      "grad_norm": 5.368649482727051,
      "learning_rate": 3.250715964857397e-05,
      "loss": 2.484805679321289,
      "memory(GiB)": 77.56,
      "step": 71630,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 3.069063022149865,
      "grad_norm": 6.015465259552002,
      "learning_rate": 3.25008553341926e-05,
      "loss": 2.111690330505371,
      "memory(GiB)": 77.56,
      "step": 71635,
      "token_acc": 0.511864406779661,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 3.069277237479114,
      "grad_norm": 6.980801105499268,
      "learning_rate": 3.249455133682464e-05,
      "loss": 2.120357894897461,
      "memory(GiB)": 77.56,
      "step": 71640,
      "token_acc": 0.55625,
      "train_speed(iter/s)": 1.437576
    },
    {
      "epoch": 3.069491452808363,
      "grad_norm": 5.813309669494629,
      "learning_rate": 3.248824765658432e-05,
      "loss": 2.312276840209961,
      "memory(GiB)": 77.56,
      "step": 71645,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.437579
    },
    {
      "epoch": 3.069705668137612,
      "grad_norm": 4.540748596191406,
      "learning_rate": 3.248194429358582e-05,
      "loss": 2.337319564819336,
      "memory(GiB)": 77.56,
      "step": 71650,
      "token_acc": 0.5084175084175084,
      "train_speed(iter/s)": 1.437602
    },
    {
      "epoch": 3.069919883466861,
      "grad_norm": 5.714628219604492,
      "learning_rate": 3.247564124794332e-05,
      "loss": 2.5630964279174804,
      "memory(GiB)": 77.56,
      "step": 71655,
      "token_acc": 0.5101214574898786,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 3.07013409879611,
      "grad_norm": 4.395263671875,
      "learning_rate": 3.246933851977101e-05,
      "loss": 1.9874814987182616,
      "memory(GiB)": 77.56,
      "step": 71660,
      "token_acc": 0.5474683544303798,
      "train_speed(iter/s)": 1.437575
    },
    {
      "epoch": 3.0703483141253587,
      "grad_norm": 6.776604175567627,
      "learning_rate": 3.246303610918309e-05,
      "loss": 2.7311637878417967,
      "memory(GiB)": 77.56,
      "step": 71665,
      "token_acc": 0.4412811387900356,
      "train_speed(iter/s)": 1.437584
    },
    {
      "epoch": 3.070562529454608,
      "grad_norm": 5.680537700653076,
      "learning_rate": 3.245673401629372e-05,
      "loss": 2.378589057922363,
      "memory(GiB)": 77.56,
      "step": 71670,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.437576
    },
    {
      "epoch": 3.0707767447838568,
      "grad_norm": 5.210697174072266,
      "learning_rate": 3.245043224121708e-05,
      "loss": 2.2265769958496096,
      "memory(GiB)": 77.56,
      "step": 71675,
      "token_acc": 0.5236363636363637,
      "train_speed(iter/s)": 1.437578
    },
    {
      "epoch": 3.0709909601131056,
      "grad_norm": 5.430446147918701,
      "learning_rate": 3.2444130784067315e-05,
      "loss": 2.5373294830322264,
      "memory(GiB)": 77.56,
      "step": 71680,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.437555
    },
    {
      "epoch": 3.071205175442355,
      "grad_norm": 5.818295955657959,
      "learning_rate": 3.243782964495859e-05,
      "loss": 2.374929428100586,
      "memory(GiB)": 77.56,
      "step": 71685,
      "token_acc": 0.5610687022900763,
      "train_speed(iter/s)": 1.437565
    },
    {
      "epoch": 3.0714193907716036,
      "grad_norm": 5.918603897094727,
      "learning_rate": 3.243152882400506e-05,
      "loss": 2.3068742752075195,
      "memory(GiB)": 77.56,
      "step": 71690,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.437555
    },
    {
      "epoch": 3.0716336061008525,
      "grad_norm": 6.759967803955078,
      "learning_rate": 3.2425228321320863e-05,
      "loss": 2.3748859405517577,
      "memory(GiB)": 77.56,
      "step": 71695,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.437555
    },
    {
      "epoch": 3.0718478214301017,
      "grad_norm": 4.13250207901001,
      "learning_rate": 3.241892813702014e-05,
      "loss": 2.2627542495727537,
      "memory(GiB)": 77.56,
      "step": 71700,
      "token_acc": 0.5175718849840255,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 3.0720620367593505,
      "grad_norm": 5.313356399536133,
      "learning_rate": 3.2412628271217034e-05,
      "loss": 2.1665061950683593,
      "memory(GiB)": 77.56,
      "step": 71705,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437597
    },
    {
      "epoch": 3.0722762520885993,
      "grad_norm": 5.155050754547119,
      "learning_rate": 3.240632872402565e-05,
      "loss": 2.4742965698242188,
      "memory(GiB)": 77.56,
      "step": 71710,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.437592
    },
    {
      "epoch": 3.0724904674178486,
      "grad_norm": 5.6786932945251465,
      "learning_rate": 3.240002949556014e-05,
      "loss": 2.2357444763183594,
      "memory(GiB)": 77.56,
      "step": 71715,
      "token_acc": 0.515527950310559,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 3.0727046827470974,
      "grad_norm": 4.857753753662109,
      "learning_rate": 3.23937305859346e-05,
      "loss": 2.176723861694336,
      "memory(GiB)": 77.56,
      "step": 71720,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.437531
    },
    {
      "epoch": 3.072918898076346,
      "grad_norm": 8.926349639892578,
      "learning_rate": 3.238743199526314e-05,
      "loss": 2.383189010620117,
      "memory(GiB)": 77.56,
      "step": 71725,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.437538
    },
    {
      "epoch": 3.0731331134055955,
      "grad_norm": 5.758974552154541,
      "learning_rate": 3.2381133723659886e-05,
      "loss": 2.0396209716796876,
      "memory(GiB)": 77.56,
      "step": 71730,
      "token_acc": 0.5607142857142857,
      "train_speed(iter/s)": 1.437565
    },
    {
      "epoch": 3.0733473287348443,
      "grad_norm": 5.930120944976807,
      "learning_rate": 3.23748357712389e-05,
      "loss": 2.563071441650391,
      "memory(GiB)": 77.56,
      "step": 71735,
      "token_acc": 0.4646153846153846,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 3.073561544064093,
      "grad_norm": 5.101224422454834,
      "learning_rate": 3.236853813811431e-05,
      "loss": 2.329065132141113,
      "memory(GiB)": 77.56,
      "step": 71740,
      "token_acc": 0.5191082802547771,
      "train_speed(iter/s)": 1.437556
    },
    {
      "epoch": 3.0737757593933424,
      "grad_norm": 6.716571807861328,
      "learning_rate": 3.236224082440019e-05,
      "loss": 2.2843505859375,
      "memory(GiB)": 77.56,
      "step": 71745,
      "token_acc": 0.47540983606557374,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 3.073989974722591,
      "grad_norm": 5.448788642883301,
      "learning_rate": 3.235594383021063e-05,
      "loss": 2.566900634765625,
      "memory(GiB)": 77.56,
      "step": 71750,
      "token_acc": 0.46706586826347307,
      "train_speed(iter/s)": 1.437554
    },
    {
      "epoch": 3.07420419005184,
      "grad_norm": 5.533188819885254,
      "learning_rate": 3.23496471556597e-05,
      "loss": 2.3816951751708983,
      "memory(GiB)": 77.56,
      "step": 71755,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.437569
    },
    {
      "epoch": 3.0744184053810892,
      "grad_norm": 5.50795316696167,
      "learning_rate": 3.234335080086146e-05,
      "loss": 2.3022790908813477,
      "memory(GiB)": 77.56,
      "step": 71760,
      "token_acc": 0.5328185328185329,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 3.074632620710338,
      "grad_norm": 7.947305679321289,
      "learning_rate": 3.233705476592999e-05,
      "loss": 2.4905216217041017,
      "memory(GiB)": 77.56,
      "step": 71765,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 3.074846836039587,
      "grad_norm": 5.364322662353516,
      "learning_rate": 3.233075905097933e-05,
      "loss": 1.9276460647583007,
      "memory(GiB)": 77.56,
      "step": 71770,
      "token_acc": 0.5478260869565217,
      "train_speed(iter/s)": 1.437616
    },
    {
      "epoch": 3.075061051368836,
      "grad_norm": 6.840217113494873,
      "learning_rate": 3.232446365612356e-05,
      "loss": 2.3881908416748048,
      "memory(GiB)": 77.56,
      "step": 71775,
      "token_acc": 0.45121951219512196,
      "train_speed(iter/s)": 1.437612
    },
    {
      "epoch": 3.075275266698085,
      "grad_norm": 5.127231597900391,
      "learning_rate": 3.231816858147672e-05,
      "loss": 2.365000534057617,
      "memory(GiB)": 77.56,
      "step": 71780,
      "token_acc": 0.4902597402597403,
      "train_speed(iter/s)": 1.437598
    },
    {
      "epoch": 3.0754894820273337,
      "grad_norm": 5.032304286956787,
      "learning_rate": 3.231187382715282e-05,
      "loss": 2.220559501647949,
      "memory(GiB)": 77.56,
      "step": 71785,
      "token_acc": 0.49812734082397003,
      "train_speed(iter/s)": 1.437606
    },
    {
      "epoch": 3.075703697356583,
      "grad_norm": 5.8689703941345215,
      "learning_rate": 3.2305579393265926e-05,
      "loss": 2.5555652618408202,
      "memory(GiB)": 77.56,
      "step": 71790,
      "token_acc": 0.48466257668711654,
      "train_speed(iter/s)": 1.437619
    },
    {
      "epoch": 3.075917912685832,
      "grad_norm": 5.924358367919922,
      "learning_rate": 3.229928527993007e-05,
      "loss": 2.620853042602539,
      "memory(GiB)": 77.56,
      "step": 71795,
      "token_acc": 0.47126436781609193,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 3.0761321280150806,
      "grad_norm": 7.015709400177002,
      "learning_rate": 3.229299148725925e-05,
      "loss": 2.032087516784668,
      "memory(GiB)": 77.56,
      "step": 71800,
      "token_acc": 0.51953125,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 3.07634634334433,
      "grad_norm": 6.854808330535889,
      "learning_rate": 3.228669801536749e-05,
      "loss": 2.7323497772216796,
      "memory(GiB)": 77.56,
      "step": 71805,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 3.0765605586735787,
      "grad_norm": 6.99651575088501,
      "learning_rate": 3.228040486436884e-05,
      "loss": 2.3545578002929686,
      "memory(GiB)": 77.56,
      "step": 71810,
      "token_acc": 0.5015479876160991,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 3.0767747740028275,
      "grad_norm": 5.134286403656006,
      "learning_rate": 3.2274112034377255e-05,
      "loss": 2.4191478729248046,
      "memory(GiB)": 77.56,
      "step": 71815,
      "token_acc": 0.5130718954248366,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 3.0769889893320768,
      "grad_norm": 5.35711145401001,
      "learning_rate": 3.2267819525506784e-05,
      "loss": 2.1645038604736326,
      "memory(GiB)": 77.56,
      "step": 71820,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.0772032046613256,
      "grad_norm": 5.819248199462891,
      "learning_rate": 3.226152733787138e-05,
      "loss": 2.5150949478149416,
      "memory(GiB)": 77.56,
      "step": 71825,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.437657
    },
    {
      "epoch": 3.0774174199905744,
      "grad_norm": 7.229367256164551,
      "learning_rate": 3.225523547158505e-05,
      "loss": 2.5439426422119142,
      "memory(GiB)": 77.56,
      "step": 71830,
      "token_acc": 0.5054151624548736,
      "train_speed(iter/s)": 1.437675
    },
    {
      "epoch": 3.0776316353198236,
      "grad_norm": 6.0406084060668945,
      "learning_rate": 3.224894392676178e-05,
      "loss": 2.1014326095581053,
      "memory(GiB)": 77.56,
      "step": 71835,
      "token_acc": 0.5352112676056338,
      "train_speed(iter/s)": 1.437663
    },
    {
      "epoch": 3.0778458506490725,
      "grad_norm": 6.239911079406738,
      "learning_rate": 3.2242652703515545e-05,
      "loss": 2.1096033096313476,
      "memory(GiB)": 77.56,
      "step": 71840,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.437672
    },
    {
      "epoch": 3.0780600659783213,
      "grad_norm": 4.778571128845215,
      "learning_rate": 3.22363618019603e-05,
      "loss": 2.174831771850586,
      "memory(GiB)": 77.56,
      "step": 71845,
      "token_acc": 0.5415282392026578,
      "train_speed(iter/s)": 1.437681
    },
    {
      "epoch": 3.0782742813075705,
      "grad_norm": 4.574695110321045,
      "learning_rate": 3.223007122221004e-05,
      "loss": 2.6403177261352537,
      "memory(GiB)": 77.56,
      "step": 71850,
      "token_acc": 0.4664429530201342,
      "train_speed(iter/s)": 1.437684
    },
    {
      "epoch": 3.0784884966368193,
      "grad_norm": 4.291148662567139,
      "learning_rate": 3.22237809643787e-05,
      "loss": 2.3742092132568358,
      "memory(GiB)": 77.56,
      "step": 71855,
      "token_acc": 0.5015384615384615,
      "train_speed(iter/s)": 1.43766
    },
    {
      "epoch": 3.078702711966068,
      "grad_norm": 6.408908367156982,
      "learning_rate": 3.221749102858025e-05,
      "loss": 2.471113586425781,
      "memory(GiB)": 77.56,
      "step": 71860,
      "token_acc": 0.46546546546546547,
      "train_speed(iter/s)": 1.437658
    },
    {
      "epoch": 3.0789169272953174,
      "grad_norm": 6.602572917938232,
      "learning_rate": 3.2211201414928634e-05,
      "loss": 2.210652542114258,
      "memory(GiB)": 77.56,
      "step": 71865,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.437669
    },
    {
      "epoch": 3.079131142624566,
      "grad_norm": 4.954071998596191,
      "learning_rate": 3.220491212353778e-05,
      "loss": 2.374320411682129,
      "memory(GiB)": 77.56,
      "step": 71870,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437684
    },
    {
      "epoch": 3.079345357953815,
      "grad_norm": 5.113256931304932,
      "learning_rate": 3.219862315452164e-05,
      "loss": 2.551711082458496,
      "memory(GiB)": 77.56,
      "step": 71875,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437701
    },
    {
      "epoch": 3.0795595732830643,
      "grad_norm": 5.2521467208862305,
      "learning_rate": 3.2192334507994125e-05,
      "loss": 2.2922691345214843,
      "memory(GiB)": 77.56,
      "step": 71880,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.437723
    },
    {
      "epoch": 3.079773788612313,
      "grad_norm": 5.62010383605957,
      "learning_rate": 3.2186046184069196e-05,
      "loss": 2.6016489028930665,
      "memory(GiB)": 77.56,
      "step": 71885,
      "token_acc": 0.45565749235474007,
      "train_speed(iter/s)": 1.43775
    },
    {
      "epoch": 3.079988003941562,
      "grad_norm": 5.3352370262146,
      "learning_rate": 3.2179758182860744e-05,
      "loss": 2.4521623611450196,
      "memory(GiB)": 77.56,
      "step": 71890,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 3.080202219270811,
      "grad_norm": 6.253001689910889,
      "learning_rate": 3.21734705044827e-05,
      "loss": 2.474909210205078,
      "memory(GiB)": 77.56,
      "step": 71895,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.437777
    },
    {
      "epoch": 3.08041643460006,
      "grad_norm": 4.9991230964660645,
      "learning_rate": 3.216718314904895e-05,
      "loss": 2.061448097229004,
      "memory(GiB)": 77.56,
      "step": 71900,
      "token_acc": 0.5533333333333333,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 3.080630649929309,
      "grad_norm": 4.501735210418701,
      "learning_rate": 3.216089611667341e-05,
      "loss": 2.2024246215820313,
      "memory(GiB)": 77.56,
      "step": 71905,
      "token_acc": 0.5275080906148867,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 3.080844865258558,
      "grad_norm": 7.868646621704102,
      "learning_rate": 3.215460940746997e-05,
      "loss": 2.4164424896240235,
      "memory(GiB)": 77.56,
      "step": 71910,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 3.081059080587807,
      "grad_norm": 5.91912317276001,
      "learning_rate": 3.214832302155252e-05,
      "loss": 2.18265438079834,
      "memory(GiB)": 77.56,
      "step": 71915,
      "token_acc": 0.5211726384364821,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 3.0812732959170557,
      "grad_norm": 6.025934219360352,
      "learning_rate": 3.214203695903495e-05,
      "loss": 1.8655845642089843,
      "memory(GiB)": 77.56,
      "step": 71920,
      "token_acc": 0.5863453815261044,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.081487511246305,
      "grad_norm": 7.082553863525391,
      "learning_rate": 3.213575122003113e-05,
      "loss": 2.5260461807250976,
      "memory(GiB)": 77.56,
      "step": 71925,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 3.0817017265755537,
      "grad_norm": 6.019506931304932,
      "learning_rate": 3.212946580465493e-05,
      "loss": 2.3202917098999025,
      "memory(GiB)": 77.56,
      "step": 71930,
      "token_acc": 0.5052264808362369,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 3.0819159419048026,
      "grad_norm": 8.1309232711792,
      "learning_rate": 3.212318071302023e-05,
      "loss": 2.1723093032836913,
      "memory(GiB)": 77.56,
      "step": 71935,
      "token_acc": 0.5107296137339056,
      "train_speed(iter/s)": 1.437823
    },
    {
      "epoch": 3.082130157234052,
      "grad_norm": 5.068955421447754,
      "learning_rate": 3.211689594524088e-05,
      "loss": 2.223263740539551,
      "memory(GiB)": 77.56,
      "step": 71940,
      "token_acc": 0.5359477124183006,
      "train_speed(iter/s)": 1.437824
    },
    {
      "epoch": 3.0823443725633006,
      "grad_norm": 5.923271179199219,
      "learning_rate": 3.211061150143072e-05,
      "loss": 2.525506019592285,
      "memory(GiB)": 77.56,
      "step": 71945,
      "token_acc": 0.4862068965517241,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.0825585878925494,
      "grad_norm": 6.4857964515686035,
      "learning_rate": 3.210432738170363e-05,
      "loss": 2.2871070861816407,
      "memory(GiB)": 77.56,
      "step": 71950,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 3.0827728032217987,
      "grad_norm": 4.484994411468506,
      "learning_rate": 3.209804358617342e-05,
      "loss": 2.260784912109375,
      "memory(GiB)": 77.56,
      "step": 71955,
      "token_acc": 0.542016806722689,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.0829870185510475,
      "grad_norm": 5.6009521484375,
      "learning_rate": 3.2091760114953945e-05,
      "loss": 2.441606140136719,
      "memory(GiB)": 77.56,
      "step": 71960,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.437807
    },
    {
      "epoch": 3.0832012338802963,
      "grad_norm": 6.694757461547852,
      "learning_rate": 3.208547696815905e-05,
      "loss": 2.440965270996094,
      "memory(GiB)": 77.56,
      "step": 71965,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437817
    },
    {
      "epoch": 3.0834154492095456,
      "grad_norm": 6.703338623046875,
      "learning_rate": 3.207919414590255e-05,
      "loss": 2.2883527755737303,
      "memory(GiB)": 77.56,
      "step": 71970,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 3.0836296645387944,
      "grad_norm": 4.850164413452148,
      "learning_rate": 3.207291164829824e-05,
      "loss": 2.2760610580444336,
      "memory(GiB)": 77.56,
      "step": 71975,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.437766
    },
    {
      "epoch": 3.083843879868043,
      "grad_norm": 5.694988250732422,
      "learning_rate": 3.2066629475459964e-05,
      "loss": 2.1303171157836913,
      "memory(GiB)": 77.56,
      "step": 71980,
      "token_acc": 0.5246478873239436,
      "train_speed(iter/s)": 1.43775
    },
    {
      "epoch": 3.0840580951972925,
      "grad_norm": 7.048049449920654,
      "learning_rate": 3.206034762750152e-05,
      "loss": 2.197726821899414,
      "memory(GiB)": 77.56,
      "step": 71985,
      "token_acc": 0.545774647887324,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.0842723105265413,
      "grad_norm": 6.287624835968018,
      "learning_rate": 3.20540661045367e-05,
      "loss": 2.2292152404785157,
      "memory(GiB)": 77.56,
      "step": 71990,
      "token_acc": 0.518796992481203,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 3.08448652585579,
      "grad_norm": 5.67429256439209,
      "learning_rate": 3.204778490667931e-05,
      "loss": 2.0574792861938476,
      "memory(GiB)": 77.56,
      "step": 71995,
      "token_acc": 0.5630498533724341,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.0847007411850393,
      "grad_norm": 5.883406639099121,
      "learning_rate": 3.2041504034043134e-05,
      "loss": 2.647998046875,
      "memory(GiB)": 77.56,
      "step": 72000,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437802
    },
    {
      "epoch": 3.0847007411850393,
      "eval_loss": 2.357974052429199,
      "eval_runtime": 14.3486,
      "eval_samples_per_second": 6.969,
      "eval_steps_per_second": 6.969,
      "eval_token_acc": 0.48531139835487663,
      "step": 72000
    },
    {
      "epoch": 3.084914956514288,
      "grad_norm": 5.997870922088623,
      "learning_rate": 3.203522348674197e-05,
      "loss": 2.483399200439453,
      "memory(GiB)": 77.56,
      "step": 72005,
      "token_acc": 0.4853700516351119,
      "train_speed(iter/s)": 1.437363
    },
    {
      "epoch": 3.085129171843537,
      "grad_norm": 6.540722370147705,
      "learning_rate": 3.202894326488958e-05,
      "loss": 2.4166095733642576,
      "memory(GiB)": 77.56,
      "step": 72010,
      "token_acc": 0.5124555160142349,
      "train_speed(iter/s)": 1.437364
    },
    {
      "epoch": 3.0853433871727862,
      "grad_norm": 5.439075946807861,
      "learning_rate": 3.2022663368599736e-05,
      "loss": 2.591163444519043,
      "memory(GiB)": 77.56,
      "step": 72015,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.437352
    },
    {
      "epoch": 3.085557602502035,
      "grad_norm": 6.601009845733643,
      "learning_rate": 3.201638379798622e-05,
      "loss": 2.4263397216796876,
      "memory(GiB)": 77.56,
      "step": 72020,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.437342
    },
    {
      "epoch": 3.085771817831284,
      "grad_norm": 5.886809349060059,
      "learning_rate": 3.2010104553162765e-05,
      "loss": 2.284076118469238,
      "memory(GiB)": 77.56,
      "step": 72025,
      "token_acc": 0.48695652173913045,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 3.085986033160533,
      "grad_norm": 6.877198696136475,
      "learning_rate": 3.200382563424314e-05,
      "loss": 2.06893253326416,
      "memory(GiB)": 77.56,
      "step": 72030,
      "token_acc": 0.5189873417721519,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 3.086200248489782,
      "grad_norm": 4.7113356590271,
      "learning_rate": 3.1997547041341105e-05,
      "loss": 2.377621078491211,
      "memory(GiB)": 77.56,
      "step": 72035,
      "token_acc": 0.5098684210526315,
      "train_speed(iter/s)": 1.437345
    },
    {
      "epoch": 3.0864144638190307,
      "grad_norm": 6.559420108795166,
      "learning_rate": 3.1991268774570396e-05,
      "loss": 2.1753616333007812,
      "memory(GiB)": 77.56,
      "step": 72040,
      "token_acc": 0.5338345864661654,
      "train_speed(iter/s)": 1.437361
    },
    {
      "epoch": 3.08662867914828,
      "grad_norm": 5.55491828918457,
      "learning_rate": 3.1984990834044746e-05,
      "loss": 2.225050354003906,
      "memory(GiB)": 77.56,
      "step": 72045,
      "token_acc": 0.5047318611987381,
      "train_speed(iter/s)": 1.437361
    },
    {
      "epoch": 3.086842894477529,
      "grad_norm": 4.979081153869629,
      "learning_rate": 3.1978713219877876e-05,
      "loss": 2.218865966796875,
      "memory(GiB)": 77.56,
      "step": 72050,
      "token_acc": 0.5252225519287834,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.0870571098067776,
      "grad_norm": 5.670287609100342,
      "learning_rate": 3.197243593218353e-05,
      "loss": 2.466337776184082,
      "memory(GiB)": 77.56,
      "step": 72055,
      "token_acc": 0.5330882352941176,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 3.087271325136027,
      "grad_norm": 5.962737560272217,
      "learning_rate": 3.196615897107542e-05,
      "loss": 2.1909963607788088,
      "memory(GiB)": 77.56,
      "step": 72060,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.437361
    },
    {
      "epoch": 3.0874855404652757,
      "grad_norm": 7.104704856872559,
      "learning_rate": 3.1959882336667254e-05,
      "loss": 2.455039405822754,
      "memory(GiB)": 77.56,
      "step": 72065,
      "token_acc": 0.478134110787172,
      "train_speed(iter/s)": 1.437363
    },
    {
      "epoch": 3.0876997557945245,
      "grad_norm": 5.077696323394775,
      "learning_rate": 3.1953606029072754e-05,
      "loss": 2.294321632385254,
      "memory(GiB)": 77.56,
      "step": 72070,
      "token_acc": 0.48024316109422494,
      "train_speed(iter/s)": 1.437377
    },
    {
      "epoch": 3.0879139711237737,
      "grad_norm": 6.672941207885742,
      "learning_rate": 3.194733004840559e-05,
      "loss": 2.2407079696655274,
      "memory(GiB)": 77.56,
      "step": 72075,
      "token_acc": 0.5618729096989966,
      "train_speed(iter/s)": 1.437405
    },
    {
      "epoch": 3.0881281864530226,
      "grad_norm": 5.932276248931885,
      "learning_rate": 3.194105439477949e-05,
      "loss": 2.081951904296875,
      "memory(GiB)": 77.56,
      "step": 72080,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.437392
    },
    {
      "epoch": 3.0883424017822714,
      "grad_norm": 6.801368713378906,
      "learning_rate": 3.193477906830813e-05,
      "loss": 2.516109275817871,
      "memory(GiB)": 77.56,
      "step": 72085,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.437413
    },
    {
      "epoch": 3.0885566171115206,
      "grad_norm": 6.3427019119262695,
      "learning_rate": 3.192850406910519e-05,
      "loss": 2.098362922668457,
      "memory(GiB)": 77.56,
      "step": 72090,
      "token_acc": 0.5807692307692308,
      "train_speed(iter/s)": 1.437407
    },
    {
      "epoch": 3.0887708324407694,
      "grad_norm": 4.7134575843811035,
      "learning_rate": 3.1922229397284355e-05,
      "loss": 2.218771743774414,
      "memory(GiB)": 77.56,
      "step": 72095,
      "token_acc": 0.5423076923076923,
      "train_speed(iter/s)": 1.437404
    },
    {
      "epoch": 3.0889850477700183,
      "grad_norm": 5.925879001617432,
      "learning_rate": 3.191595505295927e-05,
      "loss": 2.6201122283935545,
      "memory(GiB)": 77.56,
      "step": 72100,
      "token_acc": 0.461038961038961,
      "train_speed(iter/s)": 1.4374
    },
    {
      "epoch": 3.0891992630992675,
      "grad_norm": 5.819891929626465,
      "learning_rate": 3.1909681036243644e-05,
      "loss": 2.661220741271973,
      "memory(GiB)": 77.56,
      "step": 72105,
      "token_acc": 0.4638888888888889,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 3.0894134784285163,
      "grad_norm": 7.23334264755249,
      "learning_rate": 3.1903407347251114e-05,
      "loss": 2.608384895324707,
      "memory(GiB)": 77.56,
      "step": 72110,
      "token_acc": 0.45149253731343286,
      "train_speed(iter/s)": 1.437425
    },
    {
      "epoch": 3.089627693757765,
      "grad_norm": 4.465384006500244,
      "learning_rate": 3.189713398609534e-05,
      "loss": 2.405615234375,
      "memory(GiB)": 77.56,
      "step": 72115,
      "token_acc": 0.5049833887043189,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 3.0898419090870144,
      "grad_norm": 5.09502649307251,
      "learning_rate": 3.1890860952889955e-05,
      "loss": 2.4041469573974608,
      "memory(GiB)": 77.56,
      "step": 72120,
      "token_acc": 0.4769736842105263,
      "train_speed(iter/s)": 1.437406
    },
    {
      "epoch": 3.090056124416263,
      "grad_norm": 5.880337715148926,
      "learning_rate": 3.188458824774862e-05,
      "loss": 2.31864013671875,
      "memory(GiB)": 77.56,
      "step": 72125,
      "token_acc": 0.5246376811594203,
      "train_speed(iter/s)": 1.437423
    },
    {
      "epoch": 3.090270339745512,
      "grad_norm": 7.496294021606445,
      "learning_rate": 3.1878315870784975e-05,
      "loss": 2.663106918334961,
      "memory(GiB)": 77.56,
      "step": 72130,
      "token_acc": 0.47249190938511326,
      "train_speed(iter/s)": 1.437416
    },
    {
      "epoch": 3.0904845550747613,
      "grad_norm": 6.645288467407227,
      "learning_rate": 3.187204382211262e-05,
      "loss": 2.5387157440185546,
      "memory(GiB)": 77.56,
      "step": 72135,
      "token_acc": 0.4584615384615385,
      "train_speed(iter/s)": 1.437427
    },
    {
      "epoch": 3.09069877040401,
      "grad_norm": 7.943534851074219,
      "learning_rate": 3.18657721018452e-05,
      "loss": 2.1895246505737305,
      "memory(GiB)": 77.56,
      "step": 72140,
      "token_acc": 0.49480968858131485,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 3.090912985733259,
      "grad_norm": 5.230518817901611,
      "learning_rate": 3.185950071009633e-05,
      "loss": 2.2710304260253906,
      "memory(GiB)": 77.56,
      "step": 72145,
      "token_acc": 0.5492957746478874,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.091127201062508,
      "grad_norm": 6.006902694702148,
      "learning_rate": 3.185322964697963e-05,
      "loss": 2.5105504989624023,
      "memory(GiB)": 77.56,
      "step": 72150,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.091341416391757,
      "grad_norm": 6.310667037963867,
      "learning_rate": 3.184695891260869e-05,
      "loss": 2.3670064926147463,
      "memory(GiB)": 77.56,
      "step": 72155,
      "token_acc": 0.48134328358208955,
      "train_speed(iter/s)": 1.437466
    },
    {
      "epoch": 3.091555631721006,
      "grad_norm": 6.4622344970703125,
      "learning_rate": 3.184068850709711e-05,
      "loss": 2.5994964599609376,
      "memory(GiB)": 77.56,
      "step": 72160,
      "token_acc": 0.4648318042813456,
      "train_speed(iter/s)": 1.437493
    },
    {
      "epoch": 3.091769847050255,
      "grad_norm": 5.36970329284668,
      "learning_rate": 3.183441843055851e-05,
      "loss": 2.4790550231933595,
      "memory(GiB)": 77.56,
      "step": 72165,
      "token_acc": 0.4985507246376812,
      "train_speed(iter/s)": 1.437533
    },
    {
      "epoch": 3.091984062379504,
      "grad_norm": 6.095882415771484,
      "learning_rate": 3.182814868310645e-05,
      "loss": 2.3345258712768553,
      "memory(GiB)": 77.56,
      "step": 72170,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437552
    },
    {
      "epoch": 3.0921982777087527,
      "grad_norm": 7.8465657234191895,
      "learning_rate": 3.1821879264854514e-05,
      "loss": 2.7106595993041993,
      "memory(GiB)": 77.56,
      "step": 72175,
      "token_acc": 0.4707692307692308,
      "train_speed(iter/s)": 1.437541
    },
    {
      "epoch": 3.092412493038002,
      "grad_norm": 6.361175060272217,
      "learning_rate": 3.181561017591629e-05,
      "loss": 2.0068809509277346,
      "memory(GiB)": 77.56,
      "step": 72180,
      "token_acc": 0.5689655172413793,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 3.0926267083672507,
      "grad_norm": 5.485069751739502,
      "learning_rate": 3.180934141640536e-05,
      "loss": 2.086277198791504,
      "memory(GiB)": 77.56,
      "step": 72185,
      "token_acc": 0.5551601423487544,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 3.0928409236964995,
      "grad_norm": 5.531437873840332,
      "learning_rate": 3.1803072986435276e-05,
      "loss": 2.581049919128418,
      "memory(GiB)": 77.56,
      "step": 72190,
      "token_acc": 0.44668587896253603,
      "train_speed(iter/s)": 1.437598
    },
    {
      "epoch": 3.093055139025749,
      "grad_norm": 5.932785987854004,
      "learning_rate": 3.179680488611958e-05,
      "loss": 2.2052261352539064,
      "memory(GiB)": 77.56,
      "step": 72195,
      "token_acc": 0.5273311897106109,
      "train_speed(iter/s)": 1.437608
    },
    {
      "epoch": 3.0932693543549976,
      "grad_norm": 6.721252918243408,
      "learning_rate": 3.179053711557185e-05,
      "loss": 2.1347131729125977,
      "memory(GiB)": 77.56,
      "step": 72200,
      "token_acc": 0.5316901408450704,
      "train_speed(iter/s)": 1.437596
    },
    {
      "epoch": 3.0934835696842464,
      "grad_norm": 9.866572380065918,
      "learning_rate": 3.178426967490562e-05,
      "loss": 2.289981651306152,
      "memory(GiB)": 77.56,
      "step": 72205,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.437608
    },
    {
      "epoch": 3.0936977850134957,
      "grad_norm": 5.840310573577881,
      "learning_rate": 3.1778002564234435e-05,
      "loss": 2.4291318893432616,
      "memory(GiB)": 77.56,
      "step": 72210,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.437614
    },
    {
      "epoch": 3.0939120003427445,
      "grad_norm": 6.415073871612549,
      "learning_rate": 3.177173578367183e-05,
      "loss": 2.2293848037719726,
      "memory(GiB)": 77.56,
      "step": 72215,
      "token_acc": 0.5494505494505495,
      "train_speed(iter/s)": 1.437616
    },
    {
      "epoch": 3.0941262156719933,
      "grad_norm": 5.98455286026001,
      "learning_rate": 3.176546933333132e-05,
      "loss": 2.3406869888305666,
      "memory(GiB)": 77.56,
      "step": 72220,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.437571
    },
    {
      "epoch": 3.0943404310012426,
      "grad_norm": 6.014914512634277,
      "learning_rate": 3.175920321332644e-05,
      "loss": 2.2767578125,
      "memory(GiB)": 77.56,
      "step": 72225,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.437576
    },
    {
      "epoch": 3.0945546463304914,
      "grad_norm": 5.497958660125732,
      "learning_rate": 3.175293742377072e-05,
      "loss": 2.4478378295898438,
      "memory(GiB)": 77.56,
      "step": 72230,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.437583
    },
    {
      "epoch": 3.09476886165974,
      "grad_norm": 7.322943210601807,
      "learning_rate": 3.1746671964777635e-05,
      "loss": 2.2696950912475584,
      "memory(GiB)": 77.56,
      "step": 72235,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 3.0949830769889894,
      "grad_norm": 6.523723602294922,
      "learning_rate": 3.1740406836460724e-05,
      "loss": 2.157855415344238,
      "memory(GiB)": 77.56,
      "step": 72240,
      "token_acc": 0.5527426160337553,
      "train_speed(iter/s)": 1.437592
    },
    {
      "epoch": 3.0951972923182383,
      "grad_norm": 5.072347164154053,
      "learning_rate": 3.173414203893346e-05,
      "loss": 2.541507339477539,
      "memory(GiB)": 77.56,
      "step": 72245,
      "token_acc": 0.4897360703812317,
      "train_speed(iter/s)": 1.437588
    },
    {
      "epoch": 3.095411507647487,
      "grad_norm": 5.609076976776123,
      "learning_rate": 3.172787757230934e-05,
      "loss": 2.286670112609863,
      "memory(GiB)": 77.56,
      "step": 72250,
      "token_acc": 0.5284810126582279,
      "train_speed(iter/s)": 1.437599
    },
    {
      "epoch": 3.0956257229767363,
      "grad_norm": 4.251852989196777,
      "learning_rate": 3.172161343670188e-05,
      "loss": 2.4701719284057617,
      "memory(GiB)": 77.56,
      "step": 72255,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.437602
    },
    {
      "epoch": 3.095839938305985,
      "grad_norm": 7.1300153732299805,
      "learning_rate": 3.1715349632224536e-05,
      "loss": 2.5306621551513673,
      "memory(GiB)": 77.56,
      "step": 72260,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.437611
    },
    {
      "epoch": 3.096054153635234,
      "grad_norm": 4.470528602600098,
      "learning_rate": 3.1709086158990774e-05,
      "loss": 2.2525672912597656,
      "memory(GiB)": 77.56,
      "step": 72265,
      "token_acc": 0.5424836601307189,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 3.096268368964483,
      "grad_norm": 5.3731231689453125,
      "learning_rate": 3.170282301711409e-05,
      "loss": 2.0743844985961912,
      "memory(GiB)": 77.56,
      "step": 72270,
      "token_acc": 0.5321428571428571,
      "train_speed(iter/s)": 1.437665
    },
    {
      "epoch": 3.096482584293732,
      "grad_norm": 6.206145286560059,
      "learning_rate": 3.169656020670793e-05,
      "loss": 2.354397201538086,
      "memory(GiB)": 77.56,
      "step": 72275,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.096696799622981,
      "grad_norm": 5.7696990966796875,
      "learning_rate": 3.1690297727885744e-05,
      "loss": 2.3854522705078125,
      "memory(GiB)": 77.56,
      "step": 72280,
      "token_acc": 0.4712230215827338,
      "train_speed(iter/s)": 1.43768
    },
    {
      "epoch": 3.09691101495223,
      "grad_norm": 5.036414623260498,
      "learning_rate": 3.1684035580761e-05,
      "loss": 2.2238574981689454,
      "memory(GiB)": 77.56,
      "step": 72285,
      "token_acc": 0.5213675213675214,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 3.097125230281479,
      "grad_norm": 5.317901134490967,
      "learning_rate": 3.1677773765447116e-05,
      "loss": 2.3005258560180666,
      "memory(GiB)": 77.56,
      "step": 72290,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.437661
    },
    {
      "epoch": 3.0973394456107277,
      "grad_norm": 6.160380840301514,
      "learning_rate": 3.1671512282057556e-05,
      "loss": 2.5581039428710937,
      "memory(GiB)": 77.56,
      "step": 72295,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 3.097553660939977,
      "grad_norm": 6.025140762329102,
      "learning_rate": 3.166525113070575e-05,
      "loss": 2.6819684982299803,
      "memory(GiB)": 77.56,
      "step": 72300,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437699
    },
    {
      "epoch": 3.097767876269226,
      "grad_norm": 5.725876808166504,
      "learning_rate": 3.165899031150512e-05,
      "loss": 2.3693115234375,
      "memory(GiB)": 77.56,
      "step": 72305,
      "token_acc": 0.5275590551181102,
      "train_speed(iter/s)": 1.437716
    },
    {
      "epoch": 3.0979820915984746,
      "grad_norm": 9.523658752441406,
      "learning_rate": 3.165272982456908e-05,
      "loss": 2.3139448165893555,
      "memory(GiB)": 77.56,
      "step": 72310,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.437738
    },
    {
      "epoch": 3.098196306927724,
      "grad_norm": 6.289538860321045,
      "learning_rate": 3.164646967001106e-05,
      "loss": 2.5493627548217774,
      "memory(GiB)": 77.56,
      "step": 72315,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.437746
    },
    {
      "epoch": 3.0984105222569727,
      "grad_norm": 8.801753997802734,
      "learning_rate": 3.164020984794444e-05,
      "loss": 2.421823501586914,
      "memory(GiB)": 77.56,
      "step": 72320,
      "token_acc": 0.4707692307692308,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 3.0986247375862215,
      "grad_norm": 5.483835220336914,
      "learning_rate": 3.163395035848263e-05,
      "loss": 2.1594799041748045,
      "memory(GiB)": 77.56,
      "step": 72325,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.0988389529154707,
      "grad_norm": 5.4879326820373535,
      "learning_rate": 3.1627691201739075e-05,
      "loss": 2.1753717422485352,
      "memory(GiB)": 77.56,
      "step": 72330,
      "token_acc": 0.546031746031746,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 3.0990531682447195,
      "grad_norm": 5.669936180114746,
      "learning_rate": 3.162143237782711e-05,
      "loss": 2.225362205505371,
      "memory(GiB)": 77.56,
      "step": 72335,
      "token_acc": 0.5457627118644067,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 3.0992673835739684,
      "grad_norm": 7.326630115509033,
      "learning_rate": 3.161517388686014e-05,
      "loss": 2.4407577514648438,
      "memory(GiB)": 77.56,
      "step": 72340,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 3.0994815989032176,
      "grad_norm": 9.202301979064941,
      "learning_rate": 3.160891572895155e-05,
      "loss": 2.6768072128295897,
      "memory(GiB)": 77.56,
      "step": 72345,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.437784
    },
    {
      "epoch": 3.0996958142324664,
      "grad_norm": 5.857728481292725,
      "learning_rate": 3.160265790421469e-05,
      "loss": 2.355600929260254,
      "memory(GiB)": 77.56,
      "step": 72350,
      "token_acc": 0.5170068027210885,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 3.0999100295617152,
      "grad_norm": 5.486622333526611,
      "learning_rate": 3.159640041276295e-05,
      "loss": 2.291430854797363,
      "memory(GiB)": 77.56,
      "step": 72355,
      "token_acc": 0.5191082802547771,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 3.1001242448909645,
      "grad_norm": 5.8983049392700195,
      "learning_rate": 3.1590143254709684e-05,
      "loss": 2.4319149017333985,
      "memory(GiB)": 77.56,
      "step": 72360,
      "token_acc": 0.5087719298245614,
      "train_speed(iter/s)": 1.437776
    },
    {
      "epoch": 3.1003384602202133,
      "grad_norm": 7.934996604919434,
      "learning_rate": 3.158388643016823e-05,
      "loss": 2.0736560821533203,
      "memory(GiB)": 77.56,
      "step": 72365,
      "token_acc": 0.5428571428571428,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 3.100552675549462,
      "grad_norm": 4.799585342407227,
      "learning_rate": 3.157762993925196e-05,
      "loss": 2.261810302734375,
      "memory(GiB)": 77.56,
      "step": 72370,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.437769
    },
    {
      "epoch": 3.1007668908787114,
      "grad_norm": 5.184191703796387,
      "learning_rate": 3.15713737820742e-05,
      "loss": 2.720342254638672,
      "memory(GiB)": 77.56,
      "step": 72375,
      "token_acc": 0.43666666666666665,
      "train_speed(iter/s)": 1.437789
    },
    {
      "epoch": 3.10098110620796,
      "grad_norm": 6.596385478973389,
      "learning_rate": 3.156511795874829e-05,
      "loss": 2.0793190002441406,
      "memory(GiB)": 77.56,
      "step": 72380,
      "token_acc": 0.5517241379310345,
      "train_speed(iter/s)": 1.437773
    },
    {
      "epoch": 3.101195321537209,
      "grad_norm": 6.830258369445801,
      "learning_rate": 3.155886246938756e-05,
      "loss": 2.2441028594970702,
      "memory(GiB)": 77.56,
      "step": 72385,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.437779
    },
    {
      "epoch": 3.1014095368664583,
      "grad_norm": 4.584389686584473,
      "learning_rate": 3.155260731410534e-05,
      "loss": 2.112406539916992,
      "memory(GiB)": 77.56,
      "step": 72390,
      "token_acc": 0.5429553264604811,
      "train_speed(iter/s)": 1.437768
    },
    {
      "epoch": 3.101623752195707,
      "grad_norm": 5.585931301116943,
      "learning_rate": 3.1546352493014946e-05,
      "loss": 2.2400184631347657,
      "memory(GiB)": 77.56,
      "step": 72395,
      "token_acc": 0.4794007490636704,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 3.101837967524956,
      "grad_norm": 6.651171684265137,
      "learning_rate": 3.154009800622967e-05,
      "loss": 2.450203704833984,
      "memory(GiB)": 77.56,
      "step": 72400,
      "token_acc": 0.5068027210884354,
      "train_speed(iter/s)": 1.437814
    },
    {
      "epoch": 3.102052182854205,
      "grad_norm": 6.673031330108643,
      "learning_rate": 3.1533843853862856e-05,
      "loss": 2.609384536743164,
      "memory(GiB)": 77.56,
      "step": 72405,
      "token_acc": 0.4380952380952381,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 3.102266398183454,
      "grad_norm": 4.675049781799316,
      "learning_rate": 3.1527590036027766e-05,
      "loss": 2.400357246398926,
      "memory(GiB)": 77.56,
      "step": 72410,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.437847
    },
    {
      "epoch": 3.1024806135127028,
      "grad_norm": 5.692420482635498,
      "learning_rate": 3.152133655283773e-05,
      "loss": 1.7764366149902344,
      "memory(GiB)": 77.56,
      "step": 72415,
      "token_acc": 0.5944700460829493,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 3.102694828841952,
      "grad_norm": 6.678676605224609,
      "learning_rate": 3.151508340440601e-05,
      "loss": 2.6026721954345704,
      "memory(GiB)": 77.56,
      "step": 72420,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 3.102909044171201,
      "grad_norm": 5.01537561416626,
      "learning_rate": 3.150883059084588e-05,
      "loss": 2.1439157485961915,
      "memory(GiB)": 77.56,
      "step": 72425,
      "token_acc": 0.5551470588235294,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 3.1031232595004496,
      "grad_norm": 6.816878795623779,
      "learning_rate": 3.150257811227065e-05,
      "loss": 2.562957000732422,
      "memory(GiB)": 77.56,
      "step": 72430,
      "token_acc": 0.4716417910447761,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.103337474829699,
      "grad_norm": 5.693693161010742,
      "learning_rate": 3.149632596879356e-05,
      "loss": 2.0666225433349608,
      "memory(GiB)": 77.56,
      "step": 72435,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.437858
    },
    {
      "epoch": 3.1035516901589477,
      "grad_norm": 8.398011207580566,
      "learning_rate": 3.149007416052789e-05,
      "loss": 2.417411041259766,
      "memory(GiB)": 77.56,
      "step": 72440,
      "token_acc": 0.4612546125461255,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.1037659054881965,
      "grad_norm": 5.838447570800781,
      "learning_rate": 3.148382268758689e-05,
      "loss": 2.361836051940918,
      "memory(GiB)": 77.56,
      "step": 72445,
      "token_acc": 0.5050505050505051,
      "train_speed(iter/s)": 1.437816
    },
    {
      "epoch": 3.103980120817446,
      "grad_norm": 4.476272106170654,
      "learning_rate": 3.14775715500838e-05,
      "loss": 2.230912780761719,
      "memory(GiB)": 77.56,
      "step": 72450,
      "token_acc": 0.5582089552238806,
      "train_speed(iter/s)": 1.437803
    },
    {
      "epoch": 3.1041943361466946,
      "grad_norm": 5.37377405166626,
      "learning_rate": 3.147132074813189e-05,
      "loss": 2.399504280090332,
      "memory(GiB)": 77.56,
      "step": 72455,
      "token_acc": 0.5345345345345346,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 3.1044085514759434,
      "grad_norm": 6.497664451599121,
      "learning_rate": 3.146507028184438e-05,
      "loss": 2.6256196975708006,
      "memory(GiB)": 77.56,
      "step": 72460,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.1046227668051927,
      "grad_norm": 6.548216819763184,
      "learning_rate": 3.1458820151334504e-05,
      "loss": 2.2182369232177734,
      "memory(GiB)": 77.56,
      "step": 72465,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.1048369821344415,
      "grad_norm": 5.119418621063232,
      "learning_rate": 3.145257035671548e-05,
      "loss": 2.126926040649414,
      "memory(GiB)": 77.56,
      "step": 72470,
      "token_acc": 0.5177304964539007,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.1050511974636903,
      "grad_norm": 6.199021339416504,
      "learning_rate": 3.144632089810057e-05,
      "loss": 2.2834930419921875,
      "memory(GiB)": 77.56,
      "step": 72475,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 3.1052654127929396,
      "grad_norm": 4.47617244720459,
      "learning_rate": 3.144007177560295e-05,
      "loss": 2.2830326080322267,
      "memory(GiB)": 77.56,
      "step": 72480,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 3.1054796281221884,
      "grad_norm": 5.179701328277588,
      "learning_rate": 3.143382298933585e-05,
      "loss": 2.3549930572509767,
      "memory(GiB)": 77.56,
      "step": 72485,
      "token_acc": 0.5174825174825175,
      "train_speed(iter/s)": 1.437812
    },
    {
      "epoch": 3.105693843451437,
      "grad_norm": 4.490260601043701,
      "learning_rate": 3.142757453941246e-05,
      "loss": 2.3018190383911135,
      "memory(GiB)": 77.56,
      "step": 72490,
      "token_acc": 0.5058823529411764,
      "train_speed(iter/s)": 1.437833
    },
    {
      "epoch": 3.1059080587806864,
      "grad_norm": 4.446118354797363,
      "learning_rate": 3.142132642594597e-05,
      "loss": 2.4987226486206056,
      "memory(GiB)": 77.56,
      "step": 72495,
      "token_acc": 0.5169230769230769,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 3.1061222741099352,
      "grad_norm": 4.7859063148498535,
      "learning_rate": 3.141507864904959e-05,
      "loss": 2.839399719238281,
      "memory(GiB)": 77.56,
      "step": 72500,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437852
    },
    {
      "epoch": 3.1061222741099352,
      "eval_loss": 2.4001803398132324,
      "eval_runtime": 14.2707,
      "eval_samples_per_second": 7.007,
      "eval_steps_per_second": 7.007,
      "eval_token_acc": 0.4444444444444444,
      "step": 72500
    },
    {
      "epoch": 3.106336489439184,
      "grad_norm": 4.4844279289245605,
      "learning_rate": 3.1408831208836496e-05,
      "loss": 2.3198383331298826,
      "memory(GiB)": 77.56,
      "step": 72505,
      "token_acc": 0.4645030425963489,
      "train_speed(iter/s)": 1.437426
    },
    {
      "epoch": 3.1065507047684333,
      "grad_norm": 6.893157482147217,
      "learning_rate": 3.140258410541985e-05,
      "loss": 2.344624328613281,
      "memory(GiB)": 77.56,
      "step": 72510,
      "token_acc": 0.5080645161290323,
      "train_speed(iter/s)": 1.437419
    },
    {
      "epoch": 3.106764920097682,
      "grad_norm": 6.0269083976745605,
      "learning_rate": 3.139633733891285e-05,
      "loss": 2.666086196899414,
      "memory(GiB)": 77.56,
      "step": 72515,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.437405
    },
    {
      "epoch": 3.106979135426931,
      "grad_norm": 5.225742816925049,
      "learning_rate": 3.139009090942865e-05,
      "loss": 2.3564117431640623,
      "memory(GiB)": 77.56,
      "step": 72520,
      "token_acc": 0.47112462006079026,
      "train_speed(iter/s)": 1.437386
    },
    {
      "epoch": 3.10719335075618,
      "grad_norm": 6.708471298217773,
      "learning_rate": 3.138384481708041e-05,
      "loss": 2.374943161010742,
      "memory(GiB)": 77.56,
      "step": 72525,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.437378
    },
    {
      "epoch": 3.107407566085429,
      "grad_norm": 6.151369094848633,
      "learning_rate": 3.137759906198129e-05,
      "loss": 2.1034936904907227,
      "memory(GiB)": 77.56,
      "step": 72530,
      "token_acc": 0.5413793103448276,
      "train_speed(iter/s)": 1.437357
    },
    {
      "epoch": 3.107621781414678,
      "grad_norm": 7.223677635192871,
      "learning_rate": 3.137135364424442e-05,
      "loss": 2.3703739166259767,
      "memory(GiB)": 77.56,
      "step": 72535,
      "token_acc": 0.48255813953488375,
      "train_speed(iter/s)": 1.437362
    },
    {
      "epoch": 3.107835996743927,
      "grad_norm": 5.07366943359375,
      "learning_rate": 3.136510856398297e-05,
      "loss": 2.1987770080566404,
      "memory(GiB)": 77.56,
      "step": 72540,
      "token_acc": 0.5512367491166078,
      "train_speed(iter/s)": 1.437392
    },
    {
      "epoch": 3.108050212073176,
      "grad_norm": 5.923264503479004,
      "learning_rate": 3.135886382131003e-05,
      "loss": 2.3404064178466797,
      "memory(GiB)": 77.56,
      "step": 72545,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.437393
    },
    {
      "epoch": 3.1082644274024247,
      "grad_norm": 7.233532905578613,
      "learning_rate": 3.135261941633878e-05,
      "loss": 2.4130218505859373,
      "memory(GiB)": 77.56,
      "step": 72550,
      "token_acc": 0.49390243902439024,
      "train_speed(iter/s)": 1.437401
    },
    {
      "epoch": 3.108478642731674,
      "grad_norm": 4.881805896759033,
      "learning_rate": 3.134637534918231e-05,
      "loss": 1.9220348358154298,
      "memory(GiB)": 77.56,
      "step": 72555,
      "token_acc": 0.5313653136531366,
      "train_speed(iter/s)": 1.437387
    },
    {
      "epoch": 3.1086928580609228,
      "grad_norm": 6.463660717010498,
      "learning_rate": 3.134013161995377e-05,
      "loss": 2.2441144943237306,
      "memory(GiB)": 77.56,
      "step": 72560,
      "token_acc": 0.5016181229773463,
      "train_speed(iter/s)": 1.437401
    },
    {
      "epoch": 3.1089070733901716,
      "grad_norm": 5.045422077178955,
      "learning_rate": 3.133388822876624e-05,
      "loss": 2.158509063720703,
      "memory(GiB)": 77.56,
      "step": 72565,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.437406
    },
    {
      "epoch": 3.109121288719421,
      "grad_norm": 5.674724578857422,
      "learning_rate": 3.1327645175732826e-05,
      "loss": 2.216973876953125,
      "memory(GiB)": 77.56,
      "step": 72570,
      "token_acc": 0.5131578947368421,
      "train_speed(iter/s)": 1.437408
    },
    {
      "epoch": 3.1093355040486697,
      "grad_norm": 5.278008460998535,
      "learning_rate": 3.132140246096665e-05,
      "loss": 2.1971664428710938,
      "memory(GiB)": 77.56,
      "step": 72575,
      "token_acc": 0.5751633986928104,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 3.1095497193779185,
      "grad_norm": 8.706609725952148,
      "learning_rate": 3.131516008458077e-05,
      "loss": 2.2271652221679688,
      "memory(GiB)": 77.56,
      "step": 72580,
      "token_acc": 0.5434782608695652,
      "train_speed(iter/s)": 1.437427
    },
    {
      "epoch": 3.1097639347071677,
      "grad_norm": 6.113956928253174,
      "learning_rate": 3.130891804668832e-05,
      "loss": 2.478483963012695,
      "memory(GiB)": 77.56,
      "step": 72585,
      "token_acc": 0.4618320610687023,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 3.1099781500364165,
      "grad_norm": 5.216123580932617,
      "learning_rate": 3.1302676347402346e-05,
      "loss": 2.4413789749145507,
      "memory(GiB)": 77.56,
      "step": 72590,
      "token_acc": 0.509090909090909,
      "train_speed(iter/s)": 1.437454
    },
    {
      "epoch": 3.1101923653656653,
      "grad_norm": 7.272138595581055,
      "learning_rate": 3.129643498683592e-05,
      "loss": 2.2538478851318358,
      "memory(GiB)": 77.56,
      "step": 72595,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.437459
    },
    {
      "epoch": 3.1104065806949146,
      "grad_norm": 5.713901042938232,
      "learning_rate": 3.129019396510212e-05,
      "loss": 2.3358409881591795,
      "memory(GiB)": 77.56,
      "step": 72600,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.437472
    },
    {
      "epoch": 3.1106207960241634,
      "grad_norm": 5.661256313323975,
      "learning_rate": 3.128395328231401e-05,
      "loss": 2.209878349304199,
      "memory(GiB)": 77.56,
      "step": 72605,
      "token_acc": 0.5504201680672269,
      "train_speed(iter/s)": 1.43751
    },
    {
      "epoch": 3.1108350113534122,
      "grad_norm": 5.26945686340332,
      "learning_rate": 3.1277712938584646e-05,
      "loss": 2.3268457412719727,
      "memory(GiB)": 77.56,
      "step": 72610,
      "token_acc": 0.5104602510460251,
      "train_speed(iter/s)": 1.437495
    },
    {
      "epoch": 3.1110492266826615,
      "grad_norm": 6.317492485046387,
      "learning_rate": 3.127147293402708e-05,
      "loss": 2.369083213806152,
      "memory(GiB)": 77.56,
      "step": 72615,
      "token_acc": 0.5290102389078498,
      "train_speed(iter/s)": 1.437492
    },
    {
      "epoch": 3.1112634420119103,
      "grad_norm": 5.80665922164917,
      "learning_rate": 3.126523326875432e-05,
      "loss": 2.2709383010864257,
      "memory(GiB)": 77.56,
      "step": 72620,
      "token_acc": 0.5103244837758112,
      "train_speed(iter/s)": 1.437476
    },
    {
      "epoch": 3.111477657341159,
      "grad_norm": 4.686379432678223,
      "learning_rate": 3.1258993942879456e-05,
      "loss": 2.4110038757324217,
      "memory(GiB)": 77.56,
      "step": 72625,
      "token_acc": 0.496875,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 3.1116918726704084,
      "grad_norm": 5.836164951324463,
      "learning_rate": 3.125275495651551e-05,
      "loss": 2.7762956619262695,
      "memory(GiB)": 77.56,
      "step": 72630,
      "token_acc": 0.47648902821316613,
      "train_speed(iter/s)": 1.437499
    },
    {
      "epoch": 3.111906087999657,
      "grad_norm": 4.589163780212402,
      "learning_rate": 3.1246516309775484e-05,
      "loss": 2.175393486022949,
      "memory(GiB)": 77.56,
      "step": 72635,
      "token_acc": 0.5465587044534413,
      "train_speed(iter/s)": 1.437496
    },
    {
      "epoch": 3.112120303328906,
      "grad_norm": 4.759354114532471,
      "learning_rate": 3.12402780027724e-05,
      "loss": 2.202978324890137,
      "memory(GiB)": 77.56,
      "step": 72640,
      "token_acc": 0.5061349693251533,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 3.1123345186581552,
      "grad_norm": 8.271427154541016,
      "learning_rate": 3.123404003561929e-05,
      "loss": 2.4508623123168944,
      "memory(GiB)": 77.56,
      "step": 72645,
      "token_acc": 0.5019305019305019,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 3.112548733987404,
      "grad_norm": 7.143495559692383,
      "learning_rate": 3.122780240842915e-05,
      "loss": 2.3175933837890623,
      "memory(GiB)": 77.56,
      "step": 72650,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.437448
    },
    {
      "epoch": 3.112762949316653,
      "grad_norm": 7.061896324157715,
      "learning_rate": 3.122156512131497e-05,
      "loss": 2.5092161178588865,
      "memory(GiB)": 77.56,
      "step": 72655,
      "token_acc": 0.4919786096256685,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 3.112977164645902,
      "grad_norm": 6.718698024749756,
      "learning_rate": 3.1215328174389754e-05,
      "loss": 2.2886322021484373,
      "memory(GiB)": 77.56,
      "step": 72660,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437454
    },
    {
      "epoch": 3.113191379975151,
      "grad_norm": 10.471782684326172,
      "learning_rate": 3.1209091567766484e-05,
      "loss": 1.9001348495483399,
      "memory(GiB)": 77.56,
      "step": 72665,
      "token_acc": 0.5791505791505791,
      "train_speed(iter/s)": 1.437464
    },
    {
      "epoch": 3.1134055953044,
      "grad_norm": 6.2345075607299805,
      "learning_rate": 3.120285530155816e-05,
      "loss": 2.3077360153198243,
      "memory(GiB)": 77.56,
      "step": 72670,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.437463
    },
    {
      "epoch": 3.113619810633649,
      "grad_norm": 7.0577392578125,
      "learning_rate": 3.1196619375877746e-05,
      "loss": 2.246171569824219,
      "memory(GiB)": 77.56,
      "step": 72675,
      "token_acc": 0.5352112676056338,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 3.113834025962898,
      "grad_norm": 5.536130428314209,
      "learning_rate": 3.11903837908382e-05,
      "loss": 2.1795209884643554,
      "memory(GiB)": 77.56,
      "step": 72680,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.114048241292147,
      "grad_norm": 5.286458969116211,
      "learning_rate": 3.1184148546552505e-05,
      "loss": 2.627972221374512,
      "memory(GiB)": 77.56,
      "step": 72685,
      "token_acc": 0.4326923076923077,
      "train_speed(iter/s)": 1.43748
    },
    {
      "epoch": 3.114262456621396,
      "grad_norm": 5.607089042663574,
      "learning_rate": 3.117791364313361e-05,
      "loss": 2.264396095275879,
      "memory(GiB)": 77.56,
      "step": 72690,
      "token_acc": 0.5377049180327869,
      "train_speed(iter/s)": 1.437498
    },
    {
      "epoch": 3.1144766719506447,
      "grad_norm": 4.869533061981201,
      "learning_rate": 3.117167908069445e-05,
      "loss": 2.3718740463256838,
      "memory(GiB)": 77.56,
      "step": 72695,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 3.114690887279894,
      "grad_norm": 5.879380702972412,
      "learning_rate": 3.116544485934799e-05,
      "loss": 2.4864360809326174,
      "memory(GiB)": 77.56,
      "step": 72700,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.43748
    },
    {
      "epoch": 3.1149051026091428,
      "grad_norm": 9.330574989318848,
      "learning_rate": 3.115921097920718e-05,
      "loss": 2.2893978118896485,
      "memory(GiB)": 77.56,
      "step": 72705,
      "token_acc": 0.5418181818181819,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.1151193179383916,
      "grad_norm": 5.411746025085449,
      "learning_rate": 3.1152977440384927e-05,
      "loss": 2.3624856948852537,
      "memory(GiB)": 77.56,
      "step": 72710,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.115333533267641,
      "grad_norm": 6.132997989654541,
      "learning_rate": 3.114674424299416e-05,
      "loss": 2.4021377563476562,
      "memory(GiB)": 77.56,
      "step": 72715,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.1155477485968897,
      "grad_norm": 5.839630126953125,
      "learning_rate": 3.114051138714783e-05,
      "loss": 2.3691007614135744,
      "memory(GiB)": 77.56,
      "step": 72720,
      "token_acc": 0.5211726384364821,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.1157619639261385,
      "grad_norm": 5.101189136505127,
      "learning_rate": 3.1134278872958814e-05,
      "loss": 2.4583951950073244,
      "memory(GiB)": 77.56,
      "step": 72725,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.437435
    },
    {
      "epoch": 3.1159761792553877,
      "grad_norm": 7.1166768074035645,
      "learning_rate": 3.112804670054004e-05,
      "loss": 2.1673370361328126,
      "memory(GiB)": 77.56,
      "step": 72730,
      "token_acc": 0.5215686274509804,
      "train_speed(iter/s)": 1.43746
    },
    {
      "epoch": 3.1161903945846365,
      "grad_norm": 5.3363165855407715,
      "learning_rate": 3.1121814870004395e-05,
      "loss": 2.1296621322631837,
      "memory(GiB)": 77.56,
      "step": 72735,
      "token_acc": 0.5149501661129569,
      "train_speed(iter/s)": 1.437466
    },
    {
      "epoch": 3.1164046099138853,
      "grad_norm": 5.645517826080322,
      "learning_rate": 3.111558338146479e-05,
      "loss": 2.426052284240723,
      "memory(GiB)": 77.56,
      "step": 72740,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.437473
    },
    {
      "epoch": 3.1166188252431346,
      "grad_norm": 6.047229290008545,
      "learning_rate": 3.11093522350341e-05,
      "loss": 2.3883052825927735,
      "memory(GiB)": 77.56,
      "step": 72745,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.437485
    },
    {
      "epoch": 3.1168330405723834,
      "grad_norm": 5.909136772155762,
      "learning_rate": 3.1103121430825224e-05,
      "loss": 2.350218963623047,
      "memory(GiB)": 77.56,
      "step": 72750,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 3.1170472559016322,
      "grad_norm": 6.532894134521484,
      "learning_rate": 3.109689096895102e-05,
      "loss": 2.4577178955078125,
      "memory(GiB)": 77.56,
      "step": 72755,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.437487
    },
    {
      "epoch": 3.1172614712308815,
      "grad_norm": 7.467641353607178,
      "learning_rate": 3.109066084952438e-05,
      "loss": 2.582114028930664,
      "memory(GiB)": 77.56,
      "step": 72760,
      "token_acc": 0.46096654275092935,
      "train_speed(iter/s)": 1.437486
    },
    {
      "epoch": 3.1174756865601303,
      "grad_norm": 5.025373458862305,
      "learning_rate": 3.1084431072658135e-05,
      "loss": 2.70206356048584,
      "memory(GiB)": 77.56,
      "step": 72765,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 3.117689901889379,
      "grad_norm": 5.961118698120117,
      "learning_rate": 3.1078201638465164e-05,
      "loss": 2.2784103393554687,
      "memory(GiB)": 77.56,
      "step": 72770,
      "token_acc": 0.5473684210526316,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.1179041172186284,
      "grad_norm": 7.404531002044678,
      "learning_rate": 3.107197254705835e-05,
      "loss": 2.651513671875,
      "memory(GiB)": 77.56,
      "step": 72775,
      "token_acc": 0.47843137254901963,
      "train_speed(iter/s)": 1.437477
    },
    {
      "epoch": 3.118118332547877,
      "grad_norm": 5.981419563293457,
      "learning_rate": 3.1065743798550496e-05,
      "loss": 2.500725746154785,
      "memory(GiB)": 77.56,
      "step": 72780,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.118332547877126,
      "grad_norm": 6.0158772468566895,
      "learning_rate": 3.105951539305445e-05,
      "loss": 2.1025915145874023,
      "memory(GiB)": 77.56,
      "step": 72785,
      "token_acc": 0.5817490494296578,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 3.1185467632063752,
      "grad_norm": 5.096309661865234,
      "learning_rate": 3.105328733068306e-05,
      "loss": 2.2628801345825194,
      "memory(GiB)": 77.56,
      "step": 72790,
      "token_acc": 0.5278969957081545,
      "train_speed(iter/s)": 1.437459
    },
    {
      "epoch": 3.118760978535624,
      "grad_norm": 5.582957744598389,
      "learning_rate": 3.104705961154914e-05,
      "loss": 2.148325729370117,
      "memory(GiB)": 77.56,
      "step": 72795,
      "token_acc": 0.5701754385964912,
      "train_speed(iter/s)": 1.437474
    },
    {
      "epoch": 3.118975193864873,
      "grad_norm": 4.9675445556640625,
      "learning_rate": 3.1040832235765506e-05,
      "loss": 2.074441909790039,
      "memory(GiB)": 77.56,
      "step": 72800,
      "token_acc": 0.558641975308642,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.119189409194122,
      "grad_norm": 4.737760066986084,
      "learning_rate": 3.103460520344499e-05,
      "loss": 2.133156585693359,
      "memory(GiB)": 77.56,
      "step": 72805,
      "token_acc": 0.5170068027210885,
      "train_speed(iter/s)": 1.437487
    },
    {
      "epoch": 3.119403624523371,
      "grad_norm": 8.512147903442383,
      "learning_rate": 3.1028378514700386e-05,
      "loss": 2.3238494873046873,
      "memory(GiB)": 77.56,
      "step": 72810,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 3.1196178398526198,
      "grad_norm": 7.502225875854492,
      "learning_rate": 3.1022152169644515e-05,
      "loss": 2.5177003860473635,
      "memory(GiB)": 77.56,
      "step": 72815,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 3.119832055181869,
      "grad_norm": 4.5646443367004395,
      "learning_rate": 3.101592616839015e-05,
      "loss": 2.361803436279297,
      "memory(GiB)": 77.56,
      "step": 72820,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 3.120046270511118,
      "grad_norm": 6.551790714263916,
      "learning_rate": 3.100970051105009e-05,
      "loss": 2.3915557861328125,
      "memory(GiB)": 77.56,
      "step": 72825,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.43748
    },
    {
      "epoch": 3.1202604858403666,
      "grad_norm": 7.009518623352051,
      "learning_rate": 3.1003475197737125e-05,
      "loss": 2.277404022216797,
      "memory(GiB)": 77.56,
      "step": 72830,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.437473
    },
    {
      "epoch": 3.120474701169616,
      "grad_norm": 5.483964920043945,
      "learning_rate": 3.0997250228564026e-05,
      "loss": 2.1274051666259766,
      "memory(GiB)": 77.56,
      "step": 72835,
      "token_acc": 0.5833333333333334,
      "train_speed(iter/s)": 1.43749
    },
    {
      "epoch": 3.1206889164988647,
      "grad_norm": 8.037076950073242,
      "learning_rate": 3.0991025603643556e-05,
      "loss": 2.237125015258789,
      "memory(GiB)": 77.56,
      "step": 72840,
      "token_acc": 0.5192307692307693,
      "train_speed(iter/s)": 1.437492
    },
    {
      "epoch": 3.1209031318281135,
      "grad_norm": 5.772195339202881,
      "learning_rate": 3.0984801323088484e-05,
      "loss": 2.5056087493896486,
      "memory(GiB)": 77.56,
      "step": 72845,
      "token_acc": 0.47766323024054985,
      "train_speed(iter/s)": 1.437492
    },
    {
      "epoch": 3.1211173471573628,
      "grad_norm": 5.439091682434082,
      "learning_rate": 3.097857738701159e-05,
      "loss": 2.4180681228637697,
      "memory(GiB)": 77.56,
      "step": 72850,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437491
    },
    {
      "epoch": 3.1213315624866116,
      "grad_norm": 5.943234443664551,
      "learning_rate": 3.097235379552561e-05,
      "loss": 2.1390995025634765,
      "memory(GiB)": 77.56,
      "step": 72855,
      "token_acc": 0.5345454545454545,
      "train_speed(iter/s)": 1.437496
    },
    {
      "epoch": 3.1215457778158604,
      "grad_norm": 6.821101665496826,
      "learning_rate": 3.096613054874328e-05,
      "loss": 2.6187824249267577,
      "memory(GiB)": 77.56,
      "step": 72860,
      "token_acc": 0.4774774774774775,
      "train_speed(iter/s)": 1.437497
    },
    {
      "epoch": 3.1217599931451097,
      "grad_norm": 4.693805694580078,
      "learning_rate": 3.0959907646777364e-05,
      "loss": 2.488432502746582,
      "memory(GiB)": 77.56,
      "step": 72865,
      "token_acc": 0.5302491103202847,
      "train_speed(iter/s)": 1.437514
    },
    {
      "epoch": 3.1219742084743585,
      "grad_norm": 7.575918197631836,
      "learning_rate": 3.0953685089740566e-05,
      "loss": 2.244872283935547,
      "memory(GiB)": 77.56,
      "step": 72870,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 3.1221884238036073,
      "grad_norm": 9.932140350341797,
      "learning_rate": 3.094746287774564e-05,
      "loss": 2.512201499938965,
      "memory(GiB)": 77.56,
      "step": 72875,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.437541
    },
    {
      "epoch": 3.1224026391328565,
      "grad_norm": 6.540576934814453,
      "learning_rate": 3.09412410109053e-05,
      "loss": 2.538871002197266,
      "memory(GiB)": 77.56,
      "step": 72880,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.437505
    },
    {
      "epoch": 3.1226168544621054,
      "grad_norm": 5.2672929763793945,
      "learning_rate": 3.093501948933225e-05,
      "loss": 2.2062732696533205,
      "memory(GiB)": 77.56,
      "step": 72885,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 3.122831069791354,
      "grad_norm": 4.811594009399414,
      "learning_rate": 3.0928798313139206e-05,
      "loss": 2.2711551666259764,
      "memory(GiB)": 77.56,
      "step": 72890,
      "token_acc": 0.5252100840336135,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 3.1230452851206034,
      "grad_norm": 6.194910049438477,
      "learning_rate": 3.092257748243888e-05,
      "loss": 2.4762619018554686,
      "memory(GiB)": 77.56,
      "step": 72895,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 3.1232595004498522,
      "grad_norm": 4.968112945556641,
      "learning_rate": 3.0916356997343945e-05,
      "loss": 2.158442497253418,
      "memory(GiB)": 77.56,
      "step": 72900,
      "token_acc": 0.5594855305466238,
      "train_speed(iter/s)": 1.437516
    },
    {
      "epoch": 3.123473715779101,
      "grad_norm": 6.127618789672852,
      "learning_rate": 3.091013685796712e-05,
      "loss": 2.200439453125,
      "memory(GiB)": 77.56,
      "step": 72905,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.437516
    },
    {
      "epoch": 3.1236879311083503,
      "grad_norm": 6.250693321228027,
      "learning_rate": 3.0903917064421064e-05,
      "loss": 2.0903013229370115,
      "memory(GiB)": 77.56,
      "step": 72910,
      "token_acc": 0.5682819383259912,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 3.123902146437599,
      "grad_norm": 5.791238784790039,
      "learning_rate": 3.0897697616818444e-05,
      "loss": 2.281488037109375,
      "memory(GiB)": 77.56,
      "step": 72915,
      "token_acc": 0.4896142433234421,
      "train_speed(iter/s)": 1.437551
    },
    {
      "epoch": 3.124116361766848,
      "grad_norm": 7.567104339599609,
      "learning_rate": 3.0891478515271975e-05,
      "loss": 2.2617361068725588,
      "memory(GiB)": 77.56,
      "step": 72920,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.437564
    },
    {
      "epoch": 3.124330577096097,
      "grad_norm": 6.483534812927246,
      "learning_rate": 3.0885259759894306e-05,
      "loss": 2.3611730575561523,
      "memory(GiB)": 77.56,
      "step": 72925,
      "token_acc": 0.512280701754386,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 3.124544792425346,
      "grad_norm": 5.5189948081970215,
      "learning_rate": 3.0879041350798064e-05,
      "loss": 2.5294239044189455,
      "memory(GiB)": 77.56,
      "step": 72930,
      "token_acc": 0.4751552795031056,
      "train_speed(iter/s)": 1.437596
    },
    {
      "epoch": 3.124759007754595,
      "grad_norm": 6.098874092102051,
      "learning_rate": 3.0872823288095946e-05,
      "loss": 2.204116439819336,
      "memory(GiB)": 77.56,
      "step": 72935,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.437591
    },
    {
      "epoch": 3.124973223083844,
      "grad_norm": 5.665869235992432,
      "learning_rate": 3.086660557190057e-05,
      "loss": 2.14300594329834,
      "memory(GiB)": 77.56,
      "step": 72940,
      "token_acc": 0.5244755244755245,
      "train_speed(iter/s)": 1.437601
    },
    {
      "epoch": 3.125187438413093,
      "grad_norm": 6.831770420074463,
      "learning_rate": 3.086038820232458e-05,
      "loss": 2.2865468978881838,
      "memory(GiB)": 77.56,
      "step": 72945,
      "token_acc": 0.5283018867924528,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 3.1254016537423417,
      "grad_norm": 5.449192523956299,
      "learning_rate": 3.085417117948062e-05,
      "loss": 2.318807601928711,
      "memory(GiB)": 77.56,
      "step": 72950,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 3.125615869071591,
      "grad_norm": 5.110204696655273,
      "learning_rate": 3.08479545034813e-05,
      "loss": 2.0981908798217774,
      "memory(GiB)": 77.56,
      "step": 72955,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 3.1258300844008398,
      "grad_norm": 7.809922695159912,
      "learning_rate": 3.084173817443925e-05,
      "loss": 2.3389583587646485,
      "memory(GiB)": 77.56,
      "step": 72960,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.437635
    },
    {
      "epoch": 3.1260442997300886,
      "grad_norm": 6.2622833251953125,
      "learning_rate": 3.0835522192467095e-05,
      "loss": 1.8912019729614258,
      "memory(GiB)": 77.56,
      "step": 72965,
      "token_acc": 0.588,
      "train_speed(iter/s)": 1.437663
    },
    {
      "epoch": 3.126258515059338,
      "grad_norm": 5.646315574645996,
      "learning_rate": 3.082930655767742e-05,
      "loss": 2.479226493835449,
      "memory(GiB)": 77.56,
      "step": 72970,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.437689
    },
    {
      "epoch": 3.1264727303885866,
      "grad_norm": 5.672926902770996,
      "learning_rate": 3.082309127018285e-05,
      "loss": 2.086069107055664,
      "memory(GiB)": 77.56,
      "step": 72975,
      "token_acc": 0.5247524752475248,
      "train_speed(iter/s)": 1.437707
    },
    {
      "epoch": 3.1266869457178355,
      "grad_norm": 5.0447773933410645,
      "learning_rate": 3.081687633009598e-05,
      "loss": 2.194288444519043,
      "memory(GiB)": 77.56,
      "step": 72980,
      "token_acc": 0.48253968253968255,
      "train_speed(iter/s)": 1.437698
    },
    {
      "epoch": 3.1269011610470847,
      "grad_norm": 5.081331253051758,
      "learning_rate": 3.081066173752937e-05,
      "loss": 2.1215471267700194,
      "memory(GiB)": 77.56,
      "step": 72985,
      "token_acc": 0.547945205479452,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 3.1271153763763335,
      "grad_norm": 5.446280002593994,
      "learning_rate": 3.080444749259561e-05,
      "loss": 2.436065673828125,
      "memory(GiB)": 77.56,
      "step": 72990,
      "token_acc": 0.48059701492537316,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 3.1273295917055823,
      "grad_norm": 5.769184589385986,
      "learning_rate": 3.079823359540732e-05,
      "loss": 2.543348693847656,
      "memory(GiB)": 77.56,
      "step": 72995,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.1275438070348316,
      "grad_norm": 5.120479106903076,
      "learning_rate": 3.079202004607704e-05,
      "loss": 2.1759628295898437,
      "memory(GiB)": 77.56,
      "step": 73000,
      "token_acc": 0.5576208178438662,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 3.1275438070348316,
      "eval_loss": 2.2976233959198,
      "eval_runtime": 14.3954,
      "eval_samples_per_second": 6.947,
      "eval_steps_per_second": 6.947,
      "eval_token_acc": 0.44025157232704404,
      "step": 73000
    },
    {
      "epoch": 3.1277580223640804,
      "grad_norm": 4.880911827087402,
      "learning_rate": 3.0785806844717324e-05,
      "loss": 2.5289411544799805,
      "memory(GiB)": 77.56,
      "step": 73005,
      "token_acc": 0.45471349353049906,
      "train_speed(iter/s)": 1.437306
    },
    {
      "epoch": 3.127972237693329,
      "grad_norm": 7.991203784942627,
      "learning_rate": 3.077959399144075e-05,
      "loss": 2.3422672271728517,
      "memory(GiB)": 77.56,
      "step": 73010,
      "token_acc": 0.5078125,
      "train_speed(iter/s)": 1.437333
    },
    {
      "epoch": 3.1281864530225785,
      "grad_norm": 6.289660453796387,
      "learning_rate": 3.0773381486359854e-05,
      "loss": 2.2838260650634767,
      "memory(GiB)": 77.56,
      "step": 73015,
      "token_acc": 0.5243055555555556,
      "train_speed(iter/s)": 1.437308
    },
    {
      "epoch": 3.1284006683518273,
      "grad_norm": 7.014865875244141,
      "learning_rate": 3.0767169329587195e-05,
      "loss": 2.3660045623779298,
      "memory(GiB)": 77.56,
      "step": 73020,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 3.128614883681076,
      "grad_norm": 5.524535179138184,
      "learning_rate": 3.07609575212353e-05,
      "loss": 2.3686203002929687,
      "memory(GiB)": 77.56,
      "step": 73025,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 3.1288290990103254,
      "grad_norm": 6.617551326751709,
      "learning_rate": 3.0754746061416704e-05,
      "loss": 1.9826824188232421,
      "memory(GiB)": 77.56,
      "step": 73030,
      "token_acc": 0.5818815331010453,
      "train_speed(iter/s)": 1.437317
    },
    {
      "epoch": 3.129043314339574,
      "grad_norm": 6.423425674438477,
      "learning_rate": 3.074853495024395e-05,
      "loss": 2.0677026748657226,
      "memory(GiB)": 77.56,
      "step": 73035,
      "token_acc": 0.553921568627451,
      "train_speed(iter/s)": 1.437297
    },
    {
      "epoch": 3.129257529668823,
      "grad_norm": 4.993946552276611,
      "learning_rate": 3.074232418782954e-05,
      "loss": 2.353153419494629,
      "memory(GiB)": 77.56,
      "step": 73040,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 3.1294717449980722,
      "grad_norm": 5.287154197692871,
      "learning_rate": 3.0736113774285984e-05,
      "loss": 2.3713645935058594,
      "memory(GiB)": 77.56,
      "step": 73045,
      "token_acc": 0.4725609756097561,
      "train_speed(iter/s)": 1.437337
    },
    {
      "epoch": 3.129685960327321,
      "grad_norm": 5.195074558258057,
      "learning_rate": 3.07299037097258e-05,
      "loss": 2.340109443664551,
      "memory(GiB)": 77.56,
      "step": 73050,
      "token_acc": 0.512987012987013,
      "train_speed(iter/s)": 1.437341
    },
    {
      "epoch": 3.12990017565657,
      "grad_norm": 6.541009426116943,
      "learning_rate": 3.0723693994261496e-05,
      "loss": 2.3680057525634766,
      "memory(GiB)": 77.56,
      "step": 73055,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.130114390985819,
      "grad_norm": 5.066952705383301,
      "learning_rate": 3.0717484628005556e-05,
      "loss": 2.6939273834228517,
      "memory(GiB)": 77.56,
      "step": 73060,
      "token_acc": 0.43356643356643354,
      "train_speed(iter/s)": 1.437342
    },
    {
      "epoch": 3.130328606315068,
      "grad_norm": 6.5553297996521,
      "learning_rate": 3.071127561107044e-05,
      "loss": 2.595798873901367,
      "memory(GiB)": 77.56,
      "step": 73065,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.437333
    },
    {
      "epoch": 3.1305428216443167,
      "grad_norm": 6.919444561004639,
      "learning_rate": 3.07050669435687e-05,
      "loss": 2.5092817306518556,
      "memory(GiB)": 77.56,
      "step": 73070,
      "token_acc": 0.4743202416918429,
      "train_speed(iter/s)": 1.437317
    },
    {
      "epoch": 3.130757036973566,
      "grad_norm": 5.77463960647583,
      "learning_rate": 3.069885862561275e-05,
      "loss": 2.577484703063965,
      "memory(GiB)": 77.56,
      "step": 73075,
      "token_acc": 0.4768211920529801,
      "train_speed(iter/s)": 1.437327
    },
    {
      "epoch": 3.130971252302815,
      "grad_norm": 6.542135715484619,
      "learning_rate": 3.0692650657315106e-05,
      "loss": 2.0368463516235353,
      "memory(GiB)": 77.56,
      "step": 73080,
      "token_acc": 0.5256410256410257,
      "train_speed(iter/s)": 1.43733
    },
    {
      "epoch": 3.1311854676320636,
      "grad_norm": 5.323869705200195,
      "learning_rate": 3.0686443038788196e-05,
      "loss": 2.0392196655273436,
      "memory(GiB)": 77.56,
      "step": 73085,
      "token_acc": 0.5373134328358209,
      "train_speed(iter/s)": 1.43733
    },
    {
      "epoch": 3.131399682961313,
      "grad_norm": 8.747483253479004,
      "learning_rate": 3.0680235770144474e-05,
      "loss": 2.5047634124755858,
      "memory(GiB)": 77.56,
      "step": 73090,
      "token_acc": 0.46905537459283386,
      "train_speed(iter/s)": 1.437352
    },
    {
      "epoch": 3.1316138982905617,
      "grad_norm": 8.253219604492188,
      "learning_rate": 3.067402885149642e-05,
      "loss": 2.1983312606811523,
      "memory(GiB)": 77.56,
      "step": 73095,
      "token_acc": 0.528169014084507,
      "train_speed(iter/s)": 1.437373
    },
    {
      "epoch": 3.1318281136198105,
      "grad_norm": 5.186071872711182,
      "learning_rate": 3.066782228295645e-05,
      "loss": 2.43719596862793,
      "memory(GiB)": 77.56,
      "step": 73100,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.437397
    },
    {
      "epoch": 3.1320423289490598,
      "grad_norm": 6.122367858886719,
      "learning_rate": 3.0661616064637025e-05,
      "loss": 2.507410430908203,
      "memory(GiB)": 77.56,
      "step": 73105,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.437413
    },
    {
      "epoch": 3.1322565442783086,
      "grad_norm": 5.829648494720459,
      "learning_rate": 3.065541019665057e-05,
      "loss": 2.292179489135742,
      "memory(GiB)": 77.56,
      "step": 73110,
      "token_acc": 0.4935897435897436,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 3.1324707596075574,
      "grad_norm": 5.905495643615723,
      "learning_rate": 3.0649204679109494e-05,
      "loss": 2.0784915924072265,
      "memory(GiB)": 77.56,
      "step": 73115,
      "token_acc": 0.5633333333333334,
      "train_speed(iter/s)": 1.437431
    },
    {
      "epoch": 3.1326849749368066,
      "grad_norm": 4.417581558227539,
      "learning_rate": 3.064299951212624e-05,
      "loss": 2.420480728149414,
      "memory(GiB)": 77.56,
      "step": 73120,
      "token_acc": 0.4813664596273292,
      "train_speed(iter/s)": 1.437417
    },
    {
      "epoch": 3.1328991902660555,
      "grad_norm": 7.090388298034668,
      "learning_rate": 3.06367946958132e-05,
      "loss": 2.0409313201904298,
      "memory(GiB)": 77.56,
      "step": 73125,
      "token_acc": 0.5270758122743683,
      "train_speed(iter/s)": 1.437408
    },
    {
      "epoch": 3.1331134055953043,
      "grad_norm": 5.905135631561279,
      "learning_rate": 3.0630590230282783e-05,
      "loss": 2.4637382507324217,
      "memory(GiB)": 77.56,
      "step": 73130,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.437429
    },
    {
      "epoch": 3.1333276209245535,
      "grad_norm": 5.71701192855835,
      "learning_rate": 3.0624386115647386e-05,
      "loss": 2.332402801513672,
      "memory(GiB)": 77.56,
      "step": 73135,
      "token_acc": 0.5057803468208093,
      "train_speed(iter/s)": 1.437405
    },
    {
      "epoch": 3.1335418362538023,
      "grad_norm": 5.1474127769470215,
      "learning_rate": 3.0618182352019434e-05,
      "loss": 2.070578384399414,
      "memory(GiB)": 77.56,
      "step": 73140,
      "token_acc": 0.541095890410959,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.133756051583051,
      "grad_norm": 6.234938621520996,
      "learning_rate": 3.061197893951128e-05,
      "loss": 2.4143468856811525,
      "memory(GiB)": 77.56,
      "step": 73145,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 3.1339702669123004,
      "grad_norm": 8.508365631103516,
      "learning_rate": 3.060577587823531e-05,
      "loss": 2.4400894165039064,
      "memory(GiB)": 77.56,
      "step": 73150,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.43735
    },
    {
      "epoch": 3.134184482241549,
      "grad_norm": 6.250735282897949,
      "learning_rate": 3.059957316830391e-05,
      "loss": 2.329109954833984,
      "memory(GiB)": 77.56,
      "step": 73155,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.437355
    },
    {
      "epoch": 3.134398697570798,
      "grad_norm": 6.181001663208008,
      "learning_rate": 3.059337080982943e-05,
      "loss": 2.4302394866943358,
      "memory(GiB)": 77.56,
      "step": 73160,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.43733
    },
    {
      "epoch": 3.1346129129000473,
      "grad_norm": 6.349664211273193,
      "learning_rate": 3.0587168802924266e-05,
      "loss": 2.741741180419922,
      "memory(GiB)": 77.56,
      "step": 73165,
      "token_acc": 0.44745762711864406,
      "train_speed(iter/s)": 1.437327
    },
    {
      "epoch": 3.134827128229296,
      "grad_norm": 4.164112567901611,
      "learning_rate": 3.058096714770074e-05,
      "loss": 2.390481376647949,
      "memory(GiB)": 77.56,
      "step": 73170,
      "token_acc": 0.5347432024169184,
      "train_speed(iter/s)": 1.437338
    },
    {
      "epoch": 3.135041343558545,
      "grad_norm": 5.895599842071533,
      "learning_rate": 3.05747658442712e-05,
      "loss": 2.263382911682129,
      "memory(GiB)": 77.56,
      "step": 73175,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.437332
    },
    {
      "epoch": 3.135255558887794,
      "grad_norm": 7.887008190155029,
      "learning_rate": 3.0568564892748005e-05,
      "loss": 2.222603988647461,
      "memory(GiB)": 77.56,
      "step": 73180,
      "token_acc": 0.5344262295081967,
      "train_speed(iter/s)": 1.437341
    },
    {
      "epoch": 3.135469774217043,
      "grad_norm": 6.08693790435791,
      "learning_rate": 3.0562364293243496e-05,
      "loss": 2.2647979736328123,
      "memory(GiB)": 77.56,
      "step": 73185,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.437329
    },
    {
      "epoch": 3.135683989546292,
      "grad_norm": 5.291981220245361,
      "learning_rate": 3.055616404586998e-05,
      "loss": 2.1242427825927734,
      "memory(GiB)": 77.56,
      "step": 73190,
      "token_acc": 0.5610687022900763,
      "train_speed(iter/s)": 1.437334
    },
    {
      "epoch": 3.135898204875541,
      "grad_norm": 6.976221084594727,
      "learning_rate": 3.05499641507398e-05,
      "loss": 2.2210906982421874,
      "memory(GiB)": 77.56,
      "step": 73195,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 3.13611242020479,
      "grad_norm": 7.151931285858154,
      "learning_rate": 3.0543764607965256e-05,
      "loss": 2.284947967529297,
      "memory(GiB)": 77.56,
      "step": 73200,
      "token_acc": 0.5144694533762058,
      "train_speed(iter/s)": 1.437295
    },
    {
      "epoch": 3.1363266355340387,
      "grad_norm": 5.4472737312316895,
      "learning_rate": 3.053756541765867e-05,
      "loss": 2.198176383972168,
      "memory(GiB)": 77.56,
      "step": 73205,
      "token_acc": 0.5524691358024691,
      "train_speed(iter/s)": 1.437302
    },
    {
      "epoch": 3.136540850863288,
      "grad_norm": 4.815734386444092,
      "learning_rate": 3.053136657993233e-05,
      "loss": 2.405898666381836,
      "memory(GiB)": 77.56,
      "step": 73210,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.437311
    },
    {
      "epoch": 3.1367550661925367,
      "grad_norm": 5.370924472808838,
      "learning_rate": 3.0525168094898566e-05,
      "loss": 2.2557302474975587,
      "memory(GiB)": 77.56,
      "step": 73215,
      "token_acc": 0.5105105105105106,
      "train_speed(iter/s)": 1.437327
    },
    {
      "epoch": 3.1369692815217856,
      "grad_norm": 4.947432994842529,
      "learning_rate": 3.051896996266964e-05,
      "loss": 2.406699371337891,
      "memory(GiB)": 77.56,
      "step": 73220,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.437356
    },
    {
      "epoch": 3.137183496851035,
      "grad_norm": 5.838931083679199,
      "learning_rate": 3.0512772183357856e-05,
      "loss": 2.4366416931152344,
      "memory(GiB)": 77.56,
      "step": 73225,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.437358
    },
    {
      "epoch": 3.1373977121802836,
      "grad_norm": 6.927596569061279,
      "learning_rate": 3.050657475707549e-05,
      "loss": 2.6281930923461916,
      "memory(GiB)": 77.56,
      "step": 73230,
      "token_acc": 0.49538461538461537,
      "train_speed(iter/s)": 1.437373
    },
    {
      "epoch": 3.1376119275095324,
      "grad_norm": 4.943711757659912,
      "learning_rate": 3.0500377683934796e-05,
      "loss": 2.364436721801758,
      "memory(GiB)": 77.56,
      "step": 73235,
      "token_acc": 0.5149501661129569,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 3.1378261428387817,
      "grad_norm": 5.119663238525391,
      "learning_rate": 3.0494180964048057e-05,
      "loss": 2.578693962097168,
      "memory(GiB)": 77.56,
      "step": 73240,
      "token_acc": 0.49,
      "train_speed(iter/s)": 1.437432
    },
    {
      "epoch": 3.1380403581680305,
      "grad_norm": 5.716797351837158,
      "learning_rate": 3.048798459752753e-05,
      "loss": 2.2608469009399412,
      "memory(GiB)": 77.56,
      "step": 73245,
      "token_acc": 0.5413533834586466,
      "train_speed(iter/s)": 1.437414
    },
    {
      "epoch": 3.1382545734972793,
      "grad_norm": 5.7616286277771,
      "learning_rate": 3.0481788584485467e-05,
      "loss": 2.371569061279297,
      "memory(GiB)": 77.56,
      "step": 73250,
      "token_acc": 0.5224358974358975,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 3.1384687888265286,
      "grad_norm": 6.170180320739746,
      "learning_rate": 3.0475592925034112e-05,
      "loss": 2.1868541717529295,
      "memory(GiB)": 77.56,
      "step": 73255,
      "token_acc": 0.5432835820895522,
      "train_speed(iter/s)": 1.437418
    },
    {
      "epoch": 3.1386830041557774,
      "grad_norm": 5.0818772315979,
      "learning_rate": 3.04693976192857e-05,
      "loss": 2.362918663024902,
      "memory(GiB)": 77.56,
      "step": 73260,
      "token_acc": 0.46546546546546547,
      "train_speed(iter/s)": 1.437409
    },
    {
      "epoch": 3.138897219485026,
      "grad_norm": 5.2998576164245605,
      "learning_rate": 3.0463202667352487e-05,
      "loss": 2.3311336517333983,
      "memory(GiB)": 77.56,
      "step": 73265,
      "token_acc": 0.5540540540540541,
      "train_speed(iter/s)": 1.437433
    },
    {
      "epoch": 3.1391114348142755,
      "grad_norm": 5.516914367675781,
      "learning_rate": 3.0457008069346677e-05,
      "loss": 2.358369255065918,
      "memory(GiB)": 77.56,
      "step": 73270,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.437444
    },
    {
      "epoch": 3.1393256501435243,
      "grad_norm": 6.1014628410339355,
      "learning_rate": 3.0450813825380487e-05,
      "loss": 2.443255805969238,
      "memory(GiB)": 77.56,
      "step": 73275,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 3.139539865472773,
      "grad_norm": 5.917468547821045,
      "learning_rate": 3.044461993556616e-05,
      "loss": 2.254629135131836,
      "memory(GiB)": 77.56,
      "step": 73280,
      "token_acc": 0.49609375,
      "train_speed(iter/s)": 1.437458
    },
    {
      "epoch": 3.1397540808020223,
      "grad_norm": 4.6785888671875,
      "learning_rate": 3.0438426400015864e-05,
      "loss": 2.2239089965820313,
      "memory(GiB)": 77.56,
      "step": 73285,
      "token_acc": 0.5254777070063694,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 3.139968296131271,
      "grad_norm": 8.927579879760742,
      "learning_rate": 3.0432233218841845e-05,
      "loss": 2.566057014465332,
      "memory(GiB)": 77.56,
      "step": 73290,
      "token_acc": 0.49375,
      "train_speed(iter/s)": 1.437517
    },
    {
      "epoch": 3.14018251146052,
      "grad_norm": 5.384395122528076,
      "learning_rate": 3.0426040392156264e-05,
      "loss": 2.16689453125,
      "memory(GiB)": 77.56,
      "step": 73295,
      "token_acc": 0.5610561056105611,
      "train_speed(iter/s)": 1.437538
    },
    {
      "epoch": 3.140396726789769,
      "grad_norm": 7.069342136383057,
      "learning_rate": 3.0419847920071333e-05,
      "loss": 2.1762584686279296,
      "memory(GiB)": 77.56,
      "step": 73300,
      "token_acc": 0.5657894736842105,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 3.140610942119018,
      "grad_norm": 7.190010070800781,
      "learning_rate": 3.0413655802699216e-05,
      "loss": 2.361526107788086,
      "memory(GiB)": 77.56,
      "step": 73305,
      "token_acc": 0.5030674846625767,
      "train_speed(iter/s)": 1.437556
    },
    {
      "epoch": 3.140825157448267,
      "grad_norm": 5.127769470214844,
      "learning_rate": 3.040746404015211e-05,
      "loss": 2.405225372314453,
      "memory(GiB)": 77.56,
      "step": 73310,
      "token_acc": 0.5045592705167173,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 3.141039372777516,
      "grad_norm": 6.798062324523926,
      "learning_rate": 3.0401272632542172e-05,
      "loss": 2.537810516357422,
      "memory(GiB)": 77.56,
      "step": 73315,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.437549
    },
    {
      "epoch": 3.141253588106765,
      "grad_norm": 5.879473686218262,
      "learning_rate": 3.0395081579981556e-05,
      "loss": 2.3737668991088867,
      "memory(GiB)": 77.56,
      "step": 73320,
      "token_acc": 0.4418604651162791,
      "train_speed(iter/s)": 1.43756
    },
    {
      "epoch": 3.1414678034360137,
      "grad_norm": 5.051718711853027,
      "learning_rate": 3.038889088258243e-05,
      "loss": 2.57739200592041,
      "memory(GiB)": 77.56,
      "step": 73325,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437587
    },
    {
      "epoch": 3.141682018765263,
      "grad_norm": 7.146501064300537,
      "learning_rate": 3.0382700540456954e-05,
      "loss": 2.003976058959961,
      "memory(GiB)": 77.56,
      "step": 73330,
      "token_acc": 0.5789473684210527,
      "train_speed(iter/s)": 1.437572
    },
    {
      "epoch": 3.141896234094512,
      "grad_norm": 7.560466766357422,
      "learning_rate": 3.0376510553717242e-05,
      "loss": 2.1586069107055663,
      "memory(GiB)": 77.56,
      "step": 73335,
      "token_acc": 0.5400696864111498,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 3.1421104494237606,
      "grad_norm": 4.7953290939331055,
      "learning_rate": 3.0370320922475466e-05,
      "loss": 1.9031032562255858,
      "memory(GiB)": 77.56,
      "step": 73340,
      "token_acc": 0.5912162162162162,
      "train_speed(iter/s)": 1.437596
    },
    {
      "epoch": 3.14232466475301,
      "grad_norm": 7.270368576049805,
      "learning_rate": 3.0364131646843725e-05,
      "loss": 2.426093101501465,
      "memory(GiB)": 77.56,
      "step": 73345,
      "token_acc": 0.4714828897338403,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 3.1425388800822587,
      "grad_norm": 5.501646995544434,
      "learning_rate": 3.0357942726934173e-05,
      "loss": 2.221392822265625,
      "memory(GiB)": 77.56,
      "step": 73350,
      "token_acc": 0.5462555066079295,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 3.1427530954115075,
      "grad_norm": 6.587131023406982,
      "learning_rate": 3.0351754162858913e-05,
      "loss": 2.3624759674072267,
      "memory(GiB)": 77.56,
      "step": 73355,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 3.1429673107407567,
      "grad_norm": 5.908666133880615,
      "learning_rate": 3.0345565954730036e-05,
      "loss": 2.521999168395996,
      "memory(GiB)": 77.56,
      "step": 73360,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.437656
    },
    {
      "epoch": 3.1431815260700056,
      "grad_norm": 5.3235602378845215,
      "learning_rate": 3.0339378102659678e-05,
      "loss": 2.5466480255126953,
      "memory(GiB)": 77.56,
      "step": 73365,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.437652
    },
    {
      "epoch": 3.1433957413992544,
      "grad_norm": 4.571225643157959,
      "learning_rate": 3.033319060675994e-05,
      "loss": 2.1431861877441407,
      "memory(GiB)": 77.56,
      "step": 73370,
      "token_acc": 0.5582329317269076,
      "train_speed(iter/s)": 1.437652
    },
    {
      "epoch": 3.1436099567285036,
      "grad_norm": 5.0783538818359375,
      "learning_rate": 3.0327003467142907e-05,
      "loss": 2.2377185821533203,
      "memory(GiB)": 77.56,
      "step": 73375,
      "token_acc": 0.511049723756906,
      "train_speed(iter/s)": 1.437651
    },
    {
      "epoch": 3.1438241720577524,
      "grad_norm": 6.018552303314209,
      "learning_rate": 3.032081668392065e-05,
      "loss": 2.31693058013916,
      "memory(GiB)": 77.56,
      "step": 73380,
      "token_acc": 0.4983388704318937,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 3.1440383873870013,
      "grad_norm": 7.480686187744141,
      "learning_rate": 3.0314630257205266e-05,
      "loss": 2.0678888320922852,
      "memory(GiB)": 77.56,
      "step": 73385,
      "token_acc": 0.5419847328244275,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 3.1442526027162505,
      "grad_norm": 5.459778785705566,
      "learning_rate": 3.030844418710881e-05,
      "loss": 2.4423583984375,
      "memory(GiB)": 77.56,
      "step": 73390,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 3.1444668180454993,
      "grad_norm": 5.258876323699951,
      "learning_rate": 3.0302258473743374e-05,
      "loss": 2.2689821243286135,
      "memory(GiB)": 77.56,
      "step": 73395,
      "token_acc": 0.5302491103202847,
      "train_speed(iter/s)": 1.437639
    },
    {
      "epoch": 3.144681033374748,
      "grad_norm": 5.106011867523193,
      "learning_rate": 3.0296073117221004e-05,
      "loss": 2.3563161849975587,
      "memory(GiB)": 77.56,
      "step": 73400,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 3.1448952487039974,
      "grad_norm": 9.77120590209961,
      "learning_rate": 3.028988811765374e-05,
      "loss": 2.4008644104003904,
      "memory(GiB)": 77.56,
      "step": 73405,
      "token_acc": 0.5482625482625483,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 3.145109464033246,
      "grad_norm": 5.13966703414917,
      "learning_rate": 3.0283703475153656e-05,
      "loss": 2.526101303100586,
      "memory(GiB)": 77.56,
      "step": 73410,
      "token_acc": 0.48476454293628807,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 3.145323679362495,
      "grad_norm": 6.0968756675720215,
      "learning_rate": 3.0277519189832777e-05,
      "loss": 2.543075180053711,
      "memory(GiB)": 77.56,
      "step": 73415,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 3.1455378946917443,
      "grad_norm": 6.282031059265137,
      "learning_rate": 3.0271335261803124e-05,
      "loss": 2.39583797454834,
      "memory(GiB)": 77.56,
      "step": 73420,
      "token_acc": 0.4965034965034965,
      "train_speed(iter/s)": 1.437639
    },
    {
      "epoch": 3.145752110020993,
      "grad_norm": 5.595254421234131,
      "learning_rate": 3.0265151691176756e-05,
      "loss": 2.2169618606567383,
      "memory(GiB)": 77.56,
      "step": 73425,
      "token_acc": 0.5284280936454849,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 3.145966325350242,
      "grad_norm": 6.173315048217773,
      "learning_rate": 3.0258968478065665e-05,
      "loss": 2.4594953536987303,
      "memory(GiB)": 77.56,
      "step": 73430,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 3.146180540679491,
      "grad_norm": 5.420690059661865,
      "learning_rate": 3.025278562258188e-05,
      "loss": 2.7817768096923827,
      "memory(GiB)": 77.56,
      "step": 73435,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.14639475600874,
      "grad_norm": 4.7022199630737305,
      "learning_rate": 3.0246603124837414e-05,
      "loss": 2.3445222854614256,
      "memory(GiB)": 77.56,
      "step": 73440,
      "token_acc": 0.49079754601226994,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 3.146608971337989,
      "grad_norm": 5.661627769470215,
      "learning_rate": 3.024042098494426e-05,
      "loss": 2.210631561279297,
      "memory(GiB)": 77.56,
      "step": 73445,
      "token_acc": 0.5126050420168067,
      "train_speed(iter/s)": 1.437629
    },
    {
      "epoch": 3.146823186667238,
      "grad_norm": 4.719162464141846,
      "learning_rate": 3.0234239203014413e-05,
      "loss": 2.2837533950805664,
      "memory(GiB)": 77.56,
      "step": 73450,
      "token_acc": 0.5320754716981132,
      "train_speed(iter/s)": 1.437655
    },
    {
      "epoch": 3.147037401996487,
      "grad_norm": 5.719695568084717,
      "learning_rate": 3.022805777915988e-05,
      "loss": 2.3076456069946287,
      "memory(GiB)": 77.56,
      "step": 73455,
      "token_acc": 0.5110410094637224,
      "train_speed(iter/s)": 1.437654
    },
    {
      "epoch": 3.1472516173257357,
      "grad_norm": 5.263019561767578,
      "learning_rate": 3.022187671349262e-05,
      "loss": 2.395330619812012,
      "memory(GiB)": 77.56,
      "step": 73460,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 3.147465832654985,
      "grad_norm": 5.4647040367126465,
      "learning_rate": 3.0215696006124606e-05,
      "loss": 2.2821044921875,
      "memory(GiB)": 77.56,
      "step": 73465,
      "token_acc": 0.5160349854227405,
      "train_speed(iter/s)": 1.43766
    },
    {
      "epoch": 3.1476800479842337,
      "grad_norm": 6.213204383850098,
      "learning_rate": 3.020951565716783e-05,
      "loss": 2.476725959777832,
      "memory(GiB)": 77.56,
      "step": 73470,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 3.1478942633134825,
      "grad_norm": 5.980068206787109,
      "learning_rate": 3.0203335666734244e-05,
      "loss": 2.4718374252319335,
      "memory(GiB)": 77.56,
      "step": 73475,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.437684
    },
    {
      "epoch": 3.148108478642732,
      "grad_norm": 7.020590305328369,
      "learning_rate": 3.0197156034935792e-05,
      "loss": 2.5662263870239257,
      "memory(GiB)": 77.56,
      "step": 73480,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.437686
    },
    {
      "epoch": 3.1483226939719806,
      "grad_norm": 5.213085174560547,
      "learning_rate": 3.019097676188445e-05,
      "loss": 2.4975818634033202,
      "memory(GiB)": 77.56,
      "step": 73485,
      "token_acc": 0.44477611940298506,
      "train_speed(iter/s)": 1.437702
    },
    {
      "epoch": 3.1485369093012294,
      "grad_norm": 5.716728210449219,
      "learning_rate": 3.0184797847692126e-05,
      "loss": 2.3468257904052736,
      "memory(GiB)": 77.56,
      "step": 73490,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 3.1487511246304787,
      "grad_norm": 4.673936367034912,
      "learning_rate": 3.01786192924708e-05,
      "loss": 2.34005126953125,
      "memory(GiB)": 77.56,
      "step": 73495,
      "token_acc": 0.5247524752475248,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 3.1489653399597275,
      "grad_norm": 4.815449237823486,
      "learning_rate": 3.017244109633237e-05,
      "loss": 2.2413490295410154,
      "memory(GiB)": 77.56,
      "step": 73500,
      "token_acc": 0.5307443365695793,
      "train_speed(iter/s)": 1.43771
    },
    {
      "epoch": 3.1489653399597275,
      "eval_loss": 2.373683214187622,
      "eval_runtime": 13.9378,
      "eval_samples_per_second": 7.175,
      "eval_steps_per_second": 7.175,
      "eval_token_acc": 0.4917808219178082,
      "step": 73500
    },
    {
      "epoch": 3.1491795552889763,
      "grad_norm": 7.076644420623779,
      "learning_rate": 3.016626325938875e-05,
      "loss": 2.581821060180664,
      "memory(GiB)": 77.56,
      "step": 73505,
      "token_acc": 0.49080348499515974,
      "train_speed(iter/s)": 1.437291
    },
    {
      "epoch": 3.1493937706182256,
      "grad_norm": 5.052487850189209,
      "learning_rate": 3.016008578175189e-05,
      "loss": 2.4996145248413084,
      "memory(GiB)": 77.56,
      "step": 73510,
      "token_acc": 0.4747191011235955,
      "train_speed(iter/s)": 1.437282
    },
    {
      "epoch": 3.1496079859474744,
      "grad_norm": 6.75240421295166,
      "learning_rate": 3.0153908663533693e-05,
      "loss": 1.9786026000976562,
      "memory(GiB)": 77.56,
      "step": 73515,
      "token_acc": 0.5919117647058824,
      "train_speed(iter/s)": 1.437285
    },
    {
      "epoch": 3.149822201276723,
      "grad_norm": 9.193188667297363,
      "learning_rate": 3.0147731904846067e-05,
      "loss": 2.526492881774902,
      "memory(GiB)": 77.56,
      "step": 73520,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.437287
    },
    {
      "epoch": 3.1500364166059724,
      "grad_norm": 7.227169036865234,
      "learning_rate": 3.014155550580088e-05,
      "loss": 2.2694686889648437,
      "memory(GiB)": 77.56,
      "step": 73525,
      "token_acc": 0.5346153846153846,
      "train_speed(iter/s)": 1.437286
    },
    {
      "epoch": 3.1502506319352213,
      "grad_norm": 4.513391017913818,
      "learning_rate": 3.0135379466510062e-05,
      "loss": 2.4309270858764647,
      "memory(GiB)": 77.56,
      "step": 73530,
      "token_acc": 0.47794117647058826,
      "train_speed(iter/s)": 1.437296
    },
    {
      "epoch": 3.15046484726447,
      "grad_norm": 7.7047200202941895,
      "learning_rate": 3.012920378708546e-05,
      "loss": 2.283696174621582,
      "memory(GiB)": 77.56,
      "step": 73535,
      "token_acc": 0.5169230769230769,
      "train_speed(iter/s)": 1.437304
    },
    {
      "epoch": 3.1506790625937193,
      "grad_norm": 5.666827201843262,
      "learning_rate": 3.0123028467638992e-05,
      "loss": 2.3855566024780273,
      "memory(GiB)": 77.56,
      "step": 73540,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437302
    },
    {
      "epoch": 3.150893277922968,
      "grad_norm": 6.268438339233398,
      "learning_rate": 3.01168535082825e-05,
      "loss": 2.542085647583008,
      "memory(GiB)": 77.56,
      "step": 73545,
      "token_acc": 0.4773413897280967,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 3.151107493252217,
      "grad_norm": 5.260461807250977,
      "learning_rate": 3.0110678909127855e-05,
      "loss": 2.4235010147094727,
      "memory(GiB)": 77.56,
      "step": 73550,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.437287
    },
    {
      "epoch": 3.151321708581466,
      "grad_norm": 6.1962409019470215,
      "learning_rate": 3.0104504670286926e-05,
      "loss": 2.5472953796386717,
      "memory(GiB)": 77.56,
      "step": 73555,
      "token_acc": 0.450920245398773,
      "train_speed(iter/s)": 1.4373
    },
    {
      "epoch": 3.151535923910715,
      "grad_norm": 5.057364463806152,
      "learning_rate": 3.009833079187156e-05,
      "loss": 2.2611093521118164,
      "memory(GiB)": 77.56,
      "step": 73560,
      "token_acc": 0.5478547854785478,
      "train_speed(iter/s)": 1.437307
    },
    {
      "epoch": 3.151750139239964,
      "grad_norm": 6.639054775238037,
      "learning_rate": 3.009215727399359e-05,
      "loss": 2.2952062606811525,
      "memory(GiB)": 77.56,
      "step": 73565,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 3.151964354569213,
      "grad_norm": 6.600549221038818,
      "learning_rate": 3.0085984116764877e-05,
      "loss": 2.3072265625,
      "memory(GiB)": 77.56,
      "step": 73570,
      "token_acc": 0.5475285171102662,
      "train_speed(iter/s)": 1.437323
    },
    {
      "epoch": 3.152178569898462,
      "grad_norm": 4.773643493652344,
      "learning_rate": 3.0079811320297235e-05,
      "loss": 2.2260398864746094,
      "memory(GiB)": 77.56,
      "step": 73575,
      "token_acc": 0.5387323943661971,
      "train_speed(iter/s)": 1.437321
    },
    {
      "epoch": 3.1523927852277107,
      "grad_norm": 6.038623332977295,
      "learning_rate": 3.007363888470249e-05,
      "loss": 2.4065332412719727,
      "memory(GiB)": 77.56,
      "step": 73580,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 3.15260700055696,
      "grad_norm": 8.307723045349121,
      "learning_rate": 3.0067466810092475e-05,
      "loss": 2.1981456756591795,
      "memory(GiB)": 77.56,
      "step": 73585,
      "token_acc": 0.5183673469387755,
      "train_speed(iter/s)": 1.437357
    },
    {
      "epoch": 3.152821215886209,
      "grad_norm": 7.262567520141602,
      "learning_rate": 3.0061295096579008e-05,
      "loss": 2.6510921478271485,
      "memory(GiB)": 77.56,
      "step": 73590,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.1530354312154576,
      "grad_norm": 5.44913387298584,
      "learning_rate": 3.005512374427387e-05,
      "loss": 2.5405113220214846,
      "memory(GiB)": 77.56,
      "step": 73595,
      "token_acc": 0.4528985507246377,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 3.153249646544707,
      "grad_norm": 6.637559413909912,
      "learning_rate": 3.0048952753288882e-05,
      "loss": 2.5898208618164062,
      "memory(GiB)": 77.56,
      "step": 73600,
      "token_acc": 0.5059760956175299,
      "train_speed(iter/s)": 1.437355
    },
    {
      "epoch": 3.1534638618739557,
      "grad_norm": 5.489455699920654,
      "learning_rate": 3.004278212373584e-05,
      "loss": 2.2588163375854493,
      "memory(GiB)": 77.56,
      "step": 73605,
      "token_acc": 0.5392491467576792,
      "train_speed(iter/s)": 1.43736
    },
    {
      "epoch": 3.1536780772032045,
      "grad_norm": 6.767765045166016,
      "learning_rate": 3.003661185572651e-05,
      "loss": 2.50377197265625,
      "memory(GiB)": 77.56,
      "step": 73610,
      "token_acc": 0.4487534626038781,
      "train_speed(iter/s)": 1.437388
    },
    {
      "epoch": 3.1538922925324537,
      "grad_norm": 5.8337883949279785,
      "learning_rate": 3.003044194937269e-05,
      "loss": 2.4526750564575197,
      "memory(GiB)": 77.56,
      "step": 73615,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.1541065078617025,
      "grad_norm": 6.908928394317627,
      "learning_rate": 3.0024272404786146e-05,
      "loss": 2.196597862243652,
      "memory(GiB)": 77.56,
      "step": 73620,
      "token_acc": 0.5176470588235295,
      "train_speed(iter/s)": 1.437392
    },
    {
      "epoch": 3.1543207231909514,
      "grad_norm": 6.508188724517822,
      "learning_rate": 3.0018103222078643e-05,
      "loss": 2.594831848144531,
      "memory(GiB)": 77.56,
      "step": 73625,
      "token_acc": 0.4793388429752066,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.1545349385202006,
      "grad_norm": 5.630304336547852,
      "learning_rate": 3.001193440136194e-05,
      "loss": 2.3586090087890623,
      "memory(GiB)": 77.56,
      "step": 73630,
      "token_acc": 0.5272206303724928,
      "train_speed(iter/s)": 1.437349
    },
    {
      "epoch": 3.1547491538494494,
      "grad_norm": 6.659980297088623,
      "learning_rate": 3.00057659427478e-05,
      "loss": 2.2676794052124025,
      "memory(GiB)": 77.56,
      "step": 73635,
      "token_acc": 0.5325670498084292,
      "train_speed(iter/s)": 1.437357
    },
    {
      "epoch": 3.1549633691786982,
      "grad_norm": 7.744792938232422,
      "learning_rate": 2.999959784634797e-05,
      "loss": 2.2578895568847654,
      "memory(GiB)": 77.56,
      "step": 73640,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.437374
    },
    {
      "epoch": 3.1551775845079475,
      "grad_norm": 4.810945987701416,
      "learning_rate": 2.999343011227419e-05,
      "loss": 2.0620615005493166,
      "memory(GiB)": 77.56,
      "step": 73645,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 3.1553917998371963,
      "grad_norm": 5.521345138549805,
      "learning_rate": 2.9987262740638178e-05,
      "loss": 2.5106502532958985,
      "memory(GiB)": 77.56,
      "step": 73650,
      "token_acc": 0.47564469914040114,
      "train_speed(iter/s)": 1.437399
    },
    {
      "epoch": 3.155606015166445,
      "grad_norm": 7.049400329589844,
      "learning_rate": 2.9981095731551666e-05,
      "loss": 2.377778434753418,
      "memory(GiB)": 77.56,
      "step": 73655,
      "token_acc": 0.5100401606425703,
      "train_speed(iter/s)": 1.437432
    },
    {
      "epoch": 3.1558202304956944,
      "grad_norm": 5.648694038391113,
      "learning_rate": 2.99749290851264e-05,
      "loss": 2.1959930419921876,
      "memory(GiB)": 77.56,
      "step": 73660,
      "token_acc": 0.5539033457249071,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.156034445824943,
      "grad_norm": 7.972008228302002,
      "learning_rate": 2.9968762801474087e-05,
      "loss": 2.2821102142333984,
      "memory(GiB)": 77.56,
      "step": 73665,
      "token_acc": 0.506578947368421,
      "train_speed(iter/s)": 1.437455
    },
    {
      "epoch": 3.156248661154192,
      "grad_norm": 5.772468090057373,
      "learning_rate": 2.9962596880706413e-05,
      "loss": 2.42156925201416,
      "memory(GiB)": 77.56,
      "step": 73670,
      "token_acc": 0.4684014869888476,
      "train_speed(iter/s)": 1.43748
    },
    {
      "epoch": 3.1564628764834413,
      "grad_norm": 11.613487243652344,
      "learning_rate": 2.9956431322935104e-05,
      "loss": 2.612662124633789,
      "memory(GiB)": 77.56,
      "step": 73675,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.437487
    },
    {
      "epoch": 3.15667709181269,
      "grad_norm": 5.495362758636475,
      "learning_rate": 2.995026612827183e-05,
      "loss": 2.4032793045043945,
      "memory(GiB)": 77.56,
      "step": 73680,
      "token_acc": 0.501628664495114,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 3.156891307141939,
      "grad_norm": 5.41244649887085,
      "learning_rate": 2.994410129682831e-05,
      "loss": 2.6234617233276367,
      "memory(GiB)": 77.56,
      "step": 73685,
      "token_acc": 0.48135593220338985,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.157105522471188,
      "grad_norm": 5.586054801940918,
      "learning_rate": 2.99379368287162e-05,
      "loss": 2.5046566009521483,
      "memory(GiB)": 77.56,
      "step": 73690,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 3.157319737800437,
      "grad_norm": 6.418153762817383,
      "learning_rate": 2.993177272404718e-05,
      "loss": 2.412129020690918,
      "memory(GiB)": 77.56,
      "step": 73695,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 3.1575339531296858,
      "grad_norm": 6.793228626251221,
      "learning_rate": 2.9925608982932928e-05,
      "loss": 2.508030891418457,
      "memory(GiB)": 77.56,
      "step": 73700,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.437474
    },
    {
      "epoch": 3.157748168458935,
      "grad_norm": 6.082930564880371,
      "learning_rate": 2.9919445605485107e-05,
      "loss": 2.453681755065918,
      "memory(GiB)": 77.56,
      "step": 73705,
      "token_acc": 0.45288753799392095,
      "train_speed(iter/s)": 1.437499
    },
    {
      "epoch": 3.157962383788184,
      "grad_norm": 5.249345779418945,
      "learning_rate": 2.9913282591815352e-05,
      "loss": 2.444533920288086,
      "memory(GiB)": 77.56,
      "step": 73710,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 3.1581765991174326,
      "grad_norm": 5.690240859985352,
      "learning_rate": 2.9907119942035338e-05,
      "loss": 2.0600845336914064,
      "memory(GiB)": 77.56,
      "step": 73715,
      "token_acc": 0.5457627118644067,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 3.158390814446682,
      "grad_norm": 5.2061381340026855,
      "learning_rate": 2.9900957656256685e-05,
      "loss": 2.6037893295288086,
      "memory(GiB)": 77.56,
      "step": 73720,
      "token_acc": 0.46218487394957986,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 3.1586050297759307,
      "grad_norm": 5.8762054443359375,
      "learning_rate": 2.989479573459104e-05,
      "loss": 2.401332473754883,
      "memory(GiB)": 77.56,
      "step": 73725,
      "token_acc": 0.4752851711026616,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 3.1588192451051795,
      "grad_norm": 7.79713773727417,
      "learning_rate": 2.9888634177150022e-05,
      "loss": 2.3656606674194336,
      "memory(GiB)": 77.56,
      "step": 73730,
      "token_acc": 0.48638132295719844,
      "train_speed(iter/s)": 1.437551
    },
    {
      "epoch": 3.159033460434429,
      "grad_norm": 5.791051387786865,
      "learning_rate": 2.9882472984045277e-05,
      "loss": 2.2346126556396486,
      "memory(GiB)": 77.56,
      "step": 73735,
      "token_acc": 0.5340909090909091,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 3.1592476757636776,
      "grad_norm": 5.915251731872559,
      "learning_rate": 2.9876312155388398e-05,
      "loss": 2.4448097229003904,
      "memory(GiB)": 77.56,
      "step": 73740,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 3.1594618910929264,
      "grad_norm": 5.226356506347656,
      "learning_rate": 2.987015169129101e-05,
      "loss": 2.174884033203125,
      "memory(GiB)": 77.56,
      "step": 73745,
      "token_acc": 0.5226586102719033,
      "train_speed(iter/s)": 1.437563
    },
    {
      "epoch": 3.1596761064221757,
      "grad_norm": 6.0764241218566895,
      "learning_rate": 2.9863991591864705e-05,
      "loss": 2.2783248901367186,
      "memory(GiB)": 77.56,
      "step": 73750,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.437567
    },
    {
      "epoch": 3.1598903217514245,
      "grad_norm": 6.55917501449585,
      "learning_rate": 2.9857831857221075e-05,
      "loss": 2.367101860046387,
      "memory(GiB)": 77.56,
      "step": 73755,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 3.1601045370806733,
      "grad_norm": 5.933302879333496,
      "learning_rate": 2.9851672487471728e-05,
      "loss": 2.1851858139038085,
      "memory(GiB)": 77.56,
      "step": 73760,
      "token_acc": 0.5032894736842105,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 3.1603187524099225,
      "grad_norm": 4.586269378662109,
      "learning_rate": 2.9845513482728217e-05,
      "loss": 2.1989646911621095,
      "memory(GiB)": 77.56,
      "step": 73765,
      "token_acc": 0.5399361022364217,
      "train_speed(iter/s)": 1.437594
    },
    {
      "epoch": 3.1605329677391714,
      "grad_norm": 5.770150661468506,
      "learning_rate": 2.983935484310215e-05,
      "loss": 1.866630744934082,
      "memory(GiB)": 77.56,
      "step": 73770,
      "token_acc": 0.6,
      "train_speed(iter/s)": 1.437604
    },
    {
      "epoch": 3.16074718306842,
      "grad_norm": 6.130843639373779,
      "learning_rate": 2.983319656870508e-05,
      "loss": 2.2004411697387694,
      "memory(GiB)": 77.56,
      "step": 73775,
      "token_acc": 0.5251798561151079,
      "train_speed(iter/s)": 1.437605
    },
    {
      "epoch": 3.1609613983976694,
      "grad_norm": 5.802975177764893,
      "learning_rate": 2.9827038659648566e-05,
      "loss": 2.0468700408935545,
      "memory(GiB)": 77.56,
      "step": 73780,
      "token_acc": 0.552,
      "train_speed(iter/s)": 1.437619
    },
    {
      "epoch": 3.1611756137269182,
      "grad_norm": 4.6417670249938965,
      "learning_rate": 2.982088111604418e-05,
      "loss": 2.3604427337646485,
      "memory(GiB)": 77.56,
      "step": 73785,
      "token_acc": 0.5458015267175572,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 3.161389829056167,
      "grad_norm": 6.024120807647705,
      "learning_rate": 2.9814723938003463e-05,
      "loss": 2.198708152770996,
      "memory(GiB)": 77.56,
      "step": 73790,
      "token_acc": 0.5150375939849624,
      "train_speed(iter/s)": 1.437627
    },
    {
      "epoch": 3.1616040443854163,
      "grad_norm": 5.336095809936523,
      "learning_rate": 2.980856712563794e-05,
      "loss": 2.5639265060424803,
      "memory(GiB)": 77.56,
      "step": 73795,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 3.161818259714665,
      "grad_norm": 4.918575763702393,
      "learning_rate": 2.9802410679059163e-05,
      "loss": 2.261923408508301,
      "memory(GiB)": 77.56,
      "step": 73800,
      "token_acc": 0.5091463414634146,
      "train_speed(iter/s)": 1.437639
    },
    {
      "epoch": 3.162032475043914,
      "grad_norm": 5.868687152862549,
      "learning_rate": 2.9796254598378683e-05,
      "loss": 2.3623683929443358,
      "memory(GiB)": 77.56,
      "step": 73805,
      "token_acc": 0.4860335195530726,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 3.162246690373163,
      "grad_norm": 5.617335319519043,
      "learning_rate": 2.9790098883707996e-05,
      "loss": 2.1367895126342775,
      "memory(GiB)": 77.56,
      "step": 73810,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 3.162460905702412,
      "grad_norm": 6.260705947875977,
      "learning_rate": 2.978394353515862e-05,
      "loss": 2.2816699981689452,
      "memory(GiB)": 77.56,
      "step": 73815,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 3.162675121031661,
      "grad_norm": 4.895193576812744,
      "learning_rate": 2.977778855284208e-05,
      "loss": 2.186238098144531,
      "memory(GiB)": 77.56,
      "step": 73820,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 3.16288933636091,
      "grad_norm": 4.447147369384766,
      "learning_rate": 2.9771633936869863e-05,
      "loss": 2.0569997787475587,
      "memory(GiB)": 77.56,
      "step": 73825,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.437681
    },
    {
      "epoch": 3.163103551690159,
      "grad_norm": 6.592648506164551,
      "learning_rate": 2.976547968735348e-05,
      "loss": 2.391103744506836,
      "memory(GiB)": 77.56,
      "step": 73830,
      "token_acc": 0.5099337748344371,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 3.163317767019408,
      "grad_norm": 5.578714847564697,
      "learning_rate": 2.9759325804404418e-05,
      "loss": 2.606415939331055,
      "memory(GiB)": 77.56,
      "step": 73835,
      "token_acc": 0.4797507788161994,
      "train_speed(iter/s)": 1.437675
    },
    {
      "epoch": 3.163531982348657,
      "grad_norm": 4.855007171630859,
      "learning_rate": 2.9753172288134146e-05,
      "loss": 2.2682567596435548,
      "memory(GiB)": 77.56,
      "step": 73840,
      "token_acc": 0.5544554455445545,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 3.1637461976779058,
      "grad_norm": 6.128581523895264,
      "learning_rate": 2.9747019138654157e-05,
      "loss": 2.4982378005981447,
      "memory(GiB)": 77.56,
      "step": 73845,
      "token_acc": 0.47468354430379744,
      "train_speed(iter/s)": 1.437699
    },
    {
      "epoch": 3.163960413007155,
      "grad_norm": 5.198617458343506,
      "learning_rate": 2.974086635607592e-05,
      "loss": 2.2245336532592774,
      "memory(GiB)": 77.56,
      "step": 73850,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 3.164174628336404,
      "grad_norm": 6.806390285491943,
      "learning_rate": 2.9734713940510884e-05,
      "loss": 2.308121109008789,
      "memory(GiB)": 77.56,
      "step": 73855,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 3.1643888436656527,
      "grad_norm": 5.530201435089111,
      "learning_rate": 2.9728561892070518e-05,
      "loss": 2.382350730895996,
      "memory(GiB)": 77.56,
      "step": 73860,
      "token_acc": 0.5203761755485894,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 3.164603058994902,
      "grad_norm": 7.2378740310668945,
      "learning_rate": 2.972241021086627e-05,
      "loss": 2.425258445739746,
      "memory(GiB)": 77.56,
      "step": 73865,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.1648172743241507,
      "grad_norm": 6.611432075500488,
      "learning_rate": 2.9716258897009586e-05,
      "loss": 2.351436424255371,
      "memory(GiB)": 77.56,
      "step": 73870,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.1650314896533995,
      "grad_norm": 5.041432857513428,
      "learning_rate": 2.9710107950611878e-05,
      "loss": 2.4247488021850585,
      "memory(GiB)": 77.56,
      "step": 73875,
      "token_acc": 0.521311475409836,
      "train_speed(iter/s)": 1.437755
    },
    {
      "epoch": 3.165245704982649,
      "grad_norm": 6.774808406829834,
      "learning_rate": 2.970395737178462e-05,
      "loss": 2.2830532073974608,
      "memory(GiB)": 77.56,
      "step": 73880,
      "token_acc": 0.5492957746478874,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 3.1654599203118976,
      "grad_norm": 6.014718055725098,
      "learning_rate": 2.96978071606392e-05,
      "loss": 2.3167903900146483,
      "memory(GiB)": 77.56,
      "step": 73885,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 3.1656741356411464,
      "grad_norm": 5.995301246643066,
      "learning_rate": 2.9691657317287068e-05,
      "loss": 2.320265197753906,
      "memory(GiB)": 77.56,
      "step": 73890,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.1658883509703957,
      "grad_norm": 5.244626998901367,
      "learning_rate": 2.968550784183961e-05,
      "loss": 2.071191596984863,
      "memory(GiB)": 77.56,
      "step": 73895,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 3.1661025662996445,
      "grad_norm": 6.777777194976807,
      "learning_rate": 2.967935873440822e-05,
      "loss": 2.31115837097168,
      "memory(GiB)": 77.56,
      "step": 73900,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 3.1663167816288933,
      "grad_norm": 5.80156946182251,
      "learning_rate": 2.9673209995104322e-05,
      "loss": 2.234435272216797,
      "memory(GiB)": 77.56,
      "step": 73905,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.437786
    },
    {
      "epoch": 3.1665309969581426,
      "grad_norm": 6.019159317016602,
      "learning_rate": 2.9667061624039282e-05,
      "loss": 2.342307281494141,
      "memory(GiB)": 77.56,
      "step": 73910,
      "token_acc": 0.5163398692810458,
      "train_speed(iter/s)": 1.437805
    },
    {
      "epoch": 3.1667452122873914,
      "grad_norm": 5.571842670440674,
      "learning_rate": 2.9660913621324503e-05,
      "loss": 2.425028610229492,
      "memory(GiB)": 77.56,
      "step": 73915,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.16695942761664,
      "grad_norm": 5.433260917663574,
      "learning_rate": 2.965476598707136e-05,
      "loss": 2.191996383666992,
      "memory(GiB)": 77.56,
      "step": 73920,
      "token_acc": 0.5210355987055016,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 3.1671736429458894,
      "grad_norm": 5.574416637420654,
      "learning_rate": 2.9648618721391197e-05,
      "loss": 2.1054168701171876,
      "memory(GiB)": 77.56,
      "step": 73925,
      "token_acc": 0.5457627118644067,
      "train_speed(iter/s)": 1.437833
    },
    {
      "epoch": 3.1673878582751382,
      "grad_norm": 8.710840225219727,
      "learning_rate": 2.9642471824395413e-05,
      "loss": 2.3504207611083983,
      "memory(GiB)": 77.56,
      "step": 73930,
      "token_acc": 0.5137931034482759,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 3.167602073604387,
      "grad_norm": 5.431893825531006,
      "learning_rate": 2.963632529619535e-05,
      "loss": 2.458882713317871,
      "memory(GiB)": 77.56,
      "step": 73935,
      "token_acc": 0.47904191616766467,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 3.1678162889336363,
      "grad_norm": 5.714679718017578,
      "learning_rate": 2.9630179136902346e-05,
      "loss": 2.395102691650391,
      "memory(GiB)": 77.56,
      "step": 73940,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.437808
    },
    {
      "epoch": 3.168030504262885,
      "grad_norm": 5.728879451751709,
      "learning_rate": 2.9624033346627766e-05,
      "loss": 1.9084863662719727,
      "memory(GiB)": 77.56,
      "step": 73945,
      "token_acc": 0.5925925925925926,
      "train_speed(iter/s)": 1.437807
    },
    {
      "epoch": 3.168244719592134,
      "grad_norm": 5.877350330352783,
      "learning_rate": 2.9617887925482914e-05,
      "loss": 1.980107307434082,
      "memory(GiB)": 77.56,
      "step": 73950,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.437805
    },
    {
      "epoch": 3.168458934921383,
      "grad_norm": 5.895495891571045,
      "learning_rate": 2.961174287357916e-05,
      "loss": 2.2974740982055666,
      "memory(GiB)": 77.56,
      "step": 73955,
      "token_acc": 0.5183673469387755,
      "train_speed(iter/s)": 1.437799
    },
    {
      "epoch": 3.168673150250632,
      "grad_norm": 5.26740026473999,
      "learning_rate": 2.9605598191027805e-05,
      "loss": 2.205534553527832,
      "memory(GiB)": 77.56,
      "step": 73960,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437807
    },
    {
      "epoch": 3.168887365579881,
      "grad_norm": 4.823087692260742,
      "learning_rate": 2.9599453877940175e-05,
      "loss": 2.4179988861083985,
      "memory(GiB)": 77.56,
      "step": 73965,
      "token_acc": 0.5131086142322098,
      "train_speed(iter/s)": 1.437782
    },
    {
      "epoch": 3.16910158090913,
      "grad_norm": 6.26754903793335,
      "learning_rate": 2.9593309934427565e-05,
      "loss": 2.0993356704711914,
      "memory(GiB)": 77.56,
      "step": 73970,
      "token_acc": 0.5543071161048689,
      "train_speed(iter/s)": 1.437776
    },
    {
      "epoch": 3.169315796238379,
      "grad_norm": 5.71920108795166,
      "learning_rate": 2.9587166360601303e-05,
      "loss": 2.552367401123047,
      "memory(GiB)": 77.56,
      "step": 73975,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 3.1695300115676277,
      "grad_norm": 5.638586521148682,
      "learning_rate": 2.9581023156572668e-05,
      "loss": 2.2518590927124023,
      "memory(GiB)": 77.56,
      "step": 73980,
      "token_acc": 0.4948805460750853,
      "train_speed(iter/s)": 1.437818
    },
    {
      "epoch": 3.169744226896877,
      "grad_norm": 5.522692680358887,
      "learning_rate": 2.9574880322452947e-05,
      "loss": 2.0866403579711914,
      "memory(GiB)": 77.56,
      "step": 73985,
      "token_acc": 0.5428571428571428,
      "train_speed(iter/s)": 1.437838
    },
    {
      "epoch": 3.1699584422261258,
      "grad_norm": 7.261645317077637,
      "learning_rate": 2.9568737858353436e-05,
      "loss": 2.2186960220336913,
      "memory(GiB)": 77.56,
      "step": 73990,
      "token_acc": 0.5409836065573771,
      "train_speed(iter/s)": 1.437828
    },
    {
      "epoch": 3.1701726575553746,
      "grad_norm": 6.001633167266846,
      "learning_rate": 2.9562595764385405e-05,
      "loss": 2.3324558258056642,
      "memory(GiB)": 77.56,
      "step": 73995,
      "token_acc": 0.5201612903225806,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.170386872884624,
      "grad_norm": 4.552384376525879,
      "learning_rate": 2.9556454040660114e-05,
      "loss": 2.2031044006347655,
      "memory(GiB)": 77.56,
      "step": 74000,
      "token_acc": 0.5047318611987381,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.170386872884624,
      "eval_loss": 2.236055374145508,
      "eval_runtime": 14.156,
      "eval_samples_per_second": 7.064,
      "eval_steps_per_second": 7.064,
      "eval_token_acc": 0.4821882951653944,
      "step": 74000
    },
    {
      "epoch": 3.1706010882138727,
      "grad_norm": 6.383446216583252,
      "learning_rate": 2.9550312687288846e-05,
      "loss": 2.299361801147461,
      "memory(GiB)": 77.56,
      "step": 74005,
      "token_acc": 0.489945155393053,
      "train_speed(iter/s)": 1.437414
    },
    {
      "epoch": 3.1708153035431215,
      "grad_norm": 5.767241954803467,
      "learning_rate": 2.954417170438284e-05,
      "loss": 2.245487594604492,
      "memory(GiB)": 77.56,
      "step": 74010,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.43742
    },
    {
      "epoch": 3.1710295188723707,
      "grad_norm": 7.623003959655762,
      "learning_rate": 2.9538031092053354e-05,
      "loss": 2.436071014404297,
      "memory(GiB)": 77.56,
      "step": 74015,
      "token_acc": 0.5181159420289855,
      "train_speed(iter/s)": 1.437417
    },
    {
      "epoch": 3.1712437342016195,
      "grad_norm": 5.613229274749756,
      "learning_rate": 2.9531890850411637e-05,
      "loss": 2.3619190216064454,
      "memory(GiB)": 77.56,
      "step": 74020,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 3.1714579495308683,
      "grad_norm": 4.912458419799805,
      "learning_rate": 2.9525750979568895e-05,
      "loss": 2.6781641006469727,
      "memory(GiB)": 77.56,
      "step": 74025,
      "token_acc": 0.46742209631728043,
      "train_speed(iter/s)": 1.437408
    },
    {
      "epoch": 3.1716721648601176,
      "grad_norm": 6.433786869049072,
      "learning_rate": 2.951961147963639e-05,
      "loss": 2.1593416213989256,
      "memory(GiB)": 77.56,
      "step": 74030,
      "token_acc": 0.5426621160409556,
      "train_speed(iter/s)": 1.437401
    },
    {
      "epoch": 3.1718863801893664,
      "grad_norm": 5.007352828979492,
      "learning_rate": 2.9513472350725347e-05,
      "loss": 2.2726470947265627,
      "memory(GiB)": 77.56,
      "step": 74035,
      "token_acc": 0.5209003215434084,
      "train_speed(iter/s)": 1.437412
    },
    {
      "epoch": 3.1721005955186152,
      "grad_norm": 9.181182861328125,
      "learning_rate": 2.9507333592946975e-05,
      "loss": 2.4367889404296874,
      "memory(GiB)": 77.56,
      "step": 74040,
      "token_acc": 0.502092050209205,
      "train_speed(iter/s)": 1.437407
    },
    {
      "epoch": 3.1723148108478645,
      "grad_norm": 4.705042839050293,
      "learning_rate": 2.9501195206412457e-05,
      "loss": 2.103071403503418,
      "memory(GiB)": 77.56,
      "step": 74045,
      "token_acc": 0.5369774919614148,
      "train_speed(iter/s)": 1.437417
    },
    {
      "epoch": 3.1725290261771133,
      "grad_norm": 5.426323890686035,
      "learning_rate": 2.9495057191233033e-05,
      "loss": 2.3035358428955077,
      "memory(GiB)": 77.56,
      "step": 74050,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 3.172743241506362,
      "grad_norm": 4.460992336273193,
      "learning_rate": 2.9488919547519878e-05,
      "loss": 2.312650680541992,
      "memory(GiB)": 77.56,
      "step": 74055,
      "token_acc": 0.4803921568627451,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 3.1729574568356114,
      "grad_norm": 6.285937786102295,
      "learning_rate": 2.948278227538419e-05,
      "loss": 2.244205856323242,
      "memory(GiB)": 77.56,
      "step": 74060,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437422
    },
    {
      "epoch": 3.17317167216486,
      "grad_norm": 6.112287521362305,
      "learning_rate": 2.947664537493715e-05,
      "loss": 2.3975431442260744,
      "memory(GiB)": 77.56,
      "step": 74065,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 3.173385887494109,
      "grad_norm": 5.413914680480957,
      "learning_rate": 2.9470508846289924e-05,
      "loss": 2.288863754272461,
      "memory(GiB)": 77.56,
      "step": 74070,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 3.1736001028233582,
      "grad_norm": 4.8067827224731445,
      "learning_rate": 2.9464372689553692e-05,
      "loss": 2.1473098754882813,
      "memory(GiB)": 77.56,
      "step": 74075,
      "token_acc": 0.5364963503649635,
      "train_speed(iter/s)": 1.437458
    },
    {
      "epoch": 3.173814318152607,
      "grad_norm": 6.240786552429199,
      "learning_rate": 2.945823690483962e-05,
      "loss": 2.233226776123047,
      "memory(GiB)": 77.56,
      "step": 74080,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437449
    },
    {
      "epoch": 3.174028533481856,
      "grad_norm": 5.830225467681885,
      "learning_rate": 2.9452101492258843e-05,
      "loss": 2.438872528076172,
      "memory(GiB)": 77.56,
      "step": 74085,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.437453
    },
    {
      "epoch": 3.174242748811105,
      "grad_norm": 7.278238296508789,
      "learning_rate": 2.9445966451922536e-05,
      "loss": 2.0950408935546876,
      "memory(GiB)": 77.56,
      "step": 74090,
      "token_acc": 0.5433070866141733,
      "train_speed(iter/s)": 1.437439
    },
    {
      "epoch": 3.174456964140354,
      "grad_norm": 4.190337181091309,
      "learning_rate": 2.9439831783941807e-05,
      "loss": 2.219211959838867,
      "memory(GiB)": 77.56,
      "step": 74095,
      "token_acc": 0.5083056478405316,
      "train_speed(iter/s)": 1.437455
    },
    {
      "epoch": 3.1746711794696028,
      "grad_norm": 12.855223655700684,
      "learning_rate": 2.9433697488427825e-05,
      "loss": 2.272871208190918,
      "memory(GiB)": 77.56,
      "step": 74100,
      "token_acc": 0.5058365758754864,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 3.174885394798852,
      "grad_norm": 7.407022953033447,
      "learning_rate": 2.94275635654917e-05,
      "loss": 2.4018808364868165,
      "memory(GiB)": 77.56,
      "step": 74105,
      "token_acc": 0.4628975265017668,
      "train_speed(iter/s)": 1.437466
    },
    {
      "epoch": 3.175099610128101,
      "grad_norm": 4.247338771820068,
      "learning_rate": 2.9421430015244565e-05,
      "loss": 2.265575981140137,
      "memory(GiB)": 77.56,
      "step": 74110,
      "token_acc": 0.5365853658536586,
      "train_speed(iter/s)": 1.437477
    },
    {
      "epoch": 3.1753138254573496,
      "grad_norm": 5.550304412841797,
      "learning_rate": 2.941529683779753e-05,
      "loss": 1.9959667205810547,
      "memory(GiB)": 77.56,
      "step": 74115,
      "token_acc": 0.5573770491803278,
      "train_speed(iter/s)": 1.437476
    },
    {
      "epoch": 3.175528040786599,
      "grad_norm": 5.715430736541748,
      "learning_rate": 2.9409164033261704e-05,
      "loss": 2.450954627990723,
      "memory(GiB)": 77.56,
      "step": 74120,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.437475
    },
    {
      "epoch": 3.1757422561158477,
      "grad_norm": 6.707547664642334,
      "learning_rate": 2.9403031601748192e-05,
      "loss": 2.356776809692383,
      "memory(GiB)": 77.56,
      "step": 74125,
      "token_acc": 0.512280701754386,
      "train_speed(iter/s)": 1.437488
    },
    {
      "epoch": 3.1759564714450965,
      "grad_norm": 5.754292011260986,
      "learning_rate": 2.939689954336807e-05,
      "loss": 2.5592182159423826,
      "memory(GiB)": 77.56,
      "step": 74130,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 3.1761706867743458,
      "grad_norm": 6.746292591094971,
      "learning_rate": 2.939076785823246e-05,
      "loss": 2.5319929122924805,
      "memory(GiB)": 77.56,
      "step": 74135,
      "token_acc": 0.4716981132075472,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 3.1763849021035946,
      "grad_norm": 5.235986232757568,
      "learning_rate": 2.9384636546452415e-05,
      "loss": 2.2475265502929687,
      "memory(GiB)": 77.56,
      "step": 74140,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.437453
    },
    {
      "epoch": 3.1765991174328434,
      "grad_norm": 6.648682594299316,
      "learning_rate": 2.9378505608139007e-05,
      "loss": 2.495352554321289,
      "memory(GiB)": 77.56,
      "step": 74145,
      "token_acc": 0.4664310954063604,
      "train_speed(iter/s)": 1.437474
    },
    {
      "epoch": 3.1768133327620927,
      "grad_norm": 4.644205570220947,
      "learning_rate": 2.937237504340333e-05,
      "loss": 2.2440237045288085,
      "memory(GiB)": 77.56,
      "step": 74150,
      "token_acc": 0.5186335403726708,
      "train_speed(iter/s)": 1.437461
    },
    {
      "epoch": 3.1770275480913415,
      "grad_norm": 6.298264503479004,
      "learning_rate": 2.9366244852356418e-05,
      "loss": 2.146969223022461,
      "memory(GiB)": 77.56,
      "step": 74155,
      "token_acc": 0.5398550724637681,
      "train_speed(iter/s)": 1.437442
    },
    {
      "epoch": 3.1772417634205903,
      "grad_norm": 6.155228137969971,
      "learning_rate": 2.9360115035109337e-05,
      "loss": 2.1114934921264648,
      "memory(GiB)": 77.56,
      "step": 74160,
      "token_acc": 0.5394736842105263,
      "train_speed(iter/s)": 1.437476
    },
    {
      "epoch": 3.1774559787498395,
      "grad_norm": 6.646638870239258,
      "learning_rate": 2.9353985591773148e-05,
      "loss": 2.24499397277832,
      "memory(GiB)": 77.56,
      "step": 74165,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437472
    },
    {
      "epoch": 3.1776701940790884,
      "grad_norm": 5.539613246917725,
      "learning_rate": 2.934785652245885e-05,
      "loss": 2.3395923614501952,
      "memory(GiB)": 77.56,
      "step": 74170,
      "token_acc": 0.5181159420289855,
      "train_speed(iter/s)": 1.437455
    },
    {
      "epoch": 3.177884409408337,
      "grad_norm": 5.576258182525635,
      "learning_rate": 2.934172782727751e-05,
      "loss": 2.0838991165161134,
      "memory(GiB)": 77.56,
      "step": 74175,
      "token_acc": 0.5031446540880503,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 3.1780986247375864,
      "grad_norm": 5.956866264343262,
      "learning_rate": 2.9335599506340166e-05,
      "loss": 2.552425193786621,
      "memory(GiB)": 77.56,
      "step": 74180,
      "token_acc": 0.47246376811594204,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 3.1783128400668352,
      "grad_norm": 5.738117218017578,
      "learning_rate": 2.9329471559757822e-05,
      "loss": 2.1716545104980467,
      "memory(GiB)": 77.56,
      "step": 74185,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.437465
    },
    {
      "epoch": 3.178527055396084,
      "grad_norm": 5.96266508102417,
      "learning_rate": 2.932334398764148e-05,
      "loss": 2.497850227355957,
      "memory(GiB)": 77.56,
      "step": 74190,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 3.1787412707253333,
      "grad_norm": 5.279277801513672,
      "learning_rate": 2.9317216790102164e-05,
      "loss": 2.1521081924438477,
      "memory(GiB)": 77.56,
      "step": 74195,
      "token_acc": 0.525691699604743,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.178955486054582,
      "grad_norm": 7.685451507568359,
      "learning_rate": 2.931108996725086e-05,
      "loss": 2.648274040222168,
      "memory(GiB)": 77.56,
      "step": 74200,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.437492
    },
    {
      "epoch": 3.179169701383831,
      "grad_norm": 5.598180770874023,
      "learning_rate": 2.930496351919858e-05,
      "loss": 2.4224830627441407,
      "memory(GiB)": 77.56,
      "step": 74205,
      "token_acc": 0.49096385542168675,
      "train_speed(iter/s)": 1.437489
    },
    {
      "epoch": 3.17938391671308,
      "grad_norm": 6.338024139404297,
      "learning_rate": 2.9298837446056304e-05,
      "loss": 2.467112350463867,
      "memory(GiB)": 77.56,
      "step": 74210,
      "token_acc": 0.48787878787878786,
      "train_speed(iter/s)": 1.437505
    },
    {
      "epoch": 3.179598132042329,
      "grad_norm": 8.39466667175293,
      "learning_rate": 2.9292711747934993e-05,
      "loss": 2.4090782165527345,
      "memory(GiB)": 77.56,
      "step": 74215,
      "token_acc": 0.49407114624505927,
      "train_speed(iter/s)": 1.437514
    },
    {
      "epoch": 3.179812347371578,
      "grad_norm": 4.34513521194458,
      "learning_rate": 2.928658642494564e-05,
      "loss": 2.5388145446777344,
      "memory(GiB)": 77.56,
      "step": 74220,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.437517
    },
    {
      "epoch": 3.180026562700827,
      "grad_norm": 6.119314193725586,
      "learning_rate": 2.9280461477199205e-05,
      "loss": 2.5132280349731446,
      "memory(GiB)": 77.56,
      "step": 74225,
      "token_acc": 0.5328947368421053,
      "train_speed(iter/s)": 1.437528
    },
    {
      "epoch": 3.180240778030076,
      "grad_norm": 5.85638952255249,
      "learning_rate": 2.9274336904806642e-05,
      "loss": 2.2614885330200196,
      "memory(GiB)": 77.56,
      "step": 74230,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.43755
    },
    {
      "epoch": 3.1804549933593247,
      "grad_norm": 4.860142230987549,
      "learning_rate": 2.9268212707878917e-05,
      "loss": 2.096175765991211,
      "memory(GiB)": 77.56,
      "step": 74235,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.437545
    },
    {
      "epoch": 3.180669208688574,
      "grad_norm": 5.776596546173096,
      "learning_rate": 2.9262088886526962e-05,
      "loss": 2.897212028503418,
      "memory(GiB)": 77.56,
      "step": 74240,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 3.1808834240178228,
      "grad_norm": 5.11076545715332,
      "learning_rate": 2.925596544086171e-05,
      "loss": 2.604241943359375,
      "memory(GiB)": 77.56,
      "step": 74245,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.437526
    },
    {
      "epoch": 3.1810976393470716,
      "grad_norm": 7.483839988708496,
      "learning_rate": 2.9249842370994106e-05,
      "loss": 2.344086456298828,
      "memory(GiB)": 77.56,
      "step": 74250,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.4375
    },
    {
      "epoch": 3.181311854676321,
      "grad_norm": 6.672262668609619,
      "learning_rate": 2.9243719677035086e-05,
      "loss": 2.4811267852783203,
      "memory(GiB)": 77.56,
      "step": 74255,
      "token_acc": 0.46236559139784944,
      "train_speed(iter/s)": 1.437503
    },
    {
      "epoch": 3.1815260700055696,
      "grad_norm": 5.62570333480835,
      "learning_rate": 2.9237597359095547e-05,
      "loss": 2.26552791595459,
      "memory(GiB)": 77.56,
      "step": 74260,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.437534
    },
    {
      "epoch": 3.1817402853348185,
      "grad_norm": 7.12554407119751,
      "learning_rate": 2.9231475417286412e-05,
      "loss": 2.08533935546875,
      "memory(GiB)": 77.56,
      "step": 74265,
      "token_acc": 0.5353535353535354,
      "train_speed(iter/s)": 1.437563
    },
    {
      "epoch": 3.1819545006640677,
      "grad_norm": 6.84037446975708,
      "learning_rate": 2.9225353851718586e-05,
      "loss": 2.1793569564819335,
      "memory(GiB)": 77.56,
      "step": 74270,
      "token_acc": 0.5261044176706827,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 3.1821687159933165,
      "grad_norm": 5.394654273986816,
      "learning_rate": 2.9219232662502948e-05,
      "loss": 2.3360498428344725,
      "memory(GiB)": 77.56,
      "step": 74275,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.437542
    },
    {
      "epoch": 3.1823829313225653,
      "grad_norm": 5.653782844543457,
      "learning_rate": 2.9213111849750425e-05,
      "loss": 2.270713043212891,
      "memory(GiB)": 77.56,
      "step": 74280,
      "token_acc": 0.5137931034482759,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 3.1825971466518146,
      "grad_norm": 5.598612308502197,
      "learning_rate": 2.9206991413571872e-05,
      "loss": 2.1324993133544923,
      "memory(GiB)": 77.56,
      "step": 74285,
      "token_acc": 0.5211726384364821,
      "train_speed(iter/s)": 1.437555
    },
    {
      "epoch": 3.1828113619810634,
      "grad_norm": 4.982089042663574,
      "learning_rate": 2.9200871354078167e-05,
      "loss": 2.3357444763183595,
      "memory(GiB)": 77.56,
      "step": 74290,
      "token_acc": 0.5101449275362319,
      "train_speed(iter/s)": 1.437562
    },
    {
      "epoch": 3.183025577310312,
      "grad_norm": 5.30708646774292,
      "learning_rate": 2.91947516713802e-05,
      "loss": 2.3095245361328125,
      "memory(GiB)": 77.56,
      "step": 74295,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 3.1832397926395615,
      "grad_norm": 5.575718402862549,
      "learning_rate": 2.918863236558882e-05,
      "loss": 2.2222293853759765,
      "memory(GiB)": 77.56,
      "step": 74300,
      "token_acc": 0.5015479876160991,
      "train_speed(iter/s)": 1.437583
    },
    {
      "epoch": 3.1834540079688103,
      "grad_norm": 5.541013717651367,
      "learning_rate": 2.9182513436814897e-05,
      "loss": 2.4817035675048826,
      "memory(GiB)": 77.56,
      "step": 74305,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.437587
    },
    {
      "epoch": 3.183668223298059,
      "grad_norm": 6.624006748199463,
      "learning_rate": 2.9176394885169238e-05,
      "loss": 2.6079805374145506,
      "memory(GiB)": 77.56,
      "step": 74310,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.4376
    },
    {
      "epoch": 3.1838824386273084,
      "grad_norm": 6.421791076660156,
      "learning_rate": 2.9170276710762746e-05,
      "loss": 2.229039192199707,
      "memory(GiB)": 77.56,
      "step": 74315,
      "token_acc": 0.5278688524590164,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 3.184096653956557,
      "grad_norm": 5.171174049377441,
      "learning_rate": 2.9164158913706207e-05,
      "loss": 2.14434814453125,
      "memory(GiB)": 77.56,
      "step": 74320,
      "token_acc": 0.535031847133758,
      "train_speed(iter/s)": 1.437556
    },
    {
      "epoch": 3.184310869285806,
      "grad_norm": 7.4009599685668945,
      "learning_rate": 2.91580414941105e-05,
      "loss": 2.45861759185791,
      "memory(GiB)": 77.56,
      "step": 74325,
      "token_acc": 0.44964028776978415,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 3.1845250846150552,
      "grad_norm": 5.991366386413574,
      "learning_rate": 2.9151924452086428e-05,
      "loss": 2.428775596618652,
      "memory(GiB)": 77.56,
      "step": 74330,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.437555
    },
    {
      "epoch": 3.184739299944304,
      "grad_norm": 6.654433727264404,
      "learning_rate": 2.9145807787744795e-05,
      "loss": 2.4450361251831056,
      "memory(GiB)": 77.56,
      "step": 74335,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.437552
    },
    {
      "epoch": 3.184953515273553,
      "grad_norm": 8.410322189331055,
      "learning_rate": 2.91396915011964e-05,
      "loss": 2.315107727050781,
      "memory(GiB)": 77.56,
      "step": 74340,
      "token_acc": 0.5346534653465347,
      "train_speed(iter/s)": 1.437554
    },
    {
      "epoch": 3.185167730602802,
      "grad_norm": 6.487464427947998,
      "learning_rate": 2.9133575592552076e-05,
      "loss": 2.2820533752441405,
      "memory(GiB)": 77.56,
      "step": 74345,
      "token_acc": 0.4954128440366973,
      "train_speed(iter/s)": 1.437563
    },
    {
      "epoch": 3.185381945932051,
      "grad_norm": 6.504099369049072,
      "learning_rate": 2.9127460061922607e-05,
      "loss": 2.475813293457031,
      "memory(GiB)": 77.56,
      "step": 74350,
      "token_acc": 0.4884488448844885,
      "train_speed(iter/s)": 1.437562
    },
    {
      "epoch": 3.1855961612612997,
      "grad_norm": 5.420844078063965,
      "learning_rate": 2.912134490941878e-05,
      "loss": 2.141633415222168,
      "memory(GiB)": 77.56,
      "step": 74355,
      "token_acc": 0.4828897338403042,
      "train_speed(iter/s)": 1.437569
    },
    {
      "epoch": 3.185810376590549,
      "grad_norm": 6.719005107879639,
      "learning_rate": 2.911523013515137e-05,
      "loss": 2.196058654785156,
      "memory(GiB)": 77.56,
      "step": 74360,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 3.186024591919798,
      "grad_norm": 5.522883892059326,
      "learning_rate": 2.9109115739231145e-05,
      "loss": 2.181784820556641,
      "memory(GiB)": 77.56,
      "step": 74365,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 3.1862388072490466,
      "grad_norm": 8.079535484313965,
      "learning_rate": 2.910300172176891e-05,
      "loss": 2.4879117965698243,
      "memory(GiB)": 77.56,
      "step": 74370,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43762
    },
    {
      "epoch": 3.186453022578296,
      "grad_norm": 8.391444206237793,
      "learning_rate": 2.909688808287539e-05,
      "loss": 2.479786682128906,
      "memory(GiB)": 77.56,
      "step": 74375,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.437622
    },
    {
      "epoch": 3.1866672379075447,
      "grad_norm": 6.152960777282715,
      "learning_rate": 2.909077482266136e-05,
      "loss": 2.4624088287353514,
      "memory(GiB)": 77.56,
      "step": 74380,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437608
    },
    {
      "epoch": 3.1868814532367935,
      "grad_norm": 6.37076997756958,
      "learning_rate": 2.9084661941237552e-05,
      "loss": 2.2588523864746093,
      "memory(GiB)": 77.56,
      "step": 74385,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.43762
    },
    {
      "epoch": 3.1870956685660428,
      "grad_norm": 5.731784820556641,
      "learning_rate": 2.9078549438714715e-05,
      "loss": 2.25836124420166,
      "memory(GiB)": 77.56,
      "step": 74390,
      "token_acc": 0.547945205479452,
      "train_speed(iter/s)": 1.437624
    },
    {
      "epoch": 3.1873098838952916,
      "grad_norm": 6.615798473358154,
      "learning_rate": 2.907243731520356e-05,
      "loss": 2.417607879638672,
      "memory(GiB)": 77.56,
      "step": 74395,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.437637
    },
    {
      "epoch": 3.1875240992245404,
      "grad_norm": 5.539975643157959,
      "learning_rate": 2.9066325570814834e-05,
      "loss": 2.1889867782592773,
      "memory(GiB)": 77.56,
      "step": 74400,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.437654
    },
    {
      "epoch": 3.1877383145537896,
      "grad_norm": 5.761571407318115,
      "learning_rate": 2.9060214205659286e-05,
      "loss": 2.2632556915283204,
      "memory(GiB)": 77.56,
      "step": 74405,
      "token_acc": 0.47962382445141066,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 3.1879525298830385,
      "grad_norm": 5.465545177459717,
      "learning_rate": 2.9054103219847596e-05,
      "loss": 2.219378662109375,
      "memory(GiB)": 77.56,
      "step": 74410,
      "token_acc": 0.5306859205776173,
      "train_speed(iter/s)": 1.437659
    },
    {
      "epoch": 3.1881667452122873,
      "grad_norm": 6.672488212585449,
      "learning_rate": 2.904799261349047e-05,
      "loss": 2.3821691513061523,
      "memory(GiB)": 77.56,
      "step": 74415,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 3.1883809605415365,
      "grad_norm": 4.450000762939453,
      "learning_rate": 2.9041882386698616e-05,
      "loss": 2.389927101135254,
      "memory(GiB)": 77.56,
      "step": 74420,
      "token_acc": 0.4602739726027397,
      "train_speed(iter/s)": 1.437662
    },
    {
      "epoch": 3.1885951758707853,
      "grad_norm": 5.67801570892334,
      "learning_rate": 2.9035772539582706e-05,
      "loss": 2.2785919189453123,
      "memory(GiB)": 77.56,
      "step": 74425,
      "token_acc": 0.5124555160142349,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.188809391200034,
      "grad_norm": 7.345423698425293,
      "learning_rate": 2.9029663072253455e-05,
      "loss": 2.541117477416992,
      "memory(GiB)": 77.56,
      "step": 74430,
      "token_acc": 0.450354609929078,
      "train_speed(iter/s)": 1.437686
    },
    {
      "epoch": 3.1890236065292834,
      "grad_norm": 8.45646858215332,
      "learning_rate": 2.9023553984821537e-05,
      "loss": 2.4620317459106444,
      "memory(GiB)": 77.56,
      "step": 74435,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.437684
    },
    {
      "epoch": 3.189237821858532,
      "grad_norm": 6.675907135009766,
      "learning_rate": 2.9017445277397616e-05,
      "loss": 2.440706253051758,
      "memory(GiB)": 77.56,
      "step": 74440,
      "token_acc": 0.49387755102040815,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 3.189452037187781,
      "grad_norm": 6.023128032684326,
      "learning_rate": 2.901133695009235e-05,
      "loss": 2.2662500381469726,
      "memory(GiB)": 77.56,
      "step": 74445,
      "token_acc": 0.5080906148867314,
      "train_speed(iter/s)": 1.437674
    },
    {
      "epoch": 3.1896662525170303,
      "grad_norm": 7.9766082763671875,
      "learning_rate": 2.900522900301641e-05,
      "loss": 2.46661376953125,
      "memory(GiB)": 77.56,
      "step": 74450,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.437662
    },
    {
      "epoch": 3.189880467846279,
      "grad_norm": 6.445828914642334,
      "learning_rate": 2.899912143628042e-05,
      "loss": 2.1338356018066404,
      "memory(GiB)": 77.56,
      "step": 74455,
      "token_acc": 0.5261627906976745,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 3.190094683175528,
      "grad_norm": 6.048820972442627,
      "learning_rate": 2.8993014249995066e-05,
      "loss": 2.4252927780151365,
      "memory(GiB)": 77.56,
      "step": 74460,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 3.190308898504777,
      "grad_norm": 6.362627029418945,
      "learning_rate": 2.898690744427096e-05,
      "loss": 2.4332820892333986,
      "memory(GiB)": 77.56,
      "step": 74465,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437639
    },
    {
      "epoch": 3.190523113834026,
      "grad_norm": 8.85700798034668,
      "learning_rate": 2.8980801019218718e-05,
      "loss": 2.404157257080078,
      "memory(GiB)": 77.56,
      "step": 74470,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 3.190737329163275,
      "grad_norm": 8.288870811462402,
      "learning_rate": 2.8974694974949013e-05,
      "loss": 2.268733024597168,
      "memory(GiB)": 77.56,
      "step": 74475,
      "token_acc": 0.49508196721311476,
      "train_speed(iter/s)": 1.437644
    },
    {
      "epoch": 3.190951544492524,
      "grad_norm": 5.976711750030518,
      "learning_rate": 2.8968589311572426e-05,
      "loss": 2.374024772644043,
      "memory(GiB)": 77.56,
      "step": 74480,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.437647
    },
    {
      "epoch": 3.191165759821773,
      "grad_norm": 9.467687606811523,
      "learning_rate": 2.896248402919955e-05,
      "loss": 2.2588104248046874,
      "memory(GiB)": 77.56,
      "step": 74485,
      "token_acc": 0.4844961240310077,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 3.1913799751510217,
      "grad_norm": 8.094931602478027,
      "learning_rate": 2.895637912794103e-05,
      "loss": 2.4281118392944334,
      "memory(GiB)": 77.56,
      "step": 74490,
      "token_acc": 0.5085910652920962,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 3.191594190480271,
      "grad_norm": 5.958942413330078,
      "learning_rate": 2.8950274607907447e-05,
      "loss": 2.1914178848266603,
      "memory(GiB)": 77.56,
      "step": 74495,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 3.1918084058095197,
      "grad_norm": 6.1112871170043945,
      "learning_rate": 2.8944170469209386e-05,
      "loss": 2.1578176498413084,
      "memory(GiB)": 77.56,
      "step": 74500,
      "token_acc": 0.5242718446601942,
      "train_speed(iter/s)": 1.437623
    },
    {
      "epoch": 3.1918084058095197,
      "eval_loss": 2.2146124839782715,
      "eval_runtime": 14.7283,
      "eval_samples_per_second": 6.79,
      "eval_steps_per_second": 6.79,
      "eval_token_acc": 0.501432664756447,
      "step": 74500
    },
    {
      "epoch": 3.1920226211387686,
      "grad_norm": 6.2797136306762695,
      "learning_rate": 2.8938066711957423e-05,
      "loss": 2.2432228088378907,
      "memory(GiB)": 77.56,
      "step": 74505,
      "token_acc": 0.507085020242915,
      "train_speed(iter/s)": 1.437202
    },
    {
      "epoch": 3.192236836468018,
      "grad_norm": 5.658566951751709,
      "learning_rate": 2.8931963336262125e-05,
      "loss": 2.5845794677734375,
      "memory(GiB)": 77.56,
      "step": 74510,
      "token_acc": 0.4759036144578313,
      "train_speed(iter/s)": 1.437208
    },
    {
      "epoch": 3.1924510517972666,
      "grad_norm": 6.106662750244141,
      "learning_rate": 2.8925860342234092e-05,
      "loss": 2.4850736618041993,
      "memory(GiB)": 77.56,
      "step": 74515,
      "token_acc": 0.46394984326018807,
      "train_speed(iter/s)": 1.437211
    },
    {
      "epoch": 3.1926652671265154,
      "grad_norm": 5.278083801269531,
      "learning_rate": 2.8919757729983866e-05,
      "loss": 2.3318244934082033,
      "memory(GiB)": 77.56,
      "step": 74520,
      "token_acc": 0.5059171597633136,
      "train_speed(iter/s)": 1.437203
    },
    {
      "epoch": 3.1928794824557647,
      "grad_norm": 7.298023700714111,
      "learning_rate": 2.8913655499622012e-05,
      "loss": 2.3609592437744142,
      "memory(GiB)": 77.56,
      "step": 74525,
      "token_acc": 0.5150501672240803,
      "train_speed(iter/s)": 1.437179
    },
    {
      "epoch": 3.1930936977850135,
      "grad_norm": 6.219154357910156,
      "learning_rate": 2.890755365125906e-05,
      "loss": 2.2634323120117186,
      "memory(GiB)": 77.56,
      "step": 74530,
      "token_acc": 0.5291970802919708,
      "train_speed(iter/s)": 1.437175
    },
    {
      "epoch": 3.1933079131142623,
      "grad_norm": 8.251895904541016,
      "learning_rate": 2.890145218500556e-05,
      "loss": 2.311538314819336,
      "memory(GiB)": 77.56,
      "step": 74535,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.437185
    },
    {
      "epoch": 3.1935221284435116,
      "grad_norm": 5.973064422607422,
      "learning_rate": 2.8895351100972033e-05,
      "loss": 2.408467483520508,
      "memory(GiB)": 77.56,
      "step": 74540,
      "token_acc": 0.4982456140350877,
      "train_speed(iter/s)": 1.437165
    },
    {
      "epoch": 3.1937363437727604,
      "grad_norm": 7.326908111572266,
      "learning_rate": 2.888925039926902e-05,
      "loss": 2.0957464218139648,
      "memory(GiB)": 77.56,
      "step": 74545,
      "token_acc": 0.5408560311284046,
      "train_speed(iter/s)": 1.437171
    },
    {
      "epoch": 3.193950559102009,
      "grad_norm": 6.643439769744873,
      "learning_rate": 2.8883150080007047e-05,
      "loss": 2.4536102294921873,
      "memory(GiB)": 77.56,
      "step": 74550,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.437171
    },
    {
      "epoch": 3.1941647744312585,
      "grad_norm": 4.713240623474121,
      "learning_rate": 2.887705014329663e-05,
      "loss": 2.3109451293945313,
      "memory(GiB)": 77.56,
      "step": 74555,
      "token_acc": 0.5114754098360655,
      "train_speed(iter/s)": 1.437158
    },
    {
      "epoch": 3.1943789897605073,
      "grad_norm": 5.468522548675537,
      "learning_rate": 2.8870950589248256e-05,
      "loss": 2.290922927856445,
      "memory(GiB)": 77.56,
      "step": 74560,
      "token_acc": 0.5015015015015015,
      "train_speed(iter/s)": 1.437141
    },
    {
      "epoch": 3.194593205089756,
      "grad_norm": 5.981069087982178,
      "learning_rate": 2.8864851417972444e-05,
      "loss": 2.2255929946899413,
      "memory(GiB)": 77.56,
      "step": 74565,
      "token_acc": 0.5170068027210885,
      "train_speed(iter/s)": 1.437151
    },
    {
      "epoch": 3.1948074204190053,
      "grad_norm": 6.307750701904297,
      "learning_rate": 2.8858752629579648e-05,
      "loss": 2.3497251510620116,
      "memory(GiB)": 77.56,
      "step": 74570,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.437166
    },
    {
      "epoch": 3.195021635748254,
      "grad_norm": 5.7458977699279785,
      "learning_rate": 2.88526542241804e-05,
      "loss": 2.2701597213745117,
      "memory(GiB)": 77.56,
      "step": 74575,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.437177
    },
    {
      "epoch": 3.195235851077503,
      "grad_norm": 5.074203968048096,
      "learning_rate": 2.8846556201885162e-05,
      "loss": 2.4747922897338865,
      "memory(GiB)": 77.56,
      "step": 74580,
      "token_acc": 0.46206896551724136,
      "train_speed(iter/s)": 1.437175
    },
    {
      "epoch": 3.195450066406752,
      "grad_norm": 5.95542049407959,
      "learning_rate": 2.8840458562804396e-05,
      "loss": 2.2604307174682616,
      "memory(GiB)": 77.56,
      "step": 74585,
      "token_acc": 0.5485232067510548,
      "train_speed(iter/s)": 1.437179
    },
    {
      "epoch": 3.195664281736001,
      "grad_norm": 5.192944526672363,
      "learning_rate": 2.883436130704858e-05,
      "loss": 2.258067321777344,
      "memory(GiB)": 77.56,
      "step": 74590,
      "token_acc": 0.5179153094462541,
      "train_speed(iter/s)": 1.43717
    },
    {
      "epoch": 3.19587849706525,
      "grad_norm": 4.612861633300781,
      "learning_rate": 2.882826443472816e-05,
      "loss": 2.4915313720703125,
      "memory(GiB)": 77.56,
      "step": 74595,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.437139
    },
    {
      "epoch": 3.196092712394499,
      "grad_norm": 5.6387619972229,
      "learning_rate": 2.882216794595357e-05,
      "loss": 2.505149269104004,
      "memory(GiB)": 77.56,
      "step": 74600,
      "token_acc": 0.5041551246537396,
      "train_speed(iter/s)": 1.43714
    },
    {
      "epoch": 3.196306927723748,
      "grad_norm": 5.267059803009033,
      "learning_rate": 2.881607184083529e-05,
      "loss": 2.180113983154297,
      "memory(GiB)": 77.56,
      "step": 74605,
      "token_acc": 0.50814332247557,
      "train_speed(iter/s)": 1.437128
    },
    {
      "epoch": 3.1965211430529967,
      "grad_norm": 4.514616966247559,
      "learning_rate": 2.880997611948375e-05,
      "loss": 2.219822120666504,
      "memory(GiB)": 77.56,
      "step": 74610,
      "token_acc": 0.5276752767527675,
      "train_speed(iter/s)": 1.437136
    },
    {
      "epoch": 3.196735358382246,
      "grad_norm": 5.553776741027832,
      "learning_rate": 2.8803880782009347e-05,
      "loss": 2.4151798248291017,
      "memory(GiB)": 77.56,
      "step": 74615,
      "token_acc": 0.4708171206225681,
      "train_speed(iter/s)": 1.437143
    },
    {
      "epoch": 3.196949573711495,
      "grad_norm": 4.378702640533447,
      "learning_rate": 2.8797785828522543e-05,
      "loss": 2.076962471008301,
      "memory(GiB)": 77.56,
      "step": 74620,
      "token_acc": 0.5513307984790875,
      "train_speed(iter/s)": 1.437146
    },
    {
      "epoch": 3.1971637890407436,
      "grad_norm": 4.514094352722168,
      "learning_rate": 2.8791691259133745e-05,
      "loss": 2.4175682067871094,
      "memory(GiB)": 77.56,
      "step": 74625,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.437141
    },
    {
      "epoch": 3.197378004369993,
      "grad_norm": 6.597666263580322,
      "learning_rate": 2.878559707395333e-05,
      "loss": 2.5885677337646484,
      "memory(GiB)": 77.56,
      "step": 74630,
      "token_acc": 0.44785276073619634,
      "train_speed(iter/s)": 1.437155
    },
    {
      "epoch": 3.1975922196992417,
      "grad_norm": 4.421172142028809,
      "learning_rate": 2.8779503273091746e-05,
      "loss": 2.02491455078125,
      "memory(GiB)": 77.56,
      "step": 74635,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.437159
    },
    {
      "epoch": 3.1978064350284905,
      "grad_norm": 5.090878963470459,
      "learning_rate": 2.877340985665936e-05,
      "loss": 2.332314300537109,
      "memory(GiB)": 77.56,
      "step": 74640,
      "token_acc": 0.4983277591973244,
      "train_speed(iter/s)": 1.437172
    },
    {
      "epoch": 3.1980206503577397,
      "grad_norm": 5.172229766845703,
      "learning_rate": 2.8767316824766577e-05,
      "loss": 2.2965587615966796,
      "memory(GiB)": 77.56,
      "step": 74645,
      "token_acc": 0.5114754098360655,
      "train_speed(iter/s)": 1.437157
    },
    {
      "epoch": 3.1982348656869886,
      "grad_norm": 5.844092845916748,
      "learning_rate": 2.876122417752376e-05,
      "loss": 2.5554794311523437,
      "memory(GiB)": 77.56,
      "step": 74650,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437172
    },
    {
      "epoch": 3.1984490810162374,
      "grad_norm": 6.179689407348633,
      "learning_rate": 2.8755131915041277e-05,
      "loss": 2.313243865966797,
      "memory(GiB)": 77.56,
      "step": 74655,
      "token_acc": 0.5213675213675214,
      "train_speed(iter/s)": 1.437165
    },
    {
      "epoch": 3.1986632963454866,
      "grad_norm": 5.895142555236816,
      "learning_rate": 2.8749040037429527e-05,
      "loss": 2.370087814331055,
      "memory(GiB)": 77.56,
      "step": 74660,
      "token_acc": 0.5165562913907285,
      "train_speed(iter/s)": 1.437158
    },
    {
      "epoch": 3.1988775116747354,
      "grad_norm": 5.770688056945801,
      "learning_rate": 2.874294854479885e-05,
      "loss": 2.162943458557129,
      "memory(GiB)": 77.56,
      "step": 74665,
      "token_acc": 0.539622641509434,
      "train_speed(iter/s)": 1.437179
    },
    {
      "epoch": 3.1990917270039843,
      "grad_norm": 5.231703758239746,
      "learning_rate": 2.8736857437259602e-05,
      "loss": 2.577847480773926,
      "memory(GiB)": 77.56,
      "step": 74670,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.437162
    },
    {
      "epoch": 3.1993059423332335,
      "grad_norm": 7.476298809051514,
      "learning_rate": 2.8730766714922126e-05,
      "loss": 2.3307313919067383,
      "memory(GiB)": 77.56,
      "step": 74675,
      "token_acc": 0.4754601226993865,
      "train_speed(iter/s)": 1.437162
    },
    {
      "epoch": 3.1995201576624823,
      "grad_norm": 7.090126991271973,
      "learning_rate": 2.8724676377896764e-05,
      "loss": 2.261647415161133,
      "memory(GiB)": 77.56,
      "step": 74680,
      "token_acc": 0.5160142348754448,
      "train_speed(iter/s)": 1.437156
    },
    {
      "epoch": 3.199734372991731,
      "grad_norm": 6.847160339355469,
      "learning_rate": 2.8718586426293825e-05,
      "loss": 2.4295669555664063,
      "memory(GiB)": 77.56,
      "step": 74685,
      "token_acc": 0.5274725274725275,
      "train_speed(iter/s)": 1.437158
    },
    {
      "epoch": 3.1999485883209804,
      "grad_norm": 6.437007904052734,
      "learning_rate": 2.8712496860223658e-05,
      "loss": 2.369749069213867,
      "memory(GiB)": 77.56,
      "step": 74690,
      "token_acc": 0.4613003095975232,
      "train_speed(iter/s)": 1.43715
    },
    {
      "epoch": 3.200162803650229,
      "grad_norm": 8.268362045288086,
      "learning_rate": 2.8706407679796605e-05,
      "loss": 2.221266174316406,
      "memory(GiB)": 77.56,
      "step": 74695,
      "token_acc": 0.5562700964630225,
      "train_speed(iter/s)": 1.437149
    },
    {
      "epoch": 3.200377018979478,
      "grad_norm": 5.253814697265625,
      "learning_rate": 2.8700318885122944e-05,
      "loss": 2.4045394897460937,
      "memory(GiB)": 77.56,
      "step": 74700,
      "token_acc": 0.5040871934604905,
      "train_speed(iter/s)": 1.437168
    },
    {
      "epoch": 3.2005912343087273,
      "grad_norm": 7.181920051574707,
      "learning_rate": 2.869423047631299e-05,
      "loss": 2.4228822708129885,
      "memory(GiB)": 77.56,
      "step": 74705,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.437155
    },
    {
      "epoch": 3.200805449637976,
      "grad_norm": 4.873536109924316,
      "learning_rate": 2.868814245347704e-05,
      "loss": 2.4592336654663085,
      "memory(GiB)": 77.56,
      "step": 74710,
      "token_acc": 0.5015479876160991,
      "train_speed(iter/s)": 1.437164
    },
    {
      "epoch": 3.201019664967225,
      "grad_norm": 9.333537101745605,
      "learning_rate": 2.8682054816725358e-05,
      "loss": 2.4023914337158203,
      "memory(GiB)": 77.56,
      "step": 74715,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.437144
    },
    {
      "epoch": 3.201233880296474,
      "grad_norm": 6.377510070800781,
      "learning_rate": 2.867596756616827e-05,
      "loss": 2.5542667388916014,
      "memory(GiB)": 77.56,
      "step": 74720,
      "token_acc": 0.5171428571428571,
      "train_speed(iter/s)": 1.437134
    },
    {
      "epoch": 3.201448095625723,
      "grad_norm": 5.612009048461914,
      "learning_rate": 2.866988070191604e-05,
      "loss": 2.253769302368164,
      "memory(GiB)": 77.56,
      "step": 74725,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.437135
    },
    {
      "epoch": 3.201662310954972,
      "grad_norm": 5.75795841217041,
      "learning_rate": 2.8663794224078926e-05,
      "loss": 2.2889842987060547,
      "memory(GiB)": 77.56,
      "step": 74730,
      "token_acc": 0.5131086142322098,
      "train_speed(iter/s)": 1.437146
    },
    {
      "epoch": 3.201876526284221,
      "grad_norm": 4.786227226257324,
      "learning_rate": 2.865770813276719e-05,
      "loss": 2.4804405212402343,
      "memory(GiB)": 77.56,
      "step": 74735,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.437166
    },
    {
      "epoch": 3.20209074161347,
      "grad_norm": 7.432167053222656,
      "learning_rate": 2.86516224280911e-05,
      "loss": 2.599211311340332,
      "memory(GiB)": 77.56,
      "step": 74740,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.437181
    },
    {
      "epoch": 3.2023049569427187,
      "grad_norm": 6.44944953918457,
      "learning_rate": 2.8645537110160873e-05,
      "loss": 2.394886779785156,
      "memory(GiB)": 77.56,
      "step": 74745,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.437203
    },
    {
      "epoch": 3.202519172271968,
      "grad_norm": 5.755970001220703,
      "learning_rate": 2.8639452179086788e-05,
      "loss": 2.422623634338379,
      "memory(GiB)": 77.56,
      "step": 74750,
      "token_acc": 0.4738955823293173,
      "train_speed(iter/s)": 1.437193
    },
    {
      "epoch": 3.2027333876012167,
      "grad_norm": 8.042232513427734,
      "learning_rate": 2.8633367634979075e-05,
      "loss": 2.2414199829101564,
      "memory(GiB)": 77.56,
      "step": 74755,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437226
    },
    {
      "epoch": 3.2029476029304655,
      "grad_norm": 7.1323628425598145,
      "learning_rate": 2.8627283477947942e-05,
      "loss": 2.407971000671387,
      "memory(GiB)": 77.56,
      "step": 74760,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.437237
    },
    {
      "epoch": 3.203161818259715,
      "grad_norm": 6.79801082611084,
      "learning_rate": 2.8621199708103606e-05,
      "loss": 2.571414756774902,
      "memory(GiB)": 77.56,
      "step": 74765,
      "token_acc": 0.43630573248407645,
      "train_speed(iter/s)": 1.437235
    },
    {
      "epoch": 3.2033760335889636,
      "grad_norm": 7.384902477264404,
      "learning_rate": 2.8615116325556307e-05,
      "loss": 2.5754440307617186,
      "memory(GiB)": 77.56,
      "step": 74770,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.437232
    },
    {
      "epoch": 3.2035902489182124,
      "grad_norm": 4.7515764236450195,
      "learning_rate": 2.8609033330416213e-05,
      "loss": 2.2517416000366213,
      "memory(GiB)": 77.56,
      "step": 74775,
      "token_acc": 0.5015673981191222,
      "train_speed(iter/s)": 1.437231
    },
    {
      "epoch": 3.2038044642474617,
      "grad_norm": 6.234803199768066,
      "learning_rate": 2.8602950722793577e-05,
      "loss": 2.364109420776367,
      "memory(GiB)": 77.56,
      "step": 74780,
      "token_acc": 0.5096153846153846,
      "train_speed(iter/s)": 1.437246
    },
    {
      "epoch": 3.2040186795767105,
      "grad_norm": 5.362788200378418,
      "learning_rate": 2.859686850279855e-05,
      "loss": 2.067121696472168,
      "memory(GiB)": 77.56,
      "step": 74785,
      "token_acc": 0.5466666666666666,
      "train_speed(iter/s)": 1.437259
    },
    {
      "epoch": 3.2042328949059593,
      "grad_norm": 6.4607648849487305,
      "learning_rate": 2.8590786670541336e-05,
      "loss": 2.260964775085449,
      "memory(GiB)": 77.56,
      "step": 74790,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.437279
    },
    {
      "epoch": 3.2044471102352086,
      "grad_norm": 8.414200782775879,
      "learning_rate": 2.8584705226132096e-05,
      "loss": 2.3255826950073244,
      "memory(GiB)": 77.56,
      "step": 74795,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.437281
    },
    {
      "epoch": 3.2046613255644574,
      "grad_norm": 6.140413284301758,
      "learning_rate": 2.8578624169680994e-05,
      "loss": 2.3195894241333006,
      "memory(GiB)": 77.56,
      "step": 74800,
      "token_acc": 0.4707692307692308,
      "train_speed(iter/s)": 1.437286
    },
    {
      "epoch": 3.204875540893706,
      "grad_norm": 5.533069133758545,
      "learning_rate": 2.857254350129823e-05,
      "loss": 2.282148551940918,
      "memory(GiB)": 77.56,
      "step": 74805,
      "token_acc": 0.5373665480427047,
      "train_speed(iter/s)": 1.437279
    },
    {
      "epoch": 3.2050897562229554,
      "grad_norm": 4.75316047668457,
      "learning_rate": 2.8566463221093943e-05,
      "loss": 2.393589210510254,
      "memory(GiB)": 77.56,
      "step": 74810,
      "token_acc": 0.49201277955271566,
      "train_speed(iter/s)": 1.437285
    },
    {
      "epoch": 3.2053039715522043,
      "grad_norm": 7.1517333984375,
      "learning_rate": 2.8560383329178286e-05,
      "loss": 2.2573310852050783,
      "memory(GiB)": 77.56,
      "step": 74815,
      "token_acc": 0.5344827586206896,
      "train_speed(iter/s)": 1.437278
    },
    {
      "epoch": 3.205518186881453,
      "grad_norm": 8.25438404083252,
      "learning_rate": 2.8554303825661388e-05,
      "loss": 2.416891670227051,
      "memory(GiB)": 77.56,
      "step": 74820,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.437283
    },
    {
      "epoch": 3.2057324022107023,
      "grad_norm": 6.156642913818359,
      "learning_rate": 2.8548224710653393e-05,
      "loss": 2.2529838562011717,
      "memory(GiB)": 77.56,
      "step": 74825,
      "token_acc": 0.5443548387096774,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 3.205946617539951,
      "grad_norm": 6.6809306144714355,
      "learning_rate": 2.8542145984264416e-05,
      "loss": 2.8260244369506835,
      "memory(GiB)": 77.56,
      "step": 74830,
      "token_acc": 0.43478260869565216,
      "train_speed(iter/s)": 1.437274
    },
    {
      "epoch": 3.2061608328692,
      "grad_norm": 8.493712425231934,
      "learning_rate": 2.8536067646604592e-05,
      "loss": 2.6808374404907225,
      "memory(GiB)": 77.56,
      "step": 74835,
      "token_acc": 0.4332129963898917,
      "train_speed(iter/s)": 1.43728
    },
    {
      "epoch": 3.206375048198449,
      "grad_norm": 6.103187084197998,
      "learning_rate": 2.852998969778406e-05,
      "loss": 2.135850524902344,
      "memory(GiB)": 77.56,
      "step": 74840,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.43727
    },
    {
      "epoch": 3.206589263527698,
      "grad_norm": 6.107016086578369,
      "learning_rate": 2.85239121379129e-05,
      "loss": 2.1684507369995116,
      "memory(GiB)": 77.56,
      "step": 74845,
      "token_acc": 0.5606060606060606,
      "train_speed(iter/s)": 1.437264
    },
    {
      "epoch": 3.206803478856947,
      "grad_norm": 5.628046989440918,
      "learning_rate": 2.851783496710122e-05,
      "loss": 2.1371402740478516,
      "memory(GiB)": 77.56,
      "step": 74850,
      "token_acc": 0.5485074626865671,
      "train_speed(iter/s)": 1.43727
    },
    {
      "epoch": 3.207017694186196,
      "grad_norm": 6.466119766235352,
      "learning_rate": 2.8511758185459113e-05,
      "loss": 2.454837989807129,
      "memory(GiB)": 77.56,
      "step": 74855,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.437278
    },
    {
      "epoch": 3.207231909515445,
      "grad_norm": 5.993646621704102,
      "learning_rate": 2.850568179309665e-05,
      "loss": 2.453556251525879,
      "memory(GiB)": 77.56,
      "step": 74860,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437286
    },
    {
      "epoch": 3.2074461248446937,
      "grad_norm": 5.757293701171875,
      "learning_rate": 2.8499605790123935e-05,
      "loss": 2.4333534240722656,
      "memory(GiB)": 77.56,
      "step": 74865,
      "token_acc": 0.48589341692789967,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 3.207660340173943,
      "grad_norm": 6.234353542327881,
      "learning_rate": 2.8493530176651028e-05,
      "loss": 2.1053205490112306,
      "memory(GiB)": 77.56,
      "step": 74870,
      "token_acc": 0.5177865612648221,
      "train_speed(iter/s)": 1.437316
    },
    {
      "epoch": 3.207874555503192,
      "grad_norm": 4.956779956817627,
      "learning_rate": 2.8487454952788008e-05,
      "loss": 2.4120140075683594,
      "memory(GiB)": 77.56,
      "step": 74875,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 3.2080887708324406,
      "grad_norm": 5.676421165466309,
      "learning_rate": 2.848138011864491e-05,
      "loss": 2.247556686401367,
      "memory(GiB)": 77.56,
      "step": 74880,
      "token_acc": 0.523972602739726,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 3.20830298616169,
      "grad_norm": 7.058403968811035,
      "learning_rate": 2.8475305674331808e-05,
      "loss": 2.232187843322754,
      "memory(GiB)": 77.56,
      "step": 74885,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 3.2085172014909387,
      "grad_norm": 6.966588973999023,
      "learning_rate": 2.8469231619958708e-05,
      "loss": 2.1898117065429688,
      "memory(GiB)": 77.56,
      "step": 74890,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.437328
    },
    {
      "epoch": 3.2087314168201875,
      "grad_norm": 6.0842156410217285,
      "learning_rate": 2.84631579556357e-05,
      "loss": 2.247432518005371,
      "memory(GiB)": 77.56,
      "step": 74895,
      "token_acc": 0.5372670807453416,
      "train_speed(iter/s)": 1.437301
    },
    {
      "epoch": 3.2089456321494367,
      "grad_norm": 6.363500595092773,
      "learning_rate": 2.845708468147278e-05,
      "loss": 2.1914358139038086,
      "memory(GiB)": 77.56,
      "step": 74900,
      "token_acc": 0.5355648535564853,
      "train_speed(iter/s)": 1.437293
    },
    {
      "epoch": 3.2091598474786855,
      "grad_norm": 7.665328025817871,
      "learning_rate": 2.8451011797579985e-05,
      "loss": 2.3460342407226564,
      "memory(GiB)": 77.56,
      "step": 74905,
      "token_acc": 0.5058365758754864,
      "train_speed(iter/s)": 1.437277
    },
    {
      "epoch": 3.2093740628079344,
      "grad_norm": 6.302661895751953,
      "learning_rate": 2.8444939304067307e-05,
      "loss": 2.4604684829711916,
      "memory(GiB)": 77.56,
      "step": 74910,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.437296
    },
    {
      "epoch": 3.2095882781371836,
      "grad_norm": 9.885723114013672,
      "learning_rate": 2.8438867201044794e-05,
      "loss": 2.3899681091308596,
      "memory(GiB)": 77.56,
      "step": 74915,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.437319
    },
    {
      "epoch": 3.2098024934664324,
      "grad_norm": 6.052088737487793,
      "learning_rate": 2.843279548862241e-05,
      "loss": 2.3846935272216796,
      "memory(GiB)": 77.56,
      "step": 74920,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.437344
    },
    {
      "epoch": 3.2100167087956812,
      "grad_norm": 7.404747009277344,
      "learning_rate": 2.842672416691018e-05,
      "loss": 2.2464561462402344,
      "memory(GiB)": 77.56,
      "step": 74925,
      "token_acc": 0.5331010452961672,
      "train_speed(iter/s)": 1.43736
    },
    {
      "epoch": 3.2102309241249305,
      "grad_norm": 6.648402690887451,
      "learning_rate": 2.8420653236018086e-05,
      "loss": 2.375318717956543,
      "memory(GiB)": 77.56,
      "step": 74930,
      "token_acc": 0.5214521452145214,
      "train_speed(iter/s)": 1.43737
    },
    {
      "epoch": 3.2104451394541793,
      "grad_norm": 5.092421054840088,
      "learning_rate": 2.8414582696056102e-05,
      "loss": 2.628529167175293,
      "memory(GiB)": 77.56,
      "step": 74935,
      "token_acc": 0.4707792207792208,
      "train_speed(iter/s)": 1.437363
    },
    {
      "epoch": 3.210659354783428,
      "grad_norm": 5.738795757293701,
      "learning_rate": 2.840851254713419e-05,
      "loss": 2.414845275878906,
      "memory(GiB)": 77.56,
      "step": 74940,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.437377
    },
    {
      "epoch": 3.2108735701126774,
      "grad_norm": 4.775619983673096,
      "learning_rate": 2.8402442789362322e-05,
      "loss": 2.3733440399169923,
      "memory(GiB)": 77.56,
      "step": 74945,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.43739
    },
    {
      "epoch": 3.211087785441926,
      "grad_norm": 5.267678260803223,
      "learning_rate": 2.8396373422850476e-05,
      "loss": 2.30804386138916,
      "memory(GiB)": 77.56,
      "step": 74950,
      "token_acc": 0.5439739413680782,
      "train_speed(iter/s)": 1.437375
    },
    {
      "epoch": 3.211302000771175,
      "grad_norm": 6.17446756362915,
      "learning_rate": 2.8390304447708594e-05,
      "loss": 2.3438674926757814,
      "memory(GiB)": 77.56,
      "step": 74955,
      "token_acc": 0.46835443037974683,
      "train_speed(iter/s)": 1.437363
    },
    {
      "epoch": 3.2115162161004243,
      "grad_norm": 8.948515892028809,
      "learning_rate": 2.8384235864046614e-05,
      "loss": 2.4184532165527344,
      "memory(GiB)": 77.56,
      "step": 74960,
      "token_acc": 0.5122950819672131,
      "train_speed(iter/s)": 1.437391
    },
    {
      "epoch": 3.211730431429673,
      "grad_norm": 5.949466705322266,
      "learning_rate": 2.837816767197448e-05,
      "loss": 2.4688013076782225,
      "memory(GiB)": 77.56,
      "step": 74965,
      "token_acc": 0.43452380952380953,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.211944646758922,
      "grad_norm": 5.721344470977783,
      "learning_rate": 2.8372099871602127e-05,
      "loss": 2.183220863342285,
      "memory(GiB)": 77.56,
      "step": 74970,
      "token_acc": 0.5222929936305732,
      "train_speed(iter/s)": 1.437376
    },
    {
      "epoch": 3.212158862088171,
      "grad_norm": 5.342127799987793,
      "learning_rate": 2.8366032463039448e-05,
      "loss": 2.578084182739258,
      "memory(GiB)": 77.56,
      "step": 74975,
      "token_acc": 0.48942598187311176,
      "train_speed(iter/s)": 1.437363
    },
    {
      "epoch": 3.21237307741742,
      "grad_norm": 6.603104591369629,
      "learning_rate": 2.835996544639641e-05,
      "loss": 2.5666284561157227,
      "memory(GiB)": 77.56,
      "step": 74980,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.2125872927466688,
      "grad_norm": 5.945138454437256,
      "learning_rate": 2.835389882178287e-05,
      "loss": 2.424089050292969,
      "memory(GiB)": 77.56,
      "step": 74985,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.43739
    },
    {
      "epoch": 3.212801508075918,
      "grad_norm": 5.490812301635742,
      "learning_rate": 2.834783258930878e-05,
      "loss": 2.522750663757324,
      "memory(GiB)": 77.56,
      "step": 74990,
      "token_acc": 0.47527472527472525,
      "train_speed(iter/s)": 1.437373
    },
    {
      "epoch": 3.213015723405167,
      "grad_norm": 6.167250633239746,
      "learning_rate": 2.834176674908402e-05,
      "loss": 2.5450586318969726,
      "memory(GiB)": 77.56,
      "step": 74995,
      "token_acc": 0.5268456375838926,
      "train_speed(iter/s)": 1.437396
    },
    {
      "epoch": 3.2132299387344156,
      "grad_norm": 7.590094089508057,
      "learning_rate": 2.8335701301218472e-05,
      "loss": 2.250750732421875,
      "memory(GiB)": 77.56,
      "step": 75000,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 3.2132299387344156,
      "eval_loss": 2.2155067920684814,
      "eval_runtime": 14.8696,
      "eval_samples_per_second": 6.725,
      "eval_steps_per_second": 6.725,
      "eval_token_acc": 0.48567335243553006,
      "step": 75000
    },
    {
      "epoch": 3.213444154063665,
      "grad_norm": 5.956737995147705,
      "learning_rate": 2.8329636245821988e-05,
      "loss": 2.535017395019531,
      "memory(GiB)": 77.56,
      "step": 75005,
      "token_acc": 0.47834645669291337,
      "train_speed(iter/s)": 1.436969
    },
    {
      "epoch": 3.2136583693929137,
      "grad_norm": 5.7900238037109375,
      "learning_rate": 2.8323571583004505e-05,
      "loss": 2.36295223236084,
      "memory(GiB)": 77.56,
      "step": 75010,
      "token_acc": 0.4968553459119497,
      "train_speed(iter/s)": 1.436981
    },
    {
      "epoch": 3.2138725847221625,
      "grad_norm": 5.765345573425293,
      "learning_rate": 2.8317507312875845e-05,
      "loss": 2.2644847869873046,
      "memory(GiB)": 77.56,
      "step": 75015,
      "token_acc": 0.5413533834586466,
      "train_speed(iter/s)": 1.43696
    },
    {
      "epoch": 3.214086800051412,
      "grad_norm": 9.455818176269531,
      "learning_rate": 2.8311443435545892e-05,
      "loss": 2.1789180755615236,
      "memory(GiB)": 77.56,
      "step": 75020,
      "token_acc": 0.5306859205776173,
      "train_speed(iter/s)": 1.436954
    },
    {
      "epoch": 3.2143010153806606,
      "grad_norm": 5.057734966278076,
      "learning_rate": 2.8305379951124478e-05,
      "loss": 2.138809013366699,
      "memory(GiB)": 77.56,
      "step": 75025,
      "token_acc": 0.5304659498207885,
      "train_speed(iter/s)": 1.436954
    },
    {
      "epoch": 3.2145152307099094,
      "grad_norm": 11.719342231750488,
      "learning_rate": 2.8299316859721435e-05,
      "loss": 2.2895336151123047,
      "memory(GiB)": 77.56,
      "step": 75030,
      "token_acc": 0.503968253968254,
      "train_speed(iter/s)": 1.436963
    },
    {
      "epoch": 3.2147294460391587,
      "grad_norm": 4.955362796783447,
      "learning_rate": 2.829325416144665e-05,
      "loss": 2.384970283508301,
      "memory(GiB)": 77.56,
      "step": 75035,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.43699
    },
    {
      "epoch": 3.2149436613684075,
      "grad_norm": 5.784004211425781,
      "learning_rate": 2.8287191856409935e-05,
      "loss": 2.0993122100830077,
      "memory(GiB)": 77.56,
      "step": 75040,
      "token_acc": 0.539568345323741,
      "train_speed(iter/s)": 1.436986
    },
    {
      "epoch": 3.2151578766976563,
      "grad_norm": 6.29775333404541,
      "learning_rate": 2.8281129944721096e-05,
      "loss": 2.2767024993896485,
      "memory(GiB)": 77.56,
      "step": 75045,
      "token_acc": 0.5347222222222222,
      "train_speed(iter/s)": 1.436987
    },
    {
      "epoch": 3.2153720920269055,
      "grad_norm": 10.267982482910156,
      "learning_rate": 2.8275068426489965e-05,
      "loss": 2.212834358215332,
      "memory(GiB)": 77.56,
      "step": 75050,
      "token_acc": 0.5486381322957199,
      "train_speed(iter/s)": 1.43701
    },
    {
      "epoch": 3.2155863073561544,
      "grad_norm": 7.0684332847595215,
      "learning_rate": 2.826900730182635e-05,
      "loss": 2.7192367553710937,
      "memory(GiB)": 77.56,
      "step": 75055,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.437
    },
    {
      "epoch": 3.215800522685403,
      "grad_norm": 4.134343147277832,
      "learning_rate": 2.8262946570840032e-05,
      "loss": 2.333767127990723,
      "memory(GiB)": 77.56,
      "step": 75060,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.437018
    },
    {
      "epoch": 3.2160147380146524,
      "grad_norm": 5.532471179962158,
      "learning_rate": 2.8256886233640834e-05,
      "loss": 2.223998260498047,
      "memory(GiB)": 77.56,
      "step": 75065,
      "token_acc": 0.4970414201183432,
      "train_speed(iter/s)": 1.437054
    },
    {
      "epoch": 3.2162289533439012,
      "grad_norm": 4.45161771774292,
      "learning_rate": 2.825082629033855e-05,
      "loss": 2.2879053115844727,
      "memory(GiB)": 77.56,
      "step": 75070,
      "token_acc": 0.5230769230769231,
      "train_speed(iter/s)": 1.437069
    },
    {
      "epoch": 3.21644316867315,
      "grad_norm": 5.835312843322754,
      "learning_rate": 2.824476674104295e-05,
      "loss": 2.675983428955078,
      "memory(GiB)": 77.56,
      "step": 75075,
      "token_acc": 0.4334470989761092,
      "train_speed(iter/s)": 1.437073
    },
    {
      "epoch": 3.2166573840023993,
      "grad_norm": 9.023283004760742,
      "learning_rate": 2.823870758586381e-05,
      "loss": 2.3446048736572265,
      "memory(GiB)": 77.56,
      "step": 75080,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437087
    },
    {
      "epoch": 3.216871599331648,
      "grad_norm": 5.886406898498535,
      "learning_rate": 2.8232648824910902e-05,
      "loss": 2.3803035736083986,
      "memory(GiB)": 77.56,
      "step": 75085,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437043
    },
    {
      "epoch": 3.217085814660897,
      "grad_norm": 7.387778282165527,
      "learning_rate": 2.8226590458293957e-05,
      "loss": 2.4575361251831054,
      "memory(GiB)": 77.56,
      "step": 75090,
      "token_acc": 0.4541984732824427,
      "train_speed(iter/s)": 1.43706
    },
    {
      "epoch": 3.217300029990146,
      "grad_norm": 5.371993064880371,
      "learning_rate": 2.822053248612277e-05,
      "loss": 2.256745719909668,
      "memory(GiB)": 77.56,
      "step": 75095,
      "token_acc": 0.5444015444015444,
      "train_speed(iter/s)": 1.437059
    },
    {
      "epoch": 3.217514245319395,
      "grad_norm": 6.534891605377197,
      "learning_rate": 2.8214474908507066e-05,
      "loss": 2.4698951721191404,
      "memory(GiB)": 77.56,
      "step": 75100,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.437066
    },
    {
      "epoch": 3.217728460648644,
      "grad_norm": 6.713491916656494,
      "learning_rate": 2.820841772555659e-05,
      "loss": 2.290480041503906,
      "memory(GiB)": 77.56,
      "step": 75105,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.437085
    },
    {
      "epoch": 3.217942675977893,
      "grad_norm": 4.971664905548096,
      "learning_rate": 2.8202360937381066e-05,
      "loss": 2.0735591888427733,
      "memory(GiB)": 77.56,
      "step": 75110,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437073
    },
    {
      "epoch": 3.218156891307142,
      "grad_norm": 6.009228706359863,
      "learning_rate": 2.8196304544090225e-05,
      "loss": 2.1902050018310546,
      "memory(GiB)": 77.56,
      "step": 75115,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.437092
    },
    {
      "epoch": 3.2183711066363907,
      "grad_norm": 4.496587753295898,
      "learning_rate": 2.819024854579375e-05,
      "loss": 2.1751617431640624,
      "memory(GiB)": 77.56,
      "step": 75120,
      "token_acc": 0.5374149659863946,
      "train_speed(iter/s)": 1.437119
    },
    {
      "epoch": 3.21858532196564,
      "grad_norm": 5.55951452255249,
      "learning_rate": 2.8184192942601413e-05,
      "loss": 2.493456268310547,
      "memory(GiB)": 77.56,
      "step": 75125,
      "token_acc": 0.4892086330935252,
      "train_speed(iter/s)": 1.437141
    },
    {
      "epoch": 3.2187995372948888,
      "grad_norm": 6.78969144821167,
      "learning_rate": 2.8178137734622878e-05,
      "loss": 2.2150049209594727,
      "memory(GiB)": 77.56,
      "step": 75130,
      "token_acc": 0.5298013245033113,
      "train_speed(iter/s)": 1.437152
    },
    {
      "epoch": 3.2190137526241376,
      "grad_norm": 6.8576579093933105,
      "learning_rate": 2.8172082921967836e-05,
      "loss": 2.2060529708862306,
      "memory(GiB)": 77.56,
      "step": 75135,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.437164
    },
    {
      "epoch": 3.219227967953387,
      "grad_norm": 8.125765800476074,
      "learning_rate": 2.8166028504746e-05,
      "loss": 2.043033218383789,
      "memory(GiB)": 77.56,
      "step": 75140,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.437175
    },
    {
      "epoch": 3.2194421832826356,
      "grad_norm": 6.171628952026367,
      "learning_rate": 2.8159974483067047e-05,
      "loss": 2.111388397216797,
      "memory(GiB)": 77.56,
      "step": 75145,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.437188
    },
    {
      "epoch": 3.2196563986118845,
      "grad_norm": 6.2173380851745605,
      "learning_rate": 2.815392085704062e-05,
      "loss": 2.421305847167969,
      "memory(GiB)": 77.56,
      "step": 75150,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.437211
    },
    {
      "epoch": 3.2198706139411337,
      "grad_norm": 5.145447731018066,
      "learning_rate": 2.8147867626776437e-05,
      "loss": 2.236465072631836,
      "memory(GiB)": 77.56,
      "step": 75155,
      "token_acc": 0.5288753799392097,
      "train_speed(iter/s)": 1.437207
    },
    {
      "epoch": 3.2200848292703825,
      "grad_norm": 4.8295464515686035,
      "learning_rate": 2.8141814792384123e-05,
      "loss": 2.476268196105957,
      "memory(GiB)": 77.56,
      "step": 75160,
      "token_acc": 0.48546511627906974,
      "train_speed(iter/s)": 1.437178
    },
    {
      "epoch": 3.2202990445996313,
      "grad_norm": 5.567069053649902,
      "learning_rate": 2.813576235397334e-05,
      "loss": 2.286100387573242,
      "memory(GiB)": 77.56,
      "step": 75165,
      "token_acc": 0.5159010600706714,
      "train_speed(iter/s)": 1.43718
    },
    {
      "epoch": 3.2205132599288806,
      "grad_norm": 6.422302722930908,
      "learning_rate": 2.8129710311653735e-05,
      "loss": 2.7097951889038088,
      "memory(GiB)": 77.56,
      "step": 75170,
      "token_acc": 0.42724458204334365,
      "train_speed(iter/s)": 1.437193
    },
    {
      "epoch": 3.2207274752581294,
      "grad_norm": 4.312794208526611,
      "learning_rate": 2.8123658665534913e-05,
      "loss": 2.155950164794922,
      "memory(GiB)": 77.56,
      "step": 75175,
      "token_acc": 0.5409836065573771,
      "train_speed(iter/s)": 1.437198
    },
    {
      "epoch": 3.2209416905873782,
      "grad_norm": 5.756070613861084,
      "learning_rate": 2.8117607415726565e-05,
      "loss": 2.342049789428711,
      "memory(GiB)": 77.56,
      "step": 75180,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.437186
    },
    {
      "epoch": 3.2211559059166275,
      "grad_norm": 9.422969818115234,
      "learning_rate": 2.8111556562338292e-05,
      "loss": 2.0476539611816404,
      "memory(GiB)": 77.56,
      "step": 75185,
      "token_acc": 0.5247148288973384,
      "train_speed(iter/s)": 1.437206
    },
    {
      "epoch": 3.2213701212458763,
      "grad_norm": 6.327995777130127,
      "learning_rate": 2.8105506105479695e-05,
      "loss": 2.227604103088379,
      "memory(GiB)": 77.56,
      "step": 75190,
      "token_acc": 0.531055900621118,
      "train_speed(iter/s)": 1.437183
    },
    {
      "epoch": 3.221584336575125,
      "grad_norm": 6.0252556800842285,
      "learning_rate": 2.8099456045260396e-05,
      "loss": 2.4878944396972655,
      "memory(GiB)": 77.56,
      "step": 75195,
      "token_acc": 0.5313432835820896,
      "train_speed(iter/s)": 1.437192
    },
    {
      "epoch": 3.2217985519043744,
      "grad_norm": 4.432491302490234,
      "learning_rate": 2.809340638179e-05,
      "loss": 2.486052322387695,
      "memory(GiB)": 77.56,
      "step": 75200,
      "token_acc": 0.48703170028818443,
      "train_speed(iter/s)": 1.437192
    },
    {
      "epoch": 3.222012767233623,
      "grad_norm": 6.828469753265381,
      "learning_rate": 2.808735711517807e-05,
      "loss": 2.2316352844238283,
      "memory(GiB)": 77.56,
      "step": 75205,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.43719
    },
    {
      "epoch": 3.222226982562872,
      "grad_norm": 4.746007919311523,
      "learning_rate": 2.8081308245534228e-05,
      "loss": 2.373458480834961,
      "memory(GiB)": 77.56,
      "step": 75210,
      "token_acc": 0.5167173252279635,
      "train_speed(iter/s)": 1.437204
    },
    {
      "epoch": 3.2224411978921212,
      "grad_norm": 5.120621681213379,
      "learning_rate": 2.8075259772968064e-05,
      "loss": 2.481466865539551,
      "memory(GiB)": 77.56,
      "step": 75215,
      "token_acc": 0.47575757575757577,
      "train_speed(iter/s)": 1.437231
    },
    {
      "epoch": 3.22265541322137,
      "grad_norm": 6.66160249710083,
      "learning_rate": 2.8069211697589138e-05,
      "loss": 2.378527069091797,
      "memory(GiB)": 77.56,
      "step": 75220,
      "token_acc": 0.5050505050505051,
      "train_speed(iter/s)": 1.437251
    },
    {
      "epoch": 3.222869628550619,
      "grad_norm": 7.113389492034912,
      "learning_rate": 2.8063164019507017e-05,
      "loss": 2.4076568603515627,
      "memory(GiB)": 77.56,
      "step": 75225,
      "token_acc": 0.5032051282051282,
      "train_speed(iter/s)": 1.437256
    },
    {
      "epoch": 3.223083843879868,
      "grad_norm": 5.610108852386475,
      "learning_rate": 2.805711673883125e-05,
      "loss": 2.467230796813965,
      "memory(GiB)": 77.56,
      "step": 75230,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.437247
    },
    {
      "epoch": 3.223298059209117,
      "grad_norm": 4.766554355621338,
      "learning_rate": 2.8051069855671387e-05,
      "loss": 2.264450454711914,
      "memory(GiB)": 77.56,
      "step": 75235,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 3.2235122745383658,
      "grad_norm": 5.303033828735352,
      "learning_rate": 2.8045023370136996e-05,
      "loss": 2.5186840057373048,
      "memory(GiB)": 77.56,
      "step": 75240,
      "token_acc": 0.5190311418685121,
      "train_speed(iter/s)": 1.437268
    },
    {
      "epoch": 3.223726489867615,
      "grad_norm": 5.408699989318848,
      "learning_rate": 2.8038977282337607e-05,
      "loss": 2.4622920989990233,
      "memory(GiB)": 77.56,
      "step": 75245,
      "token_acc": 0.49363057324840764,
      "train_speed(iter/s)": 1.437283
    },
    {
      "epoch": 3.223940705196864,
      "grad_norm": 6.971130847930908,
      "learning_rate": 2.8032931592382743e-05,
      "loss": 2.4556713104248047,
      "memory(GiB)": 77.56,
      "step": 75250,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.43729
    },
    {
      "epoch": 3.2241549205261126,
      "grad_norm": 6.508571147918701,
      "learning_rate": 2.802688630038193e-05,
      "loss": 2.23946533203125,
      "memory(GiB)": 77.56,
      "step": 75255,
      "token_acc": 0.5377049180327869,
      "train_speed(iter/s)": 1.437303
    },
    {
      "epoch": 3.224369135855362,
      "grad_norm": 4.787049770355225,
      "learning_rate": 2.8020841406444687e-05,
      "loss": 2.3575567245483398,
      "memory(GiB)": 77.56,
      "step": 75260,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437315
    },
    {
      "epoch": 3.2245833511846107,
      "grad_norm": 5.38545560836792,
      "learning_rate": 2.8014796910680498e-05,
      "loss": 2.28045597076416,
      "memory(GiB)": 77.56,
      "step": 75265,
      "token_acc": 0.5251798561151079,
      "train_speed(iter/s)": 1.43731
    },
    {
      "epoch": 3.2247975665138595,
      "grad_norm": 6.3859076499938965,
      "learning_rate": 2.800875281319891e-05,
      "loss": 2.3727489471435548,
      "memory(GiB)": 77.56,
      "step": 75270,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.437299
    },
    {
      "epoch": 3.2250117818431088,
      "grad_norm": 6.496909141540527,
      "learning_rate": 2.8002709114109392e-05,
      "loss": 2.2590238571166994,
      "memory(GiB)": 77.56,
      "step": 75275,
      "token_acc": 0.5354609929078015,
      "train_speed(iter/s)": 1.437334
    },
    {
      "epoch": 3.2252259971723576,
      "grad_norm": 8.583075523376465,
      "learning_rate": 2.799666581352141e-05,
      "loss": 2.1971153259277343,
      "memory(GiB)": 77.56,
      "step": 75280,
      "token_acc": 0.5498154981549815,
      "train_speed(iter/s)": 1.43733
    },
    {
      "epoch": 3.2254402125016064,
      "grad_norm": 5.539919853210449,
      "learning_rate": 2.7990622911544494e-05,
      "loss": 2.2781246185302733,
      "memory(GiB)": 77.56,
      "step": 75285,
      "token_acc": 0.4983922829581994,
      "train_speed(iter/s)": 1.43734
    },
    {
      "epoch": 3.2256544278308557,
      "grad_norm": 5.015482425689697,
      "learning_rate": 2.7984580408288087e-05,
      "loss": 2.0152210235595702,
      "memory(GiB)": 77.56,
      "step": 75290,
      "token_acc": 0.5306859205776173,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 3.2258686431601045,
      "grad_norm": 5.499642848968506,
      "learning_rate": 2.7978538303861635e-05,
      "loss": 2.6526565551757812,
      "memory(GiB)": 77.56,
      "step": 75295,
      "token_acc": 0.45674740484429066,
      "train_speed(iter/s)": 1.437332
    },
    {
      "epoch": 3.2260828584893533,
      "grad_norm": 5.827627658843994,
      "learning_rate": 2.797249659837464e-05,
      "loss": 2.2477169036865234,
      "memory(GiB)": 77.56,
      "step": 75300,
      "token_acc": 0.5062111801242236,
      "train_speed(iter/s)": 1.437283
    },
    {
      "epoch": 3.2262970738186025,
      "grad_norm": 5.2502336502075195,
      "learning_rate": 2.796645529193654e-05,
      "loss": 2.3638214111328124,
      "memory(GiB)": 77.56,
      "step": 75305,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437307
    },
    {
      "epoch": 3.2265112891478513,
      "grad_norm": 6.081926345825195,
      "learning_rate": 2.7960414384656763e-05,
      "loss": 2.5635066986083985,
      "memory(GiB)": 77.56,
      "step": 75310,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.43732
    },
    {
      "epoch": 3.2267255044771,
      "grad_norm": 5.67254114151001,
      "learning_rate": 2.7954373876644756e-05,
      "loss": 2.7994102478027343,
      "memory(GiB)": 77.56,
      "step": 75315,
      "token_acc": 0.449685534591195,
      "train_speed(iter/s)": 1.437291
    },
    {
      "epoch": 3.2269397198063494,
      "grad_norm": 5.713555812835693,
      "learning_rate": 2.7948333768009932e-05,
      "loss": 2.2783538818359377,
      "memory(GiB)": 77.56,
      "step": 75320,
      "token_acc": 0.5171339563862928,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 3.2271539351355982,
      "grad_norm": 4.630133628845215,
      "learning_rate": 2.7942294058861745e-05,
      "loss": 2.0302051544189452,
      "memory(GiB)": 77.56,
      "step": 75325,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.43729
    },
    {
      "epoch": 3.227368150464847,
      "grad_norm": 10.440735816955566,
      "learning_rate": 2.7936254749309594e-05,
      "loss": 2.3708826065063477,
      "memory(GiB)": 77.56,
      "step": 75330,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.437282
    },
    {
      "epoch": 3.2275823657940963,
      "grad_norm": 6.008800506591797,
      "learning_rate": 2.7930215839462882e-05,
      "loss": 2.4622371673583983,
      "memory(GiB)": 77.56,
      "step": 75335,
      "token_acc": 0.4983388704318937,
      "train_speed(iter/s)": 1.437262
    },
    {
      "epoch": 3.227796581123345,
      "grad_norm": 6.473403453826904,
      "learning_rate": 2.792417732943101e-05,
      "loss": 2.14124755859375,
      "memory(GiB)": 77.56,
      "step": 75340,
      "token_acc": 0.5330739299610895,
      "train_speed(iter/s)": 1.437262
    },
    {
      "epoch": 3.228010796452594,
      "grad_norm": 6.101529598236084,
      "learning_rate": 2.7918139219323386e-05,
      "loss": 2.278671455383301,
      "memory(GiB)": 77.56,
      "step": 75345,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.43728
    },
    {
      "epoch": 3.228225011781843,
      "grad_norm": 7.243205547332764,
      "learning_rate": 2.7912101509249357e-05,
      "loss": 2.33863525390625,
      "memory(GiB)": 77.56,
      "step": 75350,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.437288
    },
    {
      "epoch": 3.228439227111092,
      "grad_norm": 5.071996212005615,
      "learning_rate": 2.7906064199318327e-05,
      "loss": 2.5244226455688477,
      "memory(GiB)": 77.56,
      "step": 75355,
      "token_acc": 0.49216300940438873,
      "train_speed(iter/s)": 1.43729
    },
    {
      "epoch": 3.228653442440341,
      "grad_norm": 5.749786853790283,
      "learning_rate": 2.7900027289639695e-05,
      "loss": 2.1004411697387697,
      "memory(GiB)": 77.56,
      "step": 75360,
      "token_acc": 0.5283687943262412,
      "train_speed(iter/s)": 1.437306
    },
    {
      "epoch": 3.22886765776959,
      "grad_norm": 5.683716773986816,
      "learning_rate": 2.7893990780322798e-05,
      "loss": 2.262539100646973,
      "memory(GiB)": 77.56,
      "step": 75365,
      "token_acc": 0.5447154471544715,
      "train_speed(iter/s)": 1.437333
    },
    {
      "epoch": 3.229081873098839,
      "grad_norm": 4.799079895019531,
      "learning_rate": 2.7887954671477002e-05,
      "loss": 2.414706230163574,
      "memory(GiB)": 77.56,
      "step": 75370,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.437313
    },
    {
      "epoch": 3.2292960884280877,
      "grad_norm": 4.655165195465088,
      "learning_rate": 2.7881918963211638e-05,
      "loss": 2.20379638671875,
      "memory(GiB)": 77.56,
      "step": 75375,
      "token_acc": 0.5082508250825083,
      "train_speed(iter/s)": 1.437334
    },
    {
      "epoch": 3.229510303757337,
      "grad_norm": 6.18972635269165,
      "learning_rate": 2.787588365563605e-05,
      "loss": 2.529495620727539,
      "memory(GiB)": 77.56,
      "step": 75380,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.437345
    },
    {
      "epoch": 3.2297245190865858,
      "grad_norm": 6.779815196990967,
      "learning_rate": 2.7869848748859596e-05,
      "loss": 2.612514877319336,
      "memory(GiB)": 77.56,
      "step": 75385,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.437351
    },
    {
      "epoch": 3.2299387344158346,
      "grad_norm": 7.4448065757751465,
      "learning_rate": 2.786381424299159e-05,
      "loss": 2.260715103149414,
      "memory(GiB)": 77.56,
      "step": 75390,
      "token_acc": 0.5755102040816327,
      "train_speed(iter/s)": 1.43735
    },
    {
      "epoch": 3.230152949745084,
      "grad_norm": 6.15609884262085,
      "learning_rate": 2.7857780138141365e-05,
      "loss": 2.491611862182617,
      "memory(GiB)": 77.56,
      "step": 75395,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 3.2303671650743326,
      "grad_norm": 6.680952548980713,
      "learning_rate": 2.785174643441822e-05,
      "loss": 2.2012187957763674,
      "memory(GiB)": 77.56,
      "step": 75400,
      "token_acc": 0.5331125827814569,
      "train_speed(iter/s)": 1.43735
    },
    {
      "epoch": 3.2305813804035814,
      "grad_norm": 5.765521049499512,
      "learning_rate": 2.7845713131931462e-05,
      "loss": 2.6777210235595703,
      "memory(GiB)": 77.56,
      "step": 75405,
      "token_acc": 0.5163398692810458,
      "train_speed(iter/s)": 1.43737
    },
    {
      "epoch": 3.2307955957328307,
      "grad_norm": 7.155279159545898,
      "learning_rate": 2.783968023079038e-05,
      "loss": 1.909425926208496,
      "memory(GiB)": 77.56,
      "step": 75410,
      "token_acc": 0.5464684014869888,
      "train_speed(iter/s)": 1.437389
    },
    {
      "epoch": 3.2310098110620795,
      "grad_norm": 6.343012809753418,
      "learning_rate": 2.7833647731104284e-05,
      "loss": 2.2746932983398436,
      "memory(GiB)": 77.56,
      "step": 75415,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.437408
    },
    {
      "epoch": 3.2312240263913283,
      "grad_norm": 6.070235252380371,
      "learning_rate": 2.7827615632982462e-05,
      "loss": 2.3570247650146485,
      "memory(GiB)": 77.56,
      "step": 75420,
      "token_acc": 0.5233333333333333,
      "train_speed(iter/s)": 1.437413
    },
    {
      "epoch": 3.2314382417205776,
      "grad_norm": 4.9791579246521,
      "learning_rate": 2.782158393653419e-05,
      "loss": 2.5933231353759765,
      "memory(GiB)": 77.56,
      "step": 75425,
      "token_acc": 0.48066298342541436,
      "train_speed(iter/s)": 1.437422
    },
    {
      "epoch": 3.2316524570498264,
      "grad_norm": 4.7238054275512695,
      "learning_rate": 2.7815552641868703e-05,
      "loss": 2.2195587158203125,
      "memory(GiB)": 77.56,
      "step": 75430,
      "token_acc": 0.5447470817120622,
      "train_speed(iter/s)": 1.437443
    },
    {
      "epoch": 3.231866672379075,
      "grad_norm": 5.422077655792236,
      "learning_rate": 2.78095217490953e-05,
      "loss": 2.188327598571777,
      "memory(GiB)": 77.56,
      "step": 75435,
      "token_acc": 0.5130111524163569,
      "train_speed(iter/s)": 1.437459
    },
    {
      "epoch": 3.2320808877083245,
      "grad_norm": 4.599216938018799,
      "learning_rate": 2.7803491258323224e-05,
      "loss": 2.462245559692383,
      "memory(GiB)": 77.56,
      "step": 75440,
      "token_acc": 0.5097493036211699,
      "train_speed(iter/s)": 1.437437
    },
    {
      "epoch": 3.2322951030375733,
      "grad_norm": 6.823783874511719,
      "learning_rate": 2.7797461169661735e-05,
      "loss": 2.2501415252685546,
      "memory(GiB)": 77.56,
      "step": 75445,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.437448
    },
    {
      "epoch": 3.2325093183668225,
      "grad_norm": 5.404219627380371,
      "learning_rate": 2.779143148322007e-05,
      "loss": 2.2669036865234373,
      "memory(GiB)": 77.56,
      "step": 75450,
      "token_acc": 0.5157593123209169,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 3.2327235336960713,
      "grad_norm": 6.246007442474365,
      "learning_rate": 2.7785402199107456e-05,
      "loss": 2.183239555358887,
      "memory(GiB)": 77.56,
      "step": 75455,
      "token_acc": 0.5539033457249071,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.23293774902532,
      "grad_norm": 5.3288254737854,
      "learning_rate": 2.7779373317433122e-05,
      "loss": 2.436257743835449,
      "memory(GiB)": 77.56,
      "step": 75460,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.437439
    },
    {
      "epoch": 3.2331519643545694,
      "grad_norm": 5.885639190673828,
      "learning_rate": 2.777334483830626e-05,
      "loss": 2.19609375,
      "memory(GiB)": 77.56,
      "step": 75465,
      "token_acc": 0.49517684887459806,
      "train_speed(iter/s)": 1.437452
    },
    {
      "epoch": 3.2333661796838182,
      "grad_norm": 5.724212646484375,
      "learning_rate": 2.7767316761836125e-05,
      "loss": 2.309849166870117,
      "memory(GiB)": 77.56,
      "step": 75470,
      "token_acc": 0.49812734082397003,
      "train_speed(iter/s)": 1.437474
    },
    {
      "epoch": 3.233580395013067,
      "grad_norm": 5.691561222076416,
      "learning_rate": 2.7761289088131916e-05,
      "loss": 2.219216728210449,
      "memory(GiB)": 77.56,
      "step": 75475,
      "token_acc": 0.5163398692810458,
      "train_speed(iter/s)": 1.437465
    },
    {
      "epoch": 3.2337946103423163,
      "grad_norm": 8.678397178649902,
      "learning_rate": 2.77552618173028e-05,
      "loss": 2.2646039962768554,
      "memory(GiB)": 77.56,
      "step": 75480,
      "token_acc": 0.4775510204081633,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 3.234008825671565,
      "grad_norm": 8.968337059020996,
      "learning_rate": 2.7749234949457992e-05,
      "loss": 2.227193260192871,
      "memory(GiB)": 77.56,
      "step": 75485,
      "token_acc": 0.5368852459016393,
      "train_speed(iter/s)": 1.437486
    },
    {
      "epoch": 3.234223041000814,
      "grad_norm": 6.9968485832214355,
      "learning_rate": 2.7743208484706667e-05,
      "loss": 2.116176414489746,
      "memory(GiB)": 77.56,
      "step": 75490,
      "token_acc": 0.5372549019607843,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 3.234437256330063,
      "grad_norm": 7.613739013671875,
      "learning_rate": 2.7737182423157982e-05,
      "loss": 2.484148406982422,
      "memory(GiB)": 77.56,
      "step": 75495,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.437493
    },
    {
      "epoch": 3.234651471659312,
      "grad_norm": 7.118156909942627,
      "learning_rate": 2.7731156764921112e-05,
      "loss": 2.338268280029297,
      "memory(GiB)": 77.56,
      "step": 75500,
      "token_acc": 0.5430711610486891,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 3.234651471659312,
      "eval_loss": 2.196960210800171,
      "eval_runtime": 13.6844,
      "eval_samples_per_second": 7.308,
      "eval_steps_per_second": 7.308,
      "eval_token_acc": 0.4728476821192053,
      "step": 75500
    },
    {
      "epoch": 3.234865686988561,
      "grad_norm": 6.853719711303711,
      "learning_rate": 2.772513151010525e-05,
      "loss": 2.3080514907836913,
      "memory(GiB)": 77.56,
      "step": 75505,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437084
    },
    {
      "epoch": 3.23507990231781,
      "grad_norm": 6.332302570343018,
      "learning_rate": 2.7719106658819526e-05,
      "loss": 2.214857292175293,
      "memory(GiB)": 77.56,
      "step": 75510,
      "token_acc": 0.4879518072289157,
      "train_speed(iter/s)": 1.437072
    },
    {
      "epoch": 3.235294117647059,
      "grad_norm": 7.606568336486816,
      "learning_rate": 2.771308221117309e-05,
      "loss": 2.518262481689453,
      "memory(GiB)": 77.56,
      "step": 75515,
      "token_acc": 0.45364238410596025,
      "train_speed(iter/s)": 1.437048
    },
    {
      "epoch": 3.2355083329763077,
      "grad_norm": 6.424154281616211,
      "learning_rate": 2.7707058167275075e-05,
      "loss": 2.283746337890625,
      "memory(GiB)": 77.56,
      "step": 75520,
      "token_acc": 0.5049180327868853,
      "train_speed(iter/s)": 1.437053
    },
    {
      "epoch": 3.235722548305557,
      "grad_norm": 5.926879405975342,
      "learning_rate": 2.770103452723459e-05,
      "loss": 2.4940719604492188,
      "memory(GiB)": 77.56,
      "step": 75525,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437038
    },
    {
      "epoch": 3.2359367636348058,
      "grad_norm": 5.421752452850342,
      "learning_rate": 2.769501129116081e-05,
      "loss": 2.3071590423583985,
      "memory(GiB)": 77.56,
      "step": 75530,
      "token_acc": 0.5288135593220339,
      "train_speed(iter/s)": 1.437029
    },
    {
      "epoch": 3.2361509789640546,
      "grad_norm": 6.382996082305908,
      "learning_rate": 2.7688988459162818e-05,
      "loss": 2.2297019958496094,
      "memory(GiB)": 77.56,
      "step": 75535,
      "token_acc": 0.4752851711026616,
      "train_speed(iter/s)": 1.437038
    },
    {
      "epoch": 3.236365194293304,
      "grad_norm": 5.532587051391602,
      "learning_rate": 2.7682966031349732e-05,
      "loss": 2.2956300735473634,
      "memory(GiB)": 77.56,
      "step": 75540,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.437035
    },
    {
      "epoch": 3.2365794096225526,
      "grad_norm": 10.746757507324219,
      "learning_rate": 2.7676944007830653e-05,
      "loss": 2.3829824447631838,
      "memory(GiB)": 77.56,
      "step": 75545,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.437035
    },
    {
      "epoch": 3.2367936249518015,
      "grad_norm": 4.990373134613037,
      "learning_rate": 2.767092238871467e-05,
      "loss": 2.4315380096435546,
      "memory(GiB)": 77.56,
      "step": 75550,
      "token_acc": 0.4766355140186916,
      "train_speed(iter/s)": 1.437021
    },
    {
      "epoch": 3.2370078402810507,
      "grad_norm": 5.94728422164917,
      "learning_rate": 2.7664901174110857e-05,
      "loss": 2.0546823501586915,
      "memory(GiB)": 77.56,
      "step": 75555,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.437036
    },
    {
      "epoch": 3.2372220556102995,
      "grad_norm": 6.0204362869262695,
      "learning_rate": 2.7658880364128326e-05,
      "loss": 2.3509057998657226,
      "memory(GiB)": 77.56,
      "step": 75560,
      "token_acc": 0.5015197568389058,
      "train_speed(iter/s)": 1.43703
    },
    {
      "epoch": 3.2374362709395483,
      "grad_norm": 5.720105171203613,
      "learning_rate": 2.7652859958876142e-05,
      "loss": 1.9346551895141602,
      "memory(GiB)": 77.56,
      "step": 75565,
      "token_acc": 0.5679442508710801,
      "train_speed(iter/s)": 1.437059
    },
    {
      "epoch": 3.2376504862687976,
      "grad_norm": 6.184651851654053,
      "learning_rate": 2.764683995846336e-05,
      "loss": 2.212659454345703,
      "memory(GiB)": 77.56,
      "step": 75570,
      "token_acc": 0.5273972602739726,
      "train_speed(iter/s)": 1.437073
    },
    {
      "epoch": 3.2378647015980464,
      "grad_norm": 5.279417037963867,
      "learning_rate": 2.764082036299902e-05,
      "loss": 2.3793272018432616,
      "memory(GiB)": 77.56,
      "step": 75575,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.437099
    },
    {
      "epoch": 3.238078916927295,
      "grad_norm": 6.2321648597717285,
      "learning_rate": 2.7634801172592217e-05,
      "loss": 2.485188293457031,
      "memory(GiB)": 77.56,
      "step": 75580,
      "token_acc": 0.4859154929577465,
      "train_speed(iter/s)": 1.437114
    },
    {
      "epoch": 3.2382931322565445,
      "grad_norm": 7.952552318572998,
      "learning_rate": 2.7628782387351948e-05,
      "loss": 2.165928268432617,
      "memory(GiB)": 77.56,
      "step": 75585,
      "token_acc": 0.5049833887043189,
      "train_speed(iter/s)": 1.437128
    },
    {
      "epoch": 3.2385073475857933,
      "grad_norm": 6.131654739379883,
      "learning_rate": 2.762276400738729e-05,
      "loss": 2.2802419662475586,
      "memory(GiB)": 77.56,
      "step": 75590,
      "token_acc": 0.49137931034482757,
      "train_speed(iter/s)": 1.437122
    },
    {
      "epoch": 3.238721562915042,
      "grad_norm": 6.049211025238037,
      "learning_rate": 2.7616746032807262e-05,
      "loss": 2.479566955566406,
      "memory(GiB)": 77.56,
      "step": 75595,
      "token_acc": 0.5175718849840255,
      "train_speed(iter/s)": 1.437131
    },
    {
      "epoch": 3.2389357782442914,
      "grad_norm": 7.114353656768799,
      "learning_rate": 2.761072846372087e-05,
      "loss": 2.1769914627075195,
      "memory(GiB)": 77.56,
      "step": 75600,
      "token_acc": 0.525691699604743,
      "train_speed(iter/s)": 1.437146
    },
    {
      "epoch": 3.23914999357354,
      "grad_norm": 4.610339641571045,
      "learning_rate": 2.760471130023714e-05,
      "loss": 2.1934295654296876,
      "memory(GiB)": 77.56,
      "step": 75605,
      "token_acc": 0.5233918128654971,
      "train_speed(iter/s)": 1.437147
    },
    {
      "epoch": 3.239364208902789,
      "grad_norm": 5.4529266357421875,
      "learning_rate": 2.7598694542465047e-05,
      "loss": 2.193354034423828,
      "memory(GiB)": 77.56,
      "step": 75610,
      "token_acc": 0.5176848874598071,
      "train_speed(iter/s)": 1.437152
    },
    {
      "epoch": 3.2395784242320382,
      "grad_norm": 4.35702657699585,
      "learning_rate": 2.7592678190513638e-05,
      "loss": 2.3082372665405275,
      "memory(GiB)": 77.56,
      "step": 75615,
      "token_acc": 0.49544072948328266,
      "train_speed(iter/s)": 1.437168
    },
    {
      "epoch": 3.239792639561287,
      "grad_norm": 7.161694526672363,
      "learning_rate": 2.7586662244491883e-05,
      "loss": 2.4381765365600585,
      "memory(GiB)": 77.56,
      "step": 75620,
      "token_acc": 0.48659003831417624,
      "train_speed(iter/s)": 1.437162
    },
    {
      "epoch": 3.240006854890536,
      "grad_norm": 5.5795087814331055,
      "learning_rate": 2.7580646704508772e-05,
      "loss": 2.253228759765625,
      "memory(GiB)": 77.56,
      "step": 75625,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.437175
    },
    {
      "epoch": 3.240221070219785,
      "grad_norm": 6.340299606323242,
      "learning_rate": 2.757463157067327e-05,
      "loss": 2.311410140991211,
      "memory(GiB)": 77.56,
      "step": 75630,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.437198
    },
    {
      "epoch": 3.240435285549034,
      "grad_norm": 7.961738109588623,
      "learning_rate": 2.756861684309436e-05,
      "loss": 2.2823360443115233,
      "memory(GiB)": 77.56,
      "step": 75635,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.437226
    },
    {
      "epoch": 3.2406495008782827,
      "grad_norm": 5.117207050323486,
      "learning_rate": 2.7562602521880975e-05,
      "loss": 2.46173038482666,
      "memory(GiB)": 77.56,
      "step": 75640,
      "token_acc": 0.5335570469798657,
      "train_speed(iter/s)": 1.437231
    },
    {
      "epoch": 3.240863716207532,
      "grad_norm": 5.581437587738037,
      "learning_rate": 2.7556588607142107e-05,
      "loss": 2.245204734802246,
      "memory(GiB)": 77.56,
      "step": 75645,
      "token_acc": 0.5399239543726235,
      "train_speed(iter/s)": 1.437254
    },
    {
      "epoch": 3.241077931536781,
      "grad_norm": 4.618085861206055,
      "learning_rate": 2.7550575098986688e-05,
      "loss": 2.319418716430664,
      "memory(GiB)": 77.56,
      "step": 75650,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.437223
    },
    {
      "epoch": 3.2412921468660296,
      "grad_norm": 5.7654924392700195,
      "learning_rate": 2.7544561997523665e-05,
      "loss": 2.308021354675293,
      "memory(GiB)": 77.56,
      "step": 75655,
      "token_acc": 0.5389830508474577,
      "train_speed(iter/s)": 1.437212
    },
    {
      "epoch": 3.241506362195279,
      "grad_norm": 6.728402137756348,
      "learning_rate": 2.753854930286197e-05,
      "loss": 2.2943843841552733,
      "memory(GiB)": 77.56,
      "step": 75660,
      "token_acc": 0.4885057471264368,
      "train_speed(iter/s)": 1.437206
    },
    {
      "epoch": 3.2417205775245277,
      "grad_norm": 4.96250057220459,
      "learning_rate": 2.7532537015110528e-05,
      "loss": 2.046406555175781,
      "memory(GiB)": 77.56,
      "step": 75665,
      "token_acc": 0.5176470588235295,
      "train_speed(iter/s)": 1.437225
    },
    {
      "epoch": 3.2419347928537765,
      "grad_norm": 6.787867069244385,
      "learning_rate": 2.752652513437823e-05,
      "loss": 2.4859390258789062,
      "memory(GiB)": 77.56,
      "step": 75670,
      "token_acc": 0.4734848484848485,
      "train_speed(iter/s)": 1.437244
    },
    {
      "epoch": 3.2421490081830258,
      "grad_norm": 6.609344959259033,
      "learning_rate": 2.7520513660774028e-05,
      "loss": 2.20709171295166,
      "memory(GiB)": 77.56,
      "step": 75675,
      "token_acc": 0.5398550724637681,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.2423632235122746,
      "grad_norm": 4.82760763168335,
      "learning_rate": 2.7514502594406812e-05,
      "loss": 2.2754920959472655,
      "memory(GiB)": 77.56,
      "step": 75680,
      "token_acc": 0.5175438596491229,
      "train_speed(iter/s)": 1.437196
    },
    {
      "epoch": 3.2425774388415234,
      "grad_norm": 4.4051079750061035,
      "learning_rate": 2.7508491935385478e-05,
      "loss": 2.355311965942383,
      "memory(GiB)": 77.56,
      "step": 75685,
      "token_acc": 0.5234375,
      "train_speed(iter/s)": 1.437217
    },
    {
      "epoch": 3.2427916541707726,
      "grad_norm": 5.311786651611328,
      "learning_rate": 2.750248168381891e-05,
      "loss": 2.087117385864258,
      "memory(GiB)": 77.56,
      "step": 75690,
      "token_acc": 0.5335820895522388,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.2430058695000215,
      "grad_norm": 5.390645503997803,
      "learning_rate": 2.7496471839815983e-05,
      "loss": 2.2624372482299804,
      "memory(GiB)": 77.56,
      "step": 75695,
      "token_acc": 0.5742574257425742,
      "train_speed(iter/s)": 1.437221
    },
    {
      "epoch": 3.2432200848292703,
      "grad_norm": 7.065344333648682,
      "learning_rate": 2.7490462403485574e-05,
      "loss": 2.1746971130371096,
      "memory(GiB)": 77.56,
      "step": 75700,
      "token_acc": 0.4979757085020243,
      "train_speed(iter/s)": 1.437229
    },
    {
      "epoch": 3.2434343001585195,
      "grad_norm": 4.980740547180176,
      "learning_rate": 2.7484453374936558e-05,
      "loss": 2.2688079833984376,
      "memory(GiB)": 77.56,
      "step": 75705,
      "token_acc": 0.5047923322683706,
      "train_speed(iter/s)": 1.437253
    },
    {
      "epoch": 3.2436485154877683,
      "grad_norm": 5.923549175262451,
      "learning_rate": 2.74784447542778e-05,
      "loss": 2.4600738525390624,
      "memory(GiB)": 77.56,
      "step": 75710,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.437273
    },
    {
      "epoch": 3.243862730817017,
      "grad_norm": 8.077080726623535,
      "learning_rate": 2.7472436541618142e-05,
      "loss": 2.2293310165405273,
      "memory(GiB)": 77.56,
      "step": 75715,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.437241
    },
    {
      "epoch": 3.2440769461462664,
      "grad_norm": 8.59260368347168,
      "learning_rate": 2.7466428737066426e-05,
      "loss": 2.2343297958374024,
      "memory(GiB)": 77.56,
      "step": 75720,
      "token_acc": 0.5598591549295775,
      "train_speed(iter/s)": 1.437252
    },
    {
      "epoch": 3.244291161475515,
      "grad_norm": 4.704460620880127,
      "learning_rate": 2.7460421340731468e-05,
      "loss": 2.1195789337158204,
      "memory(GiB)": 77.56,
      "step": 75725,
      "token_acc": 0.516728624535316,
      "train_speed(iter/s)": 1.437269
    },
    {
      "epoch": 3.244505376804764,
      "grad_norm": 5.434798717498779,
      "learning_rate": 2.7454414352722128e-05,
      "loss": 2.295984649658203,
      "memory(GiB)": 77.56,
      "step": 75730,
      "token_acc": 0.48877805486284287,
      "train_speed(iter/s)": 1.437281
    },
    {
      "epoch": 3.2447195921340133,
      "grad_norm": 5.225401401519775,
      "learning_rate": 2.7448407773147238e-05,
      "loss": 2.537242126464844,
      "memory(GiB)": 77.56,
      "step": 75735,
      "token_acc": 0.48773006134969327,
      "train_speed(iter/s)": 1.437301
    },
    {
      "epoch": 3.244933807463262,
      "grad_norm": 7.597643852233887,
      "learning_rate": 2.7442401602115596e-05,
      "loss": 2.012503242492676,
      "memory(GiB)": 77.56,
      "step": 75740,
      "token_acc": 0.5447761194029851,
      "train_speed(iter/s)": 1.43732
    },
    {
      "epoch": 3.245148022792511,
      "grad_norm": 5.200222969055176,
      "learning_rate": 2.7436395839736016e-05,
      "loss": 2.2515895843505858,
      "memory(GiB)": 77.56,
      "step": 75745,
      "token_acc": 0.5590062111801242,
      "train_speed(iter/s)": 1.437342
    },
    {
      "epoch": 3.24536223812176,
      "grad_norm": 5.690224647521973,
      "learning_rate": 2.7430390486117284e-05,
      "loss": 2.3349048614501955,
      "memory(GiB)": 77.56,
      "step": 75750,
      "token_acc": 0.5292207792207793,
      "train_speed(iter/s)": 1.437334
    },
    {
      "epoch": 3.245576453451009,
      "grad_norm": 5.858898162841797,
      "learning_rate": 2.742438554136818e-05,
      "loss": 2.2467977523803713,
      "memory(GiB)": 77.56,
      "step": 75755,
      "token_acc": 0.5196850393700787,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 3.245790668780258,
      "grad_norm": 5.439920425415039,
      "learning_rate": 2.741838100559754e-05,
      "loss": 2.363109588623047,
      "memory(GiB)": 77.56,
      "step": 75760,
      "token_acc": 0.5547945205479452,
      "train_speed(iter/s)": 1.437362
    },
    {
      "epoch": 3.246004884109507,
      "grad_norm": 8.374643325805664,
      "learning_rate": 2.7412376878914105e-05,
      "loss": 2.812795639038086,
      "memory(GiB)": 77.56,
      "step": 75765,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.437364
    },
    {
      "epoch": 3.246219099438756,
      "grad_norm": 5.497799396514893,
      "learning_rate": 2.7406373161426647e-05,
      "loss": 2.4468584060668945,
      "memory(GiB)": 77.56,
      "step": 75770,
      "token_acc": 0.4769736842105263,
      "train_speed(iter/s)": 1.437373
    },
    {
      "epoch": 3.2464333147680047,
      "grad_norm": 5.403831481933594,
      "learning_rate": 2.7400369853243936e-05,
      "loss": 2.315408706665039,
      "memory(GiB)": 77.56,
      "step": 75775,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437374
    },
    {
      "epoch": 3.246647530097254,
      "grad_norm": 4.9260759353637695,
      "learning_rate": 2.7394366954474726e-05,
      "loss": 2.4794679641723634,
      "memory(GiB)": 77.56,
      "step": 75780,
      "token_acc": 0.50920245398773,
      "train_speed(iter/s)": 1.437379
    },
    {
      "epoch": 3.2468617454265027,
      "grad_norm": 6.249225616455078,
      "learning_rate": 2.7388364465227746e-05,
      "loss": 2.5583587646484376,
      "memory(GiB)": 77.56,
      "step": 75785,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.437382
    },
    {
      "epoch": 3.2470759607557516,
      "grad_norm": 6.810762405395508,
      "learning_rate": 2.738236238561177e-05,
      "loss": 2.592457580566406,
      "memory(GiB)": 77.56,
      "step": 75790,
      "token_acc": 0.49846153846153846,
      "train_speed(iter/s)": 1.437378
    },
    {
      "epoch": 3.247290176085001,
      "grad_norm": 4.638576030731201,
      "learning_rate": 2.7376360715735527e-05,
      "loss": 2.196521759033203,
      "memory(GiB)": 77.56,
      "step": 75795,
      "token_acc": 0.5226480836236934,
      "train_speed(iter/s)": 1.437401
    },
    {
      "epoch": 3.2475043914142496,
      "grad_norm": 4.7867350578308105,
      "learning_rate": 2.7370359455707706e-05,
      "loss": 2.3255542755126952,
      "memory(GiB)": 77.56,
      "step": 75800,
      "token_acc": 0.5258358662613982,
      "train_speed(iter/s)": 1.437395
    },
    {
      "epoch": 3.2477186067434984,
      "grad_norm": 7.031664848327637,
      "learning_rate": 2.736435860563707e-05,
      "loss": 2.326621437072754,
      "memory(GiB)": 77.56,
      "step": 75805,
      "token_acc": 0.5344827586206896,
      "train_speed(iter/s)": 1.437382
    },
    {
      "epoch": 3.2479328220727477,
      "grad_norm": 4.80388879776001,
      "learning_rate": 2.7358358165632313e-05,
      "loss": 2.1625078201293944,
      "memory(GiB)": 77.56,
      "step": 75810,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.437385
    },
    {
      "epoch": 3.2481470374019965,
      "grad_norm": 5.1422319412231445,
      "learning_rate": 2.7352358135802113e-05,
      "loss": 2.579312324523926,
      "memory(GiB)": 77.56,
      "step": 75815,
      "token_acc": 0.4379746835443038,
      "train_speed(iter/s)": 1.43739
    },
    {
      "epoch": 3.2483612527312453,
      "grad_norm": 4.782476902008057,
      "learning_rate": 2.7346358516255222e-05,
      "loss": 2.363435745239258,
      "memory(GiB)": 77.56,
      "step": 75820,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437399
    },
    {
      "epoch": 3.2485754680604946,
      "grad_norm": 5.166233539581299,
      "learning_rate": 2.7340359307100287e-05,
      "loss": 2.384761428833008,
      "memory(GiB)": 77.56,
      "step": 75825,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.43742
    },
    {
      "epoch": 3.2487896833897434,
      "grad_norm": 7.33074951171875,
      "learning_rate": 2.7334360508446005e-05,
      "loss": 2.3004081726074217,
      "memory(GiB)": 77.56,
      "step": 75830,
      "token_acc": 0.4881889763779528,
      "train_speed(iter/s)": 1.437441
    },
    {
      "epoch": 3.249003898718992,
      "grad_norm": 4.672936916351318,
      "learning_rate": 2.732836212040104e-05,
      "loss": 2.4752323150634767,
      "memory(GiB)": 77.56,
      "step": 75835,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.437456
    },
    {
      "epoch": 3.2492181140482415,
      "grad_norm": 5.176677703857422,
      "learning_rate": 2.732236414307407e-05,
      "loss": 2.136547088623047,
      "memory(GiB)": 77.56,
      "step": 75840,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 3.2494323293774903,
      "grad_norm": 7.57281494140625,
      "learning_rate": 2.731636657657372e-05,
      "loss": 2.323530578613281,
      "memory(GiB)": 77.56,
      "step": 75845,
      "token_acc": 0.5249266862170088,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 3.249646544706739,
      "grad_norm": 4.811956405639648,
      "learning_rate": 2.731036942100869e-05,
      "loss": 2.4900440216064452,
      "memory(GiB)": 77.56,
      "step": 75850,
      "token_acc": 0.48502994011976047,
      "train_speed(iter/s)": 1.437456
    },
    {
      "epoch": 3.2498607600359883,
      "grad_norm": 4.795746803283691,
      "learning_rate": 2.7304372676487606e-05,
      "loss": 2.3269702911376955,
      "memory(GiB)": 77.56,
      "step": 75855,
      "token_acc": 0.5105105105105106,
      "train_speed(iter/s)": 1.437478
    },
    {
      "epoch": 3.250074975365237,
      "grad_norm": 6.632368564605713,
      "learning_rate": 2.72983763431191e-05,
      "loss": 2.3196826934814454,
      "memory(GiB)": 77.56,
      "step": 75860,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.437507
    },
    {
      "epoch": 3.250289190694486,
      "grad_norm": 6.483335018157959,
      "learning_rate": 2.72923804210118e-05,
      "loss": 2.1050172805786134,
      "memory(GiB)": 77.56,
      "step": 75865,
      "token_acc": 0.4979757085020243,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 3.250503406023735,
      "grad_norm": 5.677504062652588,
      "learning_rate": 2.728638491027431e-05,
      "loss": 2.305510330200195,
      "memory(GiB)": 77.56,
      "step": 75870,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 3.250717621352984,
      "grad_norm": 5.297688961029053,
      "learning_rate": 2.7280389811015267e-05,
      "loss": 2.270792770385742,
      "memory(GiB)": 77.56,
      "step": 75875,
      "token_acc": 0.4613003095975232,
      "train_speed(iter/s)": 1.437535
    },
    {
      "epoch": 3.250931836682233,
      "grad_norm": 5.259054660797119,
      "learning_rate": 2.7274395123343295e-05,
      "loss": 2.373600196838379,
      "memory(GiB)": 77.56,
      "step": 75880,
      "token_acc": 0.4889502762430939,
      "train_speed(iter/s)": 1.437534
    },
    {
      "epoch": 3.251146052011482,
      "grad_norm": 6.0703606605529785,
      "learning_rate": 2.726840084736697e-05,
      "loss": 2.506459426879883,
      "memory(GiB)": 77.56,
      "step": 75885,
      "token_acc": 0.5221843003412969,
      "train_speed(iter/s)": 1.437543
    },
    {
      "epoch": 3.251360267340731,
      "grad_norm": 6.6016764640808105,
      "learning_rate": 2.7262406983194894e-05,
      "loss": 2.2500278472900392,
      "memory(GiB)": 77.56,
      "step": 75890,
      "token_acc": 0.493006993006993,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 3.2515744826699797,
      "grad_norm": 4.968935966491699,
      "learning_rate": 2.725641353093564e-05,
      "loss": 2.325093460083008,
      "memory(GiB)": 77.56,
      "step": 75895,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.437541
    },
    {
      "epoch": 3.251788697999229,
      "grad_norm": 5.574542999267578,
      "learning_rate": 2.725042049069777e-05,
      "loss": 2.5173385620117186,
      "memory(GiB)": 77.56,
      "step": 75900,
      "token_acc": 0.46441947565543074,
      "train_speed(iter/s)": 1.437538
    },
    {
      "epoch": 3.252002913328478,
      "grad_norm": 4.985105991363525,
      "learning_rate": 2.7244427862589893e-05,
      "loss": 2.4265193939208984,
      "memory(GiB)": 77.56,
      "step": 75905,
      "token_acc": 0.5014005602240896,
      "train_speed(iter/s)": 1.437542
    },
    {
      "epoch": 3.2522171286577266,
      "grad_norm": 6.420775890350342,
      "learning_rate": 2.723843564672056e-05,
      "loss": 2.3762529373168944,
      "memory(GiB)": 77.56,
      "step": 75910,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.437541
    },
    {
      "epoch": 3.252431343986976,
      "grad_norm": 6.197560787200928,
      "learning_rate": 2.723244384319831e-05,
      "loss": 2.281586837768555,
      "memory(GiB)": 77.56,
      "step": 75915,
      "token_acc": 0.5266666666666666,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 3.2526455593162247,
      "grad_norm": 6.315161228179932,
      "learning_rate": 2.7226452452131703e-05,
      "loss": 2.466244125366211,
      "memory(GiB)": 77.56,
      "step": 75920,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.437556
    },
    {
      "epoch": 3.2528597746454735,
      "grad_norm": 6.310101509094238,
      "learning_rate": 2.7220461473629265e-05,
      "loss": 2.1732805252075194,
      "memory(GiB)": 77.56,
      "step": 75925,
      "token_acc": 0.5303643724696356,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 3.2530739899747227,
      "grad_norm": 5.555732727050781,
      "learning_rate": 2.7214470907799517e-05,
      "loss": 2.0638877868652346,
      "memory(GiB)": 77.56,
      "step": 75930,
      "token_acc": 0.5451127819548872,
      "train_speed(iter/s)": 1.437594
    },
    {
      "epoch": 3.2532882053039716,
      "grad_norm": 5.987128734588623,
      "learning_rate": 2.7208480754751033e-05,
      "loss": 2.262569808959961,
      "memory(GiB)": 77.56,
      "step": 75935,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.437573
    },
    {
      "epoch": 3.2535024206332204,
      "grad_norm": 5.220691680908203,
      "learning_rate": 2.7202491014592292e-05,
      "loss": 2.179315948486328,
      "memory(GiB)": 77.56,
      "step": 75940,
      "token_acc": 0.4819277108433735,
      "train_speed(iter/s)": 1.437549
    },
    {
      "epoch": 3.2537166359624696,
      "grad_norm": 6.325334072113037,
      "learning_rate": 2.719650168743179e-05,
      "loss": 2.200699234008789,
      "memory(GiB)": 77.56,
      "step": 75945,
      "token_acc": 0.5220588235294118,
      "train_speed(iter/s)": 1.43757
    },
    {
      "epoch": 3.2539308512917184,
      "grad_norm": 6.337735176086426,
      "learning_rate": 2.7190512773378075e-05,
      "loss": 2.4285959243774413,
      "memory(GiB)": 77.56,
      "step": 75950,
      "token_acc": 0.47262247838616717,
      "train_speed(iter/s)": 1.437592
    },
    {
      "epoch": 3.2541450666209673,
      "grad_norm": 4.491027355194092,
      "learning_rate": 2.718452427253962e-05,
      "loss": 2.283153533935547,
      "memory(GiB)": 77.56,
      "step": 75955,
      "token_acc": 0.5381818181818182,
      "train_speed(iter/s)": 1.437594
    },
    {
      "epoch": 3.2543592819502165,
      "grad_norm": 8.201942443847656,
      "learning_rate": 2.717853618502489e-05,
      "loss": 2.0722557067871095,
      "memory(GiB)": 77.56,
      "step": 75960,
      "token_acc": 0.5352112676056338,
      "train_speed(iter/s)": 1.437611
    },
    {
      "epoch": 3.2545734972794653,
      "grad_norm": 8.467689514160156,
      "learning_rate": 2.7172548510942402e-05,
      "loss": 2.3165279388427735,
      "memory(GiB)": 77.56,
      "step": 75965,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 3.254787712608714,
      "grad_norm": 5.518149375915527,
      "learning_rate": 2.7166561250400613e-05,
      "loss": 2.3943042755126953,
      "memory(GiB)": 77.56,
      "step": 75970,
      "token_acc": 0.4942084942084942,
      "train_speed(iter/s)": 1.437632
    },
    {
      "epoch": 3.2550019279379634,
      "grad_norm": 10.01694393157959,
      "learning_rate": 2.7160574403507988e-05,
      "loss": 2.4342227935791017,
      "memory(GiB)": 77.56,
      "step": 75975,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.437638
    },
    {
      "epoch": 3.255216143267212,
      "grad_norm": 6.17288064956665,
      "learning_rate": 2.715458797037298e-05,
      "loss": 2.2419620513916017,
      "memory(GiB)": 77.56,
      "step": 75980,
      "token_acc": 0.49244712990936557,
      "train_speed(iter/s)": 1.437655
    },
    {
      "epoch": 3.255430358596461,
      "grad_norm": 5.122513294219971,
      "learning_rate": 2.7148601951104046e-05,
      "loss": 2.3411544799804687,
      "memory(GiB)": 77.56,
      "step": 75985,
      "token_acc": 0.5164473684210527,
      "train_speed(iter/s)": 1.437665
    },
    {
      "epoch": 3.2556445739257103,
      "grad_norm": 5.716641426086426,
      "learning_rate": 2.71426163458096e-05,
      "loss": 2.1421302795410155,
      "memory(GiB)": 77.56,
      "step": 75990,
      "token_acc": 0.5427509293680297,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 3.255858789254959,
      "grad_norm": 8.57295036315918,
      "learning_rate": 2.7136631154598124e-05,
      "loss": 2.341967010498047,
      "memory(GiB)": 77.56,
      "step": 75995,
      "token_acc": 0.5538461538461539,
      "train_speed(iter/s)": 1.437673
    },
    {
      "epoch": 3.256073004584208,
      "grad_norm": 4.116888046264648,
      "learning_rate": 2.7130646377578018e-05,
      "loss": 2.391357421875,
      "memory(GiB)": 77.56,
      "step": 76000,
      "token_acc": 0.4882154882154882,
      "train_speed(iter/s)": 1.437664
    },
    {
      "epoch": 3.256073004584208,
      "eval_loss": 2.453305244445801,
      "eval_runtime": 13.907,
      "eval_samples_per_second": 7.191,
      "eval_steps_per_second": 7.191,
      "eval_token_acc": 0.4508301404853129,
      "step": 76000
    },
    {
      "epoch": 3.256287219913457,
      "grad_norm": 5.90869140625,
      "learning_rate": 2.712466201485771e-05,
      "loss": 2.2543615341186523,
      "memory(GiB)": 77.56,
      "step": 76005,
      "token_acc": 0.46131805157593125,
      "train_speed(iter/s)": 1.437254
    },
    {
      "epoch": 3.256501435242706,
      "grad_norm": 6.839172840118408,
      "learning_rate": 2.7118678066545607e-05,
      "loss": 2.292075347900391,
      "memory(GiB)": 77.56,
      "step": 76010,
      "token_acc": 0.5460750853242321,
      "train_speed(iter/s)": 1.437234
    },
    {
      "epoch": 3.256715650571955,
      "grad_norm": 5.877674102783203,
      "learning_rate": 2.7112694532750092e-05,
      "loss": 2.1154727935791016,
      "memory(GiB)": 77.56,
      "step": 76015,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437236
    },
    {
      "epoch": 3.256929865901204,
      "grad_norm": 6.921167373657227,
      "learning_rate": 2.7106711413579587e-05,
      "loss": 2.164701461791992,
      "memory(GiB)": 77.56,
      "step": 76020,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.437246
    },
    {
      "epoch": 3.257144081230453,
      "grad_norm": 8.60682201385498,
      "learning_rate": 2.710072870914251e-05,
      "loss": 2.1594038009643555,
      "memory(GiB)": 77.56,
      "step": 76025,
      "token_acc": 0.49377593360995853,
      "train_speed(iter/s)": 1.437255
    },
    {
      "epoch": 3.2573582965597017,
      "grad_norm": 6.01790189743042,
      "learning_rate": 2.7094746419547202e-05,
      "loss": 2.4450910568237303,
      "memory(GiB)": 77.56,
      "step": 76030,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.437247
    },
    {
      "epoch": 3.257572511888951,
      "grad_norm": 5.704244613647461,
      "learning_rate": 2.7088764544902057e-05,
      "loss": 2.2321807861328127,
      "memory(GiB)": 77.56,
      "step": 76035,
      "token_acc": 0.5465116279069767,
      "train_speed(iter/s)": 1.43721
    },
    {
      "epoch": 3.2577867272181997,
      "grad_norm": 5.279576301574707,
      "learning_rate": 2.7082783085315426e-05,
      "loss": 2.1731287002563477,
      "memory(GiB)": 77.56,
      "step": 76040,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.43721
    },
    {
      "epoch": 3.2580009425474485,
      "grad_norm": 7.912080764770508,
      "learning_rate": 2.7076802040895666e-05,
      "loss": 2.3184326171875,
      "memory(GiB)": 77.56,
      "step": 76045,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.437213
    },
    {
      "epoch": 3.258215157876698,
      "grad_norm": 4.597708225250244,
      "learning_rate": 2.7070821411751158e-05,
      "loss": 2.3820796966552735,
      "memory(GiB)": 77.56,
      "step": 76050,
      "token_acc": 0.4578313253012048,
      "train_speed(iter/s)": 1.437211
    },
    {
      "epoch": 3.2584293732059466,
      "grad_norm": 6.737573623657227,
      "learning_rate": 2.7064841197990225e-05,
      "loss": 2.0597745895385744,
      "memory(GiB)": 77.56,
      "step": 76055,
      "token_acc": 0.5372549019607843,
      "train_speed(iter/s)": 1.437224
    },
    {
      "epoch": 3.2586435885351954,
      "grad_norm": 5.4444899559021,
      "learning_rate": 2.7058861399721213e-05,
      "loss": 2.1524993896484377,
      "memory(GiB)": 77.56,
      "step": 76060,
      "token_acc": 0.5222222222222223,
      "train_speed(iter/s)": 1.437213
    },
    {
      "epoch": 3.2588578038644447,
      "grad_norm": 5.415369033813477,
      "learning_rate": 2.705288201705245e-05,
      "loss": 2.2321901321411133,
      "memory(GiB)": 77.56,
      "step": 76065,
      "token_acc": 0.5136186770428015,
      "train_speed(iter/s)": 1.43722
    },
    {
      "epoch": 3.2590720191936935,
      "grad_norm": 4.683697700500488,
      "learning_rate": 2.704690305009225e-05,
      "loss": 2.1994834899902345,
      "memory(GiB)": 77.56,
      "step": 76070,
      "token_acc": 0.501466275659824,
      "train_speed(iter/s)": 1.437221
    },
    {
      "epoch": 3.2592862345229423,
      "grad_norm": 5.0263237953186035,
      "learning_rate": 2.7040924498948918e-05,
      "loss": 2.5463226318359373,
      "memory(GiB)": 77.56,
      "step": 76075,
      "token_acc": 0.48909657320872274,
      "train_speed(iter/s)": 1.437222
    },
    {
      "epoch": 3.2595004498521916,
      "grad_norm": 8.046988487243652,
      "learning_rate": 2.703494636373079e-05,
      "loss": 2.2907894134521483,
      "memory(GiB)": 77.56,
      "step": 76080,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.437237
    },
    {
      "epoch": 3.2597146651814404,
      "grad_norm": 6.496913909912109,
      "learning_rate": 2.7028968644546154e-05,
      "loss": 2.308920478820801,
      "memory(GiB)": 77.56,
      "step": 76085,
      "token_acc": 0.48134328358208955,
      "train_speed(iter/s)": 1.437257
    },
    {
      "epoch": 3.259928880510689,
      "grad_norm": 7.8755083084106445,
      "learning_rate": 2.7022991341503302e-05,
      "loss": 2.273288345336914,
      "memory(GiB)": 77.56,
      "step": 76090,
      "token_acc": 0.5270758122743683,
      "train_speed(iter/s)": 1.43727
    },
    {
      "epoch": 3.2601430958399384,
      "grad_norm": 5.517001628875732,
      "learning_rate": 2.7017014454710492e-05,
      "loss": 2.3058496475219727,
      "memory(GiB)": 77.56,
      "step": 76095,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.437267
    },
    {
      "epoch": 3.2603573111691873,
      "grad_norm": 6.559329986572266,
      "learning_rate": 2.7011037984276043e-05,
      "loss": 2.04239559173584,
      "memory(GiB)": 77.56,
      "step": 76100,
      "token_acc": 0.5126050420168067,
      "train_speed(iter/s)": 1.437256
    },
    {
      "epoch": 3.260571526498436,
      "grad_norm": 6.379025936126709,
      "learning_rate": 2.7005061930308195e-05,
      "loss": 2.5620384216308594,
      "memory(GiB)": 77.56,
      "step": 76105,
      "token_acc": 0.4923469387755102,
      "train_speed(iter/s)": 1.437251
    },
    {
      "epoch": 3.2607857418276853,
      "grad_norm": 10.403284072875977,
      "learning_rate": 2.6999086292915234e-05,
      "loss": 2.448537254333496,
      "memory(GiB)": 77.56,
      "step": 76110,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.437265
    },
    {
      "epoch": 3.260999957156934,
      "grad_norm": 5.881539344787598,
      "learning_rate": 2.6993111072205406e-05,
      "loss": 2.5906185150146483,
      "memory(GiB)": 77.56,
      "step": 76115,
      "token_acc": 0.4664804469273743,
      "train_speed(iter/s)": 1.437236
    },
    {
      "epoch": 3.261214172486183,
      "grad_norm": 5.683322429656982,
      "learning_rate": 2.698713626828695e-05,
      "loss": 2.2533668518066405,
      "memory(GiB)": 77.56,
      "step": 76120,
      "token_acc": 0.5345454545454545,
      "train_speed(iter/s)": 1.43724
    },
    {
      "epoch": 3.261428387815432,
      "grad_norm": 5.9745965003967285,
      "learning_rate": 2.6981161881268113e-05,
      "loss": 2.428679847717285,
      "memory(GiB)": 77.56,
      "step": 76125,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.437228
    },
    {
      "epoch": 3.261642603144681,
      "grad_norm": 5.338161945343018,
      "learning_rate": 2.6975187911257116e-05,
      "loss": 2.313723564147949,
      "memory(GiB)": 77.56,
      "step": 76130,
      "token_acc": 0.511326860841424,
      "train_speed(iter/s)": 1.437243
    },
    {
      "epoch": 3.26185681847393,
      "grad_norm": 4.4195685386657715,
      "learning_rate": 2.696921435836217e-05,
      "loss": 2.33978385925293,
      "memory(GiB)": 77.56,
      "step": 76135,
      "token_acc": 0.514018691588785,
      "train_speed(iter/s)": 1.437222
    },
    {
      "epoch": 3.262071033803179,
      "grad_norm": 5.146104335784912,
      "learning_rate": 2.6963241222691533e-05,
      "loss": 2.4369155883789064,
      "memory(GiB)": 77.56,
      "step": 76140,
      "token_acc": 0.48307692307692307,
      "train_speed(iter/s)": 1.437216
    },
    {
      "epoch": 3.262285249132428,
      "grad_norm": 6.217367649078369,
      "learning_rate": 2.6957268504353394e-05,
      "loss": 2.2363914489746093,
      "memory(GiB)": 77.56,
      "step": 76145,
      "token_acc": 0.4957627118644068,
      "train_speed(iter/s)": 1.43722
    },
    {
      "epoch": 3.2624994644616767,
      "grad_norm": 4.863766193389893,
      "learning_rate": 2.6951296203455945e-05,
      "loss": 2.2368846893310548,
      "memory(GiB)": 77.56,
      "step": 76150,
      "token_acc": 0.5222929936305732,
      "train_speed(iter/s)": 1.437194
    },
    {
      "epoch": 3.262713679790926,
      "grad_norm": 6.6296610832214355,
      "learning_rate": 2.694532432010739e-05,
      "loss": 2.4694311141967775,
      "memory(GiB)": 77.56,
      "step": 76155,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.437185
    },
    {
      "epoch": 3.262927895120175,
      "grad_norm": 6.070828437805176,
      "learning_rate": 2.693935285441589e-05,
      "loss": 2.4765954971313477,
      "memory(GiB)": 77.56,
      "step": 76160,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.437183
    },
    {
      "epoch": 3.2631421104494236,
      "grad_norm": 10.059720993041992,
      "learning_rate": 2.693338180648965e-05,
      "loss": 2.475826644897461,
      "memory(GiB)": 77.56,
      "step": 76165,
      "token_acc": 0.4743083003952569,
      "train_speed(iter/s)": 1.43718
    },
    {
      "epoch": 3.263356325778673,
      "grad_norm": 4.910816669464111,
      "learning_rate": 2.6927411176436857e-05,
      "loss": 2.346239471435547,
      "memory(GiB)": 77.56,
      "step": 76170,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.437192
    },
    {
      "epoch": 3.2635705411079217,
      "grad_norm": 5.955539226531982,
      "learning_rate": 2.692144096436565e-05,
      "loss": 2.270440864562988,
      "memory(GiB)": 77.56,
      "step": 76175,
      "token_acc": 0.5146443514644351,
      "train_speed(iter/s)": 1.437211
    },
    {
      "epoch": 3.2637847564371705,
      "grad_norm": 5.154200553894043,
      "learning_rate": 2.6915471170384188e-05,
      "loss": 2.294374465942383,
      "memory(GiB)": 77.56,
      "step": 76180,
      "token_acc": 0.5130111524163569,
      "train_speed(iter/s)": 1.437219
    },
    {
      "epoch": 3.2639989717664197,
      "grad_norm": 6.375717639923096,
      "learning_rate": 2.6909501794600622e-05,
      "loss": 2.4333267211914062,
      "memory(GiB)": 77.56,
      "step": 76185,
      "token_acc": 0.5404255319148936,
      "train_speed(iter/s)": 1.43722
    },
    {
      "epoch": 3.2642131870956685,
      "grad_norm": 6.558661937713623,
      "learning_rate": 2.690353283712308e-05,
      "loss": 2.5455415725708006,
      "memory(GiB)": 77.56,
      "step": 76190,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.437214
    },
    {
      "epoch": 3.2644274024249174,
      "grad_norm": 5.703103065490723,
      "learning_rate": 2.689756429805972e-05,
      "loss": 2.2252275466918947,
      "memory(GiB)": 77.56,
      "step": 76195,
      "token_acc": 0.5080906148867314,
      "train_speed(iter/s)": 1.437222
    },
    {
      "epoch": 3.2646416177541666,
      "grad_norm": 5.283342361450195,
      "learning_rate": 2.6891596177518652e-05,
      "loss": 2.2977081298828126,
      "memory(GiB)": 77.56,
      "step": 76200,
      "token_acc": 0.5337837837837838,
      "train_speed(iter/s)": 1.437221
    },
    {
      "epoch": 3.2648558330834154,
      "grad_norm": 4.891332626342773,
      "learning_rate": 2.6885628475608006e-05,
      "loss": 2.4547088623046873,
      "memory(GiB)": 77.56,
      "step": 76205,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.437223
    },
    {
      "epoch": 3.2650700484126642,
      "grad_norm": 7.913602352142334,
      "learning_rate": 2.6879661192435866e-05,
      "loss": 2.3808162689208983,
      "memory(GiB)": 77.56,
      "step": 76210,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.437223
    },
    {
      "epoch": 3.2652842637419135,
      "grad_norm": 5.7912187576293945,
      "learning_rate": 2.6873694328110365e-05,
      "loss": 2.488439178466797,
      "memory(GiB)": 77.56,
      "step": 76215,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437243
    },
    {
      "epoch": 3.2654984790711623,
      "grad_norm": 5.645628452301025,
      "learning_rate": 2.6867727882739568e-05,
      "loss": 2.3716983795166016,
      "memory(GiB)": 77.56,
      "step": 76220,
      "token_acc": 0.5186440677966102,
      "train_speed(iter/s)": 1.437247
    },
    {
      "epoch": 3.265712694400411,
      "grad_norm": 5.816945552825928,
      "learning_rate": 2.6861761856431595e-05,
      "loss": 2.4469577789306642,
      "memory(GiB)": 77.56,
      "step": 76225,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.2659269097296604,
      "grad_norm": 6.040002822875977,
      "learning_rate": 2.6855796249294512e-05,
      "loss": 2.187143325805664,
      "memory(GiB)": 77.56,
      "step": 76230,
      "token_acc": 0.5313807531380753,
      "train_speed(iter/s)": 1.437226
    },
    {
      "epoch": 3.266141125058909,
      "grad_norm": 5.661389350891113,
      "learning_rate": 2.6849831061436394e-05,
      "loss": 2.1834054946899415,
      "memory(GiB)": 77.56,
      "step": 76235,
      "token_acc": 0.5432835820895522,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.266355340388158,
      "grad_norm": 4.8527116775512695,
      "learning_rate": 2.684386629296528e-05,
      "loss": 2.133860778808594,
      "memory(GiB)": 77.56,
      "step": 76240,
      "token_acc": 0.5390070921985816,
      "train_speed(iter/s)": 1.437244
    },
    {
      "epoch": 3.2665695557174073,
      "grad_norm": 5.037625312805176,
      "learning_rate": 2.683790194398927e-05,
      "loss": 2.1706300735473634,
      "memory(GiB)": 77.56,
      "step": 76245,
      "token_acc": 0.527972027972028,
      "train_speed(iter/s)": 1.437266
    },
    {
      "epoch": 3.266783771046656,
      "grad_norm": 6.797175884246826,
      "learning_rate": 2.6831938014616377e-05,
      "loss": 2.1708532333374024,
      "memory(GiB)": 77.56,
      "step": 76250,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.437273
    },
    {
      "epoch": 3.266997986375905,
      "grad_norm": 8.627185821533203,
      "learning_rate": 2.6825974504954676e-05,
      "loss": 2.309407424926758,
      "memory(GiB)": 77.56,
      "step": 76255,
      "token_acc": 0.5223214285714286,
      "train_speed(iter/s)": 1.437275
    },
    {
      "epoch": 3.267212201705154,
      "grad_norm": 5.070251941680908,
      "learning_rate": 2.6820011415112183e-05,
      "loss": 2.704513931274414,
      "memory(GiB)": 77.56,
      "step": 76260,
      "token_acc": 0.47249190938511326,
      "train_speed(iter/s)": 1.437292
    },
    {
      "epoch": 3.267426417034403,
      "grad_norm": 7.257891654968262,
      "learning_rate": 2.6814048745196933e-05,
      "loss": 2.2436540603637694,
      "memory(GiB)": 77.56,
      "step": 76265,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.437299
    },
    {
      "epoch": 3.2676406323636518,
      "grad_norm": 6.010847568511963,
      "learning_rate": 2.6808086495316943e-05,
      "loss": 2.2097457885742187,
      "memory(GiB)": 77.56,
      "step": 76270,
      "token_acc": 0.5020242914979757,
      "train_speed(iter/s)": 1.437297
    },
    {
      "epoch": 3.267854847692901,
      "grad_norm": 6.027689456939697,
      "learning_rate": 2.68021246655802e-05,
      "loss": 2.744255256652832,
      "memory(GiB)": 77.56,
      "step": 76275,
      "token_acc": 0.45977011494252873,
      "train_speed(iter/s)": 1.437286
    },
    {
      "epoch": 3.26806906302215,
      "grad_norm": 9.140928268432617,
      "learning_rate": 2.679616325609474e-05,
      "loss": 2.4225658416748046,
      "memory(GiB)": 77.56,
      "step": 76280,
      "token_acc": 0.5028735632183908,
      "train_speed(iter/s)": 1.437295
    },
    {
      "epoch": 3.2682832783513986,
      "grad_norm": 6.586650371551514,
      "learning_rate": 2.679020226696856e-05,
      "loss": 2.3298568725585938,
      "memory(GiB)": 77.56,
      "step": 76285,
      "token_acc": 0.5105740181268882,
      "train_speed(iter/s)": 1.437285
    },
    {
      "epoch": 3.268497493680648,
      "grad_norm": 6.096415042877197,
      "learning_rate": 2.678424169830964e-05,
      "loss": 2.3894872665405273,
      "memory(GiB)": 77.56,
      "step": 76290,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.437278
    },
    {
      "epoch": 3.2687117090098967,
      "grad_norm": 8.352396965026855,
      "learning_rate": 2.6778281550225957e-05,
      "loss": 2.2749988555908205,
      "memory(GiB)": 77.56,
      "step": 76295,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.437268
    },
    {
      "epoch": 3.2689259243391455,
      "grad_norm": 6.9940056800842285,
      "learning_rate": 2.6772321822825486e-05,
      "loss": 2.0341529846191406,
      "memory(GiB)": 77.56,
      "step": 76300,
      "token_acc": 0.5692883895131086,
      "train_speed(iter/s)": 1.43728
    },
    {
      "epoch": 3.269140139668395,
      "grad_norm": 6.248111724853516,
      "learning_rate": 2.6766362516216177e-05,
      "loss": 2.4225185394287108,
      "memory(GiB)": 77.56,
      "step": 76305,
      "token_acc": 0.4897360703812317,
      "train_speed(iter/s)": 1.437281
    },
    {
      "epoch": 3.2693543549976436,
      "grad_norm": 7.4152679443359375,
      "learning_rate": 2.6760403630506027e-05,
      "loss": 2.1001409530639648,
      "memory(GiB)": 77.56,
      "step": 76310,
      "token_acc": 0.510548523206751,
      "train_speed(iter/s)": 1.437279
    },
    {
      "epoch": 3.2695685703268924,
      "grad_norm": 6.702224254608154,
      "learning_rate": 2.6754445165802938e-05,
      "loss": 2.505830764770508,
      "memory(GiB)": 77.56,
      "step": 76315,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.437286
    },
    {
      "epoch": 3.2697827856561417,
      "grad_norm": 5.524763584136963,
      "learning_rate": 2.6748487122214905e-05,
      "loss": 2.4238554000854493,
      "memory(GiB)": 77.56,
      "step": 76320,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.43727
    },
    {
      "epoch": 3.2699970009853905,
      "grad_norm": 4.90808629989624,
      "learning_rate": 2.6742529499849846e-05,
      "loss": 2.3672775268554687,
      "memory(GiB)": 77.56,
      "step": 76325,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.437276
    },
    {
      "epoch": 3.2702112163146393,
      "grad_norm": 6.328675270080566,
      "learning_rate": 2.673657229881567e-05,
      "loss": 2.0318288803100586,
      "memory(GiB)": 77.56,
      "step": 76330,
      "token_acc": 0.516728624535316,
      "train_speed(iter/s)": 1.437292
    },
    {
      "epoch": 3.2704254316438885,
      "grad_norm": 4.75703763961792,
      "learning_rate": 2.67306155192203e-05,
      "loss": 2.1186452865600587,
      "memory(GiB)": 77.56,
      "step": 76335,
      "token_acc": 0.5594405594405595,
      "train_speed(iter/s)": 1.437306
    },
    {
      "epoch": 3.2706396469731374,
      "grad_norm": 5.916680812835693,
      "learning_rate": 2.6724659161171674e-05,
      "loss": 2.036697578430176,
      "memory(GiB)": 77.56,
      "step": 76340,
      "token_acc": 0.5353159851301115,
      "train_speed(iter/s)": 1.437321
    },
    {
      "epoch": 3.270853862302386,
      "grad_norm": 5.238081455230713,
      "learning_rate": 2.671870322477768e-05,
      "loss": 1.9471094131469726,
      "memory(GiB)": 77.56,
      "step": 76345,
      "token_acc": 0.5571428571428572,
      "train_speed(iter/s)": 1.437322
    },
    {
      "epoch": 3.2710680776316354,
      "grad_norm": 6.140990734100342,
      "learning_rate": 2.6712747710146224e-05,
      "loss": 2.4319995880126952,
      "memory(GiB)": 77.56,
      "step": 76350,
      "token_acc": 0.5304659498207885,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 3.2712822929608842,
      "grad_norm": 4.777742385864258,
      "learning_rate": 2.670679261738518e-05,
      "loss": 2.3321855545043944,
      "memory(GiB)": 77.56,
      "step": 76355,
      "token_acc": 0.5275080906148867,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 3.271496508290133,
      "grad_norm": 6.19354772567749,
      "learning_rate": 2.670083794660244e-05,
      "loss": 2.4010643005371093,
      "memory(GiB)": 77.56,
      "step": 76360,
      "token_acc": 0.49624060150375937,
      "train_speed(iter/s)": 1.43732
    },
    {
      "epoch": 3.2717107236193823,
      "grad_norm": 5.319146156311035,
      "learning_rate": 2.669488369790586e-05,
      "loss": 1.9833808898925782,
      "memory(GiB)": 77.56,
      "step": 76365,
      "token_acc": 0.5374592833876222,
      "train_speed(iter/s)": 1.437327
    },
    {
      "epoch": 3.271924938948631,
      "grad_norm": 5.4314985275268555,
      "learning_rate": 2.6688929871403346e-05,
      "loss": 2.597892189025879,
      "memory(GiB)": 77.56,
      "step": 76370,
      "token_acc": 0.4785100286532951,
      "train_speed(iter/s)": 1.437303
    },
    {
      "epoch": 3.27213915427788,
      "grad_norm": 5.978343486785889,
      "learning_rate": 2.6682976467202726e-05,
      "loss": 2.3395435333251955,
      "memory(GiB)": 77.56,
      "step": 76375,
      "token_acc": 0.5131086142322098,
      "train_speed(iter/s)": 1.437331
    },
    {
      "epoch": 3.272353369607129,
      "grad_norm": 5.326536655426025,
      "learning_rate": 2.6677023485411866e-05,
      "loss": 2.5036598205566407,
      "memory(GiB)": 77.56,
      "step": 76380,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.437308
    },
    {
      "epoch": 3.272567584936378,
      "grad_norm": 6.113457202911377,
      "learning_rate": 2.667107092613861e-05,
      "loss": 2.455510139465332,
      "memory(GiB)": 77.56,
      "step": 76385,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.437288
    },
    {
      "epoch": 3.272781800265627,
      "grad_norm": 6.942384243011475,
      "learning_rate": 2.6665118789490763e-05,
      "loss": 2.1142892837524414,
      "memory(GiB)": 77.56,
      "step": 76390,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437277
    },
    {
      "epoch": 3.272996015594876,
      "grad_norm": 5.50494909286499,
      "learning_rate": 2.6659167075576176e-05,
      "loss": 2.049399185180664,
      "memory(GiB)": 77.56,
      "step": 76395,
      "token_acc": 0.5687022900763359,
      "train_speed(iter/s)": 1.437284
    },
    {
      "epoch": 3.273210230924125,
      "grad_norm": 9.861894607543945,
      "learning_rate": 2.6653215784502693e-05,
      "loss": 2.5155601501464844,
      "memory(GiB)": 77.56,
      "step": 76400,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.437289
    },
    {
      "epoch": 3.2734244462533737,
      "grad_norm": 6.839644908905029,
      "learning_rate": 2.664726491637811e-05,
      "loss": 2.615404510498047,
      "memory(GiB)": 77.56,
      "step": 76405,
      "token_acc": 0.4921259842519685,
      "train_speed(iter/s)": 1.437302
    },
    {
      "epoch": 3.273638661582623,
      "grad_norm": 9.163671493530273,
      "learning_rate": 2.664131447131023e-05,
      "loss": 2.20394287109375,
      "memory(GiB)": 77.56,
      "step": 76410,
      "token_acc": 0.5055762081784386,
      "train_speed(iter/s)": 1.437322
    },
    {
      "epoch": 3.2738528769118718,
      "grad_norm": 6.252237319946289,
      "learning_rate": 2.6635364449406853e-05,
      "loss": 2.2719280242919924,
      "memory(GiB)": 77.56,
      "step": 76415,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.437328
    },
    {
      "epoch": 3.274067092241121,
      "grad_norm": 5.88748836517334,
      "learning_rate": 2.662941485077574e-05,
      "loss": 2.2581886291503905,
      "memory(GiB)": 77.56,
      "step": 76420,
      "token_acc": 0.4829059829059829,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 3.27428130757037,
      "grad_norm": 4.954962253570557,
      "learning_rate": 2.6623465675524728e-05,
      "loss": 2.209299087524414,
      "memory(GiB)": 77.56,
      "step": 76425,
      "token_acc": 0.5320754716981132,
      "train_speed(iter/s)": 1.437307
    },
    {
      "epoch": 3.2744955228996186,
      "grad_norm": 6.656917572021484,
      "learning_rate": 2.6617516923761553e-05,
      "loss": 2.1687402725219727,
      "memory(GiB)": 77.56,
      "step": 76430,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 3.274709738228868,
      "grad_norm": 9.126182556152344,
      "learning_rate": 2.6611568595594006e-05,
      "loss": 2.508999252319336,
      "memory(GiB)": 77.56,
      "step": 76435,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.437335
    },
    {
      "epoch": 3.2749239535581167,
      "grad_norm": 4.683877944946289,
      "learning_rate": 2.6605620691129828e-05,
      "loss": 2.7646997451782225,
      "memory(GiB)": 77.56,
      "step": 76440,
      "token_acc": 0.4645161290322581,
      "train_speed(iter/s)": 1.437334
    },
    {
      "epoch": 3.2751381688873655,
      "grad_norm": 6.770195007324219,
      "learning_rate": 2.659967321047678e-05,
      "loss": 2.205705261230469,
      "memory(GiB)": 77.56,
      "step": 76445,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 3.275352384216615,
      "grad_norm": 5.407696723937988,
      "learning_rate": 2.6593726153742575e-05,
      "loss": 2.4108112335205076,
      "memory(GiB)": 77.56,
      "step": 76450,
      "token_acc": 0.45182724252491696,
      "train_speed(iter/s)": 1.437315
    },
    {
      "epoch": 3.2755665995458636,
      "grad_norm": 4.91532039642334,
      "learning_rate": 2.6587779521035007e-05,
      "loss": 2.4610605239868164,
      "memory(GiB)": 77.56,
      "step": 76455,
      "token_acc": 0.5142045454545454,
      "train_speed(iter/s)": 1.437332
    },
    {
      "epoch": 3.2757808148751124,
      "grad_norm": 5.391237258911133,
      "learning_rate": 2.6581833312461768e-05,
      "loss": 2.3996503829956053,
      "memory(GiB)": 77.56,
      "step": 76460,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.437339
    },
    {
      "epoch": 3.2759950302043617,
      "grad_norm": 6.4376091957092285,
      "learning_rate": 2.657588752813057e-05,
      "loss": 2.4347793579101564,
      "memory(GiB)": 77.56,
      "step": 76465,
      "token_acc": 0.5358361774744027,
      "train_speed(iter/s)": 1.437355
    },
    {
      "epoch": 3.2762092455336105,
      "grad_norm": 4.6318793296813965,
      "learning_rate": 2.6569942168149165e-05,
      "loss": 2.0739295959472654,
      "memory(GiB)": 77.56,
      "step": 76470,
      "token_acc": 0.5307443365695793,
      "train_speed(iter/s)": 1.437364
    },
    {
      "epoch": 3.2764234608628593,
      "grad_norm": 6.209949970245361,
      "learning_rate": 2.6563997232625236e-05,
      "loss": 2.6977296829223634,
      "memory(GiB)": 77.56,
      "step": 76475,
      "token_acc": 0.45317220543806647,
      "train_speed(iter/s)": 1.437372
    },
    {
      "epoch": 3.2766376761921086,
      "grad_norm": 6.036227226257324,
      "learning_rate": 2.655805272166646e-05,
      "loss": 2.1383331298828123,
      "memory(GiB)": 77.56,
      "step": 76480,
      "token_acc": 0.5241935483870968,
      "train_speed(iter/s)": 1.437399
    },
    {
      "epoch": 3.2768518915213574,
      "grad_norm": 5.114137649536133,
      "learning_rate": 2.6552108635380567e-05,
      "loss": 2.2285793304443358,
      "memory(GiB)": 77.56,
      "step": 76485,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437413
    },
    {
      "epoch": 3.277066106850606,
      "grad_norm": 5.297479152679443,
      "learning_rate": 2.654616497387523e-05,
      "loss": 2.0856781005859375,
      "memory(GiB)": 77.56,
      "step": 76490,
      "token_acc": 0.5269230769230769,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 3.2772803221798554,
      "grad_norm": 6.1026411056518555,
      "learning_rate": 2.654022173725811e-05,
      "loss": 2.148396110534668,
      "memory(GiB)": 77.56,
      "step": 76495,
      "token_acc": 0.524390243902439,
      "train_speed(iter/s)": 1.437409
    },
    {
      "epoch": 3.2774945375091042,
      "grad_norm": 5.110911846160889,
      "learning_rate": 2.653427892563688e-05,
      "loss": 2.410162925720215,
      "memory(GiB)": 77.56,
      "step": 76500,
      "token_acc": 0.4740061162079511,
      "train_speed(iter/s)": 1.43739
    },
    {
      "epoch": 3.2774945375091042,
      "eval_loss": 2.140929937362671,
      "eval_runtime": 14.9721,
      "eval_samples_per_second": 6.679,
      "eval_steps_per_second": 6.679,
      "eval_token_acc": 0.4733969986357435,
      "step": 76500
    },
    {
      "epoch": 3.277708752838353,
      "grad_norm": 7.50363302230835,
      "learning_rate": 2.65283365391192e-05,
      "loss": 2.338113784790039,
      "memory(GiB)": 77.56,
      "step": 76505,
      "token_acc": 0.48080614203454897,
      "train_speed(iter/s)": 1.436962
    },
    {
      "epoch": 3.2779229681676023,
      "grad_norm": 6.472196578979492,
      "learning_rate": 2.6522394577812702e-05,
      "loss": 2.3285160064697266,
      "memory(GiB)": 77.56,
      "step": 76510,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.436988
    },
    {
      "epoch": 3.278137183496851,
      "grad_norm": 5.848712921142578,
      "learning_rate": 2.6516453041825075e-05,
      "loss": 2.4467742919921873,
      "memory(GiB)": 77.56,
      "step": 76515,
      "token_acc": 0.46688741721854304,
      "train_speed(iter/s)": 1.43698
    },
    {
      "epoch": 3.2783513988261,
      "grad_norm": 4.979660987854004,
      "learning_rate": 2.6510511931263926e-05,
      "loss": 2.132375144958496,
      "memory(GiB)": 77.56,
      "step": 76520,
      "token_acc": 0.5283582089552239,
      "train_speed(iter/s)": 1.436994
    },
    {
      "epoch": 3.278565614155349,
      "grad_norm": 5.430736541748047,
      "learning_rate": 2.6504571246236893e-05,
      "loss": 2.574811553955078,
      "memory(GiB)": 77.56,
      "step": 76525,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.437004
    },
    {
      "epoch": 3.278779829484598,
      "grad_norm": 5.978100776672363,
      "learning_rate": 2.6498630986851584e-05,
      "loss": 2.3917766571044923,
      "memory(GiB)": 77.56,
      "step": 76530,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.437012
    },
    {
      "epoch": 3.278994044813847,
      "grad_norm": 5.435888767242432,
      "learning_rate": 2.649269115321561e-05,
      "loss": 2.55991153717041,
      "memory(GiB)": 77.56,
      "step": 76535,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.437024
    },
    {
      "epoch": 3.279208260143096,
      "grad_norm": 5.174025535583496,
      "learning_rate": 2.6486751745436578e-05,
      "loss": 2.373432922363281,
      "memory(GiB)": 77.56,
      "step": 76540,
      "token_acc": 0.514018691588785,
      "train_speed(iter/s)": 1.437028
    },
    {
      "epoch": 3.279422475472345,
      "grad_norm": 5.387723922729492,
      "learning_rate": 2.648081276362212e-05,
      "loss": 2.4880332946777344,
      "memory(GiB)": 77.56,
      "step": 76545,
      "token_acc": 0.5016611295681063,
      "train_speed(iter/s)": 1.437027
    },
    {
      "epoch": 3.2796366908015937,
      "grad_norm": 5.608835220336914,
      "learning_rate": 2.6474874207879807e-05,
      "loss": 2.5733638763427735,
      "memory(GiB)": 77.56,
      "step": 76550,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.437016
    },
    {
      "epoch": 3.279850906130843,
      "grad_norm": 5.937811851501465,
      "learning_rate": 2.6468936078317207e-05,
      "loss": 2.4545013427734377,
      "memory(GiB)": 77.56,
      "step": 76555,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.437033
    },
    {
      "epoch": 3.2800651214600918,
      "grad_norm": 5.4121575355529785,
      "learning_rate": 2.6462998375041908e-05,
      "loss": 2.1308835983276366,
      "memory(GiB)": 77.56,
      "step": 76560,
      "token_acc": 0.54421768707483,
      "train_speed(iter/s)": 1.437027
    },
    {
      "epoch": 3.2802793367893406,
      "grad_norm": 6.442676067352295,
      "learning_rate": 2.645706109816145e-05,
      "loss": 2.648736763000488,
      "memory(GiB)": 77.56,
      "step": 76565,
      "token_acc": 0.4506578947368421,
      "train_speed(iter/s)": 1.437004
    },
    {
      "epoch": 3.28049355211859,
      "grad_norm": 8.892988204956055,
      "learning_rate": 2.6451124247783443e-05,
      "loss": 2.449600601196289,
      "memory(GiB)": 77.56,
      "step": 76570,
      "token_acc": 0.4779116465863454,
      "train_speed(iter/s)": 1.437007
    },
    {
      "epoch": 3.2807077674478387,
      "grad_norm": 5.588657379150391,
      "learning_rate": 2.6445187824015406e-05,
      "loss": 2.5541194915771483,
      "memory(GiB)": 77.56,
      "step": 76575,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.437004
    },
    {
      "epoch": 3.2809219827770875,
      "grad_norm": 5.2621169090271,
      "learning_rate": 2.6439251826964882e-05,
      "loss": 2.4545656204223634,
      "memory(GiB)": 77.56,
      "step": 76580,
      "token_acc": 0.4779874213836478,
      "train_speed(iter/s)": 1.436999
    },
    {
      "epoch": 3.2811361981063367,
      "grad_norm": 6.59516716003418,
      "learning_rate": 2.6433316256739417e-05,
      "loss": 2.2602039337158204,
      "memory(GiB)": 77.56,
      "step": 76585,
      "token_acc": 0.5061728395061729,
      "train_speed(iter/s)": 1.437025
    },
    {
      "epoch": 3.2813504134355855,
      "grad_norm": 5.80298376083374,
      "learning_rate": 2.6427381113446536e-05,
      "loss": 2.110616111755371,
      "memory(GiB)": 77.56,
      "step": 76590,
      "token_acc": 0.5197368421052632,
      "train_speed(iter/s)": 1.437052
    },
    {
      "epoch": 3.2815646287648343,
      "grad_norm": 6.352737903594971,
      "learning_rate": 2.642144639719374e-05,
      "loss": 2.4751840591430665,
      "memory(GiB)": 77.56,
      "step": 76595,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.437057
    },
    {
      "epoch": 3.2817788440940836,
      "grad_norm": 5.855788707733154,
      "learning_rate": 2.641551210808858e-05,
      "loss": 2.0016483306884765,
      "memory(GiB)": 77.56,
      "step": 76600,
      "token_acc": 0.54296875,
      "train_speed(iter/s)": 1.437079
    },
    {
      "epoch": 3.2819930594233324,
      "grad_norm": 7.938076496124268,
      "learning_rate": 2.640957824623854e-05,
      "loss": 2.1572898864746093,
      "memory(GiB)": 77.56,
      "step": 76605,
      "token_acc": 0.5321428571428571,
      "train_speed(iter/s)": 1.437091
    },
    {
      "epoch": 3.2822072747525812,
      "grad_norm": 5.586814880371094,
      "learning_rate": 2.6403644811751104e-05,
      "loss": 2.162569046020508,
      "memory(GiB)": 77.56,
      "step": 76610,
      "token_acc": 0.5358361774744027,
      "train_speed(iter/s)": 1.437083
    },
    {
      "epoch": 3.2824214900818305,
      "grad_norm": 4.677947998046875,
      "learning_rate": 2.63977118047338e-05,
      "loss": 2.461477279663086,
      "memory(GiB)": 77.56,
      "step": 76615,
      "token_acc": 0.4618055555555556,
      "train_speed(iter/s)": 1.437108
    },
    {
      "epoch": 3.2826357054110793,
      "grad_norm": 6.2018632888793945,
      "learning_rate": 2.6391779225294084e-05,
      "loss": 2.2991186141967774,
      "memory(GiB)": 77.56,
      "step": 76620,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437126
    },
    {
      "epoch": 3.282849920740328,
      "grad_norm": 5.7149434089660645,
      "learning_rate": 2.638584707353941e-05,
      "loss": 2.1063396453857424,
      "memory(GiB)": 77.56,
      "step": 76625,
      "token_acc": 0.553030303030303,
      "train_speed(iter/s)": 1.437132
    },
    {
      "epoch": 3.2830641360695774,
      "grad_norm": 4.538356781005859,
      "learning_rate": 2.637991534957729e-05,
      "loss": 2.4295454025268555,
      "memory(GiB)": 77.56,
      "step": 76630,
      "token_acc": 0.49146757679180886,
      "train_speed(iter/s)": 1.437122
    },
    {
      "epoch": 3.283278351398826,
      "grad_norm": 5.894742965698242,
      "learning_rate": 2.6373984053515167e-05,
      "loss": 2.2914783477783205,
      "memory(GiB)": 77.56,
      "step": 76635,
      "token_acc": 0.5016949152542373,
      "train_speed(iter/s)": 1.437129
    },
    {
      "epoch": 3.283492566728075,
      "grad_norm": 8.037131309509277,
      "learning_rate": 2.636805318546048e-05,
      "loss": 2.257550811767578,
      "memory(GiB)": 77.56,
      "step": 76640,
      "token_acc": 0.5095057034220533,
      "train_speed(iter/s)": 1.437128
    },
    {
      "epoch": 3.2837067820573242,
      "grad_norm": 4.997852325439453,
      "learning_rate": 2.636212274552068e-05,
      "loss": 2.2170669555664064,
      "memory(GiB)": 77.56,
      "step": 76645,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437133
    },
    {
      "epoch": 3.283920997386573,
      "grad_norm": 4.542362213134766,
      "learning_rate": 2.6356192733803197e-05,
      "loss": 2.2677452087402346,
      "memory(GiB)": 77.56,
      "step": 76650,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.437142
    },
    {
      "epoch": 3.284135212715822,
      "grad_norm": 5.554403305053711,
      "learning_rate": 2.6350263150415443e-05,
      "loss": 2.2887750625610352,
      "memory(GiB)": 77.56,
      "step": 76655,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.437116
    },
    {
      "epoch": 3.284349428045071,
      "grad_norm": 8.246650695800781,
      "learning_rate": 2.6344333995464875e-05,
      "loss": 2.35384521484375,
      "memory(GiB)": 77.56,
      "step": 76660,
      "token_acc": 0.46613545816733065,
      "train_speed(iter/s)": 1.437145
    },
    {
      "epoch": 3.28456364337432,
      "grad_norm": 5.06002140045166,
      "learning_rate": 2.6338405269058896e-05,
      "loss": 2.462234878540039,
      "memory(GiB)": 77.56,
      "step": 76665,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.437132
    },
    {
      "epoch": 3.2847778587035688,
      "grad_norm": 6.219139099121094,
      "learning_rate": 2.6332476971304898e-05,
      "loss": 2.5646623611450194,
      "memory(GiB)": 77.56,
      "step": 76670,
      "token_acc": 0.5119047619047619,
      "train_speed(iter/s)": 1.437123
    },
    {
      "epoch": 3.284992074032818,
      "grad_norm": 7.4939866065979,
      "learning_rate": 2.6326549102310284e-05,
      "loss": 2.5314207077026367,
      "memory(GiB)": 77.56,
      "step": 76675,
      "token_acc": 0.49612403100775193,
      "train_speed(iter/s)": 1.437129
    },
    {
      "epoch": 3.285206289362067,
      "grad_norm": 6.062115669250488,
      "learning_rate": 2.632062166218241e-05,
      "loss": 2.7366336822509765,
      "memory(GiB)": 77.56,
      "step": 76680,
      "token_acc": 0.44807121661721067,
      "train_speed(iter/s)": 1.437138
    },
    {
      "epoch": 3.2854205046913156,
      "grad_norm": 5.411129474639893,
      "learning_rate": 2.6314694651028697e-05,
      "loss": 2.3598533630371095,
      "memory(GiB)": 77.56,
      "step": 76685,
      "token_acc": 0.527027027027027,
      "train_speed(iter/s)": 1.43713
    },
    {
      "epoch": 3.285634720020565,
      "grad_norm": 6.083203315734863,
      "learning_rate": 2.630876806895653e-05,
      "loss": 2.06467342376709,
      "memory(GiB)": 77.56,
      "step": 76690,
      "token_acc": 0.5667870036101083,
      "train_speed(iter/s)": 1.437128
    },
    {
      "epoch": 3.2858489353498137,
      "grad_norm": 6.668188571929932,
      "learning_rate": 2.630284191607325e-05,
      "loss": 2.583022880554199,
      "memory(GiB)": 77.56,
      "step": 76695,
      "token_acc": 0.4645390070921986,
      "train_speed(iter/s)": 1.437149
    },
    {
      "epoch": 3.2860631506790625,
      "grad_norm": 5.002578258514404,
      "learning_rate": 2.629691619248622e-05,
      "loss": 2.5051143646240233,
      "memory(GiB)": 77.56,
      "step": 76700,
      "token_acc": 0.4674922600619195,
      "train_speed(iter/s)": 1.43717
    },
    {
      "epoch": 3.2862773660083118,
      "grad_norm": 7.534294128417969,
      "learning_rate": 2.6290990898302786e-05,
      "loss": 2.527523422241211,
      "memory(GiB)": 77.56,
      "step": 76705,
      "token_acc": 0.5036496350364964,
      "train_speed(iter/s)": 1.437165
    },
    {
      "epoch": 3.2864915813375606,
      "grad_norm": 4.781052589416504,
      "learning_rate": 2.6285066033630278e-05,
      "loss": 2.4452846527099608,
      "memory(GiB)": 77.56,
      "step": 76710,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.437174
    },
    {
      "epoch": 3.2867057966668094,
      "grad_norm": 6.081940174102783,
      "learning_rate": 2.6279141598576062e-05,
      "loss": 2.662849998474121,
      "memory(GiB)": 77.56,
      "step": 76715,
      "token_acc": 0.4779874213836478,
      "train_speed(iter/s)": 1.437179
    },
    {
      "epoch": 3.2869200119960587,
      "grad_norm": 5.028762340545654,
      "learning_rate": 2.6273217593247447e-05,
      "loss": 2.49237060546875,
      "memory(GiB)": 77.56,
      "step": 76720,
      "token_acc": 0.4970588235294118,
      "train_speed(iter/s)": 1.437202
    },
    {
      "epoch": 3.2871342273253075,
      "grad_norm": 6.06223726272583,
      "learning_rate": 2.6267294017751753e-05,
      "loss": 2.2386714935302736,
      "memory(GiB)": 77.56,
      "step": 76725,
      "token_acc": 0.5559701492537313,
      "train_speed(iter/s)": 1.437195
    },
    {
      "epoch": 3.2873484426545563,
      "grad_norm": 8.5480375289917,
      "learning_rate": 2.626137087219629e-05,
      "loss": 2.0952470779418944,
      "memory(GiB)": 77.56,
      "step": 76730,
      "token_acc": 0.5276752767527675,
      "train_speed(iter/s)": 1.437213
    },
    {
      "epoch": 3.2875626579838055,
      "grad_norm": 4.992301940917969,
      "learning_rate": 2.625544815668836e-05,
      "loss": 2.448184776306152,
      "memory(GiB)": 77.56,
      "step": 76735,
      "token_acc": 0.4783950617283951,
      "train_speed(iter/s)": 1.437226
    },
    {
      "epoch": 3.2877768733130543,
      "grad_norm": 5.484918594360352,
      "learning_rate": 2.6249525871335246e-05,
      "loss": 2.1390968322753907,
      "memory(GiB)": 77.56,
      "step": 76740,
      "token_acc": 0.5445544554455446,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.287991088642303,
      "grad_norm": 6.212332248687744,
      "learning_rate": 2.624360401624427e-05,
      "loss": 2.816640853881836,
      "memory(GiB)": 77.56,
      "step": 76745,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.437218
    },
    {
      "epoch": 3.2882053039715524,
      "grad_norm": 5.6339898109436035,
      "learning_rate": 2.6237682591522693e-05,
      "loss": 2.29492301940918,
      "memory(GiB)": 77.56,
      "step": 76750,
      "token_acc": 0.5165562913907285,
      "train_speed(iter/s)": 1.437233
    },
    {
      "epoch": 3.2884195193008012,
      "grad_norm": 5.280892372131348,
      "learning_rate": 2.6231761597277783e-05,
      "loss": 2.4200132369995115,
      "memory(GiB)": 77.56,
      "step": 76755,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.437228
    },
    {
      "epoch": 3.28863373463005,
      "grad_norm": 5.667395114898682,
      "learning_rate": 2.6225841033616794e-05,
      "loss": 2.4845457077026367,
      "memory(GiB)": 77.56,
      "step": 76760,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.43722
    },
    {
      "epoch": 3.2888479499592993,
      "grad_norm": 5.819214820861816,
      "learning_rate": 2.621992090064701e-05,
      "loss": 2.399454879760742,
      "memory(GiB)": 77.56,
      "step": 76765,
      "token_acc": 0.49393939393939396,
      "train_speed(iter/s)": 1.437222
    },
    {
      "epoch": 3.289062165288548,
      "grad_norm": 5.851830959320068,
      "learning_rate": 2.6214001198475645e-05,
      "loss": 2.3173423767089845,
      "memory(GiB)": 77.56,
      "step": 76770,
      "token_acc": 0.5259515570934256,
      "train_speed(iter/s)": 1.437209
    },
    {
      "epoch": 3.289276380617797,
      "grad_norm": 5.584312915802002,
      "learning_rate": 2.6208081927209988e-05,
      "loss": 2.316762161254883,
      "memory(GiB)": 77.56,
      "step": 76775,
      "token_acc": 0.5266666666666666,
      "train_speed(iter/s)": 1.4372
    },
    {
      "epoch": 3.289490595947046,
      "grad_norm": 6.209677219390869,
      "learning_rate": 2.6202163086957243e-05,
      "loss": 2.1203369140625,
      "memory(GiB)": 77.56,
      "step": 76780,
      "token_acc": 0.51953125,
      "train_speed(iter/s)": 1.43721
    },
    {
      "epoch": 3.289704811276295,
      "grad_norm": 4.925808429718018,
      "learning_rate": 2.6196244677824637e-05,
      "loss": 2.4465789794921875,
      "memory(GiB)": 77.56,
      "step": 76785,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.43722
    },
    {
      "epoch": 3.289919026605544,
      "grad_norm": 4.3598713874816895,
      "learning_rate": 2.6190326699919388e-05,
      "loss": 2.226450729370117,
      "memory(GiB)": 77.56,
      "step": 76790,
      "token_acc": 0.5015673981191222,
      "train_speed(iter/s)": 1.437235
    },
    {
      "epoch": 3.290133241934793,
      "grad_norm": 4.940312385559082,
      "learning_rate": 2.618440915334871e-05,
      "loss": 2.213574028015137,
      "memory(GiB)": 77.56,
      "step": 76795,
      "token_acc": 0.5278688524590164,
      "train_speed(iter/s)": 1.437224
    },
    {
      "epoch": 3.290347457264042,
      "grad_norm": 6.657808780670166,
      "learning_rate": 2.617849203821978e-05,
      "loss": 2.331351470947266,
      "memory(GiB)": 77.56,
      "step": 76800,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.437254
    },
    {
      "epoch": 3.2905616725932907,
      "grad_norm": 4.6629638671875,
      "learning_rate": 2.617257535463983e-05,
      "loss": 2.223450469970703,
      "memory(GiB)": 77.56,
      "step": 76805,
      "token_acc": 0.5031645569620253,
      "train_speed(iter/s)": 1.437265
    },
    {
      "epoch": 3.29077588792254,
      "grad_norm": 5.935922622680664,
      "learning_rate": 2.616665910271603e-05,
      "loss": 2.170707130432129,
      "memory(GiB)": 77.56,
      "step": 76810,
      "token_acc": 0.5477031802120141,
      "train_speed(iter/s)": 1.437292
    },
    {
      "epoch": 3.2909901032517888,
      "grad_norm": 9.75797176361084,
      "learning_rate": 2.616074328255556e-05,
      "loss": 2.4682870864868165,
      "memory(GiB)": 77.56,
      "step": 76815,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.437308
    },
    {
      "epoch": 3.2912043185810376,
      "grad_norm": 5.185744762420654,
      "learning_rate": 2.6154827894265588e-05,
      "loss": 2.218869209289551,
      "memory(GiB)": 77.56,
      "step": 76820,
      "token_acc": 0.5029585798816568,
      "train_speed(iter/s)": 1.437316
    },
    {
      "epoch": 3.291418533910287,
      "grad_norm": 5.338229656219482,
      "learning_rate": 2.6148912937953256e-05,
      "loss": 2.3645278930664064,
      "memory(GiB)": 77.56,
      "step": 76825,
      "token_acc": 0.5392491467576792,
      "train_speed(iter/s)": 1.437317
    },
    {
      "epoch": 3.2916327492395356,
      "grad_norm": 6.619795799255371,
      "learning_rate": 2.614299841372576e-05,
      "loss": 2.0638406753540037,
      "memory(GiB)": 77.56,
      "step": 76830,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.437329
    },
    {
      "epoch": 3.2918469645687845,
      "grad_norm": 5.00105094909668,
      "learning_rate": 2.613708432169021e-05,
      "loss": 2.5913578033447267,
      "memory(GiB)": 77.56,
      "step": 76835,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.437327
    },
    {
      "epoch": 3.2920611798980337,
      "grad_norm": 5.717588424682617,
      "learning_rate": 2.613117066195378e-05,
      "loss": 2.1996074676513673,
      "memory(GiB)": 77.56,
      "step": 76840,
      "token_acc": 0.5300353356890459,
      "train_speed(iter/s)": 1.437329
    },
    {
      "epoch": 3.2922753952272825,
      "grad_norm": 4.773777961730957,
      "learning_rate": 2.6125257434623584e-05,
      "loss": 2.459240531921387,
      "memory(GiB)": 77.56,
      "step": 76845,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 3.2924896105565313,
      "grad_norm": 7.958888530731201,
      "learning_rate": 2.6119344639806753e-05,
      "loss": 2.505276107788086,
      "memory(GiB)": 77.56,
      "step": 76850,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.43729
    },
    {
      "epoch": 3.2927038258857806,
      "grad_norm": 4.454339027404785,
      "learning_rate": 2.6113432277610367e-05,
      "loss": 1.8986431121826173,
      "memory(GiB)": 77.56,
      "step": 76855,
      "token_acc": 0.5766666666666667,
      "train_speed(iter/s)": 1.437236
    },
    {
      "epoch": 3.2929180412150294,
      "grad_norm": 5.445559501647949,
      "learning_rate": 2.6107520348141585e-05,
      "loss": 2.406661605834961,
      "memory(GiB)": 77.56,
      "step": 76860,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.437237
    },
    {
      "epoch": 3.293132256544278,
      "grad_norm": 5.683530807495117,
      "learning_rate": 2.6101608851507486e-05,
      "loss": 2.0999757766723635,
      "memory(GiB)": 77.56,
      "step": 76865,
      "token_acc": 0.5613382899628253,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.2933464718735275,
      "grad_norm": 5.385169506072998,
      "learning_rate": 2.609569778781516e-05,
      "loss": 2.2904966354370115,
      "memory(GiB)": 77.56,
      "step": 76870,
      "token_acc": 0.5050505050505051,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.2935606872027763,
      "grad_norm": 5.778371810913086,
      "learning_rate": 2.6089787157171697e-05,
      "loss": 2.4114334106445314,
      "memory(GiB)": 77.56,
      "step": 76875,
      "token_acc": 0.4757834757834758,
      "train_speed(iter/s)": 1.437234
    },
    {
      "epoch": 3.293774902532025,
      "grad_norm": 10.222197532653809,
      "learning_rate": 2.6083876959684162e-05,
      "loss": 2.0714389801025392,
      "memory(GiB)": 77.56,
      "step": 76880,
      "token_acc": 0.5693950177935944,
      "train_speed(iter/s)": 1.437265
    },
    {
      "epoch": 3.2939891178612744,
      "grad_norm": 7.302116870880127,
      "learning_rate": 2.607796719545962e-05,
      "loss": 2.097861480712891,
      "memory(GiB)": 77.56,
      "step": 76885,
      "token_acc": 0.5278810408921933,
      "train_speed(iter/s)": 1.437275
    },
    {
      "epoch": 3.294203333190523,
      "grad_norm": 7.242732524871826,
      "learning_rate": 2.6072057864605147e-05,
      "loss": 2.6891801834106444,
      "memory(GiB)": 77.56,
      "step": 76890,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.437255
    },
    {
      "epoch": 3.294417548519772,
      "grad_norm": 6.562575340270996,
      "learning_rate": 2.606614896722781e-05,
      "loss": 2.539505195617676,
      "memory(GiB)": 77.56,
      "step": 76895,
      "token_acc": 0.46130952380952384,
      "train_speed(iter/s)": 1.437234
    },
    {
      "epoch": 3.2946317638490212,
      "grad_norm": 7.772091865539551,
      "learning_rate": 2.6060240503434623e-05,
      "loss": 2.3576221466064453,
      "memory(GiB)": 77.56,
      "step": 76900,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.437245
    },
    {
      "epoch": 3.29484597917827,
      "grad_norm": 5.5433220863342285,
      "learning_rate": 2.6054332473332622e-05,
      "loss": 2.216729927062988,
      "memory(GiB)": 77.56,
      "step": 76905,
      "token_acc": 0.47491638795986624,
      "train_speed(iter/s)": 1.43725
    },
    {
      "epoch": 3.295060194507519,
      "grad_norm": 4.832674026489258,
      "learning_rate": 2.6048424877028876e-05,
      "loss": 2.3038759231567383,
      "memory(GiB)": 77.56,
      "step": 76910,
      "token_acc": 0.4843205574912892,
      "train_speed(iter/s)": 1.437237
    },
    {
      "epoch": 3.295274409836768,
      "grad_norm": 4.854836463928223,
      "learning_rate": 2.6042517714630354e-05,
      "loss": 2.394809341430664,
      "memory(GiB)": 77.56,
      "step": 76915,
      "token_acc": 0.48307692307692307,
      "train_speed(iter/s)": 1.437231
    },
    {
      "epoch": 3.295488625166017,
      "grad_norm": 8.294893264770508,
      "learning_rate": 2.6036610986244125e-05,
      "loss": 2.3702972412109373,
      "memory(GiB)": 77.56,
      "step": 76920,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.437252
    },
    {
      "epoch": 3.2957028404952657,
      "grad_norm": 4.817264556884766,
      "learning_rate": 2.6030704691977158e-05,
      "loss": 2.360240364074707,
      "memory(GiB)": 77.56,
      "step": 76925,
      "token_acc": 0.5109034267912772,
      "train_speed(iter/s)": 1.437276
    },
    {
      "epoch": 3.295917055824515,
      "grad_norm": 6.1844401359558105,
      "learning_rate": 2.602479883193647e-05,
      "loss": 2.492644500732422,
      "memory(GiB)": 77.56,
      "step": 76930,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.437247
    },
    {
      "epoch": 3.296131271153764,
      "grad_norm": 5.200422763824463,
      "learning_rate": 2.6018893406229033e-05,
      "loss": 1.9987030029296875,
      "memory(GiB)": 77.56,
      "step": 76935,
      "token_acc": 0.5754385964912281,
      "train_speed(iter/s)": 1.437253
    },
    {
      "epoch": 3.2963454864830126,
      "grad_norm": 5.031759738922119,
      "learning_rate": 2.6012988414961848e-05,
      "loss": 2.362395477294922,
      "memory(GiB)": 77.56,
      "step": 76940,
      "token_acc": 0.46710526315789475,
      "train_speed(iter/s)": 1.437253
    },
    {
      "epoch": 3.296559701812262,
      "grad_norm": 6.7006635665893555,
      "learning_rate": 2.6007083858241853e-05,
      "loss": 2.38018856048584,
      "memory(GiB)": 77.56,
      "step": 76945,
      "token_acc": 0.5032467532467533,
      "train_speed(iter/s)": 1.437243
    },
    {
      "epoch": 3.2967739171415107,
      "grad_norm": 6.165157794952393,
      "learning_rate": 2.6001179736176064e-05,
      "loss": 2.5515838623046876,
      "memory(GiB)": 77.56,
      "step": 76950,
      "token_acc": 0.4612903225806452,
      "train_speed(iter/s)": 1.437249
    },
    {
      "epoch": 3.2969881324707595,
      "grad_norm": 4.450499057769775,
      "learning_rate": 2.599527604887141e-05,
      "loss": 2.2928436279296873,
      "memory(GiB)": 77.56,
      "step": 76955,
      "token_acc": 0.5040322580645161,
      "train_speed(iter/s)": 1.437272
    },
    {
      "epoch": 3.2972023478000088,
      "grad_norm": 5.850487232208252,
      "learning_rate": 2.5989372796434854e-05,
      "loss": 2.1894216537475586,
      "memory(GiB)": 77.56,
      "step": 76960,
      "token_acc": 0.510548523206751,
      "train_speed(iter/s)": 1.437269
    },
    {
      "epoch": 3.2974165631292576,
      "grad_norm": 5.848648548126221,
      "learning_rate": 2.598346997897333e-05,
      "loss": 2.1983219146728517,
      "memory(GiB)": 77.56,
      "step": 76965,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.437258
    },
    {
      "epoch": 3.2976307784585064,
      "grad_norm": 9.699432373046875,
      "learning_rate": 2.597756759659376e-05,
      "loss": 2.3296913146972655,
      "memory(GiB)": 77.56,
      "step": 76970,
      "token_acc": 0.5291970802919708,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 3.2978449937877556,
      "grad_norm": 5.465476036071777,
      "learning_rate": 2.597166564940311e-05,
      "loss": 2.33630485534668,
      "memory(GiB)": 77.56,
      "step": 76975,
      "token_acc": 0.5016181229773463,
      "train_speed(iter/s)": 1.437259
    },
    {
      "epoch": 3.2980592091170045,
      "grad_norm": 6.026414394378662,
      "learning_rate": 2.5965764137508253e-05,
      "loss": 2.3223102569580076,
      "memory(GiB)": 77.56,
      "step": 76980,
      "token_acc": 0.5236686390532544,
      "train_speed(iter/s)": 1.437272
    },
    {
      "epoch": 3.2982734244462533,
      "grad_norm": 5.62860107421875,
      "learning_rate": 2.5959863061016144e-05,
      "loss": 2.326262855529785,
      "memory(GiB)": 77.56,
      "step": 76985,
      "token_acc": 0.4609375,
      "train_speed(iter/s)": 1.43726
    },
    {
      "epoch": 3.2984876397755025,
      "grad_norm": 5.432168960571289,
      "learning_rate": 2.5953962420033673e-05,
      "loss": 2.564494514465332,
      "memory(GiB)": 77.56,
      "step": 76990,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437255
    },
    {
      "epoch": 3.2987018551047513,
      "grad_norm": 6.185112476348877,
      "learning_rate": 2.5948062214667723e-05,
      "loss": 2.4724334716796874,
      "memory(GiB)": 77.56,
      "step": 76995,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.43725
    },
    {
      "epoch": 3.298916070434,
      "grad_norm": 5.683099746704102,
      "learning_rate": 2.5942162445025174e-05,
      "loss": 2.2289962768554688,
      "memory(GiB)": 77.56,
      "step": 77000,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 3.298916070434,
      "eval_loss": 2.3029232025146484,
      "eval_runtime": 14.7464,
      "eval_samples_per_second": 6.781,
      "eval_steps_per_second": 6.781,
      "eval_token_acc": 0.4774193548387097,
      "step": 77000
    },
    {
      "epoch": 3.2991302857632494,
      "grad_norm": 6.969137668609619,
      "learning_rate": 2.593626311121294e-05,
      "loss": 2.268109893798828,
      "memory(GiB)": 77.56,
      "step": 77005,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.436817
    },
    {
      "epoch": 3.299344501092498,
      "grad_norm": 5.646996974945068,
      "learning_rate": 2.5930364213337865e-05,
      "loss": 2.5246856689453123,
      "memory(GiB)": 77.56,
      "step": 77010,
      "token_acc": 0.45925925925925926,
      "train_speed(iter/s)": 1.436829
    },
    {
      "epoch": 3.299558716421747,
      "grad_norm": 6.769023418426514,
      "learning_rate": 2.592446575150683e-05,
      "loss": 2.1980318069458007,
      "memory(GiB)": 77.56,
      "step": 77015,
      "token_acc": 0.5271966527196653,
      "train_speed(iter/s)": 1.436815
    },
    {
      "epoch": 3.2997729317509963,
      "grad_norm": 6.5589823722839355,
      "learning_rate": 2.5918567725826682e-05,
      "loss": 2.33660888671875,
      "memory(GiB)": 77.56,
      "step": 77020,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436819
    },
    {
      "epoch": 3.299987147080245,
      "grad_norm": 6.002827167510986,
      "learning_rate": 2.591267013640427e-05,
      "loss": 2.4459890365600585,
      "memory(GiB)": 77.56,
      "step": 77025,
      "token_acc": 0.4629080118694362,
      "train_speed(iter/s)": 1.436822
    },
    {
      "epoch": 3.300201362409494,
      "grad_norm": 5.471892833709717,
      "learning_rate": 2.590677298334641e-05,
      "loss": 2.5731468200683594,
      "memory(GiB)": 77.56,
      "step": 77030,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.436841
    },
    {
      "epoch": 3.300415577738743,
      "grad_norm": 5.689818382263184,
      "learning_rate": 2.590087626675998e-05,
      "loss": 2.279067039489746,
      "memory(GiB)": 77.56,
      "step": 77035,
      "token_acc": 0.5487364620938628,
      "train_speed(iter/s)": 1.436846
    },
    {
      "epoch": 3.300629793067992,
      "grad_norm": 6.6045660972595215,
      "learning_rate": 2.589497998675179e-05,
      "loss": 2.3414470672607424,
      "memory(GiB)": 77.56,
      "step": 77040,
      "token_acc": 0.525,
      "train_speed(iter/s)": 1.436827
    },
    {
      "epoch": 3.300844008397241,
      "grad_norm": 4.153785705566406,
      "learning_rate": 2.5889084143428643e-05,
      "loss": 2.5267446517944334,
      "memory(GiB)": 77.56,
      "step": 77045,
      "token_acc": 0.5109034267912772,
      "train_speed(iter/s)": 1.436834
    },
    {
      "epoch": 3.30105822372649,
      "grad_norm": 5.836066722869873,
      "learning_rate": 2.5883188736897356e-05,
      "loss": 2.373625946044922,
      "memory(GiB)": 77.56,
      "step": 77050,
      "token_acc": 0.5080906148867314,
      "train_speed(iter/s)": 1.43681
    },
    {
      "epoch": 3.301272439055739,
      "grad_norm": 4.907026290893555,
      "learning_rate": 2.587729376726471e-05,
      "loss": 2.3365612030029297,
      "memory(GiB)": 77.56,
      "step": 77055,
      "token_acc": 0.4984423676012461,
      "train_speed(iter/s)": 1.436809
    },
    {
      "epoch": 3.3014866543849877,
      "grad_norm": 6.26133918762207,
      "learning_rate": 2.587139923463751e-05,
      "loss": 2.819825553894043,
      "memory(GiB)": 77.56,
      "step": 77060,
      "token_acc": 0.42803030303030304,
      "train_speed(iter/s)": 1.436825
    },
    {
      "epoch": 3.301700869714237,
      "grad_norm": 4.15021276473999,
      "learning_rate": 2.586550513912257e-05,
      "loss": 2.058642578125,
      "memory(GiB)": 77.56,
      "step": 77065,
      "token_acc": 0.5355805243445693,
      "train_speed(iter/s)": 1.436842
    },
    {
      "epoch": 3.3019150850434857,
      "grad_norm": 9.351841926574707,
      "learning_rate": 2.585961148082665e-05,
      "loss": 2.455078125,
      "memory(GiB)": 77.56,
      "step": 77070,
      "token_acc": 0.49375,
      "train_speed(iter/s)": 1.436862
    },
    {
      "epoch": 3.3021293003727346,
      "grad_norm": 6.883096218109131,
      "learning_rate": 2.5853718259856507e-05,
      "loss": 2.4500152587890627,
      "memory(GiB)": 77.56,
      "step": 77075,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.43686
    },
    {
      "epoch": 3.302343515701984,
      "grad_norm": 5.932377338409424,
      "learning_rate": 2.584782547631891e-05,
      "loss": 2.5991334915161133,
      "memory(GiB)": 77.56,
      "step": 77080,
      "token_acc": 0.5101214574898786,
      "train_speed(iter/s)": 1.436858
    },
    {
      "epoch": 3.3025577310312326,
      "grad_norm": 4.293975353240967,
      "learning_rate": 2.5841933130320618e-05,
      "loss": 2.4596836090087892,
      "memory(GiB)": 77.56,
      "step": 77085,
      "token_acc": 0.4937106918238994,
      "train_speed(iter/s)": 1.436856
    },
    {
      "epoch": 3.3027719463604814,
      "grad_norm": 5.631494522094727,
      "learning_rate": 2.5836041221968345e-05,
      "loss": 2.636617660522461,
      "memory(GiB)": 77.56,
      "step": 77090,
      "token_acc": 0.458041958041958,
      "train_speed(iter/s)": 1.43687
    },
    {
      "epoch": 3.3029861616897307,
      "grad_norm": 5.196710109710693,
      "learning_rate": 2.583014975136887e-05,
      "loss": 2.5874141693115233,
      "memory(GiB)": 77.56,
      "step": 77095,
      "token_acc": 0.47674418604651164,
      "train_speed(iter/s)": 1.436898
    },
    {
      "epoch": 3.3032003770189795,
      "grad_norm": 5.8815178871154785,
      "learning_rate": 2.5824258718628906e-05,
      "loss": 2.448269844055176,
      "memory(GiB)": 77.56,
      "step": 77100,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.436913
    },
    {
      "epoch": 3.3034145923482283,
      "grad_norm": 6.350180149078369,
      "learning_rate": 2.5818368123855176e-05,
      "loss": 2.6670684814453125,
      "memory(GiB)": 77.56,
      "step": 77105,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.436919
    },
    {
      "epoch": 3.3036288076774776,
      "grad_norm": 7.065086841583252,
      "learning_rate": 2.581247796715439e-05,
      "loss": 2.268832206726074,
      "memory(GiB)": 77.56,
      "step": 77110,
      "token_acc": 0.5298804780876494,
      "train_speed(iter/s)": 1.436943
    },
    {
      "epoch": 3.3038430230067264,
      "grad_norm": 5.911028861999512,
      "learning_rate": 2.5806588248633235e-05,
      "loss": 2.5288286209106445,
      "memory(GiB)": 77.56,
      "step": 77115,
      "token_acc": 0.45544554455445546,
      "train_speed(iter/s)": 1.436916
    },
    {
      "epoch": 3.304057238335975,
      "grad_norm": 6.289820194244385,
      "learning_rate": 2.580069896839845e-05,
      "loss": 2.052786445617676,
      "memory(GiB)": 77.56,
      "step": 77120,
      "token_acc": 0.5426621160409556,
      "train_speed(iter/s)": 1.436925
    },
    {
      "epoch": 3.3042714536652245,
      "grad_norm": 5.846970081329346,
      "learning_rate": 2.5794810126556707e-05,
      "loss": 2.2656833648681642,
      "memory(GiB)": 77.56,
      "step": 77125,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.436946
    },
    {
      "epoch": 3.3044856689944733,
      "grad_norm": 6.609343528747559,
      "learning_rate": 2.5788921723214664e-05,
      "loss": 2.009074401855469,
      "memory(GiB)": 77.56,
      "step": 77130,
      "token_acc": 0.5833333333333334,
      "train_speed(iter/s)": 1.436959
    },
    {
      "epoch": 3.304699884323722,
      "grad_norm": 5.0762763023376465,
      "learning_rate": 2.5783033758479035e-05,
      "loss": 2.3161203384399416,
      "memory(GiB)": 77.56,
      "step": 77135,
      "token_acc": 0.5326797385620915,
      "train_speed(iter/s)": 1.436975
    },
    {
      "epoch": 3.3049140996529713,
      "grad_norm": 6.89382266998291,
      "learning_rate": 2.5777146232456463e-05,
      "loss": 2.6329727172851562,
      "memory(GiB)": 77.56,
      "step": 77140,
      "token_acc": 0.4432624113475177,
      "train_speed(iter/s)": 1.436993
    },
    {
      "epoch": 3.30512831498222,
      "grad_norm": 7.186066150665283,
      "learning_rate": 2.577125914525359e-05,
      "loss": 2.2719356536865236,
      "memory(GiB)": 77.56,
      "step": 77145,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.436999
    },
    {
      "epoch": 3.305342530311469,
      "grad_norm": 11.30805778503418,
      "learning_rate": 2.5765372496977113e-05,
      "loss": 2.458699417114258,
      "memory(GiB)": 77.56,
      "step": 77150,
      "token_acc": 0.4681818181818182,
      "train_speed(iter/s)": 1.436996
    },
    {
      "epoch": 3.305556745640718,
      "grad_norm": 6.7140021324157715,
      "learning_rate": 2.575948628773364e-05,
      "loss": 2.263802719116211,
      "memory(GiB)": 77.56,
      "step": 77155,
      "token_acc": 0.5460992907801419,
      "train_speed(iter/s)": 1.436986
    },
    {
      "epoch": 3.305770960969967,
      "grad_norm": 6.221335411071777,
      "learning_rate": 2.5753600517629817e-05,
      "loss": 2.507166290283203,
      "memory(GiB)": 77.56,
      "step": 77160,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.436963
    },
    {
      "epoch": 3.305985176299216,
      "grad_norm": 4.820887565612793,
      "learning_rate": 2.5747715186772264e-05,
      "loss": 2.0825527191162108,
      "memory(GiB)": 77.56,
      "step": 77165,
      "token_acc": 0.5746031746031746,
      "train_speed(iter/s)": 1.436942
    },
    {
      "epoch": 3.306199391628465,
      "grad_norm": 7.407349109649658,
      "learning_rate": 2.5741830295267598e-05,
      "loss": 2.262668991088867,
      "memory(GiB)": 77.56,
      "step": 77170,
      "token_acc": 0.5163636363636364,
      "train_speed(iter/s)": 1.436935
    },
    {
      "epoch": 3.306413606957714,
      "grad_norm": 4.921755790710449,
      "learning_rate": 2.573594584322242e-05,
      "loss": 2.761060905456543,
      "memory(GiB)": 77.56,
      "step": 77175,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.436935
    },
    {
      "epoch": 3.3066278222869627,
      "grad_norm": 5.5492072105407715,
      "learning_rate": 2.5730061830743358e-05,
      "loss": 2.0439180374145507,
      "memory(GiB)": 77.56,
      "step": 77180,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.436943
    },
    {
      "epoch": 3.306842037616212,
      "grad_norm": 6.636061668395996,
      "learning_rate": 2.5724178257936992e-05,
      "loss": 2.2904075622558593,
      "memory(GiB)": 77.56,
      "step": 77185,
      "token_acc": 0.4861878453038674,
      "train_speed(iter/s)": 1.43694
    },
    {
      "epoch": 3.307056252945461,
      "grad_norm": 5.415263652801514,
      "learning_rate": 2.5718295124909913e-05,
      "loss": 2.2275161743164062,
      "memory(GiB)": 77.56,
      "step": 77190,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.436946
    },
    {
      "epoch": 3.3072704682747096,
      "grad_norm": 4.004490852355957,
      "learning_rate": 2.5712412431768696e-05,
      "loss": 2.1486921310424805,
      "memory(GiB)": 77.56,
      "step": 77195,
      "token_acc": 0.5427631578947368,
      "train_speed(iter/s)": 1.436963
    },
    {
      "epoch": 3.307484683603959,
      "grad_norm": 5.417473316192627,
      "learning_rate": 2.5706530178619893e-05,
      "loss": 2.082661819458008,
      "memory(GiB)": 77.56,
      "step": 77200,
      "token_acc": 0.5487364620938628,
      "train_speed(iter/s)": 1.436977
    },
    {
      "epoch": 3.3076988989332077,
      "grad_norm": 7.476407527923584,
      "learning_rate": 2.570064836557008e-05,
      "loss": 2.4236865997314454,
      "memory(GiB)": 77.56,
      "step": 77205,
      "token_acc": 0.498371335504886,
      "train_speed(iter/s)": 1.436991
    },
    {
      "epoch": 3.3079131142624565,
      "grad_norm": 6.389898300170898,
      "learning_rate": 2.5694766992725837e-05,
      "loss": 2.0084529876708985,
      "memory(GiB)": 77.56,
      "step": 77210,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436987
    },
    {
      "epoch": 3.3081273295917057,
      "grad_norm": 4.9215240478515625,
      "learning_rate": 2.5688886060193694e-05,
      "loss": 2.378492546081543,
      "memory(GiB)": 77.56,
      "step": 77215,
      "token_acc": 0.4774193548387097,
      "train_speed(iter/s)": 1.436993
    },
    {
      "epoch": 3.3083415449209546,
      "grad_norm": 4.335348129272461,
      "learning_rate": 2.5683005568080188e-05,
      "loss": 2.1754829406738283,
      "memory(GiB)": 77.56,
      "step": 77220,
      "token_acc": 0.5134099616858238,
      "train_speed(iter/s)": 1.437007
    },
    {
      "epoch": 3.3085557602502034,
      "grad_norm": 5.2346367835998535,
      "learning_rate": 2.567712551649184e-05,
      "loss": 2.379818916320801,
      "memory(GiB)": 77.56,
      "step": 77225,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.436985
    },
    {
      "epoch": 3.3087699755794526,
      "grad_norm": 4.546489238739014,
      "learning_rate": 2.567124590553518e-05,
      "loss": 2.0582487106323244,
      "memory(GiB)": 77.56,
      "step": 77230,
      "token_acc": 0.5415162454873647,
      "train_speed(iter/s)": 1.436993
    },
    {
      "epoch": 3.3089841909087014,
      "grad_norm": 5.561270236968994,
      "learning_rate": 2.5665366735316708e-05,
      "loss": 2.371104431152344,
      "memory(GiB)": 77.56,
      "step": 77235,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437008
    },
    {
      "epoch": 3.3091984062379503,
      "grad_norm": 6.500630855560303,
      "learning_rate": 2.565948800594296e-05,
      "loss": 2.3415708541870117,
      "memory(GiB)": 77.56,
      "step": 77240,
      "token_acc": 0.532319391634981,
      "train_speed(iter/s)": 1.437014
    },
    {
      "epoch": 3.3094126215671995,
      "grad_norm": 6.098557472229004,
      "learning_rate": 2.5653609717520423e-05,
      "loss": 2.2728961944580077,
      "memory(GiB)": 77.56,
      "step": 77245,
      "token_acc": 0.5381944444444444,
      "train_speed(iter/s)": 1.437016
    },
    {
      "epoch": 3.3096268368964483,
      "grad_norm": 5.145260334014893,
      "learning_rate": 2.5647731870155585e-05,
      "loss": 1.9735578536987304,
      "memory(GiB)": 77.56,
      "step": 77250,
      "token_acc": 0.5867768595041323,
      "train_speed(iter/s)": 1.437008
    },
    {
      "epoch": 3.309841052225697,
      "grad_norm": 5.626273155212402,
      "learning_rate": 2.5641854463954928e-05,
      "loss": 2.3705728530883787,
      "memory(GiB)": 77.56,
      "step": 77255,
      "token_acc": 0.5084175084175084,
      "train_speed(iter/s)": 1.437015
    },
    {
      "epoch": 3.3100552675549464,
      "grad_norm": 6.133059978485107,
      "learning_rate": 2.563597749902491e-05,
      "loss": 2.1151107788085937,
      "memory(GiB)": 77.56,
      "step": 77260,
      "token_acc": 0.5526315789473685,
      "train_speed(iter/s)": 1.437034
    },
    {
      "epoch": 3.310269482884195,
      "grad_norm": 5.907609939575195,
      "learning_rate": 2.5630100975472026e-05,
      "loss": 2.417312240600586,
      "memory(GiB)": 77.56,
      "step": 77265,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43702
    },
    {
      "epoch": 3.310483698213444,
      "grad_norm": 6.832525730133057,
      "learning_rate": 2.5624224893402733e-05,
      "loss": 2.3832204818725584,
      "memory(GiB)": 77.56,
      "step": 77270,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437025
    },
    {
      "epoch": 3.3106979135426933,
      "grad_norm": 5.4986090660095215,
      "learning_rate": 2.5618349252923448e-05,
      "loss": 2.671581840515137,
      "memory(GiB)": 77.56,
      "step": 77275,
      "token_acc": 0.4707692307692308,
      "train_speed(iter/s)": 1.437032
    },
    {
      "epoch": 3.310912128871942,
      "grad_norm": 6.092564105987549,
      "learning_rate": 2.5612474054140657e-05,
      "loss": 2.3200450897216798,
      "memory(GiB)": 77.56,
      "step": 77280,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.437005
    },
    {
      "epoch": 3.311126344201191,
      "grad_norm": 5.981649398803711,
      "learning_rate": 2.560659929716078e-05,
      "loss": 2.4917009353637694,
      "memory(GiB)": 77.56,
      "step": 77285,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437011
    },
    {
      "epoch": 3.31134055953044,
      "grad_norm": 4.974832534790039,
      "learning_rate": 2.560072498209022e-05,
      "loss": 2.4469219207763673,
      "memory(GiB)": 77.56,
      "step": 77290,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.437044
    },
    {
      "epoch": 3.311554774859689,
      "grad_norm": 5.801551342010498,
      "learning_rate": 2.5594851109035435e-05,
      "loss": 2.112714958190918,
      "memory(GiB)": 77.56,
      "step": 77295,
      "token_acc": 0.5286624203821656,
      "train_speed(iter/s)": 1.437063
    },
    {
      "epoch": 3.3117689901889378,
      "grad_norm": 5.231300354003906,
      "learning_rate": 2.558897767810281e-05,
      "loss": 2.416782188415527,
      "memory(GiB)": 77.56,
      "step": 77300,
      "token_acc": 0.4854014598540146,
      "train_speed(iter/s)": 1.437078
    },
    {
      "epoch": 3.311983205518187,
      "grad_norm": 7.499194145202637,
      "learning_rate": 2.5583104689398757e-05,
      "loss": 1.9685455322265626,
      "memory(GiB)": 77.56,
      "step": 77305,
      "token_acc": 0.48917748917748916,
      "train_speed(iter/s)": 1.437078
    },
    {
      "epoch": 3.312197420847436,
      "grad_norm": 7.441330432891846,
      "learning_rate": 2.5577232143029672e-05,
      "loss": 2.247381019592285,
      "memory(GiB)": 77.56,
      "step": 77310,
      "token_acc": 0.47651006711409394,
      "train_speed(iter/s)": 1.437092
    },
    {
      "epoch": 3.3124116361766847,
      "grad_norm": 6.9077301025390625,
      "learning_rate": 2.5571360039101932e-05,
      "loss": 2.178914260864258,
      "memory(GiB)": 77.56,
      "step": 77315,
      "token_acc": 0.5572519083969466,
      "train_speed(iter/s)": 1.437113
    },
    {
      "epoch": 3.312625851505934,
      "grad_norm": 7.821229934692383,
      "learning_rate": 2.5565488377721903e-05,
      "loss": 2.3870431900024416,
      "memory(GiB)": 77.56,
      "step": 77320,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.437125
    },
    {
      "epoch": 3.3128400668351827,
      "grad_norm": 4.951988697052002,
      "learning_rate": 2.555961715899599e-05,
      "loss": 2.2436527252197265,
      "memory(GiB)": 77.56,
      "step": 77325,
      "token_acc": 0.5389830508474577,
      "train_speed(iter/s)": 1.437133
    },
    {
      "epoch": 3.3130542821644315,
      "grad_norm": 5.681581020355225,
      "learning_rate": 2.555374638303054e-05,
      "loss": 2.310961151123047,
      "memory(GiB)": 77.56,
      "step": 77330,
      "token_acc": 0.5448028673835126,
      "train_speed(iter/s)": 1.437127
    },
    {
      "epoch": 3.313268497493681,
      "grad_norm": 6.0256547927856445,
      "learning_rate": 2.554787604993191e-05,
      "loss": 2.1599552154541017,
      "memory(GiB)": 77.56,
      "step": 77335,
      "token_acc": 0.5436241610738255,
      "train_speed(iter/s)": 1.437137
    },
    {
      "epoch": 3.3134827128229296,
      "grad_norm": 6.890291690826416,
      "learning_rate": 2.5542006159806444e-05,
      "loss": 2.521312713623047,
      "memory(GiB)": 77.56,
      "step": 77340,
      "token_acc": 0.4910394265232975,
      "train_speed(iter/s)": 1.437146
    },
    {
      "epoch": 3.3136969281521784,
      "grad_norm": 6.326132774353027,
      "learning_rate": 2.5536136712760457e-05,
      "loss": 2.4545194625854494,
      "memory(GiB)": 77.56,
      "step": 77345,
      "token_acc": 0.46557377049180326,
      "train_speed(iter/s)": 1.437116
    },
    {
      "epoch": 3.3139111434814277,
      "grad_norm": 8.4386625289917,
      "learning_rate": 2.5530267708900303e-05,
      "loss": 2.2343910217285154,
      "memory(GiB)": 77.56,
      "step": 77350,
      "token_acc": 0.5139318885448917,
      "train_speed(iter/s)": 1.437134
    },
    {
      "epoch": 3.3141253588106765,
      "grad_norm": 6.033736705780029,
      "learning_rate": 2.5524399148332325e-05,
      "loss": 2.4118396759033205,
      "memory(GiB)": 77.56,
      "step": 77355,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.437147
    },
    {
      "epoch": 3.3143395741399253,
      "grad_norm": 5.393014907836914,
      "learning_rate": 2.551853103116281e-05,
      "loss": 2.4604232788085936,
      "memory(GiB)": 77.56,
      "step": 77360,
      "token_acc": 0.48580441640378547,
      "train_speed(iter/s)": 1.437167
    },
    {
      "epoch": 3.3145537894691746,
      "grad_norm": 7.311291217803955,
      "learning_rate": 2.5512663357498078e-05,
      "loss": 2.2386837005615234,
      "memory(GiB)": 77.56,
      "step": 77365,
      "token_acc": 0.5298804780876494,
      "train_speed(iter/s)": 1.437162
    },
    {
      "epoch": 3.3147680047984234,
      "grad_norm": 6.575873374938965,
      "learning_rate": 2.550679612744442e-05,
      "loss": 2.5401411056518555,
      "memory(GiB)": 77.56,
      "step": 77370,
      "token_acc": 0.4270833333333333,
      "train_speed(iter/s)": 1.437193
    },
    {
      "epoch": 3.314982220127672,
      "grad_norm": 6.860659122467041,
      "learning_rate": 2.5500929341108104e-05,
      "loss": 2.5129186630249025,
      "memory(GiB)": 77.56,
      "step": 77375,
      "token_acc": 0.47808764940239046,
      "train_speed(iter/s)": 1.437208
    },
    {
      "epoch": 3.3151964354569214,
      "grad_norm": 6.716062545776367,
      "learning_rate": 2.5495062998595455e-05,
      "loss": 2.372308349609375,
      "memory(GiB)": 77.56,
      "step": 77380,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437222
    },
    {
      "epoch": 3.3154106507861703,
      "grad_norm": 5.149068355560303,
      "learning_rate": 2.548919710001273e-05,
      "loss": 2.5191085815429686,
      "memory(GiB)": 77.56,
      "step": 77385,
      "token_acc": 0.42524916943521596,
      "train_speed(iter/s)": 1.437207
    },
    {
      "epoch": 3.315624866115419,
      "grad_norm": 5.713041305541992,
      "learning_rate": 2.5483331645466192e-05,
      "loss": 2.181832504272461,
      "memory(GiB)": 77.56,
      "step": 77390,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.437197
    },
    {
      "epoch": 3.3158390814446683,
      "grad_norm": 5.911479473114014,
      "learning_rate": 2.5477466635062097e-05,
      "loss": 2.1378547668457033,
      "memory(GiB)": 77.56,
      "step": 77395,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.437203
    },
    {
      "epoch": 3.316053296773917,
      "grad_norm": 5.920670509338379,
      "learning_rate": 2.5471602068906697e-05,
      "loss": 2.1676309585571287,
      "memory(GiB)": 77.56,
      "step": 77400,
      "token_acc": 0.5368421052631579,
      "train_speed(iter/s)": 1.437241
    },
    {
      "epoch": 3.316267512103166,
      "grad_norm": 5.166963577270508,
      "learning_rate": 2.5465737947106218e-05,
      "loss": 2.1913219451904298,
      "memory(GiB)": 77.56,
      "step": 77405,
      "token_acc": 0.5324675324675324,
      "train_speed(iter/s)": 1.437257
    },
    {
      "epoch": 3.316481727432415,
      "grad_norm": 6.212830066680908,
      "learning_rate": 2.545987426976693e-05,
      "loss": 2.2177059173583986,
      "memory(GiB)": 77.56,
      "step": 77410,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 3.316695942761664,
      "grad_norm": 6.647331237792969,
      "learning_rate": 2.545401103699504e-05,
      "loss": 2.5991729736328124,
      "memory(GiB)": 77.56,
      "step": 77415,
      "token_acc": 0.48253968253968255,
      "train_speed(iter/s)": 1.437275
    },
    {
      "epoch": 3.316910158090913,
      "grad_norm": 4.814138412475586,
      "learning_rate": 2.5448148248896768e-05,
      "loss": 2.4958990097045897,
      "memory(GiB)": 77.56,
      "step": 77420,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.437267
    },
    {
      "epoch": 3.317124373420162,
      "grad_norm": 4.491227149963379,
      "learning_rate": 2.54422859055783e-05,
      "loss": 2.555317687988281,
      "memory(GiB)": 77.56,
      "step": 77425,
      "token_acc": 0.4735202492211838,
      "train_speed(iter/s)": 1.437264
    },
    {
      "epoch": 3.317338588749411,
      "grad_norm": 6.427128314971924,
      "learning_rate": 2.543642400714588e-05,
      "loss": 2.1997188568115233,
      "memory(GiB)": 77.56,
      "step": 77430,
      "token_acc": 0.5328947368421053,
      "train_speed(iter/s)": 1.43725
    },
    {
      "epoch": 3.3175528040786597,
      "grad_norm": 5.459718227386475,
      "learning_rate": 2.543056255370566e-05,
      "loss": 2.3429189682006837,
      "memory(GiB)": 77.56,
      "step": 77435,
      "token_acc": 0.534965034965035,
      "train_speed(iter/s)": 1.437232
    },
    {
      "epoch": 3.317767019407909,
      "grad_norm": 7.484043121337891,
      "learning_rate": 2.542470154536387e-05,
      "loss": 2.47094669342041,
      "memory(GiB)": 77.56,
      "step": 77440,
      "token_acc": 0.5079872204472844,
      "train_speed(iter/s)": 1.437239
    },
    {
      "epoch": 3.317981234737158,
      "grad_norm": 6.093674182891846,
      "learning_rate": 2.5418840982226667e-05,
      "loss": 2.3110321044921873,
      "memory(GiB)": 77.56,
      "step": 77445,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.437235
    },
    {
      "epoch": 3.3181954500664066,
      "grad_norm": 9.87459945678711,
      "learning_rate": 2.5412980864400217e-05,
      "loss": 2.4254985809326173,
      "memory(GiB)": 77.56,
      "step": 77450,
      "token_acc": 0.5037037037037037,
      "train_speed(iter/s)": 1.437243
    },
    {
      "epoch": 3.318409665395656,
      "grad_norm": 5.21718168258667,
      "learning_rate": 2.540712119199068e-05,
      "loss": 2.5110260009765626,
      "memory(GiB)": 77.56,
      "step": 77455,
      "token_acc": 0.5271565495207667,
      "train_speed(iter/s)": 1.437252
    },
    {
      "epoch": 3.3186238807249047,
      "grad_norm": 5.1384687423706055,
      "learning_rate": 2.5401261965104217e-05,
      "loss": 2.2521526336669924,
      "memory(GiB)": 77.56,
      "step": 77460,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 3.3188380960541535,
      "grad_norm": 6.13267183303833,
      "learning_rate": 2.5395403183846945e-05,
      "loss": 2.136837959289551,
      "memory(GiB)": 77.56,
      "step": 77465,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437278
    },
    {
      "epoch": 3.3190523113834027,
      "grad_norm": 5.374451160430908,
      "learning_rate": 2.5389544848325054e-05,
      "loss": 2.4706241607666017,
      "memory(GiB)": 77.56,
      "step": 77470,
      "token_acc": 0.4754601226993865,
      "train_speed(iter/s)": 1.437264
    },
    {
      "epoch": 3.3192665267126515,
      "grad_norm": 6.128334999084473,
      "learning_rate": 2.5383686958644636e-05,
      "loss": 2.2426692962646486,
      "memory(GiB)": 77.56,
      "step": 77475,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.437268
    },
    {
      "epoch": 3.3194807420419004,
      "grad_norm": 7.791270732879639,
      "learning_rate": 2.5377829514911822e-05,
      "loss": 2.4220884323120115,
      "memory(GiB)": 77.56,
      "step": 77480,
      "token_acc": 0.4542124542124542,
      "train_speed(iter/s)": 1.437277
    },
    {
      "epoch": 3.3196949573711496,
      "grad_norm": 4.998856067657471,
      "learning_rate": 2.537197251723272e-05,
      "loss": 2.290847587585449,
      "memory(GiB)": 77.56,
      "step": 77485,
      "token_acc": 0.5526315789473685,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 3.3199091727003984,
      "grad_norm": 5.187781810760498,
      "learning_rate": 2.5366115965713422e-05,
      "loss": 1.7143903732299806,
      "memory(GiB)": 77.56,
      "step": 77490,
      "token_acc": 0.6065573770491803,
      "train_speed(iter/s)": 1.43728
    },
    {
      "epoch": 3.3201233880296472,
      "grad_norm": 5.368714809417725,
      "learning_rate": 2.5360259860460066e-05,
      "loss": 2.231265640258789,
      "memory(GiB)": 77.56,
      "step": 77495,
      "token_acc": 0.5017064846416383,
      "train_speed(iter/s)": 1.437288
    },
    {
      "epoch": 3.3203376033588965,
      "grad_norm": 5.128393650054932,
      "learning_rate": 2.5354404201578696e-05,
      "loss": 2.5056447982788086,
      "memory(GiB)": 77.56,
      "step": 77500,
      "token_acc": 0.4984894259818731,
      "train_speed(iter/s)": 1.437299
    },
    {
      "epoch": 3.3203376033588965,
      "eval_loss": 2.105395555496216,
      "eval_runtime": 14.2759,
      "eval_samples_per_second": 7.005,
      "eval_steps_per_second": 7.005,
      "eval_token_acc": 0.4832869080779944,
      "step": 77500
    },
    {
      "epoch": 3.3205518186881453,
      "grad_norm": 4.861273765563965,
      "learning_rate": 2.5348548989175424e-05,
      "loss": 2.174689292907715,
      "memory(GiB)": 77.56,
      "step": 77505,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.436912
    },
    {
      "epoch": 3.320766034017394,
      "grad_norm": 5.340626239776611,
      "learning_rate": 2.534269422335632e-05,
      "loss": 2.338021659851074,
      "memory(GiB)": 77.56,
      "step": 77510,
      "token_acc": 0.49038461538461536,
      "train_speed(iter/s)": 1.43693
    },
    {
      "epoch": 3.3209802493466434,
      "grad_norm": 5.060118675231934,
      "learning_rate": 2.533683990422744e-05,
      "loss": 2.5441051483154298,
      "memory(GiB)": 77.56,
      "step": 77515,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436953
    },
    {
      "epoch": 3.321194464675892,
      "grad_norm": 5.681626796722412,
      "learning_rate": 2.5330986031894823e-05,
      "loss": 2.472270202636719,
      "memory(GiB)": 77.56,
      "step": 77520,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.436968
    },
    {
      "epoch": 3.321408680005141,
      "grad_norm": 5.349667549133301,
      "learning_rate": 2.5325132606464552e-05,
      "loss": 2.3636363983154296,
      "memory(GiB)": 77.56,
      "step": 77525,
      "token_acc": 0.48773006134969327,
      "train_speed(iter/s)": 1.436969
    },
    {
      "epoch": 3.3216228953343903,
      "grad_norm": 6.741828441619873,
      "learning_rate": 2.5319279628042647e-05,
      "loss": 2.293136978149414,
      "memory(GiB)": 77.56,
      "step": 77530,
      "token_acc": 0.4796511627906977,
      "train_speed(iter/s)": 1.436963
    },
    {
      "epoch": 3.321837110663639,
      "grad_norm": 5.243693828582764,
      "learning_rate": 2.5313427096735155e-05,
      "loss": 2.487224578857422,
      "memory(GiB)": 77.56,
      "step": 77535,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.436966
    },
    {
      "epoch": 3.322051325992888,
      "grad_norm": 6.401910781860352,
      "learning_rate": 2.530757501264808e-05,
      "loss": 2.330003356933594,
      "memory(GiB)": 77.56,
      "step": 77540,
      "token_acc": 0.4846153846153846,
      "train_speed(iter/s)": 1.43695
    },
    {
      "epoch": 3.322265541322137,
      "grad_norm": 5.999273300170898,
      "learning_rate": 2.5301723375887447e-05,
      "loss": 2.492059326171875,
      "memory(GiB)": 77.56,
      "step": 77545,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.436955
    },
    {
      "epoch": 3.322479756651386,
      "grad_norm": 6.247038841247559,
      "learning_rate": 2.5295872186559243e-05,
      "loss": 2.4057302474975586,
      "memory(GiB)": 77.56,
      "step": 77550,
      "token_acc": 0.4981549815498155,
      "train_speed(iter/s)": 1.43697
    },
    {
      "epoch": 3.3226939719806348,
      "grad_norm": 5.043736457824707,
      "learning_rate": 2.5290021444769517e-05,
      "loss": 1.9194206237792968,
      "memory(GiB)": 77.56,
      "step": 77555,
      "token_acc": 0.5296052631578947,
      "train_speed(iter/s)": 1.436985
    },
    {
      "epoch": 3.322908187309884,
      "grad_norm": 5.577779769897461,
      "learning_rate": 2.5284171150624225e-05,
      "loss": 2.409530448913574,
      "memory(GiB)": 77.56,
      "step": 77560,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.436998
    },
    {
      "epoch": 3.323122402639133,
      "grad_norm": 5.46181058883667,
      "learning_rate": 2.5278321304229358e-05,
      "loss": 2.1508968353271483,
      "memory(GiB)": 77.56,
      "step": 77565,
      "token_acc": 0.574468085106383,
      "train_speed(iter/s)": 1.436994
    },
    {
      "epoch": 3.3233366179683816,
      "grad_norm": 5.481894493103027,
      "learning_rate": 2.5272471905690876e-05,
      "loss": 2.2374034881591798,
      "memory(GiB)": 77.56,
      "step": 77570,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.436995
    },
    {
      "epoch": 3.323550833297631,
      "grad_norm": 4.195674419403076,
      "learning_rate": 2.526662295511478e-05,
      "loss": 2.3077556610107424,
      "memory(GiB)": 77.56,
      "step": 77575,
      "token_acc": 0.524822695035461,
      "train_speed(iter/s)": 1.437004
    },
    {
      "epoch": 3.3237650486268797,
      "grad_norm": 6.152205944061279,
      "learning_rate": 2.5260774452606993e-05,
      "loss": 2.2600879669189453,
      "memory(GiB)": 77.56,
      "step": 77580,
      "token_acc": 0.5483870967741935,
      "train_speed(iter/s)": 1.436959
    },
    {
      "epoch": 3.3239792639561285,
      "grad_norm": 5.229650020599365,
      "learning_rate": 2.5254926398273498e-05,
      "loss": 2.219290924072266,
      "memory(GiB)": 77.56,
      "step": 77585,
      "token_acc": 0.532051282051282,
      "train_speed(iter/s)": 1.436951
    },
    {
      "epoch": 3.324193479285378,
      "grad_norm": 5.283496379852295,
      "learning_rate": 2.5249078792220227e-05,
      "loss": 2.4240015029907225,
      "memory(GiB)": 77.56,
      "step": 77590,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.436958
    },
    {
      "epoch": 3.3244076946146266,
      "grad_norm": 5.911551475524902,
      "learning_rate": 2.524323163455311e-05,
      "loss": 2.318562126159668,
      "memory(GiB)": 77.56,
      "step": 77595,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.436949
    },
    {
      "epoch": 3.3246219099438754,
      "grad_norm": 6.566239356994629,
      "learning_rate": 2.5237384925378084e-05,
      "loss": 2.473470115661621,
      "memory(GiB)": 77.56,
      "step": 77600,
      "token_acc": 0.46204620462046203,
      "train_speed(iter/s)": 1.436977
    },
    {
      "epoch": 3.3248361252731247,
      "grad_norm": 5.843847751617432,
      "learning_rate": 2.5231538664801047e-05,
      "loss": 2.5067211151123048,
      "memory(GiB)": 77.56,
      "step": 77605,
      "token_acc": 0.47962382445141066,
      "train_speed(iter/s)": 1.436984
    },
    {
      "epoch": 3.3250503406023735,
      "grad_norm": 6.355237007141113,
      "learning_rate": 2.5225692852927913e-05,
      "loss": 2.354112243652344,
      "memory(GiB)": 77.56,
      "step": 77610,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.436993
    },
    {
      "epoch": 3.3252645559316223,
      "grad_norm": 6.199345588684082,
      "learning_rate": 2.5219847489864606e-05,
      "loss": 2.480083465576172,
      "memory(GiB)": 77.56,
      "step": 77615,
      "token_acc": 0.501628664495114,
      "train_speed(iter/s)": 1.437016
    },
    {
      "epoch": 3.3254787712608715,
      "grad_norm": 6.4629926681518555,
      "learning_rate": 2.521400257571701e-05,
      "loss": 2.1230697631835938,
      "memory(GiB)": 77.56,
      "step": 77620,
      "token_acc": 0.5266903914590747,
      "train_speed(iter/s)": 1.437017
    },
    {
      "epoch": 3.3256929865901204,
      "grad_norm": 6.372697830200195,
      "learning_rate": 2.5208158110591006e-05,
      "loss": 2.2228771209716798,
      "memory(GiB)": 77.56,
      "step": 77625,
      "token_acc": 0.532319391634981,
      "train_speed(iter/s)": 1.437027
    },
    {
      "epoch": 3.325907201919369,
      "grad_norm": 6.506839752197266,
      "learning_rate": 2.5202314094592478e-05,
      "loss": 2.265080451965332,
      "memory(GiB)": 77.56,
      "step": 77630,
      "token_acc": 0.5114503816793893,
      "train_speed(iter/s)": 1.437016
    },
    {
      "epoch": 3.3261214172486184,
      "grad_norm": 5.440707683563232,
      "learning_rate": 2.519647052782727e-05,
      "loss": 2.350636291503906,
      "memory(GiB)": 77.56,
      "step": 77635,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.437017
    },
    {
      "epoch": 3.3263356325778672,
      "grad_norm": 7.022124767303467,
      "learning_rate": 2.519062741040129e-05,
      "loss": 2.264799880981445,
      "memory(GiB)": 77.56,
      "step": 77640,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437026
    },
    {
      "epoch": 3.326549847907116,
      "grad_norm": 6.467678546905518,
      "learning_rate": 2.5184784742420342e-05,
      "loss": 2.4305543899536133,
      "memory(GiB)": 77.56,
      "step": 77645,
      "token_acc": 0.4896142433234421,
      "train_speed(iter/s)": 1.437042
    },
    {
      "epoch": 3.3267640632363653,
      "grad_norm": 6.843394756317139,
      "learning_rate": 2.5178942523990324e-05,
      "loss": 2.4796756744384765,
      "memory(GiB)": 77.56,
      "step": 77650,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.437054
    },
    {
      "epoch": 3.326978278565614,
      "grad_norm": 7.019276142120361,
      "learning_rate": 2.5173100755217037e-05,
      "loss": 2.522543716430664,
      "memory(GiB)": 77.56,
      "step": 77655,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.437071
    },
    {
      "epoch": 3.327192493894863,
      "grad_norm": 6.105474948883057,
      "learning_rate": 2.5167259436206325e-05,
      "loss": 2.2438526153564453,
      "memory(GiB)": 77.56,
      "step": 77660,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.437085
    },
    {
      "epoch": 3.327406709224112,
      "grad_norm": 7.286857604980469,
      "learning_rate": 2.5161418567063983e-05,
      "loss": 2.4137947082519533,
      "memory(GiB)": 77.56,
      "step": 77665,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.437077
    },
    {
      "epoch": 3.327620924553361,
      "grad_norm": 4.995393753051758,
      "learning_rate": 2.5155578147895862e-05,
      "loss": 2.5548038482666016,
      "memory(GiB)": 77.56,
      "step": 77670,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.437102
    },
    {
      "epoch": 3.32783513988261,
      "grad_norm": 5.823831081390381,
      "learning_rate": 2.5149738178807745e-05,
      "loss": 2.384229278564453,
      "memory(GiB)": 77.56,
      "step": 77675,
      "token_acc": 0.5115606936416185,
      "train_speed(iter/s)": 1.437127
    },
    {
      "epoch": 3.328049355211859,
      "grad_norm": 4.5682806968688965,
      "learning_rate": 2.5143898659905442e-05,
      "loss": 2.3219558715820314,
      "memory(GiB)": 77.56,
      "step": 77680,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.43712
    },
    {
      "epoch": 3.328263570541108,
      "grad_norm": 6.711953639984131,
      "learning_rate": 2.5138059591294727e-05,
      "loss": 2.207210922241211,
      "memory(GiB)": 77.56,
      "step": 77685,
      "token_acc": 0.492,
      "train_speed(iter/s)": 1.437116
    },
    {
      "epoch": 3.3284777858703567,
      "grad_norm": 4.569845676422119,
      "learning_rate": 2.513222097308138e-05,
      "loss": 2.1944902420043944,
      "memory(GiB)": 77.56,
      "step": 77690,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.437127
    },
    {
      "epoch": 3.328692001199606,
      "grad_norm": 4.993614673614502,
      "learning_rate": 2.512638280537117e-05,
      "loss": 2.3164234161376953,
      "memory(GiB)": 77.56,
      "step": 77695,
      "token_acc": 0.4764705882352941,
      "train_speed(iter/s)": 1.437117
    },
    {
      "epoch": 3.3289062165288548,
      "grad_norm": 5.47150182723999,
      "learning_rate": 2.5120545088269877e-05,
      "loss": 2.5082813262939454,
      "memory(GiB)": 77.56,
      "step": 77700,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.437135
    },
    {
      "epoch": 3.3291204318581036,
      "grad_norm": 5.957233428955078,
      "learning_rate": 2.5114707821883253e-05,
      "loss": 2.559680938720703,
      "memory(GiB)": 77.56,
      "step": 77705,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.43714
    },
    {
      "epoch": 3.329334647187353,
      "grad_norm": 8.189870834350586,
      "learning_rate": 2.5108871006317046e-05,
      "loss": 2.5916494369506835,
      "memory(GiB)": 77.56,
      "step": 77710,
      "token_acc": 0.47570332480818417,
      "train_speed(iter/s)": 1.43713
    },
    {
      "epoch": 3.3295488625166016,
      "grad_norm": 6.239110469818115,
      "learning_rate": 2.510303464167698e-05,
      "loss": 2.6648834228515623,
      "memory(GiB)": 77.56,
      "step": 77715,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.437139
    },
    {
      "epoch": 3.3297630778458505,
      "grad_norm": 5.674903869628906,
      "learning_rate": 2.509719872806878e-05,
      "loss": 2.4365608215332033,
      "memory(GiB)": 77.56,
      "step": 77720,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.437147
    },
    {
      "epoch": 3.3299772931750997,
      "grad_norm": 6.108583927154541,
      "learning_rate": 2.5091363265598196e-05,
      "loss": 2.6104385375976564,
      "memory(GiB)": 77.56,
      "step": 77725,
      "token_acc": 0.46048109965635736,
      "train_speed(iter/s)": 1.437157
    },
    {
      "epoch": 3.3301915085043485,
      "grad_norm": 4.625551223754883,
      "learning_rate": 2.5085528254370942e-05,
      "loss": 2.2240320205688477,
      "memory(GiB)": 77.56,
      "step": 77730,
      "token_acc": 0.5015673981191222,
      "train_speed(iter/s)": 1.437143
    },
    {
      "epoch": 3.3304057238335973,
      "grad_norm": 5.3962554931640625,
      "learning_rate": 2.5079693694492722e-05,
      "loss": 2.283380317687988,
      "memory(GiB)": 77.56,
      "step": 77735,
      "token_acc": 0.5220338983050847,
      "train_speed(iter/s)": 1.437143
    },
    {
      "epoch": 3.3306199391628466,
      "grad_norm": 6.390612602233887,
      "learning_rate": 2.507385958606922e-05,
      "loss": 2.451319122314453,
      "memory(GiB)": 77.56,
      "step": 77740,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.437138
    },
    {
      "epoch": 3.3308341544920954,
      "grad_norm": 6.468560695648193,
      "learning_rate": 2.506802592920614e-05,
      "loss": 2.5463306427001955,
      "memory(GiB)": 77.56,
      "step": 77745,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.437146
    },
    {
      "epoch": 3.3310483698213442,
      "grad_norm": 5.827782154083252,
      "learning_rate": 2.506219272400916e-05,
      "loss": 2.2985111236572267,
      "memory(GiB)": 77.56,
      "step": 77750,
      "token_acc": 0.5450980392156862,
      "train_speed(iter/s)": 1.437164
    },
    {
      "epoch": 3.3312625851505935,
      "grad_norm": 6.416467189788818,
      "learning_rate": 2.5056359970583925e-05,
      "loss": 2.2989683151245117,
      "memory(GiB)": 77.56,
      "step": 77755,
      "token_acc": 0.519434628975265,
      "train_speed(iter/s)": 1.437168
    },
    {
      "epoch": 3.3314768004798423,
      "grad_norm": 6.548943996429443,
      "learning_rate": 2.505052766903615e-05,
      "loss": 2.1714780807495115,
      "memory(GiB)": 77.56,
      "step": 77760,
      "token_acc": 0.5131086142322098,
      "train_speed(iter/s)": 1.437163
    },
    {
      "epoch": 3.331691015809091,
      "grad_norm": 4.737626075744629,
      "learning_rate": 2.5044695819471475e-05,
      "loss": 2.2847221374511717,
      "memory(GiB)": 77.56,
      "step": 77765,
      "token_acc": 0.5266272189349113,
      "train_speed(iter/s)": 1.437167
    },
    {
      "epoch": 3.3319052311383404,
      "grad_norm": 5.4486284255981445,
      "learning_rate": 2.5038864421995545e-05,
      "loss": 1.9878698348999024,
      "memory(GiB)": 77.56,
      "step": 77770,
      "token_acc": 0.5418181818181819,
      "train_speed(iter/s)": 1.437172
    },
    {
      "epoch": 3.332119446467589,
      "grad_norm": 5.901936054229736,
      "learning_rate": 2.5033033476713998e-05,
      "loss": 2.3245212554931642,
      "memory(GiB)": 77.56,
      "step": 77775,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.437131
    },
    {
      "epoch": 3.332333661796838,
      "grad_norm": 5.089118003845215,
      "learning_rate": 2.5027202983732456e-05,
      "loss": 1.9117870330810547,
      "memory(GiB)": 77.56,
      "step": 77780,
      "token_acc": 0.5826446280991735,
      "train_speed(iter/s)": 1.437137
    },
    {
      "epoch": 3.3325478771260872,
      "grad_norm": 4.50844144821167,
      "learning_rate": 2.5021372943156575e-05,
      "loss": 2.2879833221435546,
      "memory(GiB)": 77.56,
      "step": 77785,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.437146
    },
    {
      "epoch": 3.332762092455336,
      "grad_norm": 5.373756408691406,
      "learning_rate": 2.501554335509195e-05,
      "loss": 2.2437599182128904,
      "memory(GiB)": 77.56,
      "step": 77790,
      "token_acc": 0.511400651465798,
      "train_speed(iter/s)": 1.437152
    },
    {
      "epoch": 3.332976307784585,
      "grad_norm": 5.752211570739746,
      "learning_rate": 2.5009714219644175e-05,
      "loss": 2.316151237487793,
      "memory(GiB)": 77.56,
      "step": 77795,
      "token_acc": 0.4898785425101215,
      "train_speed(iter/s)": 1.437155
    },
    {
      "epoch": 3.333190523113834,
      "grad_norm": 6.007317066192627,
      "learning_rate": 2.5003885536918897e-05,
      "loss": 2.331071662902832,
      "memory(GiB)": 77.56,
      "step": 77800,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.437149
    },
    {
      "epoch": 3.333404738443083,
      "grad_norm": 8.905684471130371,
      "learning_rate": 2.4998057307021677e-05,
      "loss": 2.60831298828125,
      "memory(GiB)": 77.56,
      "step": 77805,
      "token_acc": 0.4725609756097561,
      "train_speed(iter/s)": 1.437148
    },
    {
      "epoch": 3.3336189537723317,
      "grad_norm": 7.32859468460083,
      "learning_rate": 2.4992229530058087e-05,
      "loss": 2.297551727294922,
      "memory(GiB)": 77.56,
      "step": 77810,
      "token_acc": 0.5341365461847389,
      "train_speed(iter/s)": 1.437142
    },
    {
      "epoch": 3.333833169101581,
      "grad_norm": 5.975539684295654,
      "learning_rate": 2.498640220613373e-05,
      "loss": 2.313236427307129,
      "memory(GiB)": 77.56,
      "step": 77815,
      "token_acc": 0.5112994350282486,
      "train_speed(iter/s)": 1.437141
    },
    {
      "epoch": 3.33404738443083,
      "grad_norm": 5.545083045959473,
      "learning_rate": 2.498057533535417e-05,
      "loss": 2.289481544494629,
      "memory(GiB)": 77.56,
      "step": 77820,
      "token_acc": 0.5079872204472844,
      "train_speed(iter/s)": 1.437156
    },
    {
      "epoch": 3.3342615997600786,
      "grad_norm": 7.65740966796875,
      "learning_rate": 2.4974748917824954e-05,
      "loss": 2.2598876953125,
      "memory(GiB)": 77.56,
      "step": 77825,
      "token_acc": 0.5368852459016393,
      "train_speed(iter/s)": 1.437152
    },
    {
      "epoch": 3.334475815089328,
      "grad_norm": 5.546942234039307,
      "learning_rate": 2.4968922953651635e-05,
      "loss": 2.283573341369629,
      "memory(GiB)": 77.56,
      "step": 77830,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.43715
    },
    {
      "epoch": 3.3346900304185767,
      "grad_norm": 6.100461959838867,
      "learning_rate": 2.496309744293976e-05,
      "loss": 2.337263488769531,
      "memory(GiB)": 77.56,
      "step": 77835,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.437171
    },
    {
      "epoch": 3.3349042457478255,
      "grad_norm": 5.265915870666504,
      "learning_rate": 2.495727238579484e-05,
      "loss": 2.141219711303711,
      "memory(GiB)": 77.56,
      "step": 77840,
      "token_acc": 0.556390977443609,
      "train_speed(iter/s)": 1.437176
    },
    {
      "epoch": 3.3351184610770748,
      "grad_norm": 5.829535484313965,
      "learning_rate": 2.495144778232244e-05,
      "loss": 2.5924875259399416,
      "memory(GiB)": 77.56,
      "step": 77845,
      "token_acc": 0.4746268656716418,
      "train_speed(iter/s)": 1.437192
    },
    {
      "epoch": 3.3353326764063236,
      "grad_norm": 5.743008613586426,
      "learning_rate": 2.4945623632628068e-05,
      "loss": 2.428538131713867,
      "memory(GiB)": 77.56,
      "step": 77850,
      "token_acc": 0.49014084507042255,
      "train_speed(iter/s)": 1.437194
    },
    {
      "epoch": 3.3355468917355724,
      "grad_norm": 5.537662029266357,
      "learning_rate": 2.4939799936817222e-05,
      "loss": 2.6278318405151366,
      "memory(GiB)": 77.56,
      "step": 77855,
      "token_acc": 0.4631578947368421,
      "train_speed(iter/s)": 1.437185
    },
    {
      "epoch": 3.3357611070648217,
      "grad_norm": 6.905949592590332,
      "learning_rate": 2.493397669499541e-05,
      "loss": 2.5519298553466796,
      "memory(GiB)": 77.56,
      "step": 77860,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.437167
    },
    {
      "epoch": 3.3359753223940705,
      "grad_norm": 7.764893531799316,
      "learning_rate": 2.4928153907268104e-05,
      "loss": 2.425844192504883,
      "memory(GiB)": 77.56,
      "step": 77865,
      "token_acc": 0.4923547400611621,
      "train_speed(iter/s)": 1.437178
    },
    {
      "epoch": 3.3361895377233193,
      "grad_norm": 6.55760383605957,
      "learning_rate": 2.4922331573740808e-05,
      "loss": 2.3337722778320313,
      "memory(GiB)": 77.56,
      "step": 77870,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.437154
    },
    {
      "epoch": 3.3364037530525685,
      "grad_norm": 7.696887493133545,
      "learning_rate": 2.491650969451902e-05,
      "loss": 2.0694936752319335,
      "memory(GiB)": 77.56,
      "step": 77875,
      "token_acc": 0.5202952029520295,
      "train_speed(iter/s)": 1.437177
    },
    {
      "epoch": 3.3366179683818173,
      "grad_norm": 6.950080394744873,
      "learning_rate": 2.4910688269708195e-05,
      "loss": 2.245351219177246,
      "memory(GiB)": 77.56,
      "step": 77880,
      "token_acc": 0.5017064846416383,
      "train_speed(iter/s)": 1.437179
    },
    {
      "epoch": 3.336832183711066,
      "grad_norm": 5.977203369140625,
      "learning_rate": 2.4904867299413783e-05,
      "loss": 2.2859540939331056,
      "memory(GiB)": 77.56,
      "step": 77885,
      "token_acc": 0.5176470588235295,
      "train_speed(iter/s)": 1.437181
    },
    {
      "epoch": 3.3370463990403154,
      "grad_norm": 6.971149444580078,
      "learning_rate": 2.4899046783741235e-05,
      "loss": 2.1541297912597654,
      "memory(GiB)": 77.56,
      "step": 77890,
      "token_acc": 0.5214285714285715,
      "train_speed(iter/s)": 1.437183
    },
    {
      "epoch": 3.3372606143695642,
      "grad_norm": 6.053440093994141,
      "learning_rate": 2.4893226722796008e-05,
      "loss": 2.4205205917358397,
      "memory(GiB)": 77.56,
      "step": 77895,
      "token_acc": 0.4790996784565916,
      "train_speed(iter/s)": 1.437206
    },
    {
      "epoch": 3.337474829698813,
      "grad_norm": 6.8531880378723145,
      "learning_rate": 2.4887407116683516e-05,
      "loss": 2.202609062194824,
      "memory(GiB)": 77.56,
      "step": 77900,
      "token_acc": 0.48535564853556484,
      "train_speed(iter/s)": 1.437195
    },
    {
      "epoch": 3.3376890450280623,
      "grad_norm": 5.678084850311279,
      "learning_rate": 2.488158796550921e-05,
      "loss": 2.300441360473633,
      "memory(GiB)": 77.56,
      "step": 77905,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.437164
    },
    {
      "epoch": 3.337903260357311,
      "grad_norm": 5.907774925231934,
      "learning_rate": 2.487576926937851e-05,
      "loss": 1.8783260345458985,
      "memory(GiB)": 77.56,
      "step": 77910,
      "token_acc": 0.5447470817120622,
      "train_speed(iter/s)": 1.437174
    },
    {
      "epoch": 3.33811747568656,
      "grad_norm": 6.085709095001221,
      "learning_rate": 2.4869951028396813e-05,
      "loss": 2.2033203125,
      "memory(GiB)": 77.56,
      "step": 77915,
      "token_acc": 0.5315985130111525,
      "train_speed(iter/s)": 1.43718
    },
    {
      "epoch": 3.338331691015809,
      "grad_norm": 7.500263690948486,
      "learning_rate": 2.486413324266953e-05,
      "loss": 2.3250667572021486,
      "memory(GiB)": 77.56,
      "step": 77920,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.437173
    },
    {
      "epoch": 3.338545906345058,
      "grad_norm": 5.4479241371154785,
      "learning_rate": 2.4858315912302033e-05,
      "loss": 2.252872848510742,
      "memory(GiB)": 77.56,
      "step": 77925,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437164
    },
    {
      "epoch": 3.338760121674307,
      "grad_norm": 5.195981502532959,
      "learning_rate": 2.4852499037399747e-05,
      "loss": 2.331142044067383,
      "memory(GiB)": 77.56,
      "step": 77930,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.437185
    },
    {
      "epoch": 3.338974337003556,
      "grad_norm": 7.888880252838135,
      "learning_rate": 2.484668261806803e-05,
      "loss": 2.3773981094360352,
      "memory(GiB)": 77.56,
      "step": 77935,
      "token_acc": 0.4979757085020243,
      "train_speed(iter/s)": 1.437209
    },
    {
      "epoch": 3.339188552332805,
      "grad_norm": 7.641469955444336,
      "learning_rate": 2.4840866654412232e-05,
      "loss": 2.4169261932373045,
      "memory(GiB)": 77.56,
      "step": 77940,
      "token_acc": 0.44192634560906513,
      "train_speed(iter/s)": 1.437231
    },
    {
      "epoch": 3.3394027676620537,
      "grad_norm": 5.84895658493042,
      "learning_rate": 2.483505114653776e-05,
      "loss": 2.359182357788086,
      "memory(GiB)": 77.56,
      "step": 77945,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.437234
    },
    {
      "epoch": 3.339616982991303,
      "grad_norm": 6.297906398773193,
      "learning_rate": 2.482923609454994e-05,
      "loss": 2.5064010620117188,
      "memory(GiB)": 77.56,
      "step": 77950,
      "token_acc": 0.48493975903614456,
      "train_speed(iter/s)": 1.437223
    },
    {
      "epoch": 3.3398311983205518,
      "grad_norm": 5.065043926239014,
      "learning_rate": 2.48234214985541e-05,
      "loss": 2.597334098815918,
      "memory(GiB)": 77.56,
      "step": 77955,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.437235
    },
    {
      "epoch": 3.3400454136498006,
      "grad_norm": 7.707672595977783,
      "learning_rate": 2.4817607358655614e-05,
      "loss": 2.3225643157958986,
      "memory(GiB)": 77.56,
      "step": 77960,
      "token_acc": 0.4977973568281938,
      "train_speed(iter/s)": 1.437217
    },
    {
      "epoch": 3.34025962897905,
      "grad_norm": 7.994879722595215,
      "learning_rate": 2.48117936749598e-05,
      "loss": 2.471352958679199,
      "memory(GiB)": 77.56,
      "step": 77965,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.43724
    },
    {
      "epoch": 3.3404738443082986,
      "grad_norm": 6.065828323364258,
      "learning_rate": 2.480598044757197e-05,
      "loss": 2.297563934326172,
      "memory(GiB)": 77.56,
      "step": 77970,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.43724
    },
    {
      "epoch": 3.3406880596375474,
      "grad_norm": 6.3979010581970215,
      "learning_rate": 2.4800167676597436e-05,
      "loss": 2.1008651733398436,
      "memory(GiB)": 77.56,
      "step": 77975,
      "token_acc": 0.5643939393939394,
      "train_speed(iter/s)": 1.437231
    },
    {
      "epoch": 3.3409022749667967,
      "grad_norm": 6.603244781494141,
      "learning_rate": 2.4794355362141508e-05,
      "loss": 2.4242521286010743,
      "memory(GiB)": 77.56,
      "step": 77980,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437237
    },
    {
      "epoch": 3.3411164902960455,
      "grad_norm": 5.314538955688477,
      "learning_rate": 2.4788543504309454e-05,
      "loss": 2.4956085205078127,
      "memory(GiB)": 77.56,
      "step": 77985,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.437265
    },
    {
      "epoch": 3.3413307056252943,
      "grad_norm": 6.049429416656494,
      "learning_rate": 2.4782732103206607e-05,
      "loss": 2.0845806121826174,
      "memory(GiB)": 77.56,
      "step": 77990,
      "token_acc": 0.5681818181818182,
      "train_speed(iter/s)": 1.437261
    },
    {
      "epoch": 3.3415449209545436,
      "grad_norm": 5.353680610656738,
      "learning_rate": 2.4776921158938222e-05,
      "loss": 2.4130149841308595,
      "memory(GiB)": 77.56,
      "step": 77995,
      "token_acc": 0.47278911564625853,
      "train_speed(iter/s)": 1.437271
    },
    {
      "epoch": 3.3417591362837924,
      "grad_norm": 7.042940139770508,
      "learning_rate": 2.4771110671609573e-05,
      "loss": 2.3023122787475585,
      "memory(GiB)": 77.56,
      "step": 78000,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.43729
    },
    {
      "epoch": 3.3417591362837924,
      "eval_loss": 2.0868334770202637,
      "eval_runtime": 14.2158,
      "eval_samples_per_second": 7.034,
      "eval_steps_per_second": 7.034,
      "eval_token_acc": 0.4759556103575832,
      "step": 78000
    },
    {
      "epoch": 3.341973351613041,
      "grad_norm": 5.571708679199219,
      "learning_rate": 2.4765300641325915e-05,
      "loss": 2.2235841751098633,
      "memory(GiB)": 77.56,
      "step": 78005,
      "token_acc": 0.49167397020157755,
      "train_speed(iter/s)": 1.436902
    },
    {
      "epoch": 3.3421875669422905,
      "grad_norm": 6.3985772132873535,
      "learning_rate": 2.4759491068192496e-05,
      "loss": 2.1793134689331053,
      "memory(GiB)": 77.56,
      "step": 78010,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.436896
    },
    {
      "epoch": 3.3424017822715393,
      "grad_norm": 6.703181266784668,
      "learning_rate": 2.4753681952314573e-05,
      "loss": 2.1313396453857423,
      "memory(GiB)": 77.56,
      "step": 78015,
      "token_acc": 0.5513307984790875,
      "train_speed(iter/s)": 1.436896
    },
    {
      "epoch": 3.342615997600788,
      "grad_norm": 7.6307220458984375,
      "learning_rate": 2.4747873293797396e-05,
      "loss": 2.4202816009521486,
      "memory(GiB)": 77.56,
      "step": 78020,
      "token_acc": 0.4657534246575342,
      "train_speed(iter/s)": 1.436907
    },
    {
      "epoch": 3.3428302129300373,
      "grad_norm": 4.953913688659668,
      "learning_rate": 2.474206509274619e-05,
      "loss": 2.2251657485961913,
      "memory(GiB)": 77.56,
      "step": 78025,
      "token_acc": 0.50997150997151,
      "train_speed(iter/s)": 1.436888
    },
    {
      "epoch": 3.343044428259286,
      "grad_norm": 4.519009590148926,
      "learning_rate": 2.4736257349266167e-05,
      "loss": 1.9159467697143555,
      "memory(GiB)": 77.56,
      "step": 78030,
      "token_acc": 0.5593869731800766,
      "train_speed(iter/s)": 1.436896
    },
    {
      "epoch": 3.343258643588535,
      "grad_norm": 5.637549877166748,
      "learning_rate": 2.473045006346254e-05,
      "loss": 2.184259033203125,
      "memory(GiB)": 77.56,
      "step": 78035,
      "token_acc": 0.4781021897810219,
      "train_speed(iter/s)": 1.436914
    },
    {
      "epoch": 3.3434728589177842,
      "grad_norm": 6.153614521026611,
      "learning_rate": 2.472464323544052e-05,
      "loss": 2.3234369277954103,
      "memory(GiB)": 77.56,
      "step": 78040,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.436913
    },
    {
      "epoch": 3.343687074247033,
      "grad_norm": 6.867588996887207,
      "learning_rate": 2.4718836865305274e-05,
      "loss": 2.067275047302246,
      "memory(GiB)": 77.56,
      "step": 78045,
      "token_acc": 0.5478927203065134,
      "train_speed(iter/s)": 1.436905
    },
    {
      "epoch": 3.343901289576282,
      "grad_norm": 5.022627830505371,
      "learning_rate": 2.4713030953162036e-05,
      "loss": 2.133124351501465,
      "memory(GiB)": 77.56,
      "step": 78050,
      "token_acc": 0.48046875,
      "train_speed(iter/s)": 1.436919
    },
    {
      "epoch": 3.344115504905531,
      "grad_norm": 6.870769500732422,
      "learning_rate": 2.470722549911596e-05,
      "loss": 2.483769989013672,
      "memory(GiB)": 77.56,
      "step": 78055,
      "token_acc": 0.4767932489451477,
      "train_speed(iter/s)": 1.436932
    },
    {
      "epoch": 3.34432972023478,
      "grad_norm": 4.614297389984131,
      "learning_rate": 2.470142050327222e-05,
      "loss": 1.9196325302124024,
      "memory(GiB)": 77.56,
      "step": 78060,
      "token_acc": 0.5791245791245792,
      "train_speed(iter/s)": 1.436951
    },
    {
      "epoch": 3.3445439355640287,
      "grad_norm": 5.502196311950684,
      "learning_rate": 2.4695615965735984e-05,
      "loss": 2.3422874450683593,
      "memory(GiB)": 77.56,
      "step": 78065,
      "token_acc": 0.5286624203821656,
      "train_speed(iter/s)": 1.436971
    },
    {
      "epoch": 3.344758150893278,
      "grad_norm": 7.608879566192627,
      "learning_rate": 2.468981188661238e-05,
      "loss": 2.381334686279297,
      "memory(GiB)": 77.56,
      "step": 78070,
      "token_acc": 0.48787878787878786,
      "train_speed(iter/s)": 1.436987
    },
    {
      "epoch": 3.344972366222527,
      "grad_norm": 5.68510103225708,
      "learning_rate": 2.46840082660066e-05,
      "loss": 2.297234535217285,
      "memory(GiB)": 77.56,
      "step": 78075,
      "token_acc": 0.5223880597014925,
      "train_speed(iter/s)": 1.436998
    },
    {
      "epoch": 3.3451865815517756,
      "grad_norm": 5.46876859664917,
      "learning_rate": 2.467820510402375e-05,
      "loss": 2.5136638641357423,
      "memory(GiB)": 77.56,
      "step": 78080,
      "token_acc": 0.49008498583569404,
      "train_speed(iter/s)": 1.437005
    },
    {
      "epoch": 3.345400796881025,
      "grad_norm": 5.594293594360352,
      "learning_rate": 2.4672402400768973e-05,
      "loss": 2.059544563293457,
      "memory(GiB)": 77.56,
      "step": 78085,
      "token_acc": 0.5551020408163265,
      "train_speed(iter/s)": 1.437011
    },
    {
      "epoch": 3.3456150122102737,
      "grad_norm": 5.572644233703613,
      "learning_rate": 2.466660015634737e-05,
      "loss": 2.262191581726074,
      "memory(GiB)": 77.56,
      "step": 78090,
      "token_acc": 0.5033783783783784,
      "train_speed(iter/s)": 1.437005
    },
    {
      "epoch": 3.3458292275395225,
      "grad_norm": 5.715251922607422,
      "learning_rate": 2.4660798370864086e-05,
      "loss": 2.2446441650390625,
      "memory(GiB)": 77.56,
      "step": 78095,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.437019
    },
    {
      "epoch": 3.3460434428687718,
      "grad_norm": 5.06517219543457,
      "learning_rate": 2.465499704442419e-05,
      "loss": 2.350856399536133,
      "memory(GiB)": 77.56,
      "step": 78100,
      "token_acc": 0.48286604361370716,
      "train_speed(iter/s)": 1.437038
    },
    {
      "epoch": 3.3462576581980206,
      "grad_norm": 6.508033752441406,
      "learning_rate": 2.4649196177132818e-05,
      "loss": 2.49069766998291,
      "memory(GiB)": 77.56,
      "step": 78105,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.43703
    },
    {
      "epoch": 3.3464718735272694,
      "grad_norm": 5.585472583770752,
      "learning_rate": 2.4643395769095035e-05,
      "loss": 2.807998847961426,
      "memory(GiB)": 77.56,
      "step": 78110,
      "token_acc": 0.4024767801857585,
      "train_speed(iter/s)": 1.437029
    },
    {
      "epoch": 3.3466860888565186,
      "grad_norm": 6.244565486907959,
      "learning_rate": 2.4637595820415925e-05,
      "loss": 2.553742218017578,
      "memory(GiB)": 77.56,
      "step": 78115,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.437013
    },
    {
      "epoch": 3.3469003041857674,
      "grad_norm": 4.824309825897217,
      "learning_rate": 2.4631796331200564e-05,
      "loss": 2.383909225463867,
      "memory(GiB)": 77.56,
      "step": 78120,
      "token_acc": 0.44516129032258067,
      "train_speed(iter/s)": 1.437008
    },
    {
      "epoch": 3.3471145195150163,
      "grad_norm": 5.161762237548828,
      "learning_rate": 2.4625997301554005e-05,
      "loss": 2.1975996017456056,
      "memory(GiB)": 77.56,
      "step": 78125,
      "token_acc": 0.5119047619047619,
      "train_speed(iter/s)": 1.437006
    },
    {
      "epoch": 3.3473287348442655,
      "grad_norm": 5.7215800285339355,
      "learning_rate": 2.462019873158129e-05,
      "loss": 2.4382396697998048,
      "memory(GiB)": 77.56,
      "step": 78130,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436999
    },
    {
      "epoch": 3.3475429501735143,
      "grad_norm": 5.0890727043151855,
      "learning_rate": 2.46144006213875e-05,
      "loss": 2.512197494506836,
      "memory(GiB)": 77.56,
      "step": 78135,
      "token_acc": 0.535031847133758,
      "train_speed(iter/s)": 1.437003
    },
    {
      "epoch": 3.347757165502763,
      "grad_norm": 8.426888465881348,
      "learning_rate": 2.460860297107766e-05,
      "loss": 2.3685791015625,
      "memory(GiB)": 77.56,
      "step": 78140,
      "token_acc": 0.5252918287937743,
      "train_speed(iter/s)": 1.436993
    },
    {
      "epoch": 3.3479713808320124,
      "grad_norm": 6.038755893707275,
      "learning_rate": 2.4602805780756795e-05,
      "loss": 2.1319467544555666,
      "memory(GiB)": 77.56,
      "step": 78145,
      "token_acc": 0.5353535353535354,
      "train_speed(iter/s)": 1.436972
    },
    {
      "epoch": 3.348185596161261,
      "grad_norm": 5.103981971740723,
      "learning_rate": 2.459700905052993e-05,
      "loss": 2.464715576171875,
      "memory(GiB)": 77.56,
      "step": 78150,
      "token_acc": 0.4608150470219436,
      "train_speed(iter/s)": 1.436983
    },
    {
      "epoch": 3.34839981149051,
      "grad_norm": 6.755388259887695,
      "learning_rate": 2.459121278050205e-05,
      "loss": 2.696822929382324,
      "memory(GiB)": 77.56,
      "step": 78155,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.436985
    },
    {
      "epoch": 3.3486140268197593,
      "grad_norm": 5.507720470428467,
      "learning_rate": 2.4585416970778207e-05,
      "loss": 2.312678337097168,
      "memory(GiB)": 77.56,
      "step": 78160,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.436984
    },
    {
      "epoch": 3.348828242149008,
      "grad_norm": 6.923756122589111,
      "learning_rate": 2.4579621621463362e-05,
      "loss": 2.3618371963500975,
      "memory(GiB)": 77.56,
      "step": 78165,
      "token_acc": 0.5077881619937694,
      "train_speed(iter/s)": 1.436966
    },
    {
      "epoch": 3.349042457478257,
      "grad_norm": 5.579107761383057,
      "learning_rate": 2.4573826732662537e-05,
      "loss": 2.170520210266113,
      "memory(GiB)": 77.56,
      "step": 78170,
      "token_acc": 0.5477941176470589,
      "train_speed(iter/s)": 1.436948
    },
    {
      "epoch": 3.349256672807506,
      "grad_norm": 6.676364421844482,
      "learning_rate": 2.4568032304480687e-05,
      "loss": 2.558653450012207,
      "memory(GiB)": 77.56,
      "step": 78175,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.436948
    },
    {
      "epoch": 3.349470888136755,
      "grad_norm": 4.242257118225098,
      "learning_rate": 2.4562238337022793e-05,
      "loss": 2.0917230606079102,
      "memory(GiB)": 77.56,
      "step": 78180,
      "token_acc": 0.5470383275261324,
      "train_speed(iter/s)": 1.436916
    },
    {
      "epoch": 3.349685103466004,
      "grad_norm": 6.848104953765869,
      "learning_rate": 2.455644483039381e-05,
      "loss": 2.4666683197021486,
      "memory(GiB)": 77.56,
      "step": 78185,
      "token_acc": 0.4915254237288136,
      "train_speed(iter/s)": 1.436938
    },
    {
      "epoch": 3.349899318795253,
      "grad_norm": 7.107452869415283,
      "learning_rate": 2.455065178469868e-05,
      "loss": 2.505731964111328,
      "memory(GiB)": 77.56,
      "step": 78190,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.436943
    },
    {
      "epoch": 3.350113534124502,
      "grad_norm": 7.527360916137695,
      "learning_rate": 2.4544859200042386e-05,
      "loss": 2.4773880004882813,
      "memory(GiB)": 77.56,
      "step": 78195,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.436947
    },
    {
      "epoch": 3.3503277494537507,
      "grad_norm": 5.069779396057129,
      "learning_rate": 2.4539067076529847e-05,
      "loss": 2.4517032623291017,
      "memory(GiB)": 77.56,
      "step": 78200,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.436958
    },
    {
      "epoch": 3.350541964783,
      "grad_norm": 6.344890117645264,
      "learning_rate": 2.4533275414265992e-05,
      "loss": 2.4281055450439455,
      "memory(GiB)": 77.56,
      "step": 78205,
      "token_acc": 0.4819672131147541,
      "train_speed(iter/s)": 1.436947
    },
    {
      "epoch": 3.3507561801122487,
      "grad_norm": 6.175689220428467,
      "learning_rate": 2.452748421335574e-05,
      "loss": 2.127104568481445,
      "memory(GiB)": 77.56,
      "step": 78210,
      "token_acc": 0.5390625,
      "train_speed(iter/s)": 1.436958
    },
    {
      "epoch": 3.3509703954414976,
      "grad_norm": 5.342073440551758,
      "learning_rate": 2.452169347390399e-05,
      "loss": 2.346451759338379,
      "memory(GiB)": 77.56,
      "step": 78215,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.436937
    },
    {
      "epoch": 3.351184610770747,
      "grad_norm": 7.146237850189209,
      "learning_rate": 2.4515903196015684e-05,
      "loss": 2.3414581298828123,
      "memory(GiB)": 77.56,
      "step": 78220,
      "token_acc": 0.5223880597014925,
      "train_speed(iter/s)": 1.436933
    },
    {
      "epoch": 3.3513988260999956,
      "grad_norm": 6.014937877655029,
      "learning_rate": 2.4510113379795696e-05,
      "loss": 2.262928009033203,
      "memory(GiB)": 77.56,
      "step": 78225,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436936
    },
    {
      "epoch": 3.3516130414292444,
      "grad_norm": 5.9865570068359375,
      "learning_rate": 2.4504324025348912e-05,
      "loss": 2.5818096160888673,
      "memory(GiB)": 77.56,
      "step": 78230,
      "token_acc": 0.48639455782312924,
      "train_speed(iter/s)": 1.436936
    },
    {
      "epoch": 3.3518272567584937,
      "grad_norm": 6.626688003540039,
      "learning_rate": 2.44985351327802e-05,
      "loss": 2.4460886001586912,
      "memory(GiB)": 77.56,
      "step": 78235,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.43694
    },
    {
      "epoch": 3.3520414720877425,
      "grad_norm": 5.517338275909424,
      "learning_rate": 2.4492746702194463e-05,
      "loss": 2.3007774353027344,
      "memory(GiB)": 77.56,
      "step": 78240,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.436961
    },
    {
      "epoch": 3.3522556874169913,
      "grad_norm": 5.2339186668396,
      "learning_rate": 2.448695873369653e-05,
      "loss": 2.563678169250488,
      "memory(GiB)": 77.56,
      "step": 78245,
      "token_acc": 0.4517133956386293,
      "train_speed(iter/s)": 1.436973
    },
    {
      "epoch": 3.3524699027462406,
      "grad_norm": 6.285702705383301,
      "learning_rate": 2.4481171227391293e-05,
      "loss": 2.2242761611938477,
      "memory(GiB)": 77.56,
      "step": 78250,
      "token_acc": 0.4752851711026616,
      "train_speed(iter/s)": 1.436974
    },
    {
      "epoch": 3.3526841180754894,
      "grad_norm": 6.033846855163574,
      "learning_rate": 2.4475384183383577e-05,
      "loss": 2.1046445846557615,
      "memory(GiB)": 77.56,
      "step": 78255,
      "token_acc": 0.5488721804511278,
      "train_speed(iter/s)": 1.436976
    },
    {
      "epoch": 3.352898333404738,
      "grad_norm": 6.011137962341309,
      "learning_rate": 2.4469597601778222e-05,
      "loss": 2.1895471572875977,
      "memory(GiB)": 77.56,
      "step": 78260,
      "token_acc": 0.4622356495468278,
      "train_speed(iter/s)": 1.436964
    },
    {
      "epoch": 3.3531125487339875,
      "grad_norm": 5.461022853851318,
      "learning_rate": 2.446381148268005e-05,
      "loss": 2.343461608886719,
      "memory(GiB)": 77.56,
      "step": 78265,
      "token_acc": 0.4405144694533762,
      "train_speed(iter/s)": 1.436953
    },
    {
      "epoch": 3.3533267640632363,
      "grad_norm": 8.106976509094238,
      "learning_rate": 2.445802582619389e-05,
      "loss": 2.473061752319336,
      "memory(GiB)": 77.56,
      "step": 78270,
      "token_acc": 0.43824701195219123,
      "train_speed(iter/s)": 1.436983
    },
    {
      "epoch": 3.353540979392485,
      "grad_norm": 5.615047454833984,
      "learning_rate": 2.4452240632424538e-05,
      "loss": 2.2097375869750975,
      "memory(GiB)": 77.56,
      "step": 78275,
      "token_acc": 0.5352941176470588,
      "train_speed(iter/s)": 1.436975
    },
    {
      "epoch": 3.3537551947217343,
      "grad_norm": 8.703311920166016,
      "learning_rate": 2.4446455901476828e-05,
      "loss": 2.5292709350585936,
      "memory(GiB)": 77.56,
      "step": 78280,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.436964
    },
    {
      "epoch": 3.353969410050983,
      "grad_norm": 5.164878845214844,
      "learning_rate": 2.4440671633455543e-05,
      "loss": 2.130105209350586,
      "memory(GiB)": 77.56,
      "step": 78285,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.436954
    },
    {
      "epoch": 3.3541836253802324,
      "grad_norm": 4.742149353027344,
      "learning_rate": 2.4434887828465463e-05,
      "loss": 2.4441783905029295,
      "memory(GiB)": 77.56,
      "step": 78290,
      "token_acc": 0.5072046109510087,
      "train_speed(iter/s)": 1.436975
    },
    {
      "epoch": 3.354397840709481,
      "grad_norm": 7.378182411193848,
      "learning_rate": 2.4429104486611376e-05,
      "loss": 2.41677188873291,
      "memory(GiB)": 77.56,
      "step": 78295,
      "token_acc": 0.47988505747126436,
      "train_speed(iter/s)": 1.436975
    },
    {
      "epoch": 3.35461205603873,
      "grad_norm": 5.843394756317139,
      "learning_rate": 2.4423321607998028e-05,
      "loss": 2.53665771484375,
      "memory(GiB)": 77.56,
      "step": 78300,
      "token_acc": 0.48044692737430167,
      "train_speed(iter/s)": 1.436968
    },
    {
      "epoch": 3.3548262713679793,
      "grad_norm": 6.190471649169922,
      "learning_rate": 2.4417539192730226e-05,
      "loss": 2.4405431747436523,
      "memory(GiB)": 77.56,
      "step": 78305,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.436981
    },
    {
      "epoch": 3.355040486697228,
      "grad_norm": 5.517795085906982,
      "learning_rate": 2.4411757240912675e-05,
      "loss": 2.4316272735595703,
      "memory(GiB)": 77.56,
      "step": 78310,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.436988
    },
    {
      "epoch": 3.355254702026477,
      "grad_norm": 5.0164995193481445,
      "learning_rate": 2.4405975752650168e-05,
      "loss": 2.506674575805664,
      "memory(GiB)": 77.56,
      "step": 78315,
      "token_acc": 0.48031496062992124,
      "train_speed(iter/s)": 1.436984
    },
    {
      "epoch": 3.355468917355726,
      "grad_norm": 5.901783466339111,
      "learning_rate": 2.4400194728047414e-05,
      "loss": 2.279845428466797,
      "memory(GiB)": 77.56,
      "step": 78320,
      "token_acc": 0.5140562248995983,
      "train_speed(iter/s)": 1.436978
    },
    {
      "epoch": 3.355683132684975,
      "grad_norm": 5.524072170257568,
      "learning_rate": 2.4394414167209152e-05,
      "loss": 2.553447151184082,
      "memory(GiB)": 77.56,
      "step": 78325,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.436979
    },
    {
      "epoch": 3.355897348014224,
      "grad_norm": 7.261928081512451,
      "learning_rate": 2.4388634070240097e-05,
      "loss": 2.4628835678100587,
      "memory(GiB)": 77.56,
      "step": 78330,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.436966
    },
    {
      "epoch": 3.356111563343473,
      "grad_norm": 5.276325702667236,
      "learning_rate": 2.438285443724494e-05,
      "loss": 2.3215408325195312,
      "memory(GiB)": 77.56,
      "step": 78335,
      "token_acc": 0.49544072948328266,
      "train_speed(iter/s)": 1.436985
    },
    {
      "epoch": 3.356325778672722,
      "grad_norm": 5.752867221832275,
      "learning_rate": 2.4377075268328426e-05,
      "loss": 2.396231842041016,
      "memory(GiB)": 77.56,
      "step": 78340,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.437
    },
    {
      "epoch": 3.3565399940019707,
      "grad_norm": 6.296432018280029,
      "learning_rate": 2.437129656359523e-05,
      "loss": 2.437132453918457,
      "memory(GiB)": 77.56,
      "step": 78345,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.437008
    },
    {
      "epoch": 3.35675420933122,
      "grad_norm": 6.259218215942383,
      "learning_rate": 2.4365518323150037e-05,
      "loss": 2.417165756225586,
      "memory(GiB)": 77.56,
      "step": 78350,
      "token_acc": 0.4980544747081712,
      "train_speed(iter/s)": 1.437023
    },
    {
      "epoch": 3.3569684246604687,
      "grad_norm": 6.226585865020752,
      "learning_rate": 2.4359740547097526e-05,
      "loss": 2.375764083862305,
      "memory(GiB)": 77.56,
      "step": 78355,
      "token_acc": 0.44140625,
      "train_speed(iter/s)": 1.43703
    },
    {
      "epoch": 3.3571826399897176,
      "grad_norm": 4.719928741455078,
      "learning_rate": 2.435396323554235e-05,
      "loss": 2.0290273666381835,
      "memory(GiB)": 77.56,
      "step": 78360,
      "token_acc": 0.5563380281690141,
      "train_speed(iter/s)": 1.437035
    },
    {
      "epoch": 3.357396855318967,
      "grad_norm": 6.692323684692383,
      "learning_rate": 2.4348186388589206e-05,
      "loss": 2.3463871002197267,
      "memory(GiB)": 77.56,
      "step": 78365,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.437019
    },
    {
      "epoch": 3.3576110706482156,
      "grad_norm": 7.227686882019043,
      "learning_rate": 2.4342410006342732e-05,
      "loss": 2.2718215942382813,
      "memory(GiB)": 77.56,
      "step": 78370,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.437023
    },
    {
      "epoch": 3.3578252859774644,
      "grad_norm": 5.753372669219971,
      "learning_rate": 2.4336634088907566e-05,
      "loss": 2.347666931152344,
      "memory(GiB)": 77.56,
      "step": 78375,
      "token_acc": 0.4919093851132686,
      "train_speed(iter/s)": 1.437016
    },
    {
      "epoch": 3.3580395013067137,
      "grad_norm": 6.541168689727783,
      "learning_rate": 2.4330858636388348e-05,
      "loss": 2.3220420837402345,
      "memory(GiB)": 77.56,
      "step": 78380,
      "token_acc": 0.4902597402597403,
      "train_speed(iter/s)": 1.437021
    },
    {
      "epoch": 3.3582537166359625,
      "grad_norm": 6.848973274230957,
      "learning_rate": 2.432508364888969e-05,
      "loss": 2.5263412475585936,
      "memory(GiB)": 77.56,
      "step": 78385,
      "token_acc": 0.45394736842105265,
      "train_speed(iter/s)": 1.437023
    },
    {
      "epoch": 3.3584679319652113,
      "grad_norm": 5.124996662139893,
      "learning_rate": 2.431930912651622e-05,
      "loss": 2.5684064865112304,
      "memory(GiB)": 77.56,
      "step": 78390,
      "token_acc": 0.5165165165165165,
      "train_speed(iter/s)": 1.437033
    },
    {
      "epoch": 3.3586821472944606,
      "grad_norm": 7.014379024505615,
      "learning_rate": 2.4313535069372584e-05,
      "loss": 2.1359622955322264,
      "memory(GiB)": 77.56,
      "step": 78395,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.437035
    },
    {
      "epoch": 3.3588963626237094,
      "grad_norm": 5.489606857299805,
      "learning_rate": 2.430776147756335e-05,
      "loss": 1.9601947784423828,
      "memory(GiB)": 77.56,
      "step": 78400,
      "token_acc": 0.540453074433657,
      "train_speed(iter/s)": 1.437036
    },
    {
      "epoch": 3.359110577952958,
      "grad_norm": 5.671398639678955,
      "learning_rate": 2.4301988351193117e-05,
      "loss": 2.3124685287475586,
      "memory(GiB)": 77.56,
      "step": 78405,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.437025
    },
    {
      "epoch": 3.3593247932822075,
      "grad_norm": 5.578364372253418,
      "learning_rate": 2.4296215690366476e-05,
      "loss": 2.585572052001953,
      "memory(GiB)": 77.56,
      "step": 78410,
      "token_acc": 0.4753521126760563,
      "train_speed(iter/s)": 1.437031
    },
    {
      "epoch": 3.3595390086114563,
      "grad_norm": 7.403778076171875,
      "learning_rate": 2.4290443495188e-05,
      "loss": 2.462246322631836,
      "memory(GiB)": 77.56,
      "step": 78415,
      "token_acc": 0.49612403100775193,
      "train_speed(iter/s)": 1.43705
    },
    {
      "epoch": 3.359753223940705,
      "grad_norm": 7.052006721496582,
      "learning_rate": 2.4284671765762235e-05,
      "loss": 2.1567081451416015,
      "memory(GiB)": 77.56,
      "step": 78420,
      "token_acc": 0.5536912751677853,
      "train_speed(iter/s)": 1.437065
    },
    {
      "epoch": 3.3599674392699543,
      "grad_norm": 5.3216729164123535,
      "learning_rate": 2.427890050219378e-05,
      "loss": 2.389144515991211,
      "memory(GiB)": 77.56,
      "step": 78425,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.437079
    },
    {
      "epoch": 3.360181654599203,
      "grad_norm": 5.682316303253174,
      "learning_rate": 2.427312970458718e-05,
      "loss": 2.1203407287597655,
      "memory(GiB)": 77.56,
      "step": 78430,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.437077
    },
    {
      "epoch": 3.360395869928452,
      "grad_norm": 5.355327129364014,
      "learning_rate": 2.426735937304696e-05,
      "loss": 2.4626581192016603,
      "memory(GiB)": 77.56,
      "step": 78435,
      "token_acc": 0.5014749262536873,
      "train_speed(iter/s)": 1.437081
    },
    {
      "epoch": 3.360610085257701,
      "grad_norm": 6.6465325355529785,
      "learning_rate": 2.426158950767767e-05,
      "loss": 2.3337871551513674,
      "memory(GiB)": 77.56,
      "step": 78440,
      "token_acc": 0.5211267605633803,
      "train_speed(iter/s)": 1.437074
    },
    {
      "epoch": 3.36082430058695,
      "grad_norm": 6.364837169647217,
      "learning_rate": 2.425582010858381e-05,
      "loss": 2.3513408660888673,
      "memory(GiB)": 77.56,
      "step": 78445,
      "token_acc": 0.46579804560260585,
      "train_speed(iter/s)": 1.437065
    },
    {
      "epoch": 3.361038515916199,
      "grad_norm": 5.640448093414307,
      "learning_rate": 2.4250051175869938e-05,
      "loss": 2.292061996459961,
      "memory(GiB)": 77.56,
      "step": 78450,
      "token_acc": 0.501577287066246,
      "train_speed(iter/s)": 1.43707
    },
    {
      "epoch": 3.361252731245448,
      "grad_norm": 7.223572254180908,
      "learning_rate": 2.4244282709640542e-05,
      "loss": 2.3309823989868166,
      "memory(GiB)": 77.56,
      "step": 78455,
      "token_acc": 0.5134328358208955,
      "train_speed(iter/s)": 1.437087
    },
    {
      "epoch": 3.361466946574697,
      "grad_norm": 8.981651306152344,
      "learning_rate": 2.4238514710000103e-05,
      "loss": 2.4027854919433596,
      "memory(GiB)": 77.56,
      "step": 78460,
      "token_acc": 0.5233333333333333,
      "train_speed(iter/s)": 1.43707
    },
    {
      "epoch": 3.3616811619039457,
      "grad_norm": 6.733354091644287,
      "learning_rate": 2.4232747177053167e-05,
      "loss": 2.4601390838623045,
      "memory(GiB)": 77.56,
      "step": 78465,
      "token_acc": 0.4837662337662338,
      "train_speed(iter/s)": 1.437077
    },
    {
      "epoch": 3.361895377233195,
      "grad_norm": 5.849766254425049,
      "learning_rate": 2.422698011090418e-05,
      "loss": 2.5320676803588866,
      "memory(GiB)": 77.56,
      "step": 78470,
      "token_acc": 0.4745222929936306,
      "train_speed(iter/s)": 1.43708
    },
    {
      "epoch": 3.362109592562444,
      "grad_norm": 5.1082763671875,
      "learning_rate": 2.4221213511657624e-05,
      "loss": 2.312602424621582,
      "memory(GiB)": 77.56,
      "step": 78475,
      "token_acc": 0.5246478873239436,
      "train_speed(iter/s)": 1.437095
    },
    {
      "epoch": 3.3623238078916926,
      "grad_norm": 5.328272819519043,
      "learning_rate": 2.421544737941795e-05,
      "loss": 2.1653512954711913,
      "memory(GiB)": 77.56,
      "step": 78480,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.437079
    },
    {
      "epoch": 3.362538023220942,
      "grad_norm": 5.10175895690918,
      "learning_rate": 2.4209681714289655e-05,
      "loss": 2.266076850891113,
      "memory(GiB)": 77.56,
      "step": 78485,
      "token_acc": 0.478125,
      "train_speed(iter/s)": 1.4371
    },
    {
      "epoch": 3.3627522385501907,
      "grad_norm": 7.176144123077393,
      "learning_rate": 2.4203916516377167e-05,
      "loss": 2.213001823425293,
      "memory(GiB)": 77.56,
      "step": 78490,
      "token_acc": 0.5335276967930029,
      "train_speed(iter/s)": 1.437111
    },
    {
      "epoch": 3.3629664538794395,
      "grad_norm": 6.665398120880127,
      "learning_rate": 2.4198151785784934e-05,
      "loss": 1.9317384719848634,
      "memory(GiB)": 77.56,
      "step": 78495,
      "token_acc": 0.5387596899224806,
      "train_speed(iter/s)": 1.437122
    },
    {
      "epoch": 3.3631806692086887,
      "grad_norm": 7.4047112464904785,
      "learning_rate": 2.4192387522617384e-05,
      "loss": 2.1524192810058596,
      "memory(GiB)": 77.56,
      "step": 78500,
      "token_acc": 0.50390625,
      "train_speed(iter/s)": 1.437151
    },
    {
      "epoch": 3.3631806692086887,
      "eval_loss": 2.210336208343506,
      "eval_runtime": 14.0938,
      "eval_samples_per_second": 7.095,
      "eval_steps_per_second": 7.095,
      "eval_token_acc": 0.4819277108433735,
      "step": 78500
    },
    {
      "epoch": 3.3633948845379376,
      "grad_norm": 6.0035858154296875,
      "learning_rate": 2.4186623726978925e-05,
      "loss": 2.236882972717285,
      "memory(GiB)": 77.56,
      "step": 78505,
      "token_acc": 0.4975845410628019,
      "train_speed(iter/s)": 1.436735
    },
    {
      "epoch": 3.3636090998671864,
      "grad_norm": 9.440528869628906,
      "learning_rate": 2.418086039897401e-05,
      "loss": 2.7023162841796875,
      "memory(GiB)": 77.56,
      "step": 78510,
      "token_acc": 0.4437299035369775,
      "train_speed(iter/s)": 1.436719
    },
    {
      "epoch": 3.3638233151964356,
      "grad_norm": 6.444137096405029,
      "learning_rate": 2.4175097538707025e-05,
      "loss": 2.593526268005371,
      "memory(GiB)": 77.56,
      "step": 78515,
      "token_acc": 0.45695364238410596,
      "train_speed(iter/s)": 1.436714
    },
    {
      "epoch": 3.3640375305256844,
      "grad_norm": 5.1491289138793945,
      "learning_rate": 2.4169335146282378e-05,
      "loss": 2.54056396484375,
      "memory(GiB)": 77.56,
      "step": 78520,
      "token_acc": 0.46402877697841727,
      "train_speed(iter/s)": 1.436738
    },
    {
      "epoch": 3.3642517458549333,
      "grad_norm": 9.636716842651367,
      "learning_rate": 2.4163573221804457e-05,
      "loss": 2.5109472274780273,
      "memory(GiB)": 77.56,
      "step": 78525,
      "token_acc": 0.48046875,
      "train_speed(iter/s)": 1.436726
    },
    {
      "epoch": 3.3644659611841825,
      "grad_norm": 5.885531902313232,
      "learning_rate": 2.4157811765377624e-05,
      "loss": 2.549351119995117,
      "memory(GiB)": 77.56,
      "step": 78530,
      "token_acc": 0.5077399380804953,
      "train_speed(iter/s)": 1.436724
    },
    {
      "epoch": 3.3646801765134313,
      "grad_norm": 6.32645320892334,
      "learning_rate": 2.4152050777106273e-05,
      "loss": 2.239894485473633,
      "memory(GiB)": 77.56,
      "step": 78535,
      "token_acc": 0.523972602739726,
      "train_speed(iter/s)": 1.436717
    },
    {
      "epoch": 3.36489439184268,
      "grad_norm": 5.9216461181640625,
      "learning_rate": 2.414629025709479e-05,
      "loss": 2.386954498291016,
      "memory(GiB)": 77.56,
      "step": 78540,
      "token_acc": 0.48089171974522293,
      "train_speed(iter/s)": 1.436734
    },
    {
      "epoch": 3.3651086071719294,
      "grad_norm": 8.108158111572266,
      "learning_rate": 2.414053020544751e-05,
      "loss": 2.459110641479492,
      "memory(GiB)": 77.56,
      "step": 78545,
      "token_acc": 0.4765625,
      "train_speed(iter/s)": 1.436707
    },
    {
      "epoch": 3.365322822501178,
      "grad_norm": 5.555756092071533,
      "learning_rate": 2.4134770622268783e-05,
      "loss": 2.4668888092041015,
      "memory(GiB)": 77.56,
      "step": 78550,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.436682
    },
    {
      "epoch": 3.365537037830427,
      "grad_norm": 6.063318252563477,
      "learning_rate": 2.4129011507662945e-05,
      "loss": 2.2724941253662108,
      "memory(GiB)": 77.56,
      "step": 78555,
      "token_acc": 0.51985559566787,
      "train_speed(iter/s)": 1.436696
    },
    {
      "epoch": 3.3657512531596763,
      "grad_norm": 5.631870269775391,
      "learning_rate": 2.4123252861734334e-05,
      "loss": 2.016470527648926,
      "memory(GiB)": 77.56,
      "step": 78560,
      "token_acc": 0.528125,
      "train_speed(iter/s)": 1.436689
    },
    {
      "epoch": 3.365965468488925,
      "grad_norm": 8.290855407714844,
      "learning_rate": 2.4117494684587262e-05,
      "loss": 2.462776756286621,
      "memory(GiB)": 77.56,
      "step": 78565,
      "token_acc": 0.48036253776435045,
      "train_speed(iter/s)": 1.436694
    },
    {
      "epoch": 3.366179683818174,
      "grad_norm": 8.175646781921387,
      "learning_rate": 2.4111736976326066e-05,
      "loss": 2.38049201965332,
      "memory(GiB)": 77.56,
      "step": 78570,
      "token_acc": 0.49823321554770317,
      "train_speed(iter/s)": 1.43669
    },
    {
      "epoch": 3.366393899147423,
      "grad_norm": 5.28679895401001,
      "learning_rate": 2.410597973705504e-05,
      "loss": 2.0457387924194337,
      "memory(GiB)": 77.56,
      "step": 78575,
      "token_acc": 0.5506072874493927,
      "train_speed(iter/s)": 1.43669
    },
    {
      "epoch": 3.366608114476672,
      "grad_norm": 4.147820472717285,
      "learning_rate": 2.4100222966878484e-05,
      "loss": 2.4577089309692384,
      "memory(GiB)": 77.56,
      "step": 78580,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.436695
    },
    {
      "epoch": 3.3668223298059208,
      "grad_norm": 6.195014953613281,
      "learning_rate": 2.409446666590068e-05,
      "loss": 2.477466011047363,
      "memory(GiB)": 77.56,
      "step": 78585,
      "token_acc": 0.48375451263537905,
      "train_speed(iter/s)": 1.436713
    },
    {
      "epoch": 3.36703654513517,
      "grad_norm": 8.295110702514648,
      "learning_rate": 2.4088710834225896e-05,
      "loss": 2.2955419540405275,
      "memory(GiB)": 77.56,
      "step": 78590,
      "token_acc": 0.5043859649122807,
      "train_speed(iter/s)": 1.436705
    },
    {
      "epoch": 3.367250760464419,
      "grad_norm": 5.086785316467285,
      "learning_rate": 2.408295547195844e-05,
      "loss": 2.157618522644043,
      "memory(GiB)": 77.56,
      "step": 78595,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.436709
    },
    {
      "epoch": 3.3674649757936677,
      "grad_norm": 6.766571998596191,
      "learning_rate": 2.4077200579202563e-05,
      "loss": 2.3635231018066407,
      "memory(GiB)": 77.56,
      "step": 78600,
      "token_acc": 0.4613003095975232,
      "train_speed(iter/s)": 1.436729
    },
    {
      "epoch": 3.367679191122917,
      "grad_norm": 5.771625518798828,
      "learning_rate": 2.4071446156062494e-05,
      "loss": 2.4489162445068358,
      "memory(GiB)": 77.56,
      "step": 78605,
      "token_acc": 0.512987012987013,
      "train_speed(iter/s)": 1.436725
    },
    {
      "epoch": 3.3678934064521657,
      "grad_norm": 6.445941925048828,
      "learning_rate": 2.406569220264252e-05,
      "loss": 2.2327081680297853,
      "memory(GiB)": 77.56,
      "step": 78610,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.436759
    },
    {
      "epoch": 3.3681076217814145,
      "grad_norm": 7.226221084594727,
      "learning_rate": 2.405993871904686e-05,
      "loss": 2.241598892211914,
      "memory(GiB)": 77.56,
      "step": 78615,
      "token_acc": 0.5201465201465202,
      "train_speed(iter/s)": 1.436768
    },
    {
      "epoch": 3.368321837110664,
      "grad_norm": 6.602431774139404,
      "learning_rate": 2.4054185705379724e-05,
      "loss": 2.6349117279052736,
      "memory(GiB)": 77.56,
      "step": 78620,
      "token_acc": 0.4644808743169399,
      "train_speed(iter/s)": 1.436767
    },
    {
      "epoch": 3.3685360524399126,
      "grad_norm": 8.15915298461914,
      "learning_rate": 2.404843316174537e-05,
      "loss": 1.9415018081665039,
      "memory(GiB)": 77.56,
      "step": 78625,
      "token_acc": 0.5535714285714286,
      "train_speed(iter/s)": 1.436766
    },
    {
      "epoch": 3.3687502677691614,
      "grad_norm": 6.5156779289245605,
      "learning_rate": 2.4042681088248e-05,
      "loss": 2.0290224075317385,
      "memory(GiB)": 77.56,
      "step": 78630,
      "token_acc": 0.5708661417322834,
      "train_speed(iter/s)": 1.43678
    },
    {
      "epoch": 3.3689644830984107,
      "grad_norm": 5.561295986175537,
      "learning_rate": 2.4036929484991804e-05,
      "loss": 2.150686836242676,
      "memory(GiB)": 77.56,
      "step": 78635,
      "token_acc": 0.5412541254125413,
      "train_speed(iter/s)": 1.436797
    },
    {
      "epoch": 3.3691786984276595,
      "grad_norm": 6.023956298828125,
      "learning_rate": 2.4031178352080992e-05,
      "loss": 2.2956975936889648,
      "memory(GiB)": 77.56,
      "step": 78640,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436797
    },
    {
      "epoch": 3.3693929137569083,
      "grad_norm": 4.972743034362793,
      "learning_rate": 2.402542768961974e-05,
      "loss": 2.4439579010009767,
      "memory(GiB)": 77.56,
      "step": 78645,
      "token_acc": 0.5392156862745098,
      "train_speed(iter/s)": 1.436792
    },
    {
      "epoch": 3.3696071290861576,
      "grad_norm": 6.531024932861328,
      "learning_rate": 2.4019677497712216e-05,
      "loss": 2.3612098693847656,
      "memory(GiB)": 77.56,
      "step": 78650,
      "token_acc": 0.4794952681388013,
      "train_speed(iter/s)": 1.436808
    },
    {
      "epoch": 3.3698213444154064,
      "grad_norm": 4.327822208404541,
      "learning_rate": 2.4013927776462625e-05,
      "loss": 2.5188278198242187,
      "memory(GiB)": 77.56,
      "step": 78655,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.436808
    },
    {
      "epoch": 3.370035559744655,
      "grad_norm": 5.676511764526367,
      "learning_rate": 2.4008178525975105e-05,
      "loss": 2.563261795043945,
      "memory(GiB)": 77.56,
      "step": 78660,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.436799
    },
    {
      "epoch": 3.3702497750739044,
      "grad_norm": 5.0904459953308105,
      "learning_rate": 2.4002429746353817e-05,
      "loss": 2.187174606323242,
      "memory(GiB)": 77.56,
      "step": 78665,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.436803
    },
    {
      "epoch": 3.3704639904031533,
      "grad_norm": 6.471100807189941,
      "learning_rate": 2.39966814377029e-05,
      "loss": 2.420202445983887,
      "memory(GiB)": 77.56,
      "step": 78670,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.436821
    },
    {
      "epoch": 3.370678205732402,
      "grad_norm": 5.1567702293396,
      "learning_rate": 2.3990933600126476e-05,
      "loss": 2.4047405242919924,
      "memory(GiB)": 77.56,
      "step": 78675,
      "token_acc": 0.49693251533742333,
      "train_speed(iter/s)": 1.436819
    },
    {
      "epoch": 3.3708924210616513,
      "grad_norm": 6.0861334800720215,
      "learning_rate": 2.3985186233728686e-05,
      "loss": 1.9490814208984375,
      "memory(GiB)": 77.56,
      "step": 78680,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.436843
    },
    {
      "epoch": 3.3711066363909,
      "grad_norm": 5.486029148101807,
      "learning_rate": 2.3979439338613668e-05,
      "loss": 2.255574035644531,
      "memory(GiB)": 77.56,
      "step": 78685,
      "token_acc": 0.5401459854014599,
      "train_speed(iter/s)": 1.436862
    },
    {
      "epoch": 3.371320851720149,
      "grad_norm": 6.144575119018555,
      "learning_rate": 2.397369291488552e-05,
      "loss": 2.4148992538452148,
      "memory(GiB)": 77.56,
      "step": 78690,
      "token_acc": 0.4774436090225564,
      "train_speed(iter/s)": 1.436874
    },
    {
      "epoch": 3.371535067049398,
      "grad_norm": 6.161258697509766,
      "learning_rate": 2.3967946962648334e-05,
      "loss": 2.2431365966796877,
      "memory(GiB)": 77.56,
      "step": 78695,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.436877
    },
    {
      "epoch": 3.371749282378647,
      "grad_norm": 5.4289703369140625,
      "learning_rate": 2.3962201482006215e-05,
      "loss": 2.270370292663574,
      "memory(GiB)": 77.56,
      "step": 78700,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436874
    },
    {
      "epoch": 3.371963497707896,
      "grad_norm": 5.3703436851501465,
      "learning_rate": 2.395645647306324e-05,
      "loss": 2.1858808517456056,
      "memory(GiB)": 77.56,
      "step": 78705,
      "token_acc": 0.5443425076452599,
      "train_speed(iter/s)": 1.436871
    },
    {
      "epoch": 3.372177713037145,
      "grad_norm": 7.366776466369629,
      "learning_rate": 2.3950711935923466e-05,
      "loss": 2.2356048583984376,
      "memory(GiB)": 77.56,
      "step": 78710,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.43687
    },
    {
      "epoch": 3.372391928366394,
      "grad_norm": 5.853217601776123,
      "learning_rate": 2.3944967870691003e-05,
      "loss": 2.356764221191406,
      "memory(GiB)": 77.56,
      "step": 78715,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.436894
    },
    {
      "epoch": 3.3726061436956427,
      "grad_norm": 7.239072799682617,
      "learning_rate": 2.3939224277469886e-05,
      "loss": 2.3769832611083985,
      "memory(GiB)": 77.56,
      "step": 78720,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.4369
    },
    {
      "epoch": 3.372820359024892,
      "grad_norm": 5.061244964599609,
      "learning_rate": 2.3933481156364168e-05,
      "loss": 2.1726757049560548,
      "memory(GiB)": 77.56,
      "step": 78725,
      "token_acc": 0.5395189003436426,
      "train_speed(iter/s)": 1.436909
    },
    {
      "epoch": 3.373034574354141,
      "grad_norm": 5.375277042388916,
      "learning_rate": 2.392773850747789e-05,
      "loss": 2.148880386352539,
      "memory(GiB)": 77.56,
      "step": 78730,
      "token_acc": 0.5095785440613027,
      "train_speed(iter/s)": 1.436901
    },
    {
      "epoch": 3.3732487896833896,
      "grad_norm": 5.191781520843506,
      "learning_rate": 2.3921996330915076e-05,
      "loss": 2.4782032012939452,
      "memory(GiB)": 77.56,
      "step": 78735,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.43688
    },
    {
      "epoch": 3.373463005012639,
      "grad_norm": 8.026511192321777,
      "learning_rate": 2.391625462677977e-05,
      "loss": 2.4119050979614256,
      "memory(GiB)": 77.56,
      "step": 78740,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.436898
    },
    {
      "epoch": 3.3736772203418877,
      "grad_norm": 6.124702453613281,
      "learning_rate": 2.3910513395175988e-05,
      "loss": 2.4365854263305664,
      "memory(GiB)": 77.56,
      "step": 78745,
      "token_acc": 0.43197278911564624,
      "train_speed(iter/s)": 1.436896
    },
    {
      "epoch": 3.3738914356711365,
      "grad_norm": 6.308697700500488,
      "learning_rate": 2.3904772636207723e-05,
      "loss": 2.259089469909668,
      "memory(GiB)": 77.56,
      "step": 78750,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.436903
    },
    {
      "epoch": 3.3741056510003857,
      "grad_norm": 5.427739143371582,
      "learning_rate": 2.3899032349978967e-05,
      "loss": 2.1422523498535155,
      "memory(GiB)": 77.56,
      "step": 78755,
      "token_acc": 0.5617529880478087,
      "train_speed(iter/s)": 1.436907
    },
    {
      "epoch": 3.3743198663296345,
      "grad_norm": 4.6939215660095215,
      "learning_rate": 2.389329253659374e-05,
      "loss": 2.2956966400146483,
      "memory(GiB)": 77.56,
      "step": 78760,
      "token_acc": 0.5350553505535055,
      "train_speed(iter/s)": 1.436901
    },
    {
      "epoch": 3.3745340816588834,
      "grad_norm": 6.2993316650390625,
      "learning_rate": 2.3887553196155995e-05,
      "loss": 2.2858009338378906,
      "memory(GiB)": 77.56,
      "step": 78765,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.43689
    },
    {
      "epoch": 3.3747482969881326,
      "grad_norm": 6.204442501068115,
      "learning_rate": 2.3881814328769737e-05,
      "loss": 2.245798873901367,
      "memory(GiB)": 77.56,
      "step": 78770,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.436892
    },
    {
      "epoch": 3.3749625123173814,
      "grad_norm": 5.690382957458496,
      "learning_rate": 2.387607593453891e-05,
      "loss": 2.323918914794922,
      "memory(GiB)": 77.56,
      "step": 78775,
      "token_acc": 0.4967532467532468,
      "train_speed(iter/s)": 1.436896
    },
    {
      "epoch": 3.3751767276466302,
      "grad_norm": 4.640789985656738,
      "learning_rate": 2.3870338013567474e-05,
      "loss": 2.3230365753173827,
      "memory(GiB)": 77.56,
      "step": 78780,
      "token_acc": 0.5273972602739726,
      "train_speed(iter/s)": 1.43691
    },
    {
      "epoch": 3.3753909429758795,
      "grad_norm": 5.124710559844971,
      "learning_rate": 2.3864600565959377e-05,
      "loss": 2.3819780349731445,
      "memory(GiB)": 77.56,
      "step": 78785,
      "token_acc": 0.5068870523415978,
      "train_speed(iter/s)": 1.436912
    },
    {
      "epoch": 3.3756051583051283,
      "grad_norm": 5.971935272216797,
      "learning_rate": 2.3858863591818558e-05,
      "loss": 2.1204200744628907,
      "memory(GiB)": 77.56,
      "step": 78790,
      "token_acc": 0.5396341463414634,
      "train_speed(iter/s)": 1.436887
    },
    {
      "epoch": 3.375819373634377,
      "grad_norm": 6.131096839904785,
      "learning_rate": 2.385312709124893e-05,
      "loss": 2.3459039688110352,
      "memory(GiB)": 77.56,
      "step": 78795,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.436906
    },
    {
      "epoch": 3.3760335889636264,
      "grad_norm": 6.3354058265686035,
      "learning_rate": 2.3847391064354453e-05,
      "loss": 2.379328727722168,
      "memory(GiB)": 77.56,
      "step": 78800,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.436895
    },
    {
      "epoch": 3.376247804292875,
      "grad_norm": 6.278886318206787,
      "learning_rate": 2.3841655511239013e-05,
      "loss": 2.1081916809082033,
      "memory(GiB)": 77.56,
      "step": 78805,
      "token_acc": 0.5785123966942148,
      "train_speed(iter/s)": 1.436888
    },
    {
      "epoch": 3.376462019622124,
      "grad_norm": 8.349748611450195,
      "learning_rate": 2.3835920432006527e-05,
      "loss": 2.3201984405517577,
      "memory(GiB)": 77.56,
      "step": 78810,
      "token_acc": 0.48627450980392156,
      "train_speed(iter/s)": 1.4369
    },
    {
      "epoch": 3.3766762349513733,
      "grad_norm": 6.222157955169678,
      "learning_rate": 2.3830185826760887e-05,
      "loss": 2.1716976165771484,
      "memory(GiB)": 77.56,
      "step": 78815,
      "token_acc": 0.5239852398523985,
      "train_speed(iter/s)": 1.436911
    },
    {
      "epoch": 3.376890450280622,
      "grad_norm": 5.05953311920166,
      "learning_rate": 2.3824451695605958e-05,
      "loss": 2.503479766845703,
      "memory(GiB)": 77.56,
      "step": 78820,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.43692
    },
    {
      "epoch": 3.377104665609871,
      "grad_norm": 6.064471244812012,
      "learning_rate": 2.381871803864566e-05,
      "loss": 2.3991886138916017,
      "memory(GiB)": 77.56,
      "step": 78825,
      "token_acc": 0.5250836120401338,
      "train_speed(iter/s)": 1.436921
    },
    {
      "epoch": 3.37731888093912,
      "grad_norm": 6.053960800170898,
      "learning_rate": 2.381298485598383e-05,
      "loss": 2.5158113479614257,
      "memory(GiB)": 77.56,
      "step": 78830,
      "token_acc": 0.4879032258064516,
      "train_speed(iter/s)": 1.436906
    },
    {
      "epoch": 3.377533096268369,
      "grad_norm": 6.7615814208984375,
      "learning_rate": 2.3807252147724362e-05,
      "loss": 2.529646873474121,
      "memory(GiB)": 77.56,
      "step": 78835,
      "token_acc": 0.48727272727272725,
      "train_speed(iter/s)": 1.436912
    },
    {
      "epoch": 3.3777473115976178,
      "grad_norm": 6.478045463562012,
      "learning_rate": 2.380151991397109e-05,
      "loss": 2.299068832397461,
      "memory(GiB)": 77.56,
      "step": 78840,
      "token_acc": 0.5259515570934256,
      "train_speed(iter/s)": 1.436934
    },
    {
      "epoch": 3.377961526926867,
      "grad_norm": 6.816653728485107,
      "learning_rate": 2.3795788154827862e-05,
      "loss": 2.3651493072509764,
      "memory(GiB)": 77.56,
      "step": 78845,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.436938
    },
    {
      "epoch": 3.378175742256116,
      "grad_norm": 5.127247333526611,
      "learning_rate": 2.3790056870398515e-05,
      "loss": 2.6714509963989257,
      "memory(GiB)": 77.56,
      "step": 78850,
      "token_acc": 0.4161290322580645,
      "train_speed(iter/s)": 1.436958
    },
    {
      "epoch": 3.3783899575853646,
      "grad_norm": 6.059878826141357,
      "learning_rate": 2.3784326060786855e-05,
      "loss": 2.464638519287109,
      "memory(GiB)": 77.56,
      "step": 78855,
      "token_acc": 0.4948805460750853,
      "train_speed(iter/s)": 1.43697
    },
    {
      "epoch": 3.378604172914614,
      "grad_norm": 7.489520072937012,
      "learning_rate": 2.3778595726096737e-05,
      "loss": 2.5586109161376953,
      "memory(GiB)": 77.56,
      "step": 78860,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.436979
    },
    {
      "epoch": 3.3788183882438627,
      "grad_norm": 6.65469217300415,
      "learning_rate": 2.3772865866431955e-05,
      "loss": 1.8535985946655273,
      "memory(GiB)": 77.56,
      "step": 78865,
      "token_acc": 0.594488188976378,
      "train_speed(iter/s)": 1.436989
    },
    {
      "epoch": 3.3790326035731115,
      "grad_norm": 5.638848781585693,
      "learning_rate": 2.3767136481896312e-05,
      "loss": 2.201658821105957,
      "memory(GiB)": 77.56,
      "step": 78870,
      "token_acc": 0.5124223602484472,
      "train_speed(iter/s)": 1.437
    },
    {
      "epoch": 3.379246818902361,
      "grad_norm": 6.5392889976501465,
      "learning_rate": 2.3761407572593603e-05,
      "loss": 2.2582328796386717,
      "memory(GiB)": 77.56,
      "step": 78875,
      "token_acc": 0.5461254612546126,
      "train_speed(iter/s)": 1.437007
    },
    {
      "epoch": 3.3794610342316096,
      "grad_norm": 5.483560562133789,
      "learning_rate": 2.375567913862759e-05,
      "loss": 2.305634689331055,
      "memory(GiB)": 77.56,
      "step": 78880,
      "token_acc": 0.5224358974358975,
      "train_speed(iter/s)": 1.437
    },
    {
      "epoch": 3.3796752495608584,
      "grad_norm": 6.366110801696777,
      "learning_rate": 2.3749951180102082e-05,
      "loss": 2.267227554321289,
      "memory(GiB)": 77.56,
      "step": 78885,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.436993
    },
    {
      "epoch": 3.3798894648901077,
      "grad_norm": 6.813939094543457,
      "learning_rate": 2.3744223697120836e-05,
      "loss": 2.3075139999389647,
      "memory(GiB)": 77.56,
      "step": 78890,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.436986
    },
    {
      "epoch": 3.3801036802193565,
      "grad_norm": 5.68144416809082,
      "learning_rate": 2.373849668978761e-05,
      "loss": 2.6016822814941407,
      "memory(GiB)": 77.56,
      "step": 78895,
      "token_acc": 0.4854771784232365,
      "train_speed(iter/s)": 1.436987
    },
    {
      "epoch": 3.3803178955486053,
      "grad_norm": 6.739671230316162,
      "learning_rate": 2.373277015820613e-05,
      "loss": 2.3951135635375977,
      "memory(GiB)": 77.56,
      "step": 78900,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.43699
    },
    {
      "epoch": 3.3805321108778545,
      "grad_norm": 6.346278190612793,
      "learning_rate": 2.3727044102480184e-05,
      "loss": 2.5380760192871095,
      "memory(GiB)": 77.56,
      "step": 78905,
      "token_acc": 0.4786885245901639,
      "train_speed(iter/s)": 1.436998
    },
    {
      "epoch": 3.3807463262071034,
      "grad_norm": 6.784225940704346,
      "learning_rate": 2.3721318522713453e-05,
      "loss": 2.1891901016235353,
      "memory(GiB)": 77.56,
      "step": 78910,
      "token_acc": 0.5527272727272727,
      "train_speed(iter/s)": 1.43701
    },
    {
      "epoch": 3.380960541536352,
      "grad_norm": 6.588382720947266,
      "learning_rate": 2.3715593419009714e-05,
      "loss": 2.473220443725586,
      "memory(GiB)": 77.56,
      "step": 78915,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.437022
    },
    {
      "epoch": 3.3811747568656014,
      "grad_norm": 5.927586555480957,
      "learning_rate": 2.3709868791472652e-05,
      "loss": 2.267233657836914,
      "memory(GiB)": 77.56,
      "step": 78920,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.437018
    },
    {
      "epoch": 3.3813889721948502,
      "grad_norm": 5.045629024505615,
      "learning_rate": 2.3704144640205983e-05,
      "loss": 2.1847902297973634,
      "memory(GiB)": 77.56,
      "step": 78925,
      "token_acc": 0.5515873015873016,
      "train_speed(iter/s)": 1.437031
    },
    {
      "epoch": 3.381603187524099,
      "grad_norm": 5.2909722328186035,
      "learning_rate": 2.3698420965313395e-05,
      "loss": 2.217837905883789,
      "memory(GiB)": 77.56,
      "step": 78930,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.437019
    },
    {
      "epoch": 3.3818174028533483,
      "grad_norm": 8.216493606567383,
      "learning_rate": 2.3692697766898592e-05,
      "loss": 2.493809127807617,
      "memory(GiB)": 77.56,
      "step": 78935,
      "token_acc": 0.5132743362831859,
      "train_speed(iter/s)": 1.437021
    },
    {
      "epoch": 3.382031618182597,
      "grad_norm": 4.826843738555908,
      "learning_rate": 2.3686975045065223e-05,
      "loss": 2.0823545455932617,
      "memory(GiB)": 77.56,
      "step": 78940,
      "token_acc": 0.5273224043715847,
      "train_speed(iter/s)": 1.437027
    },
    {
      "epoch": 3.382245833511846,
      "grad_norm": 4.023681163787842,
      "learning_rate": 2.3681252799917002e-05,
      "loss": 1.9837207794189453,
      "memory(GiB)": 77.56,
      "step": 78945,
      "token_acc": 0.5648854961832062,
      "train_speed(iter/s)": 1.437044
    },
    {
      "epoch": 3.382460048841095,
      "grad_norm": 4.9561991691589355,
      "learning_rate": 2.367553103155758e-05,
      "loss": 2.5363336563110352,
      "memory(GiB)": 77.56,
      "step": 78950,
      "token_acc": 0.49709302325581395,
      "train_speed(iter/s)": 1.437053
    },
    {
      "epoch": 3.382674264170344,
      "grad_norm": 5.4621381759643555,
      "learning_rate": 2.366980974009061e-05,
      "loss": 2.503839683532715,
      "memory(GiB)": 77.56,
      "step": 78955,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.437074
    },
    {
      "epoch": 3.382888479499593,
      "grad_norm": 5.743004322052002,
      "learning_rate": 2.3664088925619732e-05,
      "loss": 2.3493316650390623,
      "memory(GiB)": 77.56,
      "step": 78960,
      "token_acc": 0.5050847457627119,
      "train_speed(iter/s)": 1.43708
    },
    {
      "epoch": 3.383102694828842,
      "grad_norm": 6.767465114593506,
      "learning_rate": 2.365836858824857e-05,
      "loss": 2.1151844024658204,
      "memory(GiB)": 77.56,
      "step": 78965,
      "token_acc": 0.525096525096525,
      "train_speed(iter/s)": 1.437071
    },
    {
      "epoch": 3.383316910158091,
      "grad_norm": 6.905044078826904,
      "learning_rate": 2.365264872808079e-05,
      "loss": 2.4366338729858397,
      "memory(GiB)": 77.56,
      "step": 78970,
      "token_acc": 0.48627450980392156,
      "train_speed(iter/s)": 1.437083
    },
    {
      "epoch": 3.3835311254873397,
      "grad_norm": 6.050512790679932,
      "learning_rate": 2.3646929345219975e-05,
      "loss": 2.303232955932617,
      "memory(GiB)": 77.56,
      "step": 78975,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.437084
    },
    {
      "epoch": 3.383745340816589,
      "grad_norm": 6.511990070343018,
      "learning_rate": 2.3641210439769773e-05,
      "loss": 2.374123764038086,
      "memory(GiB)": 77.56,
      "step": 78980,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.437084
    },
    {
      "epoch": 3.3839595561458378,
      "grad_norm": 6.069528579711914,
      "learning_rate": 2.3635492011833778e-05,
      "loss": 2.2651861190795897,
      "memory(GiB)": 77.56,
      "step": 78985,
      "token_acc": 0.44722222222222224,
      "train_speed(iter/s)": 1.437086
    },
    {
      "epoch": 3.3841737714750866,
      "grad_norm": 5.74617862701416,
      "learning_rate": 2.362977406151557e-05,
      "loss": 2.408353805541992,
      "memory(GiB)": 77.56,
      "step": 78990,
      "token_acc": 0.46774193548387094,
      "train_speed(iter/s)": 1.437097
    },
    {
      "epoch": 3.384387986804336,
      "grad_norm": 6.989368915557861,
      "learning_rate": 2.362405658891874e-05,
      "loss": 2.538044548034668,
      "memory(GiB)": 77.56,
      "step": 78995,
      "token_acc": 0.43902439024390244,
      "train_speed(iter/s)": 1.437092
    },
    {
      "epoch": 3.3846022021335846,
      "grad_norm": 4.436270236968994,
      "learning_rate": 2.3618339594146853e-05,
      "loss": 2.15976619720459,
      "memory(GiB)": 77.56,
      "step": 79000,
      "token_acc": 0.5468164794007491,
      "train_speed(iter/s)": 1.437093
    },
    {
      "epoch": 3.3846022021335846,
      "eval_loss": 2.242905378341675,
      "eval_runtime": 14.5098,
      "eval_samples_per_second": 6.892,
      "eval_steps_per_second": 6.892,
      "eval_token_acc": 0.469281045751634,
      "step": 79000
    },
    {
      "epoch": 3.3848164174628335,
      "grad_norm": 5.641612529754639,
      "learning_rate": 2.3612623077303514e-05,
      "loss": 2.408580207824707,
      "memory(GiB)": 77.56,
      "step": 79005,
      "token_acc": 0.4895833333333333,
      "train_speed(iter/s)": 1.436687
    },
    {
      "epoch": 3.3850306327920827,
      "grad_norm": 5.722011089324951,
      "learning_rate": 2.360690703849226e-05,
      "loss": 2.2639083862304688,
      "memory(GiB)": 77.56,
      "step": 79010,
      "token_acc": 0.47651006711409394,
      "train_speed(iter/s)": 1.43668
    },
    {
      "epoch": 3.3852448481213315,
      "grad_norm": 6.5511980056762695,
      "learning_rate": 2.360119147781664e-05,
      "loss": 2.4920969009399414,
      "memory(GiB)": 77.56,
      "step": 79015,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.436695
    },
    {
      "epoch": 3.3854590634505803,
      "grad_norm": 4.943918228149414,
      "learning_rate": 2.35954763953802e-05,
      "loss": 2.2495140075683593,
      "memory(GiB)": 77.56,
      "step": 79020,
      "token_acc": 0.5234657039711191,
      "train_speed(iter/s)": 1.436723
    },
    {
      "epoch": 3.3856732787798296,
      "grad_norm": 5.779406547546387,
      "learning_rate": 2.3589761791286462e-05,
      "loss": 2.338222694396973,
      "memory(GiB)": 77.56,
      "step": 79025,
      "token_acc": 0.4969512195121951,
      "train_speed(iter/s)": 1.436738
    },
    {
      "epoch": 3.3858874941090784,
      "grad_norm": 4.647282123565674,
      "learning_rate": 2.3584047665638977e-05,
      "loss": 2.040752983093262,
      "memory(GiB)": 77.56,
      "step": 79030,
      "token_acc": 0.5430711610486891,
      "train_speed(iter/s)": 1.436763
    },
    {
      "epoch": 3.3861017094383272,
      "grad_norm": 4.793888568878174,
      "learning_rate": 2.3578334018541254e-05,
      "loss": 2.2106000900268556,
      "memory(GiB)": 77.56,
      "step": 79035,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.436792
    },
    {
      "epoch": 3.3863159247675765,
      "grad_norm": 6.412319183349609,
      "learning_rate": 2.3572620850096787e-05,
      "loss": 2.2116941452026366,
      "memory(GiB)": 77.56,
      "step": 79040,
      "token_acc": 0.55078125,
      "train_speed(iter/s)": 1.436776
    },
    {
      "epoch": 3.3865301400968253,
      "grad_norm": 5.8349995613098145,
      "learning_rate": 2.356690816040909e-05,
      "loss": 2.4938520431518554,
      "memory(GiB)": 77.56,
      "step": 79045,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.436785
    },
    {
      "epoch": 3.386744355426074,
      "grad_norm": 6.164633274078369,
      "learning_rate": 2.3561195949581634e-05,
      "loss": 2.3005395889282227,
      "memory(GiB)": 77.56,
      "step": 79050,
      "token_acc": 0.4854014598540146,
      "train_speed(iter/s)": 1.436786
    },
    {
      "epoch": 3.3869585707553234,
      "grad_norm": 7.676568031311035,
      "learning_rate": 2.3555484217717904e-05,
      "loss": 2.0836063385009767,
      "memory(GiB)": 77.56,
      "step": 79055,
      "token_acc": 0.5686274509803921,
      "train_speed(iter/s)": 1.436789
    },
    {
      "epoch": 3.387172786084572,
      "grad_norm": 6.290713787078857,
      "learning_rate": 2.3549772964921413e-05,
      "loss": 2.061748504638672,
      "memory(GiB)": 77.56,
      "step": 79060,
      "token_acc": 0.5567765567765568,
      "train_speed(iter/s)": 1.436802
    },
    {
      "epoch": 3.387387001413821,
      "grad_norm": 5.1710333824157715,
      "learning_rate": 2.354406219129559e-05,
      "loss": 2.359303092956543,
      "memory(GiB)": 77.56,
      "step": 79065,
      "token_acc": 0.5340501792114696,
      "train_speed(iter/s)": 1.436802
    },
    {
      "epoch": 3.3876012167430702,
      "grad_norm": 5.794148921966553,
      "learning_rate": 2.35383518969439e-05,
      "loss": 2.0363449096679687,
      "memory(GiB)": 77.56,
      "step": 79070,
      "token_acc": 0.5734265734265734,
      "train_speed(iter/s)": 1.436807
    },
    {
      "epoch": 3.387815432072319,
      "grad_norm": 7.266098976135254,
      "learning_rate": 2.353264208196979e-05,
      "loss": 2.161484146118164,
      "memory(GiB)": 77.56,
      "step": 79075,
      "token_acc": 0.5329153605015674,
      "train_speed(iter/s)": 1.436788
    },
    {
      "epoch": 3.388029647401568,
      "grad_norm": 7.2324395179748535,
      "learning_rate": 2.35269327464767e-05,
      "loss": 2.4642690658569335,
      "memory(GiB)": 77.56,
      "step": 79080,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.436778
    },
    {
      "epoch": 3.388243862730817,
      "grad_norm": 5.676534652709961,
      "learning_rate": 2.3521223890568032e-05,
      "loss": 2.6387828826904296,
      "memory(GiB)": 77.56,
      "step": 79085,
      "token_acc": 0.441747572815534,
      "train_speed(iter/s)": 1.436793
    },
    {
      "epoch": 3.388458078060066,
      "grad_norm": 5.013236999511719,
      "learning_rate": 2.3515515514347252e-05,
      "loss": 2.474356460571289,
      "memory(GiB)": 77.56,
      "step": 79090,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.436779
    },
    {
      "epoch": 3.3886722933893147,
      "grad_norm": 6.498418807983398,
      "learning_rate": 2.3509807617917757e-05,
      "loss": 2.0500114440917967,
      "memory(GiB)": 77.56,
      "step": 79095,
      "token_acc": 0.5402298850574713,
      "train_speed(iter/s)": 1.436795
    },
    {
      "epoch": 3.388886508718564,
      "grad_norm": 5.321521282196045,
      "learning_rate": 2.3504100201382945e-05,
      "loss": 2.3788131713867187,
      "memory(GiB)": 77.56,
      "step": 79100,
      "token_acc": 0.48135593220338985,
      "train_speed(iter/s)": 1.43678
    },
    {
      "epoch": 3.389100724047813,
      "grad_norm": 5.2803544998168945,
      "learning_rate": 2.3498393264846212e-05,
      "loss": 2.418195343017578,
      "memory(GiB)": 77.56,
      "step": 79105,
      "token_acc": 0.48562300319488816,
      "train_speed(iter/s)": 1.436791
    },
    {
      "epoch": 3.3893149393770616,
      "grad_norm": 6.558969974517822,
      "learning_rate": 2.349268680841093e-05,
      "loss": 2.2810951232910157,
      "memory(GiB)": 77.56,
      "step": 79110,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436816
    },
    {
      "epoch": 3.389529154706311,
      "grad_norm": 7.498997688293457,
      "learning_rate": 2.3486980832180505e-05,
      "loss": 2.3022176742553713,
      "memory(GiB)": 77.56,
      "step": 79115,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.436839
    },
    {
      "epoch": 3.3897433700355597,
      "grad_norm": 5.172027111053467,
      "learning_rate": 2.34812753362583e-05,
      "loss": 2.319327735900879,
      "memory(GiB)": 77.56,
      "step": 79120,
      "token_acc": 0.5507246376811594,
      "train_speed(iter/s)": 1.436828
    },
    {
      "epoch": 3.3899575853648085,
      "grad_norm": 6.7108154296875,
      "learning_rate": 2.3475570320747647e-05,
      "loss": 2.295004653930664,
      "memory(GiB)": 77.56,
      "step": 79125,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436824
    },
    {
      "epoch": 3.3901718006940578,
      "grad_norm": 6.020142555236816,
      "learning_rate": 2.3469865785751938e-05,
      "loss": 2.1553855895996095,
      "memory(GiB)": 77.56,
      "step": 79130,
      "token_acc": 0.4826254826254826,
      "train_speed(iter/s)": 1.436832
    },
    {
      "epoch": 3.3903860160233066,
      "grad_norm": 4.676198959350586,
      "learning_rate": 2.3464161731374496e-05,
      "loss": 2.272721862792969,
      "memory(GiB)": 77.56,
      "step": 79135,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.436829
    },
    {
      "epoch": 3.390600231352556,
      "grad_norm": 5.459369659423828,
      "learning_rate": 2.345845815771866e-05,
      "loss": 2.2400577545166014,
      "memory(GiB)": 77.56,
      "step": 79140,
      "token_acc": 0.49137931034482757,
      "train_speed(iter/s)": 1.436834
    },
    {
      "epoch": 3.3908144466818047,
      "grad_norm": 6.605140209197998,
      "learning_rate": 2.3452755064887732e-05,
      "loss": 2.378836441040039,
      "memory(GiB)": 77.56,
      "step": 79145,
      "token_acc": 0.48773006134969327,
      "train_speed(iter/s)": 1.436843
    },
    {
      "epoch": 3.3910286620110535,
      "grad_norm": 7.440648555755615,
      "learning_rate": 2.3447052452985068e-05,
      "loss": 2.395745849609375,
      "memory(GiB)": 77.56,
      "step": 79150,
      "token_acc": 0.49216300940438873,
      "train_speed(iter/s)": 1.436846
    },
    {
      "epoch": 3.3912428773403027,
      "grad_norm": 6.153088092803955,
      "learning_rate": 2.3441350322113957e-05,
      "loss": 2.0995552062988283,
      "memory(GiB)": 77.56,
      "step": 79155,
      "token_acc": 0.5402298850574713,
      "train_speed(iter/s)": 1.43685
    },
    {
      "epoch": 3.3914570926695515,
      "grad_norm": 5.468713760375977,
      "learning_rate": 2.3435648672377702e-05,
      "loss": 2.259908676147461,
      "memory(GiB)": 77.56,
      "step": 79160,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.436865
    },
    {
      "epoch": 3.3916713079988003,
      "grad_norm": 6.328258514404297,
      "learning_rate": 2.342994750387959e-05,
      "loss": 2.4397909164428713,
      "memory(GiB)": 77.56,
      "step": 79165,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.436888
    },
    {
      "epoch": 3.3918855233280496,
      "grad_norm": 6.460422515869141,
      "learning_rate": 2.3424246816722884e-05,
      "loss": 2.6259462356567385,
      "memory(GiB)": 77.56,
      "step": 79170,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.436911
    },
    {
      "epoch": 3.3920997386572984,
      "grad_norm": 5.835700988769531,
      "learning_rate": 2.3418546611010895e-05,
      "loss": 2.2085365295410155,
      "memory(GiB)": 77.56,
      "step": 79175,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.436923
    },
    {
      "epoch": 3.3923139539865472,
      "grad_norm": 5.2655181884765625,
      "learning_rate": 2.3412846886846867e-05,
      "loss": 2.112788200378418,
      "memory(GiB)": 77.56,
      "step": 79180,
      "token_acc": 0.556420233463035,
      "train_speed(iter/s)": 1.436935
    },
    {
      "epoch": 3.3925281693157965,
      "grad_norm": 5.313553333282471,
      "learning_rate": 2.3407147644334067e-05,
      "loss": 2.480325126647949,
      "memory(GiB)": 77.56,
      "step": 79185,
      "token_acc": 0.46545454545454545,
      "train_speed(iter/s)": 1.436939
    },
    {
      "epoch": 3.3927423846450453,
      "grad_norm": 5.882853984832764,
      "learning_rate": 2.340144888357572e-05,
      "loss": 2.2125625610351562,
      "memory(GiB)": 77.56,
      "step": 79190,
      "token_acc": 0.5292096219931272,
      "train_speed(iter/s)": 1.436937
    },
    {
      "epoch": 3.392956599974294,
      "grad_norm": 7.756390571594238,
      "learning_rate": 2.339575060467507e-05,
      "loss": 2.259334754943848,
      "memory(GiB)": 77.56,
      "step": 79195,
      "token_acc": 0.5813953488372093,
      "train_speed(iter/s)": 1.436899
    },
    {
      "epoch": 3.3931708153035434,
      "grad_norm": 7.951709270477295,
      "learning_rate": 2.3390052807735352e-05,
      "loss": 2.2962581634521486,
      "memory(GiB)": 77.56,
      "step": 79200,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436896
    },
    {
      "epoch": 3.393385030632792,
      "grad_norm": 6.02262544631958,
      "learning_rate": 2.338435549285981e-05,
      "loss": 2.3636016845703125,
      "memory(GiB)": 77.56,
      "step": 79205,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.436891
    },
    {
      "epoch": 3.393599245962041,
      "grad_norm": 6.503762722015381,
      "learning_rate": 2.337865866015163e-05,
      "loss": 2.3004405975341795,
      "memory(GiB)": 77.56,
      "step": 79210,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436926
    },
    {
      "epoch": 3.3938134612912902,
      "grad_norm": 4.864107608795166,
      "learning_rate": 2.3372962309714023e-05,
      "loss": 2.367596435546875,
      "memory(GiB)": 77.56,
      "step": 79215,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.436934
    },
    {
      "epoch": 3.394027676620539,
      "grad_norm": 7.189845561981201,
      "learning_rate": 2.3367266441650188e-05,
      "loss": 2.5482656478881838,
      "memory(GiB)": 77.56,
      "step": 79220,
      "token_acc": 0.5064516129032258,
      "train_speed(iter/s)": 1.436948
    },
    {
      "epoch": 3.394241891949788,
      "grad_norm": 8.789438247680664,
      "learning_rate": 2.3361571056063302e-05,
      "loss": 2.2059791564941404,
      "memory(GiB)": 77.56,
      "step": 79225,
      "token_acc": 0.5363321799307958,
      "train_speed(iter/s)": 1.436946
    },
    {
      "epoch": 3.394456107279037,
      "grad_norm": 5.332568645477295,
      "learning_rate": 2.335587615305652e-05,
      "loss": 2.3448369979858397,
      "memory(GiB)": 77.56,
      "step": 79230,
      "token_acc": 0.5020408163265306,
      "train_speed(iter/s)": 1.436941
    },
    {
      "epoch": 3.394670322608286,
      "grad_norm": 5.452345371246338,
      "learning_rate": 2.335018173273306e-05,
      "loss": 2.240138626098633,
      "memory(GiB)": 77.56,
      "step": 79235,
      "token_acc": 0.5096153846153846,
      "train_speed(iter/s)": 1.43694
    },
    {
      "epoch": 3.3948845379375348,
      "grad_norm": 6.187854290008545,
      "learning_rate": 2.3344487795196063e-05,
      "loss": 2.2077091217041014,
      "memory(GiB)": 77.56,
      "step": 79240,
      "token_acc": 0.5236486486486487,
      "train_speed(iter/s)": 1.43695
    },
    {
      "epoch": 3.395098753266784,
      "grad_norm": 6.3480305671691895,
      "learning_rate": 2.3338794340548666e-05,
      "loss": 2.5340810775756837,
      "memory(GiB)": 77.56,
      "step": 79245,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.436957
    },
    {
      "epoch": 3.395312968596033,
      "grad_norm": 5.4756059646606445,
      "learning_rate": 2.3333101368894024e-05,
      "loss": 2.3447132110595703,
      "memory(GiB)": 77.56,
      "step": 79250,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.436959
    },
    {
      "epoch": 3.3955271839252816,
      "grad_norm": 4.763704299926758,
      "learning_rate": 2.3327408880335245e-05,
      "loss": 2.172576141357422,
      "memory(GiB)": 77.56,
      "step": 79255,
      "token_acc": 0.5573248407643312,
      "train_speed(iter/s)": 1.436964
    },
    {
      "epoch": 3.395741399254531,
      "grad_norm": 5.209707736968994,
      "learning_rate": 2.3321716874975498e-05,
      "loss": 2.482649230957031,
      "memory(GiB)": 77.56,
      "step": 79260,
      "token_acc": 0.47076023391812866,
      "train_speed(iter/s)": 1.436948
    },
    {
      "epoch": 3.3959556145837797,
      "grad_norm": 6.771468162536621,
      "learning_rate": 2.331602535291787e-05,
      "loss": 2.4084396362304688,
      "memory(GiB)": 77.56,
      "step": 79265,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.436964
    },
    {
      "epoch": 3.3961698299130285,
      "grad_norm": 6.260661602020264,
      "learning_rate": 2.331033431426546e-05,
      "loss": 2.2911216735839846,
      "memory(GiB)": 77.56,
      "step": 79270,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.436974
    },
    {
      "epoch": 3.3963840452422778,
      "grad_norm": 4.987298488616943,
      "learning_rate": 2.3304643759121398e-05,
      "loss": 2.2478328704833985,
      "memory(GiB)": 77.56,
      "step": 79275,
      "token_acc": 0.5111731843575419,
      "train_speed(iter/s)": 1.436966
    },
    {
      "epoch": 3.3965982605715266,
      "grad_norm": 5.767981052398682,
      "learning_rate": 2.3298953687588753e-05,
      "loss": 2.572159767150879,
      "memory(GiB)": 77.56,
      "step": 79280,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.436941
    },
    {
      "epoch": 3.3968124759007754,
      "grad_norm": 5.535113334655762,
      "learning_rate": 2.3293264099770613e-05,
      "loss": 2.552655029296875,
      "memory(GiB)": 77.56,
      "step": 79285,
      "token_acc": 0.44907407407407407,
      "train_speed(iter/s)": 1.436965
    },
    {
      "epoch": 3.3970266912300247,
      "grad_norm": 4.84979248046875,
      "learning_rate": 2.3287574995770028e-05,
      "loss": 2.142882537841797,
      "memory(GiB)": 77.56,
      "step": 79290,
      "token_acc": 0.5377906976744186,
      "train_speed(iter/s)": 1.43697
    },
    {
      "epoch": 3.3972409065592735,
      "grad_norm": 5.1381049156188965,
      "learning_rate": 2.32818863756901e-05,
      "loss": 2.6485221862792967,
      "memory(GiB)": 77.56,
      "step": 79295,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.436986
    },
    {
      "epoch": 3.3974551218885223,
      "grad_norm": 6.498202323913574,
      "learning_rate": 2.327619823963386e-05,
      "loss": 2.624630355834961,
      "memory(GiB)": 77.56,
      "step": 79300,
      "token_acc": 0.47651006711409394,
      "train_speed(iter/s)": 1.436988
    },
    {
      "epoch": 3.3976693372177715,
      "grad_norm": 4.937193870544434,
      "learning_rate": 2.3270510587704365e-05,
      "loss": 2.212263488769531,
      "memory(GiB)": 77.56,
      "step": 79305,
      "token_acc": 0.5311475409836065,
      "train_speed(iter/s)": 1.43701
    },
    {
      "epoch": 3.3978835525470203,
      "grad_norm": 5.264317989349365,
      "learning_rate": 2.326482342000464e-05,
      "loss": 2.2468730926513674,
      "memory(GiB)": 77.56,
      "step": 79310,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.43701
    },
    {
      "epoch": 3.398097767876269,
      "grad_norm": 5.205953598022461,
      "learning_rate": 2.3259136736637697e-05,
      "loss": 2.163889503479004,
      "memory(GiB)": 77.56,
      "step": 79315,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.437023
    },
    {
      "epoch": 3.3983119832055184,
      "grad_norm": 5.57481050491333,
      "learning_rate": 2.32534505377066e-05,
      "loss": 2.367462730407715,
      "memory(GiB)": 77.56,
      "step": 79320,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.437049
    },
    {
      "epoch": 3.3985261985347672,
      "grad_norm": 6.042647361755371,
      "learning_rate": 2.324776482331434e-05,
      "loss": 2.2394073486328123,
      "memory(GiB)": 77.56,
      "step": 79325,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.437065
    },
    {
      "epoch": 3.398740413864016,
      "grad_norm": 6.816196441650391,
      "learning_rate": 2.324207959356391e-05,
      "loss": 2.3014684677124024,
      "memory(GiB)": 77.56,
      "step": 79330,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.437069
    },
    {
      "epoch": 3.3989546291932653,
      "grad_norm": 7.3775434494018555,
      "learning_rate": 2.323639484855831e-05,
      "loss": 2.4087228775024414,
      "memory(GiB)": 77.56,
      "step": 79335,
      "token_acc": 0.5139442231075697,
      "train_speed(iter/s)": 1.437078
    },
    {
      "epoch": 3.399168844522514,
      "grad_norm": 5.364716529846191,
      "learning_rate": 2.3230710588400505e-05,
      "loss": 2.52114315032959,
      "memory(GiB)": 77.56,
      "step": 79340,
      "token_acc": 0.4776536312849162,
      "train_speed(iter/s)": 1.437084
    },
    {
      "epoch": 3.399383059851763,
      "grad_norm": 5.3653564453125,
      "learning_rate": 2.322502681319349e-05,
      "loss": 2.2224849700927733,
      "memory(GiB)": 77.56,
      "step": 79345,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.437096
    },
    {
      "epoch": 3.399597275181012,
      "grad_norm": 5.870612621307373,
      "learning_rate": 2.321934352304025e-05,
      "loss": 2.585462951660156,
      "memory(GiB)": 77.56,
      "step": 79350,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.437091
    },
    {
      "epoch": 3.399811490510261,
      "grad_norm": 8.266950607299805,
      "learning_rate": 2.321366071804373e-05,
      "loss": 2.141002655029297,
      "memory(GiB)": 77.56,
      "step": 79355,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.437112
    },
    {
      "epoch": 3.40002570583951,
      "grad_norm": 5.365269660949707,
      "learning_rate": 2.320797839830686e-05,
      "loss": 2.263966751098633,
      "memory(GiB)": 77.56,
      "step": 79360,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.437118
    },
    {
      "epoch": 3.400239921168759,
      "grad_norm": 6.08232307434082,
      "learning_rate": 2.3202296563932607e-05,
      "loss": 2.504233169555664,
      "memory(GiB)": 77.56,
      "step": 79365,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.437114
    },
    {
      "epoch": 3.400454136498008,
      "grad_norm": 5.069942474365234,
      "learning_rate": 2.3196615215023886e-05,
      "loss": 2.228591728210449,
      "memory(GiB)": 77.56,
      "step": 79370,
      "token_acc": 0.5189873417721519,
      "train_speed(iter/s)": 1.437104
    },
    {
      "epoch": 3.4006683518272567,
      "grad_norm": 6.968050479888916,
      "learning_rate": 2.3190934351683602e-05,
      "loss": 2.438836860656738,
      "memory(GiB)": 77.56,
      "step": 79375,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.437123
    },
    {
      "epoch": 3.400882567156506,
      "grad_norm": 4.4127044677734375,
      "learning_rate": 2.3185253974014714e-05,
      "loss": 2.0383148193359375,
      "memory(GiB)": 77.56,
      "step": 79380,
      "token_acc": 0.5190311418685121,
      "train_speed(iter/s)": 1.437108
    },
    {
      "epoch": 3.4010967824857548,
      "grad_norm": 5.228052616119385,
      "learning_rate": 2.3179574082120105e-05,
      "loss": 2.2109039306640623,
      "memory(GiB)": 77.56,
      "step": 79385,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.437118
    },
    {
      "epoch": 3.4013109978150036,
      "grad_norm": 6.189214706420898,
      "learning_rate": 2.317389467610267e-05,
      "loss": 2.3826671600341798,
      "memory(GiB)": 77.56,
      "step": 79390,
      "token_acc": 0.5381818181818182,
      "train_speed(iter/s)": 1.437106
    },
    {
      "epoch": 3.401525213144253,
      "grad_norm": 6.784708499908447,
      "learning_rate": 2.3168215756065292e-05,
      "loss": 2.701226806640625,
      "memory(GiB)": 77.56,
      "step": 79395,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.437102
    },
    {
      "epoch": 3.4017394284735016,
      "grad_norm": 6.745068550109863,
      "learning_rate": 2.3162537322110843e-05,
      "loss": 2.35314884185791,
      "memory(GiB)": 77.56,
      "step": 79400,
      "token_acc": 0.47249190938511326,
      "train_speed(iter/s)": 1.437119
    },
    {
      "epoch": 3.4019536438027504,
      "grad_norm": 7.3597731590271,
      "learning_rate": 2.3156859374342226e-05,
      "loss": 2.168228530883789,
      "memory(GiB)": 77.56,
      "step": 79405,
      "token_acc": 0.5244755244755245,
      "train_speed(iter/s)": 1.437104
    },
    {
      "epoch": 3.4021678591319997,
      "grad_norm": 8.19350528717041,
      "learning_rate": 2.315118191286228e-05,
      "loss": 2.142755126953125,
      "memory(GiB)": 77.56,
      "step": 79410,
      "token_acc": 0.5347985347985348,
      "train_speed(iter/s)": 1.437112
    },
    {
      "epoch": 3.4023820744612485,
      "grad_norm": 5.765941143035889,
      "learning_rate": 2.314550493777386e-05,
      "loss": 2.490132713317871,
      "memory(GiB)": 77.56,
      "step": 79415,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.437096
    },
    {
      "epoch": 3.4025962897904973,
      "grad_norm": 4.746912479400635,
      "learning_rate": 2.313982844917979e-05,
      "loss": 2.188548278808594,
      "memory(GiB)": 77.56,
      "step": 79420,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.437104
    },
    {
      "epoch": 3.4028105051197466,
      "grad_norm": 4.999203205108643,
      "learning_rate": 2.3134152447182945e-05,
      "loss": 2.3510326385498046,
      "memory(GiB)": 77.56,
      "step": 79425,
      "token_acc": 0.5266666666666666,
      "train_speed(iter/s)": 1.437115
    },
    {
      "epoch": 3.4030247204489954,
      "grad_norm": 7.057901382446289,
      "learning_rate": 2.3128476931886128e-05,
      "loss": 2.490496826171875,
      "memory(GiB)": 77.56,
      "step": 79430,
      "token_acc": 0.518796992481203,
      "train_speed(iter/s)": 1.43713
    },
    {
      "epoch": 3.403238935778244,
      "grad_norm": 5.6993584632873535,
      "learning_rate": 2.3122801903392145e-05,
      "loss": 2.242514419555664,
      "memory(GiB)": 77.56,
      "step": 79435,
      "token_acc": 0.5207547169811321,
      "train_speed(iter/s)": 1.437107
    },
    {
      "epoch": 3.4034531511074935,
      "grad_norm": 6.187243938446045,
      "learning_rate": 2.311712736180383e-05,
      "loss": 2.408182907104492,
      "memory(GiB)": 77.56,
      "step": 79440,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.437102
    },
    {
      "epoch": 3.4036673664367423,
      "grad_norm": 4.937978744506836,
      "learning_rate": 2.3111453307223978e-05,
      "loss": 2.5719539642333986,
      "memory(GiB)": 77.56,
      "step": 79445,
      "token_acc": 0.46258503401360546,
      "train_speed(iter/s)": 1.437118
    },
    {
      "epoch": 3.403881581765991,
      "grad_norm": 4.627047538757324,
      "learning_rate": 2.3105779739755368e-05,
      "loss": 2.1482330322265626,
      "memory(GiB)": 77.56,
      "step": 79450,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.437121
    },
    {
      "epoch": 3.4040957970952403,
      "grad_norm": 5.4321794509887695,
      "learning_rate": 2.3100106659500794e-05,
      "loss": 2.273075485229492,
      "memory(GiB)": 77.56,
      "step": 79455,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.437112
    },
    {
      "epoch": 3.404310012424489,
      "grad_norm": 6.22972297668457,
      "learning_rate": 2.3094434066562993e-05,
      "loss": 2.2309507369995116,
      "memory(GiB)": 77.56,
      "step": 79460,
      "token_acc": 0.516728624535316,
      "train_speed(iter/s)": 1.437139
    },
    {
      "epoch": 3.404524227753738,
      "grad_norm": 7.1927103996276855,
      "learning_rate": 2.3088761961044786e-05,
      "loss": 2.387375259399414,
      "memory(GiB)": 77.56,
      "step": 79465,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.437133
    },
    {
      "epoch": 3.4047384430829872,
      "grad_norm": 8.879721641540527,
      "learning_rate": 2.3083090343048898e-05,
      "loss": 2.5375238418579102,
      "memory(GiB)": 77.56,
      "step": 79470,
      "token_acc": 0.4416058394160584,
      "train_speed(iter/s)": 1.437136
    },
    {
      "epoch": 3.404952658412236,
      "grad_norm": 6.653095245361328,
      "learning_rate": 2.3077419212678076e-05,
      "loss": 2.424617958068848,
      "memory(GiB)": 77.56,
      "step": 79475,
      "token_acc": 0.5133531157270029,
      "train_speed(iter/s)": 1.437155
    },
    {
      "epoch": 3.405166873741485,
      "grad_norm": 7.53515100479126,
      "learning_rate": 2.3071748570035063e-05,
      "loss": 2.191908264160156,
      "memory(GiB)": 77.56,
      "step": 79480,
      "token_acc": 0.5507246376811594,
      "train_speed(iter/s)": 1.437145
    },
    {
      "epoch": 3.405381089070734,
      "grad_norm": 6.5444159507751465,
      "learning_rate": 2.3066078415222563e-05,
      "loss": 2.1773199081420898,
      "memory(GiB)": 77.56,
      "step": 79485,
      "token_acc": 0.5096153846153846,
      "train_speed(iter/s)": 1.437137
    },
    {
      "epoch": 3.405595304399983,
      "grad_norm": 5.0450825691223145,
      "learning_rate": 2.306040874834334e-05,
      "loss": 2.3375301361083984,
      "memory(GiB)": 77.56,
      "step": 79490,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.43714
    },
    {
      "epoch": 3.4058095197292317,
      "grad_norm": 5.600021839141846,
      "learning_rate": 2.3054739569500066e-05,
      "loss": 2.074955368041992,
      "memory(GiB)": 77.56,
      "step": 79495,
      "token_acc": 0.5306859205776173,
      "train_speed(iter/s)": 1.437149
    },
    {
      "epoch": 3.406023735058481,
      "grad_norm": 8.0911283493042,
      "learning_rate": 2.3049070878795477e-05,
      "loss": 2.452461051940918,
      "memory(GiB)": 77.56,
      "step": 79500,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.437167
    },
    {
      "epoch": 3.406023735058481,
      "eval_loss": 2.179293155670166,
      "eval_runtime": 14.297,
      "eval_samples_per_second": 6.994,
      "eval_steps_per_second": 6.994,
      "eval_token_acc": 0.48917748917748916,
      "step": 79500
    },
    {
      "epoch": 3.40623795038773,
      "grad_norm": 6.782418727874756,
      "learning_rate": 2.3043402676332253e-05,
      "loss": 2.328179359436035,
      "memory(GiB)": 77.56,
      "step": 79505,
      "token_acc": 0.4969262295081967,
      "train_speed(iter/s)": 1.436782
    },
    {
      "epoch": 3.4064521657169786,
      "grad_norm": 5.8732075691223145,
      "learning_rate": 2.303773496221308e-05,
      "loss": 2.4902297973632814,
      "memory(GiB)": 77.56,
      "step": 79510,
      "token_acc": 0.45962732919254656,
      "train_speed(iter/s)": 1.436782
    },
    {
      "epoch": 3.406666381046228,
      "grad_norm": 5.070359706878662,
      "learning_rate": 2.3032067736540626e-05,
      "loss": 2.1516193389892577,
      "memory(GiB)": 77.56,
      "step": 79515,
      "token_acc": 0.5083056478405316,
      "train_speed(iter/s)": 1.436751
    },
    {
      "epoch": 3.4068805963754767,
      "grad_norm": 5.498624324798584,
      "learning_rate": 2.3026400999417546e-05,
      "loss": 2.2944265365600587,
      "memory(GiB)": 77.56,
      "step": 79520,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.436765
    },
    {
      "epoch": 3.4070948117047255,
      "grad_norm": 5.4773640632629395,
      "learning_rate": 2.3020734750946534e-05,
      "loss": 2.510694122314453,
      "memory(GiB)": 77.56,
      "step": 79525,
      "token_acc": 0.5030674846625767,
      "train_speed(iter/s)": 1.436785
    },
    {
      "epoch": 3.4073090270339748,
      "grad_norm": 6.929642677307129,
      "learning_rate": 2.3015068991230222e-05,
      "loss": 2.2454782485961915,
      "memory(GiB)": 77.56,
      "step": 79530,
      "token_acc": 0.5255972696245734,
      "train_speed(iter/s)": 1.436807
    },
    {
      "epoch": 3.4075232423632236,
      "grad_norm": 5.425712585449219,
      "learning_rate": 2.3009403720371247e-05,
      "loss": 2.5280765533447265,
      "memory(GiB)": 77.56,
      "step": 79535,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.436816
    },
    {
      "epoch": 3.4077374576924724,
      "grad_norm": 5.826126575469971,
      "learning_rate": 2.300373893847224e-05,
      "loss": 2.5497669219970702,
      "memory(GiB)": 77.56,
      "step": 79540,
      "token_acc": 0.4555160142348754,
      "train_speed(iter/s)": 1.436823
    },
    {
      "epoch": 3.4079516730217216,
      "grad_norm": 5.537844657897949,
      "learning_rate": 2.2998074645635815e-05,
      "loss": 2.3873971939086913,
      "memory(GiB)": 77.56,
      "step": 79545,
      "token_acc": 0.49702380952380953,
      "train_speed(iter/s)": 1.436826
    },
    {
      "epoch": 3.4081658883509705,
      "grad_norm": 7.128018379211426,
      "learning_rate": 2.299241084196461e-05,
      "loss": 2.3843105316162108,
      "memory(GiB)": 77.56,
      "step": 79550,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.436825
    },
    {
      "epoch": 3.4083801036802193,
      "grad_norm": 5.012312412261963,
      "learning_rate": 2.2986747527561227e-05,
      "loss": 2.2672367095947266,
      "memory(GiB)": 77.56,
      "step": 79555,
      "token_acc": 0.5396825396825397,
      "train_speed(iter/s)": 1.43683
    },
    {
      "epoch": 3.4085943190094685,
      "grad_norm": 7.558150768280029,
      "learning_rate": 2.2981084702528244e-05,
      "loss": 2.379343032836914,
      "memory(GiB)": 77.56,
      "step": 79560,
      "token_acc": 0.5182724252491694,
      "train_speed(iter/s)": 1.436853
    },
    {
      "epoch": 3.4088085343387173,
      "grad_norm": 5.734176158905029,
      "learning_rate": 2.297542236696824e-05,
      "loss": 2.2569149017333983,
      "memory(GiB)": 77.56,
      "step": 79565,
      "token_acc": 0.5698529411764706,
      "train_speed(iter/s)": 1.436871
    },
    {
      "epoch": 3.409022749667966,
      "grad_norm": 5.442965030670166,
      "learning_rate": 2.296976052098383e-05,
      "loss": 2.3304931640625,
      "memory(GiB)": 77.56,
      "step": 79570,
      "token_acc": 0.5276872964169381,
      "train_speed(iter/s)": 1.436894
    },
    {
      "epoch": 3.4092369649972154,
      "grad_norm": 5.321924686431885,
      "learning_rate": 2.2964099164677567e-05,
      "loss": 2.5034063339233397,
      "memory(GiB)": 77.56,
      "step": 79575,
      "token_acc": 0.49221183800623053,
      "train_speed(iter/s)": 1.436901
    },
    {
      "epoch": 3.409451180326464,
      "grad_norm": 8.341198921203613,
      "learning_rate": 2.295843829815199e-05,
      "loss": 2.216433525085449,
      "memory(GiB)": 77.56,
      "step": 79580,
      "token_acc": 0.463519313304721,
      "train_speed(iter/s)": 1.436907
    },
    {
      "epoch": 3.409665395655713,
      "grad_norm": 5.1558942794799805,
      "learning_rate": 2.295277792150969e-05,
      "loss": 2.4018110275268554,
      "memory(GiB)": 77.56,
      "step": 79585,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.436889
    },
    {
      "epoch": 3.4098796109849623,
      "grad_norm": 6.371580600738525,
      "learning_rate": 2.294711803485319e-05,
      "loss": 2.2421146392822267,
      "memory(GiB)": 77.56,
      "step": 79590,
      "token_acc": 0.5160256410256411,
      "train_speed(iter/s)": 1.436895
    },
    {
      "epoch": 3.410093826314211,
      "grad_norm": 5.7849602699279785,
      "learning_rate": 2.2941458638285017e-05,
      "loss": 2.2049732208251953,
      "memory(GiB)": 77.56,
      "step": 79595,
      "token_acc": 0.5140562248995983,
      "train_speed(iter/s)": 1.436914
    },
    {
      "epoch": 3.41030804164346,
      "grad_norm": 4.941309452056885,
      "learning_rate": 2.2935799731907707e-05,
      "loss": 2.064447593688965,
      "memory(GiB)": 77.56,
      "step": 79600,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.436922
    },
    {
      "epoch": 3.410522256972709,
      "grad_norm": 6.720141410827637,
      "learning_rate": 2.293014131582376e-05,
      "loss": 2.2994516372680662,
      "memory(GiB)": 77.56,
      "step": 79605,
      "token_acc": 0.5036231884057971,
      "train_speed(iter/s)": 1.436932
    },
    {
      "epoch": 3.410736472301958,
      "grad_norm": 5.530195236206055,
      "learning_rate": 2.2924483390135716e-05,
      "loss": 2.2740102767944337,
      "memory(GiB)": 77.56,
      "step": 79610,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.436961
    },
    {
      "epoch": 3.410950687631207,
      "grad_norm": 5.3113203048706055,
      "learning_rate": 2.291882595494605e-05,
      "loss": 2.197417449951172,
      "memory(GiB)": 77.56,
      "step": 79615,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.436943
    },
    {
      "epoch": 3.411164902960456,
      "grad_norm": 7.109493255615234,
      "learning_rate": 2.2913169010357256e-05,
      "loss": 2.252703857421875,
      "memory(GiB)": 77.56,
      "step": 79620,
      "token_acc": 0.5338345864661654,
      "train_speed(iter/s)": 1.436948
    },
    {
      "epoch": 3.411379118289705,
      "grad_norm": 8.050132751464844,
      "learning_rate": 2.2907512556471817e-05,
      "loss": 2.3119527816772463,
      "memory(GiB)": 77.56,
      "step": 79625,
      "token_acc": 0.4745762711864407,
      "train_speed(iter/s)": 1.436946
    },
    {
      "epoch": 3.4115933336189537,
      "grad_norm": 8.109003067016602,
      "learning_rate": 2.290185659339218e-05,
      "loss": 2.50976505279541,
      "memory(GiB)": 77.56,
      "step": 79630,
      "token_acc": 0.44805194805194803,
      "train_speed(iter/s)": 1.436938
    },
    {
      "epoch": 3.411807548948203,
      "grad_norm": 6.99859094619751,
      "learning_rate": 2.2896201121220856e-05,
      "loss": 2.250266456604004,
      "memory(GiB)": 77.56,
      "step": 79635,
      "token_acc": 0.5234657039711191,
      "train_speed(iter/s)": 1.436954
    },
    {
      "epoch": 3.4120217642774517,
      "grad_norm": 4.7064313888549805,
      "learning_rate": 2.289054614006025e-05,
      "loss": 2.3661777496337892,
      "memory(GiB)": 77.56,
      "step": 79640,
      "token_acc": 0.493006993006993,
      "train_speed(iter/s)": 1.436971
    },
    {
      "epoch": 3.4122359796067006,
      "grad_norm": 5.220667839050293,
      "learning_rate": 2.288489165001285e-05,
      "loss": 2.3214014053344725,
      "memory(GiB)": 77.56,
      "step": 79645,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.43699
    },
    {
      "epoch": 3.41245019493595,
      "grad_norm": 5.36407995223999,
      "learning_rate": 2.287923765118108e-05,
      "loss": 2.4953689575195312,
      "memory(GiB)": 77.56,
      "step": 79650,
      "token_acc": 0.5170068027210885,
      "train_speed(iter/s)": 1.437005
    },
    {
      "epoch": 3.4126644102651986,
      "grad_norm": 6.248805522918701,
      "learning_rate": 2.2873584143667352e-05,
      "loss": 2.323070526123047,
      "memory(GiB)": 77.56,
      "step": 79655,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.437015
    },
    {
      "epoch": 3.4128786255944474,
      "grad_norm": 5.491689682006836,
      "learning_rate": 2.28679311275741e-05,
      "loss": 2.6055953979492186,
      "memory(GiB)": 77.56,
      "step": 79660,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.436999
    },
    {
      "epoch": 3.4130928409236967,
      "grad_norm": 5.359642505645752,
      "learning_rate": 2.2862278603003707e-05,
      "loss": 2.322352409362793,
      "memory(GiB)": 77.56,
      "step": 79665,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.436995
    },
    {
      "epoch": 3.4133070562529455,
      "grad_norm": 6.347226619720459,
      "learning_rate": 2.2856626570058613e-05,
      "loss": 2.268888473510742,
      "memory(GiB)": 77.56,
      "step": 79670,
      "token_acc": 0.5245398773006135,
      "train_speed(iter/s)": 1.436996
    },
    {
      "epoch": 3.4135212715821943,
      "grad_norm": 5.04331111907959,
      "learning_rate": 2.2850975028841194e-05,
      "loss": 2.4179004669189452,
      "memory(GiB)": 77.56,
      "step": 79675,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.437002
    },
    {
      "epoch": 3.4137354869114436,
      "grad_norm": 5.191434860229492,
      "learning_rate": 2.2845323979453832e-05,
      "loss": 1.9253528594970704,
      "memory(GiB)": 77.56,
      "step": 79680,
      "token_acc": 0.5468164794007491,
      "train_speed(iter/s)": 1.436997
    },
    {
      "epoch": 3.4139497022406924,
      "grad_norm": 4.421807289123535,
      "learning_rate": 2.2839673421998892e-05,
      "loss": 2.186411476135254,
      "memory(GiB)": 77.56,
      "step": 79685,
      "token_acc": 0.5451127819548872,
      "train_speed(iter/s)": 1.436997
    },
    {
      "epoch": 3.414163917569941,
      "grad_norm": 5.415095329284668,
      "learning_rate": 2.283402335657873e-05,
      "loss": 2.6603214263916017,
      "memory(GiB)": 77.56,
      "step": 79690,
      "token_acc": 0.44785276073619634,
      "train_speed(iter/s)": 1.436996
    },
    {
      "epoch": 3.4143781328991905,
      "grad_norm": 4.832653999328613,
      "learning_rate": 2.2828373783295743e-05,
      "loss": 2.298392486572266,
      "memory(GiB)": 77.56,
      "step": 79695,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.436996
    },
    {
      "epoch": 3.4145923482284393,
      "grad_norm": 5.813657283782959,
      "learning_rate": 2.2822724702252245e-05,
      "loss": 2.100205993652344,
      "memory(GiB)": 77.56,
      "step": 79700,
      "token_acc": 0.5138461538461538,
      "train_speed(iter/s)": 1.436989
    },
    {
      "epoch": 3.414806563557688,
      "grad_norm": 6.642099380493164,
      "learning_rate": 2.281707611355059e-05,
      "loss": 2.4816299438476563,
      "memory(GiB)": 77.56,
      "step": 79705,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.437006
    },
    {
      "epoch": 3.4150207788869373,
      "grad_norm": 6.974586486816406,
      "learning_rate": 2.2811428017293096e-05,
      "loss": 2.395670509338379,
      "memory(GiB)": 77.56,
      "step": 79710,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.437018
    },
    {
      "epoch": 3.415234994216186,
      "grad_norm": 8.6386137008667,
      "learning_rate": 2.2805780413582075e-05,
      "loss": 2.3550479888916014,
      "memory(GiB)": 77.56,
      "step": 79715,
      "token_acc": 0.4744525547445255,
      "train_speed(iter/s)": 1.437016
    },
    {
      "epoch": 3.415449209545435,
      "grad_norm": 7.0342841148376465,
      "learning_rate": 2.2800133302519866e-05,
      "loss": 2.405625915527344,
      "memory(GiB)": 77.56,
      "step": 79720,
      "token_acc": 0.48466257668711654,
      "train_speed(iter/s)": 1.437036
    },
    {
      "epoch": 3.415663424874684,
      "grad_norm": 5.447117328643799,
      "learning_rate": 2.2794486684208744e-05,
      "loss": 2.42040901184082,
      "memory(GiB)": 77.56,
      "step": 79725,
      "token_acc": 0.4584717607973422,
      "train_speed(iter/s)": 1.43703
    },
    {
      "epoch": 3.415877640203933,
      "grad_norm": 6.344385147094727,
      "learning_rate": 2.278884055875103e-05,
      "loss": 2.3380762100219727,
      "memory(GiB)": 77.56,
      "step": 79730,
      "token_acc": 0.47717842323651455,
      "train_speed(iter/s)": 1.437041
    },
    {
      "epoch": 3.416091855533182,
      "grad_norm": 6.731952667236328,
      "learning_rate": 2.2783194926248996e-05,
      "loss": 2.247830390930176,
      "memory(GiB)": 77.56,
      "step": 79735,
      "token_acc": 0.5350553505535055,
      "train_speed(iter/s)": 1.437059
    },
    {
      "epoch": 3.416306070862431,
      "grad_norm": 5.4700493812561035,
      "learning_rate": 2.2777549786804918e-05,
      "loss": 2.7882902145385744,
      "memory(GiB)": 77.56,
      "step": 79740,
      "token_acc": 0.43450479233226835,
      "train_speed(iter/s)": 1.437076
    },
    {
      "epoch": 3.41652028619168,
      "grad_norm": 6.6195197105407715,
      "learning_rate": 2.2771905140521067e-05,
      "loss": 2.3725833892822266,
      "memory(GiB)": 77.56,
      "step": 79745,
      "token_acc": 0.49146757679180886,
      "train_speed(iter/s)": 1.437039
    },
    {
      "epoch": 3.4167345015209287,
      "grad_norm": 5.307271957397461,
      "learning_rate": 2.2766260987499677e-05,
      "loss": 2.4947845458984377,
      "memory(GiB)": 77.56,
      "step": 79750,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.437064
    },
    {
      "epoch": 3.416948716850178,
      "grad_norm": 5.474647521972656,
      "learning_rate": 2.276061732784303e-05,
      "loss": 2.5116575241088865,
      "memory(GiB)": 77.56,
      "step": 79755,
      "token_acc": 0.4968152866242038,
      "train_speed(iter/s)": 1.437062
    },
    {
      "epoch": 3.417162932179427,
      "grad_norm": 5.101426124572754,
      "learning_rate": 2.275497416165335e-05,
      "loss": 1.9923717498779296,
      "memory(GiB)": 77.56,
      "step": 79760,
      "token_acc": 0.5537459283387622,
      "train_speed(iter/s)": 1.437074
    },
    {
      "epoch": 3.4173771475086756,
      "grad_norm": 5.685215950012207,
      "learning_rate": 2.2749331489032884e-05,
      "loss": 2.2149404525756835,
      "memory(GiB)": 77.56,
      "step": 79765,
      "token_acc": 0.5373665480427047,
      "train_speed(iter/s)": 1.437083
    },
    {
      "epoch": 3.417591362837925,
      "grad_norm": 6.465989589691162,
      "learning_rate": 2.274368931008383e-05,
      "loss": 2.424448776245117,
      "memory(GiB)": 77.56,
      "step": 79770,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437097
    },
    {
      "epoch": 3.4178055781671737,
      "grad_norm": 4.465653896331787,
      "learning_rate": 2.27380476249084e-05,
      "loss": 2.2471393585205077,
      "memory(GiB)": 77.56,
      "step": 79775,
      "token_acc": 0.495114006514658,
      "train_speed(iter/s)": 1.437094
    },
    {
      "epoch": 3.4180197934964225,
      "grad_norm": 5.666294574737549,
      "learning_rate": 2.2732406433608826e-05,
      "loss": 2.379106330871582,
      "memory(GiB)": 77.56,
      "step": 79780,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.437108
    },
    {
      "epoch": 3.4182340088256717,
      "grad_norm": 6.225849151611328,
      "learning_rate": 2.2726765736287287e-05,
      "loss": 2.2286605834960938,
      "memory(GiB)": 77.56,
      "step": 79785,
      "token_acc": 0.5211267605633803,
      "train_speed(iter/s)": 1.437111
    },
    {
      "epoch": 3.4184482241549206,
      "grad_norm": 4.352685928344727,
      "learning_rate": 2.2721125533045955e-05,
      "loss": 2.1221166610717774,
      "memory(GiB)": 77.56,
      "step": 79790,
      "token_acc": 0.5433962264150943,
      "train_speed(iter/s)": 1.437113
    },
    {
      "epoch": 3.4186624394841694,
      "grad_norm": 7.090567588806152,
      "learning_rate": 2.2715485823987043e-05,
      "loss": 2.541061210632324,
      "memory(GiB)": 77.56,
      "step": 79795,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.437096
    },
    {
      "epoch": 3.4188766548134186,
      "grad_norm": 5.3690338134765625,
      "learning_rate": 2.27098466092127e-05,
      "loss": 2.6047725677490234,
      "memory(GiB)": 77.56,
      "step": 79800,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.437104
    },
    {
      "epoch": 3.4190908701426674,
      "grad_norm": 6.505526542663574,
      "learning_rate": 2.2704207888825086e-05,
      "loss": 2.4545711517333983,
      "memory(GiB)": 77.56,
      "step": 79805,
      "token_acc": 0.48632218844984804,
      "train_speed(iter/s)": 1.437108
    },
    {
      "epoch": 3.4193050854719163,
      "grad_norm": 7.470578193664551,
      "learning_rate": 2.2698569662926333e-05,
      "loss": 2.25128231048584,
      "memory(GiB)": 77.56,
      "step": 79810,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437119
    },
    {
      "epoch": 3.4195193008011655,
      "grad_norm": 6.0045976638793945,
      "learning_rate": 2.2692931931618622e-05,
      "loss": 2.173305320739746,
      "memory(GiB)": 77.56,
      "step": 79815,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.437133
    },
    {
      "epoch": 3.4197335161304143,
      "grad_norm": 5.753742694854736,
      "learning_rate": 2.2687294695004058e-05,
      "loss": 2.374880790710449,
      "memory(GiB)": 77.56,
      "step": 79820,
      "token_acc": 0.5052264808362369,
      "train_speed(iter/s)": 1.437153
    },
    {
      "epoch": 3.419947731459663,
      "grad_norm": 4.839774131774902,
      "learning_rate": 2.2681657953184775e-05,
      "loss": 2.262137031555176,
      "memory(GiB)": 77.56,
      "step": 79825,
      "token_acc": 0.555921052631579,
      "train_speed(iter/s)": 1.437157
    },
    {
      "epoch": 3.4201619467889124,
      "grad_norm": 5.287930011749268,
      "learning_rate": 2.267602170626289e-05,
      "loss": 2.496467399597168,
      "memory(GiB)": 77.56,
      "step": 79830,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.437156
    },
    {
      "epoch": 3.420376162118161,
      "grad_norm": 5.560831546783447,
      "learning_rate": 2.267038595434048e-05,
      "loss": 2.4401559829711914,
      "memory(GiB)": 77.56,
      "step": 79835,
      "token_acc": 0.5054151624548736,
      "train_speed(iter/s)": 1.437178
    },
    {
      "epoch": 3.42059037744741,
      "grad_norm": 6.871474742889404,
      "learning_rate": 2.266475069751969e-05,
      "loss": 2.202611541748047,
      "memory(GiB)": 77.56,
      "step": 79840,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.437173
    },
    {
      "epoch": 3.4208045927766593,
      "grad_norm": 5.315938472747803,
      "learning_rate": 2.2659115935902576e-05,
      "loss": 2.588887023925781,
      "memory(GiB)": 77.56,
      "step": 79845,
      "token_acc": 0.46953405017921146,
      "train_speed(iter/s)": 1.437177
    },
    {
      "epoch": 3.421018808105908,
      "grad_norm": 4.877957820892334,
      "learning_rate": 2.265348166959123e-05,
      "loss": 2.452315902709961,
      "memory(GiB)": 77.56,
      "step": 79850,
      "token_acc": 0.4707792207792208,
      "train_speed(iter/s)": 1.43718
    },
    {
      "epoch": 3.421233023435157,
      "grad_norm": 11.038566589355469,
      "learning_rate": 2.2647847898687712e-05,
      "loss": 2.247224044799805,
      "memory(GiB)": 77.56,
      "step": 79855,
      "token_acc": 0.5394190871369294,
      "train_speed(iter/s)": 1.437195
    },
    {
      "epoch": 3.421447238764406,
      "grad_norm": 7.7552876472473145,
      "learning_rate": 2.2642214623294073e-05,
      "loss": 2.363880157470703,
      "memory(GiB)": 77.56,
      "step": 79860,
      "token_acc": 0.5379061371841155,
      "train_speed(iter/s)": 1.437168
    },
    {
      "epoch": 3.421661454093655,
      "grad_norm": 5.875951766967773,
      "learning_rate": 2.2636581843512378e-05,
      "loss": 2.306252670288086,
      "memory(GiB)": 77.56,
      "step": 79865,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.43715
    },
    {
      "epoch": 3.4218756694229038,
      "grad_norm": 6.256127834320068,
      "learning_rate": 2.2630949559444693e-05,
      "loss": 2.12294864654541,
      "memory(GiB)": 77.56,
      "step": 79870,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.437175
    },
    {
      "epoch": 3.422089884752153,
      "grad_norm": 7.295079231262207,
      "learning_rate": 2.2625317771193027e-05,
      "loss": 2.479261779785156,
      "memory(GiB)": 77.56,
      "step": 79875,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.437183
    },
    {
      "epoch": 3.422304100081402,
      "grad_norm": 5.242221832275391,
      "learning_rate": 2.2619686478859416e-05,
      "loss": 2.3235212326049806,
      "memory(GiB)": 77.56,
      "step": 79880,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.437197
    },
    {
      "epoch": 3.4225183154106507,
      "grad_norm": 4.939266204833984,
      "learning_rate": 2.261405568254586e-05,
      "loss": 2.5372453689575196,
      "memory(GiB)": 77.56,
      "step": 79885,
      "token_acc": 0.48757763975155277,
      "train_speed(iter/s)": 1.437208
    },
    {
      "epoch": 3.4227325307399,
      "grad_norm": 5.698630332946777,
      "learning_rate": 2.2608425382354383e-05,
      "loss": 2.1325597763061523,
      "memory(GiB)": 77.56,
      "step": 79890,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.437199
    },
    {
      "epoch": 3.4229467460691487,
      "grad_norm": 4.5413360595703125,
      "learning_rate": 2.2602795578386954e-05,
      "loss": 2.3484495162963865,
      "memory(GiB)": 77.56,
      "step": 79895,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.437209
    },
    {
      "epoch": 3.4231609613983975,
      "grad_norm": 8.1338529586792,
      "learning_rate": 2.2597166270745603e-05,
      "loss": 2.586288642883301,
      "memory(GiB)": 77.56,
      "step": 79900,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437214
    },
    {
      "epoch": 3.423375176727647,
      "grad_norm": 7.2224507331848145,
      "learning_rate": 2.2591537459532287e-05,
      "loss": 2.2072099685668944,
      "memory(GiB)": 77.56,
      "step": 79905,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.437185
    },
    {
      "epoch": 3.4235893920568956,
      "grad_norm": 5.806991100311279,
      "learning_rate": 2.258590914484898e-05,
      "loss": 2.345249557495117,
      "memory(GiB)": 77.56,
      "step": 79910,
      "token_acc": 0.47041420118343197,
      "train_speed(iter/s)": 1.437195
    },
    {
      "epoch": 3.4238036073861444,
      "grad_norm": 6.364602088928223,
      "learning_rate": 2.2580281326797653e-05,
      "loss": 2.1221460342407226,
      "memory(GiB)": 77.56,
      "step": 79915,
      "token_acc": 0.5363321799307958,
      "train_speed(iter/s)": 1.437202
    },
    {
      "epoch": 3.4240178227153937,
      "grad_norm": 7.039035797119141,
      "learning_rate": 2.2574654005480232e-05,
      "loss": 2.511148452758789,
      "memory(GiB)": 77.56,
      "step": 79920,
      "token_acc": 0.4774011299435028,
      "train_speed(iter/s)": 1.437194
    },
    {
      "epoch": 3.4242320380446425,
      "grad_norm": 6.606437683105469,
      "learning_rate": 2.2569027180998697e-05,
      "loss": 2.268423652648926,
      "memory(GiB)": 77.56,
      "step": 79925,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.437214
    },
    {
      "epoch": 3.4244462533738913,
      "grad_norm": 5.700983047485352,
      "learning_rate": 2.2563400853454963e-05,
      "loss": 2.458114433288574,
      "memory(GiB)": 77.56,
      "step": 79930,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.4246604687031406,
      "grad_norm": 6.004363059997559,
      "learning_rate": 2.2557775022950948e-05,
      "loss": 2.214258575439453,
      "memory(GiB)": 77.56,
      "step": 79935,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.437238
    },
    {
      "epoch": 3.4248746840323894,
      "grad_norm": 5.716898441314697,
      "learning_rate": 2.25521496895886e-05,
      "loss": 2.24503173828125,
      "memory(GiB)": 77.56,
      "step": 79940,
      "token_acc": 0.5480769230769231,
      "train_speed(iter/s)": 1.437258
    },
    {
      "epoch": 3.425088899361638,
      "grad_norm": 5.839542865753174,
      "learning_rate": 2.2546524853469802e-05,
      "loss": 2.3188934326171875,
      "memory(GiB)": 77.56,
      "step": 79945,
      "token_acc": 0.4789156626506024,
      "train_speed(iter/s)": 1.437272
    },
    {
      "epoch": 3.4253031146908874,
      "grad_norm": 5.682053089141846,
      "learning_rate": 2.2540900514696466e-05,
      "loss": 2.1888347625732423,
      "memory(GiB)": 77.56,
      "step": 79950,
      "token_acc": 0.5428571428571428,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 3.4255173300201363,
      "grad_norm": 6.361555576324463,
      "learning_rate": 2.2535276673370453e-05,
      "loss": 2.4863655090332033,
      "memory(GiB)": 77.56,
      "step": 79955,
      "token_acc": 0.5179153094462541,
      "train_speed(iter/s)": 1.437287
    },
    {
      "epoch": 3.425731545349385,
      "grad_norm": 6.168817043304443,
      "learning_rate": 2.2529653329593686e-05,
      "loss": 2.3084014892578124,
      "memory(GiB)": 77.56,
      "step": 79960,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.437265
    },
    {
      "epoch": 3.4259457606786343,
      "grad_norm": 6.668516635894775,
      "learning_rate": 2.2524030483468024e-05,
      "loss": 2.289501953125,
      "memory(GiB)": 77.56,
      "step": 79965,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.437282
    },
    {
      "epoch": 3.426159976007883,
      "grad_norm": 7.366813659667969,
      "learning_rate": 2.251840813509532e-05,
      "loss": 2.098384666442871,
      "memory(GiB)": 77.56,
      "step": 79970,
      "token_acc": 0.5466237942122186,
      "train_speed(iter/s)": 1.437301
    },
    {
      "epoch": 3.426374191337132,
      "grad_norm": 7.163089752197266,
      "learning_rate": 2.251278628457743e-05,
      "loss": 2.420930862426758,
      "memory(GiB)": 77.56,
      "step": 79975,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 3.426588406666381,
      "grad_norm": 8.034339904785156,
      "learning_rate": 2.2507164932016188e-05,
      "loss": 2.158621406555176,
      "memory(GiB)": 77.56,
      "step": 79980,
      "token_acc": 0.4867549668874172,
      "train_speed(iter/s)": 1.437328
    },
    {
      "epoch": 3.42680262199563,
      "grad_norm": 6.220357418060303,
      "learning_rate": 2.250154407751346e-05,
      "loss": 2.057103157043457,
      "memory(GiB)": 77.56,
      "step": 79985,
      "token_acc": 0.5382059800664452,
      "train_speed(iter/s)": 1.43731
    },
    {
      "epoch": 3.427016837324879,
      "grad_norm": 5.75319766998291,
      "learning_rate": 2.2495923721171053e-05,
      "loss": 2.724925994873047,
      "memory(GiB)": 77.56,
      "step": 79990,
      "token_acc": 0.450920245398773,
      "train_speed(iter/s)": 1.437317
    },
    {
      "epoch": 3.427231052654128,
      "grad_norm": 5.656720161437988,
      "learning_rate": 2.2490303863090793e-05,
      "loss": 2.4372371673583983,
      "memory(GiB)": 77.56,
      "step": 79995,
      "token_acc": 0.49107142857142855,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 3.427445267983377,
      "grad_norm": 6.461557865142822,
      "learning_rate": 2.2484684503374487e-05,
      "loss": 2.1339502334594727,
      "memory(GiB)": 77.56,
      "step": 80000,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.437292
    },
    {
      "epoch": 3.427445267983377,
      "eval_loss": 2.102041244506836,
      "eval_runtime": 14.3997,
      "eval_samples_per_second": 6.945,
      "eval_steps_per_second": 6.945,
      "eval_token_acc": 0.4992679355783309,
      "step": 80000
    },
    {
      "epoch": 3.4276594833126257,
      "grad_norm": 4.7783098220825195,
      "learning_rate": 2.2479065642123908e-05,
      "loss": 2.2108074188232423,
      "memory(GiB)": 77.56,
      "step": 80005,
      "token_acc": 0.4952978056426332,
      "train_speed(iter/s)": 1.436891
    },
    {
      "epoch": 3.427873698641875,
      "grad_norm": 7.851231098175049,
      "learning_rate": 2.2473447279440875e-05,
      "loss": 2.437786865234375,
      "memory(GiB)": 77.56,
      "step": 80010,
      "token_acc": 0.5163934426229508,
      "train_speed(iter/s)": 1.436896
    },
    {
      "epoch": 3.428087913971124,
      "grad_norm": 5.642972946166992,
      "learning_rate": 2.246782941542718e-05,
      "loss": 2.296259880065918,
      "memory(GiB)": 77.56,
      "step": 80015,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.436911
    },
    {
      "epoch": 3.4283021293003726,
      "grad_norm": 7.360104560852051,
      "learning_rate": 2.2462212050184594e-05,
      "loss": 2.5477380752563477,
      "memory(GiB)": 77.56,
      "step": 80020,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.436923
    },
    {
      "epoch": 3.428516344629622,
      "grad_norm": 5.5510711669921875,
      "learning_rate": 2.245659518381486e-05,
      "loss": 2.2331222534179687,
      "memory(GiB)": 77.56,
      "step": 80025,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.43693
    },
    {
      "epoch": 3.4287305599588707,
      "grad_norm": 5.894035816192627,
      "learning_rate": 2.2450978816419742e-05,
      "loss": 2.2158954620361326,
      "memory(GiB)": 77.56,
      "step": 80030,
      "token_acc": 0.5469798657718121,
      "train_speed(iter/s)": 1.436942
    },
    {
      "epoch": 3.4289447752881195,
      "grad_norm": 10.432945251464844,
      "learning_rate": 2.244536294810099e-05,
      "loss": 2.4966503143310548,
      "memory(GiB)": 77.56,
      "step": 80035,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.436935
    },
    {
      "epoch": 3.4291589906173687,
      "grad_norm": 7.354784965515137,
      "learning_rate": 2.2439747578960318e-05,
      "loss": 2.4185523986816406,
      "memory(GiB)": 77.56,
      "step": 80040,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.436946
    },
    {
      "epoch": 3.4293732059466175,
      "grad_norm": 10.520761489868164,
      "learning_rate": 2.2434132709099487e-05,
      "loss": 2.436611557006836,
      "memory(GiB)": 77.56,
      "step": 80045,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.436966
    },
    {
      "epoch": 3.4295874212758664,
      "grad_norm": 5.536417484283447,
      "learning_rate": 2.24285183386202e-05,
      "loss": 2.587092399597168,
      "memory(GiB)": 77.56,
      "step": 80050,
      "token_acc": 0.4657534246575342,
      "train_speed(iter/s)": 1.436962
    },
    {
      "epoch": 3.4298016366051156,
      "grad_norm": 5.768876552581787,
      "learning_rate": 2.2422904467624172e-05,
      "loss": 2.2307018280029296,
      "memory(GiB)": 77.56,
      "step": 80055,
      "token_acc": 0.5201342281879194,
      "train_speed(iter/s)": 1.436981
    },
    {
      "epoch": 3.4300158519343644,
      "grad_norm": 9.289017677307129,
      "learning_rate": 2.24172910962131e-05,
      "loss": 2.284924125671387,
      "memory(GiB)": 77.56,
      "step": 80060,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.436978
    },
    {
      "epoch": 3.4302300672636132,
      "grad_norm": 7.217469692230225,
      "learning_rate": 2.241167822448865e-05,
      "loss": 2.5033828735351564,
      "memory(GiB)": 77.56,
      "step": 80065,
      "token_acc": 0.515748031496063,
      "train_speed(iter/s)": 1.436983
    },
    {
      "epoch": 3.4304442825928625,
      "grad_norm": 5.486662864685059,
      "learning_rate": 2.2406065852552556e-05,
      "loss": 2.2092723846435547,
      "memory(GiB)": 77.56,
      "step": 80070,
      "token_acc": 0.5272727272727272,
      "train_speed(iter/s)": 1.43698
    },
    {
      "epoch": 3.4306584979221113,
      "grad_norm": 5.81664514541626,
      "learning_rate": 2.240045398050646e-05,
      "loss": 2.5850425720214845,
      "memory(GiB)": 77.56,
      "step": 80075,
      "token_acc": 0.47720364741641336,
      "train_speed(iter/s)": 1.436981
    },
    {
      "epoch": 3.43087271325136,
      "grad_norm": 9.906838417053223,
      "learning_rate": 2.2394842608452026e-05,
      "loss": 2.4849132537841796,
      "memory(GiB)": 77.56,
      "step": 80080,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.436986
    },
    {
      "epoch": 3.4310869285806094,
      "grad_norm": 7.220390796661377,
      "learning_rate": 2.2389231736490896e-05,
      "loss": 2.329759216308594,
      "memory(GiB)": 77.56,
      "step": 80085,
      "token_acc": 0.5522388059701493,
      "train_speed(iter/s)": 1.436997
    },
    {
      "epoch": 3.431301143909858,
      "grad_norm": 8.10257625579834,
      "learning_rate": 2.2383621364724755e-05,
      "loss": 2.3808422088623047,
      "memory(GiB)": 77.56,
      "step": 80090,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.437012
    },
    {
      "epoch": 3.431515359239107,
      "grad_norm": 6.302786827087402,
      "learning_rate": 2.2378011493255225e-05,
      "loss": 2.5750173568725585,
      "memory(GiB)": 77.56,
      "step": 80095,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.437017
    },
    {
      "epoch": 3.4317295745683563,
      "grad_norm": 7.061655521392822,
      "learning_rate": 2.23724021221839e-05,
      "loss": 1.7369194030761719,
      "memory(GiB)": 77.56,
      "step": 80100,
      "token_acc": 0.6181818181818182,
      "train_speed(iter/s)": 1.437023
    },
    {
      "epoch": 3.431943789897605,
      "grad_norm": 4.812687873840332,
      "learning_rate": 2.236679325161245e-05,
      "loss": 2.0262184143066406,
      "memory(GiB)": 77.56,
      "step": 80105,
      "token_acc": 0.5303030303030303,
      "train_speed(iter/s)": 1.437031
    },
    {
      "epoch": 3.432158005226854,
      "grad_norm": 4.6677045822143555,
      "learning_rate": 2.2361184881642468e-05,
      "loss": 2.480453109741211,
      "memory(GiB)": 77.56,
      "step": 80110,
      "token_acc": 0.5033783783783784,
      "train_speed(iter/s)": 1.43703
    },
    {
      "epoch": 3.432372220556103,
      "grad_norm": 5.960814476013184,
      "learning_rate": 2.2355577012375538e-05,
      "loss": 2.2339080810546874,
      "memory(GiB)": 77.56,
      "step": 80115,
      "token_acc": 0.5304878048780488,
      "train_speed(iter/s)": 1.437044
    },
    {
      "epoch": 3.432586435885352,
      "grad_norm": 6.189528465270996,
      "learning_rate": 2.234996964391327e-05,
      "loss": 2.257598876953125,
      "memory(GiB)": 77.56,
      "step": 80120,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.437052
    },
    {
      "epoch": 3.4328006512146008,
      "grad_norm": 6.552147388458252,
      "learning_rate": 2.2344362776357218e-05,
      "loss": 2.324764633178711,
      "memory(GiB)": 77.56,
      "step": 80125,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437069
    },
    {
      "epoch": 3.43301486654385,
      "grad_norm": 6.442381858825684,
      "learning_rate": 2.2338756409809004e-05,
      "loss": 2.1429094314575194,
      "memory(GiB)": 77.56,
      "step": 80130,
      "token_acc": 0.541095890410959,
      "train_speed(iter/s)": 1.437075
    },
    {
      "epoch": 3.433229081873099,
      "grad_norm": 6.968410491943359,
      "learning_rate": 2.2333150544370157e-05,
      "loss": 2.0460203170776365,
      "memory(GiB)": 77.56,
      "step": 80135,
      "token_acc": 0.5372549019607843,
      "train_speed(iter/s)": 1.437077
    },
    {
      "epoch": 3.4334432972023476,
      "grad_norm": 6.643887042999268,
      "learning_rate": 2.2327545180142247e-05,
      "loss": 2.3601781845092775,
      "memory(GiB)": 77.56,
      "step": 80140,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.43706
    },
    {
      "epoch": 3.433657512531597,
      "grad_norm": 6.591537952423096,
      "learning_rate": 2.232194031722681e-05,
      "loss": 2.4266279220581053,
      "memory(GiB)": 77.56,
      "step": 80145,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.437034
    },
    {
      "epoch": 3.4338717278608457,
      "grad_norm": 5.752573490142822,
      "learning_rate": 2.231633595572537e-05,
      "loss": 2.5533803939819335,
      "memory(GiB)": 77.56,
      "step": 80150,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437053
    },
    {
      "epoch": 3.4340859431900945,
      "grad_norm": 5.829822063446045,
      "learning_rate": 2.2310732095739494e-05,
      "loss": 2.509723663330078,
      "memory(GiB)": 77.56,
      "step": 80155,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.437068
    },
    {
      "epoch": 3.434300158519344,
      "grad_norm": 5.743741035461426,
      "learning_rate": 2.2305128737370662e-05,
      "loss": 2.101865768432617,
      "memory(GiB)": 77.56,
      "step": 80160,
      "token_acc": 0.5474137931034483,
      "train_speed(iter/s)": 1.437026
    },
    {
      "epoch": 3.4345143738485926,
      "grad_norm": 6.74943208694458,
      "learning_rate": 2.2299525880720424e-05,
      "loss": 2.2783987045288088,
      "memory(GiB)": 77.56,
      "step": 80165,
      "token_acc": 0.5223367697594502,
      "train_speed(iter/s)": 1.437047
    },
    {
      "epoch": 3.4347285891778414,
      "grad_norm": 5.304409027099609,
      "learning_rate": 2.2293923525890258e-05,
      "loss": 2.2710968017578126,
      "memory(GiB)": 77.56,
      "step": 80170,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.437051
    },
    {
      "epoch": 3.4349428045070907,
      "grad_norm": 5.887391567230225,
      "learning_rate": 2.2288321672981655e-05,
      "loss": 2.3437952041625976,
      "memory(GiB)": 77.56,
      "step": 80175,
      "token_acc": 0.5179856115107914,
      "train_speed(iter/s)": 1.437029
    },
    {
      "epoch": 3.4351570198363395,
      "grad_norm": 7.42817497253418,
      "learning_rate": 2.22827203220961e-05,
      "loss": 2.3383115768432616,
      "memory(GiB)": 77.56,
      "step": 80180,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.437055
    },
    {
      "epoch": 3.4353712351655883,
      "grad_norm": 6.673346996307373,
      "learning_rate": 2.2277119473335058e-05,
      "loss": 2.538251495361328,
      "memory(GiB)": 77.56,
      "step": 80185,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.437071
    },
    {
      "epoch": 3.4355854504948375,
      "grad_norm": 5.708316326141357,
      "learning_rate": 2.2271519126800013e-05,
      "loss": 2.402574157714844,
      "memory(GiB)": 77.56,
      "step": 80190,
      "token_acc": 0.5234657039711191,
      "train_speed(iter/s)": 1.437039
    },
    {
      "epoch": 3.4357996658240864,
      "grad_norm": 4.448325157165527,
      "learning_rate": 2.2265919282592412e-05,
      "loss": 2.2469444274902344,
      "memory(GiB)": 77.56,
      "step": 80195,
      "token_acc": 0.5246913580246914,
      "train_speed(iter/s)": 1.437039
    },
    {
      "epoch": 3.436013881153335,
      "grad_norm": 6.499133586883545,
      "learning_rate": 2.22603199408137e-05,
      "loss": 2.289023590087891,
      "memory(GiB)": 77.56,
      "step": 80200,
      "token_acc": 0.5163398692810458,
      "train_speed(iter/s)": 1.437061
    },
    {
      "epoch": 3.4362280964825844,
      "grad_norm": 7.010324001312256,
      "learning_rate": 2.225472110156531e-05,
      "loss": 2.2964408874511717,
      "memory(GiB)": 77.56,
      "step": 80205,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.437076
    },
    {
      "epoch": 3.4364423118118332,
      "grad_norm": 7.886505126953125,
      "learning_rate": 2.2249122764948665e-05,
      "loss": 2.2595388412475588,
      "memory(GiB)": 77.56,
      "step": 80210,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.437097
    },
    {
      "epoch": 3.436656527141082,
      "grad_norm": 6.966797351837158,
      "learning_rate": 2.2243524931065206e-05,
      "loss": 2.2966806411743166,
      "memory(GiB)": 77.56,
      "step": 80215,
      "token_acc": 0.5169230769230769,
      "train_speed(iter/s)": 1.437118
    },
    {
      "epoch": 3.4368707424703313,
      "grad_norm": 6.148560523986816,
      "learning_rate": 2.2237927600016334e-05,
      "loss": 2.2613367080688476,
      "memory(GiB)": 77.56,
      "step": 80220,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.437124
    },
    {
      "epoch": 3.43708495779958,
      "grad_norm": 5.912735462188721,
      "learning_rate": 2.223233077190344e-05,
      "loss": 2.331842041015625,
      "memory(GiB)": 77.56,
      "step": 80225,
      "token_acc": 0.47904191616766467,
      "train_speed(iter/s)": 1.437129
    },
    {
      "epoch": 3.437299173128829,
      "grad_norm": 7.128683567047119,
      "learning_rate": 2.222673444682793e-05,
      "loss": 2.6025352478027344,
      "memory(GiB)": 77.56,
      "step": 80230,
      "token_acc": 0.5039370078740157,
      "train_speed(iter/s)": 1.437135
    },
    {
      "epoch": 3.437513388458078,
      "grad_norm": 6.587675094604492,
      "learning_rate": 2.2221138624891158e-05,
      "loss": 2.3038351058959963,
      "memory(GiB)": 77.56,
      "step": 80235,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.437135
    },
    {
      "epoch": 3.437727603787327,
      "grad_norm": 7.692835330963135,
      "learning_rate": 2.2215543306194535e-05,
      "loss": 2.3964452743530273,
      "memory(GiB)": 77.56,
      "step": 80240,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.437139
    },
    {
      "epoch": 3.437941819116576,
      "grad_norm": 7.405257225036621,
      "learning_rate": 2.2209948490839388e-05,
      "loss": 2.551258087158203,
      "memory(GiB)": 77.56,
      "step": 80245,
      "token_acc": 0.5054151624548736,
      "train_speed(iter/s)": 1.437139
    },
    {
      "epoch": 3.438156034445825,
      "grad_norm": 12.880958557128906,
      "learning_rate": 2.2204354178927116e-05,
      "loss": 2.5628082275390627,
      "memory(GiB)": 77.56,
      "step": 80250,
      "token_acc": 0.47651006711409394,
      "train_speed(iter/s)": 1.437156
    },
    {
      "epoch": 3.438370249775074,
      "grad_norm": 4.969246864318848,
      "learning_rate": 2.2198760370559048e-05,
      "loss": 1.9327110290527343,
      "memory(GiB)": 77.56,
      "step": 80255,
      "token_acc": 0.5895196506550219,
      "train_speed(iter/s)": 1.437161
    },
    {
      "epoch": 3.4385844651043227,
      "grad_norm": 6.226701259613037,
      "learning_rate": 2.2193167065836506e-05,
      "loss": 2.1575859069824217,
      "memory(GiB)": 77.56,
      "step": 80260,
      "token_acc": 0.5165562913907285,
      "train_speed(iter/s)": 1.437181
    },
    {
      "epoch": 3.438798680433572,
      "grad_norm": 5.530786514282227,
      "learning_rate": 2.218757426486083e-05,
      "loss": 2.1741479873657226,
      "memory(GiB)": 77.56,
      "step": 80265,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437188
    },
    {
      "epoch": 3.4390128957628208,
      "grad_norm": 6.205590724945068,
      "learning_rate": 2.2181981967733316e-05,
      "loss": 2.382988166809082,
      "memory(GiB)": 77.56,
      "step": 80270,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437158
    },
    {
      "epoch": 3.4392271110920696,
      "grad_norm": 5.49648904800415,
      "learning_rate": 2.2176390174555316e-05,
      "loss": 2.3088052749633787,
      "memory(GiB)": 77.56,
      "step": 80275,
      "token_acc": 0.5141242937853108,
      "train_speed(iter/s)": 1.437171
    },
    {
      "epoch": 3.439441326421319,
      "grad_norm": 6.080382823944092,
      "learning_rate": 2.2170798885428113e-05,
      "loss": 2.34581298828125,
      "memory(GiB)": 77.56,
      "step": 80280,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.437187
    },
    {
      "epoch": 3.4396555417505676,
      "grad_norm": 5.663379192352295,
      "learning_rate": 2.2165208100452982e-05,
      "loss": 2.0927642822265624,
      "memory(GiB)": 77.56,
      "step": 80285,
      "token_acc": 0.5470383275261324,
      "train_speed(iter/s)": 1.437202
    },
    {
      "epoch": 3.4398697570798165,
      "grad_norm": 6.5682597160339355,
      "learning_rate": 2.2159617819731226e-05,
      "loss": 2.1643686294555664,
      "memory(GiB)": 77.56,
      "step": 80290,
      "token_acc": 0.5127272727272727,
      "train_speed(iter/s)": 1.43722
    },
    {
      "epoch": 3.4400839724090657,
      "grad_norm": 5.731741428375244,
      "learning_rate": 2.215402804336409e-05,
      "loss": 2.268716239929199,
      "memory(GiB)": 77.56,
      "step": 80295,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.437241
    },
    {
      "epoch": 3.4402981877383145,
      "grad_norm": 6.32515811920166,
      "learning_rate": 2.214843877145288e-05,
      "loss": 2.220654296875,
      "memory(GiB)": 77.56,
      "step": 80300,
      "token_acc": 0.5042016806722689,
      "train_speed(iter/s)": 1.437243
    },
    {
      "epoch": 3.4405124030675633,
      "grad_norm": 7.128988742828369,
      "learning_rate": 2.2142850004098804e-05,
      "loss": 2.2158023834228517,
      "memory(GiB)": 77.56,
      "step": 80305,
      "token_acc": 0.5435684647302904,
      "train_speed(iter/s)": 1.437264
    },
    {
      "epoch": 3.4407266183968126,
      "grad_norm": 5.263505935668945,
      "learning_rate": 2.2137261741403154e-05,
      "loss": 2.6984487533569337,
      "memory(GiB)": 77.56,
      "step": 80310,
      "token_acc": 0.48324022346368717,
      "train_speed(iter/s)": 1.437282
    },
    {
      "epoch": 3.4409408337260614,
      "grad_norm": 5.217807769775391,
      "learning_rate": 2.213167398346714e-05,
      "loss": 2.5654275894165037,
      "memory(GiB)": 77.56,
      "step": 80315,
      "token_acc": 0.4356955380577428,
      "train_speed(iter/s)": 1.437275
    },
    {
      "epoch": 3.4411550490553102,
      "grad_norm": 5.2641706466674805,
      "learning_rate": 2.2126086730391997e-05,
      "loss": 2.4469430923461912,
      "memory(GiB)": 77.56,
      "step": 80320,
      "token_acc": 0.477124183006536,
      "train_speed(iter/s)": 1.437284
    },
    {
      "epoch": 3.4413692643845595,
      "grad_norm": 5.480915069580078,
      "learning_rate": 2.2120499982278942e-05,
      "loss": 2.5283790588378907,
      "memory(GiB)": 77.56,
      "step": 80325,
      "token_acc": 0.4373401534526854,
      "train_speed(iter/s)": 1.437297
    },
    {
      "epoch": 3.4415834797138083,
      "grad_norm": 6.123584270477295,
      "learning_rate": 2.2114913739229166e-05,
      "loss": 2.2348833084106445,
      "memory(GiB)": 77.56,
      "step": 80330,
      "token_acc": 0.5063694267515924,
      "train_speed(iter/s)": 1.437289
    },
    {
      "epoch": 3.441797695043057,
      "grad_norm": 6.341625690460205,
      "learning_rate": 2.2109328001343898e-05,
      "loss": 2.2139204025268553,
      "memory(GiB)": 77.56,
      "step": 80335,
      "token_acc": 0.5340909090909091,
      "train_speed(iter/s)": 1.437293
    },
    {
      "epoch": 3.4420119103723064,
      "grad_norm": 5.228277206420898,
      "learning_rate": 2.2103742768724322e-05,
      "loss": 2.1739824295043944,
      "memory(GiB)": 77.56,
      "step": 80340,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437295
    },
    {
      "epoch": 3.442226125701555,
      "grad_norm": 7.82539176940918,
      "learning_rate": 2.2098158041471617e-05,
      "loss": 2.5690967559814455,
      "memory(GiB)": 77.56,
      "step": 80345,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.437308
    },
    {
      "epoch": 3.442440341030804,
      "grad_norm": 8.47783374786377,
      "learning_rate": 2.2092573819686945e-05,
      "loss": 2.5492820739746094,
      "memory(GiB)": 77.56,
      "step": 80350,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 3.4426545563600532,
      "grad_norm": 6.251232147216797,
      "learning_rate": 2.2086990103471468e-05,
      "loss": 2.719070816040039,
      "memory(GiB)": 77.56,
      "step": 80355,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.43734
    },
    {
      "epoch": 3.442868771689302,
      "grad_norm": 6.333085536956787,
      "learning_rate": 2.2081406892926366e-05,
      "loss": 2.5249135971069334,
      "memory(GiB)": 77.56,
      "step": 80360,
      "token_acc": 0.4709897610921502,
      "train_speed(iter/s)": 1.437364
    },
    {
      "epoch": 3.443082987018551,
      "grad_norm": 4.731055736541748,
      "learning_rate": 2.2075824188152762e-05,
      "loss": 2.313634490966797,
      "memory(GiB)": 77.56,
      "step": 80365,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.437368
    },
    {
      "epoch": 3.4432972023478,
      "grad_norm": 6.09040641784668,
      "learning_rate": 2.2070241989251805e-05,
      "loss": 2.452932929992676,
      "memory(GiB)": 77.56,
      "step": 80370,
      "token_acc": 0.49221183800623053,
      "train_speed(iter/s)": 1.437355
    },
    {
      "epoch": 3.443511417677049,
      "grad_norm": 5.076159477233887,
      "learning_rate": 2.2064660296324607e-05,
      "loss": 2.124247169494629,
      "memory(GiB)": 77.56,
      "step": 80375,
      "token_acc": 0.5444444444444444,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.4437256330062977,
      "grad_norm": 7.787264823913574,
      "learning_rate": 2.205907910947228e-05,
      "loss": 2.140587043762207,
      "memory(GiB)": 77.56,
      "step": 80380,
      "token_acc": 0.5412186379928315,
      "train_speed(iter/s)": 1.437382
    },
    {
      "epoch": 3.443939848335547,
      "grad_norm": 4.084612846374512,
      "learning_rate": 2.2053498428795965e-05,
      "loss": 2.4007919311523436,
      "memory(GiB)": 77.56,
      "step": 80385,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 3.444154063664796,
      "grad_norm": 6.645567893981934,
      "learning_rate": 2.2047918254396722e-05,
      "loss": 2.06510066986084,
      "memory(GiB)": 77.56,
      "step": 80390,
      "token_acc": 0.5309090909090909,
      "train_speed(iter/s)": 1.437396
    },
    {
      "epoch": 3.4443682789940446,
      "grad_norm": 4.380014896392822,
      "learning_rate": 2.2042338586375676e-05,
      "loss": 2.4250310897827148,
      "memory(GiB)": 77.56,
      "step": 80395,
      "token_acc": 0.49173553719008267,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 3.444582494323294,
      "grad_norm": 7.077320575714111,
      "learning_rate": 2.2036759424833898e-05,
      "loss": 2.5221746444702147,
      "memory(GiB)": 77.56,
      "step": 80400,
      "token_acc": 0.4277286135693215,
      "train_speed(iter/s)": 1.437417
    },
    {
      "epoch": 3.4447967096525427,
      "grad_norm": 4.8981194496154785,
      "learning_rate": 2.2031180769872456e-05,
      "loss": 2.5382272720336916,
      "memory(GiB)": 77.56,
      "step": 80405,
      "token_acc": 0.527027027027027,
      "train_speed(iter/s)": 1.437427
    },
    {
      "epoch": 3.4450109249817915,
      "grad_norm": 4.959420680999756,
      "learning_rate": 2.2025602621592404e-05,
      "loss": 2.1550403594970704,
      "memory(GiB)": 77.56,
      "step": 80410,
      "token_acc": 0.521978021978022,
      "train_speed(iter/s)": 1.437436
    },
    {
      "epoch": 3.4452251403110408,
      "grad_norm": 6.577432155609131,
      "learning_rate": 2.2020024980094788e-05,
      "loss": 2.2065486907958984,
      "memory(GiB)": 77.56,
      "step": 80415,
      "token_acc": 0.5034722222222222,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.4454393556402896,
      "grad_norm": 5.82411003112793,
      "learning_rate": 2.2014447845480684e-05,
      "loss": 2.485791015625,
      "memory(GiB)": 77.56,
      "step": 80420,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437454
    },
    {
      "epoch": 3.4456535709695384,
      "grad_norm": 6.361300945281982,
      "learning_rate": 2.2008871217851114e-05,
      "loss": 2.391726493835449,
      "memory(GiB)": 77.56,
      "step": 80425,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.437434
    },
    {
      "epoch": 3.4458677862987876,
      "grad_norm": 5.705111503601074,
      "learning_rate": 2.2003295097307097e-05,
      "loss": 2.0951635360717775,
      "memory(GiB)": 77.56,
      "step": 80430,
      "token_acc": 0.5692883895131086,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.4460820016280365,
      "grad_norm": 6.383053779602051,
      "learning_rate": 2.1997719483949652e-05,
      "loss": 2.1492645263671877,
      "memory(GiB)": 77.56,
      "step": 80435,
      "token_acc": 0.5457627118644067,
      "train_speed(iter/s)": 1.43743
    },
    {
      "epoch": 3.4462962169572853,
      "grad_norm": 9.024420738220215,
      "learning_rate": 2.199214437787977e-05,
      "loss": 2.33245735168457,
      "memory(GiB)": 77.56,
      "step": 80440,
      "token_acc": 0.5052631578947369,
      "train_speed(iter/s)": 1.437444
    },
    {
      "epoch": 3.4465104322865345,
      "grad_norm": 6.605022430419922,
      "learning_rate": 2.1986569779198486e-05,
      "loss": 2.446097755432129,
      "memory(GiB)": 77.56,
      "step": 80445,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.437459
    },
    {
      "epoch": 3.4467246476157833,
      "grad_norm": 7.3540358543396,
      "learning_rate": 2.198099568800676e-05,
      "loss": 2.448853874206543,
      "memory(GiB)": 77.56,
      "step": 80450,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 3.446938862945032,
      "grad_norm": 6.087551116943359,
      "learning_rate": 2.1975422104405576e-05,
      "loss": 2.3472049713134764,
      "memory(GiB)": 77.56,
      "step": 80455,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 3.4471530782742814,
      "grad_norm": 5.073938846588135,
      "learning_rate": 2.1969849028495915e-05,
      "loss": 1.8209077835083007,
      "memory(GiB)": 77.56,
      "step": 80460,
      "token_acc": 0.5862068965517241,
      "train_speed(iter/s)": 1.437427
    },
    {
      "epoch": 3.4473672936035302,
      "grad_norm": 4.9377031326293945,
      "learning_rate": 2.196427646037874e-05,
      "loss": 2.3113643646240236,
      "memory(GiB)": 77.56,
      "step": 80465,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437429
    },
    {
      "epoch": 3.447581508932779,
      "grad_norm": 6.210414886474609,
      "learning_rate": 2.1958704400154995e-05,
      "loss": 2.5444915771484373,
      "memory(GiB)": 77.56,
      "step": 80470,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.4477957242620283,
      "grad_norm": 4.72358512878418,
      "learning_rate": 2.195313284792561e-05,
      "loss": 2.432740020751953,
      "memory(GiB)": 77.56,
      "step": 80475,
      "token_acc": 0.4744318181818182,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 3.448009939591277,
      "grad_norm": 6.679464817047119,
      "learning_rate": 2.194756180379155e-05,
      "loss": 2.223580741882324,
      "memory(GiB)": 77.56,
      "step": 80480,
      "token_acc": 0.5143769968051118,
      "train_speed(iter/s)": 1.437472
    },
    {
      "epoch": 3.448224154920526,
      "grad_norm": 6.284995079040527,
      "learning_rate": 2.1941991267853722e-05,
      "loss": 2.3431787490844727,
      "memory(GiB)": 77.56,
      "step": 80485,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.437477
    },
    {
      "epoch": 3.448438370249775,
      "grad_norm": 5.769493103027344,
      "learning_rate": 2.193642124021304e-05,
      "loss": 2.043992042541504,
      "memory(GiB)": 77.56,
      "step": 80490,
      "token_acc": 0.5387596899224806,
      "train_speed(iter/s)": 1.43748
    },
    {
      "epoch": 3.448652585579024,
      "grad_norm": 8.8023042678833,
      "learning_rate": 2.193085172097042e-05,
      "loss": 2.373923683166504,
      "memory(GiB)": 77.56,
      "step": 80495,
      "token_acc": 0.45980707395498394,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.448866800908273,
      "grad_norm": 4.613881587982178,
      "learning_rate": 2.192528271022673e-05,
      "loss": 2.3887168884277346,
      "memory(GiB)": 77.56,
      "step": 80500,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.437491
    },
    {
      "epoch": 3.448866800908273,
      "eval_loss": 2.26839542388916,
      "eval_runtime": 14.1365,
      "eval_samples_per_second": 7.074,
      "eval_steps_per_second": 7.074,
      "eval_token_acc": 0.47979139504563234,
      "step": 80500
    },
    {
      "epoch": 3.449081016237522,
      "grad_norm": 5.859959602355957,
      "learning_rate": 2.1919714208082898e-05,
      "loss": 2.2975589752197267,
      "memory(GiB)": 77.56,
      "step": 80505,
      "token_acc": 0.4976303317535545,
      "train_speed(iter/s)": 1.437116
    },
    {
      "epoch": 3.449295231566771,
      "grad_norm": 5.504870414733887,
      "learning_rate": 2.1914146214639782e-05,
      "loss": 2.1874622344970702,
      "memory(GiB)": 77.56,
      "step": 80510,
      "token_acc": 0.540084388185654,
      "train_speed(iter/s)": 1.437122
    },
    {
      "epoch": 3.4495094468960197,
      "grad_norm": 4.164382457733154,
      "learning_rate": 2.1908578729998258e-05,
      "loss": 2.2466367721557616,
      "memory(GiB)": 77.56,
      "step": 80515,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437137
    },
    {
      "epoch": 3.449723662225269,
      "grad_norm": 4.9224114418029785,
      "learning_rate": 2.1903011754259177e-05,
      "loss": 2.3191972732543946,
      "memory(GiB)": 77.56,
      "step": 80520,
      "token_acc": 0.47038327526132406,
      "train_speed(iter/s)": 1.437121
    },
    {
      "epoch": 3.4499378775545178,
      "grad_norm": 9.172769546508789,
      "learning_rate": 2.1897445287523384e-05,
      "loss": 2.249551773071289,
      "memory(GiB)": 77.56,
      "step": 80525,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.437142
    },
    {
      "epoch": 3.4501520928837666,
      "grad_norm": 6.565616607666016,
      "learning_rate": 2.1891879329891746e-05,
      "loss": 2.2075742721557616,
      "memory(GiB)": 77.56,
      "step": 80530,
      "token_acc": 0.5387453874538746,
      "train_speed(iter/s)": 1.437177
    },
    {
      "epoch": 3.450366308213016,
      "grad_norm": 6.820053577423096,
      "learning_rate": 2.1886313881465063e-05,
      "loss": 2.319544792175293,
      "memory(GiB)": 77.56,
      "step": 80535,
      "token_acc": 0.5274725274725275,
      "train_speed(iter/s)": 1.43718
    },
    {
      "epoch": 3.4505805235422646,
      "grad_norm": 7.024238109588623,
      "learning_rate": 2.1880748942344198e-05,
      "loss": 2.1867679595947265,
      "memory(GiB)": 77.56,
      "step": 80540,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.437191
    },
    {
      "epoch": 3.4507947388715134,
      "grad_norm": 6.06240177154541,
      "learning_rate": 2.1875184512629943e-05,
      "loss": 2.3003362655639648,
      "memory(GiB)": 77.56,
      "step": 80545,
      "token_acc": 0.5034965034965035,
      "train_speed(iter/s)": 1.437214
    },
    {
      "epoch": 3.4510089542007627,
      "grad_norm": 5.127089023590088,
      "learning_rate": 2.18696205924231e-05,
      "loss": 2.335889434814453,
      "memory(GiB)": 77.56,
      "step": 80550,
      "token_acc": 0.5154320987654321,
      "train_speed(iter/s)": 1.437234
    },
    {
      "epoch": 3.4512231695300115,
      "grad_norm": 5.436455249786377,
      "learning_rate": 2.186405718182447e-05,
      "loss": 2.5750221252441405,
      "memory(GiB)": 77.56,
      "step": 80555,
      "token_acc": 0.4600760456273764,
      "train_speed(iter/s)": 1.437237
    },
    {
      "epoch": 3.4514373848592603,
      "grad_norm": 5.467102527618408,
      "learning_rate": 2.1858494280934826e-05,
      "loss": 2.3708337783813476,
      "memory(GiB)": 77.56,
      "step": 80560,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.43724
    },
    {
      "epoch": 3.4516516001885096,
      "grad_norm": 6.265877723693848,
      "learning_rate": 2.1852931889854966e-05,
      "loss": 2.397037124633789,
      "memory(GiB)": 77.56,
      "step": 80565,
      "token_acc": 0.49809885931558934,
      "train_speed(iter/s)": 1.437243
    },
    {
      "epoch": 3.4518658155177584,
      "grad_norm": 6.674444198608398,
      "learning_rate": 2.1847370008685657e-05,
      "loss": 2.2196815490722654,
      "memory(GiB)": 77.56,
      "step": 80570,
      "token_acc": 0.519434628975265,
      "train_speed(iter/s)": 1.437253
    },
    {
      "epoch": 3.452080030847007,
      "grad_norm": 5.629004955291748,
      "learning_rate": 2.1841808637527645e-05,
      "loss": 2.157611083984375,
      "memory(GiB)": 77.56,
      "step": 80575,
      "token_acc": 0.5149501661129569,
      "train_speed(iter/s)": 1.437258
    },
    {
      "epoch": 3.4522942461762565,
      "grad_norm": 6.988944053649902,
      "learning_rate": 2.1836247776481684e-05,
      "loss": 2.4590389251708986,
      "memory(GiB)": 77.56,
      "step": 80580,
      "token_acc": 0.4573170731707317,
      "train_speed(iter/s)": 1.437261
    },
    {
      "epoch": 3.4525084615055053,
      "grad_norm": 5.524704456329346,
      "learning_rate": 2.1830687425648495e-05,
      "loss": 2.4149913787841797,
      "memory(GiB)": 77.56,
      "step": 80585,
      "token_acc": 0.5404411764705882,
      "train_speed(iter/s)": 1.437279
    },
    {
      "epoch": 3.452722676834754,
      "grad_norm": 4.795392036437988,
      "learning_rate": 2.182512758512885e-05,
      "loss": 2.159110450744629,
      "memory(GiB)": 77.56,
      "step": 80590,
      "token_acc": 0.5636363636363636,
      "train_speed(iter/s)": 1.437286
    },
    {
      "epoch": 3.4529368921640033,
      "grad_norm": 6.560310363769531,
      "learning_rate": 2.181956825502345e-05,
      "loss": 2.130175018310547,
      "memory(GiB)": 77.56,
      "step": 80595,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.437284
    },
    {
      "epoch": 3.453151107493252,
      "grad_norm": 5.127025127410889,
      "learning_rate": 2.181400943543298e-05,
      "loss": 2.2630002975463865,
      "memory(GiB)": 77.56,
      "step": 80600,
      "token_acc": 0.5249169435215947,
      "train_speed(iter/s)": 1.437304
    },
    {
      "epoch": 3.453365322822501,
      "grad_norm": 5.843238353729248,
      "learning_rate": 2.1808451126458196e-05,
      "loss": 2.145084571838379,
      "memory(GiB)": 77.56,
      "step": 80605,
      "token_acc": 0.5318352059925093,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 3.4535795381517502,
      "grad_norm": 5.5879621505737305,
      "learning_rate": 2.1802893328199763e-05,
      "loss": 2.4597858428955077,
      "memory(GiB)": 77.56,
      "step": 80610,
      "token_acc": 0.45422535211267606,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 3.453793753480999,
      "grad_norm": 8.249075889587402,
      "learning_rate": 2.1797336040758365e-05,
      "loss": 2.7451202392578127,
      "memory(GiB)": 77.56,
      "step": 80615,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.437363
    },
    {
      "epoch": 3.454007968810248,
      "grad_norm": 5.4733428955078125,
      "learning_rate": 2.179177926423467e-05,
      "loss": 2.4360809326171875,
      "memory(GiB)": 77.56,
      "step": 80620,
      "token_acc": 0.4672897196261682,
      "train_speed(iter/s)": 1.437342
    },
    {
      "epoch": 3.454222184139497,
      "grad_norm": 6.793597221374512,
      "learning_rate": 2.1786222998729367e-05,
      "loss": 2.3753881454467773,
      "memory(GiB)": 77.56,
      "step": 80625,
      "token_acc": 0.5123456790123457,
      "train_speed(iter/s)": 1.437348
    },
    {
      "epoch": 3.454436399468746,
      "grad_norm": 5.418949604034424,
      "learning_rate": 2.1780667244343106e-05,
      "loss": 2.4566730499267577,
      "memory(GiB)": 77.56,
      "step": 80630,
      "token_acc": 0.4981684981684982,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 3.4546506147979947,
      "grad_norm": 4.494617462158203,
      "learning_rate": 2.1775112001176522e-05,
      "loss": 1.8612266540527345,
      "memory(GiB)": 77.56,
      "step": 80635,
      "token_acc": 0.5905797101449275,
      "train_speed(iter/s)": 1.437357
    },
    {
      "epoch": 3.454864830127244,
      "grad_norm": 6.100369453430176,
      "learning_rate": 2.176955726933027e-05,
      "loss": 2.3583000183105467,
      "memory(GiB)": 77.56,
      "step": 80640,
      "token_acc": 0.490625,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 3.455079045456493,
      "grad_norm": 6.376121997833252,
      "learning_rate": 2.1764003048904947e-05,
      "loss": 2.413394546508789,
      "memory(GiB)": 77.56,
      "step": 80645,
      "token_acc": 0.47,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.4552932607857416,
      "grad_norm": 5.128441333770752,
      "learning_rate": 2.1758449340001213e-05,
      "loss": 2.035100746154785,
      "memory(GiB)": 77.56,
      "step": 80650,
      "token_acc": 0.5546875,
      "train_speed(iter/s)": 1.437355
    },
    {
      "epoch": 3.455507476114991,
      "grad_norm": 13.508808135986328,
      "learning_rate": 2.175289614271967e-05,
      "loss": 2.2530609130859376,
      "memory(GiB)": 77.56,
      "step": 80655,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.437349
    },
    {
      "epoch": 3.4557216914442397,
      "grad_norm": 5.964972019195557,
      "learning_rate": 2.1747343457160902e-05,
      "loss": 2.080478858947754,
      "memory(GiB)": 77.56,
      "step": 80660,
      "token_acc": 0.5409252669039146,
      "train_speed(iter/s)": 1.437373
    },
    {
      "epoch": 3.4559359067734885,
      "grad_norm": 4.962120056152344,
      "learning_rate": 2.1741791283425516e-05,
      "loss": 2.4243751525878907,
      "memory(GiB)": 77.56,
      "step": 80665,
      "token_acc": 0.48905109489051096,
      "train_speed(iter/s)": 1.437362
    },
    {
      "epoch": 3.4561501221027378,
      "grad_norm": 8.512435913085938,
      "learning_rate": 2.173623962161407e-05,
      "loss": 2.403050994873047,
      "memory(GiB)": 77.56,
      "step": 80670,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.437386
    },
    {
      "epoch": 3.4563643374319866,
      "grad_norm": 6.178690433502197,
      "learning_rate": 2.1730688471827176e-05,
      "loss": 2.334092903137207,
      "memory(GiB)": 77.56,
      "step": 80675,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.437367
    },
    {
      "epoch": 3.4565785527612354,
      "grad_norm": 7.984079360961914,
      "learning_rate": 2.1725137834165356e-05,
      "loss": 2.353314971923828,
      "memory(GiB)": 77.56,
      "step": 80680,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.437389
    },
    {
      "epoch": 3.4567927680904846,
      "grad_norm": 5.329522132873535,
      "learning_rate": 2.171958770872921e-05,
      "loss": 2.085029411315918,
      "memory(GiB)": 77.56,
      "step": 80685,
      "token_acc": 0.5220883534136547,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 3.4570069834197334,
      "grad_norm": 6.020061016082764,
      "learning_rate": 2.171403809561926e-05,
      "loss": 2.447394371032715,
      "memory(GiB)": 77.56,
      "step": 80690,
      "token_acc": 0.4823529411764706,
      "train_speed(iter/s)": 1.437379
    },
    {
      "epoch": 3.4572211987489823,
      "grad_norm": 7.538059711456299,
      "learning_rate": 2.1708488994936048e-05,
      "loss": 2.4250808715820313,
      "memory(GiB)": 77.56,
      "step": 80695,
      "token_acc": 0.5269230769230769,
      "train_speed(iter/s)": 1.437392
    },
    {
      "epoch": 3.4574354140782315,
      "grad_norm": 5.586945533752441,
      "learning_rate": 2.1702940406780097e-05,
      "loss": 2.337316703796387,
      "memory(GiB)": 77.56,
      "step": 80700,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.437385
    },
    {
      "epoch": 3.4576496294074803,
      "grad_norm": 9.51451301574707,
      "learning_rate": 2.1697392331251904e-05,
      "loss": 2.32459716796875,
      "memory(GiB)": 77.56,
      "step": 80705,
      "token_acc": 0.5461847389558233,
      "train_speed(iter/s)": 1.437389
    },
    {
      "epoch": 3.457863844736729,
      "grad_norm": 6.780073165893555,
      "learning_rate": 2.169184476845202e-05,
      "loss": 2.3071727752685547,
      "memory(GiB)": 77.56,
      "step": 80710,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.437392
    },
    {
      "epoch": 3.4580780600659784,
      "grad_norm": 5.097660541534424,
      "learning_rate": 2.168629771848092e-05,
      "loss": 2.2894062042236327,
      "memory(GiB)": 77.56,
      "step": 80715,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.437406
    },
    {
      "epoch": 3.458292275395227,
      "grad_norm": 5.0699782371521,
      "learning_rate": 2.1680751181439096e-05,
      "loss": 2.4819961547851563,
      "memory(GiB)": 77.56,
      "step": 80720,
      "token_acc": 0.45867768595041325,
      "train_speed(iter/s)": 1.437416
    },
    {
      "epoch": 3.458506490724476,
      "grad_norm": 6.098877906799316,
      "learning_rate": 2.1675205157427035e-05,
      "loss": 2.170017433166504,
      "memory(GiB)": 77.56,
      "step": 80725,
      "token_acc": 0.534965034965035,
      "train_speed(iter/s)": 1.437401
    },
    {
      "epoch": 3.4587207060537253,
      "grad_norm": 5.183691024780273,
      "learning_rate": 2.1669659646545183e-05,
      "loss": 2.038774871826172,
      "memory(GiB)": 77.56,
      "step": 80730,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.437419
    },
    {
      "epoch": 3.458934921382974,
      "grad_norm": 6.527834415435791,
      "learning_rate": 2.166411464889404e-05,
      "loss": 2.4816041946411134,
      "memory(GiB)": 77.56,
      "step": 80735,
      "token_acc": 0.4753521126760563,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 3.459149136712223,
      "grad_norm": 5.282503604888916,
      "learning_rate": 2.1658570164574044e-05,
      "loss": 2.385080337524414,
      "memory(GiB)": 77.56,
      "step": 80740,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.437423
    },
    {
      "epoch": 3.459363352041472,
      "grad_norm": 5.43533182144165,
      "learning_rate": 2.1653026193685632e-05,
      "loss": 2.1967086791992188,
      "memory(GiB)": 77.56,
      "step": 80745,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.437436
    },
    {
      "epoch": 3.459577567370721,
      "grad_norm": 5.7788286209106445,
      "learning_rate": 2.1647482736329227e-05,
      "loss": 2.2340927124023438,
      "memory(GiB)": 77.56,
      "step": 80750,
      "token_acc": 0.5378486055776892,
      "train_speed(iter/s)": 1.437453
    },
    {
      "epoch": 3.45979178269997,
      "grad_norm": 8.328045845031738,
      "learning_rate": 2.1641939792605287e-05,
      "loss": 2.5865297317504883,
      "memory(GiB)": 77.56,
      "step": 80755,
      "token_acc": 0.4343065693430657,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.460005998029219,
      "grad_norm": 6.668324947357178,
      "learning_rate": 2.163639736261421e-05,
      "loss": 2.241141128540039,
      "memory(GiB)": 77.56,
      "step": 80760,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.437423
    },
    {
      "epoch": 3.460220213358468,
      "grad_norm": 6.232999324798584,
      "learning_rate": 2.1630855446456382e-05,
      "loss": 2.140214538574219,
      "memory(GiB)": 77.56,
      "step": 80765,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 3.4604344286877167,
      "grad_norm": 8.097764015197754,
      "learning_rate": 2.1625314044232236e-05,
      "loss": 1.9963481903076172,
      "memory(GiB)": 77.56,
      "step": 80770,
      "token_acc": 0.5443548387096774,
      "train_speed(iter/s)": 1.437433
    },
    {
      "epoch": 3.460648644016966,
      "grad_norm": 4.857347011566162,
      "learning_rate": 2.161977315604215e-05,
      "loss": 2.3844308853149414,
      "memory(GiB)": 77.56,
      "step": 80775,
      "token_acc": 0.4944649446494465,
      "train_speed(iter/s)": 1.437443
    },
    {
      "epoch": 3.4608628593462147,
      "grad_norm": 6.190577983856201,
      "learning_rate": 2.1614232781986493e-05,
      "loss": 2.4740447998046875,
      "memory(GiB)": 77.56,
      "step": 80780,
      "token_acc": 0.47019867549668876,
      "train_speed(iter/s)": 1.437441
    },
    {
      "epoch": 3.4610770746754635,
      "grad_norm": 4.87861967086792,
      "learning_rate": 2.1608692922165634e-05,
      "loss": 2.3485822677612305,
      "memory(GiB)": 77.56,
      "step": 80785,
      "token_acc": 0.5110410094637224,
      "train_speed(iter/s)": 1.437419
    },
    {
      "epoch": 3.461291290004713,
      "grad_norm": 8.18832015991211,
      "learning_rate": 2.1603153576679917e-05,
      "loss": 2.623311996459961,
      "memory(GiB)": 77.56,
      "step": 80790,
      "token_acc": 0.48299319727891155,
      "train_speed(iter/s)": 1.43742
    },
    {
      "epoch": 3.4615055053339616,
      "grad_norm": 6.9606032371521,
      "learning_rate": 2.1597614745629734e-05,
      "loss": 2.4861278533935547,
      "memory(GiB)": 77.56,
      "step": 80795,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437429
    },
    {
      "epoch": 3.4617197206632104,
      "grad_norm": 5.269224166870117,
      "learning_rate": 2.1592076429115406e-05,
      "loss": 2.347925567626953,
      "memory(GiB)": 77.56,
      "step": 80800,
      "token_acc": 0.5027777777777778,
      "train_speed(iter/s)": 1.43745
    },
    {
      "epoch": 3.4619339359924597,
      "grad_norm": 4.54539680480957,
      "learning_rate": 2.1586538627237257e-05,
      "loss": 2.302774429321289,
      "memory(GiB)": 77.56,
      "step": 80805,
      "token_acc": 0.4773413897280967,
      "train_speed(iter/s)": 1.437437
    },
    {
      "epoch": 3.4621481513217085,
      "grad_norm": 5.531946182250977,
      "learning_rate": 2.158100134009562e-05,
      "loss": 2.526936340332031,
      "memory(GiB)": 77.56,
      "step": 80810,
      "token_acc": 0.4824561403508772,
      "train_speed(iter/s)": 1.437434
    },
    {
      "epoch": 3.4623623666509573,
      "grad_norm": 7.029908180236816,
      "learning_rate": 2.1575464567790778e-05,
      "loss": 2.8980453491210936,
      "memory(GiB)": 77.56,
      "step": 80815,
      "token_acc": 0.4562043795620438,
      "train_speed(iter/s)": 1.43746
    },
    {
      "epoch": 3.4625765819802066,
      "grad_norm": 6.715392589569092,
      "learning_rate": 2.1569928310423082e-05,
      "loss": 2.575844383239746,
      "memory(GiB)": 77.56,
      "step": 80820,
      "token_acc": 0.47416413373860183,
      "train_speed(iter/s)": 1.437477
    },
    {
      "epoch": 3.4627907973094554,
      "grad_norm": 5.619195461273193,
      "learning_rate": 2.1564392568092785e-05,
      "loss": 2.3543155670166014,
      "memory(GiB)": 77.56,
      "step": 80825,
      "token_acc": 0.5031446540880503,
      "train_speed(iter/s)": 1.437466
    },
    {
      "epoch": 3.463005012638704,
      "grad_norm": 6.5175628662109375,
      "learning_rate": 2.1558857340900207e-05,
      "loss": 2.494320297241211,
      "memory(GiB)": 77.56,
      "step": 80830,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.4632192279679535,
      "grad_norm": 5.518880844116211,
      "learning_rate": 2.1553322628945615e-05,
      "loss": 2.514445686340332,
      "memory(GiB)": 77.56,
      "step": 80835,
      "token_acc": 0.4902597402597403,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.4634334432972023,
      "grad_norm": 5.718807220458984,
      "learning_rate": 2.1547788432329263e-05,
      "loss": 2.5359895706176756,
      "memory(GiB)": 77.56,
      "step": 80840,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.437494
    },
    {
      "epoch": 3.463647658626451,
      "grad_norm": 5.379241943359375,
      "learning_rate": 2.1542254751151412e-05,
      "loss": 2.7938880920410156,
      "memory(GiB)": 77.56,
      "step": 80845,
      "token_acc": 0.4681528662420382,
      "train_speed(iter/s)": 1.437499
    },
    {
      "epoch": 3.4638618739557003,
      "grad_norm": 5.155433177947998,
      "learning_rate": 2.1536721585512298e-05,
      "loss": 2.273894119262695,
      "memory(GiB)": 77.56,
      "step": 80850,
      "token_acc": 0.5241157556270096,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 3.464076089284949,
      "grad_norm": 6.263373374938965,
      "learning_rate": 2.153118893551219e-05,
      "loss": 2.0479362487792967,
      "memory(GiB)": 77.56,
      "step": 80855,
      "token_acc": 0.583916083916084,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 3.464290304614198,
      "grad_norm": 4.4990363121032715,
      "learning_rate": 2.1525656801251304e-05,
      "loss": 2.322727584838867,
      "memory(GiB)": 77.56,
      "step": 80860,
      "token_acc": 0.49034749034749037,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 3.464504519943447,
      "grad_norm": 5.9686598777771,
      "learning_rate": 2.1520125182829852e-05,
      "loss": 2.562971496582031,
      "memory(GiB)": 77.56,
      "step": 80865,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.437499
    },
    {
      "epoch": 3.464718735272696,
      "grad_norm": 5.9565205574035645,
      "learning_rate": 2.151459408034805e-05,
      "loss": 2.3215253829956053,
      "memory(GiB)": 77.56,
      "step": 80870,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.437475
    },
    {
      "epoch": 3.464932950601945,
      "grad_norm": 5.3802947998046875,
      "learning_rate": 2.150906349390608e-05,
      "loss": 2.3780303955078126,
      "memory(GiB)": 77.56,
      "step": 80875,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.437467
    },
    {
      "epoch": 3.465147165931194,
      "grad_norm": 5.702744960784912,
      "learning_rate": 2.1503533423604167e-05,
      "loss": 2.1353302001953125,
      "memory(GiB)": 77.56,
      "step": 80880,
      "token_acc": 0.5698529411764706,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 3.465361381260443,
      "grad_norm": 6.480691909790039,
      "learning_rate": 2.1498003869542483e-05,
      "loss": 2.376254844665527,
      "memory(GiB)": 77.56,
      "step": 80885,
      "token_acc": 0.46794871794871795,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 3.4655755965896917,
      "grad_norm": 7.769824028015137,
      "learning_rate": 2.149247483182119e-05,
      "loss": 2.088960647583008,
      "memory(GiB)": 77.56,
      "step": 80890,
      "token_acc": 0.5125,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 3.465789811918941,
      "grad_norm": 5.404630184173584,
      "learning_rate": 2.148694631054046e-05,
      "loss": 2.281485366821289,
      "memory(GiB)": 77.56,
      "step": 80895,
      "token_acc": 0.5180327868852459,
      "train_speed(iter/s)": 1.43751
    },
    {
      "epoch": 3.46600402724819,
      "grad_norm": 5.724725723266602,
      "learning_rate": 2.1481418305800425e-05,
      "loss": 2.1865924835205077,
      "memory(GiB)": 77.56,
      "step": 80900,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.437511
    },
    {
      "epoch": 3.4662182425774386,
      "grad_norm": 6.743160724639893,
      "learning_rate": 2.1475890817701272e-05,
      "loss": 2.6283044815063477,
      "memory(GiB)": 77.56,
      "step": 80905,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.437541
    },
    {
      "epoch": 3.466432457906688,
      "grad_norm": 7.072866916656494,
      "learning_rate": 2.1470363846343096e-05,
      "loss": 2.362382507324219,
      "memory(GiB)": 77.56,
      "step": 80910,
      "token_acc": 0.496875,
      "train_speed(iter/s)": 1.437551
    },
    {
      "epoch": 3.4666466732359367,
      "grad_norm": 6.42167854309082,
      "learning_rate": 2.146483739182606e-05,
      "loss": 2.4504037857055665,
      "memory(GiB)": 77.56,
      "step": 80915,
      "token_acc": 0.47733333333333333,
      "train_speed(iter/s)": 1.437545
    },
    {
      "epoch": 3.4668608885651855,
      "grad_norm": 6.257534980773926,
      "learning_rate": 2.1459311454250258e-05,
      "loss": 2.543269157409668,
      "memory(GiB)": 77.56,
      "step": 80920,
      "token_acc": 0.44868035190615835,
      "train_speed(iter/s)": 1.437543
    },
    {
      "epoch": 3.4670751038944347,
      "grad_norm": 5.240074634552002,
      "learning_rate": 2.1453786033715807e-05,
      "loss": 2.296847343444824,
      "memory(GiB)": 77.56,
      "step": 80925,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437565
    },
    {
      "epoch": 3.4672893192236836,
      "grad_norm": 6.4981770515441895,
      "learning_rate": 2.14482611303228e-05,
      "loss": 2.4388652801513673,
      "memory(GiB)": 77.56,
      "step": 80930,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 3.4675035345529324,
      "grad_norm": 6.463721752166748,
      "learning_rate": 2.1442736744171305e-05,
      "loss": 2.4519615173339844,
      "memory(GiB)": 77.56,
      "step": 80935,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.437559
    },
    {
      "epoch": 3.4677177498821816,
      "grad_norm": 5.944515705108643,
      "learning_rate": 2.1437212875361444e-05,
      "loss": 2.3641876220703124,
      "memory(GiB)": 77.56,
      "step": 80940,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.437536
    },
    {
      "epoch": 3.4679319652114304,
      "grad_norm": 5.1539740562438965,
      "learning_rate": 2.1431689523993263e-05,
      "loss": 2.2079532623291014,
      "memory(GiB)": 77.56,
      "step": 80945,
      "token_acc": 0.5574912891986062,
      "train_speed(iter/s)": 1.437527
    },
    {
      "epoch": 3.4681461805406792,
      "grad_norm": 6.763242721557617,
      "learning_rate": 2.142616669016683e-05,
      "loss": 2.4610616683959963,
      "memory(GiB)": 77.56,
      "step": 80950,
      "token_acc": 0.4281437125748503,
      "train_speed(iter/s)": 1.437536
    },
    {
      "epoch": 3.4683603958699285,
      "grad_norm": 8.104201316833496,
      "learning_rate": 2.142064437398219e-05,
      "loss": 2.17742919921875,
      "memory(GiB)": 77.56,
      "step": 80955,
      "token_acc": 0.5119453924914675,
      "train_speed(iter/s)": 1.437524
    },
    {
      "epoch": 3.4685746111991773,
      "grad_norm": 5.250313758850098,
      "learning_rate": 2.141512257553937e-05,
      "loss": 2.2010494232177735,
      "memory(GiB)": 77.56,
      "step": 80960,
      "token_acc": 0.503125,
      "train_speed(iter/s)": 1.437526
    },
    {
      "epoch": 3.468788826528426,
      "grad_norm": 6.576564311981201,
      "learning_rate": 2.1409601294938436e-05,
      "loss": 2.3733232498168944,
      "memory(GiB)": 77.56,
      "step": 80965,
      "token_acc": 0.4968152866242038,
      "train_speed(iter/s)": 1.437507
    },
    {
      "epoch": 3.4690030418576754,
      "grad_norm": 6.809447765350342,
      "learning_rate": 2.1404080532279376e-05,
      "loss": 2.2804502487182616,
      "memory(GiB)": 77.56,
      "step": 80970,
      "token_acc": 0.5016501650165016,
      "train_speed(iter/s)": 1.437497
    },
    {
      "epoch": 3.469217257186924,
      "grad_norm": 6.49641752243042,
      "learning_rate": 2.139856028766224e-05,
      "loss": 2.490182113647461,
      "memory(GiB)": 77.56,
      "step": 80975,
      "token_acc": 0.5101214574898786,
      "train_speed(iter/s)": 1.437516
    },
    {
      "epoch": 3.469431472516173,
      "grad_norm": 7.083703994750977,
      "learning_rate": 2.139304056118701e-05,
      "loss": 2.632050323486328,
      "memory(GiB)": 77.56,
      "step": 80980,
      "token_acc": 0.4674922600619195,
      "train_speed(iter/s)": 1.437512
    },
    {
      "epoch": 3.4696456878454223,
      "grad_norm": 4.87672233581543,
      "learning_rate": 2.1387521352953695e-05,
      "loss": 2.320743751525879,
      "memory(GiB)": 77.56,
      "step": 80985,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 3.469859903174671,
      "grad_norm": 7.414366245269775,
      "learning_rate": 2.138200266306227e-05,
      "loss": 2.4225080490112303,
      "memory(GiB)": 77.56,
      "step": 80990,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 3.47007411850392,
      "grad_norm": 9.614435195922852,
      "learning_rate": 2.1376484491612685e-05,
      "loss": 2.3034950256347657,
      "memory(GiB)": 77.56,
      "step": 80995,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 3.470288333833169,
      "grad_norm": 5.982474327087402,
      "learning_rate": 2.1370966838704963e-05,
      "loss": 2.5725757598876955,
      "memory(GiB)": 77.56,
      "step": 81000,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.437498
    },
    {
      "epoch": 3.470288333833169,
      "eval_loss": 2.343449115753174,
      "eval_runtime": 14.2529,
      "eval_samples_per_second": 7.016,
      "eval_steps_per_second": 7.016,
      "eval_token_acc": 0.4537914691943128,
      "step": 81000
    },
    {
      "epoch": 3.470502549162418,
      "grad_norm": 6.501448154449463,
      "learning_rate": 2.1365449704439028e-05,
      "loss": 2.7911149978637697,
      "memory(GiB)": 77.56,
      "step": 81005,
      "token_acc": 0.44601769911504424,
      "train_speed(iter/s)": 1.437104
    },
    {
      "epoch": 3.470716764491667,
      "grad_norm": 6.5665483474731445,
      "learning_rate": 2.1359933088914836e-05,
      "loss": 2.267047119140625,
      "memory(GiB)": 77.56,
      "step": 81010,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.437125
    },
    {
      "epoch": 3.470930979820916,
      "grad_norm": 5.170248031616211,
      "learning_rate": 2.1354416992232318e-05,
      "loss": 2.2988662719726562,
      "memory(GiB)": 77.56,
      "step": 81015,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.437142
    },
    {
      "epoch": 3.471145195150165,
      "grad_norm": 4.95196533203125,
      "learning_rate": 2.1348901414491396e-05,
      "loss": 2.2833063125610353,
      "memory(GiB)": 77.56,
      "step": 81020,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.437166
    },
    {
      "epoch": 3.471359410479414,
      "grad_norm": 5.3953070640563965,
      "learning_rate": 2.1343386355792017e-05,
      "loss": 2.1579689025878905,
      "memory(GiB)": 77.56,
      "step": 81025,
      "token_acc": 0.5494880546075085,
      "train_speed(iter/s)": 1.437163
    },
    {
      "epoch": 3.471573625808663,
      "grad_norm": 4.396566390991211,
      "learning_rate": 2.1337871816234072e-05,
      "loss": 2.290094757080078,
      "memory(GiB)": 77.56,
      "step": 81030,
      "token_acc": 0.49337748344370863,
      "train_speed(iter/s)": 1.437161
    },
    {
      "epoch": 3.4717878411379117,
      "grad_norm": 6.068824768066406,
      "learning_rate": 2.1332357795917475e-05,
      "loss": 2.367625427246094,
      "memory(GiB)": 77.56,
      "step": 81035,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.437173
    },
    {
      "epoch": 3.472002056467161,
      "grad_norm": 5.833049774169922,
      "learning_rate": 2.13268442949421e-05,
      "loss": 2.22940616607666,
      "memory(GiB)": 77.56,
      "step": 81040,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.437193
    },
    {
      "epoch": 3.47221627179641,
      "grad_norm": 8.971484184265137,
      "learning_rate": 2.132133131340782e-05,
      "loss": 2.377848243713379,
      "memory(GiB)": 77.56,
      "step": 81045,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.437207
    },
    {
      "epoch": 3.4724304871256586,
      "grad_norm": 5.634929180145264,
      "learning_rate": 2.1315818851414553e-05,
      "loss": 2.1934200286865235,
      "memory(GiB)": 77.56,
      "step": 81050,
      "token_acc": 0.5266457680250783,
      "train_speed(iter/s)": 1.437196
    },
    {
      "epoch": 3.472644702454908,
      "grad_norm": 5.89577579498291,
      "learning_rate": 2.131030690906211e-05,
      "loss": 2.758171272277832,
      "memory(GiB)": 77.56,
      "step": 81055,
      "token_acc": 0.4456140350877193,
      "train_speed(iter/s)": 1.437208
    },
    {
      "epoch": 3.4728589177841567,
      "grad_norm": 6.55145788192749,
      "learning_rate": 2.1304795486450398e-05,
      "loss": 2.172046661376953,
      "memory(GiB)": 77.56,
      "step": 81060,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.437207
    },
    {
      "epoch": 3.4730731331134055,
      "grad_norm": 9.11303424835205,
      "learning_rate": 2.1299284583679224e-05,
      "loss": 2.420275115966797,
      "memory(GiB)": 77.56,
      "step": 81065,
      "token_acc": 0.47468354430379744,
      "train_speed(iter/s)": 1.43723
    },
    {
      "epoch": 3.4732873484426547,
      "grad_norm": 5.454561710357666,
      "learning_rate": 2.1293774200848437e-05,
      "loss": 2.279291343688965,
      "memory(GiB)": 77.56,
      "step": 81070,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.437238
    },
    {
      "epoch": 3.4735015637719036,
      "grad_norm": 5.810431480407715,
      "learning_rate": 2.1288264338057855e-05,
      "loss": 2.373587799072266,
      "memory(GiB)": 77.56,
      "step": 81075,
      "token_acc": 0.48184818481848185,
      "train_speed(iter/s)": 1.437246
    },
    {
      "epoch": 3.4737157791011524,
      "grad_norm": 5.706579685211182,
      "learning_rate": 2.1282754995407285e-05,
      "loss": 2.238748550415039,
      "memory(GiB)": 77.56,
      "step": 81080,
      "token_acc": 0.5390334572490706,
      "train_speed(iter/s)": 1.437261
    },
    {
      "epoch": 3.4739299944304016,
      "grad_norm": 5.767402172088623,
      "learning_rate": 2.1277246172996562e-05,
      "loss": 2.377247619628906,
      "memory(GiB)": 77.56,
      "step": 81085,
      "token_acc": 0.5013774104683195,
      "train_speed(iter/s)": 1.437258
    },
    {
      "epoch": 3.4741442097596504,
      "grad_norm": 7.721344470977783,
      "learning_rate": 2.1271737870925473e-05,
      "loss": 2.5996915817260744,
      "memory(GiB)": 77.56,
      "step": 81090,
      "token_acc": 0.47079037800687284,
      "train_speed(iter/s)": 1.437277
    },
    {
      "epoch": 3.4743584250888992,
      "grad_norm": 7.026059627532959,
      "learning_rate": 2.12662300892938e-05,
      "loss": 2.4789125442504885,
      "memory(GiB)": 77.56,
      "step": 81095,
      "token_acc": 0.449438202247191,
      "train_speed(iter/s)": 1.437296
    },
    {
      "epoch": 3.4745726404181485,
      "grad_norm": 5.499500751495361,
      "learning_rate": 2.1260722828201323e-05,
      "loss": 2.306113433837891,
      "memory(GiB)": 77.56,
      "step": 81100,
      "token_acc": 0.5252918287937743,
      "train_speed(iter/s)": 1.43729
    },
    {
      "epoch": 3.4747868557473973,
      "grad_norm": 6.0566020011901855,
      "learning_rate": 2.1255216087747792e-05,
      "loss": 2.3789741516113283,
      "memory(GiB)": 77.56,
      "step": 81105,
      "token_acc": 0.5205047318611987,
      "train_speed(iter/s)": 1.437284
    },
    {
      "epoch": 3.475001071076646,
      "grad_norm": 7.280029773712158,
      "learning_rate": 2.1249709868033007e-05,
      "loss": 2.2331844329833985,
      "memory(GiB)": 77.56,
      "step": 81110,
      "token_acc": 0.5614035087719298,
      "train_speed(iter/s)": 1.437308
    },
    {
      "epoch": 3.4752152864058954,
      "grad_norm": 6.601282119750977,
      "learning_rate": 2.1244204169156694e-05,
      "loss": 2.7776655197143554,
      "memory(GiB)": 77.56,
      "step": 81115,
      "token_acc": 0.4581818181818182,
      "train_speed(iter/s)": 1.437305
    },
    {
      "epoch": 3.475429501735144,
      "grad_norm": 5.758901119232178,
      "learning_rate": 2.123869899121858e-05,
      "loss": 2.33856143951416,
      "memory(GiB)": 77.56,
      "step": 81120,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.437315
    },
    {
      "epoch": 3.475643717064393,
      "grad_norm": 6.502024173736572,
      "learning_rate": 2.1233194334318435e-05,
      "loss": 2.5741403579711912,
      "memory(GiB)": 77.56,
      "step": 81125,
      "token_acc": 0.4867924528301887,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 3.4758579323936423,
      "grad_norm": 6.306130409240723,
      "learning_rate": 2.122769019855596e-05,
      "loss": 2.2737781524658205,
      "memory(GiB)": 77.56,
      "step": 81130,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.437309
    },
    {
      "epoch": 3.476072147722891,
      "grad_norm": 5.00747013092041,
      "learning_rate": 2.1222186584030867e-05,
      "loss": 2.4421785354614256,
      "memory(GiB)": 77.56,
      "step": 81135,
      "token_acc": 0.48562300319488816,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 3.47628636305214,
      "grad_norm": 7.00429630279541,
      "learning_rate": 2.1216683490842843e-05,
      "loss": 2.6984304428100585,
      "memory(GiB)": 77.56,
      "step": 81140,
      "token_acc": 0.4779874213836478,
      "train_speed(iter/s)": 1.437316
    },
    {
      "epoch": 3.476500578381389,
      "grad_norm": 6.758488178253174,
      "learning_rate": 2.1211180919091615e-05,
      "loss": 2.3303596496582033,
      "memory(GiB)": 77.56,
      "step": 81145,
      "token_acc": 0.5162241887905604,
      "train_speed(iter/s)": 1.437335
    },
    {
      "epoch": 3.476714793710638,
      "grad_norm": 6.930009365081787,
      "learning_rate": 2.120567886887685e-05,
      "loss": 2.501367378234863,
      "memory(GiB)": 77.56,
      "step": 81150,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437334
    },
    {
      "epoch": 3.4769290090398868,
      "grad_norm": 5.041632175445557,
      "learning_rate": 2.1200177340298222e-05,
      "loss": 2.677979278564453,
      "memory(GiB)": 77.56,
      "step": 81155,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.437336
    },
    {
      "epoch": 3.477143224369136,
      "grad_norm": 5.47381591796875,
      "learning_rate": 2.1194676333455404e-05,
      "loss": 2.7373023986816407,
      "memory(GiB)": 77.56,
      "step": 81160,
      "token_acc": 0.4352517985611511,
      "train_speed(iter/s)": 1.437328
    },
    {
      "epoch": 3.477357439698385,
      "grad_norm": 4.7734856605529785,
      "learning_rate": 2.1189175848448023e-05,
      "loss": 2.2085458755493166,
      "memory(GiB)": 77.56,
      "step": 81165,
      "token_acc": 0.5132450331125827,
      "train_speed(iter/s)": 1.437323
    },
    {
      "epoch": 3.4775716550276337,
      "grad_norm": 6.531951904296875,
      "learning_rate": 2.1183675885375764e-05,
      "loss": 2.4423059463500976,
      "memory(GiB)": 77.56,
      "step": 81170,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.437333
    },
    {
      "epoch": 3.477785870356883,
      "grad_norm": 5.582119941711426,
      "learning_rate": 2.1178176444338255e-05,
      "loss": 2.783696174621582,
      "memory(GiB)": 77.56,
      "step": 81175,
      "token_acc": 0.4440894568690096,
      "train_speed(iter/s)": 1.437331
    },
    {
      "epoch": 3.4780000856861317,
      "grad_norm": 4.885003566741943,
      "learning_rate": 2.1172677525435115e-05,
      "loss": 2.2024856567382813,
      "memory(GiB)": 77.56,
      "step": 81180,
      "token_acc": 0.5159235668789809,
      "train_speed(iter/s)": 1.437332
    },
    {
      "epoch": 3.4782143010153805,
      "grad_norm": 8.581168174743652,
      "learning_rate": 2.1167179128765958e-05,
      "loss": 2.5739604949951174,
      "memory(GiB)": 77.56,
      "step": 81185,
      "token_acc": 0.45387453874538747,
      "train_speed(iter/s)": 1.43737
    },
    {
      "epoch": 3.47842851634463,
      "grad_norm": 6.265730381011963,
      "learning_rate": 2.1161681254430387e-05,
      "loss": 2.4695104598999023,
      "memory(GiB)": 77.56,
      "step": 81190,
      "token_acc": 0.4773413897280967,
      "train_speed(iter/s)": 1.437368
    },
    {
      "epoch": 3.4786427316738786,
      "grad_norm": 7.1325764656066895,
      "learning_rate": 2.1156183902528024e-05,
      "loss": 2.5873565673828125,
      "memory(GiB)": 77.56,
      "step": 81195,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.437386
    },
    {
      "epoch": 3.4788569470031274,
      "grad_norm": 6.447802543640137,
      "learning_rate": 2.1150687073158437e-05,
      "loss": 2.517266845703125,
      "memory(GiB)": 77.56,
      "step": 81200,
      "token_acc": 0.4697406340057637,
      "train_speed(iter/s)": 1.437418
    },
    {
      "epoch": 3.4790711623323767,
      "grad_norm": 4.610739707946777,
      "learning_rate": 2.114519076642123e-05,
      "loss": 2.425057601928711,
      "memory(GiB)": 77.56,
      "step": 81205,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.437412
    },
    {
      "epoch": 3.4792853776616255,
      "grad_norm": 4.679841995239258,
      "learning_rate": 2.1139694982415965e-05,
      "loss": 2.3926328659057616,
      "memory(GiB)": 77.56,
      "step": 81210,
      "token_acc": 0.5198675496688742,
      "train_speed(iter/s)": 1.437422
    },
    {
      "epoch": 3.4794995929908743,
      "grad_norm": 8.280360221862793,
      "learning_rate": 2.1134199721242194e-05,
      "loss": 2.250014877319336,
      "memory(GiB)": 77.56,
      "step": 81215,
      "token_acc": 0.5265017667844523,
      "train_speed(iter/s)": 1.437429
    },
    {
      "epoch": 3.4797138083201236,
      "grad_norm": 4.4545793533325195,
      "learning_rate": 2.112870498299947e-05,
      "loss": 2.3474857330322267,
      "memory(GiB)": 77.56,
      "step": 81220,
      "token_acc": 0.524390243902439,
      "train_speed(iter/s)": 1.437433
    },
    {
      "epoch": 3.4799280236493724,
      "grad_norm": 6.209092140197754,
      "learning_rate": 2.112321076778733e-05,
      "loss": 2.6005908966064455,
      "memory(GiB)": 77.56,
      "step": 81225,
      "token_acc": 0.4625,
      "train_speed(iter/s)": 1.437455
    },
    {
      "epoch": 3.480142238978621,
      "grad_norm": 7.500361919403076,
      "learning_rate": 2.1117717075705324e-05,
      "loss": 2.185984420776367,
      "memory(GiB)": 77.56,
      "step": 81230,
      "token_acc": 0.5203252032520326,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.4803564543078704,
      "grad_norm": 7.425896644592285,
      "learning_rate": 2.1112223906852973e-05,
      "loss": 2.440434455871582,
      "memory(GiB)": 77.56,
      "step": 81235,
      "token_acc": 0.4898785425101215,
      "train_speed(iter/s)": 1.437485
    },
    {
      "epoch": 3.4805706696371193,
      "grad_norm": 8.427542686462402,
      "learning_rate": 2.1106731261329783e-05,
      "loss": 2.3623069763183593,
      "memory(GiB)": 77.56,
      "step": 81240,
      "token_acc": 0.5017064846416383,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 3.480784884966368,
      "grad_norm": 8.56961441040039,
      "learning_rate": 2.1101239139235263e-05,
      "loss": 2.473025894165039,
      "memory(GiB)": 77.56,
      "step": 81245,
      "token_acc": 0.4435146443514644,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 3.4809991002956173,
      "grad_norm": 5.188290119171143,
      "learning_rate": 2.109574754066888e-05,
      "loss": 2.5860151290893554,
      "memory(GiB)": 77.56,
      "step": 81250,
      "token_acc": 0.5018181818181818,
      "train_speed(iter/s)": 1.437496
    },
    {
      "epoch": 3.481213315624866,
      "grad_norm": 7.030704975128174,
      "learning_rate": 2.109025646573017e-05,
      "loss": 2.461176109313965,
      "memory(GiB)": 77.56,
      "step": 81255,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 3.481427530954115,
      "grad_norm": 5.170688152313232,
      "learning_rate": 2.108476591451858e-05,
      "loss": 2.539849853515625,
      "memory(GiB)": 77.56,
      "step": 81260,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 3.481641746283364,
      "grad_norm": 5.83721923828125,
      "learning_rate": 2.1079275887133565e-05,
      "loss": 2.1372711181640627,
      "memory(GiB)": 77.56,
      "step": 81265,
      "token_acc": 0.541095890410959,
      "train_speed(iter/s)": 1.437495
    },
    {
      "epoch": 3.481855961612613,
      "grad_norm": 5.7097272872924805,
      "learning_rate": 2.107378638367461e-05,
      "loss": 2.3553020477294924,
      "memory(GiB)": 77.56,
      "step": 81270,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.437485
    },
    {
      "epoch": 3.482070176941862,
      "grad_norm": 5.059092998504639,
      "learning_rate": 2.1068297404241156e-05,
      "loss": 2.374205780029297,
      "memory(GiB)": 77.56,
      "step": 81275,
      "token_acc": 0.49117647058823527,
      "train_speed(iter/s)": 1.437476
    },
    {
      "epoch": 3.482284392271111,
      "grad_norm": 6.398963451385498,
      "learning_rate": 2.1062808948932633e-05,
      "loss": 2.45794792175293,
      "memory(GiB)": 77.56,
      "step": 81280,
      "token_acc": 0.434375,
      "train_speed(iter/s)": 1.437485
    },
    {
      "epoch": 3.48249860760036,
      "grad_norm": 6.465716361999512,
      "learning_rate": 2.105732101784845e-05,
      "loss": 2.0750759124755858,
      "memory(GiB)": 77.56,
      "step": 81285,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.437466
    },
    {
      "epoch": 3.4827128229296087,
      "grad_norm": 4.81996488571167,
      "learning_rate": 2.105183361108807e-05,
      "loss": 2.3773626327514648,
      "memory(GiB)": 77.56,
      "step": 81290,
      "token_acc": 0.5230263157894737,
      "train_speed(iter/s)": 1.437455
    },
    {
      "epoch": 3.482927038258858,
      "grad_norm": 4.905280590057373,
      "learning_rate": 2.1046346728750872e-05,
      "loss": 2.4053321838378907,
      "memory(GiB)": 77.56,
      "step": 81295,
      "token_acc": 0.4738675958188153,
      "train_speed(iter/s)": 1.437473
    },
    {
      "epoch": 3.483141253588107,
      "grad_norm": 5.443066120147705,
      "learning_rate": 2.104086037093627e-05,
      "loss": 2.1631118774414064,
      "memory(GiB)": 77.56,
      "step": 81300,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437486
    },
    {
      "epoch": 3.4833554689173556,
      "grad_norm": 9.772308349609375,
      "learning_rate": 2.1035374537743645e-05,
      "loss": 2.531840515136719,
      "memory(GiB)": 77.56,
      "step": 81305,
      "token_acc": 0.47717842323651455,
      "train_speed(iter/s)": 1.437495
    },
    {
      "epoch": 3.483569684246605,
      "grad_norm": 6.452080249786377,
      "learning_rate": 2.1029889229272364e-05,
      "loss": 2.456101417541504,
      "memory(GiB)": 77.56,
      "step": 81310,
      "token_acc": 0.49823321554770317,
      "train_speed(iter/s)": 1.437505
    },
    {
      "epoch": 3.4837838995758537,
      "grad_norm": 6.846014499664307,
      "learning_rate": 2.1024404445621837e-05,
      "loss": 2.318460464477539,
      "memory(GiB)": 77.56,
      "step": 81315,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.437514
    },
    {
      "epoch": 3.4839981149051025,
      "grad_norm": 5.934245586395264,
      "learning_rate": 2.1018920186891407e-05,
      "loss": 2.5138072967529297,
      "memory(GiB)": 77.56,
      "step": 81320,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.437522
    },
    {
      "epoch": 3.4842123302343517,
      "grad_norm": 4.512868881225586,
      "learning_rate": 2.1013436453180417e-05,
      "loss": 1.8392452239990233,
      "memory(GiB)": 77.56,
      "step": 81325,
      "token_acc": 0.5708955223880597,
      "train_speed(iter/s)": 1.43753
    },
    {
      "epoch": 3.4844265455636005,
      "grad_norm": 7.431070804595947,
      "learning_rate": 2.1007953244588218e-05,
      "loss": 2.495029830932617,
      "memory(GiB)": 77.56,
      "step": 81330,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.437523
    },
    {
      "epoch": 3.4846407608928494,
      "grad_norm": 5.911373615264893,
      "learning_rate": 2.1002470561214126e-05,
      "loss": 2.294062042236328,
      "memory(GiB)": 77.56,
      "step": 81335,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.437527
    },
    {
      "epoch": 3.4848549762220986,
      "grad_norm": 6.209281921386719,
      "learning_rate": 2.09969884031575e-05,
      "loss": 2.4408111572265625,
      "memory(GiB)": 77.56,
      "step": 81340,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.437528
    },
    {
      "epoch": 3.4850691915513474,
      "grad_norm": 4.966249465942383,
      "learning_rate": 2.0991506770517622e-05,
      "loss": 2.3700326919555663,
      "memory(GiB)": 77.56,
      "step": 81345,
      "token_acc": 0.5015197568389058,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 3.4852834068805962,
      "grad_norm": 5.710653781890869,
      "learning_rate": 2.0986025663393823e-05,
      "loss": 2.409952926635742,
      "memory(GiB)": 77.56,
      "step": 81350,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.437543
    },
    {
      "epoch": 3.4854976222098455,
      "grad_norm": 7.252102851867676,
      "learning_rate": 2.098054508188539e-05,
      "loss": 2.2739690780639648,
      "memory(GiB)": 77.56,
      "step": 81355,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.437546
    },
    {
      "epoch": 3.4857118375390943,
      "grad_norm": 4.753846168518066,
      "learning_rate": 2.0975065026091606e-05,
      "loss": 2.398512268066406,
      "memory(GiB)": 77.56,
      "step": 81360,
      "token_acc": 0.4548611111111111,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 3.485926052868343,
      "grad_norm": 6.591948986053467,
      "learning_rate": 2.0969585496111743e-05,
      "loss": 2.8616899490356444,
      "memory(GiB)": 77.56,
      "step": 81365,
      "token_acc": 0.4485049833887043,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 3.4861402681975924,
      "grad_norm": 7.861251354217529,
      "learning_rate": 2.0964106492045055e-05,
      "loss": 2.430014419555664,
      "memory(GiB)": 77.56,
      "step": 81370,
      "token_acc": 0.540268456375839,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 3.486354483526841,
      "grad_norm": 7.088751316070557,
      "learning_rate": 2.0958628013990832e-05,
      "loss": 2.309187126159668,
      "memory(GiB)": 77.56,
      "step": 81375,
      "token_acc": 0.5220338983050847,
      "train_speed(iter/s)": 1.437599
    },
    {
      "epoch": 3.48656869885609,
      "grad_norm": 4.486125946044922,
      "learning_rate": 2.095315006204831e-05,
      "loss": 2.1281736373901365,
      "memory(GiB)": 77.56,
      "step": 81380,
      "token_acc": 0.5681818181818182,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 3.4867829141853393,
      "grad_norm": 5.6165642738342285,
      "learning_rate": 2.0947672636316723e-05,
      "loss": 2.5054168701171875,
      "memory(GiB)": 77.56,
      "step": 81385,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.437613
    },
    {
      "epoch": 3.486997129514588,
      "grad_norm": 7.254798412322998,
      "learning_rate": 2.0942195736895293e-05,
      "loss": 2.51782112121582,
      "memory(GiB)": 77.56,
      "step": 81390,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.437601
    },
    {
      "epoch": 3.487211344843837,
      "grad_norm": 5.275437831878662,
      "learning_rate": 2.0936719363883234e-05,
      "loss": 2.4329294204711913,
      "memory(GiB)": 77.56,
      "step": 81395,
      "token_acc": 0.4956772334293948,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 3.487425560173086,
      "grad_norm": 6.750207901000977,
      "learning_rate": 2.0931243517379784e-05,
      "loss": 2.5842845916748045,
      "memory(GiB)": 77.56,
      "step": 81400,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.437585
    },
    {
      "epoch": 3.487639775502335,
      "grad_norm": 5.144187927246094,
      "learning_rate": 2.0925768197484124e-05,
      "loss": 2.38464241027832,
      "memory(GiB)": 77.56,
      "step": 81405,
      "token_acc": 0.5044510385756676,
      "train_speed(iter/s)": 1.437583
    },
    {
      "epoch": 3.4878539908315838,
      "grad_norm": 4.889918327331543,
      "learning_rate": 2.092029340429545e-05,
      "loss": 1.939674186706543,
      "memory(GiB)": 77.56,
      "step": 81410,
      "token_acc": 0.5879828326180258,
      "train_speed(iter/s)": 1.437589
    },
    {
      "epoch": 3.488068206160833,
      "grad_norm": 7.400070667266846,
      "learning_rate": 2.0914819137912916e-05,
      "loss": 2.2888742446899415,
      "memory(GiB)": 77.56,
      "step": 81415,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 3.488282421490082,
      "grad_norm": 6.307043552398682,
      "learning_rate": 2.0909345398435743e-05,
      "loss": 2.389171600341797,
      "memory(GiB)": 77.56,
      "step": 81420,
      "token_acc": 0.46296296296296297,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 3.4884966368193306,
      "grad_norm": 6.655287265777588,
      "learning_rate": 2.0903872185963063e-05,
      "loss": 2.1464033126831055,
      "memory(GiB)": 77.56,
      "step": 81425,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 3.48871085214858,
      "grad_norm": 4.3760151863098145,
      "learning_rate": 2.0898399500594012e-05,
      "loss": 2.5138574600219727,
      "memory(GiB)": 77.56,
      "step": 81430,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.437588
    },
    {
      "epoch": 3.4889250674778287,
      "grad_norm": 5.804774761199951,
      "learning_rate": 2.0892927342427775e-05,
      "loss": 2.4694240570068358,
      "memory(GiB)": 77.56,
      "step": 81435,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.437592
    },
    {
      "epoch": 3.4891392828070775,
      "grad_norm": 7.967717170715332,
      "learning_rate": 2.0887455711563468e-05,
      "loss": 2.6059778213500975,
      "memory(GiB)": 77.56,
      "step": 81440,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437583
    },
    {
      "epoch": 3.489353498136327,
      "grad_norm": 5.517899036407471,
      "learning_rate": 2.0881984608100207e-05,
      "loss": 2.2054464340209963,
      "memory(GiB)": 77.56,
      "step": 81445,
      "token_acc": 0.5080906148867314,
      "train_speed(iter/s)": 1.437592
    },
    {
      "epoch": 3.4895677134655756,
      "grad_norm": 6.188095569610596,
      "learning_rate": 2.0876514032137105e-05,
      "loss": 2.292200469970703,
      "memory(GiB)": 77.56,
      "step": 81450,
      "token_acc": 0.4635036496350365,
      "train_speed(iter/s)": 1.437587
    },
    {
      "epoch": 3.4897819287948244,
      "grad_norm": 5.592284202575684,
      "learning_rate": 2.087104398377326e-05,
      "loss": 2.3133792877197266,
      "memory(GiB)": 77.56,
      "step": 81455,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.43762
    },
    {
      "epoch": 3.4899961441240737,
      "grad_norm": 8.104219436645508,
      "learning_rate": 2.0865574463107794e-05,
      "loss": 2.0908971786499024,
      "memory(GiB)": 77.56,
      "step": 81460,
      "token_acc": 0.5381679389312977,
      "train_speed(iter/s)": 1.437629
    },
    {
      "epoch": 3.4902103594533225,
      "grad_norm": 4.688964366912842,
      "learning_rate": 2.0860105470239777e-05,
      "loss": 2.451948356628418,
      "memory(GiB)": 77.56,
      "step": 81465,
      "token_acc": 0.44054054054054054,
      "train_speed(iter/s)": 1.437641
    },
    {
      "epoch": 3.4904245747825713,
      "grad_norm": 3.8236515522003174,
      "learning_rate": 2.0854637005268284e-05,
      "loss": 2.4827356338500977,
      "memory(GiB)": 77.56,
      "step": 81470,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 3.4906387901118205,
      "grad_norm": 6.5485615730285645,
      "learning_rate": 2.0849169068292386e-05,
      "loss": 2.260237121582031,
      "memory(GiB)": 77.56,
      "step": 81475,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.437658
    },
    {
      "epoch": 3.4908530054410694,
      "grad_norm": 5.4037346839904785,
      "learning_rate": 2.0843701659411112e-05,
      "loss": 2.235797882080078,
      "memory(GiB)": 77.56,
      "step": 81480,
      "token_acc": 0.5358490566037736,
      "train_speed(iter/s)": 1.437671
    },
    {
      "epoch": 3.491067220770318,
      "grad_norm": 5.696934700012207,
      "learning_rate": 2.0838234778723555e-05,
      "loss": 2.248728370666504,
      "memory(GiB)": 77.56,
      "step": 81485,
      "token_acc": 0.5472972972972973,
      "train_speed(iter/s)": 1.437675
    },
    {
      "epoch": 3.4912814360995674,
      "grad_norm": 8.324548721313477,
      "learning_rate": 2.0832768426328715e-05,
      "loss": 2.201025390625,
      "memory(GiB)": 77.56,
      "step": 81490,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.437665
    },
    {
      "epoch": 3.4914956514288162,
      "grad_norm": 5.908787250518799,
      "learning_rate": 2.082730260232565e-05,
      "loss": 2.462218475341797,
      "memory(GiB)": 77.56,
      "step": 81495,
      "token_acc": 0.4958217270194986,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 3.491709866758065,
      "grad_norm": 5.736169815063477,
      "learning_rate": 2.0821837306813364e-05,
      "loss": 2.101084518432617,
      "memory(GiB)": 77.56,
      "step": 81500,
      "token_acc": 0.5313807531380753,
      "train_speed(iter/s)": 1.437705
    },
    {
      "epoch": 3.491709866758065,
      "eval_loss": 2.1931748390197754,
      "eval_runtime": 14.3072,
      "eval_samples_per_second": 6.99,
      "eval_steps_per_second": 6.99,
      "eval_token_acc": 0.46185286103542234,
      "step": 81500
    },
    {
      "epoch": 3.4919240820873143,
      "grad_norm": 5.316596984863281,
      "learning_rate": 2.0816372539890872e-05,
      "loss": 2.1783926010131838,
      "memory(GiB)": 77.56,
      "step": 81505,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437335
    },
    {
      "epoch": 3.492138297416563,
      "grad_norm": 8.70543384552002,
      "learning_rate": 2.081090830165716e-05,
      "loss": 2.3440284729003906,
      "memory(GiB)": 77.56,
      "step": 81510,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 3.492352512745812,
      "grad_norm": 4.562430381774902,
      "learning_rate": 2.0805444592211216e-05,
      "loss": 2.322625923156738,
      "memory(GiB)": 77.56,
      "step": 81515,
      "token_acc": 0.5539358600583091,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.492566728075061,
      "grad_norm": 6.326747894287109,
      "learning_rate": 2.079998141165204e-05,
      "loss": 2.4467071533203124,
      "memory(GiB)": 77.56,
      "step": 81520,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.43735
    },
    {
      "epoch": 3.49278094340431,
      "grad_norm": 9.520673751831055,
      "learning_rate": 2.0794518760078596e-05,
      "loss": 1.9941160202026367,
      "memory(GiB)": 77.56,
      "step": 81525,
      "token_acc": 0.5585585585585585,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.492995158733559,
      "grad_norm": 6.004496097564697,
      "learning_rate": 2.0789056637589845e-05,
      "loss": 2.365845489501953,
      "memory(GiB)": 77.56,
      "step": 81530,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.437361
    },
    {
      "epoch": 3.493209374062808,
      "grad_norm": 4.822202205657959,
      "learning_rate": 2.078359504428473e-05,
      "loss": 2.146456718444824,
      "memory(GiB)": 77.56,
      "step": 81535,
      "token_acc": 0.5518518518518518,
      "train_speed(iter/s)": 1.437384
    },
    {
      "epoch": 3.493423589392057,
      "grad_norm": 5.514254093170166,
      "learning_rate": 2.077813398026218e-05,
      "loss": 2.5107051849365236,
      "memory(GiB)": 77.56,
      "step": 81540,
      "token_acc": 0.4718498659517426,
      "train_speed(iter/s)": 1.437374
    },
    {
      "epoch": 3.4936378047213057,
      "grad_norm": 7.6399946212768555,
      "learning_rate": 2.0772673445621165e-05,
      "loss": 2.2827678680419923,
      "memory(GiB)": 77.56,
      "step": 81545,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.437375
    },
    {
      "epoch": 3.493852020050555,
      "grad_norm": 5.592060565948486,
      "learning_rate": 2.076721344046059e-05,
      "loss": 2.3949600219726563,
      "memory(GiB)": 77.56,
      "step": 81550,
      "token_acc": 0.5228070175438596,
      "train_speed(iter/s)": 1.437398
    },
    {
      "epoch": 3.4940662353798038,
      "grad_norm": 6.032351493835449,
      "learning_rate": 2.0761753964879367e-05,
      "loss": 2.2148674011230467,
      "memory(GiB)": 77.56,
      "step": 81555,
      "token_acc": 0.5289855072463768,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 3.4942804507090526,
      "grad_norm": 6.091811656951904,
      "learning_rate": 2.0756295018976397e-05,
      "loss": 2.4260675430297853,
      "memory(GiB)": 77.56,
      "step": 81560,
      "token_acc": 0.5302013422818792,
      "train_speed(iter/s)": 1.437396
    },
    {
      "epoch": 3.494494666038302,
      "grad_norm": 4.107726097106934,
      "learning_rate": 2.0750836602850553e-05,
      "loss": 2.0095905303955077,
      "memory(GiB)": 77.56,
      "step": 81565,
      "token_acc": 0.5653846153846154,
      "train_speed(iter/s)": 1.437406
    },
    {
      "epoch": 3.4947088813675506,
      "grad_norm": 5.650120735168457,
      "learning_rate": 2.0745378716600766e-05,
      "loss": 2.265056610107422,
      "memory(GiB)": 77.56,
      "step": 81570,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.437411
    },
    {
      "epoch": 3.4949230966967995,
      "grad_norm": 5.963672161102295,
      "learning_rate": 2.0739921360325866e-05,
      "loss": 2.457226371765137,
      "memory(GiB)": 77.56,
      "step": 81575,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 3.4951373120260487,
      "grad_norm": 5.260590553283691,
      "learning_rate": 2.0734464534124754e-05,
      "loss": 2.2100851058959963,
      "memory(GiB)": 77.56,
      "step": 81580,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.437449
    },
    {
      "epoch": 3.4953515273552975,
      "grad_norm": 6.349801540374756,
      "learning_rate": 2.0729008238096275e-05,
      "loss": 2.6900619506835937,
      "memory(GiB)": 77.56,
      "step": 81585,
      "token_acc": 0.48773006134969327,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 3.4955657426845463,
      "grad_norm": 5.960821151733398,
      "learning_rate": 2.0723552472339268e-05,
      "loss": 2.267192840576172,
      "memory(GiB)": 77.56,
      "step": 81590,
      "token_acc": 0.5265151515151515,
      "train_speed(iter/s)": 1.43745
    },
    {
      "epoch": 3.4957799580137956,
      "grad_norm": 5.048311710357666,
      "learning_rate": 2.071809723695257e-05,
      "loss": 2.6000579833984374,
      "memory(GiB)": 77.56,
      "step": 81595,
      "token_acc": 0.4380664652567976,
      "train_speed(iter/s)": 1.437451
    },
    {
      "epoch": 3.4959941733430444,
      "grad_norm": 5.510818958282471,
      "learning_rate": 2.071264253203499e-05,
      "loss": 2.4653921127319336,
      "memory(GiB)": 77.56,
      "step": 81600,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 3.496208388672293,
      "grad_norm": 6.702828884124756,
      "learning_rate": 2.0707188357685393e-05,
      "loss": 2.6925033569335937,
      "memory(GiB)": 77.56,
      "step": 81605,
      "token_acc": 0.4732142857142857,
      "train_speed(iter/s)": 1.437451
    },
    {
      "epoch": 3.4964226040015425,
      "grad_norm": 5.453231334686279,
      "learning_rate": 2.0701734714002545e-05,
      "loss": 2.1410213470458985,
      "memory(GiB)": 77.56,
      "step": 81610,
      "token_acc": 0.590443686006826,
      "train_speed(iter/s)": 1.437451
    },
    {
      "epoch": 3.4966368193307913,
      "grad_norm": 5.779356479644775,
      "learning_rate": 2.0696281601085264e-05,
      "loss": 2.041061019897461,
      "memory(GiB)": 77.56,
      "step": 81615,
      "token_acc": 0.5533596837944664,
      "train_speed(iter/s)": 1.437461
    },
    {
      "epoch": 3.49685103466004,
      "grad_norm": 5.889928340911865,
      "learning_rate": 2.0690829019032327e-05,
      "loss": 2.3318981170654296,
      "memory(GiB)": 77.56,
      "step": 81620,
      "token_acc": 0.55,
      "train_speed(iter/s)": 1.437481
    },
    {
      "epoch": 3.4970652499892894,
      "grad_norm": 6.09144401550293,
      "learning_rate": 2.06853769679425e-05,
      "loss": 2.47115421295166,
      "memory(GiB)": 77.56,
      "step": 81625,
      "token_acc": 0.4713375796178344,
      "train_speed(iter/s)": 1.437507
    },
    {
      "epoch": 3.497279465318538,
      "grad_norm": 4.814365863800049,
      "learning_rate": 2.067992544791459e-05,
      "loss": 2.357257270812988,
      "memory(GiB)": 77.56,
      "step": 81630,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 3.497493680647787,
      "grad_norm": 6.003396511077881,
      "learning_rate": 2.067447445904731e-05,
      "loss": 2.341724395751953,
      "memory(GiB)": 77.56,
      "step": 81635,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.437505
    },
    {
      "epoch": 3.4977078959770362,
      "grad_norm": 6.591564655303955,
      "learning_rate": 2.066902400143946e-05,
      "loss": 2.381968307495117,
      "memory(GiB)": 77.56,
      "step": 81640,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.437531
    },
    {
      "epoch": 3.497922111306285,
      "grad_norm": 7.968791484832764,
      "learning_rate": 2.066357407518975e-05,
      "loss": 2.4607547760009765,
      "memory(GiB)": 77.56,
      "step": 81645,
      "token_acc": 0.46621621621621623,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 3.498136326635534,
      "grad_norm": 4.376980304718018,
      "learning_rate": 2.0658124680396918e-05,
      "loss": 2.211495208740234,
      "memory(GiB)": 77.56,
      "step": 81650,
      "token_acc": 0.5025906735751295,
      "train_speed(iter/s)": 1.437514
    },
    {
      "epoch": 3.498350541964783,
      "grad_norm": 5.4497528076171875,
      "learning_rate": 2.065267581715968e-05,
      "loss": 2.395167922973633,
      "memory(GiB)": 77.56,
      "step": 81655,
      "token_acc": 0.5243055555555556,
      "train_speed(iter/s)": 1.437517
    },
    {
      "epoch": 3.498564757294032,
      "grad_norm": 5.857762813568115,
      "learning_rate": 2.0647227485576733e-05,
      "loss": 2.256017303466797,
      "memory(GiB)": 77.56,
      "step": 81660,
      "token_acc": 0.5714285714285714,
      "train_speed(iter/s)": 1.437503
    },
    {
      "epoch": 3.4987789726232807,
      "grad_norm": 5.593625545501709,
      "learning_rate": 2.064177968574681e-05,
      "loss": 2.181442070007324,
      "memory(GiB)": 77.56,
      "step": 81665,
      "token_acc": 0.5016393442622951,
      "train_speed(iter/s)": 1.437535
    },
    {
      "epoch": 3.49899318795253,
      "grad_norm": 6.4224982261657715,
      "learning_rate": 2.0636332417768595e-05,
      "loss": 2.054220962524414,
      "memory(GiB)": 77.56,
      "step": 81670,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.43754
    },
    {
      "epoch": 3.499207403281779,
      "grad_norm": 6.219550609588623,
      "learning_rate": 2.0630885681740758e-05,
      "loss": 2.1898889541625977,
      "memory(GiB)": 77.56,
      "step": 81675,
      "token_acc": 0.5033112582781457,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 3.4994216186110276,
      "grad_norm": 7.5490922927856445,
      "learning_rate": 2.0625439477761975e-05,
      "loss": 2.2260009765625,
      "memory(GiB)": 77.56,
      "step": 81680,
      "token_acc": 0.5447470817120622,
      "train_speed(iter/s)": 1.437503
    },
    {
      "epoch": 3.499635833940277,
      "grad_norm": 4.393792152404785,
      "learning_rate": 2.0619993805930904e-05,
      "loss": 2.2037561416625975,
      "memory(GiB)": 77.56,
      "step": 81685,
      "token_acc": 0.4972067039106145,
      "train_speed(iter/s)": 1.437494
    },
    {
      "epoch": 3.4998500492695257,
      "grad_norm": 6.435640335083008,
      "learning_rate": 2.0614548666346212e-05,
      "loss": 2.566878890991211,
      "memory(GiB)": 77.56,
      "step": 81690,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.437489
    },
    {
      "epoch": 3.500064264598775,
      "grad_norm": 5.867866039276123,
      "learning_rate": 2.060910405910654e-05,
      "loss": 2.516805648803711,
      "memory(GiB)": 77.56,
      "step": 81695,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.437471
    },
    {
      "epoch": 3.5002784799280238,
      "grad_norm": 8.683708190917969,
      "learning_rate": 2.0603659984310525e-05,
      "loss": 2.34912223815918,
      "memory(GiB)": 77.56,
      "step": 81700,
      "token_acc": 0.4713656387665198,
      "train_speed(iter/s)": 1.437477
    },
    {
      "epoch": 3.5004926952572726,
      "grad_norm": 8.144967079162598,
      "learning_rate": 2.0598216442056784e-05,
      "loss": 2.360447120666504,
      "memory(GiB)": 77.56,
      "step": 81705,
      "token_acc": 0.5170068027210885,
      "train_speed(iter/s)": 1.437482
    },
    {
      "epoch": 3.500706910586522,
      "grad_norm": 4.874293327331543,
      "learning_rate": 2.0592773432443913e-05,
      "loss": 2.5154451370239257,
      "memory(GiB)": 77.56,
      "step": 81710,
      "token_acc": 0.4517133956386293,
      "train_speed(iter/s)": 1.437457
    },
    {
      "epoch": 3.5009211259157706,
      "grad_norm": 5.730638027191162,
      "learning_rate": 2.058733095557055e-05,
      "loss": 2.2578094482421873,
      "memory(GiB)": 77.56,
      "step": 81715,
      "token_acc": 0.5232974910394266,
      "train_speed(iter/s)": 1.43742
    },
    {
      "epoch": 3.5011353412450195,
      "grad_norm": 6.449235439300537,
      "learning_rate": 2.0581889011535264e-05,
      "loss": 2.304231071472168,
      "memory(GiB)": 77.56,
      "step": 81720,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.437429
    },
    {
      "epoch": 3.5013495565742687,
      "grad_norm": 5.849225044250488,
      "learning_rate": 2.0576447600436676e-05,
      "loss": 2.5435646057128904,
      "memory(GiB)": 77.56,
      "step": 81725,
      "token_acc": 0.5078740157480315,
      "train_speed(iter/s)": 1.437414
    },
    {
      "epoch": 3.5015637719035175,
      "grad_norm": 5.297560691833496,
      "learning_rate": 2.0571006722373343e-05,
      "loss": 2.2740482330322265,
      "memory(GiB)": 77.56,
      "step": 81730,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437396
    },
    {
      "epoch": 3.5017779872327663,
      "grad_norm": 6.248037338256836,
      "learning_rate": 2.0565566377443823e-05,
      "loss": 2.4441009521484376,
      "memory(GiB)": 77.56,
      "step": 81735,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.437391
    },
    {
      "epoch": 3.5019922025620156,
      "grad_norm": 5.896634578704834,
      "learning_rate": 2.0560126565746686e-05,
      "loss": 2.598617935180664,
      "memory(GiB)": 77.56,
      "step": 81740,
      "token_acc": 0.4724137931034483,
      "train_speed(iter/s)": 1.437403
    },
    {
      "epoch": 3.5022064178912644,
      "grad_norm": 4.903329849243164,
      "learning_rate": 2.0554687287380447e-05,
      "loss": 2.1694101333618163,
      "memory(GiB)": 77.56,
      "step": 81745,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437407
    },
    {
      "epoch": 3.5024206332205132,
      "grad_norm": 5.899200439453125,
      "learning_rate": 2.0549248542443695e-05,
      "loss": 2.40710391998291,
      "memory(GiB)": 77.56,
      "step": 81750,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.437433
    },
    {
      "epoch": 3.5026348485497625,
      "grad_norm": 7.038031101226807,
      "learning_rate": 2.0543810331034928e-05,
      "loss": 2.2765064239501953,
      "memory(GiB)": 77.56,
      "step": 81755,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.437414
    },
    {
      "epoch": 3.5028490638790113,
      "grad_norm": 5.879939556121826,
      "learning_rate": 2.0538372653252673e-05,
      "loss": 2.6087474822998047,
      "memory(GiB)": 77.56,
      "step": 81760,
      "token_acc": 0.46397694524495675,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 3.50306327920826,
      "grad_norm": 5.754982948303223,
      "learning_rate": 2.0532935509195432e-05,
      "loss": 2.3488487243652343,
      "memory(GiB)": 77.56,
      "step": 81765,
      "token_acc": 0.49640287769784175,
      "train_speed(iter/s)": 1.437419
    },
    {
      "epoch": 3.5032774945375094,
      "grad_norm": 5.459589958190918,
      "learning_rate": 2.0527498898961695e-05,
      "loss": 2.2230276107788085,
      "memory(GiB)": 77.56,
      "step": 81770,
      "token_acc": 0.5539568345323741,
      "train_speed(iter/s)": 1.437421
    },
    {
      "epoch": 3.503491709866758,
      "grad_norm": 5.302542209625244,
      "learning_rate": 2.0522062822649973e-05,
      "loss": 2.6603164672851562,
      "memory(GiB)": 77.56,
      "step": 81775,
      "token_acc": 0.47928994082840237,
      "train_speed(iter/s)": 1.437428
    },
    {
      "epoch": 3.503705925196007,
      "grad_norm": 8.425333976745605,
      "learning_rate": 2.0516627280358737e-05,
      "loss": 2.5466064453125,
      "memory(GiB)": 77.56,
      "step": 81780,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.437424
    },
    {
      "epoch": 3.5039201405252562,
      "grad_norm": 6.835321426391602,
      "learning_rate": 2.051119227218644e-05,
      "loss": 2.2706045150756835,
      "memory(GiB)": 77.56,
      "step": 81785,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.437458
    },
    {
      "epoch": 3.504134355854505,
      "grad_norm": 5.7809319496154785,
      "learning_rate": 2.0505757798231573e-05,
      "loss": 2.8129369735717775,
      "memory(GiB)": 77.56,
      "step": 81790,
      "token_acc": 0.44744744744744747,
      "train_speed(iter/s)": 1.437455
    },
    {
      "epoch": 3.504348571183754,
      "grad_norm": 5.822942733764648,
      "learning_rate": 2.0500323858592572e-05,
      "loss": 2.3328447341918945,
      "memory(GiB)": 77.56,
      "step": 81795,
      "token_acc": 0.490625,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 3.504562786513003,
      "grad_norm": 7.569825172424316,
      "learning_rate": 2.0494890453367875e-05,
      "loss": 2.5288734436035156,
      "memory(GiB)": 77.56,
      "step": 81800,
      "token_acc": 0.4577922077922078,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 3.504777001842252,
      "grad_norm": 7.130863189697266,
      "learning_rate": 2.0489457582655902e-05,
      "loss": 2.171244812011719,
      "memory(GiB)": 77.56,
      "step": 81805,
      "token_acc": 0.5430711610486891,
      "train_speed(iter/s)": 1.437491
    },
    {
      "epoch": 3.5049912171715008,
      "grad_norm": 5.355434894561768,
      "learning_rate": 2.0484025246555107e-05,
      "loss": 2.0865997314453124,
      "memory(GiB)": 77.56,
      "step": 81810,
      "token_acc": 0.5320754716981132,
      "train_speed(iter/s)": 1.437488
    },
    {
      "epoch": 3.50520543250075,
      "grad_norm": 4.92542028427124,
      "learning_rate": 2.047859344516388e-05,
      "loss": 2.3790191650390624,
      "memory(GiB)": 77.56,
      "step": 81815,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 3.505419647829999,
      "grad_norm": 5.839577674865723,
      "learning_rate": 2.0473162178580622e-05,
      "loss": 2.576974296569824,
      "memory(GiB)": 77.56,
      "step": 81820,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 3.5056338631592476,
      "grad_norm": 4.92479944229126,
      "learning_rate": 2.0467731446903736e-05,
      "loss": 2.3958988189697266,
      "memory(GiB)": 77.56,
      "step": 81825,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.437488
    },
    {
      "epoch": 3.505848078488497,
      "grad_norm": 6.27665376663208,
      "learning_rate": 2.046230125023158e-05,
      "loss": 2.234067916870117,
      "memory(GiB)": 77.56,
      "step": 81830,
      "token_acc": 0.5244299674267101,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 3.5060622938177457,
      "grad_norm": 5.429213047027588,
      "learning_rate": 2.0456871588662563e-05,
      "loss": 2.456011962890625,
      "memory(GiB)": 77.56,
      "step": 81835,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.437489
    },
    {
      "epoch": 3.5062765091469945,
      "grad_norm": 5.514077663421631,
      "learning_rate": 2.0451442462295028e-05,
      "loss": 2.261183738708496,
      "memory(GiB)": 77.56,
      "step": 81840,
      "token_acc": 0.45481049562682213,
      "train_speed(iter/s)": 1.437494
    },
    {
      "epoch": 3.5064907244762438,
      "grad_norm": 5.656719207763672,
      "learning_rate": 2.044601387122733e-05,
      "loss": 2.2159067153930665,
      "memory(GiB)": 77.56,
      "step": 81845,
      "token_acc": 0.5442622950819672,
      "train_speed(iter/s)": 1.437503
    },
    {
      "epoch": 3.5067049398054926,
      "grad_norm": 7.250095844268799,
      "learning_rate": 2.044058581555782e-05,
      "loss": 2.1542308807373045,
      "memory(GiB)": 77.56,
      "step": 81850,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.437502
    },
    {
      "epoch": 3.5069191551347414,
      "grad_norm": 6.025059223175049,
      "learning_rate": 2.0435158295384826e-05,
      "loss": 2.0294025421142576,
      "memory(GiB)": 77.56,
      "step": 81855,
      "token_acc": 0.5756457564575646,
      "train_speed(iter/s)": 1.437483
    },
    {
      "epoch": 3.5071333704639907,
      "grad_norm": 4.839198112487793,
      "learning_rate": 2.0429731310806655e-05,
      "loss": 2.4145973205566404,
      "memory(GiB)": 77.56,
      "step": 81860,
      "token_acc": 0.49074074074074076,
      "train_speed(iter/s)": 1.437505
    },
    {
      "epoch": 3.5073475857932395,
      "grad_norm": 6.596020221710205,
      "learning_rate": 2.042430486192164e-05,
      "loss": 2.3078685760498048,
      "memory(GiB)": 77.56,
      "step": 81865,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.437516
    },
    {
      "epoch": 3.5075618011224883,
      "grad_norm": 6.517064094543457,
      "learning_rate": 2.04188789488281e-05,
      "loss": 2.580352783203125,
      "memory(GiB)": 77.56,
      "step": 81870,
      "token_acc": 0.44242424242424244,
      "train_speed(iter/s)": 1.437511
    },
    {
      "epoch": 3.5077760164517375,
      "grad_norm": 6.77636194229126,
      "learning_rate": 2.0413453571624323e-05,
      "loss": 2.6035734176635743,
      "memory(GiB)": 77.56,
      "step": 81875,
      "token_acc": 0.4831932773109244,
      "train_speed(iter/s)": 1.437524
    },
    {
      "epoch": 3.5079902317809863,
      "grad_norm": 7.556825160980225,
      "learning_rate": 2.0408028730408584e-05,
      "loss": 2.0738378524780274,
      "memory(GiB)": 77.56,
      "step": 81880,
      "token_acc": 0.5953307392996109,
      "train_speed(iter/s)": 1.437543
    },
    {
      "epoch": 3.508204447110235,
      "grad_norm": 7.605564594268799,
      "learning_rate": 2.040260442527917e-05,
      "loss": 2.4505847930908202,
      "memory(GiB)": 77.56,
      "step": 81885,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.43755
    },
    {
      "epoch": 3.5084186624394844,
      "grad_norm": 6.6986517906188965,
      "learning_rate": 2.0397180656334318e-05,
      "loss": 2.1217586517333986,
      "memory(GiB)": 77.56,
      "step": 81890,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.437552
    },
    {
      "epoch": 3.5086328777687332,
      "grad_norm": 6.056543827056885,
      "learning_rate": 2.039175742367233e-05,
      "loss": 2.1005084991455076,
      "memory(GiB)": 77.56,
      "step": 81895,
      "token_acc": 0.5298013245033113,
      "train_speed(iter/s)": 1.437545
    },
    {
      "epoch": 3.508847093097982,
      "grad_norm": 5.604673862457275,
      "learning_rate": 2.0386334727391432e-05,
      "loss": 2.2785987854003906,
      "memory(GiB)": 77.56,
      "step": 81900,
      "token_acc": 0.5295698924731183,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 3.5090613084272313,
      "grad_norm": 4.90866756439209,
      "learning_rate": 2.038091256758986e-05,
      "loss": 2.545729064941406,
      "memory(GiB)": 77.56,
      "step": 81905,
      "token_acc": 0.4805194805194805,
      "train_speed(iter/s)": 1.437573
    },
    {
      "epoch": 3.50927552375648,
      "grad_norm": 6.3202595710754395,
      "learning_rate": 2.037549094436584e-05,
      "loss": 2.1428165435791016,
      "memory(GiB)": 77.56,
      "step": 81910,
      "token_acc": 0.5092936802973977,
      "train_speed(iter/s)": 1.437584
    },
    {
      "epoch": 3.509489739085729,
      "grad_norm": 6.21103048324585,
      "learning_rate": 2.0370069857817576e-05,
      "loss": 2.1100795745849608,
      "memory(GiB)": 77.56,
      "step": 81915,
      "token_acc": 0.5461538461538461,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 3.509703954414978,
      "grad_norm": 5.559045314788818,
      "learning_rate": 2.0364649308043303e-05,
      "loss": 2.4545820236206053,
      "memory(GiB)": 77.56,
      "step": 81920,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.437577
    },
    {
      "epoch": 3.509918169744227,
      "grad_norm": 6.621930122375488,
      "learning_rate": 2.0359229295141213e-05,
      "loss": 2.326128387451172,
      "memory(GiB)": 77.56,
      "step": 81925,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.437587
    },
    {
      "epoch": 3.510132385073476,
      "grad_norm": 5.367328643798828,
      "learning_rate": 2.035380981920949e-05,
      "loss": 2.347463607788086,
      "memory(GiB)": 77.56,
      "step": 81930,
      "token_acc": 0.5119453924914675,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 3.510346600402725,
      "grad_norm": 8.202676773071289,
      "learning_rate": 2.034839088034629e-05,
      "loss": 2.2407098770141602,
      "memory(GiB)": 77.56,
      "step": 81935,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.437585
    },
    {
      "epoch": 3.510560815731974,
      "grad_norm": 5.3595967292785645,
      "learning_rate": 2.034297247864983e-05,
      "loss": 2.0268245697021485,
      "memory(GiB)": 77.56,
      "step": 81940,
      "token_acc": 0.5631067961165048,
      "train_speed(iter/s)": 1.437559
    },
    {
      "epoch": 3.5107750310612227,
      "grad_norm": 6.485270977020264,
      "learning_rate": 2.033755461421824e-05,
      "loss": 2.417161560058594,
      "memory(GiB)": 77.56,
      "step": 81945,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.437573
    },
    {
      "epoch": 3.510989246390472,
      "grad_norm": 5.715497016906738,
      "learning_rate": 2.0332137287149654e-05,
      "loss": 2.437858772277832,
      "memory(GiB)": 77.56,
      "step": 81950,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.437598
    },
    {
      "epoch": 3.5112034617197208,
      "grad_norm": 7.498754978179932,
      "learning_rate": 2.0326720497542252e-05,
      "loss": 2.150317573547363,
      "memory(GiB)": 77.56,
      "step": 81955,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.4376
    },
    {
      "epoch": 3.5114176770489696,
      "grad_norm": 5.521462440490723,
      "learning_rate": 2.0321304245494145e-05,
      "loss": 2.2425622940063477,
      "memory(GiB)": 77.56,
      "step": 81960,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.437605
    },
    {
      "epoch": 3.511631892378219,
      "grad_norm": 7.329811096191406,
      "learning_rate": 2.031588853110345e-05,
      "loss": 2.3874088287353517,
      "memory(GiB)": 77.56,
      "step": 81965,
      "token_acc": 0.5176056338028169,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 3.5118461077074676,
      "grad_norm": 7.661202907562256,
      "learning_rate": 2.0310473354468283e-05,
      "loss": 2.218642997741699,
      "memory(GiB)": 77.56,
      "step": 81970,
      "token_acc": 0.5084175084175084,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 3.5120603230367164,
      "grad_norm": 6.89029598236084,
      "learning_rate": 2.030505871568672e-05,
      "loss": 2.4021123886108398,
      "memory(GiB)": 77.56,
      "step": 81975,
      "token_acc": 0.5210727969348659,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 3.5122745383659657,
      "grad_norm": 5.101368427276611,
      "learning_rate": 2.0299644614856895e-05,
      "loss": 2.3878353118896483,
      "memory(GiB)": 77.56,
      "step": 81980,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.437632
    },
    {
      "epoch": 3.5124887536952145,
      "grad_norm": 5.6344990730285645,
      "learning_rate": 2.0294231052076874e-05,
      "loss": 2.1968948364257814,
      "memory(GiB)": 77.56,
      "step": 81985,
      "token_acc": 0.5186335403726708,
      "train_speed(iter/s)": 1.437636
    },
    {
      "epoch": 3.5127029690244633,
      "grad_norm": 6.068678855895996,
      "learning_rate": 2.028881802744472e-05,
      "loss": 2.401434326171875,
      "memory(GiB)": 77.56,
      "step": 81990,
      "token_acc": 0.5174418604651163,
      "train_speed(iter/s)": 1.437636
    },
    {
      "epoch": 3.5129171843537126,
      "grad_norm": 7.214240074157715,
      "learning_rate": 2.0283405541058494e-05,
      "loss": 2.4219038009643556,
      "memory(GiB)": 77.56,
      "step": 81995,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 3.5131313996829614,
      "grad_norm": 5.917394638061523,
      "learning_rate": 2.0277993593016252e-05,
      "loss": 2.2924312591552733,
      "memory(GiB)": 77.56,
      "step": 82000,
      "token_acc": 0.5179153094462541,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 3.5131313996829614,
      "eval_loss": 2.2839174270629883,
      "eval_runtime": 14.06,
      "eval_samples_per_second": 7.112,
      "eval_steps_per_second": 7.112,
      "eval_token_acc": 0.4642857142857143,
      "step": 82000
    },
    {
      "epoch": 3.51334561501221,
      "grad_norm": 6.334027290344238,
      "learning_rate": 2.0272582183416027e-05,
      "loss": 2.2294769287109375,
      "memory(GiB)": 77.56,
      "step": 82005,
      "token_acc": 0.47673314339981004,
      "train_speed(iter/s)": 1.437242
    },
    {
      "epoch": 3.5135598303414595,
      "grad_norm": 6.302854061126709,
      "learning_rate": 2.026717131235586e-05,
      "loss": 2.472804641723633,
      "memory(GiB)": 77.56,
      "step": 82010,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.437249
    },
    {
      "epoch": 3.5137740456707083,
      "grad_norm": 7.024232387542725,
      "learning_rate": 2.0261760979933796e-05,
      "loss": 2.2838665008544923,
      "memory(GiB)": 77.56,
      "step": 82015,
      "token_acc": 0.4847560975609756,
      "train_speed(iter/s)": 1.437265
    },
    {
      "epoch": 3.513988260999957,
      "grad_norm": 4.253474235534668,
      "learning_rate": 2.0256351186247824e-05,
      "loss": 2.4336471557617188,
      "memory(GiB)": 77.56,
      "step": 82020,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.437267
    },
    {
      "epoch": 3.5142024763292063,
      "grad_norm": 6.034378528594971,
      "learning_rate": 2.0250941931395957e-05,
      "loss": 2.303436279296875,
      "memory(GiB)": 77.56,
      "step": 82025,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.43726
    },
    {
      "epoch": 3.514416691658455,
      "grad_norm": 5.84157657623291,
      "learning_rate": 2.024553321547618e-05,
      "loss": 2.523159980773926,
      "memory(GiB)": 77.56,
      "step": 82030,
      "token_acc": 0.4336283185840708,
      "train_speed(iter/s)": 1.437259
    },
    {
      "epoch": 3.514630906987704,
      "grad_norm": 5.684593200683594,
      "learning_rate": 2.0240125038586465e-05,
      "loss": 2.3350486755371094,
      "memory(GiB)": 77.56,
      "step": 82035,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.437264
    },
    {
      "epoch": 3.5148451223169532,
      "grad_norm": 5.939591407775879,
      "learning_rate": 2.0234717400824814e-05,
      "loss": 2.566223907470703,
      "memory(GiB)": 77.56,
      "step": 82040,
      "token_acc": 0.46905537459283386,
      "train_speed(iter/s)": 1.437263
    },
    {
      "epoch": 3.515059337646202,
      "grad_norm": 4.637340068817139,
      "learning_rate": 2.022931030228919e-05,
      "loss": 2.2540449142456054,
      "memory(GiB)": 77.56,
      "step": 82045,
      "token_acc": 0.5471698113207547,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 3.515273552975451,
      "grad_norm": 5.105705261230469,
      "learning_rate": 2.0223903743077528e-05,
      "loss": 2.1845422744750977,
      "memory(GiB)": 77.56,
      "step": 82050,
      "token_acc": 0.5435540069686411,
      "train_speed(iter/s)": 1.437312
    },
    {
      "epoch": 3.5154877683047,
      "grad_norm": 5.498888969421387,
      "learning_rate": 2.0218497723287788e-05,
      "loss": 2.3621320724487305,
      "memory(GiB)": 77.56,
      "step": 82055,
      "token_acc": 0.5112994350282486,
      "train_speed(iter/s)": 1.437325
    },
    {
      "epoch": 3.515701983633949,
      "grad_norm": 5.912207126617432,
      "learning_rate": 2.021309224301788e-05,
      "loss": 1.9903440475463867,
      "memory(GiB)": 77.56,
      "step": 82060,
      "token_acc": 0.5597269624573379,
      "train_speed(iter/s)": 1.437336
    },
    {
      "epoch": 3.5159161989631977,
      "grad_norm": 6.581540584564209,
      "learning_rate": 2.0207687302365762e-05,
      "loss": 2.4122432708740233,
      "memory(GiB)": 77.56,
      "step": 82065,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.437323
    },
    {
      "epoch": 3.516130414292447,
      "grad_norm": 5.1985673904418945,
      "learning_rate": 2.0202282901429338e-05,
      "loss": 2.209666633605957,
      "memory(GiB)": 77.56,
      "step": 82070,
      "token_acc": 0.5278969957081545,
      "train_speed(iter/s)": 1.43734
    },
    {
      "epoch": 3.516344629621696,
      "grad_norm": 6.54959774017334,
      "learning_rate": 2.019687904030651e-05,
      "loss": 2.4396503448486326,
      "memory(GiB)": 77.56,
      "step": 82075,
      "token_acc": 0.4691358024691358,
      "train_speed(iter/s)": 1.437362
    },
    {
      "epoch": 3.5165588449509446,
      "grad_norm": 6.053768634796143,
      "learning_rate": 2.0191475719095166e-05,
      "loss": 2.1498748779296877,
      "memory(GiB)": 77.56,
      "step": 82080,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.437356
    },
    {
      "epoch": 3.516773060280194,
      "grad_norm": 5.249053001403809,
      "learning_rate": 2.0186072937893212e-05,
      "loss": 2.1845705032348635,
      "memory(GiB)": 77.56,
      "step": 82085,
      "token_acc": 0.5013698630136987,
      "train_speed(iter/s)": 1.437356
    },
    {
      "epoch": 3.5169872756094427,
      "grad_norm": 6.844965934753418,
      "learning_rate": 2.0180670696798514e-05,
      "loss": 2.3761863708496094,
      "memory(GiB)": 77.56,
      "step": 82090,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.437358
    },
    {
      "epoch": 3.5172014909386915,
      "grad_norm": 6.118383884429932,
      "learning_rate": 2.0175268995908924e-05,
      "loss": 2.374035453796387,
      "memory(GiB)": 77.56,
      "step": 82095,
      "token_acc": 0.5214521452145214,
      "train_speed(iter/s)": 1.437351
    },
    {
      "epoch": 3.5174157062679408,
      "grad_norm": 6.474025249481201,
      "learning_rate": 2.0169867835322332e-05,
      "loss": 2.2998636245727537,
      "memory(GiB)": 77.56,
      "step": 82100,
      "token_acc": 0.5310344827586206,
      "train_speed(iter/s)": 1.437361
    },
    {
      "epoch": 3.5176299215971896,
      "grad_norm": 5.027404308319092,
      "learning_rate": 2.0164467215136566e-05,
      "loss": 2.2003076553344725,
      "memory(GiB)": 77.56,
      "step": 82105,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.437367
    },
    {
      "epoch": 3.5178441369264384,
      "grad_norm": 4.83841609954834,
      "learning_rate": 2.0159067135449467e-05,
      "loss": 2.483617401123047,
      "memory(GiB)": 77.56,
      "step": 82110,
      "token_acc": 0.4915254237288136,
      "train_speed(iter/s)": 1.43735
    },
    {
      "epoch": 3.5180583522556876,
      "grad_norm": 5.018257141113281,
      "learning_rate": 2.0153667596358862e-05,
      "loss": 2.6086502075195312,
      "memory(GiB)": 77.56,
      "step": 82115,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.437344
    },
    {
      "epoch": 3.5182725675849364,
      "grad_norm": 7.816195011138916,
      "learning_rate": 2.0148268597962544e-05,
      "loss": 2.1150815963745115,
      "memory(GiB)": 77.56,
      "step": 82120,
      "token_acc": 0.5488215488215489,
      "train_speed(iter/s)": 1.437343
    },
    {
      "epoch": 3.5184867829141853,
      "grad_norm": 7.717521667480469,
      "learning_rate": 2.014287014035836e-05,
      "loss": 2.6598011016845704,
      "memory(GiB)": 77.56,
      "step": 82125,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.43736
    },
    {
      "epoch": 3.5187009982434345,
      "grad_norm": 11.461997985839844,
      "learning_rate": 2.0137472223644093e-05,
      "loss": 2.650553894042969,
      "memory(GiB)": 77.56,
      "step": 82130,
      "token_acc": 0.46757679180887374,
      "train_speed(iter/s)": 1.437342
    },
    {
      "epoch": 3.5189152135726833,
      "grad_norm": 6.628066539764404,
      "learning_rate": 2.0132074847917533e-05,
      "loss": 2.4993259429931642,
      "memory(GiB)": 77.56,
      "step": 82135,
      "token_acc": 0.4854014598540146,
      "train_speed(iter/s)": 1.437328
    },
    {
      "epoch": 3.519129428901932,
      "grad_norm": 7.059751987457275,
      "learning_rate": 2.012667801327645e-05,
      "loss": 2.2334800720214845,
      "memory(GiB)": 77.56,
      "step": 82140,
      "token_acc": 0.518796992481203,
      "train_speed(iter/s)": 1.437317
    },
    {
      "epoch": 3.5193436442311814,
      "grad_norm": 6.7917304039001465,
      "learning_rate": 2.0121281719818625e-05,
      "loss": 2.2749711990356447,
      "memory(GiB)": 77.56,
      "step": 82145,
      "token_acc": 0.49032258064516127,
      "train_speed(iter/s)": 1.437319
    },
    {
      "epoch": 3.51955785956043,
      "grad_norm": 5.486000061035156,
      "learning_rate": 2.011588596764179e-05,
      "loss": 2.346443939208984,
      "memory(GiB)": 77.56,
      "step": 82150,
      "token_acc": 0.5032467532467533,
      "train_speed(iter/s)": 1.437333
    },
    {
      "epoch": 3.519772074889679,
      "grad_norm": 5.41943359375,
      "learning_rate": 2.0110490756843714e-05,
      "loss": 2.428234672546387,
      "memory(GiB)": 77.56,
      "step": 82155,
      "token_acc": 0.4793650793650794,
      "train_speed(iter/s)": 1.437359
    },
    {
      "epoch": 3.5199862902189283,
      "grad_norm": 4.3501877784729,
      "learning_rate": 2.0105096087522153e-05,
      "loss": 2.333928680419922,
      "memory(GiB)": 77.56,
      "step": 82160,
      "token_acc": 0.49363057324840764,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.520200505548177,
      "grad_norm": 6.327761650085449,
      "learning_rate": 2.009970195977482e-05,
      "loss": 2.576307487487793,
      "memory(GiB)": 77.56,
      "step": 82165,
      "token_acc": 0.4458204334365325,
      "train_speed(iter/s)": 1.437382
    },
    {
      "epoch": 3.520414720877426,
      "grad_norm": 7.110805988311768,
      "learning_rate": 2.0094308373699434e-05,
      "loss": 2.0621532440185546,
      "memory(GiB)": 77.56,
      "step": 82170,
      "token_acc": 0.5061349693251533,
      "train_speed(iter/s)": 1.437386
    },
    {
      "epoch": 3.520628936206675,
      "grad_norm": 7.577486038208008,
      "learning_rate": 2.008891532939371e-05,
      "loss": 2.73079833984375,
      "memory(GiB)": 77.56,
      "step": 82175,
      "token_acc": 0.4370629370629371,
      "train_speed(iter/s)": 1.437405
    },
    {
      "epoch": 3.520843151535924,
      "grad_norm": 7.6457719802856445,
      "learning_rate": 2.008352282695532e-05,
      "loss": 2.592006492614746,
      "memory(GiB)": 77.56,
      "step": 82180,
      "token_acc": 0.4672131147540984,
      "train_speed(iter/s)": 1.437418
    },
    {
      "epoch": 3.521057366865173,
      "grad_norm": 4.591124057769775,
      "learning_rate": 2.0078130866481998e-05,
      "loss": 2.239780044555664,
      "memory(GiB)": 77.56,
      "step": 82185,
      "token_acc": 0.4935897435897436,
      "train_speed(iter/s)": 1.437412
    },
    {
      "epoch": 3.521271582194422,
      "grad_norm": 6.016817569732666,
      "learning_rate": 2.0072739448071405e-05,
      "loss": 2.174859809875488,
      "memory(GiB)": 77.56,
      "step": 82190,
      "token_acc": 0.5176470588235295,
      "train_speed(iter/s)": 1.437417
    },
    {
      "epoch": 3.521485797523671,
      "grad_norm": 7.4195661544799805,
      "learning_rate": 2.0067348571821214e-05,
      "loss": 2.3572559356689453,
      "memory(GiB)": 77.56,
      "step": 82195,
      "token_acc": 0.46502057613168724,
      "train_speed(iter/s)": 1.437416
    },
    {
      "epoch": 3.5217000128529197,
      "grad_norm": 5.714663028717041,
      "learning_rate": 2.0061958237829075e-05,
      "loss": 2.1579971313476562,
      "memory(GiB)": 77.56,
      "step": 82200,
      "token_acc": 0.5387323943661971,
      "train_speed(iter/s)": 1.437422
    },
    {
      "epoch": 3.521914228182169,
      "grad_norm": 6.145798683166504,
      "learning_rate": 2.0056568446192636e-05,
      "loss": 2.239365005493164,
      "memory(GiB)": 77.56,
      "step": 82205,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.437388
    },
    {
      "epoch": 3.5221284435114177,
      "grad_norm": 5.390622615814209,
      "learning_rate": 2.0051179197009568e-05,
      "loss": 2.5779598236083983,
      "memory(GiB)": 77.56,
      "step": 82210,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 3.5223426588406666,
      "grad_norm": 5.716423988342285,
      "learning_rate": 2.0045790490377476e-05,
      "loss": 2.153798484802246,
      "memory(GiB)": 77.56,
      "step": 82215,
      "token_acc": 0.4983922829581994,
      "train_speed(iter/s)": 1.437385
    },
    {
      "epoch": 3.522556874169916,
      "grad_norm": 5.381307601928711,
      "learning_rate": 2.0040402326393993e-05,
      "loss": 2.4117923736572267,
      "memory(GiB)": 77.56,
      "step": 82220,
      "token_acc": 0.45054945054945056,
      "train_speed(iter/s)": 1.43736
    },
    {
      "epoch": 3.5227710894991646,
      "grad_norm": 8.643548965454102,
      "learning_rate": 2.0035014705156723e-05,
      "loss": 2.2229400634765626,
      "memory(GiB)": 77.56,
      "step": 82225,
      "token_acc": 0.5463917525773195,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 3.5229853048284134,
      "grad_norm": 7.187229633331299,
      "learning_rate": 2.002962762676326e-05,
      "loss": 2.3501636505126955,
      "memory(GiB)": 77.56,
      "step": 82230,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.437376
    },
    {
      "epoch": 3.5231995201576627,
      "grad_norm": 5.810476303100586,
      "learning_rate": 2.0024241091311218e-05,
      "loss": 2.4225767135620115,
      "memory(GiB)": 77.56,
      "step": 82235,
      "token_acc": 0.5038461538461538,
      "train_speed(iter/s)": 1.437371
    },
    {
      "epoch": 3.5234137354869115,
      "grad_norm": 7.631504535675049,
      "learning_rate": 2.0018855098898153e-05,
      "loss": 2.152008056640625,
      "memory(GiB)": 77.56,
      "step": 82240,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.437355
    },
    {
      "epoch": 3.5236279508161603,
      "grad_norm": 6.231959342956543,
      "learning_rate": 2.001346964962167e-05,
      "loss": 2.5960718154907227,
      "memory(GiB)": 77.56,
      "step": 82245,
      "token_acc": 0.4652567975830816,
      "train_speed(iter/s)": 1.437366
    },
    {
      "epoch": 3.5238421661454096,
      "grad_norm": 5.141421794891357,
      "learning_rate": 2.0008084743579313e-05,
      "loss": 2.4458024978637694,
      "memory(GiB)": 77.56,
      "step": 82250,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.437376
    },
    {
      "epoch": 3.5240563814746584,
      "grad_norm": 9.120293617248535,
      "learning_rate": 2.0002700380868633e-05,
      "loss": 2.559148406982422,
      "memory(GiB)": 77.56,
      "step": 82255,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.437377
    },
    {
      "epoch": 3.524270596803907,
      "grad_norm": 5.958583354949951,
      "learning_rate": 1.999731656158718e-05,
      "loss": 2.4650705337524412,
      "memory(GiB)": 77.56,
      "step": 82260,
      "token_acc": 0.4588607594936709,
      "train_speed(iter/s)": 1.437369
    },
    {
      "epoch": 3.5244848121331565,
      "grad_norm": 8.771286964416504,
      "learning_rate": 1.9991933285832465e-05,
      "loss": 2.351560592651367,
      "memory(GiB)": 77.56,
      "step": 82265,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.437378
    },
    {
      "epoch": 3.5246990274624053,
      "grad_norm": 5.028775691986084,
      "learning_rate": 1.9986550553702045e-05,
      "loss": 2.2397974014282225,
      "memory(GiB)": 77.56,
      "step": 82270,
      "token_acc": 0.5105633802816901,
      "train_speed(iter/s)": 1.437401
    },
    {
      "epoch": 3.524913242791654,
      "grad_norm": 6.096969127655029,
      "learning_rate": 1.998116836529341e-05,
      "loss": 2.4847530364990233,
      "memory(GiB)": 77.56,
      "step": 82275,
      "token_acc": 0.5190311418685121,
      "train_speed(iter/s)": 1.437394
    },
    {
      "epoch": 3.5251274581209033,
      "grad_norm": 6.039845943450928,
      "learning_rate": 1.9975786720704083e-05,
      "loss": 2.4454181671142576,
      "memory(GiB)": 77.56,
      "step": 82280,
      "token_acc": 0.5261538461538462,
      "train_speed(iter/s)": 1.43741
    },
    {
      "epoch": 3.525341673450152,
      "grad_norm": 6.259216785430908,
      "learning_rate": 1.9970405620031534e-05,
      "loss": 2.182073211669922,
      "memory(GiB)": 77.56,
      "step": 82285,
      "token_acc": 0.5296442687747036,
      "train_speed(iter/s)": 1.437413
    },
    {
      "epoch": 3.525555888779401,
      "grad_norm": 6.64978551864624,
      "learning_rate": 1.996502506337326e-05,
      "loss": 2.5128606796264648,
      "memory(GiB)": 77.56,
      "step": 82290,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.437434
    },
    {
      "epoch": 3.52577010410865,
      "grad_norm": 6.265961170196533,
      "learning_rate": 1.9959645050826714e-05,
      "loss": 2.4377384185791016,
      "memory(GiB)": 77.56,
      "step": 82295,
      "token_acc": 0.4507936507936508,
      "train_speed(iter/s)": 1.437447
    },
    {
      "epoch": 3.525984319437899,
      "grad_norm": 5.836155414581299,
      "learning_rate": 1.995426558248938e-05,
      "loss": 2.300048828125,
      "memory(GiB)": 77.56,
      "step": 82300,
      "token_acc": 0.5127388535031847,
      "train_speed(iter/s)": 1.437446
    },
    {
      "epoch": 3.526198534767148,
      "grad_norm": 5.279250144958496,
      "learning_rate": 1.9948886658458727e-05,
      "loss": 2.5441802978515624,
      "memory(GiB)": 77.56,
      "step": 82305,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 3.526412750096397,
      "grad_norm": 4.949451446533203,
      "learning_rate": 1.9943508278832186e-05,
      "loss": 2.505910110473633,
      "memory(GiB)": 77.56,
      "step": 82310,
      "token_acc": 0.539622641509434,
      "train_speed(iter/s)": 1.437484
    },
    {
      "epoch": 3.526626965425646,
      "grad_norm": 5.963874340057373,
      "learning_rate": 1.993813044370718e-05,
      "loss": 2.241254425048828,
      "memory(GiB)": 77.56,
      "step": 82315,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.437478
    },
    {
      "epoch": 3.5268411807548947,
      "grad_norm": 5.687797546386719,
      "learning_rate": 1.993275315318115e-05,
      "loss": 2.353618049621582,
      "memory(GiB)": 77.56,
      "step": 82320,
      "token_acc": 0.528,
      "train_speed(iter/s)": 1.437477
    },
    {
      "epoch": 3.527055396084144,
      "grad_norm": 5.25728702545166,
      "learning_rate": 1.992737640735148e-05,
      "loss": 2.342526435852051,
      "memory(GiB)": 77.56,
      "step": 82325,
      "token_acc": 0.4823848238482385,
      "train_speed(iter/s)": 1.437463
    },
    {
      "epoch": 3.527269611413393,
      "grad_norm": 5.838143825531006,
      "learning_rate": 1.992200020631561e-05,
      "loss": 2.495411491394043,
      "memory(GiB)": 77.56,
      "step": 82330,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.437468
    },
    {
      "epoch": 3.5274838267426416,
      "grad_norm": 5.741201400756836,
      "learning_rate": 1.9916624550170926e-05,
      "loss": 2.5044696807861326,
      "memory(GiB)": 77.56,
      "step": 82335,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.437485
    },
    {
      "epoch": 3.527698042071891,
      "grad_norm": 5.947779655456543,
      "learning_rate": 1.9911249439014812e-05,
      "loss": 2.3775211334228517,
      "memory(GiB)": 77.56,
      "step": 82340,
      "token_acc": 0.4898785425101215,
      "train_speed(iter/s)": 1.437505
    },
    {
      "epoch": 3.5279122574011397,
      "grad_norm": 6.149830341339111,
      "learning_rate": 1.9905874872944635e-05,
      "loss": 2.2238487243652343,
      "memory(GiB)": 77.56,
      "step": 82345,
      "token_acc": 0.5222929936305732,
      "train_speed(iter/s)": 1.437503
    },
    {
      "epoch": 3.5281264727303885,
      "grad_norm": 5.631735801696777,
      "learning_rate": 1.9900500852057747e-05,
      "loss": 2.384516143798828,
      "memory(GiB)": 77.56,
      "step": 82350,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.437516
    },
    {
      "epoch": 3.5283406880596377,
      "grad_norm": 6.412612438201904,
      "learning_rate": 1.9895127376451544e-05,
      "loss": 2.508793258666992,
      "memory(GiB)": 77.56,
      "step": 82355,
      "token_acc": 0.4416058394160584,
      "train_speed(iter/s)": 1.437523
    },
    {
      "epoch": 3.5285549033888866,
      "grad_norm": 4.521615028381348,
      "learning_rate": 1.9889754446223348e-05,
      "loss": 2.44223575592041,
      "memory(GiB)": 77.56,
      "step": 82360,
      "token_acc": 0.4837662337662338,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 3.5287691187181354,
      "grad_norm": 6.081164360046387,
      "learning_rate": 1.9884382061470492e-05,
      "loss": 2.4649074554443358,
      "memory(GiB)": 77.56,
      "step": 82365,
      "token_acc": 0.5116959064327485,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 3.5289833340473846,
      "grad_norm": 4.48116397857666,
      "learning_rate": 1.9879010222290314e-05,
      "loss": 2.298491096496582,
      "memory(GiB)": 77.56,
      "step": 82370,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.437551
    },
    {
      "epoch": 3.5291975493766334,
      "grad_norm": 7.052316665649414,
      "learning_rate": 1.9873638928780093e-05,
      "loss": 2.519794464111328,
      "memory(GiB)": 77.56,
      "step": 82375,
      "token_acc": 0.4463087248322148,
      "train_speed(iter/s)": 1.437547
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 6.021188735961914,
      "learning_rate": 1.9868268181037185e-05,
      "loss": 2.10813045501709,
      "memory(GiB)": 77.56,
      "step": 82380,
      "token_acc": 0.5544217687074829,
      "train_speed(iter/s)": 1.437558
    },
    {
      "epoch": 3.5296259800351315,
      "grad_norm": 6.387473106384277,
      "learning_rate": 1.986289797915885e-05,
      "loss": 2.3520223617553713,
      "memory(GiB)": 77.56,
      "step": 82385,
      "token_acc": 0.5210727969348659,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 3.5298401953643803,
      "grad_norm": 5.794710159301758,
      "learning_rate": 1.9857528323242407e-05,
      "loss": 2.2839115142822264,
      "memory(GiB)": 77.56,
      "step": 82390,
      "token_acc": 0.4688427299703264,
      "train_speed(iter/s)": 1.437575
    },
    {
      "epoch": 3.530054410693629,
      "grad_norm": 6.127869129180908,
      "learning_rate": 1.9852159213385113e-05,
      "loss": 2.1248931884765625,
      "memory(GiB)": 77.56,
      "step": 82395,
      "token_acc": 0.5264705882352941,
      "train_speed(iter/s)": 1.437565
    },
    {
      "epoch": 3.5302686260228784,
      "grad_norm": 5.518728733062744,
      "learning_rate": 1.9846790649684232e-05,
      "loss": 2.242991638183594,
      "memory(GiB)": 77.56,
      "step": 82400,
      "token_acc": 0.47959183673469385,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 3.530482841352127,
      "grad_norm": 5.253847122192383,
      "learning_rate": 1.9841422632237026e-05,
      "loss": 2.307224082946777,
      "memory(GiB)": 77.56,
      "step": 82405,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 3.530697056681376,
      "grad_norm": 6.456474304199219,
      "learning_rate": 1.9836055161140725e-05,
      "loss": 2.311740684509277,
      "memory(GiB)": 77.56,
      "step": 82410,
      "token_acc": 0.4828767123287671,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 3.5309112720106253,
      "grad_norm": 4.971041679382324,
      "learning_rate": 1.9830688236492602e-05,
      "loss": 2.5282386779785155,
      "memory(GiB)": 77.56,
      "step": 82415,
      "token_acc": 0.4628975265017668,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 3.531125487339874,
      "grad_norm": 4.824924945831299,
      "learning_rate": 1.9825321858389855e-05,
      "loss": 2.323630142211914,
      "memory(GiB)": 77.56,
      "step": 82420,
      "token_acc": 0.5288135593220339,
      "train_speed(iter/s)": 1.4376
    },
    {
      "epoch": 3.531339702669123,
      "grad_norm": 6.92716646194458,
      "learning_rate": 1.9819956026929715e-05,
      "loss": 2.3047618865966797,
      "memory(GiB)": 77.56,
      "step": 82425,
      "token_acc": 0.4721189591078067,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 3.531553917998372,
      "grad_norm": 6.4586501121521,
      "learning_rate": 1.9814590742209382e-05,
      "loss": 2.255636215209961,
      "memory(GiB)": 77.56,
      "step": 82430,
      "token_acc": 0.5182926829268293,
      "train_speed(iter/s)": 1.437538
    },
    {
      "epoch": 3.531768133327621,
      "grad_norm": 6.5254225730896,
      "learning_rate": 1.9809226004326032e-05,
      "loss": 2.47855281829834,
      "memory(GiB)": 77.56,
      "step": 82435,
      "token_acc": 0.4708029197080292,
      "train_speed(iter/s)": 1.437563
    },
    {
      "epoch": 3.5319823486568698,
      "grad_norm": 5.294203758239746,
      "learning_rate": 1.9803861813376895e-05,
      "loss": 2.5271537780761717,
      "memory(GiB)": 77.56,
      "step": 82440,
      "token_acc": 0.517799352750809,
      "train_speed(iter/s)": 1.437575
    },
    {
      "epoch": 3.532196563986119,
      "grad_norm": 5.014577388763428,
      "learning_rate": 1.9798498169459124e-05,
      "loss": 2.2906585693359376,
      "memory(GiB)": 77.56,
      "step": 82445,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.437563
    },
    {
      "epoch": 3.532410779315368,
      "grad_norm": 5.592288494110107,
      "learning_rate": 1.9793135072669878e-05,
      "loss": 2.267938232421875,
      "memory(GiB)": 77.56,
      "step": 82450,
      "token_acc": 0.513595166163142,
      "train_speed(iter/s)": 1.437568
    },
    {
      "epoch": 3.5326249946446167,
      "grad_norm": 5.750865936279297,
      "learning_rate": 1.978777252310634e-05,
      "loss": 2.02744083404541,
      "memory(GiB)": 77.56,
      "step": 82455,
      "token_acc": 0.569023569023569,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 3.532839209973866,
      "grad_norm": 4.834346771240234,
      "learning_rate": 1.9782410520865648e-05,
      "loss": 2.5700956344604493,
      "memory(GiB)": 77.56,
      "step": 82460,
      "token_acc": 0.47896440129449835,
      "train_speed(iter/s)": 1.437588
    },
    {
      "epoch": 3.5330534253031147,
      "grad_norm": 7.49498987197876,
      "learning_rate": 1.977704906604493e-05,
      "loss": 2.2034053802490234,
      "memory(GiB)": 77.56,
      "step": 82465,
      "token_acc": 0.5179282868525896,
      "train_speed(iter/s)": 1.437588
    },
    {
      "epoch": 3.5332676406323635,
      "grad_norm": 7.677392959594727,
      "learning_rate": 1.9771688158741302e-05,
      "loss": 2.3965681076049803,
      "memory(GiB)": 77.56,
      "step": 82470,
      "token_acc": 0.5035211267605634,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 3.533481855961613,
      "grad_norm": 7.435286998748779,
      "learning_rate": 1.976632779905192e-05,
      "loss": 2.3592693328857424,
      "memory(GiB)": 77.56,
      "step": 82475,
      "token_acc": 0.50814332247557,
      "train_speed(iter/s)": 1.437606
    },
    {
      "epoch": 3.5336960712908616,
      "grad_norm": 5.908475875854492,
      "learning_rate": 1.9760967987073876e-05,
      "loss": 2.421270179748535,
      "memory(GiB)": 77.56,
      "step": 82480,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.437609
    },
    {
      "epoch": 3.5339102866201104,
      "grad_norm": 5.820932388305664,
      "learning_rate": 1.9755608722904256e-05,
      "loss": 2.4349483489990233,
      "memory(GiB)": 77.56,
      "step": 82485,
      "token_acc": 0.48220064724919093,
      "train_speed(iter/s)": 1.43763
    },
    {
      "epoch": 3.5341245019493597,
      "grad_norm": 6.203174114227295,
      "learning_rate": 1.975025000664016e-05,
      "loss": 2.3788665771484374,
      "memory(GiB)": 77.56,
      "step": 82490,
      "token_acc": 0.49390243902439024,
      "train_speed(iter/s)": 1.437646
    },
    {
      "epoch": 3.5343387172786085,
      "grad_norm": 6.152998924255371,
      "learning_rate": 1.974489183837864e-05,
      "loss": 2.567764472961426,
      "memory(GiB)": 77.56,
      "step": 82495,
      "token_acc": 0.46264367816091956,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 3.5345529326078573,
      "grad_norm": 5.205676078796387,
      "learning_rate": 1.9739534218216805e-05,
      "loss": 2.3460786819458006,
      "memory(GiB)": 77.56,
      "step": 82500,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.437646
    },
    {
      "epoch": 3.5345529326078573,
      "eval_loss": 2.1523923873901367,
      "eval_runtime": 14.0725,
      "eval_samples_per_second": 7.106,
      "eval_steps_per_second": 7.106,
      "eval_token_acc": 0.4811320754716981,
      "step": 82500
    },
    {
      "epoch": 3.5347671479371066,
      "grad_norm": 5.5499348640441895,
      "learning_rate": 1.97341771462517e-05,
      "loss": 2.32886962890625,
      "memory(GiB)": 77.56,
      "step": 82505,
      "token_acc": 0.4863849765258216,
      "train_speed(iter/s)": 1.437254
    },
    {
      "epoch": 3.5349813632663554,
      "grad_norm": 4.896663188934326,
      "learning_rate": 1.9728820622580358e-05,
      "loss": 2.1635631561279296,
      "memory(GiB)": 77.56,
      "step": 82510,
      "token_acc": 0.5202702702702703,
      "train_speed(iter/s)": 1.437244
    },
    {
      "epoch": 3.535195578595604,
      "grad_norm": 4.505125045776367,
      "learning_rate": 1.9723464647299834e-05,
      "loss": 2.2201520919799806,
      "memory(GiB)": 77.56,
      "step": 82515,
      "token_acc": 0.5467128027681661,
      "train_speed(iter/s)": 1.437243
    },
    {
      "epoch": 3.5354097939248534,
      "grad_norm": 5.725396633148193,
      "learning_rate": 1.9718109220507147e-05,
      "loss": 2.3150224685668945,
      "memory(GiB)": 77.56,
      "step": 82520,
      "token_acc": 0.4618181818181818,
      "train_speed(iter/s)": 1.437214
    },
    {
      "epoch": 3.5356240092541023,
      "grad_norm": 6.492303371429443,
      "learning_rate": 1.97127543422993e-05,
      "loss": 2.139881896972656,
      "memory(GiB)": 77.56,
      "step": 82525,
      "token_acc": 0.5355805243445693,
      "train_speed(iter/s)": 1.437233
    },
    {
      "epoch": 3.535838224583351,
      "grad_norm": 5.474704265594482,
      "learning_rate": 1.9707400012773318e-05,
      "loss": 2.3984148025512697,
      "memory(GiB)": 77.56,
      "step": 82530,
      "token_acc": 0.45980707395498394,
      "train_speed(iter/s)": 1.437238
    },
    {
      "epoch": 3.5360524399126003,
      "grad_norm": 4.704589366912842,
      "learning_rate": 1.9702046232026222e-05,
      "loss": 2.0160974502563476,
      "memory(GiB)": 77.56,
      "step": 82535,
      "token_acc": 0.5800711743772242,
      "train_speed(iter/s)": 1.437252
    },
    {
      "epoch": 3.536266655241849,
      "grad_norm": 5.991926670074463,
      "learning_rate": 1.969669300015498e-05,
      "loss": 2.3534507751464844,
      "memory(GiB)": 77.56,
      "step": 82540,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437267
    },
    {
      "epoch": 3.536480870571098,
      "grad_norm": 4.970750331878662,
      "learning_rate": 1.9691340317256575e-05,
      "loss": 2.43546142578125,
      "memory(GiB)": 77.56,
      "step": 82545,
      "token_acc": 0.4968553459119497,
      "train_speed(iter/s)": 1.437261
    },
    {
      "epoch": 3.536695085900347,
      "grad_norm": 5.680489540100098,
      "learning_rate": 1.9685988183427968e-05,
      "loss": 2.5124191284179687,
      "memory(GiB)": 77.56,
      "step": 82550,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.437278
    },
    {
      "epoch": 3.536909301229596,
      "grad_norm": 5.884768009185791,
      "learning_rate": 1.9680636598766104e-05,
      "loss": 2.324357604980469,
      "memory(GiB)": 77.56,
      "step": 82555,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.437277
    },
    {
      "epoch": 3.537123516558845,
      "grad_norm": 5.680827617645264,
      "learning_rate": 1.967528556336797e-05,
      "loss": 2.441228675842285,
      "memory(GiB)": 77.56,
      "step": 82560,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.437284
    },
    {
      "epoch": 3.537337731888094,
      "grad_norm": 5.874486446380615,
      "learning_rate": 1.966993507733048e-05,
      "loss": 2.417146682739258,
      "memory(GiB)": 77.56,
      "step": 82565,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.437294
    },
    {
      "epoch": 3.537551947217343,
      "grad_norm": 6.918670177459717,
      "learning_rate": 1.9664585140750574e-05,
      "loss": 2.1657217025756834,
      "memory(GiB)": 77.56,
      "step": 82570,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.4373
    },
    {
      "epoch": 3.5377661625465917,
      "grad_norm": 6.989270210266113,
      "learning_rate": 1.965923575372516e-05,
      "loss": 2.3373685836791993,
      "memory(GiB)": 77.56,
      "step": 82575,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.437319
    },
    {
      "epoch": 3.537980377875841,
      "grad_norm": 5.781280994415283,
      "learning_rate": 1.965388691635114e-05,
      "loss": 2.4048690795898438,
      "memory(GiB)": 77.56,
      "step": 82580,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.437311
    },
    {
      "epoch": 3.5381945932050898,
      "grad_norm": 5.102858066558838,
      "learning_rate": 1.964853862872544e-05,
      "loss": 2.4206228256225586,
      "memory(GiB)": 77.56,
      "step": 82585,
      "token_acc": 0.473972602739726,
      "train_speed(iter/s)": 1.437307
    },
    {
      "epoch": 3.5384088085343386,
      "grad_norm": 6.523634433746338,
      "learning_rate": 1.964319089094493e-05,
      "loss": 2.548038101196289,
      "memory(GiB)": 77.56,
      "step": 82590,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.437318
    },
    {
      "epoch": 3.538623023863588,
      "grad_norm": 5.074010848999023,
      "learning_rate": 1.9637843703106503e-05,
      "loss": 2.2360279083251955,
      "memory(GiB)": 77.56,
      "step": 82595,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.437333
    },
    {
      "epoch": 3.5388372391928367,
      "grad_norm": 5.488282680511475,
      "learning_rate": 1.9632497065306993e-05,
      "loss": 2.5630460739135743,
      "memory(GiB)": 77.56,
      "step": 82600,
      "token_acc": 0.4837662337662338,
      "train_speed(iter/s)": 1.437345
    },
    {
      "epoch": 3.5390514545220855,
      "grad_norm": 6.235357761383057,
      "learning_rate": 1.962715097764331e-05,
      "loss": 2.199538230895996,
      "memory(GiB)": 77.56,
      "step": 82605,
      "token_acc": 0.5067114093959731,
      "train_speed(iter/s)": 1.437353
    },
    {
      "epoch": 3.5392656698513347,
      "grad_norm": 4.771381855010986,
      "learning_rate": 1.9621805440212275e-05,
      "loss": 2.450325775146484,
      "memory(GiB)": 77.56,
      "step": 82610,
      "token_acc": 0.44072948328267475,
      "train_speed(iter/s)": 1.43734
    },
    {
      "epoch": 3.5394798851805835,
      "grad_norm": 5.8777384757995605,
      "learning_rate": 1.9616460453110712e-05,
      "loss": 2.311336898803711,
      "memory(GiB)": 77.56,
      "step": 82615,
      "token_acc": 0.5148514851485149,
      "train_speed(iter/s)": 1.437367
    },
    {
      "epoch": 3.5396941005098324,
      "grad_norm": 6.512287616729736,
      "learning_rate": 1.9611116016435495e-05,
      "loss": 2.3543508529663084,
      "memory(GiB)": 77.56,
      "step": 82620,
      "token_acc": 0.5381944444444444,
      "train_speed(iter/s)": 1.437383
    },
    {
      "epoch": 3.5399083158390816,
      "grad_norm": 7.980329990386963,
      "learning_rate": 1.9605772130283413e-05,
      "loss": 2.5231698989868163,
      "memory(GiB)": 77.56,
      "step": 82625,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 3.5401225311683304,
      "grad_norm": 6.829824924468994,
      "learning_rate": 1.9600428794751285e-05,
      "loss": 2.416186714172363,
      "memory(GiB)": 77.56,
      "step": 82630,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.437431
    },
    {
      "epoch": 3.5403367464975792,
      "grad_norm": 5.182572364807129,
      "learning_rate": 1.9595086009935903e-05,
      "loss": 2.279319953918457,
      "memory(GiB)": 77.56,
      "step": 82635,
      "token_acc": 0.5223367697594502,
      "train_speed(iter/s)": 1.437443
    },
    {
      "epoch": 3.5405509618268285,
      "grad_norm": 5.381372928619385,
      "learning_rate": 1.9589743775934043e-05,
      "loss": 2.0162782669067383,
      "memory(GiB)": 77.56,
      "step": 82640,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437441
    },
    {
      "epoch": 3.5407651771560773,
      "grad_norm": 6.481118679046631,
      "learning_rate": 1.958440209284252e-05,
      "loss": 2.0292457580566405,
      "memory(GiB)": 77.56,
      "step": 82645,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.43742
    },
    {
      "epoch": 3.540979392485326,
      "grad_norm": 6.066073417663574,
      "learning_rate": 1.9579060960758082e-05,
      "loss": 2.274923324584961,
      "memory(GiB)": 77.56,
      "step": 82650,
      "token_acc": 0.5326460481099656,
      "train_speed(iter/s)": 1.437429
    },
    {
      "epoch": 3.5411936078145754,
      "grad_norm": 6.3374457359313965,
      "learning_rate": 1.9573720379777498e-05,
      "loss": 2.2671525955200194,
      "memory(GiB)": 77.56,
      "step": 82655,
      "token_acc": 0.504950495049505,
      "train_speed(iter/s)": 1.43744
    },
    {
      "epoch": 3.541407823143824,
      "grad_norm": 5.057811260223389,
      "learning_rate": 1.9568380349997506e-05,
      "loss": 2.4789525985717775,
      "memory(GiB)": 77.56,
      "step": 82660,
      "token_acc": 0.48404255319148937,
      "train_speed(iter/s)": 1.437427
    },
    {
      "epoch": 3.541622038473073,
      "grad_norm": 7.5626935958862305,
      "learning_rate": 1.9563040871514854e-05,
      "loss": 2.2776498794555664,
      "memory(GiB)": 77.56,
      "step": 82665,
      "token_acc": 0.5236363636363637,
      "train_speed(iter/s)": 1.437438
    },
    {
      "epoch": 3.5418362538023223,
      "grad_norm": 6.010182857513428,
      "learning_rate": 1.955770194442625e-05,
      "loss": 2.372435760498047,
      "memory(GiB)": 77.56,
      "step": 82670,
      "token_acc": 0.5117845117845118,
      "train_speed(iter/s)": 1.437422
    },
    {
      "epoch": 3.542050469131571,
      "grad_norm": 6.353419780731201,
      "learning_rate": 1.9552363568828437e-05,
      "loss": 2.2901140213012696,
      "memory(GiB)": 77.56,
      "step": 82675,
      "token_acc": 0.5410447761194029,
      "train_speed(iter/s)": 1.437439
    },
    {
      "epoch": 3.54226468446082,
      "grad_norm": 6.403958320617676,
      "learning_rate": 1.954702574481813e-05,
      "loss": 2.299019622802734,
      "memory(GiB)": 77.56,
      "step": 82680,
      "token_acc": 0.5401459854014599,
      "train_speed(iter/s)": 1.437462
    },
    {
      "epoch": 3.542478899790069,
      "grad_norm": 7.3599629402160645,
      "learning_rate": 1.954168847249202e-05,
      "loss": 2.282818603515625,
      "memory(GiB)": 77.56,
      "step": 82685,
      "token_acc": 0.5229681978798587,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.542693115119318,
      "grad_norm": 5.285392761230469,
      "learning_rate": 1.95363517519468e-05,
      "loss": 2.588500213623047,
      "memory(GiB)": 77.56,
      "step": 82690,
      "token_acc": 0.486404833836858,
      "train_speed(iter/s)": 1.437461
    },
    {
      "epoch": 3.5429073304485668,
      "grad_norm": 5.888866424560547,
      "learning_rate": 1.9531015583279143e-05,
      "loss": 2.322518730163574,
      "memory(GiB)": 77.56,
      "step": 82695,
      "token_acc": 0.5047318611987381,
      "train_speed(iter/s)": 1.437478
    },
    {
      "epoch": 3.543121545777816,
      "grad_norm": 7.0269598960876465,
      "learning_rate": 1.9525679966585703e-05,
      "loss": 2.2573759078979494,
      "memory(GiB)": 77.56,
      "step": 82700,
      "token_acc": 0.49800796812749004,
      "train_speed(iter/s)": 1.437501
    },
    {
      "epoch": 3.543335761107065,
      "grad_norm": 6.0991597175598145,
      "learning_rate": 1.9520344901963177e-05,
      "loss": 2.4632055282592775,
      "memory(GiB)": 77.56,
      "step": 82705,
      "token_acc": 0.494949494949495,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 3.5435499764363136,
      "grad_norm": 6.6377129554748535,
      "learning_rate": 1.951501038950819e-05,
      "loss": 1.9929412841796874,
      "memory(GiB)": 77.56,
      "step": 82710,
      "token_acc": 0.5725806451612904,
      "train_speed(iter/s)": 1.437512
    },
    {
      "epoch": 3.543764191765563,
      "grad_norm": 6.9814982414245605,
      "learning_rate": 1.9509676429317397e-05,
      "loss": 2.2883739471435547,
      "memory(GiB)": 77.56,
      "step": 82715,
      "token_acc": 0.4713114754098361,
      "train_speed(iter/s)": 1.437511
    },
    {
      "epoch": 3.5439784070948117,
      "grad_norm": 6.554887771606445,
      "learning_rate": 1.950434302148741e-05,
      "loss": 2.4074459075927734,
      "memory(GiB)": 77.56,
      "step": 82720,
      "token_acc": 0.4892966360856269,
      "train_speed(iter/s)": 1.43751
    },
    {
      "epoch": 3.5441926224240605,
      "grad_norm": 6.756819725036621,
      "learning_rate": 1.9499010166114838e-05,
      "loss": 2.5532487869262694,
      "memory(GiB)": 77.56,
      "step": 82725,
      "token_acc": 0.49375,
      "train_speed(iter/s)": 1.437516
    },
    {
      "epoch": 3.54440683775331,
      "grad_norm": 8.334879875183105,
      "learning_rate": 1.9493677863296332e-05,
      "loss": 2.4358564376831056,
      "memory(GiB)": 77.56,
      "step": 82730,
      "token_acc": 0.5056603773584906,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 3.5446210530825586,
      "grad_norm": 7.20154333114624,
      "learning_rate": 1.948834611312847e-05,
      "loss": 2.0781314849853514,
      "memory(GiB)": 77.56,
      "step": 82735,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.437533
    },
    {
      "epoch": 3.5448352684118074,
      "grad_norm": 4.454391956329346,
      "learning_rate": 1.9483014915707832e-05,
      "loss": 2.2311704635620115,
      "memory(GiB)": 77.56,
      "step": 82740,
      "token_acc": 0.49393939393939396,
      "train_speed(iter/s)": 1.437523
    },
    {
      "epoch": 3.5450494837410567,
      "grad_norm": 6.827240467071533,
      "learning_rate": 1.9477684271130995e-05,
      "loss": 2.402065086364746,
      "memory(GiB)": 77.56,
      "step": 82745,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.437498
    },
    {
      "epoch": 3.5452636990703055,
      "grad_norm": 5.677723407745361,
      "learning_rate": 1.9472354179494556e-05,
      "loss": 2.230828857421875,
      "memory(GiB)": 77.56,
      "step": 82750,
      "token_acc": 0.4898785425101215,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 3.5454779143995543,
      "grad_norm": 6.276457786560059,
      "learning_rate": 1.9467024640895053e-05,
      "loss": 2.5080982208251954,
      "memory(GiB)": 77.56,
      "step": 82755,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 3.5456921297288035,
      "grad_norm": 5.828501224517822,
      "learning_rate": 1.946169565542903e-05,
      "loss": 2.380128288269043,
      "memory(GiB)": 77.56,
      "step": 82760,
      "token_acc": 0.504885993485342,
      "train_speed(iter/s)": 1.43751
    },
    {
      "epoch": 3.5459063450580524,
      "grad_norm": 6.4717254638671875,
      "learning_rate": 1.9456367223193055e-05,
      "loss": 2.505108451843262,
      "memory(GiB)": 77.56,
      "step": 82765,
      "token_acc": 0.43234323432343236,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 3.546120560387301,
      "grad_norm": 5.942768096923828,
      "learning_rate": 1.945103934428364e-05,
      "loss": 2.6547855377197265,
      "memory(GiB)": 77.56,
      "step": 82770,
      "token_acc": 0.44981412639405205,
      "train_speed(iter/s)": 1.437524
    },
    {
      "epoch": 3.5463347757165504,
      "grad_norm": 5.546896457672119,
      "learning_rate": 1.9445712018797296e-05,
      "loss": 2.2827091217041016,
      "memory(GiB)": 77.56,
      "step": 82775,
      "token_acc": 0.5204081632653061,
      "train_speed(iter/s)": 1.437497
    },
    {
      "epoch": 3.5465489910457992,
      "grad_norm": 6.897151947021484,
      "learning_rate": 1.944038524683055e-05,
      "loss": 2.5003940582275392,
      "memory(GiB)": 77.56,
      "step": 82780,
      "token_acc": 0.49709302325581395,
      "train_speed(iter/s)": 1.437497
    },
    {
      "epoch": 3.546763206375048,
      "grad_norm": 7.0494489669799805,
      "learning_rate": 1.9435059028479874e-05,
      "loss": 2.2949268341064455,
      "memory(GiB)": 77.56,
      "step": 82785,
      "token_acc": 0.5035714285714286,
      "train_speed(iter/s)": 1.437497
    },
    {
      "epoch": 3.5469774217042973,
      "grad_norm": 5.7212724685668945,
      "learning_rate": 1.942973336384179e-05,
      "loss": 2.2536026000976563,
      "memory(GiB)": 77.56,
      "step": 82790,
      "token_acc": 0.4918032786885246,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 3.547191637033546,
      "grad_norm": 5.892222881317139,
      "learning_rate": 1.9424408253012765e-05,
      "loss": 2.4641143798828127,
      "memory(GiB)": 77.56,
      "step": 82795,
      "token_acc": 0.4520123839009288,
      "train_speed(iter/s)": 1.437501
    },
    {
      "epoch": 3.547405852362795,
      "grad_norm": 5.773164749145508,
      "learning_rate": 1.941908369608927e-05,
      "loss": 2.5576282501220704,
      "memory(GiB)": 77.56,
      "step": 82800,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437506
    },
    {
      "epoch": 3.547620067692044,
      "grad_norm": 5.912968635559082,
      "learning_rate": 1.9413759693167754e-05,
      "loss": 2.131065559387207,
      "memory(GiB)": 77.56,
      "step": 82805,
      "token_acc": 0.5310344827586206,
      "train_speed(iter/s)": 1.437522
    },
    {
      "epoch": 3.547834283021293,
      "grad_norm": 5.046838283538818,
      "learning_rate": 1.940843624434468e-05,
      "loss": 2.562174606323242,
      "memory(GiB)": 77.56,
      "step": 82810,
      "token_acc": 0.504,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 3.548048498350542,
      "grad_norm": 8.35932731628418,
      "learning_rate": 1.940311334971646e-05,
      "loss": 2.10133056640625,
      "memory(GiB)": 77.56,
      "step": 82815,
      "token_acc": 0.54296875,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 3.548262713679791,
      "grad_norm": 7.073241233825684,
      "learning_rate": 1.9397791009379546e-05,
      "loss": 2.3052635192871094,
      "memory(GiB)": 77.56,
      "step": 82820,
      "token_acc": 0.5211267605633803,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 3.54847692900904,
      "grad_norm": 5.813384056091309,
      "learning_rate": 1.939246922343037e-05,
      "loss": 2.3398006439208983,
      "memory(GiB)": 77.56,
      "step": 82825,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 3.5486911443382887,
      "grad_norm": 9.540542602539062,
      "learning_rate": 1.9387147991965327e-05,
      "loss": 2.3147607803344727,
      "memory(GiB)": 77.56,
      "step": 82830,
      "token_acc": 0.5136363636363637,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 3.548905359667538,
      "grad_norm": 5.839265823364258,
      "learning_rate": 1.9381827315080815e-05,
      "loss": 2.501369857788086,
      "memory(GiB)": 77.56,
      "step": 82835,
      "token_acc": 0.44891640866873067,
      "train_speed(iter/s)": 1.437566
    },
    {
      "epoch": 3.5491195749967868,
      "grad_norm": 6.068883419036865,
      "learning_rate": 1.9376507192873223e-05,
      "loss": 2.134100341796875,
      "memory(GiB)": 77.56,
      "step": 82840,
      "token_acc": 0.5419354838709678,
      "train_speed(iter/s)": 1.437555
    },
    {
      "epoch": 3.5493337903260356,
      "grad_norm": 6.481598854064941,
      "learning_rate": 1.9371187625438913e-05,
      "loss": 2.6785915374755858,
      "memory(GiB)": 77.56,
      "step": 82845,
      "token_acc": 0.4845679012345679,
      "train_speed(iter/s)": 1.437573
    },
    {
      "epoch": 3.549548005655285,
      "grad_norm": 6.411945343017578,
      "learning_rate": 1.9365868612874287e-05,
      "loss": 2.175364112854004,
      "memory(GiB)": 77.56,
      "step": 82850,
      "token_acc": 0.5524475524475524,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 3.5497622209845336,
      "grad_norm": 7.459131717681885,
      "learning_rate": 1.9360550155275693e-05,
      "loss": 2.4688262939453125,
      "memory(GiB)": 77.56,
      "step": 82855,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.437596
    },
    {
      "epoch": 3.5499764363137825,
      "grad_norm": 6.617015838623047,
      "learning_rate": 1.9355232252739474e-05,
      "loss": 2.2800247192382814,
      "memory(GiB)": 77.56,
      "step": 82860,
      "token_acc": 0.5188679245283019,
      "train_speed(iter/s)": 1.437617
    },
    {
      "epoch": 3.5501906516430317,
      "grad_norm": 6.170217514038086,
      "learning_rate": 1.9349914905361964e-05,
      "loss": 2.2226984024047853,
      "memory(GiB)": 77.56,
      "step": 82865,
      "token_acc": 0.4835164835164835,
      "train_speed(iter/s)": 1.437612
    },
    {
      "epoch": 3.5504048669722805,
      "grad_norm": 6.286918640136719,
      "learning_rate": 1.9344598113239488e-05,
      "loss": 2.3269222259521483,
      "memory(GiB)": 77.56,
      "step": 82870,
      "token_acc": 0.4673913043478261,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 3.5506190823015293,
      "grad_norm": 8.006568908691406,
      "learning_rate": 1.933928187646839e-05,
      "loss": 2.300686264038086,
      "memory(GiB)": 77.56,
      "step": 82875,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.437618
    },
    {
      "epoch": 3.5508332976307786,
      "grad_norm": 5.5448503494262695,
      "learning_rate": 1.9333966195144958e-05,
      "loss": 2.3359729766845705,
      "memory(GiB)": 77.56,
      "step": 82880,
      "token_acc": 0.48014440433212996,
      "train_speed(iter/s)": 1.43761
    },
    {
      "epoch": 3.5510475129600274,
      "grad_norm": 6.8241448402404785,
      "learning_rate": 1.93286510693655e-05,
      "loss": 2.128226470947266,
      "memory(GiB)": 77.56,
      "step": 82885,
      "token_acc": 0.515850144092219,
      "train_speed(iter/s)": 1.437582
    },
    {
      "epoch": 3.551261728289276,
      "grad_norm": 4.49894905090332,
      "learning_rate": 1.93233364992263e-05,
      "loss": 2.339345169067383,
      "memory(GiB)": 77.56,
      "step": 82890,
      "token_acc": 0.4983388704318937,
      "train_speed(iter/s)": 1.437549
    },
    {
      "epoch": 3.5514759436185255,
      "grad_norm": 6.002151966094971,
      "learning_rate": 1.9318022484823618e-05,
      "loss": 2.3803098678588865,
      "memory(GiB)": 77.56,
      "step": 82895,
      "token_acc": 0.5225563909774437,
      "train_speed(iter/s)": 1.437557
    },
    {
      "epoch": 3.5516901589477743,
      "grad_norm": 7.201699733734131,
      "learning_rate": 1.9312709026253756e-05,
      "loss": 2.199318695068359,
      "memory(GiB)": 77.56,
      "step": 82900,
      "token_acc": 0.5444839857651246,
      "train_speed(iter/s)": 1.437548
    },
    {
      "epoch": 3.551904374277023,
      "grad_norm": 9.999852180480957,
      "learning_rate": 1.9307396123612942e-05,
      "loss": 2.2409217834472654,
      "memory(GiB)": 77.56,
      "step": 82905,
      "token_acc": 0.5773584905660377,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 3.5521185896062724,
      "grad_norm": 7.5008673667907715,
      "learning_rate": 1.9302083776997454e-05,
      "loss": 2.323116111755371,
      "memory(GiB)": 77.56,
      "step": 82910,
      "token_acc": 0.5437262357414449,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 3.552332804935521,
      "grad_norm": 6.727487564086914,
      "learning_rate": 1.929677198650352e-05,
      "loss": 2.3816705703735352,
      "memory(GiB)": 77.56,
      "step": 82915,
      "token_acc": 0.48757763975155277,
      "train_speed(iter/s)": 1.437565
    },
    {
      "epoch": 3.55254702026477,
      "grad_norm": 8.107377052307129,
      "learning_rate": 1.929146075222736e-05,
      "loss": 2.2722057342529296,
      "memory(GiB)": 77.56,
      "step": 82920,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 3.5527612355940192,
      "grad_norm": 5.84694242477417,
      "learning_rate": 1.9286150074265198e-05,
      "loss": 2.4540891647338867,
      "memory(GiB)": 77.56,
      "step": 82925,
      "token_acc": 0.4557823129251701,
      "train_speed(iter/s)": 1.437598
    },
    {
      "epoch": 3.552975450923268,
      "grad_norm": 6.92732048034668,
      "learning_rate": 1.928083995271322e-05,
      "loss": 2.3148471832275392,
      "memory(GiB)": 77.56,
      "step": 82930,
      "token_acc": 0.528125,
      "train_speed(iter/s)": 1.437596
    },
    {
      "epoch": 3.553189666252517,
      "grad_norm": 7.589896202087402,
      "learning_rate": 1.9275530387667655e-05,
      "loss": 2.4585365295410155,
      "memory(GiB)": 77.56,
      "step": 82935,
      "token_acc": 0.4582043343653251,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 3.553403881581766,
      "grad_norm": 5.6032490730285645,
      "learning_rate": 1.927022137922469e-05,
      "loss": 2.193222427368164,
      "memory(GiB)": 77.56,
      "step": 82940,
      "token_acc": 0.5228758169934641,
      "train_speed(iter/s)": 1.43763
    },
    {
      "epoch": 3.553618096911015,
      "grad_norm": 6.023399353027344,
      "learning_rate": 1.9264912927480483e-05,
      "loss": 2.1307304382324217,
      "memory(GiB)": 77.56,
      "step": 82945,
      "token_acc": 0.5463917525773195,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 3.5538323122402637,
      "grad_norm": 6.18001651763916,
      "learning_rate": 1.9259605032531213e-05,
      "loss": 2.3222259521484374,
      "memory(GiB)": 77.56,
      "step": 82950,
      "token_acc": 0.460431654676259,
      "train_speed(iter/s)": 1.43765
    },
    {
      "epoch": 3.554046527569513,
      "grad_norm": 7.535070896148682,
      "learning_rate": 1.9254297694473038e-05,
      "loss": 2.488578224182129,
      "memory(GiB)": 77.56,
      "step": 82955,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.437676
    },
    {
      "epoch": 3.554260742898762,
      "grad_norm": 5.162063121795654,
      "learning_rate": 1.9248990913402083e-05,
      "loss": 2.250819206237793,
      "memory(GiB)": 77.56,
      "step": 82960,
      "token_acc": 0.5239852398523985,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 3.5544749582280106,
      "grad_norm": 7.015691757202148,
      "learning_rate": 1.9243684689414498e-05,
      "loss": 2.3662088394165037,
      "memory(GiB)": 77.56,
      "step": 82965,
      "token_acc": 0.5347222222222222,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 3.55468917355726,
      "grad_norm": 6.708771705627441,
      "learning_rate": 1.923837902260644e-05,
      "loss": 2.3405385971069337,
      "memory(GiB)": 77.56,
      "step": 82970,
      "token_acc": 0.47592067988668557,
      "train_speed(iter/s)": 1.437698
    },
    {
      "epoch": 3.5549033888865087,
      "grad_norm": 5.543859004974365,
      "learning_rate": 1.9233073913074002e-05,
      "loss": 2.3417749404907227,
      "memory(GiB)": 77.56,
      "step": 82975,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.437708
    },
    {
      "epoch": 3.5551176042157575,
      "grad_norm": 6.479108810424805,
      "learning_rate": 1.9227769360913296e-05,
      "loss": 2.2098255157470703,
      "memory(GiB)": 77.56,
      "step": 82980,
      "token_acc": 0.5317725752508361,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 3.5553318195450068,
      "grad_norm": 7.023055076599121,
      "learning_rate": 1.9222465366220406e-05,
      "loss": 2.3799482345581056,
      "memory(GiB)": 77.56,
      "step": 82985,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 3.5555460348742556,
      "grad_norm": 6.113099575042725,
      "learning_rate": 1.9217161929091415e-05,
      "loss": 2.3580018997192385,
      "memory(GiB)": 77.56,
      "step": 82990,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 3.5557602502035044,
      "grad_norm": 6.232082843780518,
      "learning_rate": 1.921185904962243e-05,
      "loss": 2.260558319091797,
      "memory(GiB)": 77.56,
      "step": 82995,
      "token_acc": 0.5028901734104047,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.5559744655327536,
      "grad_norm": 6.2291951179504395,
      "learning_rate": 1.9206556727909496e-05,
      "loss": 2.466316795349121,
      "memory(GiB)": 77.56,
      "step": 83000,
      "token_acc": 0.49795918367346936,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 3.5559744655327536,
      "eval_loss": 2.24786376953125,
      "eval_runtime": 14.5619,
      "eval_samples_per_second": 6.867,
      "eval_steps_per_second": 6.867,
      "eval_token_acc": 0.48221343873517786,
      "step": 83000
    },
    {
      "epoch": 3.5561886808620025,
      "grad_norm": 5.826775074005127,
      "learning_rate": 1.9201254964048675e-05,
      "loss": 2.264176940917969,
      "memory(GiB)": 77.56,
      "step": 83005,
      "token_acc": 0.48930232558139536,
      "train_speed(iter/s)": 1.437341
    },
    {
      "epoch": 3.5564028961912513,
      "grad_norm": 5.459835052490234,
      "learning_rate": 1.9195953758136015e-05,
      "loss": 2.2328216552734377,
      "memory(GiB)": 77.56,
      "step": 83010,
      "token_acc": 0.48928571428571427,
      "train_speed(iter/s)": 1.437346
    },
    {
      "epoch": 3.5566171115205005,
      "grad_norm": 7.009839057922363,
      "learning_rate": 1.9190653110267532e-05,
      "loss": 2.614377021789551,
      "memory(GiB)": 77.56,
      "step": 83015,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.437352
    },
    {
      "epoch": 3.5568313268497493,
      "grad_norm": 5.47695779800415,
      "learning_rate": 1.9185353020539288e-05,
      "loss": 1.952031707763672,
      "memory(GiB)": 77.56,
      "step": 83020,
      "token_acc": 0.5724907063197026,
      "train_speed(iter/s)": 1.437365
    },
    {
      "epoch": 3.557045542178998,
      "grad_norm": 4.556182861328125,
      "learning_rate": 1.9180053489047273e-05,
      "loss": 2.1064924240112304,
      "memory(GiB)": 77.56,
      "step": 83025,
      "token_acc": 0.5421686746987951,
      "train_speed(iter/s)": 1.437383
    },
    {
      "epoch": 3.5572597575082474,
      "grad_norm": 5.167880535125732,
      "learning_rate": 1.9174754515887506e-05,
      "loss": 2.242349624633789,
      "memory(GiB)": 77.56,
      "step": 83030,
      "token_acc": 0.49700598802395207,
      "train_speed(iter/s)": 1.4374
    },
    {
      "epoch": 3.5574739728374962,
      "grad_norm": 5.98271369934082,
      "learning_rate": 1.9169456101155974e-05,
      "loss": 2.3210241317749025,
      "memory(GiB)": 77.56,
      "step": 83035,
      "token_acc": 0.5031446540880503,
      "train_speed(iter/s)": 1.437396
    },
    {
      "epoch": 3.557688188166745,
      "grad_norm": 4.825085639953613,
      "learning_rate": 1.9164158244948654e-05,
      "loss": 2.364198684692383,
      "memory(GiB)": 77.56,
      "step": 83040,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437402
    },
    {
      "epoch": 3.5579024034959943,
      "grad_norm": 5.873676776885986,
      "learning_rate": 1.9158860947361545e-05,
      "loss": 2.426919937133789,
      "memory(GiB)": 77.56,
      "step": 83045,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.43738
    },
    {
      "epoch": 3.558116618825243,
      "grad_norm": 5.452817440032959,
      "learning_rate": 1.9153564208490588e-05,
      "loss": 2.241208076477051,
      "memory(GiB)": 77.56,
      "step": 83050,
      "token_acc": 0.5365853658536586,
      "train_speed(iter/s)": 1.437389
    },
    {
      "epoch": 3.558330834154492,
      "grad_norm": 7.726084232330322,
      "learning_rate": 1.9148268028431772e-05,
      "loss": 2.145909881591797,
      "memory(GiB)": 77.56,
      "step": 83055,
      "token_acc": 0.5589519650655022,
      "train_speed(iter/s)": 1.437387
    },
    {
      "epoch": 3.558545049483741,
      "grad_norm": 6.4745869636535645,
      "learning_rate": 1.9142972407281017e-05,
      "loss": 2.3085563659667967,
      "memory(GiB)": 77.56,
      "step": 83060,
      "token_acc": 0.5483870967741935,
      "train_speed(iter/s)": 1.437409
    },
    {
      "epoch": 3.55875926481299,
      "grad_norm": 5.735274314880371,
      "learning_rate": 1.9137677345134262e-05,
      "loss": 2.128959846496582,
      "memory(GiB)": 77.56,
      "step": 83065,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.437412
    },
    {
      "epoch": 3.558973480142239,
      "grad_norm": 6.817280292510986,
      "learning_rate": 1.9132382842087438e-05,
      "loss": 2.519239044189453,
      "memory(GiB)": 77.56,
      "step": 83070,
      "token_acc": 0.4491525423728814,
      "train_speed(iter/s)": 1.437408
    },
    {
      "epoch": 3.559187695471488,
      "grad_norm": 6.035449981689453,
      "learning_rate": 1.9127088898236434e-05,
      "loss": 2.375954818725586,
      "memory(GiB)": 77.56,
      "step": 83075,
      "token_acc": 0.49693251533742333,
      "train_speed(iter/s)": 1.437389
    },
    {
      "epoch": 3.559401910800737,
      "grad_norm": 5.832775592803955,
      "learning_rate": 1.912179551367719e-05,
      "loss": 2.079724884033203,
      "memory(GiB)": 77.56,
      "step": 83080,
      "token_acc": 0.48535564853556484,
      "train_speed(iter/s)": 1.437385
    },
    {
      "epoch": 3.5596161261299857,
      "grad_norm": 6.4994049072265625,
      "learning_rate": 1.9116502688505593e-05,
      "loss": 2.344584655761719,
      "memory(GiB)": 77.56,
      "step": 83085,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.437393
    },
    {
      "epoch": 3.559830341459235,
      "grad_norm": 8.411913871765137,
      "learning_rate": 1.911121042281752e-05,
      "loss": 2.6202192306518555,
      "memory(GiB)": 77.56,
      "step": 83090,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.437398
    },
    {
      "epoch": 3.5600445567884837,
      "grad_norm": 7.736565589904785,
      "learning_rate": 1.9105918716708843e-05,
      "loss": 2.2053459167480467,
      "memory(GiB)": 77.56,
      "step": 83095,
      "token_acc": 0.5505226480836237,
      "train_speed(iter/s)": 1.437434
    },
    {
      "epoch": 3.5602587721177326,
      "grad_norm": 7.399161338806152,
      "learning_rate": 1.910062757027543e-05,
      "loss": 2.0944755554199217,
      "memory(GiB)": 77.56,
      "step": 83100,
      "token_acc": 0.5578947368421052,
      "train_speed(iter/s)": 1.437452
    },
    {
      "epoch": 3.560472987446982,
      "grad_norm": 5.091325283050537,
      "learning_rate": 1.9095336983613122e-05,
      "loss": 2.383116912841797,
      "memory(GiB)": 77.56,
      "step": 83105,
      "token_acc": 0.5016611295681063,
      "train_speed(iter/s)": 1.437479
    },
    {
      "epoch": 3.5606872027762306,
      "grad_norm": 6.804299354553223,
      "learning_rate": 1.9090046956817786e-05,
      "loss": 2.755426216125488,
      "memory(GiB)": 77.56,
      "step": 83110,
      "token_acc": 0.4420731707317073,
      "train_speed(iter/s)": 1.437483
    },
    {
      "epoch": 3.5609014181054794,
      "grad_norm": 8.014354705810547,
      "learning_rate": 1.908475748998523e-05,
      "loss": 2.2549163818359377,
      "memory(GiB)": 77.56,
      "step": 83115,
      "token_acc": 0.5130434782608696,
      "train_speed(iter/s)": 1.437469
    },
    {
      "epoch": 3.5611156334347287,
      "grad_norm": 6.086384296417236,
      "learning_rate": 1.9079468583211314e-05,
      "loss": 2.5265478134155273,
      "memory(GiB)": 77.56,
      "step": 83120,
      "token_acc": 0.4745222929936306,
      "train_speed(iter/s)": 1.437464
    },
    {
      "epoch": 3.5613298487639775,
      "grad_norm": 5.863791465759277,
      "learning_rate": 1.9074180236591826e-05,
      "loss": 2.9002330780029295,
      "memory(GiB)": 77.56,
      "step": 83125,
      "token_acc": 0.42724458204334365,
      "train_speed(iter/s)": 1.437469
    },
    {
      "epoch": 3.5615440640932263,
      "grad_norm": 7.259562969207764,
      "learning_rate": 1.9068892450222577e-05,
      "loss": 2.3109451293945313,
      "memory(GiB)": 77.56,
      "step": 83130,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.437488
    },
    {
      "epoch": 3.5617582794224756,
      "grad_norm": 5.618175506591797,
      "learning_rate": 1.9063605224199333e-05,
      "loss": 2.2776432037353516,
      "memory(GiB)": 77.56,
      "step": 83135,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437501
    },
    {
      "epoch": 3.5619724947517244,
      "grad_norm": 5.945962905883789,
      "learning_rate": 1.9058318558617927e-05,
      "loss": 2.3846551895141603,
      "memory(GiB)": 77.56,
      "step": 83140,
      "token_acc": 0.4690265486725664,
      "train_speed(iter/s)": 1.437501
    },
    {
      "epoch": 3.562186710080973,
      "grad_norm": 7.259668350219727,
      "learning_rate": 1.9053032453574098e-05,
      "loss": 2.5611623764038085,
      "memory(GiB)": 77.56,
      "step": 83145,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.43751
    },
    {
      "epoch": 3.5624009254102225,
      "grad_norm": 6.636059761047363,
      "learning_rate": 1.9047746909163623e-05,
      "loss": 2.2019626617431642,
      "memory(GiB)": 77.56,
      "step": 83150,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 3.5626151407394713,
      "grad_norm": 7.4019975662231445,
      "learning_rate": 1.9042461925482248e-05,
      "loss": 2.182967948913574,
      "memory(GiB)": 77.56,
      "step": 83155,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 3.56282935606872,
      "grad_norm": 6.843671798706055,
      "learning_rate": 1.9037177502625697e-05,
      "loss": 2.595068359375,
      "memory(GiB)": 77.56,
      "step": 83160,
      "token_acc": 0.4605263157894737,
      "train_speed(iter/s)": 1.437529
    },
    {
      "epoch": 3.5630435713979693,
      "grad_norm": 6.6064453125,
      "learning_rate": 1.9031893640689735e-05,
      "loss": 2.3750823974609374,
      "memory(GiB)": 77.56,
      "step": 83165,
      "token_acc": 0.4983922829581994,
      "train_speed(iter/s)": 1.437535
    },
    {
      "epoch": 3.563257786727218,
      "grad_norm": 6.949801921844482,
      "learning_rate": 1.9026610339770074e-05,
      "loss": 2.5514530181884765,
      "memory(GiB)": 77.56,
      "step": 83170,
      "token_acc": 0.49226006191950467,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 3.563472002056467,
      "grad_norm": 4.747146129608154,
      "learning_rate": 1.902132759996242e-05,
      "loss": 2.1096385955810546,
      "memory(GiB)": 77.56,
      "step": 83175,
      "token_acc": 0.5230263157894737,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 3.5636862173857162,
      "grad_norm": 6.8741936683654785,
      "learning_rate": 1.9016045421362478e-05,
      "loss": 2.374500846862793,
      "memory(GiB)": 77.56,
      "step": 83180,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 3.563900432714965,
      "grad_norm": 5.954809188842773,
      "learning_rate": 1.901076380406594e-05,
      "loss": 2.0965930938720705,
      "memory(GiB)": 77.56,
      "step": 83185,
      "token_acc": 0.5575539568345323,
      "train_speed(iter/s)": 1.43755
    },
    {
      "epoch": 3.564114648044214,
      "grad_norm": 6.052725315093994,
      "learning_rate": 1.9005482748168468e-05,
      "loss": 2.3403459548950196,
      "memory(GiB)": 77.56,
      "step": 83190,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.437553
    },
    {
      "epoch": 3.564328863373463,
      "grad_norm": 4.475099563598633,
      "learning_rate": 1.900020225376575e-05,
      "loss": 2.392584228515625,
      "memory(GiB)": 77.56,
      "step": 83195,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.437541
    },
    {
      "epoch": 3.564543078702712,
      "grad_norm": 6.442307472229004,
      "learning_rate": 1.8994922320953462e-05,
      "loss": 2.0630943298339846,
      "memory(GiB)": 77.56,
      "step": 83200,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.43755
    },
    {
      "epoch": 3.5647572940319607,
      "grad_norm": 6.023510456085205,
      "learning_rate": 1.8989642949827246e-05,
      "loss": 2.303834915161133,
      "memory(GiB)": 77.56,
      "step": 83205,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 3.56497150936121,
      "grad_norm": 5.839447975158691,
      "learning_rate": 1.898436414048274e-05,
      "loss": 2.0025238037109374,
      "memory(GiB)": 77.56,
      "step": 83210,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.437541
    },
    {
      "epoch": 3.565185724690459,
      "grad_norm": 5.560582160949707,
      "learning_rate": 1.897908589301557e-05,
      "loss": 2.419338035583496,
      "memory(GiB)": 77.56,
      "step": 83215,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.437561
    },
    {
      "epoch": 3.5653999400197076,
      "grad_norm": 5.170771598815918,
      "learning_rate": 1.8973808207521344e-05,
      "loss": 2.455280876159668,
      "memory(GiB)": 77.56,
      "step": 83220,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 3.565614155348957,
      "grad_norm": 6.187577724456787,
      "learning_rate": 1.89685310840957e-05,
      "loss": 2.3973594665527345,
      "memory(GiB)": 77.56,
      "step": 83225,
      "token_acc": 0.4985835694050991,
      "train_speed(iter/s)": 1.437586
    },
    {
      "epoch": 3.5658283706782057,
      "grad_norm": 6.095307350158691,
      "learning_rate": 1.8963254522834227e-05,
      "loss": 1.9889350891113282,
      "memory(GiB)": 77.56,
      "step": 83230,
      "token_acc": 0.5871559633027523,
      "train_speed(iter/s)": 1.437604
    },
    {
      "epoch": 3.5660425860074545,
      "grad_norm": 6.330908298492432,
      "learning_rate": 1.8957978523832514e-05,
      "loss": 2.834647369384766,
      "memory(GiB)": 77.56,
      "step": 83235,
      "token_acc": 0.42765273311897106,
      "train_speed(iter/s)": 1.437598
    },
    {
      "epoch": 3.5662568013367038,
      "grad_norm": 5.0780181884765625,
      "learning_rate": 1.8952703087186142e-05,
      "loss": 2.5584503173828126,
      "memory(GiB)": 77.56,
      "step": 83240,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.437591
    },
    {
      "epoch": 3.5664710166659526,
      "grad_norm": 5.6616950035095215,
      "learning_rate": 1.894742821299067e-05,
      "loss": 2.273427200317383,
      "memory(GiB)": 77.56,
      "step": 83245,
      "token_acc": 0.5174603174603175,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 3.5666852319952014,
      "grad_norm": 6.130789279937744,
      "learning_rate": 1.894215390134166e-05,
      "loss": 2.090376281738281,
      "memory(GiB)": 77.56,
      "step": 83250,
      "token_acc": 0.5387453874538746,
      "train_speed(iter/s)": 1.437605
    },
    {
      "epoch": 3.5668994473244506,
      "grad_norm": 7.444840431213379,
      "learning_rate": 1.8936880152334674e-05,
      "loss": 2.194511604309082,
      "memory(GiB)": 77.56,
      "step": 83255,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 3.5671136626536994,
      "grad_norm": 5.143515110015869,
      "learning_rate": 1.8931606966065247e-05,
      "loss": 2.379649543762207,
      "memory(GiB)": 77.56,
      "step": 83260,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 3.5673278779829483,
      "grad_norm": 8.019730567932129,
      "learning_rate": 1.8926334342628888e-05,
      "loss": 2.353934478759766,
      "memory(GiB)": 77.56,
      "step": 83265,
      "token_acc": 0.55859375,
      "train_speed(iter/s)": 1.437633
    },
    {
      "epoch": 3.5675420933121975,
      "grad_norm": 5.699904918670654,
      "learning_rate": 1.892106228212115e-05,
      "loss": 2.3754390716552733,
      "memory(GiB)": 77.56,
      "step": 83270,
      "token_acc": 0.4618181818181818,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 3.5677563086414463,
      "grad_norm": 4.760214805603027,
      "learning_rate": 1.891579078463752e-05,
      "loss": 2.3147865295410157,
      "memory(GiB)": 77.56,
      "step": 83275,
      "token_acc": 0.5291666666666667,
      "train_speed(iter/s)": 1.437642
    },
    {
      "epoch": 3.567970523970695,
      "grad_norm": 6.6610541343688965,
      "learning_rate": 1.8910519850273485e-05,
      "loss": 2.324036407470703,
      "memory(GiB)": 77.56,
      "step": 83280,
      "token_acc": 0.5157894736842106,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 3.5681847392999444,
      "grad_norm": 5.243537425994873,
      "learning_rate": 1.8905249479124563e-05,
      "loss": 2.440631866455078,
      "memory(GiB)": 77.56,
      "step": 83285,
      "token_acc": 0.4498567335243553,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 3.568398954629193,
      "grad_norm": 6.568016052246094,
      "learning_rate": 1.8899979671286216e-05,
      "loss": 2.2847660064697264,
      "memory(GiB)": 77.56,
      "step": 83290,
      "token_acc": 0.5149501661129569,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 3.568613169958442,
      "grad_norm": 6.597051620483398,
      "learning_rate": 1.8894710426853913e-05,
      "loss": 2.22582893371582,
      "memory(GiB)": 77.56,
      "step": 83295,
      "token_acc": 0.49814126394052044,
      "train_speed(iter/s)": 1.437651
    },
    {
      "epoch": 3.5688273852876913,
      "grad_norm": 6.890224933624268,
      "learning_rate": 1.888944174592311e-05,
      "loss": 2.4346603393554687,
      "memory(GiB)": 77.56,
      "step": 83300,
      "token_acc": 0.484593837535014,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 3.56904160061694,
      "grad_norm": 5.900251865386963,
      "learning_rate": 1.8884173628589236e-05,
      "loss": 2.282183074951172,
      "memory(GiB)": 77.56,
      "step": 83305,
      "token_acc": 0.5055762081784386,
      "train_speed(iter/s)": 1.437659
    },
    {
      "epoch": 3.569255815946189,
      "grad_norm": 7.65071439743042,
      "learning_rate": 1.887890607494776e-05,
      "loss": 2.2001413345336913,
      "memory(GiB)": 77.56,
      "step": 83310,
      "token_acc": 0.4915254237288136,
      "train_speed(iter/s)": 1.437652
    },
    {
      "epoch": 3.569470031275438,
      "grad_norm": 6.551050662994385,
      "learning_rate": 1.8873639085094097e-05,
      "loss": 2.1010459899902343,
      "memory(GiB)": 77.56,
      "step": 83315,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.437672
    },
    {
      "epoch": 3.569684246604687,
      "grad_norm": 7.86661958694458,
      "learning_rate": 1.8868372659123655e-05,
      "loss": 2.231205368041992,
      "memory(GiB)": 77.56,
      "step": 83320,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 3.569898461933936,
      "grad_norm": 6.6978535652160645,
      "learning_rate": 1.8863106797131857e-05,
      "loss": 2.0611236572265623,
      "memory(GiB)": 77.56,
      "step": 83325,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.437682
    },
    {
      "epoch": 3.570112677263185,
      "grad_norm": 7.841678619384766,
      "learning_rate": 1.885784149921408e-05,
      "loss": 2.247226333618164,
      "memory(GiB)": 77.56,
      "step": 83330,
      "token_acc": 0.5543071161048689,
      "train_speed(iter/s)": 1.437697
    },
    {
      "epoch": 3.570326892592434,
      "grad_norm": 8.14116096496582,
      "learning_rate": 1.88525767654657e-05,
      "loss": 2.122659111022949,
      "memory(GiB)": 77.56,
      "step": 83335,
      "token_acc": 0.5230263157894737,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 3.5705411079216827,
      "grad_norm": 6.669203281402588,
      "learning_rate": 1.884731259598212e-05,
      "loss": 2.4635011672973635,
      "memory(GiB)": 77.56,
      "step": 83340,
      "token_acc": 0.4625,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 3.570755323250932,
      "grad_norm": 6.690664768218994,
      "learning_rate": 1.8842048990858706e-05,
      "loss": 2.5491796493530274,
      "memory(GiB)": 77.56,
      "step": 83345,
      "token_acc": 0.45390070921985815,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 3.5709695385801807,
      "grad_norm": 7.593526363372803,
      "learning_rate": 1.8836785950190804e-05,
      "loss": 2.4520469665527345,
      "memory(GiB)": 77.56,
      "step": 83350,
      "token_acc": 0.4723926380368098,
      "train_speed(iter/s)": 1.437707
    },
    {
      "epoch": 3.5711837539094295,
      "grad_norm": 5.425413131713867,
      "learning_rate": 1.883152347407376e-05,
      "loss": 2.4824970245361326,
      "memory(GiB)": 77.56,
      "step": 83355,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.43771
    },
    {
      "epoch": 3.571397969238679,
      "grad_norm": 7.038394927978516,
      "learning_rate": 1.8826261562602905e-05,
      "loss": 2.2883943557739257,
      "memory(GiB)": 77.56,
      "step": 83360,
      "token_acc": 0.48188405797101447,
      "train_speed(iter/s)": 1.437717
    },
    {
      "epoch": 3.5716121845679276,
      "grad_norm": 5.130452632904053,
      "learning_rate": 1.8821000215873548e-05,
      "loss": 2.492097282409668,
      "memory(GiB)": 77.56,
      "step": 83365,
      "token_acc": 0.4770408163265306,
      "train_speed(iter/s)": 1.43773
    },
    {
      "epoch": 3.5718263998971764,
      "grad_norm": 8.625472068786621,
      "learning_rate": 1.8815739433981034e-05,
      "loss": 2.171437454223633,
      "memory(GiB)": 77.56,
      "step": 83370,
      "token_acc": 0.5037878787878788,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 3.5720406152264257,
      "grad_norm": 6.226128101348877,
      "learning_rate": 1.8810479217020654e-05,
      "loss": 2.4843910217285154,
      "memory(GiB)": 77.56,
      "step": 83375,
      "token_acc": 0.49828178694158076,
      "train_speed(iter/s)": 1.437737
    },
    {
      "epoch": 3.5722548305556745,
      "grad_norm": 6.627004623413086,
      "learning_rate": 1.8805219565087707e-05,
      "loss": 2.474307060241699,
      "memory(GiB)": 77.56,
      "step": 83380,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437744
    },
    {
      "epoch": 3.5724690458849233,
      "grad_norm": 5.86984920501709,
      "learning_rate": 1.879996047827747e-05,
      "loss": 1.9881847381591797,
      "memory(GiB)": 77.56,
      "step": 83385,
      "token_acc": 0.5623100303951368,
      "train_speed(iter/s)": 1.437748
    },
    {
      "epoch": 3.5726832612141726,
      "grad_norm": 4.480538845062256,
      "learning_rate": 1.8794701956685212e-05,
      "loss": 2.3707780838012695,
      "memory(GiB)": 77.56,
      "step": 83390,
      "token_acc": 0.5095541401273885,
      "train_speed(iter/s)": 1.437748
    },
    {
      "epoch": 3.5728974765434214,
      "grad_norm": 6.057560920715332,
      "learning_rate": 1.878944400040618e-05,
      "loss": 2.2293033599853516,
      "memory(GiB)": 77.56,
      "step": 83395,
      "token_acc": 0.528052805280528,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 3.57311169187267,
      "grad_norm": 5.348850727081299,
      "learning_rate": 1.8784186609535677e-05,
      "loss": 2.4478580474853517,
      "memory(GiB)": 77.56,
      "step": 83400,
      "token_acc": 0.47126436781609193,
      "train_speed(iter/s)": 1.437773
    },
    {
      "epoch": 3.5733259072019194,
      "grad_norm": 5.241185188293457,
      "learning_rate": 1.877892978416891e-05,
      "loss": 2.0920101165771485,
      "memory(GiB)": 77.56,
      "step": 83405,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.5735401225311683,
      "grad_norm": 7.74973201751709,
      "learning_rate": 1.87736735244011e-05,
      "loss": 2.1073328018188477,
      "memory(GiB)": 77.56,
      "step": 83410,
      "token_acc": 0.5241635687732342,
      "train_speed(iter/s)": 1.437738
    },
    {
      "epoch": 3.573754337860417,
      "grad_norm": 6.204540729522705,
      "learning_rate": 1.8768417830327507e-05,
      "loss": 2.2306278228759764,
      "memory(GiB)": 77.56,
      "step": 83415,
      "token_acc": 0.519163763066202,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.5739685531896663,
      "grad_norm": 6.120555400848389,
      "learning_rate": 1.876316270204332e-05,
      "loss": 2.218754196166992,
      "memory(GiB)": 77.56,
      "step": 83420,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.574182768518915,
      "grad_norm": 5.856813907623291,
      "learning_rate": 1.875790813964372e-05,
      "loss": 2.2100473403930665,
      "memory(GiB)": 77.56,
      "step": 83425,
      "token_acc": 0.4724137931034483,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 3.574396983848164,
      "grad_norm": 5.689226150512695,
      "learning_rate": 1.8752654143223946e-05,
      "loss": 2.5029279708862306,
      "memory(GiB)": 77.56,
      "step": 83430,
      "token_acc": 0.4876325088339223,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 3.574611199177413,
      "grad_norm": 5.500398635864258,
      "learning_rate": 1.8747400712879154e-05,
      "loss": 2.360789489746094,
      "memory(GiB)": 77.56,
      "step": 83435,
      "token_acc": 0.48615384615384616,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 3.574825414506662,
      "grad_norm": 8.030378341674805,
      "learning_rate": 1.8742147848704516e-05,
      "loss": 2.179517364501953,
      "memory(GiB)": 77.56,
      "step": 83440,
      "token_acc": 0.5437956204379562,
      "train_speed(iter/s)": 1.437761
    },
    {
      "epoch": 3.575039629835911,
      "grad_norm": 5.850546836853027,
      "learning_rate": 1.873689555079519e-05,
      "loss": 2.3703914642333985,
      "memory(GiB)": 77.56,
      "step": 83445,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.57525384516516,
      "grad_norm": 5.844541072845459,
      "learning_rate": 1.8731643819246312e-05,
      "loss": 2.42591552734375,
      "memory(GiB)": 77.56,
      "step": 83450,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.437779
    },
    {
      "epoch": 3.575468060494409,
      "grad_norm": 5.354722023010254,
      "learning_rate": 1.8726392654153058e-05,
      "loss": 2.4405771255493165,
      "memory(GiB)": 77.56,
      "step": 83455,
      "token_acc": 0.49852507374631266,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 3.5756822758236577,
      "grad_norm": 7.0297675132751465,
      "learning_rate": 1.8721142055610536e-05,
      "loss": 2.3134946823120117,
      "memory(GiB)": 77.56,
      "step": 83460,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.575896491152907,
      "grad_norm": 6.780994415283203,
      "learning_rate": 1.8715892023713866e-05,
      "loss": 2.5196014404296876,
      "memory(GiB)": 77.56,
      "step": 83465,
      "token_acc": 0.48986486486486486,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.576110706482156,
      "grad_norm": 6.342161178588867,
      "learning_rate": 1.8710642558558162e-05,
      "loss": 2.0995716094970702,
      "memory(GiB)": 77.56,
      "step": 83470,
      "token_acc": 0.5252100840336135,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.5763249218114046,
      "grad_norm": 5.033740043640137,
      "learning_rate": 1.870539366023852e-05,
      "loss": 2.407251739501953,
      "memory(GiB)": 77.56,
      "step": 83475,
      "token_acc": 0.5317725752508361,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 3.576539137140654,
      "grad_norm": 5.4944658279418945,
      "learning_rate": 1.8700145328850005e-05,
      "loss": 2.302262878417969,
      "memory(GiB)": 77.56,
      "step": 83480,
      "token_acc": 0.511400651465798,
      "train_speed(iter/s)": 1.437811
    },
    {
      "epoch": 3.5767533524699027,
      "grad_norm": 6.006190299987793,
      "learning_rate": 1.8694897564487725e-05,
      "loss": 2.326752853393555,
      "memory(GiB)": 77.56,
      "step": 83485,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.437811
    },
    {
      "epoch": 3.5769675677991515,
      "grad_norm": 6.718530654907227,
      "learning_rate": 1.868965036724676e-05,
      "loss": 2.514005661010742,
      "memory(GiB)": 77.56,
      "step": 83490,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.5771817831284007,
      "grad_norm": 5.306966304779053,
      "learning_rate": 1.8684403737222144e-05,
      "loss": 2.169674301147461,
      "memory(GiB)": 77.56,
      "step": 83495,
      "token_acc": 0.5232974910394266,
      "train_speed(iter/s)": 1.43784
    },
    {
      "epoch": 3.5773959984576496,
      "grad_norm": 6.0591912269592285,
      "learning_rate": 1.8679157674508936e-05,
      "loss": 2.4086856842041016,
      "memory(GiB)": 77.56,
      "step": 83500,
      "token_acc": 0.4437869822485207,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.5773959984576496,
      "eval_loss": 2.27353572845459,
      "eval_runtime": 14.5028,
      "eval_samples_per_second": 6.895,
      "eval_steps_per_second": 6.895,
      "eval_token_acc": 0.465,
      "step": 83500
    },
    {
      "epoch": 3.5776102137868984,
      "grad_norm": 7.037579536437988,
      "learning_rate": 1.8673912179202163e-05,
      "loss": 2.461414337158203,
      "memory(GiB)": 77.56,
      "step": 83505,
      "token_acc": 0.46955719557195574,
      "train_speed(iter/s)": 1.437482
    },
    {
      "epoch": 3.5778244291161476,
      "grad_norm": 6.035539150238037,
      "learning_rate": 1.8668667251396847e-05,
      "loss": 2.124465751647949,
      "memory(GiB)": 77.56,
      "step": 83510,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437508
    },
    {
      "epoch": 3.5780386444453964,
      "grad_norm": 5.7824883460998535,
      "learning_rate": 1.8663422891188024e-05,
      "loss": 2.38531436920166,
      "memory(GiB)": 77.56,
      "step": 83515,
      "token_acc": 0.4519230769230769,
      "train_speed(iter/s)": 1.43747
    },
    {
      "epoch": 3.5782528597746452,
      "grad_norm": 7.397764205932617,
      "learning_rate": 1.8658179098670698e-05,
      "loss": 2.171657371520996,
      "memory(GiB)": 77.56,
      "step": 83520,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.437475
    },
    {
      "epoch": 3.5784670751038945,
      "grad_norm": 4.856852054595947,
      "learning_rate": 1.8652935873939858e-05,
      "loss": 2.4343599319458007,
      "memory(GiB)": 77.56,
      "step": 83525,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.43748
    },
    {
      "epoch": 3.5786812904331433,
      "grad_norm": 6.537598609924316,
      "learning_rate": 1.864769321709049e-05,
      "loss": 2.4519826889038088,
      "memory(GiB)": 77.56,
      "step": 83530,
      "token_acc": 0.48360655737704916,
      "train_speed(iter/s)": 1.437489
    },
    {
      "epoch": 3.578895505762392,
      "grad_norm": 6.61281156539917,
      "learning_rate": 1.864245112821757e-05,
      "loss": 2.3779029846191406,
      "memory(GiB)": 77.56,
      "step": 83535,
      "token_acc": 0.5114503816793893,
      "train_speed(iter/s)": 1.437487
    },
    {
      "epoch": 3.5791097210916414,
      "grad_norm": 4.975118637084961,
      "learning_rate": 1.8637209607416044e-05,
      "loss": 2.0590206146240235,
      "memory(GiB)": 77.56,
      "step": 83540,
      "token_acc": 0.5758754863813229,
      "train_speed(iter/s)": 1.437488
    },
    {
      "epoch": 3.57932393642089,
      "grad_norm": 4.7138285636901855,
      "learning_rate": 1.8631968654780906e-05,
      "loss": 2.371194267272949,
      "memory(GiB)": 77.56,
      "step": 83545,
      "token_acc": 0.5083612040133779,
      "train_speed(iter/s)": 1.437486
    },
    {
      "epoch": 3.579538151750139,
      "grad_norm": 5.548679828643799,
      "learning_rate": 1.8626728270407078e-05,
      "loss": 2.3220003128051756,
      "memory(GiB)": 77.56,
      "step": 83550,
      "token_acc": 0.5150501672240803,
      "train_speed(iter/s)": 1.437483
    },
    {
      "epoch": 3.5797523670793883,
      "grad_norm": 4.663828372955322,
      "learning_rate": 1.86214884543895e-05,
      "loss": 2.4430870056152343,
      "memory(GiB)": 77.56,
      "step": 83555,
      "token_acc": 0.4984126984126984,
      "train_speed(iter/s)": 1.437474
    },
    {
      "epoch": 3.579966582408637,
      "grad_norm": 5.698978424072266,
      "learning_rate": 1.8616249206823085e-05,
      "loss": 2.2960018157958983,
      "memory(GiB)": 77.56,
      "step": 83560,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.437477
    },
    {
      "epoch": 3.580180797737886,
      "grad_norm": 6.00016450881958,
      "learning_rate": 1.8611010527802763e-05,
      "loss": 2.573992156982422,
      "memory(GiB)": 77.56,
      "step": 83565,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.437493
    },
    {
      "epoch": 3.580395013067135,
      "grad_norm": 4.582003593444824,
      "learning_rate": 1.8605772417423417e-05,
      "loss": 2.2473644256591796,
      "memory(GiB)": 77.56,
      "step": 83570,
      "token_acc": 0.5264900662251656,
      "train_speed(iter/s)": 1.43748
    },
    {
      "epoch": 3.580609228396384,
      "grad_norm": 5.979335308074951,
      "learning_rate": 1.8600534875779967e-05,
      "loss": 2.472529983520508,
      "memory(GiB)": 77.56,
      "step": 83575,
      "token_acc": 0.46827794561933533,
      "train_speed(iter/s)": 1.437496
    },
    {
      "epoch": 3.5808234437256328,
      "grad_norm": 6.774959564208984,
      "learning_rate": 1.859529790296729e-05,
      "loss": 2.517338180541992,
      "memory(GiB)": 77.56,
      "step": 83580,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.437499
    },
    {
      "epoch": 3.581037659054882,
      "grad_norm": 7.629566192626953,
      "learning_rate": 1.8590061499080246e-05,
      "loss": 2.458251190185547,
      "memory(GiB)": 77.56,
      "step": 83585,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437486
    },
    {
      "epoch": 3.581251874384131,
      "grad_norm": 7.092397212982178,
      "learning_rate": 1.8584825664213707e-05,
      "loss": 2.5144321441650392,
      "memory(GiB)": 77.56,
      "step": 83590,
      "token_acc": 0.4789156626506024,
      "train_speed(iter/s)": 1.4375
    },
    {
      "epoch": 3.5814660897133797,
      "grad_norm": 5.519599914550781,
      "learning_rate": 1.85795903984625e-05,
      "loss": 2.537326431274414,
      "memory(GiB)": 77.56,
      "step": 83595,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.437501
    },
    {
      "epoch": 3.581680305042629,
      "grad_norm": 5.344213485717773,
      "learning_rate": 1.8574355701921502e-05,
      "loss": 2.5091175079345702,
      "memory(GiB)": 77.56,
      "step": 83600,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.437501
    },
    {
      "epoch": 3.5818945203718777,
      "grad_norm": 4.627424716949463,
      "learning_rate": 1.8569121574685538e-05,
      "loss": 2.2671415328979494,
      "memory(GiB)": 77.56,
      "step": 83605,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437504
    },
    {
      "epoch": 3.5821087357011265,
      "grad_norm": 6.110616207122803,
      "learning_rate": 1.8563888016849417e-05,
      "loss": 2.309220886230469,
      "memory(GiB)": 77.56,
      "step": 83610,
      "token_acc": 0.5130718954248366,
      "train_speed(iter/s)": 1.437516
    },
    {
      "epoch": 3.582322951030376,
      "grad_norm": 5.834411144256592,
      "learning_rate": 1.8558655028507954e-05,
      "loss": 2.2596513748168947,
      "memory(GiB)": 77.56,
      "step": 83615,
      "token_acc": 0.48404255319148937,
      "train_speed(iter/s)": 1.437533
    },
    {
      "epoch": 3.5825371663596246,
      "grad_norm": 8.132269859313965,
      "learning_rate": 1.855342260975595e-05,
      "loss": 2.5855949401855467,
      "memory(GiB)": 77.56,
      "step": 83620,
      "token_acc": 0.4381625441696113,
      "train_speed(iter/s)": 1.437525
    },
    {
      "epoch": 3.5827513816888734,
      "grad_norm": 7.2655158042907715,
      "learning_rate": 1.8548190760688176e-05,
      "loss": 1.9536001205444335,
      "memory(GiB)": 77.56,
      "step": 83625,
      "token_acc": 0.5740740740740741,
      "train_speed(iter/s)": 1.437544
    },
    {
      "epoch": 3.5829655970181227,
      "grad_norm": 6.249345779418945,
      "learning_rate": 1.8542959481399435e-05,
      "loss": 2.417573165893555,
      "memory(GiB)": 77.56,
      "step": 83630,
      "token_acc": 0.4952681388012618,
      "train_speed(iter/s)": 1.437538
    },
    {
      "epoch": 3.5831798123473715,
      "grad_norm": 5.846346378326416,
      "learning_rate": 1.8537728771984504e-05,
      "loss": 2.258195495605469,
      "memory(GiB)": 77.56,
      "step": 83635,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.437554
    },
    {
      "epoch": 3.5833940276766203,
      "grad_norm": 4.9812912940979,
      "learning_rate": 1.853249863253813e-05,
      "loss": 2.118291664123535,
      "memory(GiB)": 77.56,
      "step": 83640,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.437559
    },
    {
      "epoch": 3.5836082430058696,
      "grad_norm": 6.051755428314209,
      "learning_rate": 1.8527269063155068e-05,
      "loss": 2.199260711669922,
      "memory(GiB)": 77.56,
      "step": 83645,
      "token_acc": 0.5382059800664452,
      "train_speed(iter/s)": 1.437569
    },
    {
      "epoch": 3.5838224583351184,
      "grad_norm": 6.251136302947998,
      "learning_rate": 1.8522040063930045e-05,
      "loss": 2.2736490249633787,
      "memory(GiB)": 77.56,
      "step": 83650,
      "token_acc": 0.5340501792114696,
      "train_speed(iter/s)": 1.437586
    },
    {
      "epoch": 3.584036673664367,
      "grad_norm": 6.631663799285889,
      "learning_rate": 1.851681163495778e-05,
      "loss": 2.3653751373291017,
      "memory(GiB)": 77.56,
      "step": 83655,
      "token_acc": 0.5167785234899329,
      "train_speed(iter/s)": 1.437594
    },
    {
      "epoch": 3.5842508889936164,
      "grad_norm": 5.372085094451904,
      "learning_rate": 1.8511583776333026e-05,
      "loss": 2.081974411010742,
      "memory(GiB)": 77.56,
      "step": 83660,
      "token_acc": 0.5322033898305085,
      "train_speed(iter/s)": 1.437593
    },
    {
      "epoch": 3.5844651043228652,
      "grad_norm": 5.604437828063965,
      "learning_rate": 1.8506356488150467e-05,
      "loss": 2.400470161437988,
      "memory(GiB)": 77.56,
      "step": 83665,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.437623
    },
    {
      "epoch": 3.584679319652114,
      "grad_norm": 6.18123197555542,
      "learning_rate": 1.8501129770504805e-05,
      "loss": 2.4475830078125,
      "memory(GiB)": 77.56,
      "step": 83670,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.43762
    },
    {
      "epoch": 3.5848935349813633,
      "grad_norm": 5.402523994445801,
      "learning_rate": 1.8495903623490722e-05,
      "loss": 2.5152135848999024,
      "memory(GiB)": 77.56,
      "step": 83675,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437632
    },
    {
      "epoch": 3.585107750310612,
      "grad_norm": 6.611544132232666,
      "learning_rate": 1.8490678047202885e-05,
      "loss": 2.4221534729003906,
      "memory(GiB)": 77.56,
      "step": 83680,
      "token_acc": 0.5275080906148867,
      "train_speed(iter/s)": 1.437651
    },
    {
      "epoch": 3.585321965639861,
      "grad_norm": 5.7812018394470215,
      "learning_rate": 1.848545304173599e-05,
      "loss": 2.1626869201660157,
      "memory(GiB)": 77.56,
      "step": 83685,
      "token_acc": 0.4880952380952381,
      "train_speed(iter/s)": 1.437659
    },
    {
      "epoch": 3.58553618096911,
      "grad_norm": 6.878974914550781,
      "learning_rate": 1.8480228607184675e-05,
      "loss": 2.4956584930419923,
      "memory(GiB)": 77.56,
      "step": 83690,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.585750396298359,
      "grad_norm": 5.689792633056641,
      "learning_rate": 1.8475004743643587e-05,
      "loss": 2.375166130065918,
      "memory(GiB)": 77.56,
      "step": 83695,
      "token_acc": 0.4869281045751634,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 3.585964611627608,
      "grad_norm": 6.481431484222412,
      "learning_rate": 1.846978145120736e-05,
      "loss": 2.2234724044799803,
      "memory(GiB)": 77.56,
      "step": 83700,
      "token_acc": 0.5323741007194245,
      "train_speed(iter/s)": 1.43768
    },
    {
      "epoch": 3.586178826956857,
      "grad_norm": 5.415676116943359,
      "learning_rate": 1.8464558729970603e-05,
      "loss": 2.0002246856689454,
      "memory(GiB)": 77.56,
      "step": 83705,
      "token_acc": 0.5983935742971888,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 3.586393042286106,
      "grad_norm": 6.878890514373779,
      "learning_rate": 1.845933658002797e-05,
      "loss": 2.1780679702758787,
      "memory(GiB)": 77.56,
      "step": 83710,
      "token_acc": 0.5364431486880467,
      "train_speed(iter/s)": 1.437697
    },
    {
      "epoch": 3.5866072576153547,
      "grad_norm": 5.245037078857422,
      "learning_rate": 1.845411500147402e-05,
      "loss": 2.156040000915527,
      "memory(GiB)": 77.56,
      "step": 83715,
      "token_acc": 0.5494880546075085,
      "train_speed(iter/s)": 1.437717
    },
    {
      "epoch": 3.586821472944604,
      "grad_norm": 5.945116996765137,
      "learning_rate": 1.844889399440338e-05,
      "loss": 2.3285757064819337,
      "memory(GiB)": 77.56,
      "step": 83720,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.437737
    },
    {
      "epoch": 3.5870356882738528,
      "grad_norm": 4.118406295776367,
      "learning_rate": 1.8443673558910635e-05,
      "loss": 2.4053455352783204,
      "memory(GiB)": 77.56,
      "step": 83725,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 3.5872499036031016,
      "grad_norm": 6.067787170410156,
      "learning_rate": 1.8438453695090342e-05,
      "loss": 2.390274238586426,
      "memory(GiB)": 77.56,
      "step": 83730,
      "token_acc": 0.502906976744186,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.587464118932351,
      "grad_norm": 6.500606060028076,
      "learning_rate": 1.8433234403037065e-05,
      "loss": 2.5961103439331055,
      "memory(GiB)": 77.56,
      "step": 83735,
      "token_acc": 0.46283783783783783,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.5876783342615997,
      "grad_norm": 5.9028544425964355,
      "learning_rate": 1.842801568284534e-05,
      "loss": 2.1781572341918944,
      "memory(GiB)": 77.56,
      "step": 83740,
      "token_acc": 0.5367647058823529,
      "train_speed(iter/s)": 1.437764
    },
    {
      "epoch": 3.5878925495908485,
      "grad_norm": 6.955174446105957,
      "learning_rate": 1.8422797534609748e-05,
      "loss": 2.2484804153442384,
      "memory(GiB)": 77.56,
      "step": 83745,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 3.5881067649200977,
      "grad_norm": 5.07317590713501,
      "learning_rate": 1.8417579958424797e-05,
      "loss": 2.1796667098999025,
      "memory(GiB)": 77.56,
      "step": 83750,
      "token_acc": 0.5360824742268041,
      "train_speed(iter/s)": 1.43772
    },
    {
      "epoch": 3.5883209802493465,
      "grad_norm": 5.653574466705322,
      "learning_rate": 1.8412362954385015e-05,
      "loss": 2.3984197616577148,
      "memory(GiB)": 77.56,
      "step": 83755,
      "token_acc": 0.48424068767908307,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 3.5885351955785953,
      "grad_norm": 5.272899150848389,
      "learning_rate": 1.8407146522584906e-05,
      "loss": 2.333488655090332,
      "memory(GiB)": 77.56,
      "step": 83760,
      "token_acc": 0.4979253112033195,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 3.5887494109078446,
      "grad_norm": 7.157073497772217,
      "learning_rate": 1.8401930663118983e-05,
      "loss": 2.637564277648926,
      "memory(GiB)": 77.56,
      "step": 83765,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 3.5889636262370934,
      "grad_norm": 7.094167232513428,
      "learning_rate": 1.83967153760817e-05,
      "loss": 2.211750030517578,
      "memory(GiB)": 77.56,
      "step": 83770,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.5891778415663422,
      "grad_norm": 6.066314697265625,
      "learning_rate": 1.8391500661567584e-05,
      "loss": 2.4501609802246094,
      "memory(GiB)": 77.56,
      "step": 83775,
      "token_acc": 0.5019305019305019,
      "train_speed(iter/s)": 1.437768
    },
    {
      "epoch": 3.5893920568955915,
      "grad_norm": 7.008761882781982,
      "learning_rate": 1.838628651967107e-05,
      "loss": 2.1569086074829102,
      "memory(GiB)": 77.56,
      "step": 83780,
      "token_acc": 0.549407114624506,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 3.5896062722248403,
      "grad_norm": 5.256586074829102,
      "learning_rate": 1.838107295048665e-05,
      "loss": 2.029164695739746,
      "memory(GiB)": 77.56,
      "step": 83785,
      "token_acc": 0.5477941176470589,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 3.589820487554089,
      "grad_norm": 4.974340915679932,
      "learning_rate": 1.8375859954108753e-05,
      "loss": 2.600382614135742,
      "memory(GiB)": 77.56,
      "step": 83790,
      "token_acc": 0.4709480122324159,
      "train_speed(iter/s)": 1.437833
    },
    {
      "epoch": 3.5900347028833384,
      "grad_norm": 5.226301670074463,
      "learning_rate": 1.8370647530631818e-05,
      "loss": 2.4387868881225585,
      "memory(GiB)": 77.56,
      "step": 83795,
      "token_acc": 0.49615384615384617,
      "train_speed(iter/s)": 1.437844
    },
    {
      "epoch": 3.590248918212587,
      "grad_norm": 8.13363265991211,
      "learning_rate": 1.8365435680150256e-05,
      "loss": 2.3786300659179687,
      "memory(GiB)": 77.56,
      "step": 83800,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.43786
    },
    {
      "epoch": 3.590463133541836,
      "grad_norm": 5.606438636779785,
      "learning_rate": 1.8360224402758525e-05,
      "loss": 2.511959266662598,
      "memory(GiB)": 77.56,
      "step": 83805,
      "token_acc": 0.5016835016835017,
      "train_speed(iter/s)": 1.437854
    },
    {
      "epoch": 3.5906773488710853,
      "grad_norm": 7.411236763000488,
      "learning_rate": 1.8355013698551004e-05,
      "loss": 2.3481796264648436,
      "memory(GiB)": 77.56,
      "step": 83810,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.590891564200334,
      "grad_norm": 5.4996018409729,
      "learning_rate": 1.8349803567622094e-05,
      "loss": 2.2429292678833006,
      "memory(GiB)": 77.56,
      "step": 83815,
      "token_acc": 0.4875444839857651,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.591105779529583,
      "grad_norm": 5.252297401428223,
      "learning_rate": 1.8344594010066184e-05,
      "loss": 2.380876159667969,
      "memory(GiB)": 77.56,
      "step": 83820,
      "token_acc": 0.5062893081761006,
      "train_speed(iter/s)": 1.437862
    },
    {
      "epoch": 3.591319994858832,
      "grad_norm": 5.427470684051514,
      "learning_rate": 1.8339385025977635e-05,
      "loss": 2.228294563293457,
      "memory(GiB)": 77.56,
      "step": 83825,
      "token_acc": 0.5150375939849624,
      "train_speed(iter/s)": 1.437849
    },
    {
      "epoch": 3.591534210188081,
      "grad_norm": 5.288304805755615,
      "learning_rate": 1.833417661545084e-05,
      "loss": 2.226371955871582,
      "memory(GiB)": 77.56,
      "step": 83830,
      "token_acc": 0.5445205479452054,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.5917484255173298,
      "grad_norm": 5.43513298034668,
      "learning_rate": 1.832896877858014e-05,
      "loss": 2.363986778259277,
      "memory(GiB)": 77.56,
      "step": 83835,
      "token_acc": 0.5060975609756098,
      "train_speed(iter/s)": 1.437868
    },
    {
      "epoch": 3.591962640846579,
      "grad_norm": 6.374598979949951,
      "learning_rate": 1.832376151545988e-05,
      "loss": 2.337849235534668,
      "memory(GiB)": 77.56,
      "step": 83840,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 3.592176856175828,
      "grad_norm": 7.433907985687256,
      "learning_rate": 1.83185548261844e-05,
      "loss": 2.0900434494018554,
      "memory(GiB)": 77.56,
      "step": 83845,
      "token_acc": 0.5192307692307693,
      "train_speed(iter/s)": 1.437868
    },
    {
      "epoch": 3.5923910715050766,
      "grad_norm": 5.9668049812316895,
      "learning_rate": 1.8313348710848018e-05,
      "loss": 2.6514781951904296,
      "memory(GiB)": 77.56,
      "step": 83850,
      "token_acc": 0.4567901234567901,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.592605286834326,
      "grad_norm": 5.6707258224487305,
      "learning_rate": 1.8308143169545028e-05,
      "loss": 2.1907678604125977,
      "memory(GiB)": 77.56,
      "step": 83855,
      "token_acc": 0.51953125,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 3.5928195021635747,
      "grad_norm": 5.066697120666504,
      "learning_rate": 1.8302938202369752e-05,
      "loss": 2.246035575866699,
      "memory(GiB)": 77.56,
      "step": 83860,
      "token_acc": 0.53,
      "train_speed(iter/s)": 1.437915
    },
    {
      "epoch": 3.5930337174928235,
      "grad_norm": 5.942101001739502,
      "learning_rate": 1.8297733809416505e-05,
      "loss": 2.3774572372436524,
      "memory(GiB)": 77.56,
      "step": 83865,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 3.5932479328220728,
      "grad_norm": 4.433930397033691,
      "learning_rate": 1.829252999077955e-05,
      "loss": 2.1279762268066404,
      "memory(GiB)": 77.56,
      "step": 83870,
      "token_acc": 0.5049833887043189,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 3.5934621481513216,
      "grad_norm": 4.171961784362793,
      "learning_rate": 1.8287326746553157e-05,
      "loss": 2.4300605773925783,
      "memory(GiB)": 77.56,
      "step": 83875,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 3.5936763634805704,
      "grad_norm": 6.884018898010254,
      "learning_rate": 1.8282124076831592e-05,
      "loss": 2.5158658981323243,
      "memory(GiB)": 77.56,
      "step": 83880,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437966
    },
    {
      "epoch": 3.5938905788098197,
      "grad_norm": 8.164193153381348,
      "learning_rate": 1.8276921981709077e-05,
      "loss": 2.404288101196289,
      "memory(GiB)": 77.56,
      "step": 83885,
      "token_acc": 0.5146579804560261,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 3.5941047941390685,
      "grad_norm": 5.8763275146484375,
      "learning_rate": 1.8271720461279907e-05,
      "loss": 2.0672731399536133,
      "memory(GiB)": 77.56,
      "step": 83890,
      "token_acc": 0.575,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 3.5943190094683173,
      "grad_norm": 7.617894649505615,
      "learning_rate": 1.8266519515638277e-05,
      "loss": 2.2012426376342775,
      "memory(GiB)": 77.56,
      "step": 83895,
      "token_acc": 0.5186567164179104,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 3.5945332247975665,
      "grad_norm": 5.624991416931152,
      "learning_rate": 1.826131914487842e-05,
      "loss": 2.0258535385131835,
      "memory(GiB)": 77.56,
      "step": 83900,
      "token_acc": 0.5661764705882353,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 3.5947474401268154,
      "grad_norm": 5.233500957489014,
      "learning_rate": 1.8256119349094538e-05,
      "loss": 2.4378084182739257,
      "memory(GiB)": 77.56,
      "step": 83905,
      "token_acc": 0.49375,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.594961655456064,
      "grad_norm": 6.583970069885254,
      "learning_rate": 1.8250920128380826e-05,
      "loss": 2.4524112701416017,
      "memory(GiB)": 77.56,
      "step": 83910,
      "token_acc": 0.46417445482866043,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 3.5951758707853134,
      "grad_norm": 6.571187496185303,
      "learning_rate": 1.8245721482831468e-05,
      "loss": 2.493466377258301,
      "memory(GiB)": 77.56,
      "step": 83915,
      "token_acc": 0.43465045592705165,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 3.5953900861145622,
      "grad_norm": 4.98223352432251,
      "learning_rate": 1.8240523412540667e-05,
      "loss": 2.184986877441406,
      "memory(GiB)": 77.56,
      "step": 83920,
      "token_acc": 0.46540880503144655,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.595604301443811,
      "grad_norm": 5.9935712814331055,
      "learning_rate": 1.823532591760258e-05,
      "loss": 2.578005790710449,
      "memory(GiB)": 77.56,
      "step": 83925,
      "token_acc": 0.48089171974522293,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 3.5958185167730603,
      "grad_norm": 6.469836711883545,
      "learning_rate": 1.8230128998111346e-05,
      "loss": 2.2249731063842773,
      "memory(GiB)": 77.56,
      "step": 83930,
      "token_acc": 0.5458333333333333,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 3.596032732102309,
      "grad_norm": 5.506590843200684,
      "learning_rate": 1.8224932654161143e-05,
      "loss": 2.093449020385742,
      "memory(GiB)": 77.56,
      "step": 83935,
      "token_acc": 0.5358361774744027,
      "train_speed(iter/s)": 1.438026
    },
    {
      "epoch": 3.596246947431558,
      "grad_norm": 5.520415306091309,
      "learning_rate": 1.821973688584609e-05,
      "loss": 2.1561962127685548,
      "memory(GiB)": 77.56,
      "step": 83940,
      "token_acc": 0.5527272727272727,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 3.596461162760807,
      "grad_norm": 6.590030670166016,
      "learning_rate": 1.82145416932603e-05,
      "loss": 2.2194372177124024,
      "memory(GiB)": 77.56,
      "step": 83945,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 3.596675378090056,
      "grad_norm": 5.831902980804443,
      "learning_rate": 1.8209347076497924e-05,
      "loss": 2.4743839263916017,
      "memory(GiB)": 77.56,
      "step": 83950,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.596889593419305,
      "grad_norm": 4.735555171966553,
      "learning_rate": 1.820415303565305e-05,
      "loss": 2.321951675415039,
      "memory(GiB)": 77.56,
      "step": 83955,
      "token_acc": 0.4968944099378882,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 3.597103808748554,
      "grad_norm": 5.199782371520996,
      "learning_rate": 1.8198959570819763e-05,
      "loss": 2.5252227783203125,
      "memory(GiB)": 77.56,
      "step": 83960,
      "token_acc": 0.5027777777777778,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 3.597318024077803,
      "grad_norm": 6.447930812835693,
      "learning_rate": 1.8193766682092155e-05,
      "loss": 2.4034652709960938,
      "memory(GiB)": 77.56,
      "step": 83965,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 3.597532239407052,
      "grad_norm": 5.510873794555664,
      "learning_rate": 1.8188574369564293e-05,
      "loss": 2.4141664505004883,
      "memory(GiB)": 77.56,
      "step": 83970,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.438025
    },
    {
      "epoch": 3.597746454736301,
      "grad_norm": 8.423981666564941,
      "learning_rate": 1.8183382633330255e-05,
      "loss": 2.373681640625,
      "memory(GiB)": 77.56,
      "step": 83975,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.438047
    },
    {
      "epoch": 3.5979606700655498,
      "grad_norm": 5.218179702758789,
      "learning_rate": 1.817819147348409e-05,
      "loss": 2.500479888916016,
      "memory(GiB)": 77.56,
      "step": 83980,
      "token_acc": 0.4735202492211838,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 3.598174885394799,
      "grad_norm": 10.89646053314209,
      "learning_rate": 1.817300089011984e-05,
      "loss": 2.2559423446655273,
      "memory(GiB)": 77.56,
      "step": 83985,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.438057
    },
    {
      "epoch": 3.598389100724048,
      "grad_norm": 8.182391166687012,
      "learning_rate": 1.816781088333153e-05,
      "loss": 2.2623111724853517,
      "memory(GiB)": 77.56,
      "step": 83990,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 3.5986033160532966,
      "grad_norm": 7.0720415115356445,
      "learning_rate": 1.8162621453213197e-05,
      "loss": 2.4828422546386717,
      "memory(GiB)": 77.56,
      "step": 83995,
      "token_acc": 0.50920245398773,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 3.598817531382546,
      "grad_norm": 6.682283401489258,
      "learning_rate": 1.8157432599858814e-05,
      "loss": 2.1567333221435545,
      "memory(GiB)": 77.56,
      "step": 84000,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 3.598817531382546,
      "eval_loss": 2.127963066101074,
      "eval_runtime": 15.0222,
      "eval_samples_per_second": 6.657,
      "eval_steps_per_second": 6.657,
      "eval_token_acc": 0.4673495518565941,
      "step": 84000
    },
    {
      "epoch": 3.5990317467117947,
      "grad_norm": 7.187999248504639,
      "learning_rate": 1.8152244323362416e-05,
      "loss": 2.5996124267578127,
      "memory(GiB)": 77.56,
      "step": 84005,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.437692
    },
    {
      "epoch": 3.5992459620410435,
      "grad_norm": 6.486800193786621,
      "learning_rate": 1.8147056623818004e-05,
      "loss": 2.2904407501220705,
      "memory(GiB)": 77.56,
      "step": 84010,
      "token_acc": 0.5133079847908745,
      "train_speed(iter/s)": 1.437684
    },
    {
      "epoch": 3.599460177370293,
      "grad_norm": 5.945263862609863,
      "learning_rate": 1.814186950131954e-05,
      "loss": 2.2699140548706054,
      "memory(GiB)": 77.56,
      "step": 84015,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.437708
    },
    {
      "epoch": 3.5996743926995416,
      "grad_norm": 5.000082969665527,
      "learning_rate": 1.8136682955960997e-05,
      "loss": 2.390966796875,
      "memory(GiB)": 77.56,
      "step": 84020,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.43771
    },
    {
      "epoch": 3.5998886080287904,
      "grad_norm": 7.152932167053223,
      "learning_rate": 1.813149698783633e-05,
      "loss": 2.339006805419922,
      "memory(GiB)": 77.56,
      "step": 84025,
      "token_acc": 0.5213675213675214,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 3.6001028233580397,
      "grad_norm": 6.6526923179626465,
      "learning_rate": 1.812631159703948e-05,
      "loss": 2.447410011291504,
      "memory(GiB)": 77.56,
      "step": 84030,
      "token_acc": 0.4729241877256318,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 3.6003170386872885,
      "grad_norm": 7.088724613189697,
      "learning_rate": 1.8121126783664406e-05,
      "loss": 2.273637390136719,
      "memory(GiB)": 77.56,
      "step": 84035,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437709
    },
    {
      "epoch": 3.6005312540165373,
      "grad_norm": 5.185393810272217,
      "learning_rate": 1.8115942547805025e-05,
      "loss": 2.0088356018066404,
      "memory(GiB)": 77.56,
      "step": 84040,
      "token_acc": 0.5608108108108109,
      "train_speed(iter/s)": 1.437726
    },
    {
      "epoch": 3.6007454693457865,
      "grad_norm": 6.586428165435791,
      "learning_rate": 1.8110758889555256e-05,
      "loss": 2.476908874511719,
      "memory(GiB)": 77.56,
      "step": 84045,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 3.6009596846750354,
      "grad_norm": 5.801070690155029,
      "learning_rate": 1.8105575809009008e-05,
      "loss": 2.1285720825195313,
      "memory(GiB)": 77.56,
      "step": 84050,
      "token_acc": 0.5111821086261981,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 3.601173900004284,
      "grad_norm": 5.38275146484375,
      "learning_rate": 1.8100393306260166e-05,
      "loss": 2.160166931152344,
      "memory(GiB)": 77.56,
      "step": 84055,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.43773
    },
    {
      "epoch": 3.6013881153335334,
      "grad_norm": 6.874888896942139,
      "learning_rate": 1.809521138140261e-05,
      "loss": 2.0716358184814454,
      "memory(GiB)": 77.56,
      "step": 84060,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 3.6016023306627822,
      "grad_norm": 5.6881513595581055,
      "learning_rate": 1.8090030034530247e-05,
      "loss": 2.5533378601074217,
      "memory(GiB)": 77.56,
      "step": 84065,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.437751
    },
    {
      "epoch": 3.601816545992031,
      "grad_norm": 6.07249641418457,
      "learning_rate": 1.8084849265736927e-05,
      "loss": 2.2581628799438476,
      "memory(GiB)": 77.56,
      "step": 84070,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 3.6020307613212803,
      "grad_norm": 4.646111488342285,
      "learning_rate": 1.807966907511648e-05,
      "loss": 2.203865814208984,
      "memory(GiB)": 77.56,
      "step": 84075,
      "token_acc": 0.5290102389078498,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 3.602244976650529,
      "grad_norm": 5.254541397094727,
      "learning_rate": 1.807448946276279e-05,
      "loss": 2.1519844055175783,
      "memory(GiB)": 77.56,
      "step": 84080,
      "token_acc": 0.5390334572490706,
      "train_speed(iter/s)": 1.437792
    },
    {
      "epoch": 3.602459191979778,
      "grad_norm": 6.682652473449707,
      "learning_rate": 1.806931042876967e-05,
      "loss": 2.4928821563720702,
      "memory(GiB)": 77.56,
      "step": 84085,
      "token_acc": 0.466403162055336,
      "train_speed(iter/s)": 1.437794
    },
    {
      "epoch": 3.602673407309027,
      "grad_norm": 7.392765522003174,
      "learning_rate": 1.806413197323093e-05,
      "loss": 2.5664039611816407,
      "memory(GiB)": 77.56,
      "step": 84090,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 3.602887622638276,
      "grad_norm": 7.688574314117432,
      "learning_rate": 1.805895409624041e-05,
      "loss": 2.4077522277832033,
      "memory(GiB)": 77.56,
      "step": 84095,
      "token_acc": 0.44368600682593856,
      "train_speed(iter/s)": 1.437818
    },
    {
      "epoch": 3.603101837967525,
      "grad_norm": 8.079208374023438,
      "learning_rate": 1.80537767978919e-05,
      "loss": 2.67993106842041,
      "memory(GiB)": 77.56,
      "step": 84100,
      "token_acc": 0.494949494949495,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 3.603316053296774,
      "grad_norm": 8.550474166870117,
      "learning_rate": 1.80486000782792e-05,
      "loss": 2.0392208099365234,
      "memory(GiB)": 77.56,
      "step": 84105,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.437838
    },
    {
      "epoch": 3.603530268626023,
      "grad_norm": 4.775581359863281,
      "learning_rate": 1.804342393749608e-05,
      "loss": 2.142267036437988,
      "memory(GiB)": 77.56,
      "step": 84110,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 3.6037444839552717,
      "grad_norm": 5.96275520324707,
      "learning_rate": 1.803824837563629e-05,
      "loss": 2.696409225463867,
      "memory(GiB)": 77.56,
      "step": 84115,
      "token_acc": 0.4723926380368098,
      "train_speed(iter/s)": 1.437844
    },
    {
      "epoch": 3.603958699284521,
      "grad_norm": 5.446780204772949,
      "learning_rate": 1.803307339279363e-05,
      "loss": 2.10909481048584,
      "memory(GiB)": 77.56,
      "step": 84120,
      "token_acc": 0.5595667870036101,
      "train_speed(iter/s)": 1.437844
    },
    {
      "epoch": 3.6041729146137698,
      "grad_norm": 5.746261119842529,
      "learning_rate": 1.8027898989061838e-05,
      "loss": 2.40936222076416,
      "memory(GiB)": 77.56,
      "step": 84125,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437857
    },
    {
      "epoch": 3.6043871299430186,
      "grad_norm": 5.935001373291016,
      "learning_rate": 1.8022725164534647e-05,
      "loss": 2.2035167694091795,
      "memory(GiB)": 77.56,
      "step": 84130,
      "token_acc": 0.5310077519379846,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 3.604601345272268,
      "grad_norm": 7.3856916427612305,
      "learning_rate": 1.8017551919305787e-05,
      "loss": 2.4229381561279295,
      "memory(GiB)": 77.56,
      "step": 84135,
      "token_acc": 0.46923076923076923,
      "train_speed(iter/s)": 1.437898
    },
    {
      "epoch": 3.6048155606015166,
      "grad_norm": 5.225915431976318,
      "learning_rate": 1.8012379253468975e-05,
      "loss": 2.365320587158203,
      "memory(GiB)": 77.56,
      "step": 84140,
      "token_acc": 0.5254901960784314,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.6050297759307655,
      "grad_norm": 7.766951560974121,
      "learning_rate": 1.80072071671179e-05,
      "loss": 2.5186187744140627,
      "memory(GiB)": 77.56,
      "step": 84145,
      "token_acc": 0.4790874524714829,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 3.6052439912600147,
      "grad_norm": 4.88001012802124,
      "learning_rate": 1.8002035660346283e-05,
      "loss": 2.2217777252197264,
      "memory(GiB)": 77.56,
      "step": 84150,
      "token_acc": 0.5598455598455598,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 3.6054582065892635,
      "grad_norm": 5.7520952224731445,
      "learning_rate": 1.7996864733247822e-05,
      "loss": 2.175375556945801,
      "memory(GiB)": 77.56,
      "step": 84155,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 3.6056724219185123,
      "grad_norm": 5.575802803039551,
      "learning_rate": 1.7991694385916176e-05,
      "loss": 2.2137933731079102,
      "memory(GiB)": 77.56,
      "step": 84160,
      "token_acc": 0.5127388535031847,
      "train_speed(iter/s)": 1.437876
    },
    {
      "epoch": 3.6058866372477616,
      "grad_norm": 8.660526275634766,
      "learning_rate": 1.7986524618445017e-05,
      "loss": 2.3878547668457033,
      "memory(GiB)": 77.56,
      "step": 84165,
      "token_acc": 0.5246478873239436,
      "train_speed(iter/s)": 1.437852
    },
    {
      "epoch": 3.6061008525770104,
      "grad_norm": 4.151677131652832,
      "learning_rate": 1.7981355430927992e-05,
      "loss": 2.5714694976806642,
      "memory(GiB)": 77.56,
      "step": 84170,
      "token_acc": 0.5080906148867314,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 3.606315067906259,
      "grad_norm": 5.799408912658691,
      "learning_rate": 1.797618682345873e-05,
      "loss": 2.5114513397216798,
      "memory(GiB)": 77.56,
      "step": 84175,
      "token_acc": 0.4664179104477612,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.6065292832355085,
      "grad_norm": 4.498894691467285,
      "learning_rate": 1.7971018796130906e-05,
      "loss": 2.5082462310791014,
      "memory(GiB)": 77.56,
      "step": 84180,
      "token_acc": 0.4585987261146497,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.6067434985647573,
      "grad_norm": 5.322307109832764,
      "learning_rate": 1.796585134903812e-05,
      "loss": 2.4112205505371094,
      "memory(GiB)": 77.56,
      "step": 84185,
      "token_acc": 0.4817073170731707,
      "train_speed(iter/s)": 1.437838
    },
    {
      "epoch": 3.606957713894006,
      "grad_norm": 7.19744873046875,
      "learning_rate": 1.7960684482273986e-05,
      "loss": 2.229107666015625,
      "memory(GiB)": 77.56,
      "step": 84190,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 3.6071719292232554,
      "grad_norm": 5.487410068511963,
      "learning_rate": 1.7955518195932114e-05,
      "loss": 2.1874242782592774,
      "memory(GiB)": 77.56,
      "step": 84195,
      "token_acc": 0.5377049180327869,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 3.607386144552504,
      "grad_norm": 7.555301189422607,
      "learning_rate": 1.7950352490106083e-05,
      "loss": 2.361857032775879,
      "memory(GiB)": 77.56,
      "step": 84200,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.607600359881753,
      "grad_norm": 4.5566816329956055,
      "learning_rate": 1.794518736488947e-05,
      "loss": 2.4064708709716798,
      "memory(GiB)": 77.56,
      "step": 84205,
      "token_acc": 0.46981627296587924,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 3.6078145752110022,
      "grad_norm": 5.034022808074951,
      "learning_rate": 1.794002282037587e-05,
      "loss": 2.3505836486816407,
      "memory(GiB)": 77.56,
      "step": 84210,
      "token_acc": 0.4923547400611621,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 3.608028790540251,
      "grad_norm": 5.144019603729248,
      "learning_rate": 1.793485885665883e-05,
      "loss": 2.4347957611083983,
      "memory(GiB)": 77.56,
      "step": 84215,
      "token_acc": 0.49079754601226994,
      "train_speed(iter/s)": 1.437834
    },
    {
      "epoch": 3.6082430058695,
      "grad_norm": 6.86022424697876,
      "learning_rate": 1.792969547383191e-05,
      "loss": 2.345616912841797,
      "memory(GiB)": 77.56,
      "step": 84220,
      "token_acc": 0.5052264808362369,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.608457221198749,
      "grad_norm": 6.0889716148376465,
      "learning_rate": 1.792453267198862e-05,
      "loss": 2.384096145629883,
      "memory(GiB)": 77.56,
      "step": 84225,
      "token_acc": 0.4786885245901639,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.608671436527998,
      "grad_norm": 7.356601715087891,
      "learning_rate": 1.7919370451222523e-05,
      "loss": 2.4440357208251955,
      "memory(GiB)": 77.56,
      "step": 84230,
      "token_acc": 0.5167785234899329,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 3.6088856518572467,
      "grad_norm": 6.352375030517578,
      "learning_rate": 1.791420881162711e-05,
      "loss": 2.1649765014648437,
      "memory(GiB)": 77.56,
      "step": 84235,
      "token_acc": 0.5224913494809689,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 3.609099867186496,
      "grad_norm": 5.922575950622559,
      "learning_rate": 1.790904775329592e-05,
      "loss": 2.3507728576660156,
      "memory(GiB)": 77.56,
      "step": 84240,
      "token_acc": 0.473015873015873,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.609314082515745,
      "grad_norm": 6.288422107696533,
      "learning_rate": 1.7903887276322433e-05,
      "loss": 2.425602340698242,
      "memory(GiB)": 77.56,
      "step": 84245,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 3.6095282978449936,
      "grad_norm": 5.826930522918701,
      "learning_rate": 1.789872738080014e-05,
      "loss": 2.387664222717285,
      "memory(GiB)": 77.56,
      "step": 84250,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 3.609742513174243,
      "grad_norm": 4.966198921203613,
      "learning_rate": 1.7893568066822507e-05,
      "loss": 2.2773681640625,
      "memory(GiB)": 77.56,
      "step": 84255,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 3.6099567285034917,
      "grad_norm": 5.8261637687683105,
      "learning_rate": 1.7888409334483e-05,
      "loss": 2.4344600677490233,
      "memory(GiB)": 77.56,
      "step": 84260,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 3.6101709438327405,
      "grad_norm": 5.018042087554932,
      "learning_rate": 1.788325118387509e-05,
      "loss": 2.3421566009521486,
      "memory(GiB)": 77.56,
      "step": 84265,
      "token_acc": 0.4767025089605735,
      "train_speed(iter/s)": 1.437876
    },
    {
      "epoch": 3.6103851591619898,
      "grad_norm": 5.248198986053467,
      "learning_rate": 1.787809361509221e-05,
      "loss": 2.4936141967773438,
      "memory(GiB)": 77.56,
      "step": 84270,
      "token_acc": 0.478134110787172,
      "train_speed(iter/s)": 1.43789
    },
    {
      "epoch": 3.6105993744912386,
      "grad_norm": 5.8466715812683105,
      "learning_rate": 1.7872936628227804e-05,
      "loss": 2.4604724884033202,
      "memory(GiB)": 77.56,
      "step": 84275,
      "token_acc": 0.49085365853658536,
      "train_speed(iter/s)": 1.4379
    },
    {
      "epoch": 3.6108135898204874,
      "grad_norm": 5.019824504852295,
      "learning_rate": 1.7867780223375284e-05,
      "loss": 2.2418691635131838,
      "memory(GiB)": 77.56,
      "step": 84280,
      "token_acc": 0.5295857988165681,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 3.6110278051497366,
      "grad_norm": 5.7772111892700195,
      "learning_rate": 1.7862624400628074e-05,
      "loss": 2.338362121582031,
      "memory(GiB)": 77.56,
      "step": 84285,
      "token_acc": 0.5095367847411444,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 3.6112420204789855,
      "grad_norm": 5.309464454650879,
      "learning_rate": 1.785746916007955e-05,
      "loss": 2.358504867553711,
      "memory(GiB)": 77.56,
      "step": 84290,
      "token_acc": 0.55,
      "train_speed(iter/s)": 1.437923
    },
    {
      "epoch": 3.6114562358082343,
      "grad_norm": 5.367234706878662,
      "learning_rate": 1.785231450182314e-05,
      "loss": 2.503060531616211,
      "memory(GiB)": 77.56,
      "step": 84295,
      "token_acc": 0.4601226993865031,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 3.6116704511374835,
      "grad_norm": 4.922787189483643,
      "learning_rate": 1.7847160425952193e-05,
      "loss": 2.4450763702392577,
      "memory(GiB)": 77.56,
      "step": 84300,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437916
    },
    {
      "epoch": 3.6118846664667323,
      "grad_norm": 7.75718879699707,
      "learning_rate": 1.7842006932560113e-05,
      "loss": 2.114785003662109,
      "memory(GiB)": 77.56,
      "step": 84305,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 3.6120988817959816,
      "grad_norm": 5.380731105804443,
      "learning_rate": 1.7836854021740245e-05,
      "loss": 2.2468557357788086,
      "memory(GiB)": 77.56,
      "step": 84310,
      "token_acc": 0.5430711610486891,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 3.6123130971252304,
      "grad_norm": 5.819838523864746,
      "learning_rate": 1.783170169358594e-05,
      "loss": 2.368028259277344,
      "memory(GiB)": 77.56,
      "step": 84315,
      "token_acc": 0.5033333333333333,
      "train_speed(iter/s)": 1.437936
    },
    {
      "epoch": 3.6125273124544792,
      "grad_norm": 6.225353240966797,
      "learning_rate": 1.7826549948190514e-05,
      "loss": 2.455536460876465,
      "memory(GiB)": 77.56,
      "step": 84320,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 3.6127415277837285,
      "grad_norm": 5.648646831512451,
      "learning_rate": 1.7821398785647338e-05,
      "loss": 2.1312772750854494,
      "memory(GiB)": 77.56,
      "step": 84325,
      "token_acc": 0.5364963503649635,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 3.6129557431129773,
      "grad_norm": 7.226056098937988,
      "learning_rate": 1.7816248206049708e-05,
      "loss": 2.219803810119629,
      "memory(GiB)": 77.56,
      "step": 84330,
      "token_acc": 0.5265017667844523,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 3.613169958442226,
      "grad_norm": 5.097196578979492,
      "learning_rate": 1.7811098209490927e-05,
      "loss": 2.351849365234375,
      "memory(GiB)": 77.56,
      "step": 84335,
      "token_acc": 0.46394984326018807,
      "train_speed(iter/s)": 1.437913
    },
    {
      "epoch": 3.6133841737714754,
      "grad_norm": 9.04416561126709,
      "learning_rate": 1.7805948796064305e-05,
      "loss": 2.352223777770996,
      "memory(GiB)": 77.56,
      "step": 84340,
      "token_acc": 0.5086505190311419,
      "train_speed(iter/s)": 1.437891
    },
    {
      "epoch": 3.613598389100724,
      "grad_norm": 4.838322639465332,
      "learning_rate": 1.780079996586312e-05,
      "loss": 2.335775947570801,
      "memory(GiB)": 77.56,
      "step": 84345,
      "token_acc": 0.489247311827957,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 3.613812604429973,
      "grad_norm": 5.915621757507324,
      "learning_rate": 1.7795651718980626e-05,
      "loss": 2.7252878189086913,
      "memory(GiB)": 77.56,
      "step": 84350,
      "token_acc": 0.43661971830985913,
      "train_speed(iter/s)": 1.437933
    },
    {
      "epoch": 3.6140268197592222,
      "grad_norm": 4.800912380218506,
      "learning_rate": 1.779050405551013e-05,
      "loss": 2.3222444534301756,
      "memory(GiB)": 77.56,
      "step": 84355,
      "token_acc": 0.501577287066246,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 3.614241035088471,
      "grad_norm": 7.249481678009033,
      "learning_rate": 1.7785356975544875e-05,
      "loss": 2.730297660827637,
      "memory(GiB)": 77.56,
      "step": 84360,
      "token_acc": 0.47962382445141066,
      "train_speed(iter/s)": 1.437947
    },
    {
      "epoch": 3.61445525041772,
      "grad_norm": 5.133399963378906,
      "learning_rate": 1.7780210479178095e-05,
      "loss": 2.1141616821289064,
      "memory(GiB)": 77.56,
      "step": 84365,
      "token_acc": 0.5179856115107914,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 3.614669465746969,
      "grad_norm": 5.133951187133789,
      "learning_rate": 1.7775064566503008e-05,
      "loss": 2.4025285720825194,
      "memory(GiB)": 77.56,
      "step": 84370,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 3.614883681076218,
      "grad_norm": 5.603993892669678,
      "learning_rate": 1.776991923761287e-05,
      "loss": 2.3214759826660156,
      "memory(GiB)": 77.56,
      "step": 84375,
      "token_acc": 0.5165562913907285,
      "train_speed(iter/s)": 1.437962
    },
    {
      "epoch": 3.6150978964054667,
      "grad_norm": 4.5850725173950195,
      "learning_rate": 1.7764774492600866e-05,
      "loss": 2.2654388427734373,
      "memory(GiB)": 77.56,
      "step": 84380,
      "token_acc": 0.49846153846153846,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.615312111734716,
      "grad_norm": 5.483675003051758,
      "learning_rate": 1.775963033156023e-05,
      "loss": 2.4457557678222654,
      "memory(GiB)": 77.56,
      "step": 84385,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 3.615526327063965,
      "grad_norm": 5.81123161315918,
      "learning_rate": 1.7754486754584132e-05,
      "loss": 2.453587532043457,
      "memory(GiB)": 77.56,
      "step": 84390,
      "token_acc": 0.4887459807073955,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.6157405423932136,
      "grad_norm": 7.053272724151611,
      "learning_rate": 1.7749343761765753e-05,
      "loss": 2.4135324478149416,
      "memory(GiB)": 77.56,
      "step": 84395,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 3.615954757722463,
      "grad_norm": 5.910089492797852,
      "learning_rate": 1.7744201353198263e-05,
      "loss": 2.403470993041992,
      "memory(GiB)": 77.56,
      "step": 84400,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 3.6161689730517117,
      "grad_norm": 6.935621738433838,
      "learning_rate": 1.7739059528974815e-05,
      "loss": 2.397553253173828,
      "memory(GiB)": 77.56,
      "step": 84405,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.438046
    },
    {
      "epoch": 3.6163831883809605,
      "grad_norm": 4.800797939300537,
      "learning_rate": 1.7733918289188577e-05,
      "loss": 2.0946538925170897,
      "memory(GiB)": 77.56,
      "step": 84410,
      "token_acc": 0.54421768707483,
      "train_speed(iter/s)": 1.438052
    },
    {
      "epoch": 3.6165974037102098,
      "grad_norm": 5.917965888977051,
      "learning_rate": 1.7728777633932688e-05,
      "loss": 2.4258491516113283,
      "memory(GiB)": 77.56,
      "step": 84415,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.438047
    },
    {
      "epoch": 3.6168116190394586,
      "grad_norm": 7.704404830932617,
      "learning_rate": 1.7723637563300256e-05,
      "loss": 2.1151952743530273,
      "memory(GiB)": 77.56,
      "step": 84420,
      "token_acc": 0.553030303030303,
      "train_speed(iter/s)": 1.438062
    },
    {
      "epoch": 3.6170258343687074,
      "grad_norm": 4.804319858551025,
      "learning_rate": 1.7718498077384406e-05,
      "loss": 2.4303009033203127,
      "memory(GiB)": 77.56,
      "step": 84425,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.438057
    },
    {
      "epoch": 3.6172400496979566,
      "grad_norm": 5.899385452270508,
      "learning_rate": 1.771335917627825e-05,
      "loss": 2.1928693771362306,
      "memory(GiB)": 77.56,
      "step": 84430,
      "token_acc": 0.5243055555555556,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 3.6174542650272055,
      "grad_norm": 5.611340522766113,
      "learning_rate": 1.7708220860074858e-05,
      "loss": 2.2682941436767576,
      "memory(GiB)": 77.56,
      "step": 84435,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.438107
    },
    {
      "epoch": 3.6176684803564543,
      "grad_norm": 5.671706676483154,
      "learning_rate": 1.770308312886736e-05,
      "loss": 2.516790008544922,
      "memory(GiB)": 77.56,
      "step": 84440,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.438111
    },
    {
      "epoch": 3.6178826956857035,
      "grad_norm": 8.051642417907715,
      "learning_rate": 1.7697945982748786e-05,
      "loss": 2.4413047790527345,
      "memory(GiB)": 77.56,
      "step": 84445,
      "token_acc": 0.4766666666666667,
      "train_speed(iter/s)": 1.438125
    },
    {
      "epoch": 3.6180969110149523,
      "grad_norm": 6.651272773742676,
      "learning_rate": 1.7692809421812246e-05,
      "loss": 2.2289749145507813,
      "memory(GiB)": 77.56,
      "step": 84450,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.43812
    },
    {
      "epoch": 3.618311126344201,
      "grad_norm": 5.098021507263184,
      "learning_rate": 1.768767344615076e-05,
      "loss": 2.1872684478759767,
      "memory(GiB)": 77.56,
      "step": 84455,
      "token_acc": 0.519298245614035,
      "train_speed(iter/s)": 1.438137
    },
    {
      "epoch": 3.6185253416734504,
      "grad_norm": 6.062860488891602,
      "learning_rate": 1.768253805585739e-05,
      "loss": 2.5328508377075196,
      "memory(GiB)": 77.56,
      "step": 84460,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.438121
    },
    {
      "epoch": 3.6187395570026992,
      "grad_norm": 5.618769645690918,
      "learning_rate": 1.767740325102514e-05,
      "loss": 2.3572317123413087,
      "memory(GiB)": 77.56,
      "step": 84465,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.438114
    },
    {
      "epoch": 3.618953772331948,
      "grad_norm": 6.2001166343688965,
      "learning_rate": 1.7672269031747065e-05,
      "loss": 2.2721626281738283,
      "memory(GiB)": 77.56,
      "step": 84470,
      "token_acc": 0.5180327868852459,
      "train_speed(iter/s)": 1.438112
    },
    {
      "epoch": 3.6191679876611973,
      "grad_norm": 6.136464595794678,
      "learning_rate": 1.7667135398116163e-05,
      "loss": 2.1934186935424806,
      "memory(GiB)": 77.56,
      "step": 84475,
      "token_acc": 0.5282392026578073,
      "train_speed(iter/s)": 1.438122
    },
    {
      "epoch": 3.619382202990446,
      "grad_norm": 6.489969253540039,
      "learning_rate": 1.7662002350225437e-05,
      "loss": 2.4644866943359376,
      "memory(GiB)": 77.56,
      "step": 84480,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 3.619596418319695,
      "grad_norm": 5.997142791748047,
      "learning_rate": 1.765686988816787e-05,
      "loss": 2.6942161560058593,
      "memory(GiB)": 77.56,
      "step": 84485,
      "token_acc": 0.4426229508196721,
      "train_speed(iter/s)": 1.438074
    },
    {
      "epoch": 3.619810633648944,
      "grad_norm": 7.912371635437012,
      "learning_rate": 1.765173801203645e-05,
      "loss": 2.4193485260009764,
      "memory(GiB)": 77.56,
      "step": 84490,
      "token_acc": 0.44776119402985076,
      "train_speed(iter/s)": 1.438078
    },
    {
      "epoch": 3.620024848978193,
      "grad_norm": 6.506412029266357,
      "learning_rate": 1.7646606721924126e-05,
      "loss": 2.3007251739501955,
      "memory(GiB)": 77.56,
      "step": 84495,
      "token_acc": 0.5461538461538461,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 3.620239064307442,
      "grad_norm": 5.071538925170898,
      "learning_rate": 1.764147601792388e-05,
      "loss": 2.6214832305908202,
      "memory(GiB)": 77.56,
      "step": 84500,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438106
    },
    {
      "epoch": 3.620239064307442,
      "eval_loss": 2.177626848220825,
      "eval_runtime": 14.2723,
      "eval_samples_per_second": 7.007,
      "eval_steps_per_second": 7.007,
      "eval_token_acc": 0.4843962008141113,
      "step": 84500
    },
    {
      "epoch": 3.620453279636691,
      "grad_norm": 6.3451008796691895,
      "learning_rate": 1.7636345900128654e-05,
      "loss": 2.256543731689453,
      "memory(GiB)": 77.56,
      "step": 84505,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.62066749496594,
      "grad_norm": 5.751153945922852,
      "learning_rate": 1.763121636863138e-05,
      "loss": 2.4733028411865234,
      "memory(GiB)": 77.56,
      "step": 84510,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.6208817102951887,
      "grad_norm": 8.732616424560547,
      "learning_rate": 1.762608742352499e-05,
      "loss": 2.6257314682006836,
      "memory(GiB)": 77.56,
      "step": 84515,
      "token_acc": 0.46645367412140576,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.621095925624438,
      "grad_norm": 7.366602897644043,
      "learning_rate": 1.7620959064902376e-05,
      "loss": 2.6168935775756834,
      "memory(GiB)": 77.56,
      "step": 84520,
      "token_acc": 0.46956521739130436,
      "train_speed(iter/s)": 1.437763
    },
    {
      "epoch": 3.6213101409536868,
      "grad_norm": 4.783069610595703,
      "learning_rate": 1.7615831292856468e-05,
      "loss": 2.33459358215332,
      "memory(GiB)": 77.56,
      "step": 84525,
      "token_acc": 0.5047923322683706,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 3.6215243562829356,
      "grad_norm": 4.890158653259277,
      "learning_rate": 1.7610704107480165e-05,
      "loss": 2.083945465087891,
      "memory(GiB)": 77.56,
      "step": 84530,
      "token_acc": 0.5117647058823529,
      "train_speed(iter/s)": 1.437751
    },
    {
      "epoch": 3.621738571612185,
      "grad_norm": 6.235067844390869,
      "learning_rate": 1.760557750886634e-05,
      "loss": 2.520630645751953,
      "memory(GiB)": 77.56,
      "step": 84535,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.6219527869414336,
      "grad_norm": 5.803600311279297,
      "learning_rate": 1.7600451497107873e-05,
      "loss": 2.7453956604003906,
      "memory(GiB)": 77.56,
      "step": 84540,
      "token_acc": 0.42857142857142855,
      "train_speed(iter/s)": 1.43773
    },
    {
      "epoch": 3.6221670022706824,
      "grad_norm": 5.568145275115967,
      "learning_rate": 1.7595326072297615e-05,
      "loss": 2.4914594650268556,
      "memory(GiB)": 77.56,
      "step": 84545,
      "token_acc": 0.46496815286624205,
      "train_speed(iter/s)": 1.437738
    },
    {
      "epoch": 3.6223812175999317,
      "grad_norm": 4.958303451538086,
      "learning_rate": 1.759020123452841e-05,
      "loss": 2.3187862396240235,
      "memory(GiB)": 77.56,
      "step": 84550,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.437755
    },
    {
      "epoch": 3.6225954329291805,
      "grad_norm": 6.314785480499268,
      "learning_rate": 1.758507698389313e-05,
      "loss": 2.130345916748047,
      "memory(GiB)": 77.56,
      "step": 84555,
      "token_acc": 0.47440273037542663,
      "train_speed(iter/s)": 1.437768
    },
    {
      "epoch": 3.6228096482584293,
      "grad_norm": 4.95468807220459,
      "learning_rate": 1.757995332048458e-05,
      "loss": 2.4282060623168946,
      "memory(GiB)": 77.56,
      "step": 84560,
      "token_acc": 0.5016611295681063,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 3.6230238635876786,
      "grad_norm": 6.215773105621338,
      "learning_rate": 1.7574830244395597e-05,
      "loss": 2.0673656463623047,
      "memory(GiB)": 77.56,
      "step": 84565,
      "token_acc": 0.5537848605577689,
      "train_speed(iter/s)": 1.437761
    },
    {
      "epoch": 3.6232380789169274,
      "grad_norm": 9.88821029663086,
      "learning_rate": 1.756970775571898e-05,
      "loss": 2.43021297454834,
      "memory(GiB)": 77.56,
      "step": 84570,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.437782
    },
    {
      "epoch": 3.623452294246176,
      "grad_norm": 5.7987060546875,
      "learning_rate": 1.7564585854547522e-05,
      "loss": 2.2280540466308594,
      "memory(GiB)": 77.56,
      "step": 84575,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 3.6236665095754255,
      "grad_norm": 6.403611660003662,
      "learning_rate": 1.755946454097401e-05,
      "loss": 2.66406307220459,
      "memory(GiB)": 77.56,
      "step": 84580,
      "token_acc": 0.46408839779005523,
      "train_speed(iter/s)": 1.437788
    },
    {
      "epoch": 3.6238807249046743,
      "grad_norm": 6.059589862823486,
      "learning_rate": 1.755434381509124e-05,
      "loss": 2.095827102661133,
      "memory(GiB)": 77.56,
      "step": 84585,
      "token_acc": 0.5202952029520295,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 3.624094940233923,
      "grad_norm": 7.217907428741455,
      "learning_rate": 1.754922367699197e-05,
      "loss": 2.5607112884521483,
      "memory(GiB)": 77.56,
      "step": 84590,
      "token_acc": 0.4533333333333333,
      "train_speed(iter/s)": 1.437788
    },
    {
      "epoch": 3.6243091555631723,
      "grad_norm": 4.224649429321289,
      "learning_rate": 1.754410412676894e-05,
      "loss": 2.3494264602661135,
      "memory(GiB)": 77.56,
      "step": 84595,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 3.624523370892421,
      "grad_norm": 7.713440418243408,
      "learning_rate": 1.7538985164514925e-05,
      "loss": 2.3876441955566405,
      "memory(GiB)": 77.56,
      "step": 84600,
      "token_acc": 0.47419354838709676,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 3.62473758622167,
      "grad_norm": 4.943674087524414,
      "learning_rate": 1.7533866790322647e-05,
      "loss": 2.4379251480102537,
      "memory(GiB)": 77.56,
      "step": 84605,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.437794
    },
    {
      "epoch": 3.6249518015509192,
      "grad_norm": 5.520883560180664,
      "learning_rate": 1.7528749004284812e-05,
      "loss": 2.2836917877197265,
      "memory(GiB)": 77.56,
      "step": 84610,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 3.625166016880168,
      "grad_norm": 7.87253475189209,
      "learning_rate": 1.752363180649416e-05,
      "loss": 2.5776390075683593,
      "memory(GiB)": 77.56,
      "step": 84615,
      "token_acc": 0.5272727272727272,
      "train_speed(iter/s)": 1.437794
    },
    {
      "epoch": 3.625380232209417,
      "grad_norm": 7.26462459564209,
      "learning_rate": 1.7518515197043383e-05,
      "loss": 2.5163551330566407,
      "memory(GiB)": 77.56,
      "step": 84620,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 3.625594447538666,
      "grad_norm": 4.841714382171631,
      "learning_rate": 1.751339917602518e-05,
      "loss": 2.1523181915283205,
      "memory(GiB)": 77.56,
      "step": 84625,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 3.625808662867915,
      "grad_norm": 5.723125457763672,
      "learning_rate": 1.7508283743532223e-05,
      "loss": 2.596650505065918,
      "memory(GiB)": 77.56,
      "step": 84630,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 3.6260228781971637,
      "grad_norm": 5.5952467918396,
      "learning_rate": 1.7503168899657184e-05,
      "loss": 2.447065544128418,
      "memory(GiB)": 77.56,
      "step": 84635,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.437814
    },
    {
      "epoch": 3.626237093526413,
      "grad_norm": 5.607433319091797,
      "learning_rate": 1.749805464449271e-05,
      "loss": 2.4382219314575195,
      "memory(GiB)": 77.56,
      "step": 84640,
      "token_acc": 0.46417445482866043,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 3.626451308855662,
      "grad_norm": 7.0985918045043945,
      "learning_rate": 1.7492940978131478e-05,
      "loss": 2.4026002883911133,
      "memory(GiB)": 77.56,
      "step": 84645,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 3.6266655241849106,
      "grad_norm": 6.141039848327637,
      "learning_rate": 1.748782790066612e-05,
      "loss": 2.002517318725586,
      "memory(GiB)": 77.56,
      "step": 84650,
      "token_acc": 0.5990338164251208,
      "train_speed(iter/s)": 1.437823
    },
    {
      "epoch": 3.62687973951416,
      "grad_norm": 5.390111446380615,
      "learning_rate": 1.7482715412189253e-05,
      "loss": 2.1562688827514647,
      "memory(GiB)": 77.56,
      "step": 84655,
      "token_acc": 0.546031746031746,
      "train_speed(iter/s)": 1.437824
    },
    {
      "epoch": 3.6270939548434087,
      "grad_norm": 6.4040303230285645,
      "learning_rate": 1.7477603512793504e-05,
      "loss": 2.0433748245239256,
      "memory(GiB)": 77.56,
      "step": 84660,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.4378
    },
    {
      "epoch": 3.6273081701726575,
      "grad_norm": 6.350438594818115,
      "learning_rate": 1.747249220257146e-05,
      "loss": 2.2895782470703123,
      "memory(GiB)": 77.56,
      "step": 84665,
      "token_acc": 0.536741214057508,
      "train_speed(iter/s)": 1.437808
    },
    {
      "epoch": 3.6275223855019068,
      "grad_norm": 5.343311786651611,
      "learning_rate": 1.746738148161573e-05,
      "loss": 2.2991817474365233,
      "memory(GiB)": 77.56,
      "step": 84670,
      "token_acc": 0.5176056338028169,
      "train_speed(iter/s)": 1.437817
    },
    {
      "epoch": 3.6277366008311556,
      "grad_norm": 5.636126518249512,
      "learning_rate": 1.746227135001892e-05,
      "loss": 2.261979675292969,
      "memory(GiB)": 77.56,
      "step": 84675,
      "token_acc": 0.5311572700296736,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 3.6279508161604044,
      "grad_norm": 5.327690124511719,
      "learning_rate": 1.745716180787359e-05,
      "loss": 2.309015083312988,
      "memory(GiB)": 77.56,
      "step": 84680,
      "token_acc": 0.5230263157894737,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 3.6281650314896536,
      "grad_norm": 5.751166343688965,
      "learning_rate": 1.7452052855272298e-05,
      "loss": 2.213973617553711,
      "memory(GiB)": 77.56,
      "step": 84685,
      "token_acc": 0.48514851485148514,
      "train_speed(iter/s)": 1.437855
    },
    {
      "epoch": 3.6283792468189024,
      "grad_norm": 5.247895240783691,
      "learning_rate": 1.7446944492307605e-05,
      "loss": 2.385271453857422,
      "memory(GiB)": 77.56,
      "step": 84690,
      "token_acc": 0.5174825174825175,
      "train_speed(iter/s)": 1.437852
    },
    {
      "epoch": 3.6285934621481513,
      "grad_norm": 7.145521640777588,
      "learning_rate": 1.744183671907203e-05,
      "loss": 2.483266830444336,
      "memory(GiB)": 77.56,
      "step": 84695,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 3.6288076774774005,
      "grad_norm": 6.7839789390563965,
      "learning_rate": 1.743672953565814e-05,
      "loss": 2.3552433013916017,
      "memory(GiB)": 77.56,
      "step": 84700,
      "token_acc": 0.5140845070422535,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 3.6290218928066493,
      "grad_norm": 7.0621442794799805,
      "learning_rate": 1.7431622942158445e-05,
      "loss": 2.123543357849121,
      "memory(GiB)": 77.56,
      "step": 84705,
      "token_acc": 0.5403225806451613,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.629236108135898,
      "grad_norm": 6.648751258850098,
      "learning_rate": 1.742651693866545e-05,
      "loss": 2.3292205810546873,
      "memory(GiB)": 77.56,
      "step": 84710,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.437844
    },
    {
      "epoch": 3.6294503234651474,
      "grad_norm": 4.871461391448975,
      "learning_rate": 1.742141152527166e-05,
      "loss": 2.1896108627319335,
      "memory(GiB)": 77.56,
      "step": 84715,
      "token_acc": 0.5236486486486487,
      "train_speed(iter/s)": 1.437863
    },
    {
      "epoch": 3.629664538794396,
      "grad_norm": 5.25560188293457,
      "learning_rate": 1.7416306702069558e-05,
      "loss": 2.3394748687744142,
      "memory(GiB)": 77.56,
      "step": 84720,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.629878754123645,
      "grad_norm": 7.446652412414551,
      "learning_rate": 1.7411202469151606e-05,
      "loss": 2.706114959716797,
      "memory(GiB)": 77.56,
      "step": 84725,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 3.6300929694528943,
      "grad_norm": 10.241846084594727,
      "learning_rate": 1.7406098826610313e-05,
      "loss": 2.4451271057128907,
      "memory(GiB)": 77.56,
      "step": 84730,
      "token_acc": 0.5074074074074074,
      "train_speed(iter/s)": 1.437922
    },
    {
      "epoch": 3.630307184782143,
      "grad_norm": 5.8900933265686035,
      "learning_rate": 1.7400995774538116e-05,
      "loss": 2.5738042831420898,
      "memory(GiB)": 77.56,
      "step": 84735,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 3.630521400111392,
      "grad_norm": 6.171445846557617,
      "learning_rate": 1.739589331302744e-05,
      "loss": 2.5489086151123046,
      "memory(GiB)": 77.56,
      "step": 84740,
      "token_acc": 0.4461538461538462,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 3.630735615440641,
      "grad_norm": 9.676703453063965,
      "learning_rate": 1.7390791442170763e-05,
      "loss": 2.5274946212768556,
      "memory(GiB)": 77.56,
      "step": 84745,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.437922
    },
    {
      "epoch": 3.63094983076989,
      "grad_norm": 6.300182819366455,
      "learning_rate": 1.7385690162060487e-05,
      "loss": 2.36279296875,
      "memory(GiB)": 77.56,
      "step": 84750,
      "token_acc": 0.5207547169811321,
      "train_speed(iter/s)": 1.437939
    },
    {
      "epoch": 3.631164046099139,
      "grad_norm": 7.311707973480225,
      "learning_rate": 1.7380589472789015e-05,
      "loss": 2.4290876388549805,
      "memory(GiB)": 77.56,
      "step": 84755,
      "token_acc": 0.5038461538461538,
      "train_speed(iter/s)": 1.437954
    },
    {
      "epoch": 3.631378261428388,
      "grad_norm": 5.435907363891602,
      "learning_rate": 1.7375489374448776e-05,
      "loss": 2.5090909957885743,
      "memory(GiB)": 77.56,
      "step": 84760,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437943
    },
    {
      "epoch": 3.631592476757637,
      "grad_norm": 6.9370222091674805,
      "learning_rate": 1.7370389867132154e-05,
      "loss": 2.445126533508301,
      "memory(GiB)": 77.56,
      "step": 84765,
      "token_acc": 0.4734982332155477,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 3.6318066920868857,
      "grad_norm": 5.734540939331055,
      "learning_rate": 1.7365290950931527e-05,
      "loss": 2.336766242980957,
      "memory(GiB)": 77.56,
      "step": 84770,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.43794
    },
    {
      "epoch": 3.632020907416135,
      "grad_norm": 6.48061990737915,
      "learning_rate": 1.736019262593927e-05,
      "loss": 2.2358119964599608,
      "memory(GiB)": 77.56,
      "step": 84775,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.437922
    },
    {
      "epoch": 3.6322351227453837,
      "grad_norm": 6.097958564758301,
      "learning_rate": 1.7355094892247725e-05,
      "loss": 2.314980316162109,
      "memory(GiB)": 77.56,
      "step": 84780,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 3.6324493380746325,
      "grad_norm": 4.919453144073486,
      "learning_rate": 1.7349997749949277e-05,
      "loss": 2.359080696105957,
      "memory(GiB)": 77.56,
      "step": 84785,
      "token_acc": 0.49842271293375395,
      "train_speed(iter/s)": 1.437904
    },
    {
      "epoch": 3.632663553403882,
      "grad_norm": 5.391432285308838,
      "learning_rate": 1.734490119913625e-05,
      "loss": 2.3844417572021483,
      "memory(GiB)": 77.56,
      "step": 84790,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.43789
    },
    {
      "epoch": 3.6328777687331306,
      "grad_norm": 6.963891983032227,
      "learning_rate": 1.7339805239900964e-05,
      "loss": 2.0690610885620115,
      "memory(GiB)": 77.56,
      "step": 84795,
      "token_acc": 0.562015503875969,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 3.6330919840623794,
      "grad_norm": 4.100226879119873,
      "learning_rate": 1.7334709872335746e-05,
      "loss": 2.102535629272461,
      "memory(GiB)": 77.56,
      "step": 84800,
      "token_acc": 0.5482866043613707,
      "train_speed(iter/s)": 1.43789
    },
    {
      "epoch": 3.6333061993916287,
      "grad_norm": 5.69957971572876,
      "learning_rate": 1.7329615096532905e-05,
      "loss": 2.4984720230102537,
      "memory(GiB)": 77.56,
      "step": 84805,
      "token_acc": 0.47151898734177217,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 3.6335204147208775,
      "grad_norm": 6.949384689331055,
      "learning_rate": 1.7324520912584713e-05,
      "loss": 2.5807977676391602,
      "memory(GiB)": 77.56,
      "step": 84810,
      "token_acc": 0.48615384615384616,
      "train_speed(iter/s)": 1.437915
    },
    {
      "epoch": 3.6337346300501263,
      "grad_norm": 6.165522575378418,
      "learning_rate": 1.731942732058348e-05,
      "loss": 2.399928855895996,
      "memory(GiB)": 77.56,
      "step": 84815,
      "token_acc": 0.5015105740181269,
      "train_speed(iter/s)": 1.437905
    },
    {
      "epoch": 3.6339488453793756,
      "grad_norm": 5.983424663543701,
      "learning_rate": 1.7314334320621494e-05,
      "loss": 2.0674776077270507,
      "memory(GiB)": 77.56,
      "step": 84820,
      "token_acc": 0.5826086956521739,
      "train_speed(iter/s)": 1.43791
    },
    {
      "epoch": 3.6341630607086244,
      "grad_norm": 5.636471271514893,
      "learning_rate": 1.7309241912791e-05,
      "loss": 2.504780578613281,
      "memory(GiB)": 77.56,
      "step": 84825,
      "token_acc": 0.49032258064516127,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 3.634377276037873,
      "grad_norm": 8.242146492004395,
      "learning_rate": 1.7304150097184257e-05,
      "loss": 2.1323854446411135,
      "memory(GiB)": 77.56,
      "step": 84830,
      "token_acc": 0.4957983193277311,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 3.6345914913671225,
      "grad_norm": 6.046113967895508,
      "learning_rate": 1.7299058873893507e-05,
      "loss": 2.3913726806640625,
      "memory(GiB)": 77.56,
      "step": 84835,
      "token_acc": 0.4923547400611621,
      "train_speed(iter/s)": 1.437949
    },
    {
      "epoch": 3.6348057066963713,
      "grad_norm": 6.416695594787598,
      "learning_rate": 1.7293968243010967e-05,
      "loss": 2.4320699691772463,
      "memory(GiB)": 77.56,
      "step": 84840,
      "token_acc": 0.5387453874538746,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 3.63501992202562,
      "grad_norm": 5.738819599151611,
      "learning_rate": 1.728887820462889e-05,
      "loss": 2.441440963745117,
      "memory(GiB)": 77.56,
      "step": 84845,
      "token_acc": 0.4676923076923077,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.6352341373548693,
      "grad_norm": 5.451039791107178,
      "learning_rate": 1.7283788758839463e-05,
      "loss": 2.337528419494629,
      "memory(GiB)": 77.56,
      "step": 84850,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.437993
    },
    {
      "epoch": 3.635448352684118,
      "grad_norm": 6.397180080413818,
      "learning_rate": 1.7278699905734897e-05,
      "loss": 2.381456184387207,
      "memory(GiB)": 77.56,
      "step": 84855,
      "token_acc": 0.5138461538461538,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 3.635662568013367,
      "grad_norm": 5.56934118270874,
      "learning_rate": 1.727361164540738e-05,
      "loss": 2.3892595291137697,
      "memory(GiB)": 77.56,
      "step": 84860,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.635876783342616,
      "grad_norm": 7.340134620666504,
      "learning_rate": 1.7268523977949084e-05,
      "loss": 2.379293441772461,
      "memory(GiB)": 77.56,
      "step": 84865,
      "token_acc": 0.45180722891566266,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 3.636090998671865,
      "grad_norm": 7.421058177947998,
      "learning_rate": 1.7263436903452162e-05,
      "loss": 2.2533777236938475,
      "memory(GiB)": 77.56,
      "step": 84870,
      "token_acc": 0.5164473684210527,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 3.636305214001114,
      "grad_norm": 6.16189432144165,
      "learning_rate": 1.7258350422008813e-05,
      "loss": 2.3028507232666016,
      "memory(GiB)": 77.56,
      "step": 84875,
      "token_acc": 0.5419847328244275,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 3.636519429330363,
      "grad_norm": 9.566047668457031,
      "learning_rate": 1.7253264533711155e-05,
      "loss": 2.36781005859375,
      "memory(GiB)": 77.56,
      "step": 84880,
      "token_acc": 0.5265151515151515,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.636733644659612,
      "grad_norm": 4.463603973388672,
      "learning_rate": 1.7248179238651334e-05,
      "loss": 2.5948728561401366,
      "memory(GiB)": 77.56,
      "step": 84885,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.6369478599888607,
      "grad_norm": 9.498019218444824,
      "learning_rate": 1.724309453692145e-05,
      "loss": 2.3666923522949217,
      "memory(GiB)": 77.56,
      "step": 84890,
      "token_acc": 0.4807017543859649,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 3.63716207531811,
      "grad_norm": 5.189406394958496,
      "learning_rate": 1.7238010428613654e-05,
      "loss": 2.6930814743041993,
      "memory(GiB)": 77.56,
      "step": 84895,
      "token_acc": 0.47214076246334313,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 3.637376290647359,
      "grad_norm": 6.054851531982422,
      "learning_rate": 1.7232926913820015e-05,
      "loss": 2.129806709289551,
      "memory(GiB)": 77.56,
      "step": 84900,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 3.6375905059766076,
      "grad_norm": 6.123406887054443,
      "learning_rate": 1.7227843992632658e-05,
      "loss": 2.1641653060913084,
      "memory(GiB)": 77.56,
      "step": 84905,
      "token_acc": 0.5573770491803278,
      "train_speed(iter/s)": 1.43804
    },
    {
      "epoch": 3.637804721305857,
      "grad_norm": 5.806746482849121,
      "learning_rate": 1.7222761665143654e-05,
      "loss": 2.186139678955078,
      "memory(GiB)": 77.56,
      "step": 84910,
      "token_acc": 0.49851632047477745,
      "train_speed(iter/s)": 1.438034
    },
    {
      "epoch": 3.6380189366351057,
      "grad_norm": 5.628856182098389,
      "learning_rate": 1.7217679931445068e-05,
      "loss": 2.370722770690918,
      "memory(GiB)": 77.56,
      "step": 84915,
      "token_acc": 0.5288461538461539,
      "train_speed(iter/s)": 1.438036
    },
    {
      "epoch": 3.6382331519643545,
      "grad_norm": 5.779271602630615,
      "learning_rate": 1.721259879162896e-05,
      "loss": 2.1186973571777346,
      "memory(GiB)": 77.56,
      "step": 84920,
      "token_acc": 0.5443425076452599,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 3.6384473672936037,
      "grad_norm": 5.749781608581543,
      "learning_rate": 1.720751824578737e-05,
      "loss": 1.9772274017333984,
      "memory(GiB)": 77.56,
      "step": 84925,
      "token_acc": 0.5932835820895522,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 3.6386615826228526,
      "grad_norm": 6.004026889801025,
      "learning_rate": 1.7202438294012363e-05,
      "loss": 2.396786117553711,
      "memory(GiB)": 77.56,
      "step": 84930,
      "token_acc": 0.4972067039106145,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 3.6388757979521014,
      "grad_norm": 5.7338433265686035,
      "learning_rate": 1.719735893639595e-05,
      "loss": 2.284738540649414,
      "memory(GiB)": 77.56,
      "step": 84935,
      "token_acc": 0.46846846846846846,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 3.6390900132813506,
      "grad_norm": 5.394063949584961,
      "learning_rate": 1.7192280173030156e-05,
      "loss": 2.44345703125,
      "memory(GiB)": 77.56,
      "step": 84940,
      "token_acc": 0.4364820846905538,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 3.6393042286105994,
      "grad_norm": 6.275918006896973,
      "learning_rate": 1.718720200400698e-05,
      "loss": 2.5629690170288084,
      "memory(GiB)": 77.56,
      "step": 84945,
      "token_acc": 0.4477124183006536,
      "train_speed(iter/s)": 1.438007
    },
    {
      "epoch": 3.6395184439398482,
      "grad_norm": 5.952244281768799,
      "learning_rate": 1.718212442941842e-05,
      "loss": 2.2209205627441406,
      "memory(GiB)": 77.56,
      "step": 84950,
      "token_acc": 0.5536332179930796,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 3.6397326592690975,
      "grad_norm": 6.568378448486328,
      "learning_rate": 1.7177047449356447e-05,
      "loss": 2.4038665771484373,
      "memory(GiB)": 77.56,
      "step": 84955,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.438035
    },
    {
      "epoch": 3.6399468745983463,
      "grad_norm": 6.705170154571533,
      "learning_rate": 1.7171971063913066e-05,
      "loss": 2.2902069091796875,
      "memory(GiB)": 77.56,
      "step": 84960,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438019
    },
    {
      "epoch": 3.640161089927595,
      "grad_norm": 5.389596939086914,
      "learning_rate": 1.716689527318021e-05,
      "loss": 2.3979835510253906,
      "memory(GiB)": 77.56,
      "step": 84965,
      "token_acc": 0.5327102803738317,
      "train_speed(iter/s)": 1.438026
    },
    {
      "epoch": 3.6403753052568444,
      "grad_norm": 6.294544696807861,
      "learning_rate": 1.7161820077249856e-05,
      "loss": 2.026536750793457,
      "memory(GiB)": 77.56,
      "step": 84970,
      "token_acc": 0.5480427046263345,
      "train_speed(iter/s)": 1.43804
    },
    {
      "epoch": 3.640589520586093,
      "grad_norm": 5.4064621925354,
      "learning_rate": 1.715674547621394e-05,
      "loss": 1.979637908935547,
      "memory(GiB)": 77.56,
      "step": 84975,
      "token_acc": 0.5625,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 3.640803735915342,
      "grad_norm": 7.404888153076172,
      "learning_rate": 1.7151671470164392e-05,
      "loss": 2.1301578521728515,
      "memory(GiB)": 77.56,
      "step": 84980,
      "token_acc": 0.5020746887966805,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 3.6410179512445913,
      "grad_norm": 7.268360614776611,
      "learning_rate": 1.7146598059193113e-05,
      "loss": 2.24462890625,
      "memory(GiB)": 77.56,
      "step": 84985,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.438032
    },
    {
      "epoch": 3.64123216657384,
      "grad_norm": 6.248690128326416,
      "learning_rate": 1.714152524339204e-05,
      "loss": 2.2621837615966798,
      "memory(GiB)": 77.56,
      "step": 84990,
      "token_acc": 0.5340501792114696,
      "train_speed(iter/s)": 1.438039
    },
    {
      "epoch": 3.641446381903089,
      "grad_norm": 5.92620849609375,
      "learning_rate": 1.7136453022853067e-05,
      "loss": 2.410206413269043,
      "memory(GiB)": 77.56,
      "step": 84995,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 3.641660597232338,
      "grad_norm": 6.842756271362305,
      "learning_rate": 1.7131381397668072e-05,
      "loss": 2.3542247772216798,
      "memory(GiB)": 77.56,
      "step": 85000,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 3.641660597232338,
      "eval_loss": 2.196579694747925,
      "eval_runtime": 14.3309,
      "eval_samples_per_second": 6.978,
      "eval_steps_per_second": 6.978,
      "eval_token_acc": 0.48005698005698005,
      "step": 85000
    },
    {
      "epoch": 3.641874812561587,
      "grad_norm": 6.053701877593994,
      "learning_rate": 1.7126310367928937e-05,
      "loss": 2.034068298339844,
      "memory(GiB)": 77.56,
      "step": 85005,
      "token_acc": 0.49589322381930184,
      "train_speed(iter/s)": 1.437723
    },
    {
      "epoch": 3.6420890278908358,
      "grad_norm": 6.0152764320373535,
      "learning_rate": 1.712123993372753e-05,
      "loss": 2.191161346435547,
      "memory(GiB)": 77.56,
      "step": 85010,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 3.642303243220085,
      "grad_norm": 6.775215148925781,
      "learning_rate": 1.7116170095155687e-05,
      "loss": 2.4492094039916994,
      "memory(GiB)": 77.56,
      "step": 85015,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 3.642517458549334,
      "grad_norm": 6.784506320953369,
      "learning_rate": 1.7111100852305283e-05,
      "loss": 2.437163543701172,
      "memory(GiB)": 77.56,
      "step": 85020,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.437746
    },
    {
      "epoch": 3.6427316738785827,
      "grad_norm": 4.5357160568237305,
      "learning_rate": 1.710603220526814e-05,
      "loss": 2.5728427886962892,
      "memory(GiB)": 77.56,
      "step": 85025,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.642945889207832,
      "grad_norm": 5.133059024810791,
      "learning_rate": 1.7100964154136086e-05,
      "loss": 2.4496696472167967,
      "memory(GiB)": 77.56,
      "step": 85030,
      "token_acc": 0.5015479876160991,
      "train_speed(iter/s)": 1.437733
    },
    {
      "epoch": 3.6431601045370807,
      "grad_norm": 7.43684720993042,
      "learning_rate": 1.7095896699000908e-05,
      "loss": 2.41947078704834,
      "memory(GiB)": 77.56,
      "step": 85035,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 3.6433743198663295,
      "grad_norm": 5.835127830505371,
      "learning_rate": 1.709082983995444e-05,
      "loss": 2.408224868774414,
      "memory(GiB)": 77.56,
      "step": 85040,
      "token_acc": 0.47262247838616717,
      "train_speed(iter/s)": 1.437742
    },
    {
      "epoch": 3.643588535195579,
      "grad_norm": 7.0789995193481445,
      "learning_rate": 1.7085763577088453e-05,
      "loss": 2.179537773132324,
      "memory(GiB)": 77.56,
      "step": 85045,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.437763
    },
    {
      "epoch": 3.6438027505248276,
      "grad_norm": 4.962024688720703,
      "learning_rate": 1.708069791049475e-05,
      "loss": 2.0523859024047852,
      "memory(GiB)": 77.56,
      "step": 85050,
      "token_acc": 0.5477178423236515,
      "train_speed(iter/s)": 1.437765
    },
    {
      "epoch": 3.6440169658540764,
      "grad_norm": 6.24876594543457,
      "learning_rate": 1.7075632840265083e-05,
      "loss": 2.6402164459228517,
      "memory(GiB)": 77.56,
      "step": 85055,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 3.6442311811833257,
      "grad_norm": 5.73577356338501,
      "learning_rate": 1.7070568366491208e-05,
      "loss": 2.1512435913085937,
      "memory(GiB)": 77.56,
      "step": 85060,
      "token_acc": 0.5322033898305085,
      "train_speed(iter/s)": 1.437788
    },
    {
      "epoch": 3.6444453965125745,
      "grad_norm": 5.112829685211182,
      "learning_rate": 1.7065504489264884e-05,
      "loss": 2.4789907455444338,
      "memory(GiB)": 77.56,
      "step": 85065,
      "token_acc": 0.5043731778425656,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 3.6446596118418233,
      "grad_norm": 4.949240684509277,
      "learning_rate": 1.7060441208677818e-05,
      "loss": 2.287200927734375,
      "memory(GiB)": 77.56,
      "step": 85070,
      "token_acc": 0.5449101796407185,
      "train_speed(iter/s)": 1.437785
    },
    {
      "epoch": 3.6448738271710726,
      "grad_norm": 4.728801727294922,
      "learning_rate": 1.705537852482178e-05,
      "loss": 2.389324188232422,
      "memory(GiB)": 77.56,
      "step": 85075,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.437767
    },
    {
      "epoch": 3.6450880425003214,
      "grad_norm": 5.657744884490967,
      "learning_rate": 1.7050316437788466e-05,
      "loss": 2.286519241333008,
      "memory(GiB)": 77.56,
      "step": 85080,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 3.64530225782957,
      "grad_norm": 7.365163326263428,
      "learning_rate": 1.704525494766958e-05,
      "loss": 2.3439968109130858,
      "memory(GiB)": 77.56,
      "step": 85085,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 3.6455164731588194,
      "grad_norm": 6.09114933013916,
      "learning_rate": 1.7040194054556813e-05,
      "loss": 2.3588485717773438,
      "memory(GiB)": 77.56,
      "step": 85090,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.6457306884880682,
      "grad_norm": 4.982217788696289,
      "learning_rate": 1.703513375854185e-05,
      "loss": 2.2544700622558596,
      "memory(GiB)": 77.56,
      "step": 85095,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.437784
    },
    {
      "epoch": 3.645944903817317,
      "grad_norm": 6.5849609375,
      "learning_rate": 1.7030074059716338e-05,
      "loss": 2.165945625305176,
      "memory(GiB)": 77.56,
      "step": 85100,
      "token_acc": 0.5065359477124183,
      "train_speed(iter/s)": 1.4378
    },
    {
      "epoch": 3.6461591191465663,
      "grad_norm": 6.173681259155273,
      "learning_rate": 1.7025014958171993e-05,
      "loss": 2.4783203125,
      "memory(GiB)": 77.56,
      "step": 85105,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.437786
    },
    {
      "epoch": 3.646373334475815,
      "grad_norm": 5.841230869293213,
      "learning_rate": 1.7019956454000406e-05,
      "loss": 2.4201770782470704,
      "memory(GiB)": 77.56,
      "step": 85110,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.437771
    },
    {
      "epoch": 3.646587549805064,
      "grad_norm": 6.165286064147949,
      "learning_rate": 1.7014898547293266e-05,
      "loss": 2.2794252395629884,
      "memory(GiB)": 77.56,
      "step": 85115,
      "token_acc": 0.5527156549520766,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 3.646801765134313,
      "grad_norm": 8.642879486083984,
      "learning_rate": 1.7009841238142188e-05,
      "loss": 2.401658821105957,
      "memory(GiB)": 77.56,
      "step": 85120,
      "token_acc": 0.5316455696202531,
      "train_speed(iter/s)": 1.437816
    },
    {
      "epoch": 3.647015980463562,
      "grad_norm": 6.1140828132629395,
      "learning_rate": 1.7004784526638777e-05,
      "loss": 2.490953254699707,
      "memory(GiB)": 77.56,
      "step": 85125,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 3.647230195792811,
      "grad_norm": 5.107372283935547,
      "learning_rate": 1.6999728412874632e-05,
      "loss": 2.122241973876953,
      "memory(GiB)": 77.56,
      "step": 85130,
      "token_acc": 0.5328467153284672,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.64744441112206,
      "grad_norm": 7.959492206573486,
      "learning_rate": 1.6994672896941377e-05,
      "loss": 2.4018396377563476,
      "memory(GiB)": 77.56,
      "step": 85135,
      "token_acc": 0.5146443514644351,
      "train_speed(iter/s)": 1.437852
    },
    {
      "epoch": 3.647658626451309,
      "grad_norm": 6.337654113769531,
      "learning_rate": 1.6989617978930587e-05,
      "loss": 2.513269805908203,
      "memory(GiB)": 77.56,
      "step": 85140,
      "token_acc": 0.447887323943662,
      "train_speed(iter/s)": 1.437858
    },
    {
      "epoch": 3.6478728417805577,
      "grad_norm": 7.01861572265625,
      "learning_rate": 1.6984563658933834e-05,
      "loss": 2.5734073638916017,
      "memory(GiB)": 77.56,
      "step": 85145,
      "token_acc": 0.4402730375426621,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 3.648087057109807,
      "grad_norm": 6.410262584686279,
      "learning_rate": 1.697950993704268e-05,
      "loss": 2.494732666015625,
      "memory(GiB)": 77.56,
      "step": 85150,
      "token_acc": 0.5169811320754717,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 3.6483012724390558,
      "grad_norm": 6.100495338439941,
      "learning_rate": 1.6974456813348678e-05,
      "loss": 2.44498291015625,
      "memory(GiB)": 77.56,
      "step": 85155,
      "token_acc": 0.48299319727891155,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.6485154877683046,
      "grad_norm": 5.960434436798096,
      "learning_rate": 1.6969404287943356e-05,
      "loss": 2.3965576171875,
      "memory(GiB)": 77.56,
      "step": 85160,
      "token_acc": 0.5145985401459854,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 3.648729703097554,
      "grad_norm": 5.544156074523926,
      "learning_rate": 1.6964352360918277e-05,
      "loss": 2.4419994354248047,
      "memory(GiB)": 77.56,
      "step": 85165,
      "token_acc": 0.4925925925925926,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 3.6489439184268027,
      "grad_norm": 7.668413162231445,
      "learning_rate": 1.6959301032364944e-05,
      "loss": 2.1129375457763673,
      "memory(GiB)": 77.56,
      "step": 85170,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 3.6491581337560515,
      "grad_norm": 8.6959867477417,
      "learning_rate": 1.695425030237487e-05,
      "loss": 2.2139163970947267,
      "memory(GiB)": 77.56,
      "step": 85175,
      "token_acc": 0.5070422535211268,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.6493723490853007,
      "grad_norm": 6.109446048736572,
      "learning_rate": 1.6949200171039547e-05,
      "loss": 2.3366411209106444,
      "memory(GiB)": 77.56,
      "step": 85180,
      "token_acc": 0.4899328859060403,
      "train_speed(iter/s)": 1.437862
    },
    {
      "epoch": 3.6495865644145495,
      "grad_norm": 5.506080150604248,
      "learning_rate": 1.6944150638450457e-05,
      "loss": 2.542353057861328,
      "memory(GiB)": 77.56,
      "step": 85185,
      "token_acc": 0.446875,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 3.6498007797437984,
      "grad_norm": 5.778169631958008,
      "learning_rate": 1.6939101704699077e-05,
      "loss": 2.217597007751465,
      "memory(GiB)": 77.56,
      "step": 85190,
      "token_acc": 0.5412186379928315,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 3.6500149950730476,
      "grad_norm": 5.550141334533691,
      "learning_rate": 1.6934053369876912e-05,
      "loss": 2.356995964050293,
      "memory(GiB)": 77.56,
      "step": 85195,
      "token_acc": 0.4859154929577465,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.6502292104022964,
      "grad_norm": 4.3431715965271,
      "learning_rate": 1.692900563407538e-05,
      "loss": 2.5239662170410155,
      "memory(GiB)": 77.56,
      "step": 85200,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.437874
    },
    {
      "epoch": 3.6504434257315452,
      "grad_norm": 5.0271172523498535,
      "learning_rate": 1.692395849738594e-05,
      "loss": 2.307295227050781,
      "memory(GiB)": 77.56,
      "step": 85205,
      "token_acc": 0.5115511551155115,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 3.6506576410607945,
      "grad_norm": 5.528509616851807,
      "learning_rate": 1.6918911959900015e-05,
      "loss": 2.1722105026245115,
      "memory(GiB)": 77.56,
      "step": 85210,
      "token_acc": 0.5127118644067796,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 3.6508718563900433,
      "grad_norm": 5.842065811157227,
      "learning_rate": 1.691386602170902e-05,
      "loss": 2.196075439453125,
      "memory(GiB)": 77.56,
      "step": 85215,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.437903
    },
    {
      "epoch": 3.651086071719292,
      "grad_norm": 5.357964992523193,
      "learning_rate": 1.6908820682904387e-05,
      "loss": 2.408620071411133,
      "memory(GiB)": 77.56,
      "step": 85220,
      "token_acc": 0.4984126984126984,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 3.6513002870485414,
      "grad_norm": 6.5117106437683105,
      "learning_rate": 1.6903775943577516e-05,
      "loss": 2.6389820098876955,
      "memory(GiB)": 77.56,
      "step": 85225,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 3.65151450237779,
      "grad_norm": 4.985795021057129,
      "learning_rate": 1.689873180381979e-05,
      "loss": 2.1541330337524416,
      "memory(GiB)": 77.56,
      "step": 85230,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 3.651728717707039,
      "grad_norm": 6.050500869750977,
      "learning_rate": 1.689368826372259e-05,
      "loss": 2.3743358612060548,
      "memory(GiB)": 77.56,
      "step": 85235,
      "token_acc": 0.5136186770428015,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.6519429330362883,
      "grad_norm": 4.540327072143555,
      "learning_rate": 1.6888645323377283e-05,
      "loss": 2.4179924011230467,
      "memory(GiB)": 77.56,
      "step": 85240,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.437913
    },
    {
      "epoch": 3.652157148365537,
      "grad_norm": 5.712406635284424,
      "learning_rate": 1.6883602982875207e-05,
      "loss": 2.0408262252807616,
      "memory(GiB)": 77.56,
      "step": 85245,
      "token_acc": 0.5481171548117155,
      "train_speed(iter/s)": 1.43791
    },
    {
      "epoch": 3.652371363694786,
      "grad_norm": 3.920994758605957,
      "learning_rate": 1.687856124230775e-05,
      "loss": 2.556990051269531,
      "memory(GiB)": 77.56,
      "step": 85250,
      "token_acc": 0.45931758530183725,
      "train_speed(iter/s)": 1.437913
    },
    {
      "epoch": 3.652585579024035,
      "grad_norm": 6.97440767288208,
      "learning_rate": 1.6873520101766223e-05,
      "loss": 2.2614572525024412,
      "memory(GiB)": 77.56,
      "step": 85255,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.4379
    },
    {
      "epoch": 3.652799794353284,
      "grad_norm": 4.942814826965332,
      "learning_rate": 1.6868479561341936e-05,
      "loss": 2.188820457458496,
      "memory(GiB)": 77.56,
      "step": 85260,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.437921
    },
    {
      "epoch": 3.6530140096825328,
      "grad_norm": 9.058539390563965,
      "learning_rate": 1.6863439621126243e-05,
      "loss": 2.0449899673461913,
      "memory(GiB)": 77.56,
      "step": 85265,
      "token_acc": 0.5117845117845118,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 3.653228225011782,
      "grad_norm": 6.4220733642578125,
      "learning_rate": 1.6858400281210423e-05,
      "loss": 2.203816032409668,
      "memory(GiB)": 77.56,
      "step": 85270,
      "token_acc": 0.5280898876404494,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 3.653442440341031,
      "grad_norm": 6.192942142486572,
      "learning_rate": 1.685336154168576e-05,
      "loss": 2.2622230529785154,
      "memory(GiB)": 77.56,
      "step": 85275,
      "token_acc": 0.5174603174603175,
      "train_speed(iter/s)": 1.437935
    },
    {
      "epoch": 3.6536566556702796,
      "grad_norm": 5.6832661628723145,
      "learning_rate": 1.6848323402643558e-05,
      "loss": 2.5135812759399414,
      "memory(GiB)": 77.56,
      "step": 85280,
      "token_acc": 0.4937888198757764,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 3.653870870999529,
      "grad_norm": 6.137045860290527,
      "learning_rate": 1.684328586417508e-05,
      "loss": 2.4684810638427734,
      "memory(GiB)": 77.56,
      "step": 85285,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 3.6540850863287777,
      "grad_norm": 5.741428852081299,
      "learning_rate": 1.6838248926371575e-05,
      "loss": 2.2643291473388674,
      "memory(GiB)": 77.56,
      "step": 85290,
      "token_acc": 0.5149501661129569,
      "train_speed(iter/s)": 1.437959
    },
    {
      "epoch": 3.6542993016580265,
      "grad_norm": 4.961923122406006,
      "learning_rate": 1.6833212589324304e-05,
      "loss": 2.289890098571777,
      "memory(GiB)": 77.56,
      "step": 85295,
      "token_acc": 0.5037878787878788,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 3.654513516987276,
      "grad_norm": 5.9168171882629395,
      "learning_rate": 1.68281768531245e-05,
      "loss": 2.321359634399414,
      "memory(GiB)": 77.56,
      "step": 85300,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 3.6547277323165246,
      "grad_norm": 8.691854476928711,
      "learning_rate": 1.682314171786337e-05,
      "loss": 2.42355899810791,
      "memory(GiB)": 77.56,
      "step": 85305,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 3.6549419476457734,
      "grad_norm": 5.665338039398193,
      "learning_rate": 1.6818107183632176e-05,
      "loss": 2.434714698791504,
      "memory(GiB)": 77.56,
      "step": 85310,
      "token_acc": 0.48,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 3.6551561629750227,
      "grad_norm": 7.878963947296143,
      "learning_rate": 1.681307325052209e-05,
      "loss": 2.2658031463623045,
      "memory(GiB)": 77.56,
      "step": 85315,
      "token_acc": 0.510548523206751,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 3.6553703783042715,
      "grad_norm": 5.9954071044921875,
      "learning_rate": 1.6808039918624313e-05,
      "loss": 2.3356298446655273,
      "memory(GiB)": 77.56,
      "step": 85320,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.437962
    },
    {
      "epoch": 3.6555845936335203,
      "grad_norm": 6.986223220825195,
      "learning_rate": 1.6803007188030036e-05,
      "loss": 2.3986270904541014,
      "memory(GiB)": 77.56,
      "step": 85325,
      "token_acc": 0.4968152866242038,
      "train_speed(iter/s)": 1.437966
    },
    {
      "epoch": 3.6557988089627695,
      "grad_norm": 4.978462219238281,
      "learning_rate": 1.67979750588304e-05,
      "loss": 2.442804145812988,
      "memory(GiB)": 77.56,
      "step": 85330,
      "token_acc": 0.511326860841424,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.6560130242920184,
      "grad_norm": 5.887653350830078,
      "learning_rate": 1.6792943531116594e-05,
      "loss": 2.3946210861206056,
      "memory(GiB)": 77.56,
      "step": 85335,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 3.656227239621267,
      "grad_norm": 4.989517688751221,
      "learning_rate": 1.678791260497979e-05,
      "loss": 2.163958740234375,
      "memory(GiB)": 77.56,
      "step": 85340,
      "token_acc": 0.5298245614035088,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 3.6564414549505164,
      "grad_norm": 5.966470241546631,
      "learning_rate": 1.6782882280511097e-05,
      "loss": 2.6729021072387695,
      "memory(GiB)": 77.56,
      "step": 85345,
      "token_acc": 0.4709480122324159,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 3.6566556702797652,
      "grad_norm": 9.256418228149414,
      "learning_rate": 1.6777852557801655e-05,
      "loss": 2.7239952087402344,
      "memory(GiB)": 77.56,
      "step": 85350,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 3.656869885609014,
      "grad_norm": 6.300131797790527,
      "learning_rate": 1.6772823436942582e-05,
      "loss": 2.5542022705078127,
      "memory(GiB)": 77.56,
      "step": 85355,
      "token_acc": 0.49101796407185627,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.6570841009382633,
      "grad_norm": 5.623291969299316,
      "learning_rate": 1.676779491802496e-05,
      "loss": 2.6017124176025392,
      "memory(GiB)": 77.56,
      "step": 85360,
      "token_acc": 0.5032258064516129,
      "train_speed(iter/s)": 1.437973
    },
    {
      "epoch": 3.657298316267512,
      "grad_norm": 5.452791690826416,
      "learning_rate": 1.676276700113992e-05,
      "loss": 2.444034385681152,
      "memory(GiB)": 77.56,
      "step": 85365,
      "token_acc": 0.44879518072289154,
      "train_speed(iter/s)": 1.437982
    },
    {
      "epoch": 3.657512531596761,
      "grad_norm": 4.567123889923096,
      "learning_rate": 1.6757739686378543e-05,
      "loss": 2.0532804489135743,
      "memory(GiB)": 77.56,
      "step": 85370,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 3.65772674692601,
      "grad_norm": 7.203108787536621,
      "learning_rate": 1.6752712973831898e-05,
      "loss": 2.244427299499512,
      "memory(GiB)": 77.56,
      "step": 85375,
      "token_acc": 0.5497835497835498,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 3.657940962255259,
      "grad_norm": 6.722211837768555,
      "learning_rate": 1.674768686359104e-05,
      "loss": 2.307176971435547,
      "memory(GiB)": 77.56,
      "step": 85380,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.658155177584508,
      "grad_norm": 4.666999816894531,
      "learning_rate": 1.6742661355747026e-05,
      "loss": 2.216294860839844,
      "memory(GiB)": 77.56,
      "step": 85385,
      "token_acc": 0.5133531157270029,
      "train_speed(iter/s)": 1.437975
    },
    {
      "epoch": 3.658369392913757,
      "grad_norm": 6.5407609939575195,
      "learning_rate": 1.6737636450390887e-05,
      "loss": 2.6667196273803713,
      "memory(GiB)": 77.56,
      "step": 85390,
      "token_acc": 0.460431654676259,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 3.658583608243006,
      "grad_norm": 5.101613998413086,
      "learning_rate": 1.673261214761368e-05,
      "loss": 2.109746551513672,
      "memory(GiB)": 77.56,
      "step": 85395,
      "token_acc": 0.5445205479452054,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.6587978235722547,
      "grad_norm": 6.342676639556885,
      "learning_rate": 1.6727588447506408e-05,
      "loss": 2.3776596069335936,
      "memory(GiB)": 77.56,
      "step": 85400,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.659012038901504,
      "grad_norm": 5.256192684173584,
      "learning_rate": 1.6722565350160075e-05,
      "loss": 2.369375801086426,
      "memory(GiB)": 77.56,
      "step": 85405,
      "token_acc": 0.519163763066202,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.6592262542307528,
      "grad_norm": 5.332834720611572,
      "learning_rate": 1.6717542855665702e-05,
      "loss": 2.258794403076172,
      "memory(GiB)": 77.56,
      "step": 85410,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.438007
    },
    {
      "epoch": 3.6594404695600016,
      "grad_norm": 5.113658905029297,
      "learning_rate": 1.671252096411426e-05,
      "loss": 2.2794450759887694,
      "memory(GiB)": 77.56,
      "step": 85415,
      "token_acc": 0.5015673981191222,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 3.659654684889251,
      "grad_norm": 5.8600029945373535,
      "learning_rate": 1.670749967559671e-05,
      "loss": 2.317844009399414,
      "memory(GiB)": 77.56,
      "step": 85420,
      "token_acc": 0.5089820359281437,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 3.6598689002184996,
      "grad_norm": 5.123144626617432,
      "learning_rate": 1.670247899020405e-05,
      "loss": 2.144133186340332,
      "memory(GiB)": 77.56,
      "step": 85425,
      "token_acc": 0.5224913494809689,
      "train_speed(iter/s)": 1.438042
    },
    {
      "epoch": 3.6600831155477485,
      "grad_norm": 5.756959915161133,
      "learning_rate": 1.669745890802722e-05,
      "loss": 2.2468536376953123,
      "memory(GiB)": 77.56,
      "step": 85430,
      "token_acc": 0.5348101265822784,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 3.6602973308769977,
      "grad_norm": 5.73333215713501,
      "learning_rate": 1.6692439429157164e-05,
      "loss": 2.167403221130371,
      "memory(GiB)": 77.56,
      "step": 85435,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.43807
    },
    {
      "epoch": 3.6605115462062465,
      "grad_norm": 7.354944705963135,
      "learning_rate": 1.6687420553684808e-05,
      "loss": 2.488819885253906,
      "memory(GiB)": 77.56,
      "step": 85440,
      "token_acc": 0.4389438943894389,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 3.6607257615354953,
      "grad_norm": 8.660374641418457,
      "learning_rate": 1.668240228170108e-05,
      "loss": 2.0837434768676757,
      "memory(GiB)": 77.56,
      "step": 85445,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.438057
    },
    {
      "epoch": 3.6609399768647446,
      "grad_norm": 5.56748628616333,
      "learning_rate": 1.6677384613296875e-05,
      "loss": 2.0525686264038088,
      "memory(GiB)": 77.56,
      "step": 85450,
      "token_acc": 0.5536332179930796,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 3.6611541921939934,
      "grad_norm": 6.961643218994141,
      "learning_rate": 1.6672367548563124e-05,
      "loss": 2.2162675857543945,
      "memory(GiB)": 77.56,
      "step": 85455,
      "token_acc": 0.5703703703703704,
      "train_speed(iter/s)": 1.43806
    },
    {
      "epoch": 3.661368407523242,
      "grad_norm": 5.4471845626831055,
      "learning_rate": 1.6667351087590693e-05,
      "loss": 2.3374263763427736,
      "memory(GiB)": 77.56,
      "step": 85460,
      "token_acc": 0.5460992907801419,
      "train_speed(iter/s)": 1.438057
    },
    {
      "epoch": 3.6615826228524915,
      "grad_norm": 6.084554672241211,
      "learning_rate": 1.6662335230470472e-05,
      "loss": 2.0502607345581056,
      "memory(GiB)": 77.56,
      "step": 85465,
      "token_acc": 0.5099337748344371,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 3.6617968381817403,
      "grad_norm": 5.148804187774658,
      "learning_rate": 1.665731997729332e-05,
      "loss": 2.5644710540771483,
      "memory(GiB)": 77.56,
      "step": 85470,
      "token_acc": 0.487012987012987,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 3.662011053510989,
      "grad_norm": 5.40590763092041,
      "learning_rate": 1.6652305328150074e-05,
      "loss": 2.338912582397461,
      "memory(GiB)": 77.56,
      "step": 85475,
      "token_acc": 0.49107142857142855,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 3.6622252688402384,
      "grad_norm": 5.661278247833252,
      "learning_rate": 1.6647291283131606e-05,
      "loss": 2.632363700866699,
      "memory(GiB)": 77.56,
      "step": 85480,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.438083
    },
    {
      "epoch": 3.662439484169487,
      "grad_norm": 11.32270622253418,
      "learning_rate": 1.664227784232876e-05,
      "loss": 2.5806529998779295,
      "memory(GiB)": 77.56,
      "step": 85485,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.438093
    },
    {
      "epoch": 3.662653699498736,
      "grad_norm": 6.08767032623291,
      "learning_rate": 1.6637265005832335e-05,
      "loss": 2.323574256896973,
      "memory(GiB)": 77.56,
      "step": 85490,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 3.6628679148279852,
      "grad_norm": 5.566212177276611,
      "learning_rate": 1.6632252773733154e-05,
      "loss": 2.2979804992675783,
      "memory(GiB)": 77.56,
      "step": 85495,
      "token_acc": 0.47865853658536583,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 3.663082130157234,
      "grad_norm": 5.454259872436523,
      "learning_rate": 1.6627241146122017e-05,
      "loss": 2.145988655090332,
      "memory(GiB)": 77.56,
      "step": 85500,
      "token_acc": 0.5095785440613027,
      "train_speed(iter/s)": 1.438092
    },
    {
      "epoch": 3.663082130157234,
      "eval_loss": 2.3091440200805664,
      "eval_runtime": 15.9932,
      "eval_samples_per_second": 6.253,
      "eval_steps_per_second": 6.253,
      "eval_token_acc": 0.4723404255319149,
      "step": 85500
    },
    {
      "epoch": 3.663296345486483,
      "grad_norm": 5.737945556640625,
      "learning_rate": 1.66222301230897e-05,
      "loss": 2.3726125717163087,
      "memory(GiB)": 77.56,
      "step": 85505,
      "token_acc": 0.48023715415019763,
      "train_speed(iter/s)": 1.437671
    },
    {
      "epoch": 3.663510560815732,
      "grad_norm": 5.8969316482543945,
      "learning_rate": 1.6617219704727004e-05,
      "loss": 2.004317855834961,
      "memory(GiB)": 77.56,
      "step": 85510,
      "token_acc": 0.5270758122743683,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 3.663724776144981,
      "grad_norm": 8.292634010314941,
      "learning_rate": 1.6612209891124696e-05,
      "loss": 2.125049018859863,
      "memory(GiB)": 77.56,
      "step": 85515,
      "token_acc": 0.5236363636363637,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 3.6639389914742297,
      "grad_norm": 5.804937362670898,
      "learning_rate": 1.6607200682373526e-05,
      "loss": 2.5256513595581054,
      "memory(GiB)": 77.56,
      "step": 85520,
      "token_acc": 0.48562300319488816,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 3.664153206803479,
      "grad_norm": 7.040986061096191,
      "learning_rate": 1.6602192078564237e-05,
      "loss": 2.322521781921387,
      "memory(GiB)": 77.56,
      "step": 85525,
      "token_acc": 0.4899598393574297,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 3.664367422132728,
      "grad_norm": 5.853885650634766,
      "learning_rate": 1.659718407978757e-05,
      "loss": 2.239628219604492,
      "memory(GiB)": 77.56,
      "step": 85530,
      "token_acc": 0.5227272727272727,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 3.6645816374619766,
      "grad_norm": 5.997447490692139,
      "learning_rate": 1.6592176686134232e-05,
      "loss": 2.4890716552734373,
      "memory(GiB)": 77.56,
      "step": 85535,
      "token_acc": 0.4453125,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 3.664795852791226,
      "grad_norm": 6.228614330291748,
      "learning_rate": 1.6587169897694965e-05,
      "loss": 2.4528972625732424,
      "memory(GiB)": 77.56,
      "step": 85540,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 3.6650100681204747,
      "grad_norm": 5.8728413581848145,
      "learning_rate": 1.658216371456046e-05,
      "loss": 2.0404239654541017,
      "memory(GiB)": 77.56,
      "step": 85545,
      "token_acc": 0.5670995670995671,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 3.6652242834497235,
      "grad_norm": 6.953040599822998,
      "learning_rate": 1.6577158136821408e-05,
      "loss": 2.360011672973633,
      "memory(GiB)": 77.56,
      "step": 85550,
      "token_acc": 0.47075208913649025,
      "train_speed(iter/s)": 1.437706
    },
    {
      "epoch": 3.6654384987789728,
      "grad_norm": 6.1802077293396,
      "learning_rate": 1.6572153164568478e-05,
      "loss": 2.2948246002197266,
      "memory(GiB)": 77.56,
      "step": 85555,
      "token_acc": 0.49146757679180886,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 3.6656527141082216,
      "grad_norm": 6.816431999206543,
      "learning_rate": 1.6567148797892362e-05,
      "loss": 2.474551963806152,
      "memory(GiB)": 77.56,
      "step": 85560,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.437683
    },
    {
      "epoch": 3.6658669294374704,
      "grad_norm": 6.462721347808838,
      "learning_rate": 1.6562145036883692e-05,
      "loss": 2.2315439224243163,
      "memory(GiB)": 77.56,
      "step": 85565,
      "token_acc": 0.5615384615384615,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 3.6660811447667196,
      "grad_norm": 6.0261616706848145,
      "learning_rate": 1.6557141881633152e-05,
      "loss": 2.2492610931396486,
      "memory(GiB)": 77.56,
      "step": 85570,
      "token_acc": 0.5070422535211268,
      "train_speed(iter/s)": 1.437711
    },
    {
      "epoch": 3.6662953600959685,
      "grad_norm": 5.113628387451172,
      "learning_rate": 1.6552139332231354e-05,
      "loss": 2.332870292663574,
      "memory(GiB)": 77.56,
      "step": 85575,
      "token_acc": 0.5232974910394266,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 3.6665095754252173,
      "grad_norm": 4.488736152648926,
      "learning_rate": 1.6547137388768937e-05,
      "loss": 2.224553871154785,
      "memory(GiB)": 77.56,
      "step": 85580,
      "token_acc": 0.547550432276657,
      "train_speed(iter/s)": 1.437751
    },
    {
      "epoch": 3.6667237907544665,
      "grad_norm": 7.389726638793945,
      "learning_rate": 1.65421360513365e-05,
      "loss": 2.3076539993286134,
      "memory(GiB)": 77.56,
      "step": 85585,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 3.6669380060837153,
      "grad_norm": 6.888086318969727,
      "learning_rate": 1.6537135320024654e-05,
      "loss": 2.440826416015625,
      "memory(GiB)": 77.56,
      "step": 85590,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.437716
    },
    {
      "epoch": 3.667152221412964,
      "grad_norm": 6.0712385177612305,
      "learning_rate": 1.6532135194923982e-05,
      "loss": 2.1879508972167967,
      "memory(GiB)": 77.56,
      "step": 85595,
      "token_acc": 0.49140893470790376,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 3.6673664367422134,
      "grad_norm": 5.9911603927612305,
      "learning_rate": 1.652713567612509e-05,
      "loss": 2.3903751373291016,
      "memory(GiB)": 77.56,
      "step": 85600,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 3.667580652071462,
      "grad_norm": 7.412249565124512,
      "learning_rate": 1.6522136763718533e-05,
      "loss": 2.385560989379883,
      "memory(GiB)": 77.56,
      "step": 85605,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 3.667794867400711,
      "grad_norm": 6.102590084075928,
      "learning_rate": 1.6517138457794873e-05,
      "loss": 2.241424560546875,
      "memory(GiB)": 77.56,
      "step": 85610,
      "token_acc": 0.5075528700906344,
      "train_speed(iter/s)": 1.437753
    },
    {
      "epoch": 3.6680090827299603,
      "grad_norm": 6.744265079498291,
      "learning_rate": 1.651214075844466e-05,
      "loss": 2.0103885650634767,
      "memory(GiB)": 77.56,
      "step": 85615,
      "token_acc": 0.5524193548387096,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.668223298059209,
      "grad_norm": 6.396752834320068,
      "learning_rate": 1.6507143665758417e-05,
      "loss": 2.261327362060547,
      "memory(GiB)": 77.56,
      "step": 85620,
      "token_acc": 0.49480968858131485,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.668437513388458,
      "grad_norm": 6.04730749130249,
      "learning_rate": 1.65021471798267e-05,
      "loss": 2.2555437088012695,
      "memory(GiB)": 77.56,
      "step": 85625,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.668651728717707,
      "grad_norm": 5.6761698722839355,
      "learning_rate": 1.6497151300739994e-05,
      "loss": 2.5593610763549806,
      "memory(GiB)": 77.56,
      "step": 85630,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437773
    },
    {
      "epoch": 3.668865944046956,
      "grad_norm": 6.154073238372803,
      "learning_rate": 1.6492156028588833e-05,
      "loss": 2.2233081817626954,
      "memory(GiB)": 77.56,
      "step": 85635,
      "token_acc": 0.5236686390532544,
      "train_speed(iter/s)": 1.437767
    },
    {
      "epoch": 3.669080159376205,
      "grad_norm": 6.3652024269104,
      "learning_rate": 1.6487161363463705e-05,
      "loss": 2.3743534088134766,
      "memory(GiB)": 77.56,
      "step": 85640,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.437769
    },
    {
      "epoch": 3.669294374705454,
      "grad_norm": 4.999141693115234,
      "learning_rate": 1.6482167305455086e-05,
      "loss": 2.3318323135375976,
      "memory(GiB)": 77.56,
      "step": 85645,
      "token_acc": 0.4804270462633452,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 3.669508590034703,
      "grad_norm": 5.1600518226623535,
      "learning_rate": 1.6477173854653433e-05,
      "loss": 2.5484424591064454,
      "memory(GiB)": 77.56,
      "step": 85650,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.437755
    },
    {
      "epoch": 3.6697228053639517,
      "grad_norm": 7.605493068695068,
      "learning_rate": 1.6472181011149233e-05,
      "loss": 2.486475944519043,
      "memory(GiB)": 77.56,
      "step": 85655,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.43775
    },
    {
      "epoch": 3.669937020693201,
      "grad_norm": 6.143488883972168,
      "learning_rate": 1.6467188775032925e-05,
      "loss": 2.3266164779663088,
      "memory(GiB)": 77.56,
      "step": 85660,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 3.6701512360224497,
      "grad_norm": 6.265586853027344,
      "learning_rate": 1.646219714639496e-05,
      "loss": 2.1769920349121095,
      "memory(GiB)": 77.56,
      "step": 85665,
      "token_acc": 0.5898617511520737,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.6703654513516986,
      "grad_norm": 6.088485240936279,
      "learning_rate": 1.6457206125325747e-05,
      "loss": 2.313176727294922,
      "memory(GiB)": 77.56,
      "step": 85670,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.437742
    },
    {
      "epoch": 3.670579666680948,
      "grad_norm": 5.356929779052734,
      "learning_rate": 1.6452215711915708e-05,
      "loss": 2.166522216796875,
      "memory(GiB)": 77.56,
      "step": 85675,
      "token_acc": 0.5205479452054794,
      "train_speed(iter/s)": 1.437741
    },
    {
      "epoch": 3.6707938820101966,
      "grad_norm": 7.862729549407959,
      "learning_rate": 1.6447225906255237e-05,
      "loss": 2.045585060119629,
      "memory(GiB)": 77.56,
      "step": 85680,
      "token_acc": 0.5629139072847682,
      "train_speed(iter/s)": 1.43775
    },
    {
      "epoch": 3.6710080973394454,
      "grad_norm": 10.510821342468262,
      "learning_rate": 1.6442236708434767e-05,
      "loss": 2.264939880371094,
      "memory(GiB)": 77.56,
      "step": 85685,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.6712223126686947,
      "grad_norm": 4.985856056213379,
      "learning_rate": 1.643724811854465e-05,
      "loss": 2.147949981689453,
      "memory(GiB)": 77.56,
      "step": 85690,
      "token_acc": 0.5183946488294314,
      "train_speed(iter/s)": 1.43775
    },
    {
      "epoch": 3.6714365279979435,
      "grad_norm": 4.9990363121032715,
      "learning_rate": 1.643226013667527e-05,
      "loss": 2.415330505371094,
      "memory(GiB)": 77.56,
      "step": 85695,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 3.6716507433271923,
      "grad_norm": 5.873000144958496,
      "learning_rate": 1.6427272762916973e-05,
      "loss": 2.521626663208008,
      "memory(GiB)": 77.56,
      "step": 85700,
      "token_acc": 0.5477178423236515,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 3.6718649586564416,
      "grad_norm": 5.7365922927856445,
      "learning_rate": 1.6422285997360136e-05,
      "loss": 2.1732915878295898,
      "memory(GiB)": 77.56,
      "step": 85705,
      "token_acc": 0.5225563909774437,
      "train_speed(iter/s)": 1.437766
    },
    {
      "epoch": 3.6720791739856904,
      "grad_norm": 6.276196479797363,
      "learning_rate": 1.6417299840095073e-05,
      "loss": 2.2730649948120116,
      "memory(GiB)": 77.56,
      "step": 85710,
      "token_acc": 0.5182724252491694,
      "train_speed(iter/s)": 1.437765
    },
    {
      "epoch": 3.672293389314939,
      "grad_norm": 5.621336936950684,
      "learning_rate": 1.6412314291212143e-05,
      "loss": 2.4270160675048826,
      "memory(GiB)": 77.56,
      "step": 85715,
      "token_acc": 0.46601941747572817,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.6725076046441885,
      "grad_norm": 6.542372226715088,
      "learning_rate": 1.6407329350801647e-05,
      "loss": 2.3589384078979494,
      "memory(GiB)": 77.56,
      "step": 85720,
      "token_acc": 0.4901315789473684,
      "train_speed(iter/s)": 1.437763
    },
    {
      "epoch": 3.6727218199734373,
      "grad_norm": 4.933358669281006,
      "learning_rate": 1.640234501895389e-05,
      "loss": 2.369892120361328,
      "memory(GiB)": 77.56,
      "step": 85725,
      "token_acc": 0.5014836795252225,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 3.672936035302686,
      "grad_norm": 6.267177104949951,
      "learning_rate": 1.6397361295759172e-05,
      "loss": 2.6145603179931642,
      "memory(GiB)": 77.56,
      "step": 85730,
      "token_acc": 0.4326241134751773,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 3.6731502506319353,
      "grad_norm": 7.575082778930664,
      "learning_rate": 1.6392378181307777e-05,
      "loss": 2.1985549926757812,
      "memory(GiB)": 77.56,
      "step": 85735,
      "token_acc": 0.5052264808362369,
      "train_speed(iter/s)": 1.437789
    },
    {
      "epoch": 3.673364465961184,
      "grad_norm": 5.519351959228516,
      "learning_rate": 1.6387395675689964e-05,
      "loss": 2.317628288269043,
      "memory(GiB)": 77.56,
      "step": 85740,
      "token_acc": 0.5098684210526315,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 3.673578681290433,
      "grad_norm": 6.22172737121582,
      "learning_rate": 1.6382413778996018e-05,
      "loss": 2.307122230529785,
      "memory(GiB)": 77.56,
      "step": 85745,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.437807
    },
    {
      "epoch": 3.6737928966196822,
      "grad_norm": 6.005603790283203,
      "learning_rate": 1.637743249131619e-05,
      "loss": 2.0492132186889647,
      "memory(GiB)": 77.56,
      "step": 85750,
      "token_acc": 0.5583333333333333,
      "train_speed(iter/s)": 1.437797
    },
    {
      "epoch": 3.674007111948931,
      "grad_norm": 6.747081756591797,
      "learning_rate": 1.637245181274071e-05,
      "loss": 2.4059064865112303,
      "memory(GiB)": 77.56,
      "step": 85755,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 3.67422132727818,
      "grad_norm": 5.078756332397461,
      "learning_rate": 1.6367471743359814e-05,
      "loss": 2.4050907135009765,
      "memory(GiB)": 77.56,
      "step": 85760,
      "token_acc": 0.49818181818181817,
      "train_speed(iter/s)": 1.437824
    },
    {
      "epoch": 3.674435542607429,
      "grad_norm": 4.663774013519287,
      "learning_rate": 1.636249228326369e-05,
      "loss": 2.221445083618164,
      "memory(GiB)": 77.56,
      "step": 85765,
      "token_acc": 0.5337620578778135,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.674649757936678,
      "grad_norm": 7.007661819458008,
      "learning_rate": 1.63575134325426e-05,
      "loss": 2.309585952758789,
      "memory(GiB)": 77.56,
      "step": 85770,
      "token_acc": 0.4943820224719101,
      "train_speed(iter/s)": 1.437836
    },
    {
      "epoch": 3.6748639732659267,
      "grad_norm": 6.709556579589844,
      "learning_rate": 1.6352535191286693e-05,
      "loss": 2.2447441101074217,
      "memory(GiB)": 77.56,
      "step": 85775,
      "token_acc": 0.5080385852090032,
      "train_speed(iter/s)": 1.437826
    },
    {
      "epoch": 3.675078188595176,
      "grad_norm": 6.195633411407471,
      "learning_rate": 1.6347557559586192e-05,
      "loss": 2.3480348587036133,
      "memory(GiB)": 77.56,
      "step": 85780,
      "token_acc": 0.48909657320872274,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 3.675292403924425,
      "grad_norm": 5.72022008895874,
      "learning_rate": 1.6342580537531256e-05,
      "loss": 2.477133560180664,
      "memory(GiB)": 77.56,
      "step": 85785,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.6755066192536736,
      "grad_norm": 7.236176013946533,
      "learning_rate": 1.633760412521204e-05,
      "loss": 2.2890499114990233,
      "memory(GiB)": 77.56,
      "step": 85790,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.675720834582923,
      "grad_norm": 6.310311317443848,
      "learning_rate": 1.6332628322718696e-05,
      "loss": 2.393000030517578,
      "memory(GiB)": 77.56,
      "step": 85795,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437858
    },
    {
      "epoch": 3.6759350499121717,
      "grad_norm": 6.815532684326172,
      "learning_rate": 1.6327653130141385e-05,
      "loss": 2.3370025634765623,
      "memory(GiB)": 77.56,
      "step": 85800,
      "token_acc": 0.5369127516778524,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 3.6761492652414205,
      "grad_norm": 7.008269309997559,
      "learning_rate": 1.6322678547570224e-05,
      "loss": 2.1994056701660156,
      "memory(GiB)": 77.56,
      "step": 85805,
      "token_acc": 0.5241935483870968,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.6763634805706698,
      "grad_norm": 8.137584686279297,
      "learning_rate": 1.6317704575095333e-05,
      "loss": 2.4131784439086914,
      "memory(GiB)": 77.56,
      "step": 85810,
      "token_acc": 0.46215139442231074,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 3.6765776958999186,
      "grad_norm": 4.737418174743652,
      "learning_rate": 1.631273121280682e-05,
      "loss": 2.221118927001953,
      "memory(GiB)": 77.56,
      "step": 85815,
      "token_acc": 0.5223367697594502,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.6767919112291674,
      "grad_norm": 6.895242691040039,
      "learning_rate": 1.630775846079478e-05,
      "loss": 2.405704879760742,
      "memory(GiB)": 77.56,
      "step": 85820,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.6770061265584166,
      "grad_norm": 5.875631332397461,
      "learning_rate": 1.630278631914929e-05,
      "loss": 2.21718807220459,
      "memory(GiB)": 77.56,
      "step": 85825,
      "token_acc": 0.5268456375838926,
      "train_speed(iter/s)": 1.437858
    },
    {
      "epoch": 3.6772203418876654,
      "grad_norm": 5.645164489746094,
      "learning_rate": 1.6297814787960453e-05,
      "loss": 2.2740612030029297,
      "memory(GiB)": 77.56,
      "step": 85830,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.437858
    },
    {
      "epoch": 3.6774345572169143,
      "grad_norm": 4.974506378173828,
      "learning_rate": 1.6292843867318307e-05,
      "loss": 2.0515157699584963,
      "memory(GiB)": 77.56,
      "step": 85835,
      "token_acc": 0.5476190476190477,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.6776487725461635,
      "grad_norm": 7.125292778015137,
      "learning_rate": 1.6287873557312927e-05,
      "loss": 2.3242237091064455,
      "memory(GiB)": 77.56,
      "step": 85840,
      "token_acc": 0.5258064516129032,
      "train_speed(iter/s)": 1.437854
    },
    {
      "epoch": 3.6778629878754123,
      "grad_norm": 5.172096252441406,
      "learning_rate": 1.628290385803433e-05,
      "loss": 1.9990224838256836,
      "memory(GiB)": 77.56,
      "step": 85845,
      "token_acc": 0.5179856115107914,
      "train_speed(iter/s)": 1.437888
    },
    {
      "epoch": 3.678077203204661,
      "grad_norm": 6.257838249206543,
      "learning_rate": 1.6277934769572552e-05,
      "loss": 2.3137283325195312,
      "memory(GiB)": 77.56,
      "step": 85850,
      "token_acc": 0.516,
      "train_speed(iter/s)": 1.437893
    },
    {
      "epoch": 3.6782914185339104,
      "grad_norm": 6.082180976867676,
      "learning_rate": 1.6272966292017616e-05,
      "loss": 2.568686866760254,
      "memory(GiB)": 77.56,
      "step": 85855,
      "token_acc": 0.45938375350140054,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.678505633863159,
      "grad_norm": 6.004775524139404,
      "learning_rate": 1.6267998425459552e-05,
      "loss": 2.5092388153076173,
      "memory(GiB)": 77.56,
      "step": 85860,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 3.678719849192408,
      "grad_norm": 7.182103157043457,
      "learning_rate": 1.6263031169988337e-05,
      "loss": 1.9721778869628905,
      "memory(GiB)": 77.56,
      "step": 85865,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.437892
    },
    {
      "epoch": 3.6789340645216573,
      "grad_norm": 5.52949857711792,
      "learning_rate": 1.625806452569396e-05,
      "loss": 2.7242692947387694,
      "memory(GiB)": 77.56,
      "step": 85870,
      "token_acc": 0.4539877300613497,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.679148279850906,
      "grad_norm": 6.061544895172119,
      "learning_rate": 1.6253098492666397e-05,
      "loss": 2.4043949127197264,
      "memory(GiB)": 77.56,
      "step": 85875,
      "token_acc": 0.49642857142857144,
      "train_speed(iter/s)": 1.437888
    },
    {
      "epoch": 3.679362495180155,
      "grad_norm": 5.408303260803223,
      "learning_rate": 1.6248133070995613e-05,
      "loss": 2.3352306365966795,
      "memory(GiB)": 77.56,
      "step": 85880,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 3.679576710509404,
      "grad_norm": 5.1435866355896,
      "learning_rate": 1.6243168260771547e-05,
      "loss": 2.291178512573242,
      "memory(GiB)": 77.56,
      "step": 85885,
      "token_acc": 0.554006968641115,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 3.679790925838653,
      "grad_norm": 6.436489582061768,
      "learning_rate": 1.623820406208417e-05,
      "loss": 2.3680397033691407,
      "memory(GiB)": 77.56,
      "step": 85890,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.437884
    },
    {
      "epoch": 3.680005141167902,
      "grad_norm": 5.247585296630859,
      "learning_rate": 1.6233240475023394e-05,
      "loss": 2.155375862121582,
      "memory(GiB)": 77.56,
      "step": 85895,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 3.680219356497151,
      "grad_norm": 7.1036376953125,
      "learning_rate": 1.622827749967914e-05,
      "loss": 2.5256114959716798,
      "memory(GiB)": 77.56,
      "step": 85900,
      "token_acc": 0.4979253112033195,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 3.6804335718264,
      "grad_norm": 6.2489013671875,
      "learning_rate": 1.6223315136141327e-05,
      "loss": 2.473739242553711,
      "memory(GiB)": 77.56,
      "step": 85905,
      "token_acc": 0.458041958041958,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 3.6806477871556487,
      "grad_norm": 5.340621471405029,
      "learning_rate": 1.6218353384499824e-05,
      "loss": 2.2262123107910154,
      "memory(GiB)": 77.56,
      "step": 85910,
      "token_acc": 0.5264705882352941,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.680862002484898,
      "grad_norm": 5.009217262268066,
      "learning_rate": 1.6213392244844554e-05,
      "loss": 2.2785408020019533,
      "memory(GiB)": 77.56,
      "step": 85915,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.6810762178141467,
      "grad_norm": 5.049722671508789,
      "learning_rate": 1.6208431717265382e-05,
      "loss": 2.5480831146240233,
      "memory(GiB)": 77.56,
      "step": 85920,
      "token_acc": 0.46686746987951805,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.6812904331433955,
      "grad_norm": 4.7372870445251465,
      "learning_rate": 1.6203471801852148e-05,
      "loss": 2.421305465698242,
      "memory(GiB)": 77.56,
      "step": 85925,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.437892
    },
    {
      "epoch": 3.681504648472645,
      "grad_norm": 8.707403182983398,
      "learning_rate": 1.6198512498694744e-05,
      "loss": 1.9206993103027343,
      "memory(GiB)": 77.56,
      "step": 85930,
      "token_acc": 0.56640625,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 3.6817188638018936,
      "grad_norm": 4.758282661437988,
      "learning_rate": 1.6193553807882998e-05,
      "loss": 2.6549762725830077,
      "memory(GiB)": 77.56,
      "step": 85935,
      "token_acc": 0.4513715710723192,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 3.6819330791311424,
      "grad_norm": 5.072307109832764,
      "learning_rate": 1.6188595729506722e-05,
      "loss": 2.264400291442871,
      "memory(GiB)": 77.56,
      "step": 85940,
      "token_acc": 0.5040214477211796,
      "train_speed(iter/s)": 1.437904
    },
    {
      "epoch": 3.6821472944603917,
      "grad_norm": 6.071218490600586,
      "learning_rate": 1.6183638263655765e-05,
      "loss": 2.5853748321533203,
      "memory(GiB)": 77.56,
      "step": 85945,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 3.6823615097896405,
      "grad_norm": 6.709692001342773,
      "learning_rate": 1.617868141041992e-05,
      "loss": 2.2545841217041014,
      "memory(GiB)": 77.56,
      "step": 85950,
      "token_acc": 0.5157894736842106,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 3.6825757251188893,
      "grad_norm": 4.431982040405273,
      "learning_rate": 1.6173725169888994e-05,
      "loss": 2.0270526885986326,
      "memory(GiB)": 77.56,
      "step": 85955,
      "token_acc": 0.5346534653465347,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.6827899404481386,
      "grad_norm": 5.886152744293213,
      "learning_rate": 1.6168769542152767e-05,
      "loss": 2.2336679458618165,
      "memory(GiB)": 77.56,
      "step": 85960,
      "token_acc": 0.5344827586206896,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 3.6830041557773874,
      "grad_norm": 5.202999114990234,
      "learning_rate": 1.6163814527301013e-05,
      "loss": 2.024606704711914,
      "memory(GiB)": 77.56,
      "step": 85965,
      "token_acc": 0.5525291828793775,
      "train_speed(iter/s)": 1.437858
    },
    {
      "epoch": 3.683218371106636,
      "grad_norm": 5.611460208892822,
      "learning_rate": 1.6158860125423485e-05,
      "loss": 2.552311134338379,
      "memory(GiB)": 77.56,
      "step": 85970,
      "token_acc": 0.45857988165680474,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 3.6834325864358854,
      "grad_norm": 5.390848159790039,
      "learning_rate": 1.615390633660997e-05,
      "loss": 2.3400020599365234,
      "memory(GiB)": 77.56,
      "step": 85975,
      "token_acc": 0.48314606741573035,
      "train_speed(iter/s)": 1.437876
    },
    {
      "epoch": 3.6836468017651343,
      "grad_norm": 7.039914131164551,
      "learning_rate": 1.6148953160950187e-05,
      "loss": 2.2798843383789062,
      "memory(GiB)": 77.56,
      "step": 85980,
      "token_acc": 0.47878787878787876,
      "train_speed(iter/s)": 1.437852
    },
    {
      "epoch": 3.683861017094383,
      "grad_norm": 6.332791328430176,
      "learning_rate": 1.6144000598533877e-05,
      "loss": 2.4170612335205077,
      "memory(GiB)": 77.56,
      "step": 85985,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.6840752324236323,
      "grad_norm": 5.428787708282471,
      "learning_rate": 1.6139048649450756e-05,
      "loss": 2.3463172912597656,
      "memory(GiB)": 77.56,
      "step": 85990,
      "token_acc": 0.4701492537313433,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.684289447752881,
      "grad_norm": 5.748325824737549,
      "learning_rate": 1.6134097313790515e-05,
      "loss": 2.4689104080200197,
      "memory(GiB)": 77.56,
      "step": 85995,
      "token_acc": 0.4847560975609756,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.68450366308213,
      "grad_norm": 6.762945175170898,
      "learning_rate": 1.6129146591642868e-05,
      "loss": 2.4168350219726564,
      "memory(GiB)": 77.56,
      "step": 86000,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 3.68450366308213,
      "eval_loss": 2.258465051651001,
      "eval_runtime": 14.8893,
      "eval_samples_per_second": 6.716,
      "eval_steps_per_second": 6.716,
      "eval_token_acc": 0.4717741935483871,
      "step": 86000
    },
    {
      "epoch": 3.684717878411379,
      "grad_norm": 5.709500312805176,
      "learning_rate": 1.612419648309752e-05,
      "loss": 2.455639457702637,
      "memory(GiB)": 77.56,
      "step": 86005,
      "token_acc": 0.4690522243713733,
      "train_speed(iter/s)": 1.437482
    },
    {
      "epoch": 3.684932093740628,
      "grad_norm": 5.775025844573975,
      "learning_rate": 1.6119246988244136e-05,
      "loss": 2.1570785522460936,
      "memory(GiB)": 77.56,
      "step": 86010,
      "token_acc": 0.5567375886524822,
      "train_speed(iter/s)": 1.437494
    },
    {
      "epoch": 3.685146309069877,
      "grad_norm": 5.722115516662598,
      "learning_rate": 1.6114298107172372e-05,
      "loss": 2.342878723144531,
      "memory(GiB)": 77.56,
      "step": 86015,
      "token_acc": 0.4703389830508475,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 3.685360524399126,
      "grad_norm": 5.323507785797119,
      "learning_rate": 1.6109349839971883e-05,
      "loss": 2.594965362548828,
      "memory(GiB)": 77.56,
      "step": 86020,
      "token_acc": 0.45345345345345345,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 3.685574739728375,
      "grad_norm": 5.1564717292785645,
      "learning_rate": 1.61044021867323e-05,
      "loss": 2.1144077301025392,
      "memory(GiB)": 77.56,
      "step": 86025,
      "token_acc": 0.5341614906832298,
      "train_speed(iter/s)": 1.437509
    },
    {
      "epoch": 3.6857889550576237,
      "grad_norm": 5.714221000671387,
      "learning_rate": 1.6099455147543284e-05,
      "loss": 2.1551139831542967,
      "memory(GiB)": 77.56,
      "step": 86030,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.437513
    },
    {
      "epoch": 3.686003170386873,
      "grad_norm": 10.161355018615723,
      "learning_rate": 1.609450872249444e-05,
      "loss": 2.054778289794922,
      "memory(GiB)": 77.56,
      "step": 86035,
      "token_acc": 0.5247148288973384,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 3.686217385716122,
      "grad_norm": 5.03427791595459,
      "learning_rate": 1.6089562911675377e-05,
      "loss": 2.21038761138916,
      "memory(GiB)": 77.56,
      "step": 86040,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437538
    },
    {
      "epoch": 3.6864316010453706,
      "grad_norm": 3.9941518306732178,
      "learning_rate": 1.6084617715175686e-05,
      "loss": 2.171978759765625,
      "memory(GiB)": 77.56,
      "step": 86045,
      "token_acc": 0.5197368421052632,
      "train_speed(iter/s)": 1.437556
    },
    {
      "epoch": 3.68664581637462,
      "grad_norm": 5.85705041885376,
      "learning_rate": 1.6079673133084967e-05,
      "loss": 2.461191940307617,
      "memory(GiB)": 77.56,
      "step": 86050,
      "token_acc": 0.4713375796178344,
      "train_speed(iter/s)": 1.437547
    },
    {
      "epoch": 3.6868600317038687,
      "grad_norm": 5.6941142082214355,
      "learning_rate": 1.607472916549277e-05,
      "loss": 2.5281028747558594,
      "memory(GiB)": 77.56,
      "step": 86055,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.437546
    },
    {
      "epoch": 3.6870742470331175,
      "grad_norm": 5.876923084259033,
      "learning_rate": 1.6069785812488696e-05,
      "loss": 2.085440444946289,
      "memory(GiB)": 77.56,
      "step": 86060,
      "token_acc": 0.5675675675675675,
      "train_speed(iter/s)": 1.43756
    },
    {
      "epoch": 3.6872884623623667,
      "grad_norm": 5.975595951080322,
      "learning_rate": 1.6064843074162285e-05,
      "loss": 2.307909393310547,
      "memory(GiB)": 77.56,
      "step": 86065,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.437543
    },
    {
      "epoch": 3.6875026776916155,
      "grad_norm": 9.323893547058105,
      "learning_rate": 1.6059900950603052e-05,
      "loss": 2.2341079711914062,
      "memory(GiB)": 77.56,
      "step": 86070,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.437542
    },
    {
      "epoch": 3.6877168930208644,
      "grad_norm": 5.0056562423706055,
      "learning_rate": 1.6054959441900574e-05,
      "loss": 2.4665449142456053,
      "memory(GiB)": 77.56,
      "step": 86075,
      "token_acc": 0.4955223880597015,
      "train_speed(iter/s)": 1.437528
    },
    {
      "epoch": 3.6879311083501136,
      "grad_norm": 6.423130512237549,
      "learning_rate": 1.605001854814434e-05,
      "loss": 2.401065635681152,
      "memory(GiB)": 77.56,
      "step": 86080,
      "token_acc": 0.5060728744939271,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 3.6881453236793624,
      "grad_norm": 5.566985130310059,
      "learning_rate": 1.6045078269423864e-05,
      "loss": 2.0344518661499023,
      "memory(GiB)": 77.56,
      "step": 86085,
      "token_acc": 0.5658914728682171,
      "train_speed(iter/s)": 1.43752
    },
    {
      "epoch": 3.6883595390086112,
      "grad_norm": 7.832764625549316,
      "learning_rate": 1.6040138605828653e-05,
      "loss": 2.373078727722168,
      "memory(GiB)": 77.56,
      "step": 86090,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.4375
    },
    {
      "epoch": 3.6885737543378605,
      "grad_norm": 4.745559215545654,
      "learning_rate": 1.6035199557448194e-05,
      "loss": 2.058857727050781,
      "memory(GiB)": 77.56,
      "step": 86095,
      "token_acc": 0.5244755244755245,
      "train_speed(iter/s)": 1.43751
    },
    {
      "epoch": 3.6887879696671093,
      "grad_norm": 5.882087707519531,
      "learning_rate": 1.6030261124371953e-05,
      "loss": 2.2976215362548826,
      "memory(GiB)": 77.56,
      "step": 86100,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.437522
    },
    {
      "epoch": 3.689002184996358,
      "grad_norm": 5.777613162994385,
      "learning_rate": 1.60253233066894e-05,
      "loss": 2.481217956542969,
      "memory(GiB)": 77.56,
      "step": 86105,
      "token_acc": 0.486404833836858,
      "train_speed(iter/s)": 1.437518
    },
    {
      "epoch": 3.6892164003256074,
      "grad_norm": 7.4842939376831055,
      "learning_rate": 1.6020386104489986e-05,
      "loss": 2.264832878112793,
      "memory(GiB)": 77.56,
      "step": 86110,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.437519
    },
    {
      "epoch": 3.689430615654856,
      "grad_norm": 5.6255083084106445,
      "learning_rate": 1.6015449517863133e-05,
      "loss": 2.1634780883789064,
      "memory(GiB)": 77.56,
      "step": 86115,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.437521
    },
    {
      "epoch": 3.689644830984105,
      "grad_norm": 6.240344524383545,
      "learning_rate": 1.6010513546898316e-05,
      "loss": 2.3222721099853514,
      "memory(GiB)": 77.56,
      "step": 86120,
      "token_acc": 0.546031746031746,
      "train_speed(iter/s)": 1.43753
    },
    {
      "epoch": 3.6898590463133543,
      "grad_norm": 6.719814300537109,
      "learning_rate": 1.600557819168493e-05,
      "loss": 2.174650955200195,
      "memory(GiB)": 77.56,
      "step": 86125,
      "token_acc": 0.5230125523012552,
      "train_speed(iter/s)": 1.437532
    },
    {
      "epoch": 3.690073261642603,
      "grad_norm": 5.500310897827148,
      "learning_rate": 1.600064345231238e-05,
      "loss": 2.5471914291381834,
      "memory(GiB)": 77.56,
      "step": 86130,
      "token_acc": 0.4562334217506631,
      "train_speed(iter/s)": 1.437539
    },
    {
      "epoch": 3.690287476971852,
      "grad_norm": 4.539886951446533,
      "learning_rate": 1.5995709328870067e-05,
      "loss": 2.1795543670654296,
      "memory(GiB)": 77.56,
      "step": 86135,
      "token_acc": 0.5437956204379562,
      "train_speed(iter/s)": 1.437531
    },
    {
      "epoch": 3.690501692301101,
      "grad_norm": 5.665561199188232,
      "learning_rate": 1.5990775821447363e-05,
      "loss": 2.526131248474121,
      "memory(GiB)": 77.56,
      "step": 86140,
      "token_acc": 0.44984802431610943,
      "train_speed(iter/s)": 1.437537
    },
    {
      "epoch": 3.69071590763035,
      "grad_norm": 6.106988906860352,
      "learning_rate": 1.598584293013366e-05,
      "loss": 2.3974456787109375,
      "memory(GiB)": 77.56,
      "step": 86145,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.437559
    },
    {
      "epoch": 3.6909301229595988,
      "grad_norm": 6.5132598876953125,
      "learning_rate": 1.5980910655018332e-05,
      "loss": 2.2827762603759765,
      "memory(GiB)": 77.56,
      "step": 86150,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.43758
    },
    {
      "epoch": 3.691144338288848,
      "grad_norm": 6.577299118041992,
      "learning_rate": 1.5975978996190727e-05,
      "loss": 2.463894081115723,
      "memory(GiB)": 77.56,
      "step": 86155,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.437586
    },
    {
      "epoch": 3.691358553618097,
      "grad_norm": 4.83075475692749,
      "learning_rate": 1.5971047953740174e-05,
      "loss": 2.4222484588623048,
      "memory(GiB)": 77.56,
      "step": 86160,
      "token_acc": 0.47953216374269003,
      "train_speed(iter/s)": 1.437574
    },
    {
      "epoch": 3.6915727689473457,
      "grad_norm": 6.76957893371582,
      "learning_rate": 1.5966117527756013e-05,
      "loss": 2.2134328842163087,
      "memory(GiB)": 77.56,
      "step": 86165,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.437585
    },
    {
      "epoch": 3.691786984276595,
      "grad_norm": 4.991066932678223,
      "learning_rate": 1.596118771832754e-05,
      "loss": 2.489685821533203,
      "memory(GiB)": 77.56,
      "step": 86170,
      "token_acc": 0.5229681978798587,
      "train_speed(iter/s)": 1.437597
    },
    {
      "epoch": 3.6920011996058437,
      "grad_norm": 7.68670129776001,
      "learning_rate": 1.59562585255441e-05,
      "loss": 2.3423063278198244,
      "memory(GiB)": 77.56,
      "step": 86175,
      "token_acc": 0.4626865671641791,
      "train_speed(iter/s)": 1.437614
    },
    {
      "epoch": 3.6922154149350925,
      "grad_norm": 4.818495750427246,
      "learning_rate": 1.5951329949494976e-05,
      "loss": 2.311749076843262,
      "memory(GiB)": 77.56,
      "step": 86180,
      "token_acc": 0.504424778761062,
      "train_speed(iter/s)": 1.437615
    },
    {
      "epoch": 3.692429630264342,
      "grad_norm": 6.807466506958008,
      "learning_rate": 1.5946401990269444e-05,
      "loss": 2.1719968795776365,
      "memory(GiB)": 77.56,
      "step": 86185,
      "token_acc": 0.5176056338028169,
      "train_speed(iter/s)": 1.437624
    },
    {
      "epoch": 3.6926438455935906,
      "grad_norm": 6.530341625213623,
      "learning_rate": 1.5941474647956788e-05,
      "loss": 2.429329681396484,
      "memory(GiB)": 77.56,
      "step": 86190,
      "token_acc": 0.47674418604651164,
      "train_speed(iter/s)": 1.43763
    },
    {
      "epoch": 3.6928580609228394,
      "grad_norm": 6.639760494232178,
      "learning_rate": 1.5936547922646268e-05,
      "loss": 2.506332015991211,
      "memory(GiB)": 77.56,
      "step": 86195,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.437623
    },
    {
      "epoch": 3.6930722762520887,
      "grad_norm": 6.662483215332031,
      "learning_rate": 1.5931621814427118e-05,
      "loss": 2.110951614379883,
      "memory(GiB)": 77.56,
      "step": 86200,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.437631
    },
    {
      "epoch": 3.6932864915813375,
      "grad_norm": 4.36579704284668,
      "learning_rate": 1.5926696323388618e-05,
      "loss": 2.3908714294433593,
      "memory(GiB)": 77.56,
      "step": 86205,
      "token_acc": 0.5097402597402597,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 3.6935007069105863,
      "grad_norm": 6.042758941650391,
      "learning_rate": 1.5921771449619977e-05,
      "loss": 2.6326574325561523,
      "memory(GiB)": 77.56,
      "step": 86210,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.437611
    },
    {
      "epoch": 3.6937149222398356,
      "grad_norm": 6.384288311004639,
      "learning_rate": 1.591684719321041e-05,
      "loss": 2.0806385040283204,
      "memory(GiB)": 77.56,
      "step": 86215,
      "token_acc": 0.5269230769230769,
      "train_speed(iter/s)": 1.437621
    },
    {
      "epoch": 3.6939291375690844,
      "grad_norm": 5.6147379875183105,
      "learning_rate": 1.5911923554249113e-05,
      "loss": 2.5784053802490234,
      "memory(GiB)": 77.56,
      "step": 86220,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 3.694143352898333,
      "grad_norm": 6.724876880645752,
      "learning_rate": 1.590700053282532e-05,
      "loss": 2.571179962158203,
      "memory(GiB)": 77.56,
      "step": 86225,
      "token_acc": 0.4939759036144578,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 3.6943575682275824,
      "grad_norm": 4.180628299713135,
      "learning_rate": 1.590207812902817e-05,
      "loss": 2.3681308746337892,
      "memory(GiB)": 77.56,
      "step": 86230,
      "token_acc": 0.5068027210884354,
      "train_speed(iter/s)": 1.437635
    },
    {
      "epoch": 3.6945717835568312,
      "grad_norm": 5.39839506149292,
      "learning_rate": 1.5897156342946883e-05,
      "loss": 2.4671966552734377,
      "memory(GiB)": 77.56,
      "step": 86235,
      "token_acc": 0.44981412639405205,
      "train_speed(iter/s)": 1.437657
    },
    {
      "epoch": 3.69478599888608,
      "grad_norm": 5.495696067810059,
      "learning_rate": 1.58922351746706e-05,
      "loss": 2.736591911315918,
      "memory(GiB)": 77.56,
      "step": 86240,
      "token_acc": 0.4702194357366771,
      "train_speed(iter/s)": 1.437674
    },
    {
      "epoch": 3.6950002142153293,
      "grad_norm": 7.17864465713501,
      "learning_rate": 1.5887314624288467e-05,
      "loss": 2.408966636657715,
      "memory(GiB)": 77.56,
      "step": 86245,
      "token_acc": 0.450354609929078,
      "train_speed(iter/s)": 1.437687
    },
    {
      "epoch": 3.695214429544578,
      "grad_norm": 6.017739295959473,
      "learning_rate": 1.5882394691889636e-05,
      "loss": 2.2471899032592773,
      "memory(GiB)": 77.56,
      "step": 86250,
      "token_acc": 0.5348101265822784,
      "train_speed(iter/s)": 1.437697
    },
    {
      "epoch": 3.695428644873827,
      "grad_norm": 6.043412685394287,
      "learning_rate": 1.587747537756323e-05,
      "loss": 2.664326858520508,
      "memory(GiB)": 77.56,
      "step": 86255,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.437672
    },
    {
      "epoch": 3.695642860203076,
      "grad_norm": 5.255742073059082,
      "learning_rate": 1.5872556681398342e-05,
      "loss": 2.5010602951049803,
      "memory(GiB)": 77.56,
      "step": 86260,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.437673
    },
    {
      "epoch": 3.695857075532325,
      "grad_norm": 4.185830593109131,
      "learning_rate": 1.5867638603484132e-05,
      "loss": 2.6247385025024412,
      "memory(GiB)": 77.56,
      "step": 86265,
      "token_acc": 0.48255813953488375,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 3.696071290861574,
      "grad_norm": 6.7779998779296875,
      "learning_rate": 1.5862721143909658e-05,
      "loss": 2.274894905090332,
      "memory(GiB)": 77.56,
      "step": 86270,
      "token_acc": 0.5076335877862596,
      "train_speed(iter/s)": 1.43771
    },
    {
      "epoch": 3.696285506190823,
      "grad_norm": 6.136999607086182,
      "learning_rate": 1.585780430276402e-05,
      "loss": 2.0463333129882812,
      "memory(GiB)": 77.56,
      "step": 86275,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.437705
    },
    {
      "epoch": 3.696499721520072,
      "grad_norm": 4.349707126617432,
      "learning_rate": 1.585288808013628e-05,
      "loss": 2.291848564147949,
      "memory(GiB)": 77.56,
      "step": 86280,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.437726
    },
    {
      "epoch": 3.6967139368493207,
      "grad_norm": 6.033775329589844,
      "learning_rate": 1.584797247611549e-05,
      "loss": 2.2331987380981446,
      "memory(GiB)": 77.56,
      "step": 86285,
      "token_acc": 0.5394736842105263,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.69692815217857,
      "grad_norm": 7.3697662353515625,
      "learning_rate": 1.5843057490790737e-05,
      "loss": 2.4206491470336915,
      "memory(GiB)": 77.56,
      "step": 86290,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 3.6971423675078188,
      "grad_norm": 6.867831707000732,
      "learning_rate": 1.583814312425102e-05,
      "loss": 2.323303985595703,
      "memory(GiB)": 77.56,
      "step": 86295,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.437759
    },
    {
      "epoch": 3.6973565828370676,
      "grad_norm": 6.913043975830078,
      "learning_rate": 1.5833229376585402e-05,
      "loss": 2.6895069122314452,
      "memory(GiB)": 77.56,
      "step": 86300,
      "token_acc": 0.47307692307692306,
      "train_speed(iter/s)": 1.437766
    },
    {
      "epoch": 3.697570798166317,
      "grad_norm": 4.569516181945801,
      "learning_rate": 1.5828316247882884e-05,
      "loss": 2.1337255477905273,
      "memory(GiB)": 77.56,
      "step": 86305,
      "token_acc": 0.5206611570247934,
      "train_speed(iter/s)": 1.437753
    },
    {
      "epoch": 3.6977850134955657,
      "grad_norm": 7.96969747543335,
      "learning_rate": 1.582340373823248e-05,
      "loss": 2.3565837860107424,
      "memory(GiB)": 77.56,
      "step": 86310,
      "token_acc": 0.47249190938511326,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 3.6979992288248145,
      "grad_norm": 4.90771484375,
      "learning_rate": 1.581849184772315e-05,
      "loss": 2.3359630584716795,
      "memory(GiB)": 77.56,
      "step": 86315,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 3.6982134441540637,
      "grad_norm": 5.176599979400635,
      "learning_rate": 1.581358057644393e-05,
      "loss": 2.3966684341430664,
      "memory(GiB)": 77.56,
      "step": 86320,
      "token_acc": 0.504950495049505,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.6984276594833125,
      "grad_norm": 4.707461357116699,
      "learning_rate": 1.5808669924483765e-05,
      "loss": 2.5782777786254885,
      "memory(GiB)": 77.56,
      "step": 86325,
      "token_acc": 0.5122699386503068,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 3.6986418748125613,
      "grad_norm": 6.585312366485596,
      "learning_rate": 1.5803759891931613e-05,
      "loss": 2.290018081665039,
      "memory(GiB)": 77.56,
      "step": 86330,
      "token_acc": 0.5380577427821522,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 3.6988560901418106,
      "grad_norm": 5.981418609619141,
      "learning_rate": 1.579885047887644e-05,
      "loss": 2.3021392822265625,
      "memory(GiB)": 77.56,
      "step": 86335,
      "token_acc": 0.5068027210884354,
      "train_speed(iter/s)": 1.437797
    },
    {
      "epoch": 3.6990703054710594,
      "grad_norm": 5.6736674308776855,
      "learning_rate": 1.5793941685407165e-05,
      "loss": 2.3263540267944336,
      "memory(GiB)": 77.56,
      "step": 86340,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 3.6992845208003082,
      "grad_norm": 4.543986797332764,
      "learning_rate": 1.5789033511612712e-05,
      "loss": 2.0895456314086913,
      "memory(GiB)": 77.56,
      "step": 86345,
      "token_acc": 0.5611510791366906,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 3.6994987361295575,
      "grad_norm": 5.240317344665527,
      "learning_rate": 1.5784125957582025e-05,
      "loss": 2.2390132904052735,
      "memory(GiB)": 77.56,
      "step": 86350,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.6997129514588063,
      "grad_norm": 6.601179599761963,
      "learning_rate": 1.577921902340399e-05,
      "loss": 2.477798271179199,
      "memory(GiB)": 77.56,
      "step": 86355,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 3.699927166788055,
      "grad_norm": 4.889268398284912,
      "learning_rate": 1.5774312709167505e-05,
      "loss": 2.4026519775390627,
      "memory(GiB)": 77.56,
      "step": 86360,
      "token_acc": 0.46839080459770116,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 3.7001413821173044,
      "grad_norm": 4.572332382202148,
      "learning_rate": 1.576940701496144e-05,
      "loss": 2.560747528076172,
      "memory(GiB)": 77.56,
      "step": 86365,
      "token_acc": 0.44642857142857145,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 3.700355597446553,
      "grad_norm": 6.059507369995117,
      "learning_rate": 1.5764501940874688e-05,
      "loss": 2.1256687164306642,
      "memory(GiB)": 77.56,
      "step": 86370,
      "token_acc": 0.5311475409836065,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 3.700569812775802,
      "grad_norm": 4.444861888885498,
      "learning_rate": 1.5759597486996086e-05,
      "loss": 2.604166030883789,
      "memory(GiB)": 77.56,
      "step": 86375,
      "token_acc": 0.4954128440366973,
      "train_speed(iter/s)": 1.437802
    },
    {
      "epoch": 3.7007840281050512,
      "grad_norm": 7.19088888168335,
      "learning_rate": 1.5754693653414515e-05,
      "loss": 2.1890735626220703,
      "memory(GiB)": 77.56,
      "step": 86380,
      "token_acc": 0.5582329317269076,
      "train_speed(iter/s)": 1.437823
    },
    {
      "epoch": 3.7009982434343,
      "grad_norm": 7.390120029449463,
      "learning_rate": 1.5749790440218787e-05,
      "loss": 2.507330322265625,
      "memory(GiB)": 77.56,
      "step": 86385,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 3.701212458763549,
      "grad_norm": 5.598202228546143,
      "learning_rate": 1.5744887847497735e-05,
      "loss": 2.129720687866211,
      "memory(GiB)": 77.56,
      "step": 86390,
      "token_acc": 0.536,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 3.701426674092798,
      "grad_norm": 5.509608268737793,
      "learning_rate": 1.573998587534018e-05,
      "loss": 2.264794921875,
      "memory(GiB)": 77.56,
      "step": 86395,
      "token_acc": 0.551948051948052,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 3.701640889422047,
      "grad_norm": 8.979785919189453,
      "learning_rate": 1.5735084523834913e-05,
      "loss": 2.2470375061035157,
      "memory(GiB)": 77.56,
      "step": 86400,
      "token_acc": 0.4612794612794613,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 3.7018551047512958,
      "grad_norm": 5.396533012390137,
      "learning_rate": 1.573018379307072e-05,
      "loss": 2.436020851135254,
      "memory(GiB)": 77.56,
      "step": 86405,
      "token_acc": 0.5244755244755245,
      "train_speed(iter/s)": 1.437893
    },
    {
      "epoch": 3.702069320080545,
      "grad_norm": 5.912084579467773,
      "learning_rate": 1.572528368313641e-05,
      "loss": 1.9977581024169921,
      "memory(GiB)": 77.56,
      "step": 86410,
      "token_acc": 0.5304659498207885,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 3.702283535409794,
      "grad_norm": 6.608034610748291,
      "learning_rate": 1.572038419412074e-05,
      "loss": 2.1554019927978514,
      "memory(GiB)": 77.56,
      "step": 86415,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.4379
    },
    {
      "epoch": 3.7024977507390426,
      "grad_norm": 6.69238805770874,
      "learning_rate": 1.5715485326112467e-05,
      "loss": 2.2238237380981447,
      "memory(GiB)": 77.56,
      "step": 86420,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 3.702711966068292,
      "grad_norm": 7.633655071258545,
      "learning_rate": 1.5710587079200335e-05,
      "loss": 2.3384639739990236,
      "memory(GiB)": 77.56,
      "step": 86425,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437914
    },
    {
      "epoch": 3.7029261813975407,
      "grad_norm": 5.715300559997559,
      "learning_rate": 1.5705689453473072e-05,
      "loss": 2.399251937866211,
      "memory(GiB)": 77.56,
      "step": 86430,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.437926
    },
    {
      "epoch": 3.7031403967267895,
      "grad_norm": 5.963386535644531,
      "learning_rate": 1.570079244901943e-05,
      "loss": 2.3137426376342773,
      "memory(GiB)": 77.56,
      "step": 86435,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 3.7033546120560388,
      "grad_norm": 11.06846809387207,
      "learning_rate": 1.569589606592809e-05,
      "loss": 2.2921144485473635,
      "memory(GiB)": 77.56,
      "step": 86440,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.437936
    },
    {
      "epoch": 3.7035688273852876,
      "grad_norm": 8.007438659667969,
      "learning_rate": 1.569100030428779e-05,
      "loss": 2.45764274597168,
      "memory(GiB)": 77.56,
      "step": 86445,
      "token_acc": 0.4900398406374502,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 3.7037830427145364,
      "grad_norm": 6.153940200805664,
      "learning_rate": 1.568610516418721e-05,
      "loss": 2.4899049758911134,
      "memory(GiB)": 77.56,
      "step": 86450,
      "token_acc": 0.509375,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 3.7039972580437857,
      "grad_norm": 6.3502397537231445,
      "learning_rate": 1.5681210645715017e-05,
      "loss": 2.237340545654297,
      "memory(GiB)": 77.56,
      "step": 86455,
      "token_acc": 0.5144694533762058,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 3.7042114733730345,
      "grad_norm": 7.710557460784912,
      "learning_rate": 1.567631674895987e-05,
      "loss": 2.398152160644531,
      "memory(GiB)": 77.56,
      "step": 86460,
      "token_acc": 0.5063829787234042,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 3.7044256887022833,
      "grad_norm": 4.601808547973633,
      "learning_rate": 1.567142347401046e-05,
      "loss": 2.2292415618896486,
      "memory(GiB)": 77.56,
      "step": 86465,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437932
    },
    {
      "epoch": 3.7046399040315325,
      "grad_norm": 5.665591716766357,
      "learning_rate": 1.566653082095542e-05,
      "loss": 2.300118637084961,
      "memory(GiB)": 77.56,
      "step": 86470,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 3.7048541193607814,
      "grad_norm": 5.024373531341553,
      "learning_rate": 1.566163878988338e-05,
      "loss": 2.373953628540039,
      "memory(GiB)": 77.56,
      "step": 86475,
      "token_acc": 0.5047923322683706,
      "train_speed(iter/s)": 1.437906
    },
    {
      "epoch": 3.70506833469003,
      "grad_norm": 6.110803604125977,
      "learning_rate": 1.5656747380882965e-05,
      "loss": 2.3358633041381838,
      "memory(GiB)": 77.56,
      "step": 86480,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.43793
    },
    {
      "epoch": 3.7052825500192794,
      "grad_norm": 7.4865403175354,
      "learning_rate": 1.565185659404279e-05,
      "loss": 2.391068458557129,
      "memory(GiB)": 77.56,
      "step": 86485,
      "token_acc": 0.4601449275362319,
      "train_speed(iter/s)": 1.437942
    },
    {
      "epoch": 3.7054967653485282,
      "grad_norm": 5.674342632293701,
      "learning_rate": 1.564696642945143e-05,
      "loss": 2.8004451751708985,
      "memory(GiB)": 77.56,
      "step": 86490,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.437942
    },
    {
      "epoch": 3.705710980677777,
      "grad_norm": 7.742238998413086,
      "learning_rate": 1.5642076887197527e-05,
      "loss": 2.6121700286865233,
      "memory(GiB)": 77.56,
      "step": 86495,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 3.7059251960070263,
      "grad_norm": 5.112504959106445,
      "learning_rate": 1.563718796736962e-05,
      "loss": 2.178631401062012,
      "memory(GiB)": 77.56,
      "step": 86500,
      "token_acc": 0.546031746031746,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.7059251960070263,
      "eval_loss": 2.3018243312835693,
      "eval_runtime": 14.2816,
      "eval_samples_per_second": 7.002,
      "eval_steps_per_second": 7.002,
      "eval_token_acc": 0.4769433465085639,
      "step": 86500
    },
    {
      "epoch": 3.706139411336275,
      "grad_norm": 7.568201065063477,
      "learning_rate": 1.5632299670056296e-05,
      "loss": 2.3550209045410155,
      "memory(GiB)": 77.56,
      "step": 86505,
      "token_acc": 0.4878286270691334,
      "train_speed(iter/s)": 1.437619
    },
    {
      "epoch": 3.706353626665524,
      "grad_norm": 5.062838554382324,
      "learning_rate": 1.5627411995346105e-05,
      "loss": 2.4407291412353516,
      "memory(GiB)": 77.56,
      "step": 86510,
      "token_acc": 0.5201342281879194,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 3.706567841994773,
      "grad_norm": 5.365954875946045,
      "learning_rate": 1.5622524943327567e-05,
      "loss": 2.4155345916748048,
      "memory(GiB)": 77.56,
      "step": 86515,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437625
    },
    {
      "epoch": 3.706782057324022,
      "grad_norm": 5.632654190063477,
      "learning_rate": 1.5617638514089237e-05,
      "loss": 2.4387001037597655,
      "memory(GiB)": 77.56,
      "step": 86520,
      "token_acc": 0.5190311418685121,
      "train_speed(iter/s)": 1.437617
    },
    {
      "epoch": 3.706996272653271,
      "grad_norm": 5.162031173706055,
      "learning_rate": 1.561275270771966e-05,
      "loss": 2.467332458496094,
      "memory(GiB)": 77.56,
      "step": 86525,
      "token_acc": 0.4874551971326165,
      "train_speed(iter/s)": 1.437609
    },
    {
      "epoch": 3.70721048798252,
      "grad_norm": 5.403102397918701,
      "learning_rate": 1.5607867524307323e-05,
      "loss": 2.348347473144531,
      "memory(GiB)": 77.56,
      "step": 86530,
      "token_acc": 0.4747191011235955,
      "train_speed(iter/s)": 1.437607
    },
    {
      "epoch": 3.707424703311769,
      "grad_norm": 5.386386394500732,
      "learning_rate": 1.5602982963940726e-05,
      "loss": 1.9832290649414062,
      "memory(GiB)": 77.56,
      "step": 86535,
      "token_acc": 0.5397350993377483,
      "train_speed(iter/s)": 1.437617
    },
    {
      "epoch": 3.7076389186410177,
      "grad_norm": 8.20964527130127,
      "learning_rate": 1.559809902670836e-05,
      "loss": 2.6929420471191405,
      "memory(GiB)": 77.56,
      "step": 86540,
      "token_acc": 0.45064377682403434,
      "train_speed(iter/s)": 1.437603
    },
    {
      "epoch": 3.707853133970267,
      "grad_norm": 4.948558807373047,
      "learning_rate": 1.5593215712698705e-05,
      "loss": 2.3050155639648438,
      "memory(GiB)": 77.56,
      "step": 86545,
      "token_acc": 0.47678018575851394,
      "train_speed(iter/s)": 1.437601
    },
    {
      "epoch": 3.7080673492995158,
      "grad_norm": 5.862262725830078,
      "learning_rate": 1.5588333022000212e-05,
      "loss": 2.6217864990234374,
      "memory(GiB)": 77.56,
      "step": 86550,
      "token_acc": 0.4527027027027027,
      "train_speed(iter/s)": 1.437608
    },
    {
      "epoch": 3.7082815646287646,
      "grad_norm": 5.878262996673584,
      "learning_rate": 1.5583450954701356e-05,
      "loss": 2.3322378158569337,
      "memory(GiB)": 77.56,
      "step": 86555,
      "token_acc": 0.497907949790795,
      "train_speed(iter/s)": 1.437634
    },
    {
      "epoch": 3.708495779958014,
      "grad_norm": 7.144289493560791,
      "learning_rate": 1.5578569510890574e-05,
      "loss": 2.5948171615600586,
      "memory(GiB)": 77.56,
      "step": 86560,
      "token_acc": 0.4423076923076923,
      "train_speed(iter/s)": 1.437636
    },
    {
      "epoch": 3.7087099952872626,
      "grad_norm": 4.589893341064453,
      "learning_rate": 1.557368869065629e-05,
      "loss": 2.1795501708984375,
      "memory(GiB)": 77.56,
      "step": 86565,
      "token_acc": 0.49340369393139843,
      "train_speed(iter/s)": 1.437626
    },
    {
      "epoch": 3.7089242106165115,
      "grad_norm": 5.418041229248047,
      "learning_rate": 1.5568808494086933e-05,
      "loss": 2.303356742858887,
      "memory(GiB)": 77.56,
      "step": 86570,
      "token_acc": 0.5408805031446541,
      "train_speed(iter/s)": 1.437649
    },
    {
      "epoch": 3.7091384259457607,
      "grad_norm": 6.835787296295166,
      "learning_rate": 1.5563928921270898e-05,
      "loss": 2.056692123413086,
      "memory(GiB)": 77.56,
      "step": 86575,
      "token_acc": 0.5284810126582279,
      "train_speed(iter/s)": 1.437652
    },
    {
      "epoch": 3.7093526412750095,
      "grad_norm": 5.318356037139893,
      "learning_rate": 1.5559049972296606e-05,
      "loss": 2.4325708389282226,
      "memory(GiB)": 77.56,
      "step": 86580,
      "token_acc": 0.5016835016835017,
      "train_speed(iter/s)": 1.437662
    },
    {
      "epoch": 3.7095668566042583,
      "grad_norm": 5.4713311195373535,
      "learning_rate": 1.555417164725243e-05,
      "loss": 2.4357158660888674,
      "memory(GiB)": 77.56,
      "step": 86585,
      "token_acc": 0.4835820895522388,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 3.7097810719335076,
      "grad_norm": 6.023533821105957,
      "learning_rate": 1.5549293946226734e-05,
      "loss": 2.059641456604004,
      "memory(GiB)": 77.56,
      "step": 86590,
      "token_acc": 0.5669014084507042,
      "train_speed(iter/s)": 1.437688
    },
    {
      "epoch": 3.7099952872627564,
      "grad_norm": 7.49127721786499,
      "learning_rate": 1.5544416869307915e-05,
      "loss": 2.3057533264160157,
      "memory(GiB)": 77.56,
      "step": 86595,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 3.710209502592005,
      "grad_norm": 6.071860313415527,
      "learning_rate": 1.5539540416584304e-05,
      "loss": 2.3259809494018553,
      "memory(GiB)": 77.56,
      "step": 86600,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.437672
    },
    {
      "epoch": 3.7104237179212545,
      "grad_norm": 4.546608924865723,
      "learning_rate": 1.5534664588144226e-05,
      "loss": 2.314710235595703,
      "memory(GiB)": 77.56,
      "step": 86605,
      "token_acc": 0.5258064516129032,
      "train_speed(iter/s)": 1.437664
    },
    {
      "epoch": 3.7106379332505033,
      "grad_norm": 11.2535982131958,
      "learning_rate": 1.552978938407605e-05,
      "loss": 2.4088861465454103,
      "memory(GiB)": 77.56,
      "step": 86610,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.710852148579752,
      "grad_norm": 5.248917102813721,
      "learning_rate": 1.552491480446807e-05,
      "loss": 2.4195835113525392,
      "memory(GiB)": 77.56,
      "step": 86615,
      "token_acc": 0.49673202614379086,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 3.7110663639090014,
      "grad_norm": 6.9785051345825195,
      "learning_rate": 1.55200408494086e-05,
      "loss": 2.130632781982422,
      "memory(GiB)": 77.56,
      "step": 86620,
      "token_acc": 0.5714285714285714,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 3.71128057923825,
      "grad_norm": 5.133209228515625,
      "learning_rate": 1.5515167518985933e-05,
      "loss": 2.139729690551758,
      "memory(GiB)": 77.56,
      "step": 86625,
      "token_acc": 0.5462555066079295,
      "train_speed(iter/s)": 1.437708
    },
    {
      "epoch": 3.711494794567499,
      "grad_norm": 5.0500359535217285,
      "learning_rate": 1.551029481328836e-05,
      "loss": 2.1852933883666994,
      "memory(GiB)": 77.56,
      "step": 86630,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437708
    },
    {
      "epoch": 3.7117090098967482,
      "grad_norm": 5.2850518226623535,
      "learning_rate": 1.5505422732404134e-05,
      "loss": 2.4500160217285156,
      "memory(GiB)": 77.56,
      "step": 86635,
      "token_acc": 0.49458483754512633,
      "train_speed(iter/s)": 1.43771
    },
    {
      "epoch": 3.711923225225997,
      "grad_norm": 5.476200103759766,
      "learning_rate": 1.550055127642155e-05,
      "loss": 2.3460809707641603,
      "memory(GiB)": 77.56,
      "step": 86640,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 3.712137440555246,
      "grad_norm": 4.466678142547607,
      "learning_rate": 1.549568044542884e-05,
      "loss": 2.41572265625,
      "memory(GiB)": 77.56,
      "step": 86645,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 3.712351655884495,
      "grad_norm": 9.083475112915039,
      "learning_rate": 1.5490810239514254e-05,
      "loss": 2.4300527572631836,
      "memory(GiB)": 77.56,
      "step": 86650,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 3.712565871213744,
      "grad_norm": 5.925046920776367,
      "learning_rate": 1.5485940658766e-05,
      "loss": 2.495985984802246,
      "memory(GiB)": 77.56,
      "step": 86655,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.437746
    },
    {
      "epoch": 3.7127800865429927,
      "grad_norm": 7.50604248046875,
      "learning_rate": 1.5481071703272304e-05,
      "loss": 2.566284942626953,
      "memory(GiB)": 77.56,
      "step": 86660,
      "token_acc": 0.4526627218934911,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.712994301872242,
      "grad_norm": 5.756147384643555,
      "learning_rate": 1.547620337312137e-05,
      "loss": 2.1826284408569334,
      "memory(GiB)": 77.56,
      "step": 86665,
      "token_acc": 0.5674603174603174,
      "train_speed(iter/s)": 1.437744
    },
    {
      "epoch": 3.713208517201491,
      "grad_norm": 4.954468727111816,
      "learning_rate": 1.547133566840141e-05,
      "loss": 2.473418045043945,
      "memory(GiB)": 77.56,
      "step": 86670,
      "token_acc": 0.5060975609756098,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.7134227325307396,
      "grad_norm": 8.725044250488281,
      "learning_rate": 1.54664685892006e-05,
      "loss": 2.4450586318969725,
      "memory(GiB)": 77.56,
      "step": 86675,
      "token_acc": 0.4748427672955975,
      "train_speed(iter/s)": 1.437756
    },
    {
      "epoch": 3.713636947859989,
      "grad_norm": 6.033003807067871,
      "learning_rate": 1.5461602135607105e-05,
      "loss": 2.224970245361328,
      "memory(GiB)": 77.56,
      "step": 86680,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.7138511631892377,
      "grad_norm": 5.881270408630371,
      "learning_rate": 1.545673630770909e-05,
      "loss": 2.233582305908203,
      "memory(GiB)": 77.56,
      "step": 86685,
      "token_acc": 0.4980694980694981,
      "train_speed(iter/s)": 1.437779
    },
    {
      "epoch": 3.714065378518487,
      "grad_norm": 5.8360514640808105,
      "learning_rate": 1.5451871105594696e-05,
      "loss": 2.558456611633301,
      "memory(GiB)": 77.56,
      "step": 86690,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.437799
    },
    {
      "epoch": 3.7142795938477358,
      "grad_norm": 5.247663974761963,
      "learning_rate": 1.5447006529352054e-05,
      "loss": 2.6439434051513673,
      "memory(GiB)": 77.56,
      "step": 86695,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.437828
    },
    {
      "epoch": 3.7144938091769846,
      "grad_norm": 7.151357173919678,
      "learning_rate": 1.5442142579069315e-05,
      "loss": 2.2906425476074217,
      "memory(GiB)": 77.56,
      "step": 86700,
      "token_acc": 0.5016835016835017,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 3.714708024506234,
      "grad_norm": 6.617393493652344,
      "learning_rate": 1.543727925483458e-05,
      "loss": 2.353643226623535,
      "memory(GiB)": 77.56,
      "step": 86705,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 3.7149222398354826,
      "grad_norm": 6.362409591674805,
      "learning_rate": 1.5432416556735957e-05,
      "loss": 2.3665643692016602,
      "memory(GiB)": 77.56,
      "step": 86710,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 3.7151364551647315,
      "grad_norm": 6.409313201904297,
      "learning_rate": 1.5427554484861532e-05,
      "loss": 2.2252857208251955,
      "memory(GiB)": 77.56,
      "step": 86715,
      "token_acc": 0.5326460481099656,
      "train_speed(iter/s)": 1.437803
    },
    {
      "epoch": 3.7153506704939807,
      "grad_norm": 5.6080241203308105,
      "learning_rate": 1.542269303929938e-05,
      "loss": 2.3943241119384764,
      "memory(GiB)": 77.56,
      "step": 86720,
      "token_acc": 0.5373665480427047,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 3.7155648858232295,
      "grad_norm": 6.60405158996582,
      "learning_rate": 1.541783222013759e-05,
      "loss": 2.6751174926757812,
      "memory(GiB)": 77.56,
      "step": 86725,
      "token_acc": 0.45993031358885017,
      "train_speed(iter/s)": 1.437835
    },
    {
      "epoch": 3.7157791011524783,
      "grad_norm": 6.042786121368408,
      "learning_rate": 1.5412972027464213e-05,
      "loss": 2.3408435821533202,
      "memory(GiB)": 77.56,
      "step": 86730,
      "token_acc": 0.5267379679144385,
      "train_speed(iter/s)": 1.437817
    },
    {
      "epoch": 3.7159933164817276,
      "grad_norm": 6.136815547943115,
      "learning_rate": 1.5408112461367268e-05,
      "loss": 2.488810348510742,
      "memory(GiB)": 77.56,
      "step": 86735,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.437833
    },
    {
      "epoch": 3.7162075318109764,
      "grad_norm": 6.867037296295166,
      "learning_rate": 1.540325352193484e-05,
      "loss": 2.4635574340820314,
      "memory(GiB)": 77.56,
      "step": 86740,
      "token_acc": 0.46630727762803237,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.716421747140225,
      "grad_norm": 7.39055871963501,
      "learning_rate": 1.539839520925493e-05,
      "loss": 2.1807117462158203,
      "memory(GiB)": 77.56,
      "step": 86745,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 3.7166359624694745,
      "grad_norm": 5.075364589691162,
      "learning_rate": 1.539353752341553e-05,
      "loss": 2.7658105850219727,
      "memory(GiB)": 77.56,
      "step": 86750,
      "token_acc": 0.4906832298136646,
      "train_speed(iter/s)": 1.43784
    },
    {
      "epoch": 3.7168501777987233,
      "grad_norm": 5.643764495849609,
      "learning_rate": 1.538868046450468e-05,
      "loss": 2.5875982284545898,
      "memory(GiB)": 77.56,
      "step": 86755,
      "token_acc": 0.4692556634304207,
      "train_speed(iter/s)": 1.437847
    },
    {
      "epoch": 3.717064393127972,
      "grad_norm": 6.702514171600342,
      "learning_rate": 1.538382403261035e-05,
      "loss": 2.336283493041992,
      "memory(GiB)": 77.56,
      "step": 86760,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.437855
    },
    {
      "epoch": 3.7172786084572214,
      "grad_norm": 4.063947677612305,
      "learning_rate": 1.5378968227820518e-05,
      "loss": 2.2570259094238283,
      "memory(GiB)": 77.56,
      "step": 86765,
      "token_acc": 0.4835820895522388,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 3.71749282378647,
      "grad_norm": 5.202474117279053,
      "learning_rate": 1.5374113050223153e-05,
      "loss": 2.0593425750732424,
      "memory(GiB)": 77.56,
      "step": 86770,
      "token_acc": 0.5598591549295775,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 3.717707039115719,
      "grad_norm": 5.654252052307129,
      "learning_rate": 1.5369258499906215e-05,
      "loss": 2.3037891387939453,
      "memory(GiB)": 77.56,
      "step": 86775,
      "token_acc": 0.5032051282051282,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 3.7179212544449682,
      "grad_norm": 5.633025646209717,
      "learning_rate": 1.536440457695763e-05,
      "loss": 2.3663572311401366,
      "memory(GiB)": 77.56,
      "step": 86780,
      "token_acc": 0.4901315789473684,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 3.718135469774217,
      "grad_norm": 5.767129421234131,
      "learning_rate": 1.5359551281465363e-05,
      "loss": 2.3048080444335937,
      "memory(GiB)": 77.56,
      "step": 86785,
      "token_acc": 0.49850746268656715,
      "train_speed(iter/s)": 1.437849
    },
    {
      "epoch": 3.718349685103466,
      "grad_norm": 6.4442057609558105,
      "learning_rate": 1.5354698613517316e-05,
      "loss": 2.4966196060180663,
      "memory(GiB)": 77.56,
      "step": 86790,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.437847
    },
    {
      "epoch": 3.718563900432715,
      "grad_norm": 5.920304298400879,
      "learning_rate": 1.534984657320141e-05,
      "loss": 2.462950897216797,
      "memory(GiB)": 77.56,
      "step": 86795,
      "token_acc": 0.4792332268370607,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 3.718778115761964,
      "grad_norm": 4.767179012298584,
      "learning_rate": 1.534499516060553e-05,
      "loss": 2.178111267089844,
      "memory(GiB)": 77.56,
      "step": 86800,
      "token_acc": 0.5137931034482759,
      "train_speed(iter/s)": 1.437882
    },
    {
      "epoch": 3.7189923310912127,
      "grad_norm": 5.359717845916748,
      "learning_rate": 1.534014437581756e-05,
      "loss": 2.2563379287719725,
      "memory(GiB)": 77.56,
      "step": 86805,
      "token_acc": 0.5348101265822784,
      "train_speed(iter/s)": 1.437902
    },
    {
      "epoch": 3.719206546420462,
      "grad_norm": 7.186066150665283,
      "learning_rate": 1.5335294218925388e-05,
      "loss": 2.443486213684082,
      "memory(GiB)": 77.56,
      "step": 86810,
      "token_acc": 0.5141955835962145,
      "train_speed(iter/s)": 1.43792
    },
    {
      "epoch": 3.719420761749711,
      "grad_norm": 6.606617450714111,
      "learning_rate": 1.53304446900169e-05,
      "loss": 2.1772224426269533,
      "memory(GiB)": 77.56,
      "step": 86815,
      "token_acc": 0.5358649789029536,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 3.7196349770789596,
      "grad_norm": 5.413238525390625,
      "learning_rate": 1.532559578917992e-05,
      "loss": 2.5105783462524416,
      "memory(GiB)": 77.56,
      "step": 86820,
      "token_acc": 0.4965034965034965,
      "train_speed(iter/s)": 1.437936
    },
    {
      "epoch": 3.719849192408209,
      "grad_norm": 7.256813049316406,
      "learning_rate": 1.5320747516502303e-05,
      "loss": 2.2638626098632812,
      "memory(GiB)": 77.56,
      "step": 86825,
      "token_acc": 0.5338078291814946,
      "train_speed(iter/s)": 1.437947
    },
    {
      "epoch": 3.7200634077374577,
      "grad_norm": 5.982698440551758,
      "learning_rate": 1.5315899872071882e-05,
      "loss": 2.3710479736328125,
      "memory(GiB)": 77.56,
      "step": 86830,
      "token_acc": 0.47093023255813954,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.7202776230667065,
      "grad_norm": 5.053171634674072,
      "learning_rate": 1.5311052855976465e-05,
      "loss": 2.4563880920410157,
      "memory(GiB)": 77.56,
      "step": 86835,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 3.7204918383959558,
      "grad_norm": 4.804139137268066,
      "learning_rate": 1.530620646830385e-05,
      "loss": 2.5838596343994142,
      "memory(GiB)": 77.56,
      "step": 86840,
      "token_acc": 0.4943181818181818,
      "train_speed(iter/s)": 1.437948
    },
    {
      "epoch": 3.7207060537252046,
      "grad_norm": 6.840320110321045,
      "learning_rate": 1.530136070914187e-05,
      "loss": 2.3054986953735352,
      "memory(GiB)": 77.56,
      "step": 86845,
      "token_acc": 0.48214285714285715,
      "train_speed(iter/s)": 1.437955
    },
    {
      "epoch": 3.7209202690544534,
      "grad_norm": 5.580686092376709,
      "learning_rate": 1.5296515578578285e-05,
      "loss": 2.542148208618164,
      "memory(GiB)": 77.56,
      "step": 86850,
      "token_acc": 0.48441926345609065,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 3.7211344843837026,
      "grad_norm": 5.10511589050293,
      "learning_rate": 1.5291671076700882e-05,
      "loss": 1.8831974029541017,
      "memory(GiB)": 77.56,
      "step": 86855,
      "token_acc": 0.5597269624573379,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.7213486997129515,
      "grad_norm": 6.026157379150391,
      "learning_rate": 1.5286827203597414e-05,
      "loss": 2.4431621551513674,
      "memory(GiB)": 77.56,
      "step": 86860,
      "token_acc": 0.4731182795698925,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 3.7215629150422003,
      "grad_norm": 5.7920098304748535,
      "learning_rate": 1.5281983959355615e-05,
      "loss": 2.1135372161865233,
      "memory(GiB)": 77.56,
      "step": 86865,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.7217771303714495,
      "grad_norm": 4.803773403167725,
      "learning_rate": 1.5277141344063267e-05,
      "loss": 2.298733711242676,
      "memory(GiB)": 77.56,
      "step": 86870,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.7219913457006983,
      "grad_norm": 6.743514537811279,
      "learning_rate": 1.5272299357808072e-05,
      "loss": 2.7616195678710938,
      "memory(GiB)": 77.56,
      "step": 86875,
      "token_acc": 0.46735395189003437,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 3.722205561029947,
      "grad_norm": 5.747242450714111,
      "learning_rate": 1.5267458000677753e-05,
      "loss": 2.2895477294921873,
      "memory(GiB)": 77.56,
      "step": 86880,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 3.7224197763591964,
      "grad_norm": 4.466111660003662,
      "learning_rate": 1.5262617272759993e-05,
      "loss": 2.348099708557129,
      "memory(GiB)": 77.56,
      "step": 86885,
      "token_acc": 0.506896551724138,
      "train_speed(iter/s)": 1.438021
    },
    {
      "epoch": 3.722633991688445,
      "grad_norm": 6.285738468170166,
      "learning_rate": 1.5257777174142529e-05,
      "loss": 2.3186534881591796,
      "memory(GiB)": 77.56,
      "step": 86890,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438019
    },
    {
      "epoch": 3.722848207017694,
      "grad_norm": 5.389781475067139,
      "learning_rate": 1.5252937704913006e-05,
      "loss": 2.3748050689697267,
      "memory(GiB)": 77.56,
      "step": 86895,
      "token_acc": 0.4520547945205479,
      "train_speed(iter/s)": 1.438004
    },
    {
      "epoch": 3.7230624223469433,
      "grad_norm": 8.105585098266602,
      "learning_rate": 1.5248098865159127e-05,
      "loss": 2.4184099197387696,
      "memory(GiB)": 77.56,
      "step": 86900,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 3.723276637676192,
      "grad_norm": 7.014865875244141,
      "learning_rate": 1.5243260654968539e-05,
      "loss": 2.164982223510742,
      "memory(GiB)": 77.56,
      "step": 86905,
      "token_acc": 0.540650406504065,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 3.723490853005441,
      "grad_norm": 7.801687717437744,
      "learning_rate": 1.5238423074428888e-05,
      "loss": 2.7186166763305666,
      "memory(GiB)": 77.56,
      "step": 86910,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 3.72370506833469,
      "grad_norm": 6.729595184326172,
      "learning_rate": 1.5233586123627807e-05,
      "loss": 2.218162727355957,
      "memory(GiB)": 77.56,
      "step": 86915,
      "token_acc": 0.5181518151815182,
      "train_speed(iter/s)": 1.438007
    },
    {
      "epoch": 3.723919283663939,
      "grad_norm": 5.841404914855957,
      "learning_rate": 1.5228749802652932e-05,
      "loss": 2.3726444244384766,
      "memory(GiB)": 77.56,
      "step": 86920,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438001
    },
    {
      "epoch": 3.724133498993188,
      "grad_norm": 5.373078346252441,
      "learning_rate": 1.5223914111591853e-05,
      "loss": 2.351149559020996,
      "memory(GiB)": 77.56,
      "step": 86925,
      "token_acc": 0.5031847133757962,
      "train_speed(iter/s)": 1.438016
    },
    {
      "epoch": 3.724347714322437,
      "grad_norm": 4.831430912017822,
      "learning_rate": 1.5219079050532209e-05,
      "loss": 2.024088478088379,
      "memory(GiB)": 77.56,
      "step": 86930,
      "token_acc": 0.5424836601307189,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 3.724561929651686,
      "grad_norm": 5.509981155395508,
      "learning_rate": 1.521424461956158e-05,
      "loss": 2.1668506622314454,
      "memory(GiB)": 77.56,
      "step": 86935,
      "token_acc": 0.5132450331125827,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.7247761449809347,
      "grad_norm": 4.685031890869141,
      "learning_rate": 1.5209410818767539e-05,
      "loss": 2.3187753677368166,
      "memory(GiB)": 77.56,
      "step": 86940,
      "token_acc": 0.5018587360594795,
      "train_speed(iter/s)": 1.437981
    },
    {
      "epoch": 3.724990360310184,
      "grad_norm": 5.680825233459473,
      "learning_rate": 1.5204577648237656e-05,
      "loss": 2.278506278991699,
      "memory(GiB)": 77.56,
      "step": 86945,
      "token_acc": 0.4752475247524752,
      "train_speed(iter/s)": 1.437987
    },
    {
      "epoch": 3.7252045756394327,
      "grad_norm": 5.809874057769775,
      "learning_rate": 1.519974510805947e-05,
      "loss": 2.27001838684082,
      "memory(GiB)": 77.56,
      "step": 86950,
      "token_acc": 0.5139318885448917,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.7254187909686816,
      "grad_norm": 5.0600266456604,
      "learning_rate": 1.519491319832057e-05,
      "loss": 2.3480995178222654,
      "memory(GiB)": 77.56,
      "step": 86955,
      "token_acc": 0.495114006514658,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 3.725633006297931,
      "grad_norm": 5.149813175201416,
      "learning_rate": 1.5190081919108445e-05,
      "loss": 2.400051307678223,
      "memory(GiB)": 77.56,
      "step": 86960,
      "token_acc": 0.5016835016835017,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.7258472216271796,
      "grad_norm": 6.670567989349365,
      "learning_rate": 1.5185251270510659e-05,
      "loss": 2.245740509033203,
      "memory(GiB)": 77.56,
      "step": 86965,
      "token_acc": 0.51953125,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.7260614369564284,
      "grad_norm": 7.417488098144531,
      "learning_rate": 1.5180421252614707e-05,
      "loss": 2.519170951843262,
      "memory(GiB)": 77.56,
      "step": 86970,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 3.7262756522856777,
      "grad_norm": 5.339607238769531,
      "learning_rate": 1.5175591865508082e-05,
      "loss": 2.417881965637207,
      "memory(GiB)": 77.56,
      "step": 86975,
      "token_acc": 0.50814332247557,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 3.7264898676149265,
      "grad_norm": 5.728578567504883,
      "learning_rate": 1.5170763109278285e-05,
      "loss": 2.128196144104004,
      "memory(GiB)": 77.56,
      "step": 86980,
      "token_acc": 0.5648535564853556,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 3.7267040829441753,
      "grad_norm": 5.405481338500977,
      "learning_rate": 1.5165934984012769e-05,
      "loss": 2.3688072204589843,
      "memory(GiB)": 77.56,
      "step": 86985,
      "token_acc": 0.5227272727272727,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 3.7269182982734246,
      "grad_norm": 6.180614471435547,
      "learning_rate": 1.5161107489799032e-05,
      "loss": 2.276973915100098,
      "memory(GiB)": 77.56,
      "step": 86990,
      "token_acc": 0.5278688524590164,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 3.7271325136026734,
      "grad_norm": 5.205055236816406,
      "learning_rate": 1.5156280626724512e-05,
      "loss": 1.963690185546875,
      "memory(GiB)": 77.56,
      "step": 86995,
      "token_acc": 0.5903614457831325,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 3.727346728931922,
      "grad_norm": 7.002117156982422,
      "learning_rate": 1.5151454394876658e-05,
      "loss": 2.2104793548583985,
      "memory(GiB)": 77.56,
      "step": 87000,
      "token_acc": 0.5405405405405406,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 3.727346728931922,
      "eval_loss": 2.308195114135742,
      "eval_runtime": 14.3281,
      "eval_samples_per_second": 6.979,
      "eval_steps_per_second": 6.979,
      "eval_token_acc": 0.4654696132596685,
      "step": 87000
    },
    {
      "epoch": 3.7275609442611715,
      "grad_norm": 7.07882833480835,
      "learning_rate": 1.5146628794342898e-05,
      "loss": 2.3575599670410154,
      "memory(GiB)": 77.56,
      "step": 87005,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.437699
    },
    {
      "epoch": 3.7277751595904203,
      "grad_norm": 5.945185661315918,
      "learning_rate": 1.5141803825210637e-05,
      "loss": 2.127720832824707,
      "memory(GiB)": 77.56,
      "step": 87010,
      "token_acc": 0.5450819672131147,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 3.727989374919669,
      "grad_norm": 5.8800435066223145,
      "learning_rate": 1.5136979487567315e-05,
      "loss": 2.278525543212891,
      "memory(GiB)": 77.56,
      "step": 87015,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 3.7282035902489183,
      "grad_norm": 5.666601181030273,
      "learning_rate": 1.5132155781500313e-05,
      "loss": 2.3060993194580077,
      "memory(GiB)": 77.56,
      "step": 87020,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.437723
    },
    {
      "epoch": 3.728417805578167,
      "grad_norm": 5.1239705085754395,
      "learning_rate": 1.5127332707097013e-05,
      "loss": 2.422837257385254,
      "memory(GiB)": 77.56,
      "step": 87025,
      "token_acc": 0.486646884272997,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 3.7286320209074164,
      "grad_norm": 7.413276672363281,
      "learning_rate": 1.5122510264444784e-05,
      "loss": 2.3034963607788086,
      "memory(GiB)": 77.56,
      "step": 87030,
      "token_acc": 0.5133333333333333,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 3.7288462362366652,
      "grad_norm": 5.47199821472168,
      "learning_rate": 1.5117688453631019e-05,
      "loss": 2.3623912811279295,
      "memory(GiB)": 77.56,
      "step": 87035,
      "token_acc": 0.48955223880597015,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 3.729060451565914,
      "grad_norm": 6.05703592300415,
      "learning_rate": 1.5112867274743026e-05,
      "loss": 2.0504739761352537,
      "memory(GiB)": 77.56,
      "step": 87040,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.437733
    },
    {
      "epoch": 3.7292746668951633,
      "grad_norm": 7.642071723937988,
      "learning_rate": 1.5108046727868186e-05,
      "loss": 2.7841732025146486,
      "memory(GiB)": 77.56,
      "step": 87045,
      "token_acc": 0.4701195219123506,
      "train_speed(iter/s)": 1.437711
    },
    {
      "epoch": 3.729488882224412,
      "grad_norm": 5.824583053588867,
      "learning_rate": 1.5103226813093813e-05,
      "loss": 2.2717166900634767,
      "memory(GiB)": 77.56,
      "step": 87050,
      "token_acc": 0.4891304347826087,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 3.729703097553661,
      "grad_norm": 5.409212589263916,
      "learning_rate": 1.5098407530507225e-05,
      "loss": 2.5261409759521483,
      "memory(GiB)": 77.56,
      "step": 87055,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 3.72991731288291,
      "grad_norm": 8.594744682312012,
      "learning_rate": 1.509358888019572e-05,
      "loss": 2.2982528686523436,
      "memory(GiB)": 77.56,
      "step": 87060,
      "token_acc": 0.5095057034220533,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.730131528212159,
      "grad_norm": 8.690960884094238,
      "learning_rate": 1.5088770862246599e-05,
      "loss": 2.225796127319336,
      "memory(GiB)": 77.56,
      "step": 87065,
      "token_acc": 0.5285714285714286,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 3.730345743541408,
      "grad_norm": 5.013505935668945,
      "learning_rate": 1.508395347674713e-05,
      "loss": 2.222603988647461,
      "memory(GiB)": 77.56,
      "step": 87070,
      "token_acc": 0.535593220338983,
      "train_speed(iter/s)": 1.437769
    },
    {
      "epoch": 3.730559958870657,
      "grad_norm": 5.395937442779541,
      "learning_rate": 1.507913672378461e-05,
      "loss": 2.2606481552124023,
      "memory(GiB)": 77.56,
      "step": 87075,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.437777
    },
    {
      "epoch": 3.730774174199906,
      "grad_norm": 6.910282135009766,
      "learning_rate": 1.507432060344629e-05,
      "loss": 2.3298110961914062,
      "memory(GiB)": 77.56,
      "step": 87080,
      "token_acc": 0.5164473684210527,
      "train_speed(iter/s)": 1.437781
    },
    {
      "epoch": 3.7309883895291547,
      "grad_norm": 5.100796699523926,
      "learning_rate": 1.506950511581941e-05,
      "loss": 2.4390256881713865,
      "memory(GiB)": 77.56,
      "step": 87085,
      "token_acc": 0.5014577259475219,
      "train_speed(iter/s)": 1.437799
    },
    {
      "epoch": 3.731202604858404,
      "grad_norm": 6.4759626388549805,
      "learning_rate": 1.5064690260991215e-05,
      "loss": 2.3322307586669924,
      "memory(GiB)": 77.56,
      "step": 87090,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.7314168201876527,
      "grad_norm": 6.136115550994873,
      "learning_rate": 1.5059876039048914e-05,
      "loss": 2.7116325378417967,
      "memory(GiB)": 77.56,
      "step": 87095,
      "token_acc": 0.4405797101449275,
      "train_speed(iter/s)": 1.437811
    },
    {
      "epoch": 3.7316310355169016,
      "grad_norm": 5.605592250823975,
      "learning_rate": 1.5055062450079749e-05,
      "loss": 2.574116516113281,
      "memory(GiB)": 77.56,
      "step": 87100,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 3.731845250846151,
      "grad_norm": 9.748913764953613,
      "learning_rate": 1.5050249494170893e-05,
      "loss": 2.479766845703125,
      "memory(GiB)": 77.56,
      "step": 87105,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.7320594661753996,
      "grad_norm": 5.765749931335449,
      "learning_rate": 1.5045437171409571e-05,
      "loss": 2.371507263183594,
      "memory(GiB)": 77.56,
      "step": 87110,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 3.7322736815046484,
      "grad_norm": 6.0616559982299805,
      "learning_rate": 1.5040625481882942e-05,
      "loss": 2.424093246459961,
      "memory(GiB)": 77.56,
      "step": 87115,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 3.7324878968338977,
      "grad_norm": 4.634643077850342,
      "learning_rate": 1.5035814425678174e-05,
      "loss": 2.2948190689086916,
      "memory(GiB)": 77.56,
      "step": 87120,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.7327021121631465,
      "grad_norm": 5.670689582824707,
      "learning_rate": 1.5031004002882431e-05,
      "loss": 2.2508188247680665,
      "memory(GiB)": 77.56,
      "step": 87125,
      "token_acc": 0.5309446254071661,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 3.7329163274923953,
      "grad_norm": 7.1844801902771,
      "learning_rate": 1.502619421358284e-05,
      "loss": 2.478338623046875,
      "memory(GiB)": 77.56,
      "step": 87130,
      "token_acc": 0.48297213622291024,
      "train_speed(iter/s)": 1.437867
    },
    {
      "epoch": 3.7331305428216446,
      "grad_norm": 5.184489727020264,
      "learning_rate": 1.502138505786656e-05,
      "loss": 2.2576826095581053,
      "memory(GiB)": 77.56,
      "step": 87135,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 3.7333447581508934,
      "grad_norm": 5.93429708480835,
      "learning_rate": 1.5016576535820708e-05,
      "loss": 2.3770168304443358,
      "memory(GiB)": 77.56,
      "step": 87140,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.437874
    },
    {
      "epoch": 3.733558973480142,
      "grad_norm": 7.028170585632324,
      "learning_rate": 1.5011768647532382e-05,
      "loss": 2.397247314453125,
      "memory(GiB)": 77.56,
      "step": 87145,
      "token_acc": 0.4647058823529412,
      "train_speed(iter/s)": 1.437867
    },
    {
      "epoch": 3.7337731888093915,
      "grad_norm": 5.135748386383057,
      "learning_rate": 1.5006961393088692e-05,
      "loss": 2.169891357421875,
      "memory(GiB)": 77.56,
      "step": 87150,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 3.7339874041386403,
      "grad_norm": 5.808743953704834,
      "learning_rate": 1.5002154772576709e-05,
      "loss": 2.238406753540039,
      "memory(GiB)": 77.56,
      "step": 87155,
      "token_acc": 0.5483870967741935,
      "train_speed(iter/s)": 1.43786
    },
    {
      "epoch": 3.734201619467889,
      "grad_norm": 6.619450569152832,
      "learning_rate": 1.4997348786083537e-05,
      "loss": 2.3706283569335938,
      "memory(GiB)": 77.56,
      "step": 87160,
      "token_acc": 0.5167173252279635,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 3.7344158347971383,
      "grad_norm": 4.534168243408203,
      "learning_rate": 1.4992543433696228e-05,
      "loss": 2.4198450088500976,
      "memory(GiB)": 77.56,
      "step": 87165,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.437872
    },
    {
      "epoch": 3.734630050126387,
      "grad_norm": 5.226425647735596,
      "learning_rate": 1.4987738715501832e-05,
      "loss": 2.453997993469238,
      "memory(GiB)": 77.56,
      "step": 87170,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 3.734844265455636,
      "grad_norm": 7.435589790344238,
      "learning_rate": 1.4982934631587391e-05,
      "loss": 2.308641815185547,
      "memory(GiB)": 77.56,
      "step": 87175,
      "token_acc": 0.4691358024691358,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 3.7350584807848852,
      "grad_norm": 4.9824910163879395,
      "learning_rate": 1.4978131182039928e-05,
      "loss": 2.2796201705932617,
      "memory(GiB)": 77.56,
      "step": 87180,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.437884
    },
    {
      "epoch": 3.735272696114134,
      "grad_norm": 5.569031715393066,
      "learning_rate": 1.4973328366946471e-05,
      "loss": 2.3210727691650392,
      "memory(GiB)": 77.56,
      "step": 87185,
      "token_acc": 0.5121212121212121,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.735486911443383,
      "grad_norm": 7.1741132736206055,
      "learning_rate": 1.4968526186394039e-05,
      "loss": 2.4107852935791017,
      "memory(GiB)": 77.56,
      "step": 87190,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.735701126772632,
      "grad_norm": 8.392370223999023,
      "learning_rate": 1.4963724640469622e-05,
      "loss": 2.515874481201172,
      "memory(GiB)": 77.56,
      "step": 87195,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.735915342101881,
      "grad_norm": 5.084343910217285,
      "learning_rate": 1.4958923729260198e-05,
      "loss": 2.220355415344238,
      "memory(GiB)": 77.56,
      "step": 87200,
      "token_acc": 0.490625,
      "train_speed(iter/s)": 1.437868
    },
    {
      "epoch": 3.7361295574311297,
      "grad_norm": 5.194983005523682,
      "learning_rate": 1.4954123452852742e-05,
      "loss": 2.4692991256713865,
      "memory(GiB)": 77.56,
      "step": 87205,
      "token_acc": 0.4507462686567164,
      "train_speed(iter/s)": 1.437884
    },
    {
      "epoch": 3.736343772760379,
      "grad_norm": 5.243771553039551,
      "learning_rate": 1.4949323811334214e-05,
      "loss": 2.344961166381836,
      "memory(GiB)": 77.56,
      "step": 87210,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.437902
    },
    {
      "epoch": 3.736557988089628,
      "grad_norm": 5.781335830688477,
      "learning_rate": 1.4944524804791554e-05,
      "loss": 2.2208852767944336,
      "memory(GiB)": 77.56,
      "step": 87215,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.437898
    },
    {
      "epoch": 3.7367722034188766,
      "grad_norm": 6.047597885131836,
      "learning_rate": 1.4939726433311723e-05,
      "loss": 2.0977304458618162,
      "memory(GiB)": 77.56,
      "step": 87220,
      "token_acc": 0.5391304347826087,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 3.736986418748126,
      "grad_norm": 6.536790370941162,
      "learning_rate": 1.4934928696981636e-05,
      "loss": 2.4487327575683593,
      "memory(GiB)": 77.56,
      "step": 87225,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.43792
    },
    {
      "epoch": 3.7372006340773747,
      "grad_norm": 6.325838565826416,
      "learning_rate": 1.4930131595888213e-05,
      "loss": 2.370090866088867,
      "memory(GiB)": 77.56,
      "step": 87230,
      "token_acc": 0.5381818181818182,
      "train_speed(iter/s)": 1.437948
    },
    {
      "epoch": 3.7374148494066235,
      "grad_norm": 5.113385200500488,
      "learning_rate": 1.4925335130118357e-05,
      "loss": 2.2227386474609374,
      "memory(GiB)": 77.56,
      "step": 87235,
      "token_acc": 0.5339805825242718,
      "train_speed(iter/s)": 1.437962
    },
    {
      "epoch": 3.7376290647358728,
      "grad_norm": 6.892491340637207,
      "learning_rate": 1.4920539299758935e-05,
      "loss": 2.34505615234375,
      "memory(GiB)": 77.56,
      "step": 87240,
      "token_acc": 0.517799352750809,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 3.7378432800651216,
      "grad_norm": 6.933368682861328,
      "learning_rate": 1.4915744104896872e-05,
      "loss": 2.5074304580688476,
      "memory(GiB)": 77.56,
      "step": 87245,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 3.7380574953943704,
      "grad_norm": 5.056584358215332,
      "learning_rate": 1.4910949545619013e-05,
      "loss": 2.4061588287353515,
      "memory(GiB)": 77.56,
      "step": 87250,
      "token_acc": 0.44135802469135804,
      "train_speed(iter/s)": 1.437959
    },
    {
      "epoch": 3.7382717107236196,
      "grad_norm": 5.506441116333008,
      "learning_rate": 1.4906155622012202e-05,
      "loss": 1.9020082473754882,
      "memory(GiB)": 77.56,
      "step": 87255,
      "token_acc": 0.6223175965665236,
      "train_speed(iter/s)": 1.437959
    },
    {
      "epoch": 3.7384859260528684,
      "grad_norm": 4.79023551940918,
      "learning_rate": 1.490136233416332e-05,
      "loss": 2.349620056152344,
      "memory(GiB)": 77.56,
      "step": 87260,
      "token_acc": 0.5220338983050847,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 3.7387001413821173,
      "grad_norm": 6.383564472198486,
      "learning_rate": 1.489656968215919e-05,
      "loss": 2.4896823883056642,
      "memory(GiB)": 77.56,
      "step": 87265,
      "token_acc": 0.45263157894736844,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 3.7389143567113665,
      "grad_norm": 7.178621292114258,
      "learning_rate": 1.4891777666086609e-05,
      "loss": 1.8542938232421875,
      "memory(GiB)": 77.56,
      "step": 87270,
      "token_acc": 0.5921052631578947,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 3.7391285720406153,
      "grad_norm": 6.1213603019714355,
      "learning_rate": 1.4886986286032423e-05,
      "loss": 2.352415657043457,
      "memory(GiB)": 77.56,
      "step": 87275,
      "token_acc": 0.538961038961039,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 3.739342787369864,
      "grad_norm": 6.763651371002197,
      "learning_rate": 1.4882195542083421e-05,
      "loss": 2.3152748107910157,
      "memory(GiB)": 77.56,
      "step": 87280,
      "token_acc": 0.5413223140495868,
      "train_speed(iter/s)": 1.437957
    },
    {
      "epoch": 3.7395570026991134,
      "grad_norm": 6.121639728546143,
      "learning_rate": 1.487740543432639e-05,
      "loss": 2.3437719345092773,
      "memory(GiB)": 77.56,
      "step": 87285,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.739771218028362,
      "grad_norm": 6.855044364929199,
      "learning_rate": 1.4872615962848113e-05,
      "loss": 2.464218521118164,
      "memory(GiB)": 77.56,
      "step": 87290,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 3.739985433357611,
      "grad_norm": 7.222046375274658,
      "learning_rate": 1.4867827127735346e-05,
      "loss": 2.4066207885742186,
      "memory(GiB)": 77.56,
      "step": 87295,
      "token_acc": 0.5090252707581228,
      "train_speed(iter/s)": 1.43794
    },
    {
      "epoch": 3.7401996486868603,
      "grad_norm": 5.3208231925964355,
      "learning_rate": 1.486303892907483e-05,
      "loss": 2.6841299057006838,
      "memory(GiB)": 77.56,
      "step": 87300,
      "token_acc": 0.498371335504886,
      "train_speed(iter/s)": 1.43792
    },
    {
      "epoch": 3.740413864016109,
      "grad_norm": 5.993175983428955,
      "learning_rate": 1.485825136695334e-05,
      "loss": 2.2455379486083986,
      "memory(GiB)": 77.56,
      "step": 87305,
      "token_acc": 0.527027027027027,
      "train_speed(iter/s)": 1.437946
    },
    {
      "epoch": 3.740628079345358,
      "grad_norm": 5.692388534545898,
      "learning_rate": 1.4853464441457593e-05,
      "loss": 2.338824653625488,
      "memory(GiB)": 77.56,
      "step": 87310,
      "token_acc": 0.4880546075085324,
      "train_speed(iter/s)": 1.437949
    },
    {
      "epoch": 3.740842294674607,
      "grad_norm": 6.645580768585205,
      "learning_rate": 1.4848678152674311e-05,
      "loss": 2.336594009399414,
      "memory(GiB)": 77.56,
      "step": 87315,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.437962
    },
    {
      "epoch": 3.741056510003856,
      "grad_norm": 3.9664742946624756,
      "learning_rate": 1.4843892500690193e-05,
      "loss": 2.3229135513305663,
      "memory(GiB)": 77.56,
      "step": 87320,
      "token_acc": 0.5066225165562914,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 3.741270725333105,
      "grad_norm": 5.8854780197143555,
      "learning_rate": 1.483910748559193e-05,
      "loss": 2.331338310241699,
      "memory(GiB)": 77.56,
      "step": 87325,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.741484940662354,
      "grad_norm": 8.000292778015137,
      "learning_rate": 1.4834323107466218e-05,
      "loss": 2.4718648910522463,
      "memory(GiB)": 77.56,
      "step": 87330,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.741699155991603,
      "grad_norm": 6.634707927703857,
      "learning_rate": 1.4829539366399747e-05,
      "loss": 2.270786666870117,
      "memory(GiB)": 77.56,
      "step": 87335,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.7419133713208517,
      "grad_norm": 5.0419416427612305,
      "learning_rate": 1.4824756262479161e-05,
      "loss": 2.4486883163452147,
      "memory(GiB)": 77.56,
      "step": 87340,
      "token_acc": 0.47560975609756095,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 3.742127586650101,
      "grad_norm": 6.889967918395996,
      "learning_rate": 1.4819973795791115e-05,
      "loss": 2.3378862380981444,
      "memory(GiB)": 77.56,
      "step": 87345,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 3.7423418019793497,
      "grad_norm": 6.417646884918213,
      "learning_rate": 1.4815191966422243e-05,
      "loss": 2.720916748046875,
      "memory(GiB)": 77.56,
      "step": 87350,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.7425560173085985,
      "grad_norm": 5.461677551269531,
      "learning_rate": 1.4810410774459171e-05,
      "loss": 2.0022937774658205,
      "memory(GiB)": 77.56,
      "step": 87355,
      "token_acc": 0.556420233463035,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 3.742770232637848,
      "grad_norm": 5.625781059265137,
      "learning_rate": 1.4805630219988508e-05,
      "loss": 2.521144485473633,
      "memory(GiB)": 77.56,
      "step": 87360,
      "token_acc": 0.47266881028938906,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 3.7429844479670966,
      "grad_norm": 7.273965358734131,
      "learning_rate": 1.4800850303096885e-05,
      "loss": 2.439421463012695,
      "memory(GiB)": 77.56,
      "step": 87365,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 3.7431986632963454,
      "grad_norm": 5.6455912590026855,
      "learning_rate": 1.4796071023870872e-05,
      "loss": 2.403054046630859,
      "memory(GiB)": 77.56,
      "step": 87370,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 3.7434128786255947,
      "grad_norm": 5.489974021911621,
      "learning_rate": 1.4791292382397064e-05,
      "loss": 2.015095901489258,
      "memory(GiB)": 77.56,
      "step": 87375,
      "token_acc": 0.5542168674698795,
      "train_speed(iter/s)": 1.438039
    },
    {
      "epoch": 3.7436270939548435,
      "grad_norm": 5.249879837036133,
      "learning_rate": 1.4786514378762017e-05,
      "loss": 2.2564437866210936,
      "memory(GiB)": 77.56,
      "step": 87380,
      "token_acc": 0.525,
      "train_speed(iter/s)": 1.438031
    },
    {
      "epoch": 3.7438413092840923,
      "grad_norm": 7.277451038360596,
      "learning_rate": 1.4781737013052282e-05,
      "loss": 2.1843563079833985,
      "memory(GiB)": 77.56,
      "step": 87385,
      "token_acc": 0.4796380090497738,
      "train_speed(iter/s)": 1.438038
    },
    {
      "epoch": 3.7440555246133416,
      "grad_norm": 6.090884685516357,
      "learning_rate": 1.4776960285354436e-05,
      "loss": 2.3574953079223633,
      "memory(GiB)": 77.56,
      "step": 87390,
      "token_acc": 0.5264900662251656,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 3.7442697399425904,
      "grad_norm": 5.9975361824035645,
      "learning_rate": 1.4772184195754996e-05,
      "loss": 2.3810352325439452,
      "memory(GiB)": 77.56,
      "step": 87395,
      "token_acc": 0.49230769230769234,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 3.744483955271839,
      "grad_norm": 6.648539066314697,
      "learning_rate": 1.4767408744340466e-05,
      "loss": 2.2519874572753906,
      "memory(GiB)": 77.56,
      "step": 87400,
      "token_acc": 0.504424778761062,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 3.7446981706010884,
      "grad_norm": 7.1107964515686035,
      "learning_rate": 1.4762633931197395e-05,
      "loss": 2.27294807434082,
      "memory(GiB)": 77.56,
      "step": 87405,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 3.7449123859303373,
      "grad_norm": 6.525959014892578,
      "learning_rate": 1.4757859756412268e-05,
      "loss": 2.2874944686889647,
      "memory(GiB)": 77.56,
      "step": 87410,
      "token_acc": 0.4928571428571429,
      "train_speed(iter/s)": 1.438074
    },
    {
      "epoch": 3.745126601259586,
      "grad_norm": 6.147908687591553,
      "learning_rate": 1.475308622007155e-05,
      "loss": 2.3837844848632814,
      "memory(GiB)": 77.56,
      "step": 87415,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.438082
    },
    {
      "epoch": 3.7453408165888353,
      "grad_norm": 7.439899921417236,
      "learning_rate": 1.4748313322261758e-05,
      "loss": 2.318966293334961,
      "memory(GiB)": 77.56,
      "step": 87420,
      "token_acc": 0.525691699604743,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 3.745555031918084,
      "grad_norm": 5.130218982696533,
      "learning_rate": 1.4743541063069339e-05,
      "loss": 2.0219400405883787,
      "memory(GiB)": 77.56,
      "step": 87425,
      "token_acc": 0.5078125,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 3.745769247247333,
      "grad_norm": 5.853952407836914,
      "learning_rate": 1.4738769442580746e-05,
      "loss": 2.1071950912475588,
      "memory(GiB)": 77.56,
      "step": 87430,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 3.745983462576582,
      "grad_norm": 5.421910762786865,
      "learning_rate": 1.4733998460882425e-05,
      "loss": 2.549492073059082,
      "memory(GiB)": 77.56,
      "step": 87435,
      "token_acc": 0.4258241758241758,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 3.746197677905831,
      "grad_norm": 7.759028434753418,
      "learning_rate": 1.4729228118060807e-05,
      "loss": 2.4368078231811525,
      "memory(GiB)": 77.56,
      "step": 87440,
      "token_acc": 0.49851632047477745,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 3.74641189323508,
      "grad_norm": 10.274950981140137,
      "learning_rate": 1.4724458414202285e-05,
      "loss": 2.1819526672363283,
      "memory(GiB)": 77.56,
      "step": 87445,
      "token_acc": 0.5057034220532319,
      "train_speed(iter/s)": 1.438104
    },
    {
      "epoch": 3.746626108564329,
      "grad_norm": 6.6081132888793945,
      "learning_rate": 1.4719689349393312e-05,
      "loss": 2.124695587158203,
      "memory(GiB)": 77.56,
      "step": 87450,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.438104
    },
    {
      "epoch": 3.746840323893578,
      "grad_norm": 5.244287967681885,
      "learning_rate": 1.4714920923720254e-05,
      "loss": 2.07525577545166,
      "memory(GiB)": 77.56,
      "step": 87455,
      "token_acc": 0.5389830508474577,
      "train_speed(iter/s)": 1.438088
    },
    {
      "epoch": 3.7470545392228267,
      "grad_norm": 6.249682426452637,
      "learning_rate": 1.471015313726951e-05,
      "loss": 2.4163894653320312,
      "memory(GiB)": 77.56,
      "step": 87460,
      "token_acc": 0.45126353790613716,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 3.747268754552076,
      "grad_norm": 5.447098255157471,
      "learning_rate": 1.4705385990127446e-05,
      "loss": 2.197842025756836,
      "memory(GiB)": 77.56,
      "step": 87465,
      "token_acc": 0.5285171102661597,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 3.747482969881325,
      "grad_norm": 6.029251575469971,
      "learning_rate": 1.4700619482380406e-05,
      "loss": 2.540346145629883,
      "memory(GiB)": 77.56,
      "step": 87470,
      "token_acc": 0.4720496894409938,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 3.7476971852105736,
      "grad_norm": 5.29820442199707,
      "learning_rate": 1.4695853614114763e-05,
      "loss": 2.3586051940917967,
      "memory(GiB)": 77.56,
      "step": 87475,
      "token_acc": 0.4812680115273775,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 3.747911400539823,
      "grad_norm": 6.930717945098877,
      "learning_rate": 1.4691088385416857e-05,
      "loss": 2.41739559173584,
      "memory(GiB)": 77.56,
      "step": 87480,
      "token_acc": 0.49809885931558934,
      "train_speed(iter/s)": 1.438081
    },
    {
      "epoch": 3.7481256158690717,
      "grad_norm": 7.671302318572998,
      "learning_rate": 1.4686323796373014e-05,
      "loss": 2.2959407806396483,
      "memory(GiB)": 77.56,
      "step": 87485,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 3.7483398311983205,
      "grad_norm": 6.608582973480225,
      "learning_rate": 1.4681559847069537e-05,
      "loss": 2.269548034667969,
      "memory(GiB)": 77.56,
      "step": 87490,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 3.7485540465275697,
      "grad_norm": 6.8144965171813965,
      "learning_rate": 1.467679653759274e-05,
      "loss": 2.2238227844238283,
      "memory(GiB)": 77.56,
      "step": 87495,
      "token_acc": 0.5275590551181102,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 3.7487682618568186,
      "grad_norm": 6.36323356628418,
      "learning_rate": 1.4672033868028907e-05,
      "loss": 2.4278650283813477,
      "memory(GiB)": 77.56,
      "step": 87500,
      "token_acc": 0.4948805460750853,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 3.7487682618568186,
      "eval_loss": 2.2121448516845703,
      "eval_runtime": 14.2801,
      "eval_samples_per_second": 7.003,
      "eval_steps_per_second": 7.003,
      "eval_token_acc": 0.49159663865546216,
      "step": 87500
    },
    {
      "epoch": 3.7489824771860674,
      "grad_norm": 5.867633819580078,
      "learning_rate": 1.4667271838464303e-05,
      "loss": 2.487856864929199,
      "memory(GiB)": 77.56,
      "step": 87505,
      "token_acc": 0.4824824824824825,
      "train_speed(iter/s)": 1.437726
    },
    {
      "epoch": 3.7491966925153166,
      "grad_norm": 5.005818843841553,
      "learning_rate": 1.4662510448985234e-05,
      "loss": 2.4117219924926756,
      "memory(GiB)": 77.56,
      "step": 87510,
      "token_acc": 0.44765342960288806,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.7494109078445654,
      "grad_norm": 7.557612419128418,
      "learning_rate": 1.4657749699677937e-05,
      "loss": 2.1620864868164062,
      "memory(GiB)": 77.56,
      "step": 87515,
      "token_acc": 0.46715328467153283,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 3.7496251231738142,
      "grad_norm": 5.773110866546631,
      "learning_rate": 1.4652989590628658e-05,
      "loss": 2.3477474212646485,
      "memory(GiB)": 77.56,
      "step": 87520,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 3.7498393385030635,
      "grad_norm": 6.7726006507873535,
      "learning_rate": 1.4648230121923629e-05,
      "loss": 2.340291404724121,
      "memory(GiB)": 77.56,
      "step": 87525,
      "token_acc": 0.5203252032520326,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 3.7500535538323123,
      "grad_norm": 5.539238929748535,
      "learning_rate": 1.4643471293649059e-05,
      "loss": 2.4184734344482424,
      "memory(GiB)": 77.56,
      "step": 87530,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.750267769161561,
      "grad_norm": 6.12601900100708,
      "learning_rate": 1.4638713105891188e-05,
      "loss": 2.4049610137939452,
      "memory(GiB)": 77.56,
      "step": 87535,
      "token_acc": 0.5032467532467533,
      "train_speed(iter/s)": 1.437781
    },
    {
      "epoch": 3.7504819844908104,
      "grad_norm": 5.433905601501465,
      "learning_rate": 1.4633955558736201e-05,
      "loss": 2.3850824356079103,
      "memory(GiB)": 77.56,
      "step": 87540,
      "token_acc": 0.47079037800687284,
      "train_speed(iter/s)": 1.437799
    },
    {
      "epoch": 3.750696199820059,
      "grad_norm": 8.146924018859863,
      "learning_rate": 1.4629198652270288e-05,
      "loss": 2.3787258148193358,
      "memory(GiB)": 77.56,
      "step": 87545,
      "token_acc": 0.4983388704318937,
      "train_speed(iter/s)": 1.437794
    },
    {
      "epoch": 3.750910415149308,
      "grad_norm": 5.818040370941162,
      "learning_rate": 1.4624442386579601e-05,
      "loss": 2.308137893676758,
      "memory(GiB)": 77.56,
      "step": 87550,
      "token_acc": 0.5331491712707183,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 3.7511246304785573,
      "grad_norm": 6.6491780281066895,
      "learning_rate": 1.4619686761750345e-05,
      "loss": 2.322003936767578,
      "memory(GiB)": 77.56,
      "step": 87555,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.751338845807806,
      "grad_norm": 7.675992965698242,
      "learning_rate": 1.4614931777868634e-05,
      "loss": 2.6120983123779298,
      "memory(GiB)": 77.56,
      "step": 87560,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.437825
    },
    {
      "epoch": 3.751553061137055,
      "grad_norm": 6.975713729858398,
      "learning_rate": 1.4610177435020645e-05,
      "loss": 2.3559783935546874,
      "memory(GiB)": 77.56,
      "step": 87565,
      "token_acc": 0.48132780082987553,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.751767276466304,
      "grad_norm": 4.986622333526611,
      "learning_rate": 1.4605423733292494e-05,
      "loss": 2.052254295349121,
      "memory(GiB)": 77.56,
      "step": 87570,
      "token_acc": 0.5254777070063694,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.751981491795553,
      "grad_norm": 5.038890838623047,
      "learning_rate": 1.460067067277029e-05,
      "loss": 2.4341768264770507,
      "memory(GiB)": 77.56,
      "step": 87575,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.7521957071248018,
      "grad_norm": 5.075716018676758,
      "learning_rate": 1.4595918253540147e-05,
      "loss": 2.3425437927246096,
      "memory(GiB)": 77.56,
      "step": 87580,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.437835
    },
    {
      "epoch": 3.752409922454051,
      "grad_norm": 6.498806953430176,
      "learning_rate": 1.459116647568815e-05,
      "loss": 2.683580207824707,
      "memory(GiB)": 77.56,
      "step": 87585,
      "token_acc": 0.4840989399293286,
      "train_speed(iter/s)": 1.437836
    },
    {
      "epoch": 3.7526241377833,
      "grad_norm": 5.05894136428833,
      "learning_rate": 1.458641533930038e-05,
      "loss": 2.3216506958007814,
      "memory(GiB)": 77.56,
      "step": 87590,
      "token_acc": 0.5089605734767025,
      "train_speed(iter/s)": 1.437835
    },
    {
      "epoch": 3.7528383531125487,
      "grad_norm": 4.817282676696777,
      "learning_rate": 1.4581664844462929e-05,
      "loss": 2.2896682739257814,
      "memory(GiB)": 77.56,
      "step": 87595,
      "token_acc": 0.5276073619631901,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.753052568441798,
      "grad_norm": 5.829711437225342,
      "learning_rate": 1.4576914991261848e-05,
      "loss": 2.559900093078613,
      "memory(GiB)": 77.56,
      "step": 87600,
      "token_acc": 0.5046439628482973,
      "train_speed(iter/s)": 1.437854
    },
    {
      "epoch": 3.7532667837710467,
      "grad_norm": 4.7632527351379395,
      "learning_rate": 1.4572165779783176e-05,
      "loss": 2.05426082611084,
      "memory(GiB)": 77.56,
      "step": 87605,
      "token_acc": 0.5518672199170125,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.7534809991002955,
      "grad_norm": 7.520813465118408,
      "learning_rate": 1.456741721011296e-05,
      "loss": 2.5381019592285154,
      "memory(GiB)": 77.56,
      "step": 87610,
      "token_acc": 0.432258064516129,
      "train_speed(iter/s)": 1.437828
    },
    {
      "epoch": 3.753695214429545,
      "grad_norm": 5.405460357666016,
      "learning_rate": 1.4562669282337198e-05,
      "loss": 2.052898406982422,
      "memory(GiB)": 77.56,
      "step": 87615,
      "token_acc": 0.5346153846153846,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 3.7539094297587936,
      "grad_norm": 5.803069114685059,
      "learning_rate": 1.4557921996541946e-05,
      "loss": 2.5014766693115233,
      "memory(GiB)": 77.56,
      "step": 87620,
      "token_acc": 0.45592705167173253,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 3.7541236450880424,
      "grad_norm": 5.596368312835693,
      "learning_rate": 1.455317535281317e-05,
      "loss": 2.3184619903564454,
      "memory(GiB)": 77.56,
      "step": 87625,
      "token_acc": 0.512280701754386,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.7543378604172917,
      "grad_norm": 6.604919910430908,
      "learning_rate": 1.4548429351236886e-05,
      "loss": 2.2990365982055665,
      "memory(GiB)": 77.56,
      "step": 87630,
      "token_acc": 0.525096525096525,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.7545520757465405,
      "grad_norm": 6.015281677246094,
      "learning_rate": 1.454368399189906e-05,
      "loss": 2.173585891723633,
      "memory(GiB)": 77.56,
      "step": 87635,
      "token_acc": 0.4948453608247423,
      "train_speed(iter/s)": 1.43789
    },
    {
      "epoch": 3.7547662910757893,
      "grad_norm": 6.053415775299072,
      "learning_rate": 1.4538939274885665e-05,
      "loss": 2.6174060821533205,
      "memory(GiB)": 77.56,
      "step": 87640,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.437882
    },
    {
      "epoch": 3.7549805064050386,
      "grad_norm": 6.9501051902771,
      "learning_rate": 1.4534195200282646e-05,
      "loss": 2.0804874420166017,
      "memory(GiB)": 77.56,
      "step": 87645,
      "token_acc": 0.5176991150442478,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.7551947217342874,
      "grad_norm": 5.198156356811523,
      "learning_rate": 1.4529451768175933e-05,
      "loss": 2.2638015747070312,
      "memory(GiB)": 77.56,
      "step": 87650,
      "token_acc": 0.5221238938053098,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 3.755408937063536,
      "grad_norm": 4.92119836807251,
      "learning_rate": 1.4524708978651491e-05,
      "loss": 2.3373477935791014,
      "memory(GiB)": 77.56,
      "step": 87655,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 3.7556231523927854,
      "grad_norm": 7.089959621429443,
      "learning_rate": 1.4519966831795228e-05,
      "loss": 2.260489845275879,
      "memory(GiB)": 77.56,
      "step": 87660,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 3.7558373677220342,
      "grad_norm": 5.744592189788818,
      "learning_rate": 1.4515225327693049e-05,
      "loss": 2.413991928100586,
      "memory(GiB)": 77.56,
      "step": 87665,
      "token_acc": 0.4691780821917808,
      "train_speed(iter/s)": 1.437908
    },
    {
      "epoch": 3.756051583051283,
      "grad_norm": 7.479886054992676,
      "learning_rate": 1.4510484466430846e-05,
      "loss": 2.0419031143188477,
      "memory(GiB)": 77.56,
      "step": 87670,
      "token_acc": 0.5051194539249146,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 3.7562657983805323,
      "grad_norm": 6.407444953918457,
      "learning_rate": 1.45057442480945e-05,
      "loss": 2.071546936035156,
      "memory(GiB)": 77.56,
      "step": 87675,
      "token_acc": 0.5359712230215827,
      "train_speed(iter/s)": 1.43791
    },
    {
      "epoch": 3.756480013709781,
      "grad_norm": 8.050786972045898,
      "learning_rate": 1.4501004672769903e-05,
      "loss": 2.2287151336669924,
      "memory(GiB)": 77.56,
      "step": 87680,
      "token_acc": 0.5112540192926045,
      "train_speed(iter/s)": 1.437922
    },
    {
      "epoch": 3.75669422903903,
      "grad_norm": 7.456993579864502,
      "learning_rate": 1.4496265740542908e-05,
      "loss": 2.198100280761719,
      "memory(GiB)": 77.56,
      "step": 87685,
      "token_acc": 0.5625,
      "train_speed(iter/s)": 1.437923
    },
    {
      "epoch": 3.756908444368279,
      "grad_norm": 5.820798873901367,
      "learning_rate": 1.4491527451499365e-05,
      "loss": 2.502303123474121,
      "memory(GiB)": 77.56,
      "step": 87690,
      "token_acc": 0.48698884758364314,
      "train_speed(iter/s)": 1.437923
    },
    {
      "epoch": 3.757122659697528,
      "grad_norm": 5.178523540496826,
      "learning_rate": 1.448678980572511e-05,
      "loss": 2.2209280014038084,
      "memory(GiB)": 77.56,
      "step": 87695,
      "token_acc": 0.5325077399380805,
      "train_speed(iter/s)": 1.43794
    },
    {
      "epoch": 3.757336875026777,
      "grad_norm": 6.158017158508301,
      "learning_rate": 1.4482052803305962e-05,
      "loss": 2.2875513076782226,
      "memory(GiB)": 77.56,
      "step": 87700,
      "token_acc": 0.5181159420289855,
      "train_speed(iter/s)": 1.437954
    },
    {
      "epoch": 3.757551090356026,
      "grad_norm": 10.388846397399902,
      "learning_rate": 1.4477316444327738e-05,
      "loss": 2.5663766860961914,
      "memory(GiB)": 77.56,
      "step": 87705,
      "token_acc": 0.4813664596273292,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 3.757765305685275,
      "grad_norm": 5.621857166290283,
      "learning_rate": 1.4472580728876272e-05,
      "loss": 2.128612518310547,
      "memory(GiB)": 77.56,
      "step": 87710,
      "token_acc": 0.5427509293680297,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 3.7579795210145237,
      "grad_norm": 5.60006046295166,
      "learning_rate": 1.4467845657037332e-05,
      "loss": 2.4085105895996093,
      "memory(GiB)": 77.56,
      "step": 87715,
      "token_acc": 0.46204620462046203,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.758193736343773,
      "grad_norm": 7.108492851257324,
      "learning_rate": 1.4463111228896697e-05,
      "loss": 2.2842512130737305,
      "memory(GiB)": 77.56,
      "step": 87720,
      "token_acc": 0.49416342412451364,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 3.7584079516730218,
      "grad_norm": 5.585237979888916,
      "learning_rate": 1.4458377444540139e-05,
      "loss": 2.5472055435180665,
      "memory(GiB)": 77.56,
      "step": 87725,
      "token_acc": 0.48589341692789967,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.7586221670022706,
      "grad_norm": 7.364243507385254,
      "learning_rate": 1.4453644304053415e-05,
      "loss": 2.332600402832031,
      "memory(GiB)": 77.56,
      "step": 87730,
      "token_acc": 0.5050847457627119,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 3.75883638233152,
      "grad_norm": 5.167562961578369,
      "learning_rate": 1.4448911807522253e-05,
      "loss": 2.5560848236083986,
      "memory(GiB)": 77.56,
      "step": 87735,
      "token_acc": 0.46875,
      "train_speed(iter/s)": 1.437993
    },
    {
      "epoch": 3.7590505976607687,
      "grad_norm": 6.318556308746338,
      "learning_rate": 1.4444179955032422e-05,
      "loss": 2.283896636962891,
      "memory(GiB)": 77.56,
      "step": 87740,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 3.7592648129900175,
      "grad_norm": 6.268924713134766,
      "learning_rate": 1.443944874666962e-05,
      "loss": 2.2712738037109377,
      "memory(GiB)": 77.56,
      "step": 87745,
      "token_acc": 0.48936170212765956,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 3.7594790283192667,
      "grad_norm": 4.683320045471191,
      "learning_rate": 1.443471818251957e-05,
      "loss": 2.562276077270508,
      "memory(GiB)": 77.56,
      "step": 87750,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 3.7596932436485155,
      "grad_norm": 6.328271865844727,
      "learning_rate": 1.4429988262667959e-05,
      "loss": 2.044296646118164,
      "memory(GiB)": 77.56,
      "step": 87755,
      "token_acc": 0.5521885521885522,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 3.7599074589777643,
      "grad_norm": 4.791431903839111,
      "learning_rate": 1.4425258987200463e-05,
      "loss": 2.4454938888549806,
      "memory(GiB)": 77.56,
      "step": 87760,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 3.7601216743070136,
      "grad_norm": 5.1604437828063965,
      "learning_rate": 1.4420530356202783e-05,
      "loss": 2.459474563598633,
      "memory(GiB)": 77.56,
      "step": 87765,
      "token_acc": 0.5032051282051282,
      "train_speed(iter/s)": 1.437973
    },
    {
      "epoch": 3.7603358896362624,
      "grad_norm": 5.239433765411377,
      "learning_rate": 1.4415802369760562e-05,
      "loss": 2.512373352050781,
      "memory(GiB)": 77.56,
      "step": 87770,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.7605501049655112,
      "grad_norm": 5.4469075202941895,
      "learning_rate": 1.4411075027959475e-05,
      "loss": 2.4335300445556642,
      "memory(GiB)": 77.56,
      "step": 87775,
      "token_acc": 0.4260355029585799,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 3.7607643202947605,
      "grad_norm": 5.976019859313965,
      "learning_rate": 1.440634833088515e-05,
      "loss": 2.391189384460449,
      "memory(GiB)": 77.56,
      "step": 87780,
      "token_acc": 0.5182724252491694,
      "train_speed(iter/s)": 1.438032
    },
    {
      "epoch": 3.7609785356240093,
      "grad_norm": 7.111571788787842,
      "learning_rate": 1.4401622278623217e-05,
      "loss": 2.307042694091797,
      "memory(GiB)": 77.56,
      "step": 87785,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.438038
    },
    {
      "epoch": 3.761192750953258,
      "grad_norm": 7.658970832824707,
      "learning_rate": 1.4396896871259286e-05,
      "loss": 2.1998964309692384,
      "memory(GiB)": 77.56,
      "step": 87790,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.438025
    },
    {
      "epoch": 3.7614069662825074,
      "grad_norm": 5.72540283203125,
      "learning_rate": 1.4392172108878954e-05,
      "loss": 2.263543891906738,
      "memory(GiB)": 77.56,
      "step": 87795,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.438016
    },
    {
      "epoch": 3.761621181611756,
      "grad_norm": 6.018642902374268,
      "learning_rate": 1.4387447991567838e-05,
      "loss": 2.223993110656738,
      "memory(GiB)": 77.56,
      "step": 87800,
      "token_acc": 0.49508196721311476,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 3.761835396941005,
      "grad_norm": 5.77709436416626,
      "learning_rate": 1.438272451941151e-05,
      "loss": 2.0517341613769533,
      "memory(GiB)": 77.56,
      "step": 87805,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 3.7620496122702543,
      "grad_norm": 7.062778472900391,
      "learning_rate": 1.4378001692495546e-05,
      "loss": 2.3862701416015626,
      "memory(GiB)": 77.56,
      "step": 87810,
      "token_acc": 0.5347432024169184,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 3.762263827599503,
      "grad_norm": 5.769493103027344,
      "learning_rate": 1.437327951090549e-05,
      "loss": 2.4958463668823243,
      "memory(GiB)": 77.56,
      "step": 87815,
      "token_acc": 0.46511627906976744,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 3.762478042928752,
      "grad_norm": 7.23477840423584,
      "learning_rate": 1.4368557974726882e-05,
      "loss": 2.6645326614379883,
      "memory(GiB)": 77.56,
      "step": 87820,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 3.762692258258001,
      "grad_norm": 6.248107433319092,
      "learning_rate": 1.436383708404529e-05,
      "loss": 2.606052017211914,
      "memory(GiB)": 77.56,
      "step": 87825,
      "token_acc": 0.45565749235474007,
      "train_speed(iter/s)": 1.438031
    },
    {
      "epoch": 3.76290647358725,
      "grad_norm": 4.963569641113281,
      "learning_rate": 1.4359116838946219e-05,
      "loss": 2.334998893737793,
      "memory(GiB)": 77.56,
      "step": 87830,
      "token_acc": 0.4965034965034965,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 3.7631206889164988,
      "grad_norm": 7.157711029052734,
      "learning_rate": 1.4354397239515177e-05,
      "loss": 2.630050468444824,
      "memory(GiB)": 77.56,
      "step": 87835,
      "token_acc": 0.45808383233532934,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 3.763334904245748,
      "grad_norm": 5.773626327514648,
      "learning_rate": 1.434967828583767e-05,
      "loss": 2.3394601821899412,
      "memory(GiB)": 77.56,
      "step": 87840,
      "token_acc": 0.5028735632183908,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 3.763549119574997,
      "grad_norm": 6.297607898712158,
      "learning_rate": 1.4344959977999162e-05,
      "loss": 2.4435955047607423,
      "memory(GiB)": 77.56,
      "step": 87845,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 3.7637633349042456,
      "grad_norm": 5.704437732696533,
      "learning_rate": 1.4340242316085156e-05,
      "loss": 2.295665168762207,
      "memory(GiB)": 77.56,
      "step": 87850,
      "token_acc": 0.5428571428571428,
      "train_speed(iter/s)": 1.438082
    },
    {
      "epoch": 3.763977550233495,
      "grad_norm": 4.977346897125244,
      "learning_rate": 1.433552530018113e-05,
      "loss": 2.2148517608642577,
      "memory(GiB)": 77.56,
      "step": 87855,
      "token_acc": 0.5225225225225225,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 3.7641917655627437,
      "grad_norm": 5.197224140167236,
      "learning_rate": 1.4330808930372514e-05,
      "loss": 2.2645254135131836,
      "memory(GiB)": 77.56,
      "step": 87860,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.438095
    },
    {
      "epoch": 3.7644059808919925,
      "grad_norm": 7.71748161315918,
      "learning_rate": 1.4326093206744756e-05,
      "loss": 2.5463199615478516,
      "memory(GiB)": 77.56,
      "step": 87865,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.438092
    },
    {
      "epoch": 3.7646201962212418,
      "grad_norm": 8.733891487121582,
      "learning_rate": 1.432137812938329e-05,
      "loss": 2.3446420669555663,
      "memory(GiB)": 77.56,
      "step": 87870,
      "token_acc": 0.5114155251141552,
      "train_speed(iter/s)": 1.438111
    },
    {
      "epoch": 3.7648344115504906,
      "grad_norm": 5.362898826599121,
      "learning_rate": 1.4316663698373522e-05,
      "loss": 2.435167694091797,
      "memory(GiB)": 77.56,
      "step": 87875,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 3.7650486268797394,
      "grad_norm": 5.609925270080566,
      "learning_rate": 1.4311949913800848e-05,
      "loss": 2.3758773803710938,
      "memory(GiB)": 77.56,
      "step": 87880,
      "token_acc": 0.48125,
      "train_speed(iter/s)": 1.438113
    },
    {
      "epoch": 3.7652628422089887,
      "grad_norm": 7.326563358306885,
      "learning_rate": 1.4307236775750699e-05,
      "loss": 2.249185562133789,
      "memory(GiB)": 77.56,
      "step": 87885,
      "token_acc": 0.5126582278481012,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 3.7654770575382375,
      "grad_norm": 6.238636016845703,
      "learning_rate": 1.4302524284308438e-05,
      "loss": 2.2153274536132814,
      "memory(GiB)": 77.56,
      "step": 87890,
      "token_acc": 0.5344827586206896,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 3.7656912728674863,
      "grad_norm": 4.449050426483154,
      "learning_rate": 1.4297812439559439e-05,
      "loss": 2.5964075088500977,
      "memory(GiB)": 77.56,
      "step": 87895,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.438109
    },
    {
      "epoch": 3.7659054881967355,
      "grad_norm": 6.41837215423584,
      "learning_rate": 1.4293101241589051e-05,
      "loss": 2.1801387786865236,
      "memory(GiB)": 77.56,
      "step": 87900,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 3.7661197035259844,
      "grad_norm": 5.493099212646484,
      "learning_rate": 1.4288390690482622e-05,
      "loss": 2.3284868240356444,
      "memory(GiB)": 77.56,
      "step": 87905,
      "token_acc": 0.5111821086261981,
      "train_speed(iter/s)": 1.438085
    },
    {
      "epoch": 3.766333918855233,
      "grad_norm": 6.2421674728393555,
      "learning_rate": 1.428368078632551e-05,
      "loss": 2.0088367462158203,
      "memory(GiB)": 77.56,
      "step": 87910,
      "token_acc": 0.5704697986577181,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 3.7665481341844824,
      "grad_norm": 6.400420188903809,
      "learning_rate": 1.4278971529203023e-05,
      "loss": 2.339704895019531,
      "memory(GiB)": 77.56,
      "step": 87915,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 3.7667623495137312,
      "grad_norm": 5.7013115882873535,
      "learning_rate": 1.427426291920046e-05,
      "loss": 2.461294937133789,
      "memory(GiB)": 77.56,
      "step": 87920,
      "token_acc": 0.49491525423728816,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 3.76697656484298,
      "grad_norm": 5.306664943695068,
      "learning_rate": 1.4269554956403153e-05,
      "loss": 2.3485349655151366,
      "memory(GiB)": 77.56,
      "step": 87925,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.438071
    },
    {
      "epoch": 3.7671907801722293,
      "grad_norm": 6.84415340423584,
      "learning_rate": 1.4264847640896378e-05,
      "loss": 2.546147346496582,
      "memory(GiB)": 77.56,
      "step": 87930,
      "token_acc": 0.47761194029850745,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 3.767404995501478,
      "grad_norm": 6.763495445251465,
      "learning_rate": 1.4260140972765407e-05,
      "loss": 2.5340614318847656,
      "memory(GiB)": 77.56,
      "step": 87935,
      "token_acc": 0.46689895470383275,
      "train_speed(iter/s)": 1.438093
    },
    {
      "epoch": 3.767619210830727,
      "grad_norm": 5.658900260925293,
      "learning_rate": 1.4255434952095498e-05,
      "loss": 2.629825210571289,
      "memory(GiB)": 77.56,
      "step": 87940,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.438097
    },
    {
      "epoch": 3.767833426159976,
      "grad_norm": 4.1563801765441895,
      "learning_rate": 1.4250729578971927e-05,
      "loss": 2.2303306579589846,
      "memory(GiB)": 77.56,
      "step": 87945,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 3.768047641489225,
      "grad_norm": 8.799057960510254,
      "learning_rate": 1.4246024853479928e-05,
      "loss": 2.4237850189208983,
      "memory(GiB)": 77.56,
      "step": 87950,
      "token_acc": 0.4870967741935484,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 3.768261856818474,
      "grad_norm": 4.2870988845825195,
      "learning_rate": 1.4241320775704725e-05,
      "loss": 2.3504398345947264,
      "memory(GiB)": 77.56,
      "step": 87955,
      "token_acc": 0.4937888198757764,
      "train_speed(iter/s)": 1.438111
    },
    {
      "epoch": 3.768476072147723,
      "grad_norm": 7.406762599945068,
      "learning_rate": 1.4236617345731546e-05,
      "loss": 2.437298393249512,
      "memory(GiB)": 77.56,
      "step": 87960,
      "token_acc": 0.5194805194805194,
      "train_speed(iter/s)": 1.438132
    },
    {
      "epoch": 3.768690287476972,
      "grad_norm": 6.2104597091674805,
      "learning_rate": 1.4231914563645576e-05,
      "loss": 2.2796226501464845,
      "memory(GiB)": 77.56,
      "step": 87965,
      "token_acc": 0.5352697095435685,
      "train_speed(iter/s)": 1.438133
    },
    {
      "epoch": 3.7689045028062207,
      "grad_norm": 5.935835361480713,
      "learning_rate": 1.4227212429532038e-05,
      "loss": 2.283091926574707,
      "memory(GiB)": 77.56,
      "step": 87970,
      "token_acc": 0.5296167247386759,
      "train_speed(iter/s)": 1.438121
    },
    {
      "epoch": 3.76911871813547,
      "grad_norm": 5.743374347686768,
      "learning_rate": 1.4222510943476109e-05,
      "loss": 2.2220298767089846,
      "memory(GiB)": 77.56,
      "step": 87975,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.438121
    },
    {
      "epoch": 3.7693329334647188,
      "grad_norm": 5.496264934539795,
      "learning_rate": 1.4217810105562957e-05,
      "loss": 2.2633298873901366,
      "memory(GiB)": 77.56,
      "step": 87980,
      "token_acc": 0.5345911949685535,
      "train_speed(iter/s)": 1.43812
    },
    {
      "epoch": 3.7695471487939676,
      "grad_norm": 6.349199295043945,
      "learning_rate": 1.4213109915877736e-05,
      "loss": 2.056529426574707,
      "memory(GiB)": 77.56,
      "step": 87985,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 3.769761364123217,
      "grad_norm": 4.894119739532471,
      "learning_rate": 1.4208410374505587e-05,
      "loss": 2.4239370346069338,
      "memory(GiB)": 77.56,
      "step": 87990,
      "token_acc": 0.5160142348754448,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 3.7699755794524656,
      "grad_norm": 5.657999038696289,
      "learning_rate": 1.4203711481531662e-05,
      "loss": 2.5051578521728515,
      "memory(GiB)": 77.56,
      "step": 87995,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.438118
    },
    {
      "epoch": 3.7701897947817145,
      "grad_norm": 6.166049957275391,
      "learning_rate": 1.4199013237041093e-05,
      "loss": 2.047438049316406,
      "memory(GiB)": 77.56,
      "step": 88000,
      "token_acc": 0.5769230769230769,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 3.7701897947817145,
      "eval_loss": 2.046614646911621,
      "eval_runtime": 14.8782,
      "eval_samples_per_second": 6.721,
      "eval_steps_per_second": 6.721,
      "eval_token_acc": 0.4853146853146853,
      "step": 88000
    },
    {
      "epoch": 3.7704040101109637,
      "grad_norm": 6.574954032897949,
      "learning_rate": 1.4194315641118989e-05,
      "loss": 2.451774787902832,
      "memory(GiB)": 77.56,
      "step": 88005,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.7706182254402125,
      "grad_norm": 5.201181888580322,
      "learning_rate": 1.4189618693850444e-05,
      "loss": 2.38657169342041,
      "memory(GiB)": 77.56,
      "step": 88010,
      "token_acc": 0.5119047619047619,
      "train_speed(iter/s)": 1.437777
    },
    {
      "epoch": 3.7708324407694613,
      "grad_norm": 4.677194118499756,
      "learning_rate": 1.4184922395320543e-05,
      "loss": 2.497384452819824,
      "memory(GiB)": 77.56,
      "step": 88015,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.7710466560987106,
      "grad_norm": 7.468648910522461,
      "learning_rate": 1.4180226745614379e-05,
      "loss": 2.3117725372314455,
      "memory(GiB)": 77.56,
      "step": 88020,
      "token_acc": 0.5095541401273885,
      "train_speed(iter/s)": 1.437771
    },
    {
      "epoch": 3.7712608714279594,
      "grad_norm": 6.285245895385742,
      "learning_rate": 1.4175531744816989e-05,
      "loss": 2.35408992767334,
      "memory(GiB)": 77.56,
      "step": 88025,
      "token_acc": 0.493006993006993,
      "train_speed(iter/s)": 1.437794
    },
    {
      "epoch": 3.771475086757208,
      "grad_norm": 6.402484893798828,
      "learning_rate": 1.4170837393013465e-05,
      "loss": 2.3732528686523438,
      "memory(GiB)": 77.56,
      "step": 88030,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.437807
    },
    {
      "epoch": 3.7716893020864575,
      "grad_norm": 5.333975315093994,
      "learning_rate": 1.416614369028883e-05,
      "loss": 2.1019189834594725,
      "memory(GiB)": 77.56,
      "step": 88035,
      "token_acc": 0.5189873417721519,
      "train_speed(iter/s)": 1.437818
    },
    {
      "epoch": 3.7719035174157063,
      "grad_norm": 5.57673978805542,
      "learning_rate": 1.4161450636728119e-05,
      "loss": 2.4872100830078123,
      "memory(GiB)": 77.56,
      "step": 88040,
      "token_acc": 0.48405797101449277,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 3.772117732744955,
      "grad_norm": 5.9250168800354,
      "learning_rate": 1.4156758232416345e-05,
      "loss": 2.40474910736084,
      "memory(GiB)": 77.56,
      "step": 88045,
      "token_acc": 0.5087719298245614,
      "train_speed(iter/s)": 1.437815
    },
    {
      "epoch": 3.7723319480742044,
      "grad_norm": 5.628423690795898,
      "learning_rate": 1.4152066477438508e-05,
      "loss": 2.221597099304199,
      "memory(GiB)": 77.56,
      "step": 88050,
      "token_acc": 0.5175718849840255,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 3.772546163403453,
      "grad_norm": 9.546796798706055,
      "learning_rate": 1.4147375371879628e-05,
      "loss": 2.32208137512207,
      "memory(GiB)": 77.56,
      "step": 88055,
      "token_acc": 0.49814126394052044,
      "train_speed(iter/s)": 1.437816
    },
    {
      "epoch": 3.772760378732702,
      "grad_norm": 5.79128360748291,
      "learning_rate": 1.4142684915824678e-05,
      "loss": 2.3078880310058594,
      "memory(GiB)": 77.56,
      "step": 88060,
      "token_acc": 0.5148148148148148,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.7729745940619512,
      "grad_norm": 6.553867816925049,
      "learning_rate": 1.413799510935861e-05,
      "loss": 2.237506866455078,
      "memory(GiB)": 77.56,
      "step": 88065,
      "token_acc": 0.5245901639344263,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 3.7731888093912,
      "grad_norm": 7.1847662925720215,
      "learning_rate": 1.4133305952566416e-05,
      "loss": 2.5017158508300783,
      "memory(GiB)": 77.56,
      "step": 88070,
      "token_acc": 0.5145985401459854,
      "train_speed(iter/s)": 1.437852
    },
    {
      "epoch": 3.773403024720449,
      "grad_norm": 5.464165687561035,
      "learning_rate": 1.4128617445533037e-05,
      "loss": 2.3889543533325197,
      "memory(GiB)": 77.56,
      "step": 88075,
      "token_acc": 0.5503875968992248,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 3.773617240049698,
      "grad_norm": 5.415406227111816,
      "learning_rate": 1.4123929588343398e-05,
      "loss": 2.1404388427734373,
      "memory(GiB)": 77.56,
      "step": 88080,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 3.773831455378947,
      "grad_norm": 7.293788433074951,
      "learning_rate": 1.4119242381082415e-05,
      "loss": 2.301029014587402,
      "memory(GiB)": 77.56,
      "step": 88085,
      "token_acc": 0.5402298850574713,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.7740456707081957,
      "grad_norm": 7.515618324279785,
      "learning_rate": 1.411455582383503e-05,
      "loss": 2.2081398010253905,
      "memory(GiB)": 77.56,
      "step": 88090,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 3.774259886037445,
      "grad_norm": 6.695063591003418,
      "learning_rate": 1.4109869916686125e-05,
      "loss": 2.241544723510742,
      "memory(GiB)": 77.56,
      "step": 88095,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 3.774474101366694,
      "grad_norm": 4.579066753387451,
      "learning_rate": 1.41051846597206e-05,
      "loss": 2.43978214263916,
      "memory(GiB)": 77.56,
      "step": 88100,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.437888
    },
    {
      "epoch": 3.7746883166959426,
      "grad_norm": 10.251058578491211,
      "learning_rate": 1.4100500053023324e-05,
      "loss": 2.345491409301758,
      "memory(GiB)": 77.56,
      "step": 88105,
      "token_acc": 0.5241935483870968,
      "train_speed(iter/s)": 1.437908
    },
    {
      "epoch": 3.774902532025192,
      "grad_norm": 5.597171306610107,
      "learning_rate": 1.4095816096679155e-05,
      "loss": 2.4285993576049805,
      "memory(GiB)": 77.56,
      "step": 88110,
      "token_acc": 0.46607669616519176,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.7751167473544407,
      "grad_norm": 5.961618423461914,
      "learning_rate": 1.409113279077297e-05,
      "loss": 2.603607940673828,
      "memory(GiB)": 77.56,
      "step": 88115,
      "token_acc": 0.44086021505376344,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 3.7753309626836895,
      "grad_norm": 5.908853054046631,
      "learning_rate": 1.40864501353896e-05,
      "loss": 2.3162328720092775,
      "memory(GiB)": 77.56,
      "step": 88120,
      "token_acc": 0.46179401993355484,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.7755451780129388,
      "grad_norm": 5.879225254058838,
      "learning_rate": 1.4081768130613877e-05,
      "loss": 2.304482650756836,
      "memory(GiB)": 77.56,
      "step": 88125,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 3.7757593933421876,
      "grad_norm": 6.7951436042785645,
      "learning_rate": 1.4077086776530618e-05,
      "loss": 2.4009750366210936,
      "memory(GiB)": 77.56,
      "step": 88130,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.7759736086714364,
      "grad_norm": 5.3484086990356445,
      "learning_rate": 1.4072406073224608e-05,
      "loss": 2.2523502349853515,
      "memory(GiB)": 77.56,
      "step": 88135,
      "token_acc": 0.5014925373134328,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 3.7761878240006856,
      "grad_norm": 4.904800891876221,
      "learning_rate": 1.4067726020780675e-05,
      "loss": 2.3921941757202148,
      "memory(GiB)": 77.56,
      "step": 88140,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.437836
    },
    {
      "epoch": 3.7764020393299345,
      "grad_norm": 5.1627349853515625,
      "learning_rate": 1.4063046619283604e-05,
      "loss": 2.3879465103149413,
      "memory(GiB)": 77.56,
      "step": 88145,
      "token_acc": 0.4854368932038835,
      "train_speed(iter/s)": 1.437838
    },
    {
      "epoch": 3.7766162546591833,
      "grad_norm": 6.091917514801025,
      "learning_rate": 1.4058367868818156e-05,
      "loss": 2.4942670822143556,
      "memory(GiB)": 77.56,
      "step": 88150,
      "token_acc": 0.4634920634920635,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.7768304699884325,
      "grad_norm": 6.140024662017822,
      "learning_rate": 1.4053689769469085e-05,
      "loss": 2.0714902877807617,
      "memory(GiB)": 77.56,
      "step": 88155,
      "token_acc": 0.5246478873239436,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.7770446853176813,
      "grad_norm": 6.162842750549316,
      "learning_rate": 1.4049012321321147e-05,
      "loss": 2.60401611328125,
      "memory(GiB)": 77.56,
      "step": 88160,
      "token_acc": 0.479020979020979,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 3.77725890064693,
      "grad_norm": 6.035883903503418,
      "learning_rate": 1.4044335524459078e-05,
      "loss": 2.2452037811279295,
      "memory(GiB)": 77.56,
      "step": 88165,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.437876
    },
    {
      "epoch": 3.7774731159761794,
      "grad_norm": 5.5866193771362305,
      "learning_rate": 1.4039659378967584e-05,
      "loss": 2.4295591354370116,
      "memory(GiB)": 77.56,
      "step": 88170,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.777687331305428,
      "grad_norm": 7.191069602966309,
      "learning_rate": 1.4034983884931407e-05,
      "loss": 2.3771324157714844,
      "memory(GiB)": 77.56,
      "step": 88175,
      "token_acc": 0.5099601593625498,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.777901546634677,
      "grad_norm": 8.728489875793457,
      "learning_rate": 1.4030309042435236e-05,
      "loss": 2.326151466369629,
      "memory(GiB)": 77.56,
      "step": 88180,
      "token_acc": 0.506896551724138,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.7781157619639263,
      "grad_norm": 5.650343418121338,
      "learning_rate": 1.4025634851563762e-05,
      "loss": 2.416168975830078,
      "memory(GiB)": 77.56,
      "step": 88185,
      "token_acc": 0.4622093023255814,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 3.778329977293175,
      "grad_norm": 5.094841957092285,
      "learning_rate": 1.4020961312401653e-05,
      "loss": 2.3069131851196287,
      "memory(GiB)": 77.56,
      "step": 88190,
      "token_acc": 0.5106382978723404,
      "train_speed(iter/s)": 1.437893
    },
    {
      "epoch": 3.778544192622424,
      "grad_norm": 8.147745132446289,
      "learning_rate": 1.4016288425033574e-05,
      "loss": 2.1368404388427735,
      "memory(GiB)": 77.56,
      "step": 88195,
      "token_acc": 0.5117845117845118,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 3.778758407951673,
      "grad_norm": 8.104848861694336,
      "learning_rate": 1.4011616189544197e-05,
      "loss": 2.40368709564209,
      "memory(GiB)": 77.56,
      "step": 88200,
      "token_acc": 0.4521072796934866,
      "train_speed(iter/s)": 1.437892
    },
    {
      "epoch": 3.778972623280922,
      "grad_norm": 7.374303817749023,
      "learning_rate": 1.4006944606018147e-05,
      "loss": 2.216788101196289,
      "memory(GiB)": 77.56,
      "step": 88205,
      "token_acc": 0.45016077170418006,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 3.779186838610171,
      "grad_norm": 4.357128620147705,
      "learning_rate": 1.4002273674540067e-05,
      "loss": 2.1201887130737305,
      "memory(GiB)": 77.56,
      "step": 88210,
      "token_acc": 0.5645756457564576,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.77940105393942,
      "grad_norm": 5.161882400512695,
      "learning_rate": 1.3997603395194548e-05,
      "loss": 2.2386753082275392,
      "memory(GiB)": 77.56,
      "step": 88215,
      "token_acc": 0.5095057034220533,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.779615269268669,
      "grad_norm": 4.354639530181885,
      "learning_rate": 1.3992933768066235e-05,
      "loss": 2.0564577102661135,
      "memory(GiB)": 77.56,
      "step": 88220,
      "token_acc": 0.541095890410959,
      "train_speed(iter/s)": 1.43786
    },
    {
      "epoch": 3.7798294845979177,
      "grad_norm": 4.925902843475342,
      "learning_rate": 1.3988264793239702e-05,
      "loss": 2.4225868225097655,
      "memory(GiB)": 77.56,
      "step": 88225,
      "token_acc": 0.46745562130177515,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 3.780043699927167,
      "grad_norm": 5.911545276641846,
      "learning_rate": 1.398359647079952e-05,
      "loss": 2.2908609390258787,
      "memory(GiB)": 77.56,
      "step": 88230,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 3.7802579152564157,
      "grad_norm": 4.6984381675720215,
      "learning_rate": 1.3978928800830288e-05,
      "loss": 2.4360822677612304,
      "memory(GiB)": 77.56,
      "step": 88235,
      "token_acc": 0.5042492917847026,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.7804721305856646,
      "grad_norm": 5.815092086791992,
      "learning_rate": 1.3974261783416554e-05,
      "loss": 2.421206474304199,
      "memory(GiB)": 77.56,
      "step": 88240,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.437901
    },
    {
      "epoch": 3.780686345914914,
      "grad_norm": 5.670464038848877,
      "learning_rate": 1.3969595418642862e-05,
      "loss": 2.2015037536621094,
      "memory(GiB)": 77.56,
      "step": 88245,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.7809005612441626,
      "grad_norm": 5.267561912536621,
      "learning_rate": 1.396492970659375e-05,
      "loss": 2.372382164001465,
      "memory(GiB)": 77.56,
      "step": 88250,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 3.7811147765734114,
      "grad_norm": 6.126773357391357,
      "learning_rate": 1.3960264647353721e-05,
      "loss": 2.099675941467285,
      "memory(GiB)": 77.56,
      "step": 88255,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 3.7813289919026607,
      "grad_norm": 5.632686138153076,
      "learning_rate": 1.3955600241007322e-05,
      "loss": 2.4065338134765626,
      "memory(GiB)": 77.56,
      "step": 88260,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.7815432072319095,
      "grad_norm": 6.635392189025879,
      "learning_rate": 1.3950936487639039e-05,
      "loss": 2.5109167098999023,
      "memory(GiB)": 77.56,
      "step": 88265,
      "token_acc": 0.47297297297297297,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 3.7817574225611583,
      "grad_norm": 6.3631744384765625,
      "learning_rate": 1.3946273387333352e-05,
      "loss": 2.2108951568603517,
      "memory(GiB)": 77.56,
      "step": 88270,
      "token_acc": 0.5228215767634855,
      "train_speed(iter/s)": 1.437905
    },
    {
      "epoch": 3.7819716378904076,
      "grad_norm": 5.487986087799072,
      "learning_rate": 1.3941610940174748e-05,
      "loss": 2.2736156463623045,
      "memory(GiB)": 77.56,
      "step": 88275,
      "token_acc": 0.5179153094462541,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.7821858532196564,
      "grad_norm": 5.851731777191162,
      "learning_rate": 1.3936949146247675e-05,
      "loss": 2.303765869140625,
      "memory(GiB)": 77.56,
      "step": 88280,
      "token_acc": 0.5181159420289855,
      "train_speed(iter/s)": 1.437905
    },
    {
      "epoch": 3.782400068548905,
      "grad_norm": 5.260596752166748,
      "learning_rate": 1.3932288005636607e-05,
      "loss": 2.221438407897949,
      "memory(GiB)": 77.56,
      "step": 88285,
      "token_acc": 0.5355805243445693,
      "train_speed(iter/s)": 1.437912
    },
    {
      "epoch": 3.7826142838781545,
      "grad_norm": 8.172385215759277,
      "learning_rate": 1.3927627518425967e-05,
      "loss": 2.1053960800170897,
      "memory(GiB)": 77.56,
      "step": 88290,
      "token_acc": 0.5342960288808665,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 3.7828284992074033,
      "grad_norm": 6.352456569671631,
      "learning_rate": 1.3922967684700206e-05,
      "loss": 2.5014036178588865,
      "memory(GiB)": 77.56,
      "step": 88295,
      "token_acc": 0.5148148148148148,
      "train_speed(iter/s)": 1.437936
    },
    {
      "epoch": 3.783042714536652,
      "grad_norm": 6.75061559677124,
      "learning_rate": 1.3918308504543725e-05,
      "loss": 2.3760402679443358,
      "memory(GiB)": 77.56,
      "step": 88300,
      "token_acc": 0.5050167224080268,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 3.7832569298659013,
      "grad_norm": 7.905178546905518,
      "learning_rate": 1.3913649978040939e-05,
      "loss": 2.3905990600585936,
      "memory(GiB)": 77.56,
      "step": 88305,
      "token_acc": 0.4831081081081081,
      "train_speed(iter/s)": 1.437921
    },
    {
      "epoch": 3.78347114519515,
      "grad_norm": 8.97487735748291,
      "learning_rate": 1.390899210527623e-05,
      "loss": 2.5207069396972654,
      "memory(GiB)": 77.56,
      "step": 88310,
      "token_acc": 0.44656488549618323,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 3.783685360524399,
      "grad_norm": 4.976228713989258,
      "learning_rate": 1.3904334886333975e-05,
      "loss": 2.1794254302978517,
      "memory(GiB)": 77.56,
      "step": 88315,
      "token_acc": 0.5314465408805031,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 3.7838995758536482,
      "grad_norm": 5.2152204513549805,
      "learning_rate": 1.3899678321298565e-05,
      "loss": 2.1674293518066405,
      "memory(GiB)": 77.56,
      "step": 88320,
      "token_acc": 0.5873015873015873,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 3.784113791182897,
      "grad_norm": 5.7985382080078125,
      "learning_rate": 1.3895022410254348e-05,
      "loss": 2.329416275024414,
      "memory(GiB)": 77.56,
      "step": 88325,
      "token_acc": 0.5117845117845118,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 3.784328006512146,
      "grad_norm": 4.425474643707275,
      "learning_rate": 1.3890367153285672e-05,
      "loss": 2.517572021484375,
      "memory(GiB)": 77.56,
      "step": 88330,
      "token_acc": 0.5168539325842697,
      "train_speed(iter/s)": 1.437966
    },
    {
      "epoch": 3.784542221841395,
      "grad_norm": 7.013037204742432,
      "learning_rate": 1.3885712550476864e-05,
      "loss": 2.269378662109375,
      "memory(GiB)": 77.56,
      "step": 88335,
      "token_acc": 0.5147540983606558,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 3.784756437170644,
      "grad_norm": 4.694156169891357,
      "learning_rate": 1.388105860191224e-05,
      "loss": 2.3789909362792967,
      "memory(GiB)": 77.56,
      "step": 88340,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 3.7849706524998927,
      "grad_norm": 9.157238006591797,
      "learning_rate": 1.3876405307676133e-05,
      "loss": 2.2227705001831053,
      "memory(GiB)": 77.56,
      "step": 88345,
      "token_acc": 0.5450980392156862,
      "train_speed(iter/s)": 1.437961
    },
    {
      "epoch": 3.785184867829142,
      "grad_norm": 5.300726890563965,
      "learning_rate": 1.3871752667852833e-05,
      "loss": 2.2756267547607423,
      "memory(GiB)": 77.56,
      "step": 88350,
      "token_acc": 0.47151898734177217,
      "train_speed(iter/s)": 1.437964
    },
    {
      "epoch": 3.785399083158391,
      "grad_norm": 7.77399206161499,
      "learning_rate": 1.3867100682526624e-05,
      "loss": 2.1836641311645506,
      "memory(GiB)": 77.56,
      "step": 88355,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.437932
    },
    {
      "epoch": 3.7856132984876396,
      "grad_norm": 6.013106346130371,
      "learning_rate": 1.386244935178178e-05,
      "loss": 2.4205345153808593,
      "memory(GiB)": 77.56,
      "step": 88360,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.437947
    },
    {
      "epoch": 3.785827513816889,
      "grad_norm": 6.620067119598389,
      "learning_rate": 1.3857798675702555e-05,
      "loss": 2.2716690063476563,
      "memory(GiB)": 77.56,
      "step": 88365,
      "token_acc": 0.5402298850574713,
      "train_speed(iter/s)": 1.43791
    },
    {
      "epoch": 3.7860417291461377,
      "grad_norm": 5.683618068695068,
      "learning_rate": 1.385314865437322e-05,
      "loss": 2.1879451751708983,
      "memory(GiB)": 77.56,
      "step": 88370,
      "token_acc": 0.5567765567765568,
      "train_speed(iter/s)": 1.437921
    },
    {
      "epoch": 3.7862559444753865,
      "grad_norm": 7.9503679275512695,
      "learning_rate": 1.3848499287877998e-05,
      "loss": 2.2505825042724608,
      "memory(GiB)": 77.56,
      "step": 88375,
      "token_acc": 0.47635135135135137,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.7864701598046357,
      "grad_norm": 8.003528594970703,
      "learning_rate": 1.3843850576301137e-05,
      "loss": 2.463016891479492,
      "memory(GiB)": 77.56,
      "step": 88380,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.437913
    },
    {
      "epoch": 3.7866843751338846,
      "grad_norm": 4.881189823150635,
      "learning_rate": 1.3839202519726835e-05,
      "loss": 2.2944822311401367,
      "memory(GiB)": 77.56,
      "step": 88385,
      "token_acc": 0.547945205479452,
      "train_speed(iter/s)": 1.437924
    },
    {
      "epoch": 3.7868985904631334,
      "grad_norm": 7.816769123077393,
      "learning_rate": 1.3834555118239306e-05,
      "loss": 2.4817783355712892,
      "memory(GiB)": 77.56,
      "step": 88390,
      "token_acc": 0.49480968858131485,
      "train_speed(iter/s)": 1.437934
    },
    {
      "epoch": 3.7871128057923826,
      "grad_norm": 6.9696221351623535,
      "learning_rate": 1.3829908371922734e-05,
      "loss": 2.532012176513672,
      "memory(GiB)": 77.56,
      "step": 88395,
      "token_acc": 0.5125786163522013,
      "train_speed(iter/s)": 1.437959
    },
    {
      "epoch": 3.7873270211216314,
      "grad_norm": 6.02717399597168,
      "learning_rate": 1.382526228086129e-05,
      "loss": 2.4970827102661133,
      "memory(GiB)": 77.56,
      "step": 88400,
      "token_acc": 0.4826388888888889,
      "train_speed(iter/s)": 1.437964
    },
    {
      "epoch": 3.7875412364508803,
      "grad_norm": 10.718117713928223,
      "learning_rate": 1.3820616845139168e-05,
      "loss": 2.0928085327148436,
      "memory(GiB)": 77.56,
      "step": 88405,
      "token_acc": 0.5300353356890459,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.7877554517801295,
      "grad_norm": 5.691068172454834,
      "learning_rate": 1.381597206484051e-05,
      "loss": 2.388553237915039,
      "memory(GiB)": 77.56,
      "step": 88410,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.7879696671093783,
      "grad_norm": 6.725213527679443,
      "learning_rate": 1.3811327940049462e-05,
      "loss": 2.387480926513672,
      "memory(GiB)": 77.56,
      "step": 88415,
      "token_acc": 0.5059760956175299,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 3.788183882438627,
      "grad_norm": 6.097580432891846,
      "learning_rate": 1.3806684470850156e-05,
      "loss": 2.1837238311767577,
      "memory(GiB)": 77.56,
      "step": 88420,
      "token_acc": 0.5320754716981132,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 3.7883980977678764,
      "grad_norm": 7.747653007507324,
      "learning_rate": 1.3802041657326698e-05,
      "loss": 2.3544782638549804,
      "memory(GiB)": 77.56,
      "step": 88425,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.437997
    },
    {
      "epoch": 3.788612313097125,
      "grad_norm": 6.273040294647217,
      "learning_rate": 1.3797399499563229e-05,
      "loss": 2.4027923583984374,
      "memory(GiB)": 77.56,
      "step": 88430,
      "token_acc": 0.4602649006622517,
      "train_speed(iter/s)": 1.438004
    },
    {
      "epoch": 3.788826528426374,
      "grad_norm": 5.949589729309082,
      "learning_rate": 1.3792757997643806e-05,
      "loss": 2.433192825317383,
      "memory(GiB)": 77.56,
      "step": 88435,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 3.7890407437556233,
      "grad_norm": 5.2899861335754395,
      "learning_rate": 1.3788117151652563e-05,
      "loss": 2.3401731491088866,
      "memory(GiB)": 77.56,
      "step": 88440,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.438032
    },
    {
      "epoch": 3.789254959084872,
      "grad_norm": 6.018672943115234,
      "learning_rate": 1.3783476961673541e-05,
      "loss": 2.3252368927001954,
      "memory(GiB)": 77.56,
      "step": 88445,
      "token_acc": 0.49544072948328266,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 3.789469174414121,
      "grad_norm": 6.2637858390808105,
      "learning_rate": 1.3778837427790809e-05,
      "loss": 2.417571258544922,
      "memory(GiB)": 77.56,
      "step": 88450,
      "token_acc": 0.4671280276816609,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 3.78968338974337,
      "grad_norm": 6.149107456207275,
      "learning_rate": 1.3774198550088413e-05,
      "loss": 2.4179168701171876,
      "memory(GiB)": 77.56,
      "step": 88455,
      "token_acc": 0.5160256410256411,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 3.789897605072619,
      "grad_norm": 7.0848894119262695,
      "learning_rate": 1.3769560328650383e-05,
      "loss": 2.5141664505004884,
      "memory(GiB)": 77.56,
      "step": 88460,
      "token_acc": 0.4911660777385159,
      "train_speed(iter/s)": 1.438042
    },
    {
      "epoch": 3.790111820401868,
      "grad_norm": 5.412406921386719,
      "learning_rate": 1.3764922763560767e-05,
      "loss": 2.7039491653442385,
      "memory(GiB)": 77.56,
      "step": 88465,
      "token_acc": 0.46742209631728043,
      "train_speed(iter/s)": 1.438044
    },
    {
      "epoch": 3.790326035731117,
      "grad_norm": 8.049283027648926,
      "learning_rate": 1.3760285854903566e-05,
      "loss": 2.669767379760742,
      "memory(GiB)": 77.56,
      "step": 88470,
      "token_acc": 0.476038338658147,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 3.790540251060366,
      "grad_norm": 6.3365044593811035,
      "learning_rate": 1.3755649602762777e-05,
      "loss": 2.130634880065918,
      "memory(GiB)": 77.56,
      "step": 88475,
      "token_acc": 0.5358361774744027,
      "train_speed(iter/s)": 1.438081
    },
    {
      "epoch": 3.7907544663896147,
      "grad_norm": 5.440357208251953,
      "learning_rate": 1.3751014007222402e-05,
      "loss": 2.2859848022460936,
      "memory(GiB)": 77.56,
      "step": 88480,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438092
    },
    {
      "epoch": 3.790968681718864,
      "grad_norm": 4.9004998207092285,
      "learning_rate": 1.3746379068366394e-05,
      "loss": 2.5116214752197266,
      "memory(GiB)": 77.56,
      "step": 88485,
      "token_acc": 0.46419098143236076,
      "train_speed(iter/s)": 1.438115
    },
    {
      "epoch": 3.7911828970481127,
      "grad_norm": 6.595088005065918,
      "learning_rate": 1.374174478627876e-05,
      "loss": 2.4399547576904297,
      "memory(GiB)": 77.56,
      "step": 88490,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.438113
    },
    {
      "epoch": 3.7913971123773615,
      "grad_norm": 5.438348293304443,
      "learning_rate": 1.3737111161043426e-05,
      "loss": 2.25557861328125,
      "memory(GiB)": 77.56,
      "step": 88495,
      "token_acc": 0.5272108843537415,
      "train_speed(iter/s)": 1.438127
    },
    {
      "epoch": 3.791611327706611,
      "grad_norm": 5.495586395263672,
      "learning_rate": 1.3732478192744342e-05,
      "loss": 2.315243148803711,
      "memory(GiB)": 77.56,
      "step": 88500,
      "token_acc": 0.5622317596566524,
      "train_speed(iter/s)": 1.438139
    },
    {
      "epoch": 3.791611327706611,
      "eval_loss": 2.179872512817383,
      "eval_runtime": 14.6266,
      "eval_samples_per_second": 6.837,
      "eval_steps_per_second": 6.837,
      "eval_token_acc": 0.4811443433029909,
      "step": 88500
    },
    {
      "epoch": 3.7918255430358596,
      "grad_norm": 5.515460014343262,
      "learning_rate": 1.3727845881465434e-05,
      "loss": 2.4595081329345705,
      "memory(GiB)": 77.56,
      "step": 88505,
      "token_acc": 0.48402255639097747,
      "train_speed(iter/s)": 1.437765
    },
    {
      "epoch": 3.7920397583651084,
      "grad_norm": 6.716407299041748,
      "learning_rate": 1.3723214227290615e-05,
      "loss": 2.259731101989746,
      "memory(GiB)": 77.56,
      "step": 88510,
      "token_acc": 0.5371024734982333,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 3.7922539736943577,
      "grad_norm": 4.956208229064941,
      "learning_rate": 1.3718583230303794e-05,
      "loss": 1.9834444046020507,
      "memory(GiB)": 77.56,
      "step": 88515,
      "token_acc": 0.5389221556886228,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 3.7924681890236065,
      "grad_norm": 5.995739936828613,
      "learning_rate": 1.3713952890588894e-05,
      "loss": 2.6465744018554687,
      "memory(GiB)": 77.56,
      "step": 88520,
      "token_acc": 0.46846846846846846,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 3.7926824043528553,
      "grad_norm": 5.004510879516602,
      "learning_rate": 1.370932320822977e-05,
      "loss": 2.5779163360595705,
      "memory(GiB)": 77.56,
      "step": 88525,
      "token_acc": 0.4966216216216216,
      "train_speed(iter/s)": 1.437748
    },
    {
      "epoch": 3.7928966196821046,
      "grad_norm": 5.810140132904053,
      "learning_rate": 1.3704694183310301e-05,
      "loss": 2.3363718032836913,
      "memory(GiB)": 77.56,
      "step": 88530,
      "token_acc": 0.5487804878048781,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 3.7931108350113534,
      "grad_norm": 4.239035606384277,
      "learning_rate": 1.3700065815914348e-05,
      "loss": 2.233226776123047,
      "memory(GiB)": 77.56,
      "step": 88535,
      "token_acc": 0.5046153846153846,
      "train_speed(iter/s)": 1.437796
    },
    {
      "epoch": 3.793325050340602,
      "grad_norm": 6.235622406005859,
      "learning_rate": 1.3695438106125757e-05,
      "loss": 2.3486305236816407,
      "memory(GiB)": 77.56,
      "step": 88540,
      "token_acc": 0.5176056338028169,
      "train_speed(iter/s)": 1.437793
    },
    {
      "epoch": 3.7935392656698514,
      "grad_norm": 5.882605075836182,
      "learning_rate": 1.369081105402834e-05,
      "loss": 2.4421903610229494,
      "memory(GiB)": 77.56,
      "step": 88545,
      "token_acc": 0.5093167701863354,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 3.7937534809991003,
      "grad_norm": 6.603424072265625,
      "learning_rate": 1.3686184659705959e-05,
      "loss": 2.723403549194336,
      "memory(GiB)": 77.56,
      "step": 88550,
      "token_acc": 0.4377224199288256,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 3.793967696328349,
      "grad_norm": 6.497887134552002,
      "learning_rate": 1.368155892324241e-05,
      "loss": 2.0882404327392576,
      "memory(GiB)": 77.56,
      "step": 88555,
      "token_acc": 0.5849056603773585,
      "train_speed(iter/s)": 1.437823
    },
    {
      "epoch": 3.7941819116575983,
      "grad_norm": 5.324814796447754,
      "learning_rate": 1.3676933844721484e-05,
      "loss": 2.221200942993164,
      "memory(GiB)": 77.56,
      "step": 88560,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.794396126986847,
      "grad_norm": 6.296723365783691,
      "learning_rate": 1.3672309424226981e-05,
      "loss": 2.426207733154297,
      "memory(GiB)": 77.56,
      "step": 88565,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 3.794610342316096,
      "grad_norm": 5.804182529449463,
      "learning_rate": 1.3667685661842655e-05,
      "loss": 2.2590354919433593,
      "memory(GiB)": 77.56,
      "step": 88570,
      "token_acc": 0.5131578947368421,
      "train_speed(iter/s)": 1.437823
    },
    {
      "epoch": 3.794824557645345,
      "grad_norm": 5.601439476013184,
      "learning_rate": 1.36630625576523e-05,
      "loss": 2.4765195846557617,
      "memory(GiB)": 77.56,
      "step": 88575,
      "token_acc": 0.49085365853658536,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.795038772974594,
      "grad_norm": 6.061161518096924,
      "learning_rate": 1.3658440111739657e-05,
      "loss": 2.0731300354003905,
      "memory(GiB)": 77.56,
      "step": 88580,
      "token_acc": 0.5270758122743683,
      "train_speed(iter/s)": 1.437826
    },
    {
      "epoch": 3.795252988303843,
      "grad_norm": 6.686573505401611,
      "learning_rate": 1.3653818324188444e-05,
      "loss": 2.359030532836914,
      "memory(GiB)": 77.56,
      "step": 88585,
      "token_acc": 0.4840764331210191,
      "train_speed(iter/s)": 1.437836
    },
    {
      "epoch": 3.795467203633092,
      "grad_norm": 6.2744364738464355,
      "learning_rate": 1.364919719508242e-05,
      "loss": 2.588063430786133,
      "memory(GiB)": 77.56,
      "step": 88590,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.437836
    },
    {
      "epoch": 3.795681418962341,
      "grad_norm": 7.436146259307861,
      "learning_rate": 1.3644576724505292e-05,
      "loss": 2.6732555389404298,
      "memory(GiB)": 77.56,
      "step": 88595,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.7958956342915897,
      "grad_norm": 5.4840288162231445,
      "learning_rate": 1.3639956912540764e-05,
      "loss": 2.4047895431518556,
      "memory(GiB)": 77.56,
      "step": 88600,
      "token_acc": 0.5222222222222223,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.796109849620839,
      "grad_norm": 5.893927574157715,
      "learning_rate": 1.3635337759272503e-05,
      "loss": 2.2417497634887695,
      "memory(GiB)": 77.56,
      "step": 88605,
      "token_acc": 0.5235109717868338,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.796324064950088,
      "grad_norm": 5.517001628875732,
      "learning_rate": 1.363071926478423e-05,
      "loss": 2.401530647277832,
      "memory(GiB)": 77.56,
      "step": 88610,
      "token_acc": 0.471976401179941,
      "train_speed(iter/s)": 1.437857
    },
    {
      "epoch": 3.7965382802793366,
      "grad_norm": 6.061408519744873,
      "learning_rate": 1.3626101429159593e-05,
      "loss": 2.487696075439453,
      "memory(GiB)": 77.56,
      "step": 88615,
      "token_acc": 0.46619217081850534,
      "train_speed(iter/s)": 1.437872
    },
    {
      "epoch": 3.796752495608586,
      "grad_norm": 6.266184329986572,
      "learning_rate": 1.3621484252482252e-05,
      "loss": 2.2935035705566404,
      "memory(GiB)": 77.56,
      "step": 88620,
      "token_acc": 0.5343511450381679,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.7969667109378347,
      "grad_norm": 5.447831630706787,
      "learning_rate": 1.3616867734835854e-05,
      "loss": 2.3129610061645507,
      "memory(GiB)": 77.56,
      "step": 88625,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.437891
    },
    {
      "epoch": 3.7971809262670835,
      "grad_norm": 5.633322238922119,
      "learning_rate": 1.3612251876304005e-05,
      "loss": 2.3474416732788086,
      "memory(GiB)": 77.56,
      "step": 88630,
      "token_acc": 0.5258358662613982,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 3.7973951415963327,
      "grad_norm": 6.463729381561279,
      "learning_rate": 1.3607636676970365e-05,
      "loss": 2.555350494384766,
      "memory(GiB)": 77.56,
      "step": 88635,
      "token_acc": 0.46688741721854304,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 3.7976093569255815,
      "grad_norm": 7.855784893035889,
      "learning_rate": 1.3603022136918526e-05,
      "loss": 2.4156044006347654,
      "memory(GiB)": 77.56,
      "step": 88640,
      "token_acc": 0.4944649446494465,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 3.7978235722548304,
      "grad_norm": 4.906198024749756,
      "learning_rate": 1.3598408256232082e-05,
      "loss": 2.0094844818115236,
      "memory(GiB)": 77.56,
      "step": 88645,
      "token_acc": 0.554006968641115,
      "train_speed(iter/s)": 1.43789
    },
    {
      "epoch": 3.7980377875840796,
      "grad_norm": 5.530091285705566,
      "learning_rate": 1.3593795034994622e-05,
      "loss": 2.4955734252929687,
      "memory(GiB)": 77.56,
      "step": 88650,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.7982520029133284,
      "grad_norm": 5.036495685577393,
      "learning_rate": 1.3589182473289703e-05,
      "loss": 2.515034866333008,
      "memory(GiB)": 77.56,
      "step": 88655,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 3.7984662182425772,
      "grad_norm": 6.222428798675537,
      "learning_rate": 1.3584570571200895e-05,
      "loss": 2.318907928466797,
      "memory(GiB)": 77.56,
      "step": 88660,
      "token_acc": 0.48348348348348347,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 3.7986804335718265,
      "grad_norm": 5.604467868804932,
      "learning_rate": 1.357995932881177e-05,
      "loss": 2.1846290588378907,
      "memory(GiB)": 77.56,
      "step": 88665,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.437912
    },
    {
      "epoch": 3.7988946489010753,
      "grad_norm": 5.464486122131348,
      "learning_rate": 1.3575348746205847e-05,
      "loss": 2.1534345626831053,
      "memory(GiB)": 77.56,
      "step": 88670,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.437913
    },
    {
      "epoch": 3.799108864230324,
      "grad_norm": 5.152063846588135,
      "learning_rate": 1.3570738823466645e-05,
      "loss": 2.215536117553711,
      "memory(GiB)": 77.56,
      "step": 88675,
      "token_acc": 0.556910569105691,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 3.7993230795595734,
      "grad_norm": 6.006877899169922,
      "learning_rate": 1.3566129560677687e-05,
      "loss": 2.5528993606567383,
      "memory(GiB)": 77.56,
      "step": 88680,
      "token_acc": 0.46355685131195334,
      "train_speed(iter/s)": 1.437908
    },
    {
      "epoch": 3.799537294888822,
      "grad_norm": 5.099438667297363,
      "learning_rate": 1.3561520957922469e-05,
      "loss": 2.391332244873047,
      "memory(GiB)": 77.56,
      "step": 88685,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 3.799751510218071,
      "grad_norm": 6.159640312194824,
      "learning_rate": 1.3556913015284467e-05,
      "loss": 2.172776222229004,
      "memory(GiB)": 77.56,
      "step": 88690,
      "token_acc": 0.4931506849315068,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 3.7999657255473203,
      "grad_norm": 4.8566718101501465,
      "learning_rate": 1.3552305732847186e-05,
      "loss": 2.5580873489379883,
      "memory(GiB)": 77.56,
      "step": 88695,
      "token_acc": 0.4956268221574344,
      "train_speed(iter/s)": 1.437955
    },
    {
      "epoch": 3.800179940876569,
      "grad_norm": 9.976699829101562,
      "learning_rate": 1.3547699110694079e-05,
      "loss": 2.597652626037598,
      "memory(GiB)": 77.56,
      "step": 88700,
      "token_acc": 0.4948805460750853,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 3.800394156205818,
      "grad_norm": 5.338417053222656,
      "learning_rate": 1.3543093148908597e-05,
      "loss": 2.345846176147461,
      "memory(GiB)": 77.56,
      "step": 88705,
      "token_acc": 0.5101351351351351,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 3.800608371535067,
      "grad_norm": 8.344489097595215,
      "learning_rate": 1.3538487847574183e-05,
      "loss": 2.6195497512817383,
      "memory(GiB)": 77.56,
      "step": 88710,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 3.800822586864316,
      "grad_norm": 6.010809421539307,
      "learning_rate": 1.353388320677425e-05,
      "loss": 2.454133987426758,
      "memory(GiB)": 77.56,
      "step": 88715,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.8010368021935648,
      "grad_norm": 5.5802226066589355,
      "learning_rate": 1.3529279226592245e-05,
      "loss": 2.148936653137207,
      "memory(GiB)": 77.56,
      "step": 88720,
      "token_acc": 0.5433962264150943,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.801251017522814,
      "grad_norm": 5.8756208419799805,
      "learning_rate": 1.3524675907111555e-05,
      "loss": 2.5548810958862305,
      "memory(GiB)": 77.56,
      "step": 88725,
      "token_acc": 0.4807121661721068,
      "train_speed(iter/s)": 1.438007
    },
    {
      "epoch": 3.801465232852063,
      "grad_norm": 6.6932172775268555,
      "learning_rate": 1.3520073248415565e-05,
      "loss": 2.5040332794189455,
      "memory(GiB)": 77.56,
      "step": 88730,
      "token_acc": 0.4564459930313589,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.8016794481813116,
      "grad_norm": 6.522132873535156,
      "learning_rate": 1.3515471250587686e-05,
      "loss": 2.134455108642578,
      "memory(GiB)": 77.56,
      "step": 88735,
      "token_acc": 0.5510204081632653,
      "train_speed(iter/s)": 1.437982
    },
    {
      "epoch": 3.801893663510561,
      "grad_norm": 5.06507682800293,
      "learning_rate": 1.3510869913711272e-05,
      "loss": 2.4543106079101564,
      "memory(GiB)": 77.56,
      "step": 88740,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 3.8021078788398097,
      "grad_norm": 5.750798225402832,
      "learning_rate": 1.3506269237869679e-05,
      "loss": 2.3681476593017576,
      "memory(GiB)": 77.56,
      "step": 88745,
      "token_acc": 0.5127118644067796,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.8023220941690585,
      "grad_norm": 6.647711277008057,
      "learning_rate": 1.3501669223146236e-05,
      "loss": 2.441213035583496,
      "memory(GiB)": 77.56,
      "step": 88750,
      "token_acc": 0.5298013245033113,
      "train_speed(iter/s)": 1.438001
    },
    {
      "epoch": 3.802536309498308,
      "grad_norm": 6.817627429962158,
      "learning_rate": 1.3497069869624307e-05,
      "loss": 2.697922134399414,
      "memory(GiB)": 77.56,
      "step": 88755,
      "token_acc": 0.45689655172413796,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 3.8027505248275566,
      "grad_norm": 6.088765621185303,
      "learning_rate": 1.34924711773872e-05,
      "loss": 2.2126121520996094,
      "memory(GiB)": 77.56,
      "step": 88760,
      "token_acc": 0.5101351351351351,
      "train_speed(iter/s)": 1.438019
    },
    {
      "epoch": 3.8029647401568054,
      "grad_norm": 6.300668716430664,
      "learning_rate": 1.348787314651822e-05,
      "loss": 2.510703468322754,
      "memory(GiB)": 77.56,
      "step": 88765,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 3.8031789554860547,
      "grad_norm": 6.420996189117432,
      "learning_rate": 1.3483275777100679e-05,
      "loss": 2.6871437072753905,
      "memory(GiB)": 77.56,
      "step": 88770,
      "token_acc": 0.4365079365079365,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 3.8033931708153035,
      "grad_norm": 4.696811676025391,
      "learning_rate": 1.3478679069217831e-05,
      "loss": 2.2402273178100587,
      "memory(GiB)": 77.56,
      "step": 88775,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 3.8036073861445523,
      "grad_norm": 5.776646614074707,
      "learning_rate": 1.3474083022952987e-05,
      "loss": 1.9888893127441407,
      "memory(GiB)": 77.56,
      "step": 88780,
      "token_acc": 0.5607843137254902,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 3.8038216014738016,
      "grad_norm": 4.437472820281982,
      "learning_rate": 1.3469487638389394e-05,
      "loss": 2.373304557800293,
      "memory(GiB)": 77.56,
      "step": 88785,
      "token_acc": 0.4756944444444444,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 3.8040358168030504,
      "grad_norm": 6.797004222869873,
      "learning_rate": 1.3464892915610305e-05,
      "loss": 2.1475845336914063,
      "memory(GiB)": 77.56,
      "step": 88790,
      "token_acc": 0.49812734082397003,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 3.804250032132299,
      "grad_norm": 5.936126232147217,
      "learning_rate": 1.346029885469895e-05,
      "loss": 2.401422882080078,
      "memory(GiB)": 77.56,
      "step": 88795,
      "token_acc": 0.45821325648414984,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 3.8044642474615484,
      "grad_norm": 6.225060939788818,
      "learning_rate": 1.3455705455738543e-05,
      "loss": 2.261306381225586,
      "memory(GiB)": 77.56,
      "step": 88800,
      "token_acc": 0.4778481012658228,
      "train_speed(iter/s)": 1.438081
    },
    {
      "epoch": 3.8046784627907972,
      "grad_norm": 5.280925273895264,
      "learning_rate": 1.3451112718812308e-05,
      "loss": 2.239191436767578,
      "memory(GiB)": 77.56,
      "step": 88805,
      "token_acc": 0.5064102564102564,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 3.804892678120046,
      "grad_norm": 5.71987247467041,
      "learning_rate": 1.3446520644003474e-05,
      "loss": 2.279434585571289,
      "memory(GiB)": 77.56,
      "step": 88810,
      "token_acc": 0.5105105105105106,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 3.8051068934492953,
      "grad_norm": 5.1327805519104,
      "learning_rate": 1.3441929231395207e-05,
      "loss": 2.4025634765625,
      "memory(GiB)": 77.56,
      "step": 88815,
      "token_acc": 0.5124223602484472,
      "train_speed(iter/s)": 1.43807
    },
    {
      "epoch": 3.805321108778544,
      "grad_norm": 6.3011016845703125,
      "learning_rate": 1.3437338481070689e-05,
      "loss": 2.3850595474243166,
      "memory(GiB)": 77.56,
      "step": 88820,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.438074
    },
    {
      "epoch": 3.805535324107793,
      "grad_norm": 5.278100490570068,
      "learning_rate": 1.3432748393113075e-05,
      "loss": 2.1128021240234376,
      "memory(GiB)": 77.56,
      "step": 88825,
      "token_acc": 0.5419354838709678,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 3.805749539437042,
      "grad_norm": 5.199338912963867,
      "learning_rate": 1.342815896760553e-05,
      "loss": 2.3207975387573243,
      "memory(GiB)": 77.56,
      "step": 88830,
      "token_acc": 0.494949494949495,
      "train_speed(iter/s)": 1.438074
    },
    {
      "epoch": 3.805963754766291,
      "grad_norm": 6.1854729652404785,
      "learning_rate": 1.342357020463118e-05,
      "loss": 2.428904724121094,
      "memory(GiB)": 77.56,
      "step": 88835,
      "token_acc": 0.49615384615384617,
      "train_speed(iter/s)": 1.438088
    },
    {
      "epoch": 3.80617797009554,
      "grad_norm": 5.287379741668701,
      "learning_rate": 1.3418982104273182e-05,
      "loss": 2.2304576873779296,
      "memory(GiB)": 77.56,
      "step": 88840,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.438052
    },
    {
      "epoch": 3.806392185424789,
      "grad_norm": 6.720423698425293,
      "learning_rate": 1.3414394666614638e-05,
      "loss": 2.480998229980469,
      "memory(GiB)": 77.56,
      "step": 88845,
      "token_acc": 0.4606741573033708,
      "train_speed(iter/s)": 1.438055
    },
    {
      "epoch": 3.806606400754038,
      "grad_norm": 7.8951592445373535,
      "learning_rate": 1.3409807891738652e-05,
      "loss": 2.660314178466797,
      "memory(GiB)": 77.56,
      "step": 88850,
      "token_acc": 0.45987654320987653,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 3.8068206160832867,
      "grad_norm": 5.180807590484619,
      "learning_rate": 1.3405221779728327e-05,
      "loss": 2.4456287384033204,
      "memory(GiB)": 77.56,
      "step": 88855,
      "token_acc": 0.45588235294117646,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 3.807034831412536,
      "grad_norm": 6.398998260498047,
      "learning_rate": 1.3400636330666716e-05,
      "loss": 2.436860466003418,
      "memory(GiB)": 77.56,
      "step": 88860,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.438055
    },
    {
      "epoch": 3.8072490467417848,
      "grad_norm": 4.665395736694336,
      "learning_rate": 1.3396051544636928e-05,
      "loss": 2.452156639099121,
      "memory(GiB)": 77.56,
      "step": 88865,
      "token_acc": 0.4533333333333333,
      "train_speed(iter/s)": 1.438062
    },
    {
      "epoch": 3.8074632620710336,
      "grad_norm": 8.299912452697754,
      "learning_rate": 1.3391467421722004e-05,
      "loss": 2.2736478805541993,
      "memory(GiB)": 77.56,
      "step": 88870,
      "token_acc": 0.5089820359281437,
      "train_speed(iter/s)": 1.438085
    },
    {
      "epoch": 3.807677477400283,
      "grad_norm": 4.993086814880371,
      "learning_rate": 1.3386883962004992e-05,
      "loss": 1.9588802337646485,
      "memory(GiB)": 77.56,
      "step": 88875,
      "token_acc": 0.5851528384279476,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 3.8078916927295317,
      "grad_norm": 4.616026401519775,
      "learning_rate": 1.3382301165568905e-05,
      "loss": 2.0985668182373045,
      "memory(GiB)": 77.56,
      "step": 88880,
      "token_acc": 0.5548780487804879,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 3.8081059080587805,
      "grad_norm": 6.2056121826171875,
      "learning_rate": 1.33777190324968e-05,
      "loss": 2.611598587036133,
      "memory(GiB)": 77.56,
      "step": 88885,
      "token_acc": 0.45045045045045046,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 3.8083201233880297,
      "grad_norm": 5.870640754699707,
      "learning_rate": 1.3373137562871668e-05,
      "loss": 2.4061965942382812,
      "memory(GiB)": 77.56,
      "step": 88890,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 3.8085343387172785,
      "grad_norm": 5.236010551452637,
      "learning_rate": 1.336855675677649e-05,
      "loss": 2.472321701049805,
      "memory(GiB)": 77.56,
      "step": 88895,
      "token_acc": 0.4588607594936709,
      "train_speed(iter/s)": 1.438088
    },
    {
      "epoch": 3.8087485540465273,
      "grad_norm": 6.35364294052124,
      "learning_rate": 1.3363976614294288e-05,
      "loss": 2.48392333984375,
      "memory(GiB)": 77.56,
      "step": 88900,
      "token_acc": 0.47468354430379744,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 3.8089627693757766,
      "grad_norm": 4.6643781661987305,
      "learning_rate": 1.3359397135508017e-05,
      "loss": 2.2396944046020506,
      "memory(GiB)": 77.56,
      "step": 88905,
      "token_acc": 0.5016611295681063,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 3.8091769847050254,
      "grad_norm": 5.817845344543457,
      "learning_rate": 1.3354818320500634e-05,
      "loss": 2.1948976516723633,
      "memory(GiB)": 77.56,
      "step": 88910,
      "token_acc": 0.5444444444444444,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 3.8093912000342742,
      "grad_norm": 6.068362236022949,
      "learning_rate": 1.3350240169355099e-05,
      "loss": 2.5008708953857424,
      "memory(GiB)": 77.56,
      "step": 88915,
      "token_acc": 0.4828571428571429,
      "train_speed(iter/s)": 1.438052
    },
    {
      "epoch": 3.8096054153635235,
      "grad_norm": 5.653886795043945,
      "learning_rate": 1.3345662682154326e-05,
      "loss": 2.330107307434082,
      "memory(GiB)": 77.56,
      "step": 88920,
      "token_acc": 0.4967741935483871,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 3.8098196306927723,
      "grad_norm": 6.844162464141846,
      "learning_rate": 1.3341085858981273e-05,
      "loss": 2.1554405212402346,
      "memory(GiB)": 77.56,
      "step": 88925,
      "token_acc": 0.5158730158730159,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 3.810033846022021,
      "grad_norm": 5.717306137084961,
      "learning_rate": 1.3336509699918837e-05,
      "loss": 2.688567543029785,
      "memory(GiB)": 77.56,
      "step": 88930,
      "token_acc": 0.4588607594936709,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 3.8102480613512704,
      "grad_norm": 7.388886451721191,
      "learning_rate": 1.3331934205049924e-05,
      "loss": 2.192700004577637,
      "memory(GiB)": 77.56,
      "step": 88935,
      "token_acc": 0.5033112582781457,
      "train_speed(iter/s)": 1.4381
    },
    {
      "epoch": 3.810462276680519,
      "grad_norm": 4.905337333679199,
      "learning_rate": 1.3327359374457415e-05,
      "loss": 2.3151618957519533,
      "memory(GiB)": 77.56,
      "step": 88940,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.438115
    },
    {
      "epoch": 3.810676492009768,
      "grad_norm": 6.05534029006958,
      "learning_rate": 1.3322785208224181e-05,
      "loss": 2.3507307052612303,
      "memory(GiB)": 77.56,
      "step": 88945,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.438125
    },
    {
      "epoch": 3.8108907073390172,
      "grad_norm": 5.012078285217285,
      "learning_rate": 1.3318211706433114e-05,
      "loss": 2.1554950714111327,
      "memory(GiB)": 77.56,
      "step": 88950,
      "token_acc": 0.5415162454873647,
      "train_speed(iter/s)": 1.438131
    },
    {
      "epoch": 3.811104922668266,
      "grad_norm": 5.602942943572998,
      "learning_rate": 1.331363886916704e-05,
      "loss": 2.5066082000732424,
      "memory(GiB)": 77.56,
      "step": 88955,
      "token_acc": 0.49544072948328266,
      "train_speed(iter/s)": 1.438123
    },
    {
      "epoch": 3.811319137997515,
      "grad_norm": 6.411240100860596,
      "learning_rate": 1.3309066696508826e-05,
      "loss": 2.576996612548828,
      "memory(GiB)": 77.56,
      "step": 88960,
      "token_acc": 0.504950495049505,
      "train_speed(iter/s)": 1.438137
    },
    {
      "epoch": 3.811533353326764,
      "grad_norm": 4.697442054748535,
      "learning_rate": 1.3304495188541283e-05,
      "loss": 2.3949050903320312,
      "memory(GiB)": 77.56,
      "step": 88965,
      "token_acc": 0.5083798882681564,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 3.811747568656013,
      "grad_norm": 6.031418323516846,
      "learning_rate": 1.3299924345347237e-05,
      "loss": 2.269063949584961,
      "memory(GiB)": 77.56,
      "step": 88970,
      "token_acc": 0.5346534653465347,
      "train_speed(iter/s)": 1.4381
    },
    {
      "epoch": 3.8119617839852618,
      "grad_norm": 5.140743732452393,
      "learning_rate": 1.3295354167009489e-05,
      "loss": 2.3306386947631834,
      "memory(GiB)": 77.56,
      "step": 88975,
      "token_acc": 0.49707602339181284,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 3.812175999314511,
      "grad_norm": 6.081470012664795,
      "learning_rate": 1.3290784653610817e-05,
      "loss": 2.3037765502929686,
      "memory(GiB)": 77.56,
      "step": 88980,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 3.81239021464376,
      "grad_norm": 7.274415493011475,
      "learning_rate": 1.328621580523403e-05,
      "loss": 2.193290138244629,
      "memory(GiB)": 77.56,
      "step": 88985,
      "token_acc": 0.5348837209302325,
      "train_speed(iter/s)": 1.438109
    },
    {
      "epoch": 3.8126044299730086,
      "grad_norm": 6.503421306610107,
      "learning_rate": 1.3281647621961885e-05,
      "loss": 1.9087289810180663,
      "memory(GiB)": 77.56,
      "step": 88990,
      "token_acc": 0.5529953917050692,
      "train_speed(iter/s)": 1.438108
    },
    {
      "epoch": 3.812818645302258,
      "grad_norm": 5.625267028808594,
      "learning_rate": 1.3277080103877138e-05,
      "loss": 2.590570831298828,
      "memory(GiB)": 77.56,
      "step": 88995,
      "token_acc": 0.46788990825688076,
      "train_speed(iter/s)": 1.438119
    },
    {
      "epoch": 3.8130328606315067,
      "grad_norm": 6.017382621765137,
      "learning_rate": 1.3272513251062535e-05,
      "loss": 2.3766555786132812,
      "memory(GiB)": 77.56,
      "step": 89000,
      "token_acc": 0.498371335504886,
      "train_speed(iter/s)": 1.438116
    },
    {
      "epoch": 3.8130328606315067,
      "eval_loss": 2.049919843673706,
      "eval_runtime": 13.7157,
      "eval_samples_per_second": 7.291,
      "eval_steps_per_second": 7.291,
      "eval_token_acc": 0.4951590594744122,
      "step": 89000
    },
    {
      "epoch": 3.8132470759607555,
      "grad_norm": 4.624563217163086,
      "learning_rate": 1.3267947063600794e-05,
      "loss": 2.003373146057129,
      "memory(GiB)": 77.56,
      "step": 89005,
      "token_acc": 0.517,
      "train_speed(iter/s)": 1.437786
    },
    {
      "epoch": 3.8134612912900048,
      "grad_norm": 5.108028888702393,
      "learning_rate": 1.3263381541574659e-05,
      "loss": 2.4233119964599608,
      "memory(GiB)": 77.56,
      "step": 89010,
      "token_acc": 0.5030120481927711,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 3.8136755066192536,
      "grad_norm": 5.897947311401367,
      "learning_rate": 1.3258816685066832e-05,
      "loss": 2.368221664428711,
      "memory(GiB)": 77.56,
      "step": 89015,
      "token_acc": 0.4956521739130435,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 3.8138897219485024,
      "grad_norm": 5.4629011154174805,
      "learning_rate": 1.3254252494160007e-05,
      "loss": 2.1324527740478514,
      "memory(GiB)": 77.56,
      "step": 89020,
      "token_acc": 0.5413533834586466,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 3.8141039372777517,
      "grad_norm": 6.022658348083496,
      "learning_rate": 1.3249688968936869e-05,
      "loss": 2.297369194030762,
      "memory(GiB)": 77.56,
      "step": 89025,
      "token_acc": 0.47101449275362317,
      "train_speed(iter/s)": 1.437815
    },
    {
      "epoch": 3.8143181526070005,
      "grad_norm": 4.894735336303711,
      "learning_rate": 1.3245126109480076e-05,
      "loss": 2.674047088623047,
      "memory(GiB)": 77.56,
      "step": 89030,
      "token_acc": 0.4447852760736196,
      "train_speed(iter/s)": 1.437828
    },
    {
      "epoch": 3.8145323679362493,
      "grad_norm": 5.074991703033447,
      "learning_rate": 1.3240563915872316e-05,
      "loss": 2.2101449966430664,
      "memory(GiB)": 77.56,
      "step": 89035,
      "token_acc": 0.5601374570446735,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 3.8147465832654985,
      "grad_norm": 4.689578056335449,
      "learning_rate": 1.3236002388196206e-05,
      "loss": 2.1054061889648437,
      "memory(GiB)": 77.56,
      "step": 89040,
      "token_acc": 0.5363636363636364,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 3.8149607985947473,
      "grad_norm": 6.7528886795043945,
      "learning_rate": 1.3231441526534422e-05,
      "loss": 2.1909435272216795,
      "memory(GiB)": 77.56,
      "step": 89045,
      "token_acc": 0.5655737704918032,
      "train_speed(iter/s)": 1.437804
    },
    {
      "epoch": 3.815175013923996,
      "grad_norm": 6.765122413635254,
      "learning_rate": 1.3226881330969559e-05,
      "loss": 2.552584648132324,
      "memory(GiB)": 77.56,
      "step": 89050,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.8153892292532454,
      "grad_norm": 6.50324821472168,
      "learning_rate": 1.3222321801584242e-05,
      "loss": 2.1903121948242186,
      "memory(GiB)": 77.56,
      "step": 89055,
      "token_acc": 0.5274725274725275,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.8156034445824942,
      "grad_norm": 6.8679890632629395,
      "learning_rate": 1.3217762938461065e-05,
      "loss": 2.3075584411621093,
      "memory(GiB)": 77.56,
      "step": 89060,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.815817659911743,
      "grad_norm": 5.843669891357422,
      "learning_rate": 1.32132047416826e-05,
      "loss": 2.451921844482422,
      "memory(GiB)": 77.56,
      "step": 89065,
      "token_acc": 0.4984894259818731,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.8160318752409923,
      "grad_norm": 4.5095367431640625,
      "learning_rate": 1.3208647211331454e-05,
      "loss": 2.275189208984375,
      "memory(GiB)": 77.56,
      "step": 89070,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.437857
    },
    {
      "epoch": 3.816246090570241,
      "grad_norm": 5.4449968338012695,
      "learning_rate": 1.3204090347490178e-05,
      "loss": 2.4361053466796876,
      "memory(GiB)": 77.56,
      "step": 89075,
      "token_acc": 0.503448275862069,
      "train_speed(iter/s)": 1.437847
    },
    {
      "epoch": 3.81646030589949,
      "grad_norm": 5.6485514640808105,
      "learning_rate": 1.3199534150241321e-05,
      "loss": 2.132061767578125,
      "memory(GiB)": 77.56,
      "step": 89080,
      "token_acc": 0.5708955223880597,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 3.816674521228739,
      "grad_norm": 5.526097297668457,
      "learning_rate": 1.319497861966743e-05,
      "loss": 2.0641841888427734,
      "memory(GiB)": 77.56,
      "step": 89085,
      "token_acc": 0.5393258426966292,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.816888736557988,
      "grad_norm": 5.482429027557373,
      "learning_rate": 1.3190423755851005e-05,
      "loss": 2.203058052062988,
      "memory(GiB)": 77.56,
      "step": 89090,
      "token_acc": 0.5186335403726708,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 3.817102951887237,
      "grad_norm": 8.089812278747559,
      "learning_rate": 1.3185869558874602e-05,
      "loss": 2.386006164550781,
      "memory(GiB)": 77.56,
      "step": 89095,
      "token_acc": 0.49185667752442996,
      "train_speed(iter/s)": 1.437884
    },
    {
      "epoch": 3.817317167216486,
      "grad_norm": 5.60097599029541,
      "learning_rate": 1.3181316028820695e-05,
      "loss": 2.5724180221557615,
      "memory(GiB)": 77.56,
      "step": 89100,
      "token_acc": 0.4306784660766962,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 3.817531382545735,
      "grad_norm": 11.27708625793457,
      "learning_rate": 1.31767631657718e-05,
      "loss": 2.1788700103759764,
      "memory(GiB)": 77.56,
      "step": 89105,
      "token_acc": 0.5639097744360902,
      "train_speed(iter/s)": 1.4379
    },
    {
      "epoch": 3.8177455978749837,
      "grad_norm": 5.431806564331055,
      "learning_rate": 1.3172210969810378e-05,
      "loss": 2.6047641754150392,
      "memory(GiB)": 77.56,
      "step": 89110,
      "token_acc": 0.47104247104247104,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 3.817959813204233,
      "grad_norm": 5.5743088722229,
      "learning_rate": 1.3167659441018909e-05,
      "loss": 2.4898139953613283,
      "memory(GiB)": 77.56,
      "step": 89115,
      "token_acc": 0.4542483660130719,
      "train_speed(iter/s)": 1.437923
    },
    {
      "epoch": 3.8181740285334818,
      "grad_norm": 5.825106143951416,
      "learning_rate": 1.3163108579479838e-05,
      "loss": 2.3108795166015623,
      "memory(GiB)": 77.56,
      "step": 89120,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 3.8183882438627306,
      "grad_norm": 8.026784896850586,
      "learning_rate": 1.3158558385275594e-05,
      "loss": 2.4738929748535154,
      "memory(GiB)": 77.56,
      "step": 89125,
      "token_acc": 0.4610169491525424,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 3.81860245919198,
      "grad_norm": 4.974749565124512,
      "learning_rate": 1.3154008858488643e-05,
      "loss": 2.300914001464844,
      "memory(GiB)": 77.56,
      "step": 89130,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 3.8188166745212286,
      "grad_norm": 5.531448841094971,
      "learning_rate": 1.314945999920138e-05,
      "loss": 2.135707664489746,
      "memory(GiB)": 77.56,
      "step": 89135,
      "token_acc": 0.5703422053231939,
      "train_speed(iter/s)": 1.437964
    },
    {
      "epoch": 3.8190308898504775,
      "grad_norm": 5.284693717956543,
      "learning_rate": 1.3144911807496224e-05,
      "loss": 2.478932571411133,
      "memory(GiB)": 77.56,
      "step": 89140,
      "token_acc": 0.5316455696202531,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 3.8192451051797267,
      "grad_norm": 6.723596096038818,
      "learning_rate": 1.3140364283455558e-05,
      "loss": 2.5755098342895506,
      "memory(GiB)": 77.56,
      "step": 89145,
      "token_acc": 0.4872611464968153,
      "train_speed(iter/s)": 1.437947
    },
    {
      "epoch": 3.8194593205089755,
      "grad_norm": 6.281871318817139,
      "learning_rate": 1.3135817427161762e-05,
      "loss": 2.6059696197509767,
      "memory(GiB)": 77.56,
      "step": 89150,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.437936
    },
    {
      "epoch": 3.8196735358382243,
      "grad_norm": 6.7038702964782715,
      "learning_rate": 1.3131271238697223e-05,
      "loss": 2.353596305847168,
      "memory(GiB)": 77.56,
      "step": 89155,
      "token_acc": 0.5077399380804953,
      "train_speed(iter/s)": 1.43795
    },
    {
      "epoch": 3.8198877511674736,
      "grad_norm": 5.308282375335693,
      "learning_rate": 1.3126725718144295e-05,
      "loss": 2.651363563537598,
      "memory(GiB)": 77.56,
      "step": 89160,
      "token_acc": 0.44025157232704404,
      "train_speed(iter/s)": 1.437957
    },
    {
      "epoch": 3.8201019664967224,
      "grad_norm": 4.589012622833252,
      "learning_rate": 1.312218086558532e-05,
      "loss": 2.337087631225586,
      "memory(GiB)": 77.56,
      "step": 89165,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 3.820316181825971,
      "grad_norm": 7.262591361999512,
      "learning_rate": 1.3117636681102635e-05,
      "loss": 2.4449745178222657,
      "memory(GiB)": 77.56,
      "step": 89170,
      "token_acc": 0.4916387959866221,
      "train_speed(iter/s)": 1.437981
    },
    {
      "epoch": 3.8205303971552205,
      "grad_norm": 8.451205253601074,
      "learning_rate": 1.311309316477854e-05,
      "loss": 2.1115385055541993,
      "memory(GiB)": 77.56,
      "step": 89175,
      "token_acc": 0.5211726384364821,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 3.8207446124844693,
      "grad_norm": 5.48556661605835,
      "learning_rate": 1.3108550316695383e-05,
      "loss": 2.4338003158569337,
      "memory(GiB)": 77.56,
      "step": 89180,
      "token_acc": 0.48138297872340424,
      "train_speed(iter/s)": 1.437987
    },
    {
      "epoch": 3.820958827813718,
      "grad_norm": 5.887182712554932,
      "learning_rate": 1.3104008136935431e-05,
      "loss": 2.2330379486083984,
      "memory(GiB)": 77.56,
      "step": 89185,
      "token_acc": 0.525,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.8211730431429674,
      "grad_norm": 5.910205841064453,
      "learning_rate": 1.3099466625580992e-05,
      "loss": 1.8738798141479491,
      "memory(GiB)": 77.56,
      "step": 89190,
      "token_acc": 0.5563380281690141,
      "train_speed(iter/s)": 1.437997
    },
    {
      "epoch": 3.821387258472216,
      "grad_norm": 8.237466812133789,
      "learning_rate": 1.3094925782714335e-05,
      "loss": 2.2999019622802734,
      "memory(GiB)": 77.56,
      "step": 89195,
      "token_acc": 0.483271375464684,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 3.821601473801465,
      "grad_norm": 5.856817245483398,
      "learning_rate": 1.3090385608417716e-05,
      "loss": 2.5226097106933594,
      "memory(GiB)": 77.56,
      "step": 89200,
      "token_acc": 0.49063670411985016,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.8218156891307142,
      "grad_norm": 5.237555503845215,
      "learning_rate": 1.3085846102773386e-05,
      "loss": 2.376255989074707,
      "memory(GiB)": 77.56,
      "step": 89205,
      "token_acc": 0.48253968253968255,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 3.822029904459963,
      "grad_norm": 5.246608734130859,
      "learning_rate": 1.3081307265863568e-05,
      "loss": 2.372589111328125,
      "memory(GiB)": 77.56,
      "step": 89210,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.438001
    },
    {
      "epoch": 3.822244119789212,
      "grad_norm": 5.699592590332031,
      "learning_rate": 1.3076769097770514e-05,
      "loss": 2.2755014419555666,
      "memory(GiB)": 77.56,
      "step": 89215,
      "token_acc": 0.5176848874598071,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 3.822458335118461,
      "grad_norm": 5.594287872314453,
      "learning_rate": 1.3072231598576423e-05,
      "loss": 2.566649627685547,
      "memory(GiB)": 77.56,
      "step": 89220,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.82267255044771,
      "grad_norm": 5.3272881507873535,
      "learning_rate": 1.3067694768363502e-05,
      "loss": 2.299101638793945,
      "memory(GiB)": 77.56,
      "step": 89225,
      "token_acc": 0.5030120481927711,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 3.8228867657769587,
      "grad_norm": 7.6558661460876465,
      "learning_rate": 1.3063158607213938e-05,
      "loss": 2.296774482727051,
      "memory(GiB)": 77.56,
      "step": 89230,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.438001
    },
    {
      "epoch": 3.823100981106208,
      "grad_norm": 5.500546455383301,
      "learning_rate": 1.3058623115209884e-05,
      "loss": 2.2983016967773438,
      "memory(GiB)": 77.56,
      "step": 89235,
      "token_acc": 0.516245487364621,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 3.823315196435457,
      "grad_norm": 7.9366044998168945,
      "learning_rate": 1.3054088292433542e-05,
      "loss": 2.2561222076416017,
      "memory(GiB)": 77.56,
      "step": 89240,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438031
    },
    {
      "epoch": 3.8235294117647056,
      "grad_norm": 6.5696258544921875,
      "learning_rate": 1.3049554138967051e-05,
      "loss": 1.9168405532836914,
      "memory(GiB)": 77.56,
      "step": 89245,
      "token_acc": 0.5601374570446735,
      "train_speed(iter/s)": 1.438027
    },
    {
      "epoch": 3.823743627093955,
      "grad_norm": 7.222368240356445,
      "learning_rate": 1.3045020654892537e-05,
      "loss": 2.2084190368652346,
      "memory(GiB)": 77.56,
      "step": 89250,
      "token_acc": 0.5207667731629393,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 3.8239578424232037,
      "grad_norm": 6.450833797454834,
      "learning_rate": 1.3040487840292148e-05,
      "loss": 2.4299835205078124,
      "memory(GiB)": 77.56,
      "step": 89255,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 3.8241720577524525,
      "grad_norm": 5.956780910491943,
      "learning_rate": 1.3035955695247998e-05,
      "loss": 2.3340259552001954,
      "memory(GiB)": 77.56,
      "step": 89260,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438061
    },
    {
      "epoch": 3.8243862730817018,
      "grad_norm": 6.589070796966553,
      "learning_rate": 1.3031424219842186e-05,
      "loss": 2.348227691650391,
      "memory(GiB)": 77.56,
      "step": 89265,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.43806
    },
    {
      "epoch": 3.8246004884109506,
      "grad_norm": 6.042869567871094,
      "learning_rate": 1.3026893414156783e-05,
      "loss": 2.578312301635742,
      "memory(GiB)": 77.56,
      "step": 89270,
      "token_acc": 0.47904191616766467,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 3.8248147037401994,
      "grad_norm": 5.407898426055908,
      "learning_rate": 1.3022363278273908e-05,
      "loss": 2.3430669784545897,
      "memory(GiB)": 77.56,
      "step": 89275,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 3.8250289190694486,
      "grad_norm": 6.5342535972595215,
      "learning_rate": 1.3017833812275609e-05,
      "loss": 2.171712112426758,
      "memory(GiB)": 77.56,
      "step": 89280,
      "token_acc": 0.5381944444444444,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 3.8252431343986975,
      "grad_norm": 5.928194046020508,
      "learning_rate": 1.3013305016243949e-05,
      "loss": 2.6452568054199217,
      "memory(GiB)": 77.56,
      "step": 89285,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.438082
    },
    {
      "epoch": 3.8254573497279463,
      "grad_norm": 7.538603782653809,
      "learning_rate": 1.300877689026096e-05,
      "loss": 2.165818786621094,
      "memory(GiB)": 77.56,
      "step": 89290,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 3.8256715650571955,
      "grad_norm": 5.8745903968811035,
      "learning_rate": 1.300424943440866e-05,
      "loss": 2.473246192932129,
      "memory(GiB)": 77.56,
      "step": 89295,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.438085
    },
    {
      "epoch": 3.8258857803864443,
      "grad_norm": 6.3269805908203125,
      "learning_rate": 1.2999722648769103e-05,
      "loss": 2.387524223327637,
      "memory(GiB)": 77.56,
      "step": 89300,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 3.826099995715693,
      "grad_norm": 6.30497932434082,
      "learning_rate": 1.299519653342428e-05,
      "loss": 2.3783422470092774,
      "memory(GiB)": 77.56,
      "step": 89305,
      "token_acc": 0.5168539325842697,
      "train_speed(iter/s)": 1.4381
    },
    {
      "epoch": 3.8263142110449424,
      "grad_norm": 5.20912504196167,
      "learning_rate": 1.2990671088456186e-05,
      "loss": 2.3772661209106447,
      "memory(GiB)": 77.56,
      "step": 89310,
      "token_acc": 0.5408805031446541,
      "train_speed(iter/s)": 1.438119
    },
    {
      "epoch": 3.826528426374191,
      "grad_norm": 6.4588117599487305,
      "learning_rate": 1.2986146313946807e-05,
      "loss": 2.344211959838867,
      "memory(GiB)": 77.56,
      "step": 89315,
      "token_acc": 0.5180327868852459,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 3.82674264170344,
      "grad_norm": 5.583101272583008,
      "learning_rate": 1.298162220997809e-05,
      "loss": 2.261591339111328,
      "memory(GiB)": 77.56,
      "step": 89320,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.438104
    },
    {
      "epoch": 3.8269568570326893,
      "grad_norm": 4.968259334564209,
      "learning_rate": 1.2977098776632023e-05,
      "loss": 2.3683284759521483,
      "memory(GiB)": 77.56,
      "step": 89325,
      "token_acc": 0.47752808988764045,
      "train_speed(iter/s)": 1.438092
    },
    {
      "epoch": 3.827171072361938,
      "grad_norm": 7.383714199066162,
      "learning_rate": 1.2972576013990534e-05,
      "loss": 2.296504020690918,
      "memory(GiB)": 77.56,
      "step": 89330,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 3.827385287691187,
      "grad_norm": 6.4993577003479,
      "learning_rate": 1.2968053922135576e-05,
      "loss": 2.5657075881958007,
      "memory(GiB)": 77.56,
      "step": 89335,
      "token_acc": 0.47388059701492535,
      "train_speed(iter/s)": 1.438081
    },
    {
      "epoch": 3.827599503020436,
      "grad_norm": 5.624339580535889,
      "learning_rate": 1.2963532501149062e-05,
      "loss": 2.41269588470459,
      "memory(GiB)": 77.56,
      "step": 89340,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 3.827813718349685,
      "grad_norm": 8.805702209472656,
      "learning_rate": 1.2959011751112898e-05,
      "loss": 2.1869312286376954,
      "memory(GiB)": 77.56,
      "step": 89345,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.438124
    },
    {
      "epoch": 3.828027933678934,
      "grad_norm": 6.802753925323486,
      "learning_rate": 1.2954491672108987e-05,
      "loss": 2.471541976928711,
      "memory(GiB)": 77.56,
      "step": 89350,
      "token_acc": 0.47720364741641336,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 3.828242149008183,
      "grad_norm": 6.480319976806641,
      "learning_rate": 1.294997226421919e-05,
      "loss": 2.4483322143554687,
      "memory(GiB)": 77.56,
      "step": 89355,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.438121
    },
    {
      "epoch": 3.828456364337432,
      "grad_norm": 7.437613487243652,
      "learning_rate": 1.294545352752542e-05,
      "loss": 2.380429267883301,
      "memory(GiB)": 77.56,
      "step": 89360,
      "token_acc": 0.48548812664907653,
      "train_speed(iter/s)": 1.438122
    },
    {
      "epoch": 3.8286705796666807,
      "grad_norm": 8.145204544067383,
      "learning_rate": 1.2940935462109522e-05,
      "loss": 2.637093353271484,
      "memory(GiB)": 77.56,
      "step": 89365,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.438116
    },
    {
      "epoch": 3.82888479499593,
      "grad_norm": 7.535129547119141,
      "learning_rate": 1.2936418068053335e-05,
      "loss": 2.110073280334473,
      "memory(GiB)": 77.56,
      "step": 89370,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.438116
    },
    {
      "epoch": 3.8290990103251787,
      "grad_norm": 4.610408782958984,
      "learning_rate": 1.2931901345438708e-05,
      "loss": 2.3601507186889648,
      "memory(GiB)": 77.56,
      "step": 89375,
      "token_acc": 0.49193548387096775,
      "train_speed(iter/s)": 1.438141
    },
    {
      "epoch": 3.8293132256544276,
      "grad_norm": 7.503290176391602,
      "learning_rate": 1.2927385294347444e-05,
      "loss": 2.346113586425781,
      "memory(GiB)": 77.56,
      "step": 89380,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.438149
    },
    {
      "epoch": 3.829527440983677,
      "grad_norm": 6.240098476409912,
      "learning_rate": 1.292286991486139e-05,
      "loss": 2.2731552124023438,
      "memory(GiB)": 77.56,
      "step": 89385,
      "token_acc": 0.5641891891891891,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 3.8297416563129256,
      "grad_norm": 6.197078227996826,
      "learning_rate": 1.2918355207062327e-05,
      "loss": 2.187102699279785,
      "memory(GiB)": 77.56,
      "step": 89390,
      "token_acc": 0.5328185328185329,
      "train_speed(iter/s)": 1.438128
    },
    {
      "epoch": 3.829955871642175,
      "grad_norm": 7.338923931121826,
      "learning_rate": 1.2913841171032033e-05,
      "loss": 2.3082998275756834,
      "memory(GiB)": 77.56,
      "step": 89395,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.438131
    },
    {
      "epoch": 3.8301700869714237,
      "grad_norm": 5.5944623947143555,
      "learning_rate": 1.2909327806852305e-05,
      "loss": 2.19866886138916,
      "memory(GiB)": 77.56,
      "step": 89400,
      "token_acc": 0.5860655737704918,
      "train_speed(iter/s)": 1.438131
    },
    {
      "epoch": 3.8303843023006725,
      "grad_norm": 5.738093852996826,
      "learning_rate": 1.2904815114604901e-05,
      "loss": 2.348261260986328,
      "memory(GiB)": 77.56,
      "step": 89405,
      "token_acc": 0.5079365079365079,
      "train_speed(iter/s)": 1.438123
    },
    {
      "epoch": 3.8305985176299218,
      "grad_norm": 5.5591511726379395,
      "learning_rate": 1.2900303094371568e-05,
      "loss": 1.9802574157714843,
      "memory(GiB)": 77.56,
      "step": 89410,
      "token_acc": 0.5656108597285068,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 3.8308127329591706,
      "grad_norm": 7.392581939697266,
      "learning_rate": 1.2895791746234032e-05,
      "loss": 2.502816390991211,
      "memory(GiB)": 77.56,
      "step": 89415,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.438112
    },
    {
      "epoch": 3.8310269482884194,
      "grad_norm": 6.14442777633667,
      "learning_rate": 1.2891281070274052e-05,
      "loss": 2.414131927490234,
      "memory(GiB)": 77.56,
      "step": 89420,
      "token_acc": 0.49696969696969695,
      "train_speed(iter/s)": 1.438098
    },
    {
      "epoch": 3.8312411636176686,
      "grad_norm": 6.514246463775635,
      "learning_rate": 1.2886771066573323e-05,
      "loss": 2.4701881408691406,
      "memory(GiB)": 77.56,
      "step": 89425,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 3.8314553789469175,
      "grad_norm": 6.374442100524902,
      "learning_rate": 1.2882261735213552e-05,
      "loss": 2.2698043823242187,
      "memory(GiB)": 77.56,
      "step": 89430,
      "token_acc": 0.5106382978723404,
      "train_speed(iter/s)": 1.43812
    },
    {
      "epoch": 3.8316695942761663,
      "grad_norm": 6.142033576965332,
      "learning_rate": 1.2877753076276434e-05,
      "loss": 2.425623321533203,
      "memory(GiB)": 77.56,
      "step": 89435,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.438113
    },
    {
      "epoch": 3.8318838096054155,
      "grad_norm": 6.489376068115234,
      "learning_rate": 1.2873245089843621e-05,
      "loss": 2.5443748474121093,
      "memory(GiB)": 77.56,
      "step": 89440,
      "token_acc": 0.4691358024691358,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 3.8320980249346643,
      "grad_norm": 5.969970226287842,
      "learning_rate": 1.286873777599682e-05,
      "loss": 2.1370635986328126,
      "memory(GiB)": 77.56,
      "step": 89445,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 3.832312240263913,
      "grad_norm": 4.6389594078063965,
      "learning_rate": 1.2864231134817661e-05,
      "loss": 2.4667181015014648,
      "memory(GiB)": 77.56,
      "step": 89450,
      "token_acc": 0.43564356435643564,
      "train_speed(iter/s)": 1.438099
    },
    {
      "epoch": 3.8325264555931624,
      "grad_norm": 8.391698837280273,
      "learning_rate": 1.2859725166387798e-05,
      "loss": 2.4705230712890627,
      "memory(GiB)": 77.56,
      "step": 89455,
      "token_acc": 0.5170068027210885,
      "train_speed(iter/s)": 1.438104
    },
    {
      "epoch": 3.832740670922411,
      "grad_norm": 6.3625922203063965,
      "learning_rate": 1.2855219870788849e-05,
      "loss": 2.1654340744018556,
      "memory(GiB)": 77.56,
      "step": 89460,
      "token_acc": 0.5425101214574899,
      "train_speed(iter/s)": 1.438109
    },
    {
      "epoch": 3.83295488625166,
      "grad_norm": 5.567348957061768,
      "learning_rate": 1.2850715248102425e-05,
      "loss": 2.2136323928833006,
      "memory(GiB)": 77.56,
      "step": 89465,
      "token_acc": 0.49523809523809526,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 3.8331691015809093,
      "grad_norm": 5.1307759284973145,
      "learning_rate": 1.2846211298410155e-05,
      "loss": 2.3576023101806642,
      "memory(GiB)": 77.56,
      "step": 89470,
      "token_acc": 0.48031496062992124,
      "train_speed(iter/s)": 1.438129
    },
    {
      "epoch": 3.833383316910158,
      "grad_norm": 6.851169109344482,
      "learning_rate": 1.2841708021793608e-05,
      "loss": 2.3290983200073243,
      "memory(GiB)": 77.56,
      "step": 89475,
      "token_acc": 0.48226950354609927,
      "train_speed(iter/s)": 1.438127
    },
    {
      "epoch": 3.833597532239407,
      "grad_norm": 5.531679630279541,
      "learning_rate": 1.2837205418334392e-05,
      "loss": 2.3891551971435545,
      "memory(GiB)": 77.56,
      "step": 89480,
      "token_acc": 0.5366666666666666,
      "train_speed(iter/s)": 1.43812
    },
    {
      "epoch": 3.833811747568656,
      "grad_norm": 9.474173545837402,
      "learning_rate": 1.2832703488114055e-05,
      "loss": 2.4122554779052736,
      "memory(GiB)": 77.56,
      "step": 89485,
      "token_acc": 0.5347222222222222,
      "train_speed(iter/s)": 1.438121
    },
    {
      "epoch": 3.834025962897905,
      "grad_norm": 8.35528564453125,
      "learning_rate": 1.2828202231214171e-05,
      "loss": 2.1345138549804688,
      "memory(GiB)": 77.56,
      "step": 89490,
      "token_acc": 0.5197368421052632,
      "train_speed(iter/s)": 1.438097
    },
    {
      "epoch": 3.834240178227154,
      "grad_norm": 4.959789276123047,
      "learning_rate": 1.2823701647716269e-05,
      "loss": 2.350536346435547,
      "memory(GiB)": 77.56,
      "step": 89495,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438088
    },
    {
      "epoch": 3.834454393556403,
      "grad_norm": 7.224593162536621,
      "learning_rate": 1.281920173770187e-05,
      "loss": 2.2063188552856445,
      "memory(GiB)": 77.56,
      "step": 89500,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.438095
    },
    {
      "epoch": 3.834454393556403,
      "eval_loss": 2.2880513668060303,
      "eval_runtime": 14.1879,
      "eval_samples_per_second": 7.048,
      "eval_steps_per_second": 7.048,
      "eval_token_acc": 0.4837837837837838,
      "step": 89500
    },
    {
      "epoch": 3.834668608885652,
      "grad_norm": 5.8053412437438965,
      "learning_rate": 1.2814702501252524e-05,
      "loss": 2.4667795181274412,
      "memory(GiB)": 77.56,
      "step": 89505,
      "token_acc": 0.48141263940520446,
      "train_speed(iter/s)": 1.437732
    },
    {
      "epoch": 3.8348828242149007,
      "grad_norm": 6.08140754699707,
      "learning_rate": 1.2810203938449721e-05,
      "loss": 2.016112518310547,
      "memory(GiB)": 77.56,
      "step": 89510,
      "token_acc": 0.5461538461538461,
      "train_speed(iter/s)": 1.437748
    },
    {
      "epoch": 3.83509703954415,
      "grad_norm": 8.658079147338867,
      "learning_rate": 1.2805706049374967e-05,
      "loss": 2.3258127212524413,
      "memory(GiB)": 77.56,
      "step": 89515,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 3.8353112548733987,
      "grad_norm": 5.1748151779174805,
      "learning_rate": 1.280120883410974e-05,
      "loss": 2.5754547119140625,
      "memory(GiB)": 77.56,
      "step": 89520,
      "token_acc": 0.4305555555555556,
      "train_speed(iter/s)": 1.437773
    },
    {
      "epoch": 3.8355254702026476,
      "grad_norm": 19.345731735229492,
      "learning_rate": 1.2796712292735491e-05,
      "loss": 2.574457550048828,
      "memory(GiB)": 77.56,
      "step": 89525,
      "token_acc": 0.4670846394984326,
      "train_speed(iter/s)": 1.437775
    },
    {
      "epoch": 3.835739685531897,
      "grad_norm": 6.9523138999938965,
      "learning_rate": 1.2792216425333715e-05,
      "loss": 2.2901451110839846,
      "memory(GiB)": 77.56,
      "step": 89530,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.437781
    },
    {
      "epoch": 3.8359539008611456,
      "grad_norm": 6.754496097564697,
      "learning_rate": 1.2787721231985845e-05,
      "loss": 2.398523712158203,
      "memory(GiB)": 77.56,
      "step": 89535,
      "token_acc": 0.4557377049180328,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 3.8361681161903944,
      "grad_norm": 5.389326572418213,
      "learning_rate": 1.2783226712773305e-05,
      "loss": 2.414544105529785,
      "memory(GiB)": 77.56,
      "step": 89540,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.8363823315196437,
      "grad_norm": 6.015725612640381,
      "learning_rate": 1.2778732867777515e-05,
      "loss": 2.461769866943359,
      "memory(GiB)": 77.56,
      "step": 89545,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.8365965468488925,
      "grad_norm": 6.404618740081787,
      "learning_rate": 1.2774239697079915e-05,
      "loss": 2.221240425109863,
      "memory(GiB)": 77.56,
      "step": 89550,
      "token_acc": 0.5198412698412699,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.8368107621781413,
      "grad_norm": 6.301713466644287,
      "learning_rate": 1.2769747200761878e-05,
      "loss": 2.2808141708374023,
      "memory(GiB)": 77.56,
      "step": 89555,
      "token_acc": 0.4872881355932203,
      "train_speed(iter/s)": 1.437784
    },
    {
      "epoch": 3.8370249775073906,
      "grad_norm": 6.150553226470947,
      "learning_rate": 1.2765255378904778e-05,
      "loss": 2.2991519927978517,
      "memory(GiB)": 77.56,
      "step": 89560,
      "token_acc": 0.5177304964539007,
      "train_speed(iter/s)": 1.437789
    },
    {
      "epoch": 3.8372391928366394,
      "grad_norm": 6.455543518066406,
      "learning_rate": 1.2760764231590027e-05,
      "loss": 1.9785884857177733,
      "memory(GiB)": 77.56,
      "step": 89565,
      "token_acc": 0.5962962962962963,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 3.837453408165888,
      "grad_norm": 5.895117282867432,
      "learning_rate": 1.2756273758898962e-05,
      "loss": 2.1818775177001952,
      "memory(GiB)": 77.56,
      "step": 89570,
      "token_acc": 0.5244755244755245,
      "train_speed(iter/s)": 1.437811
    },
    {
      "epoch": 3.8376676234951375,
      "grad_norm": 4.784303665161133,
      "learning_rate": 1.2751783960912933e-05,
      "loss": 2.532892417907715,
      "memory(GiB)": 77.56,
      "step": 89575,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 3.8378818388243863,
      "grad_norm": 5.0392913818359375,
      "learning_rate": 1.2747294837713285e-05,
      "loss": 2.581121253967285,
      "memory(GiB)": 77.56,
      "step": 89580,
      "token_acc": 0.517799352750809,
      "train_speed(iter/s)": 1.437817
    },
    {
      "epoch": 3.838096054153635,
      "grad_norm": 5.183769702911377,
      "learning_rate": 1.2742806389381318e-05,
      "loss": 2.112078094482422,
      "memory(GiB)": 77.56,
      "step": 89585,
      "token_acc": 0.5377049180327869,
      "train_speed(iter/s)": 1.437817
    },
    {
      "epoch": 3.8383102694828843,
      "grad_norm": 5.146134853363037,
      "learning_rate": 1.2738318615998379e-05,
      "loss": 2.3564823150634764,
      "memory(GiB)": 77.56,
      "step": 89590,
      "token_acc": 0.5131964809384164,
      "train_speed(iter/s)": 1.437825
    },
    {
      "epoch": 3.838524484812133,
      "grad_norm": 6.693490505218506,
      "learning_rate": 1.273383151764575e-05,
      "loss": 2.170207977294922,
      "memory(GiB)": 77.56,
      "step": 89595,
      "token_acc": 0.5518394648829431,
      "train_speed(iter/s)": 1.437847
    },
    {
      "epoch": 3.838738700141382,
      "grad_norm": 5.216087341308594,
      "learning_rate": 1.2729345094404727e-05,
      "loss": 2.466537857055664,
      "memory(GiB)": 77.56,
      "step": 89600,
      "token_acc": 0.49846153846153846,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 3.838952915470631,
      "grad_norm": 4.496460914611816,
      "learning_rate": 1.2724859346356577e-05,
      "loss": 2.5224403381347655,
      "memory(GiB)": 77.56,
      "step": 89605,
      "token_acc": 0.49707602339181284,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.83916713079988,
      "grad_norm": 4.567159175872803,
      "learning_rate": 1.2720374273582553e-05,
      "loss": 2.3182855606079102,
      "memory(GiB)": 77.56,
      "step": 89610,
      "token_acc": 0.5242718446601942,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 3.839381346129129,
      "grad_norm": 5.612655162811279,
      "learning_rate": 1.2715889876163933e-05,
      "loss": 2.3745309829711916,
      "memory(GiB)": 77.56,
      "step": 89615,
      "token_acc": 0.4806451612903226,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.839595561458378,
      "grad_norm": 7.709139347076416,
      "learning_rate": 1.2711406154181933e-05,
      "loss": 2.1623903274536134,
      "memory(GiB)": 77.56,
      "step": 89620,
      "token_acc": 0.5205882352941177,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 3.839809776787627,
      "grad_norm": 5.680721759796143,
      "learning_rate": 1.27069231077178e-05,
      "loss": 1.8974676132202148,
      "memory(GiB)": 77.56,
      "step": 89625,
      "token_acc": 0.5204460966542751,
      "train_speed(iter/s)": 1.437847
    },
    {
      "epoch": 3.8400239921168757,
      "grad_norm": 5.09709358215332,
      "learning_rate": 1.270244073685274e-05,
      "loss": 2.3554325103759766,
      "memory(GiB)": 77.56,
      "step": 89630,
      "token_acc": 0.528169014084507,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.840238207446125,
      "grad_norm": 4.189144134521484,
      "learning_rate": 1.269795904166795e-05,
      "loss": 2.1337955474853514,
      "memory(GiB)": 77.56,
      "step": 89635,
      "token_acc": 0.5225563909774437,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 3.840452422775374,
      "grad_norm": 4.511162281036377,
      "learning_rate": 1.2693478022244631e-05,
      "loss": 2.325242042541504,
      "memory(GiB)": 77.56,
      "step": 89640,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.8406666381046226,
      "grad_norm": 5.370865821838379,
      "learning_rate": 1.268899767866394e-05,
      "loss": 2.1646139144897463,
      "memory(GiB)": 77.56,
      "step": 89645,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 3.840880853433872,
      "grad_norm": 5.54640007019043,
      "learning_rate": 1.2684518011007062e-05,
      "loss": 2.289965057373047,
      "memory(GiB)": 77.56,
      "step": 89650,
      "token_acc": 0.4831804281345566,
      "train_speed(iter/s)": 1.437883
    },
    {
      "epoch": 3.8410950687631207,
      "grad_norm": 6.399964809417725,
      "learning_rate": 1.2680039019355156e-05,
      "loss": 2.238947105407715,
      "memory(GiB)": 77.56,
      "step": 89655,
      "token_acc": 0.5369127516778524,
      "train_speed(iter/s)": 1.437884
    },
    {
      "epoch": 3.8413092840923695,
      "grad_norm": 6.106026649475098,
      "learning_rate": 1.2675560703789347e-05,
      "loss": 2.300439643859863,
      "memory(GiB)": 77.56,
      "step": 89660,
      "token_acc": 0.5192878338278932,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.8415234994216187,
      "grad_norm": 6.112517356872559,
      "learning_rate": 1.2671083064390765e-05,
      "loss": 2.6393333435058595,
      "memory(GiB)": 77.56,
      "step": 89665,
      "token_acc": 0.47572815533980584,
      "train_speed(iter/s)": 1.437898
    },
    {
      "epoch": 3.8417377147508676,
      "grad_norm": 6.421717166900635,
      "learning_rate": 1.2666606101240525e-05,
      "loss": 2.2042654037475584,
      "memory(GiB)": 77.56,
      "step": 89670,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 3.8419519300801164,
      "grad_norm": 4.967819690704346,
      "learning_rate": 1.2662129814419742e-05,
      "loss": 2.5706480026245115,
      "memory(GiB)": 77.56,
      "step": 89675,
      "token_acc": 0.4739583333333333,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 3.8421661454093656,
      "grad_norm": 5.273562431335449,
      "learning_rate": 1.2657654204009512e-05,
      "loss": 2.1526462554931642,
      "memory(GiB)": 77.56,
      "step": 89680,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.8423803607386144,
      "grad_norm": 5.4288554191589355,
      "learning_rate": 1.2653179270090897e-05,
      "loss": 2.146344757080078,
      "memory(GiB)": 77.56,
      "step": 89685,
      "token_acc": 0.5298804780876494,
      "train_speed(iter/s)": 1.43794
    },
    {
      "epoch": 3.8425945760678633,
      "grad_norm": 4.961272239685059,
      "learning_rate": 1.2648705012744983e-05,
      "loss": 2.5688634872436524,
      "memory(GiB)": 77.56,
      "step": 89690,
      "token_acc": 0.43197278911564624,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 3.8428087913971125,
      "grad_norm": 5.9478068351745605,
      "learning_rate": 1.2644231432052794e-05,
      "loss": 2.3850337982177736,
      "memory(GiB)": 77.56,
      "step": 89695,
      "token_acc": 0.5167173252279635,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 3.8430230067263613,
      "grad_norm": 6.486205101013184,
      "learning_rate": 1.2639758528095414e-05,
      "loss": 2.3262929916381836,
      "memory(GiB)": 77.56,
      "step": 89700,
      "token_acc": 0.5409252669039146,
      "train_speed(iter/s)": 1.43791
    },
    {
      "epoch": 3.84323722205561,
      "grad_norm": 5.780078887939453,
      "learning_rate": 1.2635286300953841e-05,
      "loss": 2.70859375,
      "memory(GiB)": 77.56,
      "step": 89705,
      "token_acc": 0.49142857142857144,
      "train_speed(iter/s)": 1.437912
    },
    {
      "epoch": 3.8434514373848594,
      "grad_norm": 7.040881633758545,
      "learning_rate": 1.2630814750709119e-05,
      "loss": 2.346533012390137,
      "memory(GiB)": 77.56,
      "step": 89710,
      "token_acc": 0.4789272030651341,
      "train_speed(iter/s)": 1.437922
    },
    {
      "epoch": 3.843665652714108,
      "grad_norm": 7.47244119644165,
      "learning_rate": 1.2626343877442242e-05,
      "loss": 2.394953155517578,
      "memory(GiB)": 77.56,
      "step": 89715,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.843879868043357,
      "grad_norm": 5.17090368270874,
      "learning_rate": 1.2621873681234208e-05,
      "loss": 2.286424255371094,
      "memory(GiB)": 77.56,
      "step": 89720,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437908
    },
    {
      "epoch": 3.8440940833726063,
      "grad_norm": 5.325435161590576,
      "learning_rate": 1.2617404162165997e-05,
      "loss": 2.4823562622070314,
      "memory(GiB)": 77.56,
      "step": 89725,
      "token_acc": 0.4807121661721068,
      "train_speed(iter/s)": 1.437914
    },
    {
      "epoch": 3.844308298701855,
      "grad_norm": 6.1892619132995605,
      "learning_rate": 1.2612935320318558e-05,
      "loss": 1.9413860321044922,
      "memory(GiB)": 77.56,
      "step": 89730,
      "token_acc": 0.584192439862543,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 3.8445225140311043,
      "grad_norm": 4.755699157714844,
      "learning_rate": 1.2608467155772885e-05,
      "loss": 2.2821407318115234,
      "memory(GiB)": 77.56,
      "step": 89735,
      "token_acc": 0.5317725752508361,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 3.844736729360353,
      "grad_norm": 5.3083271980285645,
      "learning_rate": 1.2603999668609906e-05,
      "loss": 2.456380844116211,
      "memory(GiB)": 77.56,
      "step": 89740,
      "token_acc": 0.47076023391812866,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 3.844950944689602,
      "grad_norm": 6.417150020599365,
      "learning_rate": 1.2599532858910556e-05,
      "loss": 2.306637001037598,
      "memory(GiB)": 77.56,
      "step": 89745,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.437903
    },
    {
      "epoch": 3.8451651600188512,
      "grad_norm": 6.258324146270752,
      "learning_rate": 1.259506672675575e-05,
      "loss": 2.401631736755371,
      "memory(GiB)": 77.56,
      "step": 89750,
      "token_acc": 0.4981684981684982,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.8453793753481,
      "grad_norm": 5.937260150909424,
      "learning_rate": 1.2590601272226383e-05,
      "loss": 2.233934783935547,
      "memory(GiB)": 77.56,
      "step": 89755,
      "token_acc": 0.5360824742268041,
      "train_speed(iter/s)": 1.43792
    },
    {
      "epoch": 3.845593590677349,
      "grad_norm": 6.198304653167725,
      "learning_rate": 1.2586136495403384e-05,
      "loss": 1.954640769958496,
      "memory(GiB)": 77.56,
      "step": 89760,
      "token_acc": 0.596,
      "train_speed(iter/s)": 1.437917
    },
    {
      "epoch": 3.845807806006598,
      "grad_norm": 6.78243350982666,
      "learning_rate": 1.25816723963676e-05,
      "loss": 2.422947883605957,
      "memory(GiB)": 77.56,
      "step": 89765,
      "token_acc": 0.5179856115107914,
      "train_speed(iter/s)": 1.43794
    },
    {
      "epoch": 3.846022021335847,
      "grad_norm": 6.163959980010986,
      "learning_rate": 1.257720897519994e-05,
      "loss": 2.6398540496826173,
      "memory(GiB)": 77.56,
      "step": 89770,
      "token_acc": 0.45671641791044776,
      "train_speed(iter/s)": 1.437946
    },
    {
      "epoch": 3.8462362366650957,
      "grad_norm": 6.760027885437012,
      "learning_rate": 1.2572746231981247e-05,
      "loss": 2.0449737548828124,
      "memory(GiB)": 77.56,
      "step": 89775,
      "token_acc": 0.5080645161290323,
      "train_speed(iter/s)": 1.437962
    },
    {
      "epoch": 3.846450451994345,
      "grad_norm": 7.761449337005615,
      "learning_rate": 1.256828416679236e-05,
      "loss": 2.4073207855224608,
      "memory(GiB)": 77.56,
      "step": 89780,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437966
    },
    {
      "epoch": 3.846664667323594,
      "grad_norm": 5.966902256011963,
      "learning_rate": 1.2563822779714125e-05,
      "loss": 2.4390159606933595,
      "memory(GiB)": 77.56,
      "step": 89785,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 3.8468788826528426,
      "grad_norm": 7.9838995933532715,
      "learning_rate": 1.2559362070827341e-05,
      "loss": 2.053150939941406,
      "memory(GiB)": 77.56,
      "step": 89790,
      "token_acc": 0.5392953929539296,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.847093097982092,
      "grad_norm": 5.227750778198242,
      "learning_rate": 1.2554902040212846e-05,
      "loss": 2.272739219665527,
      "memory(GiB)": 77.56,
      "step": 89795,
      "token_acc": 0.5291970802919708,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 3.8473073133113407,
      "grad_norm": 7.420921325683594,
      "learning_rate": 1.2550442687951431e-05,
      "loss": 2.548366928100586,
      "memory(GiB)": 77.56,
      "step": 89800,
      "token_acc": 0.4601226993865031,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 3.8475215286405895,
      "grad_norm": 6.045298099517822,
      "learning_rate": 1.2545984014123879e-05,
      "loss": 2.469654083251953,
      "memory(GiB)": 77.56,
      "step": 89805,
      "token_acc": 0.4668769716088328,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 3.8477357439698388,
      "grad_norm": 5.813005447387695,
      "learning_rate": 1.2541526018810956e-05,
      "loss": 2.4589685440063476,
      "memory(GiB)": 77.56,
      "step": 89810,
      "token_acc": 0.4934640522875817,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.8479499592990876,
      "grad_norm": 7.683053493499756,
      "learning_rate": 1.253706870209342e-05,
      "loss": 2.4205291748046873,
      "memory(GiB)": 77.56,
      "step": 89815,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 3.8481641746283364,
      "grad_norm": 5.512688159942627,
      "learning_rate": 1.2532612064052036e-05,
      "loss": 2.481915283203125,
      "memory(GiB)": 77.56,
      "step": 89820,
      "token_acc": 0.4746268656716418,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.8483783899575856,
      "grad_norm": 6.90846586227417,
      "learning_rate": 1.2528156104767536e-05,
      "loss": 2.2012306213378907,
      "memory(GiB)": 77.56,
      "step": 89825,
      "token_acc": 0.5254901960784314,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 3.8485926052868344,
      "grad_norm": 5.92537784576416,
      "learning_rate": 1.2523700824320638e-05,
      "loss": 2.336228942871094,
      "memory(GiB)": 77.56,
      "step": 89830,
      "token_acc": 0.5422077922077922,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 3.8488068206160833,
      "grad_norm": 5.809115886688232,
      "learning_rate": 1.2519246222792053e-05,
      "loss": 2.655837821960449,
      "memory(GiB)": 77.56,
      "step": 89835,
      "token_acc": 0.47315436241610737,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.8490210359453325,
      "grad_norm": 4.90449333190918,
      "learning_rate": 1.2514792300262474e-05,
      "loss": 2.62183837890625,
      "memory(GiB)": 77.56,
      "step": 89840,
      "token_acc": 0.4768392370572207,
      "train_speed(iter/s)": 1.437993
    },
    {
      "epoch": 3.8492352512745813,
      "grad_norm": 6.031124591827393,
      "learning_rate": 1.2510339056812603e-05,
      "loss": 2.1152647018432615,
      "memory(GiB)": 77.56,
      "step": 89845,
      "token_acc": 0.568,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 3.84944946660383,
      "grad_norm": 6.151834011077881,
      "learning_rate": 1.25058864925231e-05,
      "loss": 2.4015045166015625,
      "memory(GiB)": 77.56,
      "step": 89850,
      "token_acc": 0.5239852398523985,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.8496636819330794,
      "grad_norm": 5.381189823150635,
      "learning_rate": 1.2501434607474644e-05,
      "loss": 2.150976371765137,
      "memory(GiB)": 77.56,
      "step": 89855,
      "token_acc": 0.5086705202312138,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 3.849877897262328,
      "grad_norm": 7.384481906890869,
      "learning_rate": 1.249698340174788e-05,
      "loss": 2.34448127746582,
      "memory(GiB)": 77.56,
      "step": 89860,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 3.850092112591577,
      "grad_norm": 4.894954204559326,
      "learning_rate": 1.2492532875423445e-05,
      "loss": 1.7909559249877929,
      "memory(GiB)": 77.56,
      "step": 89865,
      "token_acc": 0.5582329317269076,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 3.8503063279208263,
      "grad_norm": 6.887175559997559,
      "learning_rate": 1.248808302858196e-05,
      "loss": 2.219059371948242,
      "memory(GiB)": 77.56,
      "step": 89870,
      "token_acc": 0.5166666666666667,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 3.850520543250075,
      "grad_norm": 7.515207290649414,
      "learning_rate": 1.2483633861304023e-05,
      "loss": 2.2352569580078123,
      "memory(GiB)": 77.56,
      "step": 89875,
      "token_acc": 0.5451388888888888,
      "train_speed(iter/s)": 1.438016
    },
    {
      "epoch": 3.850734758579324,
      "grad_norm": 6.325348854064941,
      "learning_rate": 1.2479185373670272e-05,
      "loss": 2.1009830474853515,
      "memory(GiB)": 77.56,
      "step": 89880,
      "token_acc": 0.48672566371681414,
      "train_speed(iter/s)": 1.438027
    },
    {
      "epoch": 3.850948973908573,
      "grad_norm": 6.187824726104736,
      "learning_rate": 1.2474737565761268e-05,
      "loss": 2.4757076263427735,
      "memory(GiB)": 77.56,
      "step": 89885,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 3.851163189237822,
      "grad_norm": 5.819389343261719,
      "learning_rate": 1.24702904376576e-05,
      "loss": 2.7836130142211912,
      "memory(GiB)": 77.56,
      "step": 89890,
      "token_acc": 0.41919191919191917,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 3.851377404567071,
      "grad_norm": 5.284809589385986,
      "learning_rate": 1.2465843989439824e-05,
      "loss": 2.3953598022460936,
      "memory(GiB)": 77.56,
      "step": 89895,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.438032
    },
    {
      "epoch": 3.85159161989632,
      "grad_norm": 5.492016792297363,
      "learning_rate": 1.246139822118848e-05,
      "loss": 2.2401538848876954,
      "memory(GiB)": 77.56,
      "step": 89900,
      "token_acc": 0.5939849624060151,
      "train_speed(iter/s)": 1.438042
    },
    {
      "epoch": 3.851805835225569,
      "grad_norm": 6.976442337036133,
      "learning_rate": 1.2456953132984133e-05,
      "loss": 2.1454036712646483,
      "memory(GiB)": 77.56,
      "step": 89905,
      "token_acc": 0.5045045045045045,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 3.8520200505548177,
      "grad_norm": 5.561842918395996,
      "learning_rate": 1.2452508724907303e-05,
      "loss": 2.2576921463012694,
      "memory(GiB)": 77.56,
      "step": 89910,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.438052
    },
    {
      "epoch": 3.852234265884067,
      "grad_norm": 6.199827194213867,
      "learning_rate": 1.2448064997038483e-05,
      "loss": 2.152077484130859,
      "memory(GiB)": 77.56,
      "step": 89915,
      "token_acc": 0.5053763440860215,
      "train_speed(iter/s)": 1.438052
    },
    {
      "epoch": 3.8524484812133157,
      "grad_norm": 8.509712219238281,
      "learning_rate": 1.2443621949458206e-05,
      "loss": 2.496550178527832,
      "memory(GiB)": 77.56,
      "step": 89920,
      "token_acc": 0.48464163822525597,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 3.8526626965425645,
      "grad_norm": 6.870933532714844,
      "learning_rate": 1.2439179582246952e-05,
      "loss": 2.1269506454467773,
      "memory(GiB)": 77.56,
      "step": 89925,
      "token_acc": 0.5291970802919708,
      "train_speed(iter/s)": 1.438049
    },
    {
      "epoch": 3.852876911871814,
      "grad_norm": 9.983698844909668,
      "learning_rate": 1.2434737895485188e-05,
      "loss": 2.5557470321655273,
      "memory(GiB)": 77.56,
      "step": 89930,
      "token_acc": 0.4612903225806452,
      "train_speed(iter/s)": 1.438062
    },
    {
      "epoch": 3.8530911272010626,
      "grad_norm": 7.707968711853027,
      "learning_rate": 1.2430296889253378e-05,
      "loss": 2.474230194091797,
      "memory(GiB)": 77.56,
      "step": 89935,
      "token_acc": 0.4962121212121212,
      "train_speed(iter/s)": 1.43806
    },
    {
      "epoch": 3.8533053425303114,
      "grad_norm": 5.554915904998779,
      "learning_rate": 1.2425856563631993e-05,
      "loss": 2.271873664855957,
      "memory(GiB)": 77.56,
      "step": 89940,
      "token_acc": 0.5535168195718655,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 3.8535195578595607,
      "grad_norm": 7.130796432495117,
      "learning_rate": 1.2421416918701467e-05,
      "loss": 2.307566261291504,
      "memory(GiB)": 77.56,
      "step": 89945,
      "token_acc": 0.5368421052631579,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 3.8537337731888095,
      "grad_norm": 6.80617618560791,
      "learning_rate": 1.2416977954542225e-05,
      "loss": 2.395248031616211,
      "memory(GiB)": 77.56,
      "step": 89950,
      "token_acc": 0.5075757575757576,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 3.8539479885180583,
      "grad_norm": 6.161218166351318,
      "learning_rate": 1.241253967123468e-05,
      "loss": 2.507639694213867,
      "memory(GiB)": 77.56,
      "step": 89955,
      "token_acc": 0.4621212121212121,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 3.8541622038473076,
      "grad_norm": 4.90719747543335,
      "learning_rate": 1.2408102068859228e-05,
      "loss": 2.6448198318481446,
      "memory(GiB)": 77.56,
      "step": 89960,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.438078
    },
    {
      "epoch": 3.8543764191765564,
      "grad_norm": 8.31674575805664,
      "learning_rate": 1.2403665147496285e-05,
      "loss": 2.3424072265625,
      "memory(GiB)": 77.56,
      "step": 89965,
      "token_acc": 0.4765625,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 3.854590634505805,
      "grad_norm": 5.838057994842529,
      "learning_rate": 1.2399228907226212e-05,
      "loss": 2.4374057769775392,
      "memory(GiB)": 77.56,
      "step": 89970,
      "token_acc": 0.48623853211009177,
      "train_speed(iter/s)": 1.438097
    },
    {
      "epoch": 3.8548048498350544,
      "grad_norm": 4.9149956703186035,
      "learning_rate": 1.2394793348129385e-05,
      "loss": 2.411842155456543,
      "memory(GiB)": 77.56,
      "step": 89975,
      "token_acc": 0.48546511627906974,
      "train_speed(iter/s)": 1.438115
    },
    {
      "epoch": 3.8550190651643033,
      "grad_norm": 7.946128845214844,
      "learning_rate": 1.2390358470286151e-05,
      "loss": 2.6184871673583983,
      "memory(GiB)": 77.56,
      "step": 89980,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.438118
    },
    {
      "epoch": 3.855233280493552,
      "grad_norm": 11.506965637207031,
      "learning_rate": 1.2385924273776839e-05,
      "loss": 2.282339859008789,
      "memory(GiB)": 77.56,
      "step": 89985,
      "token_acc": 0.4962686567164179,
      "train_speed(iter/s)": 1.438117
    },
    {
      "epoch": 3.8554474958228013,
      "grad_norm": 4.797327041625977,
      "learning_rate": 1.2381490758681808e-05,
      "loss": 2.4253971099853517,
      "memory(GiB)": 77.56,
      "step": 89990,
      "token_acc": 0.534965034965035,
      "train_speed(iter/s)": 1.438104
    },
    {
      "epoch": 3.85566171115205,
      "grad_norm": 6.015229225158691,
      "learning_rate": 1.2377057925081342e-05,
      "loss": 2.24902286529541,
      "memory(GiB)": 77.56,
      "step": 89995,
      "token_acc": 0.5140845070422535,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 3.855875926481299,
      "grad_norm": 6.512338161468506,
      "learning_rate": 1.2372625773055784e-05,
      "loss": 2.1010799407958984,
      "memory(GiB)": 77.56,
      "step": 90000,
      "token_acc": 0.5125,
      "train_speed(iter/s)": 1.438098
    },
    {
      "epoch": 3.855875926481299,
      "eval_loss": 2.1195545196533203,
      "eval_runtime": 14.1336,
      "eval_samples_per_second": 7.075,
      "eval_steps_per_second": 7.075,
      "eval_token_acc": 0.47844228094575797,
      "step": 90000
    },
    {
      "epoch": 3.856090141810548,
      "grad_norm": 5.665398597717285,
      "learning_rate": 1.236819430268541e-05,
      "loss": 2.787649726867676,
      "memory(GiB)": 77.56,
      "step": 90005,
      "token_acc": 0.485456369107322,
      "train_speed(iter/s)": 1.437773
    },
    {
      "epoch": 3.856304357139797,
      "grad_norm": 5.269782066345215,
      "learning_rate": 1.2363763514050496e-05,
      "loss": 2.5828184127807616,
      "memory(GiB)": 77.56,
      "step": 90010,
      "token_acc": 0.4969512195121951,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.856518572469046,
      "grad_norm": 6.276030540466309,
      "learning_rate": 1.235933340723131e-05,
      "loss": 2.0275732040405274,
      "memory(GiB)": 77.56,
      "step": 90015,
      "token_acc": 0.575,
      "train_speed(iter/s)": 1.437781
    },
    {
      "epoch": 3.856732787798295,
      "grad_norm": 6.831111431121826,
      "learning_rate": 1.2354903982308098e-05,
      "loss": 2.207332992553711,
      "memory(GiB)": 77.56,
      "step": 90020,
      "token_acc": 0.4889867841409692,
      "train_speed(iter/s)": 1.437796
    },
    {
      "epoch": 3.856947003127544,
      "grad_norm": 7.143763542175293,
      "learning_rate": 1.2350475239361126e-05,
      "loss": 2.3331787109375,
      "memory(GiB)": 77.56,
      "step": 90025,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.437801
    },
    {
      "epoch": 3.8571612184567927,
      "grad_norm": 6.3429436683654785,
      "learning_rate": 1.234604717847061e-05,
      "loss": 2.441649627685547,
      "memory(GiB)": 77.56,
      "step": 90030,
      "token_acc": 0.48344370860927155,
      "train_speed(iter/s)": 1.437811
    },
    {
      "epoch": 3.857375433786042,
      "grad_norm": 5.202672481536865,
      "learning_rate": 1.2341619799716775e-05,
      "loss": 2.5058883666992187,
      "memory(GiB)": 77.56,
      "step": 90035,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.437814
    },
    {
      "epoch": 3.857589649115291,
      "grad_norm": 6.648130893707275,
      "learning_rate": 1.2337193103179823e-05,
      "loss": 2.1851377487182617,
      "memory(GiB)": 77.56,
      "step": 90040,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437812
    },
    {
      "epoch": 3.8578038644445396,
      "grad_norm": 6.269743919372559,
      "learning_rate": 1.2332767088939935e-05,
      "loss": 2.31964168548584,
      "memory(GiB)": 77.56,
      "step": 90045,
      "token_acc": 0.5300751879699248,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 3.858018079773789,
      "grad_norm": 4.7640604972839355,
      "learning_rate": 1.2328341757077322e-05,
      "loss": 2.395347785949707,
      "memory(GiB)": 77.56,
      "step": 90050,
      "token_acc": 0.5288753799392097,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 3.8582322951030377,
      "grad_norm": 6.6956377029418945,
      "learning_rate": 1.2323917107672134e-05,
      "loss": 2.1309988021850588,
      "memory(GiB)": 77.56,
      "step": 90055,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.437808
    },
    {
      "epoch": 3.8584465104322865,
      "grad_norm": 5.487354278564453,
      "learning_rate": 1.2319493140804534e-05,
      "loss": 2.440011978149414,
      "memory(GiB)": 77.56,
      "step": 90060,
      "token_acc": 0.4425087108013937,
      "train_speed(iter/s)": 1.437796
    },
    {
      "epoch": 3.8586607257615357,
      "grad_norm": 7.776607513427734,
      "learning_rate": 1.2315069856554645e-05,
      "loss": 2.5134429931640625,
      "memory(GiB)": 77.56,
      "step": 90065,
      "token_acc": 0.4956521739130435,
      "train_speed(iter/s)": 1.437802
    },
    {
      "epoch": 3.8588749410907845,
      "grad_norm": 7.0737409591674805,
      "learning_rate": 1.2310647255002634e-05,
      "loss": 2.5011013031005858,
      "memory(GiB)": 77.56,
      "step": 90070,
      "token_acc": 0.5029411764705882,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 3.8590891564200334,
      "grad_norm": 7.587624549865723,
      "learning_rate": 1.23062253362286e-05,
      "loss": 2.2718746185302736,
      "memory(GiB)": 77.56,
      "step": 90075,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 3.8593033717492826,
      "grad_norm": 5.467663288116455,
      "learning_rate": 1.2301804100312642e-05,
      "loss": 2.3961555480957033,
      "memory(GiB)": 77.56,
      "step": 90080,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 3.8595175870785314,
      "grad_norm": 5.291189193725586,
      "learning_rate": 1.229738354733488e-05,
      "loss": 2.373830795288086,
      "memory(GiB)": 77.56,
      "step": 90085,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 3.8597318024077802,
      "grad_norm": 4.977430820465088,
      "learning_rate": 1.229296367737538e-05,
      "loss": 2.430828666687012,
      "memory(GiB)": 77.56,
      "step": 90090,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.8599460177370295,
      "grad_norm": 6.836342811584473,
      "learning_rate": 1.228854449051422e-05,
      "loss": 2.0693782806396483,
      "memory(GiB)": 77.56,
      "step": 90095,
      "token_acc": 0.5222672064777328,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.8601602330662783,
      "grad_norm": 4.87190055847168,
      "learning_rate": 1.2284125986831452e-05,
      "loss": 2.5876340866088867,
      "memory(GiB)": 77.56,
      "step": 90100,
      "token_acc": 0.46229508196721314,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 3.860374448395527,
      "grad_norm": 5.780139923095703,
      "learning_rate": 1.2279708166407106e-05,
      "loss": 2.474539947509766,
      "memory(GiB)": 77.56,
      "step": 90105,
      "token_acc": 0.4962962962962963,
      "train_speed(iter/s)": 1.437823
    },
    {
      "epoch": 3.8605886637247764,
      "grad_norm": 4.127979278564453,
      "learning_rate": 1.2275291029321246e-05,
      "loss": 2.362444305419922,
      "memory(GiB)": 77.56,
      "step": 90110,
      "token_acc": 0.4801223241590214,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 3.860802879054025,
      "grad_norm": 5.123636245727539,
      "learning_rate": 1.227087457565388e-05,
      "loss": 2.400991439819336,
      "memory(GiB)": 77.56,
      "step": 90115,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.437828
    },
    {
      "epoch": 3.861017094383274,
      "grad_norm": 4.854820251464844,
      "learning_rate": 1.2266458805485009e-05,
      "loss": 2.28357048034668,
      "memory(GiB)": 77.56,
      "step": 90120,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 3.8612313097125233,
      "grad_norm": 5.11182975769043,
      "learning_rate": 1.2262043718894633e-05,
      "loss": 2.468713569641113,
      "memory(GiB)": 77.56,
      "step": 90125,
      "token_acc": 0.47096774193548385,
      "train_speed(iter/s)": 1.437826
    },
    {
      "epoch": 3.861445525041772,
      "grad_norm": 6.18702507019043,
      "learning_rate": 1.2257629315962726e-05,
      "loss": 2.417119598388672,
      "memory(GiB)": 77.56,
      "step": 90130,
      "token_acc": 0.4983388704318937,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 3.861659740371021,
      "grad_norm": 5.6934051513671875,
      "learning_rate": 1.2253215596769285e-05,
      "loss": 2.258602523803711,
      "memory(GiB)": 77.56,
      "step": 90135,
      "token_acc": 0.4820359281437126,
      "train_speed(iter/s)": 1.437835
    },
    {
      "epoch": 3.86187395570027,
      "grad_norm": 5.434886455535889,
      "learning_rate": 1.2248802561394234e-05,
      "loss": 2.2361404418945314,
      "memory(GiB)": 77.56,
      "step": 90140,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.437855
    },
    {
      "epoch": 3.862088171029519,
      "grad_norm": 7.693119525909424,
      "learning_rate": 1.2244390209917555e-05,
      "loss": 2.6497882843017577,
      "memory(GiB)": 77.56,
      "step": 90145,
      "token_acc": 0.5072463768115942,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 3.8623023863587678,
      "grad_norm": 9.227799415588379,
      "learning_rate": 1.2239978542419162e-05,
      "loss": 2.240663528442383,
      "memory(GiB)": 77.56,
      "step": 90150,
      "token_acc": 0.5346153846153846,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 3.862516601688017,
      "grad_norm": 6.612579822540283,
      "learning_rate": 1.2235567558978983e-05,
      "loss": 2.1815208435058593,
      "memory(GiB)": 77.56,
      "step": 90155,
      "token_acc": 0.5258358662613982,
      "train_speed(iter/s)": 1.437849
    },
    {
      "epoch": 3.862730817017266,
      "grad_norm": 4.85897970199585,
      "learning_rate": 1.2231157259676923e-05,
      "loss": 2.305440330505371,
      "memory(GiB)": 77.56,
      "step": 90160,
      "token_acc": 0.5220338983050847,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.8629450323465147,
      "grad_norm": 8.610416412353516,
      "learning_rate": 1.2226747644592862e-05,
      "loss": 2.2540578842163086,
      "memory(GiB)": 77.56,
      "step": 90165,
      "token_acc": 0.4816053511705686,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.863159247675764,
      "grad_norm": 7.405158042907715,
      "learning_rate": 1.2222338713806718e-05,
      "loss": 2.297007751464844,
      "memory(GiB)": 77.56,
      "step": 90170,
      "token_acc": 0.5201238390092879,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 3.8633734630050127,
      "grad_norm": 7.38932466506958,
      "learning_rate": 1.2217930467398348e-05,
      "loss": 2.5066806793212892,
      "memory(GiB)": 77.56,
      "step": 90175,
      "token_acc": 0.46885245901639344,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 3.8635876783342615,
      "grad_norm": 6.427505016326904,
      "learning_rate": 1.2213522905447612e-05,
      "loss": 2.128851890563965,
      "memory(GiB)": 77.56,
      "step": 90180,
      "token_acc": 0.5157232704402516,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 3.863801893663511,
      "grad_norm": 5.607212066650391,
      "learning_rate": 1.2209116028034351e-05,
      "loss": 2.5071102142333985,
      "memory(GiB)": 77.56,
      "step": 90185,
      "token_acc": 0.4980544747081712,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 3.8640161089927596,
      "grad_norm": 6.2990264892578125,
      "learning_rate": 1.2204709835238388e-05,
      "loss": 2.4463857650756835,
      "memory(GiB)": 77.56,
      "step": 90190,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.8642303243220084,
      "grad_norm": 6.721704006195068,
      "learning_rate": 1.2200304327139578e-05,
      "loss": 2.769192123413086,
      "memory(GiB)": 77.56,
      "step": 90195,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.437883
    },
    {
      "epoch": 3.8644445396512577,
      "grad_norm": 6.095344543457031,
      "learning_rate": 1.2195899503817715e-05,
      "loss": 2.329058074951172,
      "memory(GiB)": 77.56,
      "step": 90200,
      "token_acc": 0.5134099616858238,
      "train_speed(iter/s)": 1.437893
    },
    {
      "epoch": 3.8646587549805065,
      "grad_norm": 6.428568363189697,
      "learning_rate": 1.2191495365352596e-05,
      "loss": 2.379052925109863,
      "memory(GiB)": 77.56,
      "step": 90205,
      "token_acc": 0.4953271028037383,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 3.8648729703097553,
      "grad_norm": 8.200699806213379,
      "learning_rate": 1.2187091911823989e-05,
      "loss": 2.381650924682617,
      "memory(GiB)": 77.56,
      "step": 90210,
      "token_acc": 0.545816733067729,
      "train_speed(iter/s)": 1.437905
    },
    {
      "epoch": 3.8650871856390046,
      "grad_norm": 6.794419765472412,
      "learning_rate": 1.21826891433117e-05,
      "loss": 2.5461925506591796,
      "memory(GiB)": 77.56,
      "step": 90215,
      "token_acc": 0.46017699115044247,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 3.8653014009682534,
      "grad_norm": 7.485188961029053,
      "learning_rate": 1.2178287059895471e-05,
      "loss": 2.2980983734130858,
      "memory(GiB)": 77.56,
      "step": 90220,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.437883
    },
    {
      "epoch": 3.865515616297502,
      "grad_norm": 5.096858024597168,
      "learning_rate": 1.2173885661655038e-05,
      "loss": 2.5063623428344726,
      "memory(GiB)": 77.56,
      "step": 90225,
      "token_acc": 0.5015479876160991,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 3.8657298316267514,
      "grad_norm": 5.671842575073242,
      "learning_rate": 1.2169484948670162e-05,
      "loss": 1.9846601486206055,
      "memory(GiB)": 77.56,
      "step": 90230,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.437882
    },
    {
      "epoch": 3.8659440469560002,
      "grad_norm": 5.549198627471924,
      "learning_rate": 1.2165084921020558e-05,
      "loss": 2.1314945220947266,
      "memory(GiB)": 77.56,
      "step": 90235,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 3.866158262285249,
      "grad_norm": 6.721835613250732,
      "learning_rate": 1.216068557878593e-05,
      "loss": 2.602103042602539,
      "memory(GiB)": 77.56,
      "step": 90240,
      "token_acc": 0.4551083591331269,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 3.8663724776144983,
      "grad_norm": 4.766403675079346,
      "learning_rate": 1.215628692204598e-05,
      "loss": 2.385386085510254,
      "memory(GiB)": 77.56,
      "step": 90245,
      "token_acc": 0.4765625,
      "train_speed(iter/s)": 1.437912
    },
    {
      "epoch": 3.866586692943747,
      "grad_norm": 7.1068596839904785,
      "learning_rate": 1.2151888950880374e-05,
      "loss": 2.377451705932617,
      "memory(GiB)": 77.56,
      "step": 90250,
      "token_acc": 0.5252100840336135,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 3.866800908272996,
      "grad_norm": 8.329933166503906,
      "learning_rate": 1.2147491665368826e-05,
      "loss": 2.2984407424926756,
      "memory(GiB)": 77.56,
      "step": 90255,
      "token_acc": 0.496,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 3.867015123602245,
      "grad_norm": 8.43309211730957,
      "learning_rate": 1.214309506559097e-05,
      "loss": 2.626738166809082,
      "memory(GiB)": 77.56,
      "step": 90260,
      "token_acc": 0.47232472324723246,
      "train_speed(iter/s)": 1.437939
    },
    {
      "epoch": 3.867229338931494,
      "grad_norm": 7.081747531890869,
      "learning_rate": 1.213869915162646e-05,
      "loss": 2.3862384796142577,
      "memory(GiB)": 77.56,
      "step": 90265,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.437934
    },
    {
      "epoch": 3.867443554260743,
      "grad_norm": 10.188179016113281,
      "learning_rate": 1.2134303923554935e-05,
      "loss": 2.241412353515625,
      "memory(GiB)": 77.56,
      "step": 90270,
      "token_acc": 0.47333333333333333,
      "train_speed(iter/s)": 1.437944
    },
    {
      "epoch": 3.867657769589992,
      "grad_norm": 5.070733070373535,
      "learning_rate": 1.2129909381455996e-05,
      "loss": 2.3196861267089846,
      "memory(GiB)": 77.56,
      "step": 90275,
      "token_acc": 0.5448028673835126,
      "train_speed(iter/s)": 1.437943
    },
    {
      "epoch": 3.867871984919241,
      "grad_norm": 9.10721492767334,
      "learning_rate": 1.2125515525409293e-05,
      "loss": 2.4494951248168944,
      "memory(GiB)": 77.56,
      "step": 90280,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.437957
    },
    {
      "epoch": 3.8680862002484897,
      "grad_norm": 4.724442958831787,
      "learning_rate": 1.2121122355494385e-05,
      "loss": 2.010164260864258,
      "memory(GiB)": 77.56,
      "step": 90285,
      "token_acc": 0.5574324324324325,
      "train_speed(iter/s)": 1.437947
    },
    {
      "epoch": 3.868300415577739,
      "grad_norm": 6.833724021911621,
      "learning_rate": 1.21167298717909e-05,
      "loss": 2.4033851623535156,
      "memory(GiB)": 77.56,
      "step": 90290,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 3.8685146309069878,
      "grad_norm": 6.595108985900879,
      "learning_rate": 1.2112338074378381e-05,
      "loss": 2.6386344909667967,
      "memory(GiB)": 77.56,
      "step": 90295,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.437946
    },
    {
      "epoch": 3.8687288462362366,
      "grad_norm": 4.952311992645264,
      "learning_rate": 1.210794696333641e-05,
      "loss": 2.2181528091430662,
      "memory(GiB)": 77.56,
      "step": 90300,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437955
    },
    {
      "epoch": 3.868943061565486,
      "grad_norm": 5.8525567054748535,
      "learning_rate": 1.210355653874452e-05,
      "loss": 2.2108102798461915,
      "memory(GiB)": 77.56,
      "step": 90305,
      "token_acc": 0.5225806451612903,
      "train_speed(iter/s)": 1.437944
    },
    {
      "epoch": 3.8691572768947347,
      "grad_norm": 6.393714427947998,
      "learning_rate": 1.2099166800682238e-05,
      "loss": 2.5033590316772463,
      "memory(GiB)": 77.56,
      "step": 90310,
      "token_acc": 0.4558303886925795,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 3.8693714922239835,
      "grad_norm": 5.960321426391602,
      "learning_rate": 1.209477774922912e-05,
      "loss": 2.2494022369384767,
      "memory(GiB)": 77.56,
      "step": 90315,
      "token_acc": 0.545816733067729,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 3.8695857075532327,
      "grad_norm": 7.149394989013672,
      "learning_rate": 1.2090389384464662e-05,
      "loss": 2.303862380981445,
      "memory(GiB)": 77.56,
      "step": 90320,
      "token_acc": 0.5406360424028268,
      "train_speed(iter/s)": 1.437944
    },
    {
      "epoch": 3.8697999228824815,
      "grad_norm": 7.303092956542969,
      "learning_rate": 1.2086001706468358e-05,
      "loss": 2.2746608734130858,
      "memory(GiB)": 77.56,
      "step": 90325,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 3.8700141382117303,
      "grad_norm": 5.146965980529785,
      "learning_rate": 1.2081614715319706e-05,
      "loss": 2.027683639526367,
      "memory(GiB)": 77.56,
      "step": 90330,
      "token_acc": 0.5528455284552846,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 3.8702283535409796,
      "grad_norm": 6.2376203536987305,
      "learning_rate": 1.207722841109815e-05,
      "loss": 2.205574798583984,
      "memory(GiB)": 77.56,
      "step": 90335,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 3.8704425688702284,
      "grad_norm": 6.945105075836182,
      "learning_rate": 1.2072842793883199e-05,
      "loss": 2.2663864135742187,
      "memory(GiB)": 77.56,
      "step": 90340,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 3.8706567841994772,
      "grad_norm": 7.281012058258057,
      "learning_rate": 1.2068457863754273e-05,
      "loss": 2.123605728149414,
      "memory(GiB)": 77.56,
      "step": 90345,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.437934
    },
    {
      "epoch": 3.8708709995287265,
      "grad_norm": 5.775146007537842,
      "learning_rate": 1.2064073620790823e-05,
      "loss": 2.2103017807006835,
      "memory(GiB)": 77.56,
      "step": 90350,
      "token_acc": 0.5179153094462541,
      "train_speed(iter/s)": 1.437912
    },
    {
      "epoch": 3.8710852148579753,
      "grad_norm": 5.414769172668457,
      "learning_rate": 1.205969006507226e-05,
      "loss": 2.1214412689208983,
      "memory(GiB)": 77.56,
      "step": 90355,
      "token_acc": 0.5110294117647058,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.871299430187224,
      "grad_norm": 6.277258396148682,
      "learning_rate": 1.205530719667799e-05,
      "loss": 2.454164505004883,
      "memory(GiB)": 77.56,
      "step": 90360,
      "token_acc": 0.4867021276595745,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.8715136455164734,
      "grad_norm": 5.787603378295898,
      "learning_rate": 1.2050925015687437e-05,
      "loss": 2.143338203430176,
      "memory(GiB)": 77.56,
      "step": 90365,
      "token_acc": 0.554006968641115,
      "train_speed(iter/s)": 1.437942
    },
    {
      "epoch": 3.871727860845722,
      "grad_norm": 4.927225112915039,
      "learning_rate": 1.2046543522179966e-05,
      "loss": 2.135004997253418,
      "memory(GiB)": 77.56,
      "step": 90370,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.437949
    },
    {
      "epoch": 3.871942076174971,
      "grad_norm": 5.743471622467041,
      "learning_rate": 1.2042162716234972e-05,
      "loss": 2.243510437011719,
      "memory(GiB)": 77.56,
      "step": 90375,
      "token_acc": 0.5340501792114696,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 3.8721562915042202,
      "grad_norm": 6.920787811279297,
      "learning_rate": 1.2037782597931812e-05,
      "loss": 2.5593738555908203,
      "memory(GiB)": 77.56,
      "step": 90380,
      "token_acc": 0.47474747474747475,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 3.872370506833469,
      "grad_norm": 5.611932754516602,
      "learning_rate": 1.2033403167349833e-05,
      "loss": 2.5393428802490234,
      "memory(GiB)": 77.56,
      "step": 90385,
      "token_acc": 0.45794392523364486,
      "train_speed(iter/s)": 1.437982
    },
    {
      "epoch": 3.872584722162718,
      "grad_norm": 7.454343318939209,
      "learning_rate": 1.2029024424568363e-05,
      "loss": 2.1756834030151366,
      "memory(GiB)": 77.56,
      "step": 90390,
      "token_acc": 0.5405405405405406,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.872798937491967,
      "grad_norm": 7.0108256340026855,
      "learning_rate": 1.2024646369666731e-05,
      "loss": 2.5703283309936524,
      "memory(GiB)": 77.56,
      "step": 90395,
      "token_acc": 0.516245487364621,
      "train_speed(iter/s)": 1.437959
    },
    {
      "epoch": 3.873013152821216,
      "grad_norm": 6.987051486968994,
      "learning_rate": 1.202026900272426e-05,
      "loss": 2.0040735244750976,
      "memory(GiB)": 77.56,
      "step": 90400,
      "token_acc": 0.5283842794759825,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 3.8732273681504648,
      "grad_norm": 4.960381031036377,
      "learning_rate": 1.201589232382025e-05,
      "loss": 2.2686237335205077,
      "memory(GiB)": 77.56,
      "step": 90405,
      "token_acc": 0.5289855072463768,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 3.873441583479714,
      "grad_norm": 6.1224684715271,
      "learning_rate": 1.2011516333033979e-05,
      "loss": 2.3726947784423826,
      "memory(GiB)": 77.56,
      "step": 90410,
      "token_acc": 0.47058823529411764,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 3.873655798808963,
      "grad_norm": 7.679008960723877,
      "learning_rate": 1.2007141030444723e-05,
      "loss": 2.2848957061767576,
      "memory(GiB)": 77.56,
      "step": 90415,
      "token_acc": 0.5062240663900415,
      "train_speed(iter/s)": 1.437973
    },
    {
      "epoch": 3.8738700141382116,
      "grad_norm": 5.951594352722168,
      "learning_rate": 1.2002766416131739e-05,
      "loss": 2.4400360107421877,
      "memory(GiB)": 77.56,
      "step": 90420,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 3.874084229467461,
      "grad_norm": 6.499587059020996,
      "learning_rate": 1.1998392490174299e-05,
      "loss": 2.2784664154052736,
      "memory(GiB)": 77.56,
      "step": 90425,
      "token_acc": 0.5274261603375527,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 3.8742984447967097,
      "grad_norm": 5.811992645263672,
      "learning_rate": 1.1994019252651611e-05,
      "loss": 2.1423919677734373,
      "memory(GiB)": 77.56,
      "step": 90430,
      "token_acc": 0.5662251655629139,
      "train_speed(iter/s)": 1.437981
    },
    {
      "epoch": 3.8745126601259585,
      "grad_norm": 5.947575092315674,
      "learning_rate": 1.1989646703642931e-05,
      "loss": 2.2671558380126955,
      "memory(GiB)": 77.56,
      "step": 90435,
      "token_acc": 0.5408805031446541,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 3.8747268754552078,
      "grad_norm": 6.770403861999512,
      "learning_rate": 1.1985274843227456e-05,
      "loss": 2.4048460006713865,
      "memory(GiB)": 77.56,
      "step": 90440,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 3.8749410907844566,
      "grad_norm": 6.454441547393799,
      "learning_rate": 1.1980903671484389e-05,
      "loss": 2.4436107635498048,
      "memory(GiB)": 77.56,
      "step": 90445,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 3.8751553061137054,
      "grad_norm": 7.0906548500061035,
      "learning_rate": 1.1976533188492922e-05,
      "loss": 2.0948753356933594,
      "memory(GiB)": 77.56,
      "step": 90450,
      "token_acc": 0.5934065934065934,
      "train_speed(iter/s)": 1.438021
    },
    {
      "epoch": 3.8753695214429547,
      "grad_norm": 5.974190711975098,
      "learning_rate": 1.1972163394332203e-05,
      "loss": 2.400987243652344,
      "memory(GiB)": 77.56,
      "step": 90455,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 3.8755837367722035,
      "grad_norm": 5.458359241485596,
      "learning_rate": 1.1967794289081436e-05,
      "loss": 2.177106475830078,
      "memory(GiB)": 77.56,
      "step": 90460,
      "token_acc": 0.5110410094637224,
      "train_speed(iter/s)": 1.438016
    },
    {
      "epoch": 3.8757979521014523,
      "grad_norm": 6.068882465362549,
      "learning_rate": 1.1963425872819755e-05,
      "loss": 2.225819206237793,
      "memory(GiB)": 77.56,
      "step": 90465,
      "token_acc": 0.4752475247524752,
      "train_speed(iter/s)": 1.43802
    },
    {
      "epoch": 3.8760121674307015,
      "grad_norm": 5.950623989105225,
      "learning_rate": 1.1959058145626289e-05,
      "loss": 2.334121322631836,
      "memory(GiB)": 77.56,
      "step": 90470,
      "token_acc": 0.5429553264604811,
      "train_speed(iter/s)": 1.438034
    },
    {
      "epoch": 3.8762263827599504,
      "grad_norm": 6.657140254974365,
      "learning_rate": 1.1954691107580174e-05,
      "loss": 2.392098808288574,
      "memory(GiB)": 77.56,
      "step": 90475,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438037
    },
    {
      "epoch": 3.876440598089199,
      "grad_norm": 5.313343048095703,
      "learning_rate": 1.1950324758760507e-05,
      "loss": 2.3941287994384766,
      "memory(GiB)": 77.56,
      "step": 90480,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 3.8766548134184484,
      "grad_norm": 7.793916702270508,
      "learning_rate": 1.1945959099246407e-05,
      "loss": 2.2018304824829102,
      "memory(GiB)": 77.56,
      "step": 90485,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 3.8768690287476972,
      "grad_norm": 7.034725666046143,
      "learning_rate": 1.1941594129116962e-05,
      "loss": 2.4903772354125975,
      "memory(GiB)": 77.56,
      "step": 90490,
      "token_acc": 0.5047318611987381,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 3.877083244076946,
      "grad_norm": 5.782181262969971,
      "learning_rate": 1.1937229848451237e-05,
      "loss": 2.2202796936035156,
      "memory(GiB)": 77.56,
      "step": 90495,
      "token_acc": 0.5233333333333333,
      "train_speed(iter/s)": 1.438035
    },
    {
      "epoch": 3.8772974594061953,
      "grad_norm": 4.895261764526367,
      "learning_rate": 1.1932866257328302e-05,
      "loss": 2.280509376525879,
      "memory(GiB)": 77.56,
      "step": 90500,
      "token_acc": 0.4983164983164983,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 3.8772974594061953,
      "eval_loss": 2.1616945266723633,
      "eval_runtime": 14.2274,
      "eval_samples_per_second": 7.029,
      "eval_steps_per_second": 7.029,
      "eval_token_acc": 0.4872881355932203,
      "step": 90500
    },
    {
      "epoch": 3.877511674735444,
      "grad_norm": 6.078415393829346,
      "learning_rate": 1.1928503355827192e-05,
      "loss": 2.3752138137817385,
      "memory(GiB)": 77.56,
      "step": 90505,
      "token_acc": 0.49950641658440276,
      "train_speed(iter/s)": 1.437681
    },
    {
      "epoch": 3.877725890064693,
      "grad_norm": 8.73322582244873,
      "learning_rate": 1.1924141144026969e-05,
      "loss": 2.322062873840332,
      "memory(GiB)": 77.56,
      "step": 90510,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.437668
    },
    {
      "epoch": 3.877940105393942,
      "grad_norm": 6.695369243621826,
      "learning_rate": 1.1919779622006632e-05,
      "loss": 2.640370750427246,
      "memory(GiB)": 77.56,
      "step": 90515,
      "token_acc": 0.4626334519572954,
      "train_speed(iter/s)": 1.437669
    },
    {
      "epoch": 3.878154320723191,
      "grad_norm": 4.382381916046143,
      "learning_rate": 1.1915418789845229e-05,
      "loss": 2.3129093170166017,
      "memory(GiB)": 77.56,
      "step": 90520,
      "token_acc": 0.46706586826347307,
      "train_speed(iter/s)": 1.437676
    },
    {
      "epoch": 3.87836853605244,
      "grad_norm": 6.578730583190918,
      "learning_rate": 1.1911058647621737e-05,
      "loss": 2.3238664627075196,
      "memory(GiB)": 77.56,
      "step": 90525,
      "token_acc": 0.5413793103448276,
      "train_speed(iter/s)": 1.437672
    },
    {
      "epoch": 3.878582751381689,
      "grad_norm": 6.443264484405518,
      "learning_rate": 1.1906699195415144e-05,
      "loss": 1.986534309387207,
      "memory(GiB)": 77.56,
      "step": 90530,
      "token_acc": 0.5461254612546126,
      "train_speed(iter/s)": 1.437662
    },
    {
      "epoch": 3.878796966710938,
      "grad_norm": 5.691235542297363,
      "learning_rate": 1.1902340433304431e-05,
      "loss": 2.3310657501220704,
      "memory(GiB)": 77.56,
      "step": 90535,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.43767
    },
    {
      "epoch": 3.8790111820401867,
      "grad_norm": 6.651580333709717,
      "learning_rate": 1.1897982361368548e-05,
      "loss": 2.007834053039551,
      "memory(GiB)": 77.56,
      "step": 90540,
      "token_acc": 0.5470383275261324,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 3.879225397369436,
      "grad_norm": 4.5281195640563965,
      "learning_rate": 1.1893624979686474e-05,
      "loss": 2.4487594604492187,
      "memory(GiB)": 77.56,
      "step": 90545,
      "token_acc": 0.4781420765027322,
      "train_speed(iter/s)": 1.437674
    },
    {
      "epoch": 3.8794396126986848,
      "grad_norm": 5.499750137329102,
      "learning_rate": 1.1889268288337124e-05,
      "loss": 2.4119199752807616,
      "memory(GiB)": 77.56,
      "step": 90550,
      "token_acc": 0.4783861671469741,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 3.8796538280279336,
      "grad_norm": 5.003000736236572,
      "learning_rate": 1.1884912287399436e-05,
      "loss": 2.289682960510254,
      "memory(GiB)": 77.56,
      "step": 90555,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.437682
    },
    {
      "epoch": 3.879868043357183,
      "grad_norm": 4.916240215301514,
      "learning_rate": 1.1880556976952312e-05,
      "loss": 2.5092018127441404,
      "memory(GiB)": 77.56,
      "step": 90560,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 3.8800822586864316,
      "grad_norm": 5.766725063323975,
      "learning_rate": 1.1876202357074645e-05,
      "loss": 2.514948844909668,
      "memory(GiB)": 77.56,
      "step": 90565,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 3.8802964740156805,
      "grad_norm": 5.398946285247803,
      "learning_rate": 1.187184842784535e-05,
      "loss": 2.502707290649414,
      "memory(GiB)": 77.56,
      "step": 90570,
      "token_acc": 0.46449704142011833,
      "train_speed(iter/s)": 1.437676
    },
    {
      "epoch": 3.8805106893449297,
      "grad_norm": 6.373363494873047,
      "learning_rate": 1.1867495189343286e-05,
      "loss": 2.2389551162719727,
      "memory(GiB)": 77.56,
      "step": 90575,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 3.8807249046741785,
      "grad_norm": 6.4101338386535645,
      "learning_rate": 1.1863142641647307e-05,
      "loss": 2.285857009887695,
      "memory(GiB)": 77.56,
      "step": 90580,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.437711
    },
    {
      "epoch": 3.8809391200034273,
      "grad_norm": 4.727756500244141,
      "learning_rate": 1.1858790784836282e-05,
      "loss": 2.316763687133789,
      "memory(GiB)": 77.56,
      "step": 90585,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 3.8811533353326766,
      "grad_norm": 5.3445820808410645,
      "learning_rate": 1.185443961898905e-05,
      "loss": 2.4759702682495117,
      "memory(GiB)": 77.56,
      "step": 90590,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.437706
    },
    {
      "epoch": 3.8813675506619254,
      "grad_norm": 6.157519340515137,
      "learning_rate": 1.1850089144184423e-05,
      "loss": 2.279096984863281,
      "memory(GiB)": 77.56,
      "step": 90595,
      "token_acc": 0.47586206896551725,
      "train_speed(iter/s)": 1.437723
    },
    {
      "epoch": 3.881581765991174,
      "grad_norm": 6.4327874183654785,
      "learning_rate": 1.18457393605012e-05,
      "loss": 2.179935646057129,
      "memory(GiB)": 77.56,
      "step": 90600,
      "token_acc": 0.5080645161290323,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.8817959813204235,
      "grad_norm": 6.15175724029541,
      "learning_rate": 1.184139026801822e-05,
      "loss": 2.4754737854003905,
      "memory(GiB)": 77.56,
      "step": 90605,
      "token_acc": 0.48846153846153845,
      "train_speed(iter/s)": 1.43773
    },
    {
      "epoch": 3.8820101966496723,
      "grad_norm": 5.153484344482422,
      "learning_rate": 1.1837041866814252e-05,
      "loss": 2.2230756759643553,
      "memory(GiB)": 77.56,
      "step": 90610,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437726
    },
    {
      "epoch": 3.882224411978921,
      "grad_norm": 9.239290237426758,
      "learning_rate": 1.1832694156968066e-05,
      "loss": 2.6743162155151365,
      "memory(GiB)": 77.56,
      "step": 90615,
      "token_acc": 0.4978723404255319,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 3.8824386273081704,
      "grad_norm": 6.394015789031982,
      "learning_rate": 1.1828347138558432e-05,
      "loss": 2.467215347290039,
      "memory(GiB)": 77.56,
      "step": 90620,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 3.882652842637419,
      "grad_norm": 6.586813449859619,
      "learning_rate": 1.1824000811664083e-05,
      "loss": 2.1649574279785155,
      "memory(GiB)": 77.56,
      "step": 90625,
      "token_acc": 0.4968944099378882,
      "train_speed(iter/s)": 1.437735
    },
    {
      "epoch": 3.882867057966668,
      "grad_norm": 5.818645000457764,
      "learning_rate": 1.1819655176363786e-05,
      "loss": 2.3574411392211916,
      "memory(GiB)": 77.56,
      "step": 90630,
      "token_acc": 0.5404411764705882,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 3.8830812732959172,
      "grad_norm": 8.16355037689209,
      "learning_rate": 1.1815310232736249e-05,
      "loss": 2.251815605163574,
      "memory(GiB)": 77.56,
      "step": 90635,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 3.883295488625166,
      "grad_norm": 6.806107997894287,
      "learning_rate": 1.1810965980860189e-05,
      "loss": 2.5406349182128904,
      "memory(GiB)": 77.56,
      "step": 90640,
      "token_acc": 0.4793650793650794,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 3.883509703954415,
      "grad_norm": 8.11182975769043,
      "learning_rate": 1.1806622420814306e-05,
      "loss": 2.552924919128418,
      "memory(GiB)": 77.56,
      "step": 90645,
      "token_acc": 0.5103448275862069,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.883723919283664,
      "grad_norm": 10.275968551635742,
      "learning_rate": 1.1802279552677282e-05,
      "loss": 2.2463611602783202,
      "memory(GiB)": 77.56,
      "step": 90650,
      "token_acc": 0.5120274914089347,
      "train_speed(iter/s)": 1.437773
    },
    {
      "epoch": 3.883938134612913,
      "grad_norm": 6.112062454223633,
      "learning_rate": 1.1797937376527784e-05,
      "loss": 2.1707592010498047,
      "memory(GiB)": 77.56,
      "step": 90655,
      "token_acc": 0.5350553505535055,
      "train_speed(iter/s)": 1.437761
    },
    {
      "epoch": 3.8841523499421617,
      "grad_norm": 5.287235260009766,
      "learning_rate": 1.1793595892444492e-05,
      "loss": 2.559587860107422,
      "memory(GiB)": 77.56,
      "step": 90660,
      "token_acc": 0.49107142857142855,
      "train_speed(iter/s)": 1.437764
    },
    {
      "epoch": 3.884366565271411,
      "grad_norm": 6.164012908935547,
      "learning_rate": 1.1789255100506057e-05,
      "loss": 2.5242151260375976,
      "memory(GiB)": 77.56,
      "step": 90665,
      "token_acc": 0.49146757679180886,
      "train_speed(iter/s)": 1.437764
    },
    {
      "epoch": 3.88458078060066,
      "grad_norm": 4.857330799102783,
      "learning_rate": 1.1784915000791114e-05,
      "loss": 2.275299072265625,
      "memory(GiB)": 77.56,
      "step": 90670,
      "token_acc": 0.5425531914893617,
      "train_speed(iter/s)": 1.437778
    },
    {
      "epoch": 3.8847949959299086,
      "grad_norm": 6.191040992736816,
      "learning_rate": 1.1780575593378284e-05,
      "loss": 2.5952877044677733,
      "memory(GiB)": 77.56,
      "step": 90675,
      "token_acc": 0.46923076923076923,
      "train_speed(iter/s)": 1.437777
    },
    {
      "epoch": 3.885009211259158,
      "grad_norm": 4.9005303382873535,
      "learning_rate": 1.177623687834618e-05,
      "loss": 2.3590728759765627,
      "memory(GiB)": 77.56,
      "step": 90680,
      "token_acc": 0.4954954954954955,
      "train_speed(iter/s)": 1.437765
    },
    {
      "epoch": 3.8852234265884067,
      "grad_norm": 6.755948066711426,
      "learning_rate": 1.1771898855773388e-05,
      "loss": 2.129837226867676,
      "memory(GiB)": 77.56,
      "step": 90685,
      "token_acc": 0.5787671232876712,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 3.8854376419176555,
      "grad_norm": 5.061296463012695,
      "learning_rate": 1.1767561525738525e-05,
      "loss": 1.975636863708496,
      "memory(GiB)": 77.56,
      "step": 90690,
      "token_acc": 0.5418326693227091,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 3.8856518572469048,
      "grad_norm": 5.949632167816162,
      "learning_rate": 1.1763224888320145e-05,
      "loss": 2.3912097930908205,
      "memory(GiB)": 77.56,
      "step": 90695,
      "token_acc": 0.47388059701492535,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 3.8858660725761536,
      "grad_norm": 5.824096202850342,
      "learning_rate": 1.1758888943596818e-05,
      "loss": 2.5480459213256834,
      "memory(GiB)": 77.56,
      "step": 90700,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.437781
    },
    {
      "epoch": 3.8860802879054024,
      "grad_norm": 6.116512775421143,
      "learning_rate": 1.1754553691647092e-05,
      "loss": 2.3607730865478516,
      "memory(GiB)": 77.56,
      "step": 90705,
      "token_acc": 0.5434083601286174,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.8862945032346516,
      "grad_norm": 6.583705902099609,
      "learning_rate": 1.1750219132549489e-05,
      "loss": 2.4126081466674805,
      "memory(GiB)": 77.56,
      "step": 90710,
      "token_acc": 0.44482758620689655,
      "train_speed(iter/s)": 1.437786
    },
    {
      "epoch": 3.8865087185639005,
      "grad_norm": 5.833676338195801,
      "learning_rate": 1.1745885266382561e-05,
      "loss": 2.072280693054199,
      "memory(GiB)": 77.56,
      "step": 90715,
      "token_acc": 0.5229681978798587,
      "train_speed(iter/s)": 1.437781
    },
    {
      "epoch": 3.8867229338931493,
      "grad_norm": 5.618956089019775,
      "learning_rate": 1.1741552093224805e-05,
      "loss": 2.498998260498047,
      "memory(GiB)": 77.56,
      "step": 90720,
      "token_acc": 0.45017182130584193,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.8869371492223985,
      "grad_norm": 5.718469619750977,
      "learning_rate": 1.1737219613154727e-05,
      "loss": 2.206536293029785,
      "memory(GiB)": 77.56,
      "step": 90725,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437787
    },
    {
      "epoch": 3.8871513645516473,
      "grad_norm": 5.852831840515137,
      "learning_rate": 1.1732887826250787e-05,
      "loss": 2.202661895751953,
      "memory(GiB)": 77.56,
      "step": 90730,
      "token_acc": 0.5168539325842697,
      "train_speed(iter/s)": 1.437808
    },
    {
      "epoch": 3.887365579880896,
      "grad_norm": 7.852920055389404,
      "learning_rate": 1.1728556732591501e-05,
      "loss": 2.517822265625,
      "memory(GiB)": 77.56,
      "step": 90735,
      "token_acc": 0.528957528957529,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 3.8875797952101454,
      "grad_norm": 6.729727268218994,
      "learning_rate": 1.172422633225531e-05,
      "loss": 2.0631500244140626,
      "memory(GiB)": 77.56,
      "step": 90740,
      "token_acc": 0.5487804878048781,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 3.887794010539394,
      "grad_norm": 5.046313285827637,
      "learning_rate": 1.1719896625320654e-05,
      "loss": 2.3603225708007813,
      "memory(GiB)": 77.56,
      "step": 90745,
      "token_acc": 0.5032258064516129,
      "train_speed(iter/s)": 1.437836
    },
    {
      "epoch": 3.888008225868643,
      "grad_norm": 6.943776607513428,
      "learning_rate": 1.1715567611865991e-05,
      "loss": 2.4751752853393554,
      "memory(GiB)": 77.56,
      "step": 90750,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 3.8882224411978923,
      "grad_norm": 5.720273494720459,
      "learning_rate": 1.1711239291969733e-05,
      "loss": 2.7610435485839844,
      "memory(GiB)": 77.56,
      "step": 90755,
      "token_acc": 0.4967532467532468,
      "train_speed(iter/s)": 1.437816
    },
    {
      "epoch": 3.888436656527141,
      "grad_norm": 5.783408164978027,
      "learning_rate": 1.1706911665710296e-05,
      "loss": 2.680044937133789,
      "memory(GiB)": 77.56,
      "step": 90760,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.88865087185639,
      "grad_norm": 6.125732898712158,
      "learning_rate": 1.1702584733166073e-05,
      "loss": 2.442182922363281,
      "memory(GiB)": 77.56,
      "step": 90765,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.888865087185639,
      "grad_norm": 6.588894367218018,
      "learning_rate": 1.1698258494415443e-05,
      "loss": 2.413641357421875,
      "memory(GiB)": 77.56,
      "step": 90770,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.889079302514888,
      "grad_norm": 7.6249895095825195,
      "learning_rate": 1.1693932949536801e-05,
      "loss": 2.6100582122802733,
      "memory(GiB)": 77.56,
      "step": 90775,
      "token_acc": 0.43956043956043955,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.889293517844137,
      "grad_norm": 5.18500280380249,
      "learning_rate": 1.1689608098608495e-05,
      "loss": 2.8090087890625,
      "memory(GiB)": 77.56,
      "step": 90780,
      "token_acc": 0.4809688581314879,
      "train_speed(iter/s)": 1.43786
    },
    {
      "epoch": 3.889507733173386,
      "grad_norm": 8.355817794799805,
      "learning_rate": 1.168528394170888e-05,
      "loss": 2.4034042358398438,
      "memory(GiB)": 77.56,
      "step": 90785,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 3.889721948502635,
      "grad_norm": 5.919632434844971,
      "learning_rate": 1.1680960478916292e-05,
      "loss": 2.3702533721923826,
      "memory(GiB)": 77.56,
      "step": 90790,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.8899361638318837,
      "grad_norm": 7.269247531890869,
      "learning_rate": 1.1676637710309047e-05,
      "loss": 2.7270795822143556,
      "memory(GiB)": 77.56,
      "step": 90795,
      "token_acc": 0.4540229885057471,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 3.890150379161133,
      "grad_norm": 6.30020809173584,
      "learning_rate": 1.1672315635965447e-05,
      "loss": 2.520612335205078,
      "memory(GiB)": 77.56,
      "step": 90800,
      "token_acc": 0.47840531561461797,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.8903645944903817,
      "grad_norm": 7.151754379272461,
      "learning_rate": 1.1667994255963805e-05,
      "loss": 2.196531867980957,
      "memory(GiB)": 77.56,
      "step": 90805,
      "token_acc": 0.5354609929078015,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 3.8905788098196306,
      "grad_norm": 5.397918701171875,
      "learning_rate": 1.1663673570382416e-05,
      "loss": 2.2160144805908204,
      "memory(GiB)": 77.56,
      "step": 90810,
      "token_acc": 0.5105105105105106,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 3.89079302514888,
      "grad_norm": 6.126784801483154,
      "learning_rate": 1.1659353579299543e-05,
      "loss": 2.443135452270508,
      "memory(GiB)": 77.56,
      "step": 90815,
      "token_acc": 0.5065359477124183,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 3.8910072404781286,
      "grad_norm": 7.230842590332031,
      "learning_rate": 1.1655034282793448e-05,
      "loss": 2.0986690521240234,
      "memory(GiB)": 77.56,
      "step": 90820,
      "token_acc": 0.5157894736842106,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.8912214558073774,
      "grad_norm": 5.642878532409668,
      "learning_rate": 1.1650715680942381e-05,
      "loss": 2.4565019607543945,
      "memory(GiB)": 77.56,
      "step": 90825,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 3.8914356711366267,
      "grad_norm": 5.745751857757568,
      "learning_rate": 1.1646397773824553e-05,
      "loss": 2.2658226013183596,
      "memory(GiB)": 77.56,
      "step": 90830,
      "token_acc": 0.5300751879699248,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 3.8916498864658755,
      "grad_norm": 5.3980512619018555,
      "learning_rate": 1.1642080561518226e-05,
      "loss": 2.2125831604003907,
      "memory(GiB)": 77.56,
      "step": 90835,
      "token_acc": 0.5086705202312138,
      "train_speed(iter/s)": 1.437857
    },
    {
      "epoch": 3.8918641017951243,
      "grad_norm": 6.2098517417907715,
      "learning_rate": 1.163776404410159e-05,
      "loss": 2.423111152648926,
      "memory(GiB)": 77.56,
      "step": 90840,
      "token_acc": 0.4944649446494465,
      "train_speed(iter/s)": 1.437852
    },
    {
      "epoch": 3.8920783171243736,
      "grad_norm": 6.400033473968506,
      "learning_rate": 1.1633448221652848e-05,
      "loss": 1.8792903900146485,
      "memory(GiB)": 77.56,
      "step": 90845,
      "token_acc": 0.591743119266055,
      "train_speed(iter/s)": 1.437853
    },
    {
      "epoch": 3.8922925324536224,
      "grad_norm": 6.435670852661133,
      "learning_rate": 1.1629133094250183e-05,
      "loss": 2.0793975830078124,
      "memory(GiB)": 77.56,
      "step": 90850,
      "token_acc": 0.5631399317406144,
      "train_speed(iter/s)": 1.437867
    },
    {
      "epoch": 3.892506747782871,
      "grad_norm": 5.783316135406494,
      "learning_rate": 1.1624818661971747e-05,
      "loss": 2.5071504592895506,
      "memory(GiB)": 77.56,
      "step": 90855,
      "token_acc": 0.5216049382716049,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 3.8927209631121205,
      "grad_norm": 5.509244918823242,
      "learning_rate": 1.1620504924895737e-05,
      "loss": 2.2704851150512697,
      "memory(GiB)": 77.56,
      "step": 90860,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.8929351784413693,
      "grad_norm": 5.796581745147705,
      "learning_rate": 1.1616191883100285e-05,
      "loss": 2.470877265930176,
      "memory(GiB)": 77.56,
      "step": 90865,
      "token_acc": 0.46355685131195334,
      "train_speed(iter/s)": 1.437888
    },
    {
      "epoch": 3.893149393770618,
      "grad_norm": 6.25365686416626,
      "learning_rate": 1.1611879536663523e-05,
      "loss": 2.378495979309082,
      "memory(GiB)": 77.56,
      "step": 90870,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 3.8933636090998673,
      "grad_norm": 6.276653289794922,
      "learning_rate": 1.1607567885663555e-05,
      "loss": 2.371576690673828,
      "memory(GiB)": 77.56,
      "step": 90875,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437901
    },
    {
      "epoch": 3.893577824429116,
      "grad_norm": 5.147952556610107,
      "learning_rate": 1.160325693017853e-05,
      "loss": 2.3934024810791015,
      "memory(GiB)": 77.56,
      "step": 90880,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 3.893792039758365,
      "grad_norm": 6.9699578285217285,
      "learning_rate": 1.1598946670286525e-05,
      "loss": 2.1941316604614256,
      "memory(GiB)": 77.56,
      "step": 90885,
      "token_acc": 0.54421768707483,
      "train_speed(iter/s)": 1.437915
    },
    {
      "epoch": 3.894006255087614,
      "grad_norm": 4.8469390869140625,
      "learning_rate": 1.1594637106065608e-05,
      "loss": 2.3624467849731445,
      "memory(GiB)": 77.56,
      "step": 90890,
      "token_acc": 0.518796992481203,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 3.894220470416863,
      "grad_norm": 5.040021896362305,
      "learning_rate": 1.1590328237593878e-05,
      "loss": 2.3942983627319334,
      "memory(GiB)": 77.56,
      "step": 90895,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.894434685746112,
      "grad_norm": 6.077275276184082,
      "learning_rate": 1.1586020064949387e-05,
      "loss": 2.603710174560547,
      "memory(GiB)": 77.56,
      "step": 90900,
      "token_acc": 0.47878787878787876,
      "train_speed(iter/s)": 1.437942
    },
    {
      "epoch": 3.894648901075361,
      "grad_norm": 7.9805498123168945,
      "learning_rate": 1.1581712588210181e-05,
      "loss": 2.2039958953857424,
      "memory(GiB)": 77.56,
      "step": 90905,
      "token_acc": 0.5404411764705882,
      "train_speed(iter/s)": 1.437964
    },
    {
      "epoch": 3.89486311640461,
      "grad_norm": 5.799496650695801,
      "learning_rate": 1.1577405807454283e-05,
      "loss": 2.0657562255859374,
      "memory(GiB)": 77.56,
      "step": 90910,
      "token_acc": 0.5458015267175572,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.8950773317338587,
      "grad_norm": 6.915517330169678,
      "learning_rate": 1.1573099722759712e-05,
      "loss": 2.485106658935547,
      "memory(GiB)": 77.56,
      "step": 90915,
      "token_acc": 0.49469964664310956,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 3.895291547063108,
      "grad_norm": 7.066059589385986,
      "learning_rate": 1.1568794334204502e-05,
      "loss": 2.0547019958496096,
      "memory(GiB)": 77.56,
      "step": 90920,
      "token_acc": 0.5720930232558139,
      "train_speed(iter/s)": 1.437993
    },
    {
      "epoch": 3.895505762392357,
      "grad_norm": 5.3980393409729,
      "learning_rate": 1.1564489641866633e-05,
      "loss": 2.7890256881713866,
      "memory(GiB)": 77.56,
      "step": 90925,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 3.8957199777216056,
      "grad_norm": 5.78582239151001,
      "learning_rate": 1.156018564582409e-05,
      "loss": 2.256192207336426,
      "memory(GiB)": 77.56,
      "step": 90930,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 3.895934193050855,
      "grad_norm": 6.013241767883301,
      "learning_rate": 1.1555882346154845e-05,
      "loss": 2.1658794403076174,
      "memory(GiB)": 77.56,
      "step": 90935,
      "token_acc": 0.5761316872427984,
      "train_speed(iter/s)": 1.438035
    },
    {
      "epoch": 3.8961484083801037,
      "grad_norm": 6.3571906089782715,
      "learning_rate": 1.1551579742936852e-05,
      "loss": 2.250625228881836,
      "memory(GiB)": 77.56,
      "step": 90940,
      "token_acc": 0.5018867924528302,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 3.8963626237093525,
      "grad_norm": 4.944669723510742,
      "learning_rate": 1.1547277836248043e-05,
      "loss": 2.237442398071289,
      "memory(GiB)": 77.56,
      "step": 90945,
      "token_acc": 0.5398550724637681,
      "train_speed(iter/s)": 1.438053
    },
    {
      "epoch": 3.8965768390386017,
      "grad_norm": 8.360963821411133,
      "learning_rate": 1.1542976626166374e-05,
      "loss": 2.467558670043945,
      "memory(GiB)": 77.56,
      "step": 90950,
      "token_acc": 0.5190839694656488,
      "train_speed(iter/s)": 1.438058
    },
    {
      "epoch": 3.8967910543678506,
      "grad_norm": 6.620720863342285,
      "learning_rate": 1.153867611276977e-05,
      "loss": 2.363943099975586,
      "memory(GiB)": 77.56,
      "step": 90955,
      "token_acc": 0.5241635687732342,
      "train_speed(iter/s)": 1.43806
    },
    {
      "epoch": 3.8970052696970994,
      "grad_norm": 6.948920726776123,
      "learning_rate": 1.1534376296136124e-05,
      "loss": 2.6090906143188475,
      "memory(GiB)": 77.56,
      "step": 90960,
      "token_acc": 0.48616600790513836,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 3.8972194850263486,
      "grad_norm": 6.90181303024292,
      "learning_rate": 1.153007717634334e-05,
      "loss": 2.002796745300293,
      "memory(GiB)": 77.56,
      "step": 90965,
      "token_acc": 0.5675675675675675,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 3.8974337003555974,
      "grad_norm": 8.537093162536621,
      "learning_rate": 1.1525778753469297e-05,
      "loss": 2.2156164169311525,
      "memory(GiB)": 77.56,
      "step": 90970,
      "token_acc": 0.5229681978798587,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 3.8976479156848463,
      "grad_norm": 5.488020420074463,
      "learning_rate": 1.1521481027591847e-05,
      "loss": 2.207258605957031,
      "memory(GiB)": 77.56,
      "step": 90975,
      "token_acc": 0.5037037037037037,
      "train_speed(iter/s)": 1.438042
    },
    {
      "epoch": 3.8978621310140955,
      "grad_norm": 5.44802188873291,
      "learning_rate": 1.1517183998788877e-05,
      "loss": 2.4842533111572265,
      "memory(GiB)": 77.56,
      "step": 90980,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 3.8980763463433443,
      "grad_norm": 7.082776069641113,
      "learning_rate": 1.1512887667138217e-05,
      "loss": 2.6597930908203127,
      "memory(GiB)": 77.56,
      "step": 90985,
      "token_acc": 0.4798657718120805,
      "train_speed(iter/s)": 1.438049
    },
    {
      "epoch": 3.898290561672593,
      "grad_norm": 6.406974792480469,
      "learning_rate": 1.1508592032717701e-05,
      "loss": 2.706146240234375,
      "memory(GiB)": 77.56,
      "step": 90990,
      "token_acc": 0.4409937888198758,
      "train_speed(iter/s)": 1.438055
    },
    {
      "epoch": 3.8985047770018424,
      "grad_norm": 5.782268047332764,
      "learning_rate": 1.1504297095605154e-05,
      "loss": 2.374405288696289,
      "memory(GiB)": 77.56,
      "step": 90995,
      "token_acc": 0.5223367697594502,
      "train_speed(iter/s)": 1.438071
    },
    {
      "epoch": 3.898718992331091,
      "grad_norm": 8.141237258911133,
      "learning_rate": 1.1500002855878362e-05,
      "loss": 2.6752086639404298,
      "memory(GiB)": 77.56,
      "step": 91000,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 3.898718992331091,
      "eval_loss": 2.138394594192505,
      "eval_runtime": 14.4932,
      "eval_samples_per_second": 6.9,
      "eval_steps_per_second": 6.9,
      "eval_token_acc": 0.46196403872752423,
      "step": 91000
    },
    {
      "epoch": 3.89893320766034,
      "grad_norm": 4.859127044677734,
      "learning_rate": 1.1495709313615143e-05,
      "loss": 2.604720687866211,
      "memory(GiB)": 77.56,
      "step": 91005,
      "token_acc": 0.46509341199606685,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 3.8991474229895893,
      "grad_norm": 6.947793006896973,
      "learning_rate": 1.1491416468893274e-05,
      "loss": 2.5877716064453127,
      "memory(GiB)": 77.56,
      "step": 91010,
      "token_acc": 0.471875,
      "train_speed(iter/s)": 1.43773
    },
    {
      "epoch": 3.899361638318838,
      "grad_norm": 7.862221717834473,
      "learning_rate": 1.1487124321790515e-05,
      "loss": 2.2432907104492186,
      "memory(GiB)": 77.56,
      "step": 91015,
      "token_acc": 0.5390070921985816,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 3.899575853648087,
      "grad_norm": 7.062439918518066,
      "learning_rate": 1.148283287238463e-05,
      "loss": 2.412496566772461,
      "memory(GiB)": 77.56,
      "step": 91020,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 3.899790068977336,
      "grad_norm": 6.912678241729736,
      "learning_rate": 1.1478542120753344e-05,
      "loss": 2.351667785644531,
      "memory(GiB)": 77.56,
      "step": 91025,
      "token_acc": 0.49814126394052044,
      "train_speed(iter/s)": 1.437781
    },
    {
      "epoch": 3.900004284306585,
      "grad_norm": 5.951408863067627,
      "learning_rate": 1.1474252066974422e-05,
      "loss": 2.1137380599975586,
      "memory(GiB)": 77.56,
      "step": 91030,
      "token_acc": 0.558303886925795,
      "train_speed(iter/s)": 1.437777
    },
    {
      "epoch": 3.900218499635834,
      "grad_norm": 7.651566505432129,
      "learning_rate": 1.1469962711125548e-05,
      "loss": 2.350306510925293,
      "memory(GiB)": 77.56,
      "step": 91035,
      "token_acc": 0.5070422535211268,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 3.900432714965083,
      "grad_norm": 6.07174015045166,
      "learning_rate": 1.1465674053284452e-05,
      "loss": 2.4441177368164064,
      "memory(GiB)": 77.56,
      "step": 91040,
      "token_acc": 0.4965034965034965,
      "train_speed(iter/s)": 1.437812
    },
    {
      "epoch": 3.900646930294332,
      "grad_norm": 8.041930198669434,
      "learning_rate": 1.1461386093528826e-05,
      "loss": 2.739452362060547,
      "memory(GiB)": 77.56,
      "step": 91045,
      "token_acc": 0.4337748344370861,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 3.9008611456235807,
      "grad_norm": 6.146678447723389,
      "learning_rate": 1.1457098831936342e-05,
      "loss": 2.389677810668945,
      "memory(GiB)": 77.56,
      "step": 91050,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 3.90107536095283,
      "grad_norm": 5.0948686599731445,
      "learning_rate": 1.1452812268584667e-05,
      "loss": 2.227981758117676,
      "memory(GiB)": 77.56,
      "step": 91055,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 3.9012895762820787,
      "grad_norm": 5.616827011108398,
      "learning_rate": 1.1448526403551441e-05,
      "loss": 2.4123701095581054,
      "memory(GiB)": 77.56,
      "step": 91060,
      "token_acc": 0.5112540192926045,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.9015037916113275,
      "grad_norm": 5.040546894073486,
      "learning_rate": 1.1444241236914343e-05,
      "loss": 2.4961627960205077,
      "memory(GiB)": 77.56,
      "step": 91065,
      "token_acc": 0.4969512195121951,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.901718006940577,
      "grad_norm": 5.902012825012207,
      "learning_rate": 1.143995676875098e-05,
      "loss": 2.435586357116699,
      "memory(GiB)": 77.56,
      "step": 91070,
      "token_acc": 0.5278810408921933,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 3.9019322222698256,
      "grad_norm": 5.462906360626221,
      "learning_rate": 1.1435672999138975e-05,
      "loss": 2.4297624588012696,
      "memory(GiB)": 77.56,
      "step": 91075,
      "token_acc": 0.5132450331125827,
      "train_speed(iter/s)": 1.437892
    },
    {
      "epoch": 3.9021464375990744,
      "grad_norm": 5.44817590713501,
      "learning_rate": 1.1431389928155922e-05,
      "loss": 2.2389373779296875,
      "memory(GiB)": 77.56,
      "step": 91080,
      "token_acc": 0.5409836065573771,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 3.9023606529283237,
      "grad_norm": 6.258960723876953,
      "learning_rate": 1.1427107555879412e-05,
      "loss": 2.303700637817383,
      "memory(GiB)": 77.56,
      "step": 91085,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 3.9025748682575725,
      "grad_norm": 4.524411201477051,
      "learning_rate": 1.142282588238705e-05,
      "loss": 2.0361377716064455,
      "memory(GiB)": 77.56,
      "step": 91090,
      "token_acc": 0.5348837209302325,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 3.9027890835868213,
      "grad_norm": 5.329421043395996,
      "learning_rate": 1.1418544907756368e-05,
      "loss": 2.173225975036621,
      "memory(GiB)": 77.56,
      "step": 91095,
      "token_acc": 0.5286195286195287,
      "train_speed(iter/s)": 1.437907
    },
    {
      "epoch": 3.9030032989160706,
      "grad_norm": 6.68475866317749,
      "learning_rate": 1.1414264632064952e-05,
      "loss": 2.1077871322631836,
      "memory(GiB)": 77.56,
      "step": 91100,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 3.9032175142453194,
      "grad_norm": 5.613337993621826,
      "learning_rate": 1.1409985055390332e-05,
      "loss": 2.537759208679199,
      "memory(GiB)": 77.56,
      "step": 91105,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.437924
    },
    {
      "epoch": 3.903431729574568,
      "grad_norm": 5.314393043518066,
      "learning_rate": 1.1405706177810027e-05,
      "loss": 2.444041633605957,
      "memory(GiB)": 77.56,
      "step": 91110,
      "token_acc": 0.49146757679180886,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 3.9036459449038174,
      "grad_norm": 6.4607744216918945,
      "learning_rate": 1.1401427999401565e-05,
      "loss": 2.474997901916504,
      "memory(GiB)": 77.56,
      "step": 91115,
      "token_acc": 0.45325779036827196,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 3.9038601602330663,
      "grad_norm": 5.1067585945129395,
      "learning_rate": 1.1397150520242422e-05,
      "loss": 2.2475372314453126,
      "memory(GiB)": 77.56,
      "step": 91120,
      "token_acc": 0.5034722222222222,
      "train_speed(iter/s)": 1.437957
    },
    {
      "epoch": 3.904074375562315,
      "grad_norm": 6.846627712249756,
      "learning_rate": 1.1392873740410132e-05,
      "loss": 2.4637718200683594,
      "memory(GiB)": 77.56,
      "step": 91125,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 3.9042885908915643,
      "grad_norm": 4.901951789855957,
      "learning_rate": 1.1388597659982148e-05,
      "loss": 2.3763832092285155,
      "memory(GiB)": 77.56,
      "step": 91130,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.437969
    },
    {
      "epoch": 3.904502806220813,
      "grad_norm": 5.6291728019714355,
      "learning_rate": 1.138432227903593e-05,
      "loss": 2.178856086730957,
      "memory(GiB)": 77.56,
      "step": 91135,
      "token_acc": 0.5525291828793775,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 3.904717021550062,
      "grad_norm": 5.030359745025635,
      "learning_rate": 1.1380047597648946e-05,
      "loss": 1.9573993682861328,
      "memory(GiB)": 77.56,
      "step": 91140,
      "token_acc": 0.5483870967741935,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 3.904931236879311,
      "grad_norm": 6.37743616104126,
      "learning_rate": 1.1375773615898617e-05,
      "loss": 2.0662460327148438,
      "memory(GiB)": 77.56,
      "step": 91145,
      "token_acc": 0.5060240963855421,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.90514545220856,
      "grad_norm": 6.363149166107178,
      "learning_rate": 1.137150033386239e-05,
      "loss": 2.193059539794922,
      "memory(GiB)": 77.56,
      "step": 91150,
      "token_acc": 0.510548523206751,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 3.905359667537809,
      "grad_norm": 4.6900434494018555,
      "learning_rate": 1.1367227751617671e-05,
      "loss": 2.1674001693725584,
      "memory(GiB)": 77.56,
      "step": 91155,
      "token_acc": 0.546031746031746,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 3.905573882867058,
      "grad_norm": 9.145657539367676,
      "learning_rate": 1.1362955869241865e-05,
      "loss": 2.56341609954834,
      "memory(GiB)": 77.56,
      "step": 91160,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 3.905788098196307,
      "grad_norm": 5.5173845291137695,
      "learning_rate": 1.1358684686812354e-05,
      "loss": 2.37689208984375,
      "memory(GiB)": 77.56,
      "step": 91165,
      "token_acc": 0.504,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 3.9060023135255557,
      "grad_norm": 6.67338752746582,
      "learning_rate": 1.1354414204406505e-05,
      "loss": 2.1870487213134764,
      "memory(GiB)": 77.56,
      "step": 91170,
      "token_acc": 0.5338645418326693,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 3.906216528854805,
      "grad_norm": 5.9065656661987305,
      "learning_rate": 1.135014442210171e-05,
      "loss": 2.274699020385742,
      "memory(GiB)": 77.56,
      "step": 91175,
      "token_acc": 0.5146579804560261,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 3.906430744184054,
      "grad_norm": 7.112194061279297,
      "learning_rate": 1.1345875339975292e-05,
      "loss": 2.0654006958007813,
      "memory(GiB)": 77.56,
      "step": 91180,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 3.9066449595133026,
      "grad_norm": 9.399968147277832,
      "learning_rate": 1.1341606958104616e-05,
      "loss": 2.1300832748413088,
      "memory(GiB)": 77.56,
      "step": 91185,
      "token_acc": 0.5107142857142857,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.906859174842552,
      "grad_norm": 5.284914970397949,
      "learning_rate": 1.1337339276566993e-05,
      "loss": 2.343291473388672,
      "memory(GiB)": 77.56,
      "step": 91190,
      "token_acc": 0.5097402597402597,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 3.9070733901718007,
      "grad_norm": 5.782185077667236,
      "learning_rate": 1.1333072295439739e-05,
      "loss": 2.6387989044189455,
      "memory(GiB)": 77.56,
      "step": 91195,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 3.9072876055010495,
      "grad_norm": 6.247674942016602,
      "learning_rate": 1.1328806014800158e-05,
      "loss": 2.450332260131836,
      "memory(GiB)": 77.56,
      "step": 91200,
      "token_acc": 0.5131578947368421,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 3.9075018208302987,
      "grad_norm": 5.871550559997559,
      "learning_rate": 1.132454043472551e-05,
      "loss": 2.271565818786621,
      "memory(GiB)": 77.56,
      "step": 91205,
      "token_acc": 0.476027397260274,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 3.9077160361595475,
      "grad_norm": 4.952987194061279,
      "learning_rate": 1.1320275555293113e-05,
      "loss": 2.229094314575195,
      "memory(GiB)": 77.56,
      "step": 91210,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.437967
    },
    {
      "epoch": 3.9079302514887964,
      "grad_norm": 6.041511535644531,
      "learning_rate": 1.1316011376580204e-05,
      "loss": 2.387455177307129,
      "memory(GiB)": 77.56,
      "step": 91215,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 3.9081444668180456,
      "grad_norm": 5.558559417724609,
      "learning_rate": 1.1311747898664038e-05,
      "loss": 2.146357536315918,
      "memory(GiB)": 77.56,
      "step": 91220,
      "token_acc": 0.503125,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 3.9083586821472944,
      "grad_norm": 6.989698886871338,
      "learning_rate": 1.1307485121621858e-05,
      "loss": 2.5166481018066404,
      "memory(GiB)": 77.56,
      "step": 91225,
      "token_acc": 0.47653429602888087,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.9085728974765432,
      "grad_norm": 5.662220001220703,
      "learning_rate": 1.1303223045530859e-05,
      "loss": 2.1686641693115236,
      "memory(GiB)": 77.56,
      "step": 91230,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 3.9087871128057925,
      "grad_norm": 5.485433101654053,
      "learning_rate": 1.1298961670468294e-05,
      "loss": 2.0881086349487306,
      "memory(GiB)": 77.56,
      "step": 91235,
      "token_acc": 0.5036231884057971,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 3.9090013281350413,
      "grad_norm": 5.765471935272217,
      "learning_rate": 1.1294700996511342e-05,
      "loss": 2.393413543701172,
      "memory(GiB)": 77.56,
      "step": 91240,
      "token_acc": 0.5106382978723404,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.90921554346429,
      "grad_norm": 7.634309768676758,
      "learning_rate": 1.1290441023737175e-05,
      "loss": 2.218937301635742,
      "memory(GiB)": 77.56,
      "step": 91245,
      "token_acc": 0.5186440677966102,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 3.9094297587935394,
      "grad_norm": 5.171670436859131,
      "learning_rate": 1.1286181752222996e-05,
      "loss": 2.226266860961914,
      "memory(GiB)": 77.56,
      "step": 91250,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 3.909643974122788,
      "grad_norm": 4.843559265136719,
      "learning_rate": 1.128192318204595e-05,
      "loss": 2.491107177734375,
      "memory(GiB)": 77.56,
      "step": 91255,
      "token_acc": 0.459214501510574,
      "train_speed(iter/s)": 1.437973
    },
    {
      "epoch": 3.909858189452037,
      "grad_norm": 7.1399149894714355,
      "learning_rate": 1.1277665313283187e-05,
      "loss": 2.610714149475098,
      "memory(GiB)": 77.56,
      "step": 91260,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 3.9100724047812863,
      "grad_norm": 6.221587657928467,
      "learning_rate": 1.1273408146011827e-05,
      "loss": 2.4690521240234373,
      "memory(GiB)": 77.56,
      "step": 91265,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.437987
    },
    {
      "epoch": 3.910286620110535,
      "grad_norm": 6.32486629486084,
      "learning_rate": 1.1269151680309021e-05,
      "loss": 2.415798568725586,
      "memory(GiB)": 77.56,
      "step": 91270,
      "token_acc": 0.46788990825688076,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 3.910500835439784,
      "grad_norm": 7.332726001739502,
      "learning_rate": 1.126489591625186e-05,
      "loss": 2.42496337890625,
      "memory(GiB)": 77.56,
      "step": 91275,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 3.910715050769033,
      "grad_norm": 5.7802205085754395,
      "learning_rate": 1.1260640853917453e-05,
      "loss": 2.511993980407715,
      "memory(GiB)": 77.56,
      "step": 91280,
      "token_acc": 0.4842105263157895,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 3.910929266098282,
      "grad_norm": 5.3685832023620605,
      "learning_rate": 1.1256386493382876e-05,
      "loss": 2.085711860656738,
      "memory(GiB)": 77.56,
      "step": 91285,
      "token_acc": 0.5077519379844961,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 3.9111434814275308,
      "grad_norm": 5.690767765045166,
      "learning_rate": 1.1252132834725187e-05,
      "loss": 2.3168817520141602,
      "memory(GiB)": 77.56,
      "step": 91290,
      "token_acc": 0.5285714285714286,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 3.91135769675678,
      "grad_norm": 6.205465316772461,
      "learning_rate": 1.1247879878021472e-05,
      "loss": 2.419499969482422,
      "memory(GiB)": 77.56,
      "step": 91295,
      "token_acc": 0.5162337662337663,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 3.911571912086029,
      "grad_norm": 5.981412410736084,
      "learning_rate": 1.1243627623348769e-05,
      "loss": 2.463705635070801,
      "memory(GiB)": 77.56,
      "step": 91300,
      "token_acc": 0.5028409090909091,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 3.9117861274152776,
      "grad_norm": 6.188590049743652,
      "learning_rate": 1.1239376070784108e-05,
      "loss": 2.2529827117919923,
      "memory(GiB)": 77.56,
      "step": 91305,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.43803
    },
    {
      "epoch": 3.912000342744527,
      "grad_norm": 5.830347537994385,
      "learning_rate": 1.1235125220404507e-05,
      "loss": 2.4111351013183593,
      "memory(GiB)": 77.56,
      "step": 91310,
      "token_acc": 0.5080645161290323,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 3.9122145580737757,
      "grad_norm": 7.293909549713135,
      "learning_rate": 1.1230875072286979e-05,
      "loss": 2.2350406646728516,
      "memory(GiB)": 77.56,
      "step": 91315,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 3.9124287734030245,
      "grad_norm": 7.565934181213379,
      "learning_rate": 1.1226625626508502e-05,
      "loss": 2.6229209899902344,
      "memory(GiB)": 77.56,
      "step": 91320,
      "token_acc": 0.4280701754385965,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 3.912642988732274,
      "grad_norm": 7.400323867797852,
      "learning_rate": 1.1222376883146079e-05,
      "loss": 2.283451461791992,
      "memory(GiB)": 77.56,
      "step": 91325,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 3.9128572040615226,
      "grad_norm": 5.257184028625488,
      "learning_rate": 1.1218128842276688e-05,
      "loss": 2.4757291793823244,
      "memory(GiB)": 77.56,
      "step": 91330,
      "token_acc": 0.4444444444444444,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 3.9130714193907714,
      "grad_norm": 6.685637950897217,
      "learning_rate": 1.121388150397727e-05,
      "loss": 2.557989501953125,
      "memory(GiB)": 77.56,
      "step": 91335,
      "token_acc": 0.4774193548387097,
      "train_speed(iter/s)": 1.43809
    },
    {
      "epoch": 3.9132856347200207,
      "grad_norm": 5.636646747589111,
      "learning_rate": 1.120963486832477e-05,
      "loss": 2.63091983795166,
      "memory(GiB)": 77.56,
      "step": 91340,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 3.9134998500492695,
      "grad_norm": 10.529409408569336,
      "learning_rate": 1.1205388935396127e-05,
      "loss": 2.3805414199829102,
      "memory(GiB)": 77.56,
      "step": 91345,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 3.9137140653785183,
      "grad_norm": 7.061811447143555,
      "learning_rate": 1.120114370526824e-05,
      "loss": 2.4556888580322265,
      "memory(GiB)": 77.56,
      "step": 91350,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 3.9139282807077675,
      "grad_norm": 6.291705131530762,
      "learning_rate": 1.119689917801805e-05,
      "loss": 2.3719932556152346,
      "memory(GiB)": 77.56,
      "step": 91355,
      "token_acc": 0.49049429657794674,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 3.9141424960370164,
      "grad_norm": 5.4585280418396,
      "learning_rate": 1.1192655353722425e-05,
      "loss": 2.46896915435791,
      "memory(GiB)": 77.56,
      "step": 91360,
      "token_acc": 0.47865853658536583,
      "train_speed(iter/s)": 1.438106
    },
    {
      "epoch": 3.914356711366265,
      "grad_norm": 5.959686279296875,
      "learning_rate": 1.1188412232458256e-05,
      "loss": 2.0860803604125975,
      "memory(GiB)": 77.56,
      "step": 91365,
      "token_acc": 0.5576923076923077,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 3.9145709266955144,
      "grad_norm": 6.7772698402404785,
      "learning_rate": 1.1184169814302409e-05,
      "loss": 1.9571212768554687,
      "memory(GiB)": 77.56,
      "step": 91370,
      "token_acc": 0.576271186440678,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 3.9147851420247632,
      "grad_norm": 5.908357620239258,
      "learning_rate": 1.117992809933172e-05,
      "loss": 2.389080619812012,
      "memory(GiB)": 77.56,
      "step": 91375,
      "token_acc": 0.5202952029520295,
      "train_speed(iter/s)": 1.438099
    },
    {
      "epoch": 3.914999357354012,
      "grad_norm": 5.820104598999023,
      "learning_rate": 1.1175687087623066e-05,
      "loss": 2.241306686401367,
      "memory(GiB)": 77.56,
      "step": 91380,
      "token_acc": 0.5448504983388704,
      "train_speed(iter/s)": 1.438097
    },
    {
      "epoch": 3.9152135726832613,
      "grad_norm": 4.487586498260498,
      "learning_rate": 1.1171446779253258e-05,
      "loss": 2.4218076705932616,
      "memory(GiB)": 77.56,
      "step": 91385,
      "token_acc": 0.5049833887043189,
      "train_speed(iter/s)": 1.438117
    },
    {
      "epoch": 3.91542778801251,
      "grad_norm": 5.582045078277588,
      "learning_rate": 1.116720717429912e-05,
      "loss": 2.506919097900391,
      "memory(GiB)": 77.56,
      "step": 91390,
      "token_acc": 0.49324324324324326,
      "train_speed(iter/s)": 1.438127
    },
    {
      "epoch": 3.915642003341759,
      "grad_norm": 5.908291816711426,
      "learning_rate": 1.1162968272837438e-05,
      "loss": 2.4093696594238283,
      "memory(GiB)": 77.56,
      "step": 91395,
      "token_acc": 0.4693140794223827,
      "train_speed(iter/s)": 1.438115
    },
    {
      "epoch": 3.915856218671008,
      "grad_norm": 5.138401508331299,
      "learning_rate": 1.1158730074945029e-05,
      "loss": 2.358805274963379,
      "memory(GiB)": 77.56,
      "step": 91400,
      "token_acc": 0.5389408099688473,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 3.916070434000257,
      "grad_norm": 5.670218467712402,
      "learning_rate": 1.1154492580698661e-05,
      "loss": 2.3675537109375,
      "memory(GiB)": 77.56,
      "step": 91405,
      "token_acc": 0.5181818181818182,
      "train_speed(iter/s)": 1.438118
    },
    {
      "epoch": 3.916284649329506,
      "grad_norm": 5.988509178161621,
      "learning_rate": 1.1150255790175086e-05,
      "loss": 2.559895324707031,
      "memory(GiB)": 77.56,
      "step": 91410,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 3.916498864658755,
      "grad_norm": 5.357651710510254,
      "learning_rate": 1.1146019703451083e-05,
      "loss": 2.346920394897461,
      "memory(GiB)": 77.56,
      "step": 91415,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.438124
    },
    {
      "epoch": 3.916713079988004,
      "grad_norm": 6.376716613769531,
      "learning_rate": 1.1141784320603382e-05,
      "loss": 2.4762502670288087,
      "memory(GiB)": 77.56,
      "step": 91420,
      "token_acc": 0.4859154929577465,
      "train_speed(iter/s)": 1.438118
    },
    {
      "epoch": 3.9169272953172527,
      "grad_norm": 5.306493282318115,
      "learning_rate": 1.1137549641708717e-05,
      "loss": 2.338175964355469,
      "memory(GiB)": 77.56,
      "step": 91425,
      "token_acc": 0.505338078291815,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 3.917141510646502,
      "grad_norm": 6.561230182647705,
      "learning_rate": 1.1133315666843791e-05,
      "loss": 2.5423145294189453,
      "memory(GiB)": 77.56,
      "step": 91430,
      "token_acc": 0.4937106918238994,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 3.9173557259757508,
      "grad_norm": 6.005884647369385,
      "learning_rate": 1.1129082396085294e-05,
      "loss": 2.125341606140137,
      "memory(GiB)": 77.56,
      "step": 91435,
      "token_acc": 0.5596707818930041,
      "train_speed(iter/s)": 1.438155
    },
    {
      "epoch": 3.9175699413049996,
      "grad_norm": 6.431205749511719,
      "learning_rate": 1.1124849829509953e-05,
      "loss": 2.228074836730957,
      "memory(GiB)": 77.56,
      "step": 91440,
      "token_acc": 0.525,
      "train_speed(iter/s)": 1.43816
    },
    {
      "epoch": 3.917784156634249,
      "grad_norm": 6.310107231140137,
      "learning_rate": 1.1120617967194425e-05,
      "loss": 2.296904182434082,
      "memory(GiB)": 77.56,
      "step": 91445,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438188
    },
    {
      "epoch": 3.9179983719634977,
      "grad_norm": 5.776144027709961,
      "learning_rate": 1.1116386809215368e-05,
      "loss": 2.4614768981933595,
      "memory(GiB)": 77.56,
      "step": 91450,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.438189
    },
    {
      "epoch": 3.9182125872927465,
      "grad_norm": 5.911812782287598,
      "learning_rate": 1.1112156355649444e-05,
      "loss": 2.2575763702392577,
      "memory(GiB)": 77.56,
      "step": 91455,
      "token_acc": 0.5047923322683706,
      "train_speed(iter/s)": 1.438197
    },
    {
      "epoch": 3.9184268026219957,
      "grad_norm": 4.981050491333008,
      "learning_rate": 1.1107926606573287e-05,
      "loss": 2.4948247909545898,
      "memory(GiB)": 77.56,
      "step": 91460,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.438206
    },
    {
      "epoch": 3.9186410179512445,
      "grad_norm": 5.084926128387451,
      "learning_rate": 1.1103697562063513e-05,
      "loss": 2.058063507080078,
      "memory(GiB)": 77.56,
      "step": 91465,
      "token_acc": 0.5505226480836237,
      "train_speed(iter/s)": 1.438204
    },
    {
      "epoch": 3.9188552332804933,
      "grad_norm": 5.6330132484436035,
      "learning_rate": 1.109946922219674e-05,
      "loss": 2.3477508544921877,
      "memory(GiB)": 77.56,
      "step": 91470,
      "token_acc": 0.5253164556962026,
      "train_speed(iter/s)": 1.438212
    },
    {
      "epoch": 3.9190694486097426,
      "grad_norm": 6.594236373901367,
      "learning_rate": 1.1095241587049593e-05,
      "loss": 2.533837890625,
      "memory(GiB)": 77.56,
      "step": 91475,
      "token_acc": 0.4872611464968153,
      "train_speed(iter/s)": 1.438226
    },
    {
      "epoch": 3.9192836639389914,
      "grad_norm": 7.738058567047119,
      "learning_rate": 1.1091014656698634e-05,
      "loss": 2.3503273010253904,
      "memory(GiB)": 77.56,
      "step": 91480,
      "token_acc": 0.4830188679245283,
      "train_speed(iter/s)": 1.438223
    },
    {
      "epoch": 3.9194978792682402,
      "grad_norm": 6.766422271728516,
      "learning_rate": 1.1086788431220446e-05,
      "loss": 2.350724792480469,
      "memory(GiB)": 77.56,
      "step": 91485,
      "token_acc": 0.531986531986532,
      "train_speed(iter/s)": 1.438213
    },
    {
      "epoch": 3.9197120945974895,
      "grad_norm": 9.74452018737793,
      "learning_rate": 1.108256291069159e-05,
      "loss": 2.315723419189453,
      "memory(GiB)": 77.56,
      "step": 91490,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438222
    },
    {
      "epoch": 3.9199263099267383,
      "grad_norm": 6.988160133361816,
      "learning_rate": 1.1078338095188601e-05,
      "loss": 2.458779716491699,
      "memory(GiB)": 77.56,
      "step": 91495,
      "token_acc": 0.4929078014184397,
      "train_speed(iter/s)": 1.438215
    },
    {
      "epoch": 3.920140525255987,
      "grad_norm": 6.007402420043945,
      "learning_rate": 1.1074113984788043e-05,
      "loss": 2.2051992416381836,
      "memory(GiB)": 77.56,
      "step": 91500,
      "token_acc": 0.5134099616858238,
      "train_speed(iter/s)": 1.43822
    },
    {
      "epoch": 3.920140525255987,
      "eval_loss": 2.230638027191162,
      "eval_runtime": 14.1072,
      "eval_samples_per_second": 7.089,
      "eval_steps_per_second": 7.089,
      "eval_token_acc": 0.4462915601023018,
      "step": 91500
    },
    {
      "epoch": 3.9203547405852364,
      "grad_norm": 6.819450378417969,
      "learning_rate": 1.1069890579566427e-05,
      "loss": 2.343904495239258,
      "memory(GiB)": 77.56,
      "step": 91505,
      "token_acc": 0.4606946983546618,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 3.920568955914485,
      "grad_norm": 8.59701156616211,
      "learning_rate": 1.106566787960026e-05,
      "loss": 2.2338972091674805,
      "memory(GiB)": 77.56,
      "step": 91510,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 3.920783171243734,
      "grad_norm": 5.157470703125,
      "learning_rate": 1.1061445884966042e-05,
      "loss": 2.673126983642578,
      "memory(GiB)": 77.56,
      "step": 91515,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.437872
    },
    {
      "epoch": 3.9209973865729832,
      "grad_norm": 5.757594585418701,
      "learning_rate": 1.1057224595740246e-05,
      "loss": 2.493881607055664,
      "memory(GiB)": 77.56,
      "step": 91520,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 3.921211601902232,
      "grad_norm": 6.437959671020508,
      "learning_rate": 1.1053004011999374e-05,
      "loss": 2.334260940551758,
      "memory(GiB)": 77.56,
      "step": 91525,
      "token_acc": 0.5458333333333333,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 3.921425817231481,
      "grad_norm": 4.9855875968933105,
      "learning_rate": 1.1048784133819867e-05,
      "loss": 2.3796356201171873,
      "memory(GiB)": 77.56,
      "step": 91530,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.92164003256073,
      "grad_norm": 6.146213531494141,
      "learning_rate": 1.1044564961278175e-05,
      "loss": 2.1525810241699217,
      "memory(GiB)": 77.56,
      "step": 91535,
      "token_acc": 0.5321428571428571,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.921854247889979,
      "grad_norm": 5.222489356994629,
      "learning_rate": 1.104034649445072e-05,
      "loss": 2.2984035491943358,
      "memory(GiB)": 77.56,
      "step": 91540,
      "token_acc": 0.48348348348348347,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 3.9220684632192278,
      "grad_norm": 6.565844535827637,
      "learning_rate": 1.1036128733413948e-05,
      "loss": 2.125788688659668,
      "memory(GiB)": 77.56,
      "step": 91545,
      "token_acc": 0.5517241379310345,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 3.922282678548477,
      "grad_norm": 5.705295085906982,
      "learning_rate": 1.1031911678244255e-05,
      "loss": 2.2338294982910156,
      "memory(GiB)": 77.56,
      "step": 91550,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437868
    },
    {
      "epoch": 3.922496893877726,
      "grad_norm": 6.459292411804199,
      "learning_rate": 1.1027695329018023e-05,
      "loss": 2.3540958404541015,
      "memory(GiB)": 77.56,
      "step": 91555,
      "token_acc": 0.4932249322493225,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 3.9227111092069746,
      "grad_norm": 7.275162220001221,
      "learning_rate": 1.102347968581166e-05,
      "loss": 1.8346271514892578,
      "memory(GiB)": 77.56,
      "step": 91560,
      "token_acc": 0.5487179487179488,
      "train_speed(iter/s)": 1.437884
    },
    {
      "epoch": 3.922925324536224,
      "grad_norm": 6.538516044616699,
      "learning_rate": 1.101926474870153e-05,
      "loss": 2.2621707916259766,
      "memory(GiB)": 77.56,
      "step": 91565,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 3.9231395398654727,
      "grad_norm": 6.015599727630615,
      "learning_rate": 1.1015050517763987e-05,
      "loss": 2.421868324279785,
      "memory(GiB)": 77.56,
      "step": 91570,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.9233537551947215,
      "grad_norm": 5.973509311676025,
      "learning_rate": 1.101083699307537e-05,
      "loss": 2.1814422607421875,
      "memory(GiB)": 77.56,
      "step": 91575,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 3.9235679705239708,
      "grad_norm": 7.527085781097412,
      "learning_rate": 1.1006624174711999e-05,
      "loss": 2.491049385070801,
      "memory(GiB)": 77.56,
      "step": 91580,
      "token_acc": 0.4536082474226804,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.9237821858532196,
      "grad_norm": 5.906025409698486,
      "learning_rate": 1.1002412062750233e-05,
      "loss": 2.557822418212891,
      "memory(GiB)": 77.56,
      "step": 91585,
      "token_acc": 0.4448669201520912,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 3.9239964011824684,
      "grad_norm": 5.4656829833984375,
      "learning_rate": 1.099820065726635e-05,
      "loss": 2.3741434097290037,
      "memory(GiB)": 77.56,
      "step": 91590,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.437857
    },
    {
      "epoch": 3.9242106165117177,
      "grad_norm": 6.288724422454834,
      "learning_rate": 1.0993989958336648e-05,
      "loss": 2.0333003997802734,
      "memory(GiB)": 77.56,
      "step": 91595,
      "token_acc": 0.536,
      "train_speed(iter/s)": 1.437849
    },
    {
      "epoch": 3.9244248318409665,
      "grad_norm": 6.021236896514893,
      "learning_rate": 1.098977996603741e-05,
      "loss": 2.583403778076172,
      "memory(GiB)": 77.56,
      "step": 91600,
      "token_acc": 0.46254071661237783,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 3.9246390471702153,
      "grad_norm": 4.439737319946289,
      "learning_rate": 1.09855706804449e-05,
      "loss": 2.3207374572753907,
      "memory(GiB)": 77.56,
      "step": 91605,
      "token_acc": 0.4585635359116022,
      "train_speed(iter/s)": 1.437884
    },
    {
      "epoch": 3.9248532624994645,
      "grad_norm": 7.106756687164307,
      "learning_rate": 1.0981362101635367e-05,
      "loss": 2.422466278076172,
      "memory(GiB)": 77.56,
      "step": 91610,
      "token_acc": 0.47854785478547857,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 3.9250674778287133,
      "grad_norm": 7.427601337432861,
      "learning_rate": 1.0977154229685055e-05,
      "loss": 2.144119644165039,
      "memory(GiB)": 77.56,
      "step": 91615,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.437906
    },
    {
      "epoch": 3.925281693157962,
      "grad_norm": 6.106443881988525,
      "learning_rate": 1.097294706467022e-05,
      "loss": 2.0666793823242187,
      "memory(GiB)": 77.56,
      "step": 91620,
      "token_acc": 0.5321428571428571,
      "train_speed(iter/s)": 1.437906
    },
    {
      "epoch": 3.9254959084872114,
      "grad_norm": 7.301298141479492,
      "learning_rate": 1.096874060666706e-05,
      "loss": 2.153365135192871,
      "memory(GiB)": 77.56,
      "step": 91625,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.437913
    },
    {
      "epoch": 3.9257101238164602,
      "grad_norm": 8.789388656616211,
      "learning_rate": 1.0964534855751774e-05,
      "loss": 2.432993698120117,
      "memory(GiB)": 77.56,
      "step": 91630,
      "token_acc": 0.4847328244274809,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 3.925924339145709,
      "grad_norm": 7.694841384887695,
      "learning_rate": 1.0960329812000557e-05,
      "loss": 2.319536018371582,
      "memory(GiB)": 77.56,
      "step": 91635,
      "token_acc": 0.5403508771929825,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 3.9261385544749583,
      "grad_norm": 7.1682820320129395,
      "learning_rate": 1.0956125475489575e-05,
      "loss": 2.5165098190307615,
      "memory(GiB)": 77.56,
      "step": 91640,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 3.926352769804207,
      "grad_norm": 5.862178802490234,
      "learning_rate": 1.095192184629502e-05,
      "loss": 2.5554582595825197,
      "memory(GiB)": 77.56,
      "step": 91645,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437943
    },
    {
      "epoch": 3.926566985133456,
      "grad_norm": 7.210686206817627,
      "learning_rate": 1.0947718924493034e-05,
      "loss": 2.2724714279174805,
      "memory(GiB)": 77.56,
      "step": 91650,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 3.926781200462705,
      "grad_norm": 5.732788562774658,
      "learning_rate": 1.0943516710159751e-05,
      "loss": 2.31829833984375,
      "memory(GiB)": 77.56,
      "step": 91655,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 3.926995415791954,
      "grad_norm": 5.065240383148193,
      "learning_rate": 1.0939315203371303e-05,
      "loss": 2.5158546447753904,
      "memory(GiB)": 77.56,
      "step": 91660,
      "token_acc": 0.46956521739130436,
      "train_speed(iter/s)": 1.437972
    },
    {
      "epoch": 3.927209631121203,
      "grad_norm": 6.3762736320495605,
      "learning_rate": 1.0935114404203783e-05,
      "loss": 2.3252967834472655,
      "memory(GiB)": 77.56,
      "step": 91665,
      "token_acc": 0.4711864406779661,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 3.927423846450452,
      "grad_norm": 4.754208087921143,
      "learning_rate": 1.0930914312733332e-05,
      "loss": 2.14044132232666,
      "memory(GiB)": 77.56,
      "step": 91670,
      "token_acc": 0.5359477124183006,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 3.927638061779701,
      "grad_norm": 5.205066680908203,
      "learning_rate": 1.0926714929036014e-05,
      "loss": 2.4429603576660157,
      "memory(GiB)": 77.56,
      "step": 91675,
      "token_acc": 0.4630225080385852,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 3.9278522771089497,
      "grad_norm": 5.333390235900879,
      "learning_rate": 1.0922516253187909e-05,
      "loss": 2.2542984008789064,
      "memory(GiB)": 77.56,
      "step": 91680,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.928066492438199,
      "grad_norm": 5.958066463470459,
      "learning_rate": 1.0918318285265078e-05,
      "loss": 2.263372802734375,
      "memory(GiB)": 77.56,
      "step": 91685,
      "token_acc": 0.5607142857142857,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.9282807077674478,
      "grad_norm": 5.380476474761963,
      "learning_rate": 1.091412102534356e-05,
      "loss": 2.413420867919922,
      "memory(GiB)": 77.56,
      "step": 91690,
      "token_acc": 0.49698795180722893,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 3.9284949230966966,
      "grad_norm": 4.951543807983398,
      "learning_rate": 1.0909924473499423e-05,
      "loss": 2.372220993041992,
      "memory(GiB)": 77.56,
      "step": 91695,
      "token_acc": 0.47701149425287354,
      "train_speed(iter/s)": 1.437997
    },
    {
      "epoch": 3.928709138425946,
      "grad_norm": 6.767967224121094,
      "learning_rate": 1.0905728629808654e-05,
      "loss": 2.1728693008422852,
      "memory(GiB)": 77.56,
      "step": 91700,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.437975
    },
    {
      "epoch": 3.9289233537551946,
      "grad_norm": 7.530840873718262,
      "learning_rate": 1.0901533494347304e-05,
      "loss": 2.5382482528686525,
      "memory(GiB)": 77.56,
      "step": 91705,
      "token_acc": 0.4788273615635179,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 3.9291375690844434,
      "grad_norm": 5.2336554527282715,
      "learning_rate": 1.0897339067191347e-05,
      "loss": 2.3763662338256837,
      "memory(GiB)": 77.56,
      "step": 91710,
      "token_acc": 0.503731343283582,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 3.9293517844136927,
      "grad_norm": 4.730438232421875,
      "learning_rate": 1.0893145348416778e-05,
      "loss": 2.222860336303711,
      "memory(GiB)": 77.56,
      "step": 91715,
      "token_acc": 0.491869918699187,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 3.9295659997429415,
      "grad_norm": 5.695847034454346,
      "learning_rate": 1.0888952338099561e-05,
      "loss": 2.4839725494384766,
      "memory(GiB)": 77.56,
      "step": 91720,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 3.9297802150721903,
      "grad_norm": 5.649783611297607,
      "learning_rate": 1.088476003631565e-05,
      "loss": 2.373153305053711,
      "memory(GiB)": 77.56,
      "step": 91725,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 3.9299944304014396,
      "grad_norm": 14.608512878417969,
      "learning_rate": 1.088056844314102e-05,
      "loss": 2.4498228073120116,
      "memory(GiB)": 77.56,
      "step": 91730,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 3.9302086457306884,
      "grad_norm": 7.321146011352539,
      "learning_rate": 1.087637755865159e-05,
      "loss": 2.3396718978881834,
      "memory(GiB)": 77.56,
      "step": 91735,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 3.930422861059937,
      "grad_norm": 5.216070175170898,
      "learning_rate": 1.0872187382923277e-05,
      "loss": 2.3084171295166014,
      "memory(GiB)": 77.56,
      "step": 91740,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438032
    },
    {
      "epoch": 3.9306370763891865,
      "grad_norm": 6.924057483673096,
      "learning_rate": 1.0867997916031997e-05,
      "loss": 2.1397930145263673,
      "memory(GiB)": 77.56,
      "step": 91745,
      "token_acc": 0.5617977528089888,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 3.9308512917184353,
      "grad_norm": 6.8544769287109375,
      "learning_rate": 1.0863809158053645e-05,
      "loss": 2.2442209243774416,
      "memory(GiB)": 77.56,
      "step": 91750,
      "token_acc": 0.5267175572519084,
      "train_speed(iter/s)": 1.438042
    },
    {
      "epoch": 3.931065507047684,
      "grad_norm": 5.260197639465332,
      "learning_rate": 1.0859621109064089e-05,
      "loss": 2.656650733947754,
      "memory(GiB)": 77.56,
      "step": 91755,
      "token_acc": 0.46418338108882523,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 3.9312797223769333,
      "grad_norm": 6.848014831542969,
      "learning_rate": 1.0855433769139223e-05,
      "loss": 2.2859222412109377,
      "memory(GiB)": 77.56,
      "step": 91760,
      "token_acc": 0.5126353790613718,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 3.931493937706182,
      "grad_norm": 5.351479530334473,
      "learning_rate": 1.0851247138354886e-05,
      "loss": 2.648614501953125,
      "memory(GiB)": 77.56,
      "step": 91765,
      "token_acc": 0.46407185628742514,
      "train_speed(iter/s)": 1.438044
    },
    {
      "epoch": 3.931708153035431,
      "grad_norm": 6.796504497528076,
      "learning_rate": 1.084706121678694e-05,
      "loss": 2.270510673522949,
      "memory(GiB)": 77.56,
      "step": 91770,
      "token_acc": 0.5157894736842106,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 3.9319223683646802,
      "grad_norm": 7.183225631713867,
      "learning_rate": 1.0842876004511215e-05,
      "loss": 2.47519474029541,
      "memory(GiB)": 77.56,
      "step": 91775,
      "token_acc": 0.45878136200716846,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 3.932136583693929,
      "grad_norm": 5.062354564666748,
      "learning_rate": 1.083869150160352e-05,
      "loss": 2.3480819702148437,
      "memory(GiB)": 77.56,
      "step": 91780,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.438058
    },
    {
      "epoch": 3.932350799023178,
      "grad_norm": 5.611119747161865,
      "learning_rate": 1.0834507708139651e-05,
      "loss": 2.3739879608154295,
      "memory(GiB)": 77.56,
      "step": 91785,
      "token_acc": 0.5139442231075697,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 3.932565014352427,
      "grad_norm": 4.909784317016602,
      "learning_rate": 1.083032462419543e-05,
      "loss": 2.1236831665039064,
      "memory(GiB)": 77.56,
      "step": 91790,
      "token_acc": 0.5348101265822784,
      "train_speed(iter/s)": 1.438083
    },
    {
      "epoch": 3.932779229681676,
      "grad_norm": 5.592144966125488,
      "learning_rate": 1.0826142249846621e-05,
      "loss": 2.3189544677734375,
      "memory(GiB)": 77.56,
      "step": 91795,
      "token_acc": 0.5062893081761006,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 3.9329934450109247,
      "grad_norm": 6.355260372161865,
      "learning_rate": 1.0821960585168994e-05,
      "loss": 2.4610824584960938,
      "memory(GiB)": 77.56,
      "step": 91800,
      "token_acc": 0.48638132295719844,
      "train_speed(iter/s)": 1.438099
    },
    {
      "epoch": 3.933207660340174,
      "grad_norm": 5.820274353027344,
      "learning_rate": 1.0817779630238301e-05,
      "loss": 2.5243928909301756,
      "memory(GiB)": 77.56,
      "step": 91805,
      "token_acc": 0.45231607629427795,
      "train_speed(iter/s)": 1.438099
    },
    {
      "epoch": 3.933421875669423,
      "grad_norm": 7.225492000579834,
      "learning_rate": 1.0813599385130274e-05,
      "loss": 2.2347366333007814,
      "memory(GiB)": 77.56,
      "step": 91810,
      "token_acc": 0.5486381322957199,
      "train_speed(iter/s)": 1.438098
    },
    {
      "epoch": 3.9336360909986716,
      "grad_norm": 6.190768241882324,
      "learning_rate": 1.0809419849920671e-05,
      "loss": 2.237508010864258,
      "memory(GiB)": 77.56,
      "step": 91815,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.438114
    },
    {
      "epoch": 3.933850306327921,
      "grad_norm": 5.9196624755859375,
      "learning_rate": 1.0805241024685186e-05,
      "loss": 2.134809112548828,
      "memory(GiB)": 77.56,
      "step": 91820,
      "token_acc": 0.5567765567765568,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 3.9340645216571697,
      "grad_norm": 6.857304573059082,
      "learning_rate": 1.0801062909499526e-05,
      "loss": 2.034542274475098,
      "memory(GiB)": 77.56,
      "step": 91825,
      "token_acc": 0.5677655677655677,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 3.9342787369864185,
      "grad_norm": 5.46055793762207,
      "learning_rate": 1.0796885504439391e-05,
      "loss": 2.2587892532348635,
      "memory(GiB)": 77.56,
      "step": 91830,
      "token_acc": 0.5606694560669456,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 3.9344929523156678,
      "grad_norm": 6.6425700187683105,
      "learning_rate": 1.0792708809580431e-05,
      "loss": 2.245984649658203,
      "memory(GiB)": 77.56,
      "step": 91835,
      "token_acc": 0.49415204678362573,
      "train_speed(iter/s)": 1.438038
    },
    {
      "epoch": 3.9347071676449166,
      "grad_norm": 5.84804630279541,
      "learning_rate": 1.0788532824998343e-05,
      "loss": 2.185314178466797,
      "memory(GiB)": 77.56,
      "step": 91840,
      "token_acc": 0.5203761755485894,
      "train_speed(iter/s)": 1.438053
    },
    {
      "epoch": 3.9349213829741654,
      "grad_norm": 6.235397815704346,
      "learning_rate": 1.0784357550768753e-05,
      "loss": 2.2381511688232423,
      "memory(GiB)": 77.56,
      "step": 91845,
      "token_acc": 0.5126582278481012,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 3.9351355983034146,
      "grad_norm": 5.820837020874023,
      "learning_rate": 1.078018298696733e-05,
      "loss": 2.3817996978759766,
      "memory(GiB)": 77.56,
      "step": 91850,
      "token_acc": 0.5508771929824562,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 3.9353498136326635,
      "grad_norm": 5.505218982696533,
      "learning_rate": 1.0776009133669684e-05,
      "loss": 2.1380382537841798,
      "memory(GiB)": 77.56,
      "step": 91855,
      "token_acc": 0.5127388535031847,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 3.9355640289619123,
      "grad_norm": 7.809144973754883,
      "learning_rate": 1.0771835990951424e-05,
      "loss": 2.6247671127319334,
      "memory(GiB)": 77.56,
      "step": 91860,
      "token_acc": 0.47194719471947194,
      "train_speed(iter/s)": 1.438113
    },
    {
      "epoch": 3.9357782442911615,
      "grad_norm": 6.382263660430908,
      "learning_rate": 1.0767663558888159e-05,
      "loss": 2.4069332122802733,
      "memory(GiB)": 77.56,
      "step": 91865,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.438123
    },
    {
      "epoch": 3.9359924596204103,
      "grad_norm": 7.607944488525391,
      "learning_rate": 1.076349183755545e-05,
      "loss": 2.3623115539550783,
      "memory(GiB)": 77.56,
      "step": 91870,
      "token_acc": 0.4962962962962963,
      "train_speed(iter/s)": 1.438129
    },
    {
      "epoch": 3.936206674949659,
      "grad_norm": 6.977862358093262,
      "learning_rate": 1.0759320827028913e-05,
      "loss": 2.205623435974121,
      "memory(GiB)": 77.56,
      "step": 91875,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.43813
    },
    {
      "epoch": 3.9364208902789084,
      "grad_norm": 10.510127067565918,
      "learning_rate": 1.0755150527384089e-05,
      "loss": 2.475603485107422,
      "memory(GiB)": 77.56,
      "step": 91880,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 3.936635105608157,
      "grad_norm": 5.402792930603027,
      "learning_rate": 1.0750980938696525e-05,
      "loss": 2.612287902832031,
      "memory(GiB)": 77.56,
      "step": 91885,
      "token_acc": 0.483974358974359,
      "train_speed(iter/s)": 1.438141
    },
    {
      "epoch": 3.936849320937406,
      "grad_norm": 5.810826778411865,
      "learning_rate": 1.0746812061041761e-05,
      "loss": 2.2637868881225587,
      "memory(GiB)": 77.56,
      "step": 91890,
      "token_acc": 0.5059760956175299,
      "train_speed(iter/s)": 1.438147
    },
    {
      "epoch": 3.9370635362666553,
      "grad_norm": 6.126474857330322,
      "learning_rate": 1.0742643894495319e-05,
      "loss": 2.1765499114990234,
      "memory(GiB)": 77.56,
      "step": 91895,
      "token_acc": 0.53515625,
      "train_speed(iter/s)": 1.438148
    },
    {
      "epoch": 3.937277751595904,
      "grad_norm": 7.01622200012207,
      "learning_rate": 1.07384764391327e-05,
      "loss": 2.138622283935547,
      "memory(GiB)": 77.56,
      "step": 91900,
      "token_acc": 0.5236363636363637,
      "train_speed(iter/s)": 1.438157
    },
    {
      "epoch": 3.937491966925153,
      "grad_norm": 5.674077033996582,
      "learning_rate": 1.0734309695029422e-05,
      "loss": 2.2281938552856446,
      "memory(GiB)": 77.56,
      "step": 91905,
      "token_acc": 0.49612403100775193,
      "train_speed(iter/s)": 1.438173
    },
    {
      "epoch": 3.937706182254402,
      "grad_norm": 6.738650321960449,
      "learning_rate": 1.0730143662260938e-05,
      "loss": 2.4492055892944338,
      "memory(GiB)": 77.56,
      "step": 91910,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.438183
    },
    {
      "epoch": 3.937920397583651,
      "grad_norm": 8.028721809387207,
      "learning_rate": 1.0725978340902758e-05,
      "loss": 2.5546566009521485,
      "memory(GiB)": 77.56,
      "step": 91915,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438192
    },
    {
      "epoch": 3.9381346129129,
      "grad_norm": 5.925221920013428,
      "learning_rate": 1.0721813731030328e-05,
      "loss": 2.3093683242797853,
      "memory(GiB)": 77.56,
      "step": 91920,
      "token_acc": 0.5096774193548387,
      "train_speed(iter/s)": 1.4382
    },
    {
      "epoch": 3.938348828242149,
      "grad_norm": 6.920371055603027,
      "learning_rate": 1.071764983271908e-05,
      "loss": 2.6063796997070314,
      "memory(GiB)": 77.56,
      "step": 91925,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.438227
    },
    {
      "epoch": 3.938563043571398,
      "grad_norm": 6.748383045196533,
      "learning_rate": 1.0713486646044446e-05,
      "loss": 2.2467294692993165,
      "memory(GiB)": 77.56,
      "step": 91930,
      "token_acc": 0.5182724252491694,
      "train_speed(iter/s)": 1.438218
    },
    {
      "epoch": 3.9387772589006467,
      "grad_norm": 5.717504501342773,
      "learning_rate": 1.0709324171081863e-05,
      "loss": 2.2911935806274415,
      "memory(GiB)": 77.56,
      "step": 91935,
      "token_acc": 0.541958041958042,
      "train_speed(iter/s)": 1.438223
    },
    {
      "epoch": 3.938991474229896,
      "grad_norm": 5.653698921203613,
      "learning_rate": 1.0705162407906739e-05,
      "loss": 2.173803520202637,
      "memory(GiB)": 77.56,
      "step": 91940,
      "token_acc": 0.5189873417721519,
      "train_speed(iter/s)": 1.43822
    },
    {
      "epoch": 3.9392056895591447,
      "grad_norm": 6.838394641876221,
      "learning_rate": 1.0701001356594453e-05,
      "loss": 2.5400789260864256,
      "memory(GiB)": 77.56,
      "step": 91945,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.438236
    },
    {
      "epoch": 3.9394199048883936,
      "grad_norm": 6.167585372924805,
      "learning_rate": 1.0696841017220394e-05,
      "loss": 2.295496940612793,
      "memory(GiB)": 77.56,
      "step": 91950,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.438235
    },
    {
      "epoch": 3.939634120217643,
      "grad_norm": 5.840410232543945,
      "learning_rate": 1.0692681389859916e-05,
      "loss": 2.1997032165527344,
      "memory(GiB)": 77.56,
      "step": 91955,
      "token_acc": 0.5061728395061729,
      "train_speed(iter/s)": 1.438241
    },
    {
      "epoch": 3.9398483355468916,
      "grad_norm": 7.691062927246094,
      "learning_rate": 1.0688522474588403e-05,
      "loss": 2.2673009872436523,
      "memory(GiB)": 77.56,
      "step": 91960,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438242
    },
    {
      "epoch": 3.9400625508761404,
      "grad_norm": 4.867900371551514,
      "learning_rate": 1.0684364271481184e-05,
      "loss": 2.280361557006836,
      "memory(GiB)": 77.56,
      "step": 91965,
      "token_acc": 0.5337423312883436,
      "train_speed(iter/s)": 1.438237
    },
    {
      "epoch": 3.9402767662053897,
      "grad_norm": 6.0499091148376465,
      "learning_rate": 1.0680206780613583e-05,
      "loss": 2.5694808959960938,
      "memory(GiB)": 77.56,
      "step": 91970,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.438242
    },
    {
      "epoch": 3.9404909815346385,
      "grad_norm": 7.973316192626953,
      "learning_rate": 1.0676050002060922e-05,
      "loss": 2.3693368911743162,
      "memory(GiB)": 77.56,
      "step": 91975,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.438235
    },
    {
      "epoch": 3.9407051968638873,
      "grad_norm": 5.942612648010254,
      "learning_rate": 1.0671893935898509e-05,
      "loss": 2.1118722915649415,
      "memory(GiB)": 77.56,
      "step": 91980,
      "token_acc": 0.5290102389078498,
      "train_speed(iter/s)": 1.438227
    },
    {
      "epoch": 3.9409194121931366,
      "grad_norm": 5.290731906890869,
      "learning_rate": 1.0667738582201608e-05,
      "loss": 2.2850690841674806,
      "memory(GiB)": 77.56,
      "step": 91985,
      "token_acc": 0.5451127819548872,
      "train_speed(iter/s)": 1.438227
    },
    {
      "epoch": 3.9411336275223854,
      "grad_norm": 6.314393997192383,
      "learning_rate": 1.0663583941045525e-05,
      "loss": 2.239452362060547,
      "memory(GiB)": 77.56,
      "step": 91990,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 3.941347842851634,
      "grad_norm": 6.988499641418457,
      "learning_rate": 1.0659430012505534e-05,
      "loss": 2.1616661071777346,
      "memory(GiB)": 77.56,
      "step": 91995,
      "token_acc": 0.5269230769230769,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 3.9415620581808835,
      "grad_norm": 6.156588554382324,
      "learning_rate": 1.0655276796656871e-05,
      "loss": 2.474592399597168,
      "memory(GiB)": 77.56,
      "step": 92000,
      "token_acc": 0.504950495049505,
      "train_speed(iter/s)": 1.438251
    },
    {
      "epoch": 3.9415620581808835,
      "eval_loss": 2.3511781692504883,
      "eval_runtime": 14.3955,
      "eval_samples_per_second": 6.947,
      "eval_steps_per_second": 6.947,
      "eval_token_acc": 0.4568764568764569,
      "step": 92000
    },
    {
      "epoch": 3.9417762735101323,
      "grad_norm": 5.084593772888184,
      "learning_rate": 1.0651124293574777e-05,
      "loss": 2.261703300476074,
      "memory(GiB)": 77.56,
      "step": 92005,
      "token_acc": 0.4794168096054888,
      "train_speed(iter/s)": 1.437878
    },
    {
      "epoch": 3.941990488839381,
      "grad_norm": 5.4553680419921875,
      "learning_rate": 1.0646972503334474e-05,
      "loss": 2.407316207885742,
      "memory(GiB)": 77.56,
      "step": 92010,
      "token_acc": 0.46153846153846156,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 3.9422047041686303,
      "grad_norm": 5.671087265014648,
      "learning_rate": 1.0642821426011174e-05,
      "loss": 2.3135040283203123,
      "memory(GiB)": 77.56,
      "step": 92015,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.437872
    },
    {
      "epoch": 3.942418919497879,
      "grad_norm": 6.357110023498535,
      "learning_rate": 1.0638671061680095e-05,
      "loss": 2.278458023071289,
      "memory(GiB)": 77.56,
      "step": 92020,
      "token_acc": 0.4909090909090909,
      "train_speed(iter/s)": 1.437862
    },
    {
      "epoch": 3.942633134827128,
      "grad_norm": 7.281513214111328,
      "learning_rate": 1.063452141041641e-05,
      "loss": 2.2567901611328125,
      "memory(GiB)": 77.56,
      "step": 92025,
      "token_acc": 0.5102564102564102,
      "train_speed(iter/s)": 1.437856
    },
    {
      "epoch": 3.942847350156377,
      "grad_norm": 6.84661340713501,
      "learning_rate": 1.0630372472295303e-05,
      "loss": 2.1634395599365233,
      "memory(GiB)": 77.56,
      "step": 92030,
      "token_acc": 0.5177304964539007,
      "train_speed(iter/s)": 1.437867
    },
    {
      "epoch": 3.943061565485626,
      "grad_norm": 7.569800853729248,
      "learning_rate": 1.0626224247391924e-05,
      "loss": 2.3042369842529298,
      "memory(GiB)": 77.56,
      "step": 92035,
      "token_acc": 0.5029761904761905,
      "train_speed(iter/s)": 1.437861
    },
    {
      "epoch": 3.943275780814875,
      "grad_norm": 7.012302398681641,
      "learning_rate": 1.0622076735781433e-05,
      "loss": 2.3331539154052736,
      "memory(GiB)": 77.56,
      "step": 92040,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 3.943489996144124,
      "grad_norm": 7.040589809417725,
      "learning_rate": 1.0617929937538951e-05,
      "loss": 2.262278366088867,
      "memory(GiB)": 77.56,
      "step": 92045,
      "token_acc": 0.5362903225806451,
      "train_speed(iter/s)": 1.437886
    },
    {
      "epoch": 3.943704211473373,
      "grad_norm": 5.6781415939331055,
      "learning_rate": 1.0613783852739617e-05,
      "loss": 2.512264442443848,
      "memory(GiB)": 77.56,
      "step": 92050,
      "token_acc": 0.5016949152542373,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 3.9439184268026217,
      "grad_norm": 5.332988262176514,
      "learning_rate": 1.0609638481458545e-05,
      "loss": 2.597708511352539,
      "memory(GiB)": 77.56,
      "step": 92055,
      "token_acc": 0.47474747474747475,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 3.944132642131871,
      "grad_norm": 5.484576225280762,
      "learning_rate": 1.0605493823770801e-05,
      "loss": 2.546600914001465,
      "memory(GiB)": 77.56,
      "step": 92060,
      "token_acc": 0.48059701492537316,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 3.94434685746112,
      "grad_norm": 4.486135005950928,
      "learning_rate": 1.060134987975151e-05,
      "loss": 2.3396984100341798,
      "memory(GiB)": 77.56,
      "step": 92065,
      "token_acc": 0.49683544303797467,
      "train_speed(iter/s)": 1.43789
    },
    {
      "epoch": 3.9445610727903686,
      "grad_norm": 7.204801559448242,
      "learning_rate": 1.0597206649475721e-05,
      "loss": 2.7304920196533202,
      "memory(GiB)": 77.56,
      "step": 92070,
      "token_acc": 0.45195729537366547,
      "train_speed(iter/s)": 1.437891
    },
    {
      "epoch": 3.944775288119618,
      "grad_norm": 4.8624138832092285,
      "learning_rate": 1.0593064133018488e-05,
      "loss": 2.402364158630371,
      "memory(GiB)": 77.56,
      "step": 92075,
      "token_acc": 0.5105105105105106,
      "train_speed(iter/s)": 1.437888
    },
    {
      "epoch": 3.9449895034488667,
      "grad_norm": 5.270648002624512,
      "learning_rate": 1.0588922330454876e-05,
      "loss": 2.148133087158203,
      "memory(GiB)": 77.56,
      "step": 92080,
      "token_acc": 0.5163398692810458,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 3.9452037187781155,
      "grad_norm": 5.261557579040527,
      "learning_rate": 1.0584781241859903e-05,
      "loss": 2.6369842529296874,
      "memory(GiB)": 77.56,
      "step": 92085,
      "token_acc": 0.48494983277591974,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.9454179341073647,
      "grad_norm": 5.2815704345703125,
      "learning_rate": 1.05806408673086e-05,
      "loss": 2.385141372680664,
      "memory(GiB)": 77.56,
      "step": 92090,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.437882
    },
    {
      "epoch": 3.9456321494366136,
      "grad_norm": 5.734361171722412,
      "learning_rate": 1.057650120687596e-05,
      "loss": 2.3255592346191407,
      "memory(GiB)": 77.56,
      "step": 92095,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.437843
    },
    {
      "epoch": 3.9458463647658624,
      "grad_norm": 7.409005641937256,
      "learning_rate": 1.057236226063697e-05,
      "loss": 2.0789005279541017,
      "memory(GiB)": 77.56,
      "step": 92100,
      "token_acc": 0.5568862275449101,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 3.9460605800951116,
      "grad_norm": 7.198379039764404,
      "learning_rate": 1.056822402866664e-05,
      "loss": 2.493305969238281,
      "memory(GiB)": 77.56,
      "step": 92105,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.9462747954243604,
      "grad_norm": 6.27956485748291,
      "learning_rate": 1.0564086511039917e-05,
      "loss": 2.3432601928710937,
      "memory(GiB)": 77.56,
      "step": 92110,
      "token_acc": 0.5124223602484472,
      "train_speed(iter/s)": 1.437863
    },
    {
      "epoch": 3.9464890107536097,
      "grad_norm": 6.296996593475342,
      "learning_rate": 1.0559949707831762e-05,
      "loss": 2.2563201904296877,
      "memory(GiB)": 77.56,
      "step": 92115,
      "token_acc": 0.5283018867924528,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.9467032260828585,
      "grad_norm": 11.899503707885742,
      "learning_rate": 1.0555813619117122e-05,
      "loss": 2.2604516983032226,
      "memory(GiB)": 77.56,
      "step": 92120,
      "token_acc": 0.5367647058823529,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 3.9469174414121073,
      "grad_norm": 7.101419925689697,
      "learning_rate": 1.0551678244970914e-05,
      "loss": 2.4868316650390625,
      "memory(GiB)": 77.56,
      "step": 92125,
      "token_acc": 0.47079037800687284,
      "train_speed(iter/s)": 1.437873
    },
    {
      "epoch": 3.9471316567413566,
      "grad_norm": 6.956063747406006,
      "learning_rate": 1.0547543585468044e-05,
      "loss": 2.3549430847167967,
      "memory(GiB)": 77.56,
      "step": 92130,
      "token_acc": 0.4855305466237942,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 3.9473458720706054,
      "grad_norm": 5.81870174407959,
      "learning_rate": 1.0543409640683439e-05,
      "loss": 2.129019927978516,
      "memory(GiB)": 77.56,
      "step": 92135,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.437888
    },
    {
      "epoch": 3.947560087399854,
      "grad_norm": 6.7818989753723145,
      "learning_rate": 1.053927641069199e-05,
      "loss": 2.2010318756103517,
      "memory(GiB)": 77.56,
      "step": 92140,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.437868
    },
    {
      "epoch": 3.9477743027291035,
      "grad_norm": 5.320273399353027,
      "learning_rate": 1.0535143895568566e-05,
      "loss": 2.4698165893554687,
      "memory(GiB)": 77.56,
      "step": 92145,
      "token_acc": 0.47701149425287354,
      "train_speed(iter/s)": 1.437861
    },
    {
      "epoch": 3.9479885180583523,
      "grad_norm": 6.498271942138672,
      "learning_rate": 1.0531012095388032e-05,
      "loss": 2.415632629394531,
      "memory(GiB)": 77.56,
      "step": 92150,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.948202733387601,
      "grad_norm": 6.241724967956543,
      "learning_rate": 1.0526881010225242e-05,
      "loss": 2.2276315689086914,
      "memory(GiB)": 77.56,
      "step": 92155,
      "token_acc": 0.5381944444444444,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 3.9484169487168503,
      "grad_norm": 5.565863132476807,
      "learning_rate": 1.0522750640155016e-05,
      "loss": 2.213591384887695,
      "memory(GiB)": 77.56,
      "step": 92160,
      "token_acc": 0.5524691358024691,
      "train_speed(iter/s)": 1.437872
    },
    {
      "epoch": 3.948631164046099,
      "grad_norm": 11.722769737243652,
      "learning_rate": 1.0518620985252208e-05,
      "loss": 2.277585029602051,
      "memory(GiB)": 77.56,
      "step": 92165,
      "token_acc": 0.5126353790613718,
      "train_speed(iter/s)": 1.437863
    },
    {
      "epoch": 3.948845379375348,
      "grad_norm": 6.0761237144470215,
      "learning_rate": 1.0514492045591617e-05,
      "loss": 2.217298126220703,
      "memory(GiB)": 77.56,
      "step": 92170,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.43784
    },
    {
      "epoch": 3.949059594704597,
      "grad_norm": 6.448444843292236,
      "learning_rate": 1.0510363821248042e-05,
      "loss": 2.4353891372680665,
      "memory(GiB)": 77.56,
      "step": 92175,
      "token_acc": 0.4965034965034965,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 3.949273810033846,
      "grad_norm": 4.6907958984375,
      "learning_rate": 1.050623631229627e-05,
      "loss": 2.1077457427978517,
      "memory(GiB)": 77.56,
      "step": 92180,
      "token_acc": 0.5296296296296297,
      "train_speed(iter/s)": 1.437862
    },
    {
      "epoch": 3.949488025363095,
      "grad_norm": 7.817314624786377,
      "learning_rate": 1.0502109518811065e-05,
      "loss": 2.628652572631836,
      "memory(GiB)": 77.56,
      "step": 92185,
      "token_acc": 0.48507462686567165,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 3.949702240692344,
      "grad_norm": 4.980259418487549,
      "learning_rate": 1.049798344086721e-05,
      "loss": 2.049240303039551,
      "memory(GiB)": 77.56,
      "step": 92190,
      "token_acc": 0.5265306122448979,
      "train_speed(iter/s)": 1.437875
    },
    {
      "epoch": 3.949916456021593,
      "grad_norm": 4.446688175201416,
      "learning_rate": 1.0493858078539442e-05,
      "loss": 2.167164993286133,
      "memory(GiB)": 77.56,
      "step": 92195,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.9501306713508417,
      "grad_norm": 5.079987049102783,
      "learning_rate": 1.0489733431902494e-05,
      "loss": 2.47097110748291,
      "memory(GiB)": 77.56,
      "step": 92200,
      "token_acc": 0.5143769968051118,
      "train_speed(iter/s)": 1.437855
    },
    {
      "epoch": 3.950344886680091,
      "grad_norm": 6.935995578765869,
      "learning_rate": 1.0485609501031074e-05,
      "loss": 2.2259145736694337,
      "memory(GiB)": 77.56,
      "step": 92205,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.43785
    },
    {
      "epoch": 3.95055910200934,
      "grad_norm": 4.849509239196777,
      "learning_rate": 1.048148628599992e-05,
      "loss": 2.2015085220336914,
      "memory(GiB)": 77.56,
      "step": 92210,
      "token_acc": 0.5746268656716418,
      "train_speed(iter/s)": 1.437835
    },
    {
      "epoch": 3.9507733173385886,
      "grad_norm": 4.599462985992432,
      "learning_rate": 1.0477363786883709e-05,
      "loss": 2.5279514312744142,
      "memory(GiB)": 77.56,
      "step": 92215,
      "token_acc": 0.4881656804733728,
      "train_speed(iter/s)": 1.437825
    },
    {
      "epoch": 3.950987532667838,
      "grad_norm": 5.7188401222229,
      "learning_rate": 1.0473242003757123e-05,
      "loss": 2.378571319580078,
      "memory(GiB)": 77.56,
      "step": 92220,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 3.9512017479970867,
      "grad_norm": 6.001686096191406,
      "learning_rate": 1.0469120936694843e-05,
      "loss": 2.128190803527832,
      "memory(GiB)": 77.56,
      "step": 92225,
      "token_acc": 0.5301204819277109,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 3.9514159633263355,
      "grad_norm": 6.048079013824463,
      "learning_rate": 1.0465000585771522e-05,
      "loss": 2.2242752075195313,
      "memory(GiB)": 77.56,
      "step": 92230,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.437848
    },
    {
      "epoch": 3.9516301786555847,
      "grad_norm": 4.913375377655029,
      "learning_rate": 1.0460880951061807e-05,
      "loss": 2.4028850555419923,
      "memory(GiB)": 77.56,
      "step": 92235,
      "token_acc": 0.476056338028169,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 3.9518443939848336,
      "grad_norm": 4.643843173980713,
      "learning_rate": 1.0456762032640322e-05,
      "loss": 2.3900989532470702,
      "memory(GiB)": 77.56,
      "step": 92240,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.437863
    },
    {
      "epoch": 3.9520586093140824,
      "grad_norm": 7.745403289794922,
      "learning_rate": 1.0452643830581672e-05,
      "loss": 2.1904064178466798,
      "memory(GiB)": 77.56,
      "step": 92245,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.437882
    },
    {
      "epoch": 3.9522728246433316,
      "grad_norm": 5.026733875274658,
      "learning_rate": 1.0448526344960491e-05,
      "loss": 2.5203929901123048,
      "memory(GiB)": 77.56,
      "step": 92250,
      "token_acc": 0.5031645569620253,
      "train_speed(iter/s)": 1.4379
    },
    {
      "epoch": 3.9524870399725804,
      "grad_norm": 7.009357929229736,
      "learning_rate": 1.0444409575851355e-05,
      "loss": 2.217828369140625,
      "memory(GiB)": 77.56,
      "step": 92255,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 3.9527012553018293,
      "grad_norm": 6.079318046569824,
      "learning_rate": 1.0440293523328847e-05,
      "loss": 2.33267822265625,
      "memory(GiB)": 77.56,
      "step": 92260,
      "token_acc": 0.5086206896551724,
      "train_speed(iter/s)": 1.437908
    },
    {
      "epoch": 3.9529154706310785,
      "grad_norm": 6.795281410217285,
      "learning_rate": 1.0436178187467538e-05,
      "loss": 2.4412967681884767,
      "memory(GiB)": 77.56,
      "step": 92265,
      "token_acc": 0.5018050541516246,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 3.9531296859603273,
      "grad_norm": 5.610692501068115,
      "learning_rate": 1.043206356834197e-05,
      "loss": 2.1294960021972655,
      "memory(GiB)": 77.56,
      "step": 92270,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 3.953343901289576,
      "grad_norm": 9.385631561279297,
      "learning_rate": 1.0427949666026677e-05,
      "loss": 2.221689987182617,
      "memory(GiB)": 77.56,
      "step": 92275,
      "token_acc": 0.5192307692307693,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 3.9535581166188254,
      "grad_norm": 5.482063293457031,
      "learning_rate": 1.0423836480596195e-05,
      "loss": 2.5405521392822266,
      "memory(GiB)": 77.56,
      "step": 92280,
      "token_acc": 0.468013468013468,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 3.953772331948074,
      "grad_norm": 8.201769828796387,
      "learning_rate": 1.0419724012125053e-05,
      "loss": 2.4377384185791016,
      "memory(GiB)": 77.56,
      "step": 92285,
      "token_acc": 0.4758842443729904,
      "train_speed(iter/s)": 1.437944
    },
    {
      "epoch": 3.953986547277323,
      "grad_norm": 6.548315048217773,
      "learning_rate": 1.041561226068774e-05,
      "loss": 2.113050842285156,
      "memory(GiB)": 77.56,
      "step": 92290,
      "token_acc": 0.53156146179402,
      "train_speed(iter/s)": 1.437955
    },
    {
      "epoch": 3.9542007626065723,
      "grad_norm": 5.980075836181641,
      "learning_rate": 1.041150122635875e-05,
      "loss": 2.5393014907836915,
      "memory(GiB)": 77.56,
      "step": 92295,
      "token_acc": 0.4628099173553719,
      "train_speed(iter/s)": 1.437961
    },
    {
      "epoch": 3.954414977935821,
      "grad_norm": 5.306587219238281,
      "learning_rate": 1.0407390909212544e-05,
      "loss": 2.3953643798828126,
      "memory(GiB)": 77.56,
      "step": 92300,
      "token_acc": 0.5115511551155115,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 3.95462919326507,
      "grad_norm": 4.704372882843018,
      "learning_rate": 1.0403281309323587e-05,
      "loss": 2.6436405181884766,
      "memory(GiB)": 77.56,
      "step": 92305,
      "token_acc": 0.43440233236151604,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 3.954843408594319,
      "grad_norm": 5.839838027954102,
      "learning_rate": 1.0399172426766341e-05,
      "loss": 2.354899597167969,
      "memory(GiB)": 77.56,
      "step": 92310,
      "token_acc": 0.5105740181268882,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.955057623923568,
      "grad_norm": 7.160571575164795,
      "learning_rate": 1.039506426161524e-05,
      "loss": 2.512831115722656,
      "memory(GiB)": 77.56,
      "step": 92315,
      "token_acc": 0.49070631970260226,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.955271839252817,
      "grad_norm": 6.297356605529785,
      "learning_rate": 1.0390956813944697e-05,
      "loss": 2.2826522827148437,
      "memory(GiB)": 77.56,
      "step": 92320,
      "token_acc": 0.5016077170418006,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 3.955486054582066,
      "grad_norm": 6.495451927185059,
      "learning_rate": 1.038685008382913e-05,
      "loss": 2.4069850921630858,
      "memory(GiB)": 77.56,
      "step": 92325,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 3.955700269911315,
      "grad_norm": 5.984065055847168,
      "learning_rate": 1.038274407134292e-05,
      "loss": 2.331298065185547,
      "memory(GiB)": 77.56,
      "step": 92330,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.437974
    },
    {
      "epoch": 3.9559144852405637,
      "grad_norm": 12.185142517089844,
      "learning_rate": 1.0378638776560483e-05,
      "loss": 2.315629577636719,
      "memory(GiB)": 77.56,
      "step": 92335,
      "token_acc": 0.48616600790513836,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.956128700569813,
      "grad_norm": 5.467767715454102,
      "learning_rate": 1.0374534199556168e-05,
      "loss": 2.4507892608642576,
      "memory(GiB)": 77.56,
      "step": 92340,
      "token_acc": 0.5047619047619047,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 3.9563429158990617,
      "grad_norm": 7.182713985443115,
      "learning_rate": 1.0370430340404336e-05,
      "loss": 2.116982078552246,
      "memory(GiB)": 77.56,
      "step": 92345,
      "token_acc": 0.5634920634920635,
      "train_speed(iter/s)": 1.437966
    },
    {
      "epoch": 3.9565571312283105,
      "grad_norm": 5.4907755851745605,
      "learning_rate": 1.0366327199179338e-05,
      "loss": 2.619407844543457,
      "memory(GiB)": 77.56,
      "step": 92350,
      "token_acc": 0.45652173913043476,
      "train_speed(iter/s)": 1.437969
    },
    {
      "epoch": 3.95677134655756,
      "grad_norm": 5.2013421058654785,
      "learning_rate": 1.0362224775955486e-05,
      "loss": 2.4770137786865236,
      "memory(GiB)": 77.56,
      "step": 92355,
      "token_acc": 0.43727598566308246,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 3.9569855618868086,
      "grad_norm": 5.84731912612915,
      "learning_rate": 1.035812307080713e-05,
      "loss": 2.2171327590942385,
      "memory(GiB)": 77.56,
      "step": 92360,
      "token_acc": 0.5080385852090032,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 3.9571997772160574,
      "grad_norm": 7.181849479675293,
      "learning_rate": 1.0354022083808551e-05,
      "loss": 2.052858352661133,
      "memory(GiB)": 77.56,
      "step": 92365,
      "token_acc": 0.549520766773163,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 3.9574139925453067,
      "grad_norm": 5.451623916625977,
      "learning_rate": 1.0349921815034063e-05,
      "loss": 2.466476821899414,
      "memory(GiB)": 77.56,
      "step": 92370,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 3.9576282078745555,
      "grad_norm": 5.480225086212158,
      "learning_rate": 1.0345822264557936e-05,
      "loss": 2.6819793701171877,
      "memory(GiB)": 77.56,
      "step": 92375,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 3.9578424232038043,
      "grad_norm": 6.137053966522217,
      "learning_rate": 1.0341723432454437e-05,
      "loss": 2.3277715682983398,
      "memory(GiB)": 77.56,
      "step": 92380,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.437987
    },
    {
      "epoch": 3.9580566385330536,
      "grad_norm": 6.884665489196777,
      "learning_rate": 1.0337625318797822e-05,
      "loss": 2.2621673583984374,
      "memory(GiB)": 77.56,
      "step": 92385,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.9582708538623024,
      "grad_norm": 5.110538005828857,
      "learning_rate": 1.0333527923662318e-05,
      "loss": 2.251437187194824,
      "memory(GiB)": 77.56,
      "step": 92390,
      "token_acc": 0.5226586102719033,
      "train_speed(iter/s)": 1.437959
    },
    {
      "epoch": 3.958485069191551,
      "grad_norm": 7.737443923950195,
      "learning_rate": 1.0329431247122179e-05,
      "loss": 2.5528079986572267,
      "memory(GiB)": 77.56,
      "step": 92395,
      "token_acc": 0.45104895104895104,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 3.9586992845208004,
      "grad_norm": 6.978601932525635,
      "learning_rate": 1.0325335289251602e-05,
      "loss": 2.841652679443359,
      "memory(GiB)": 77.56,
      "step": 92400,
      "token_acc": 0.4268292682926829,
      "train_speed(iter/s)": 1.437964
    },
    {
      "epoch": 3.9589134998500493,
      "grad_norm": 5.924506664276123,
      "learning_rate": 1.0321240050124797e-05,
      "loss": 2.1365234375,
      "memory(GiB)": 77.56,
      "step": 92405,
      "token_acc": 0.5305164319248826,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.959127715179298,
      "grad_norm": 8.098037719726562,
      "learning_rate": 1.0317145529815952e-05,
      "loss": 2.109018898010254,
      "memory(GiB)": 77.56,
      "step": 92410,
      "token_acc": 0.548,
      "train_speed(iter/s)": 1.437993
    },
    {
      "epoch": 3.9593419305085473,
      "grad_norm": 6.439405918121338,
      "learning_rate": 1.0313051728399243e-05,
      "loss": 2.1036882400512695,
      "memory(GiB)": 77.56,
      "step": 92415,
      "token_acc": 0.5017064846416383,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 3.959556145837796,
      "grad_norm": 6.563499927520752,
      "learning_rate": 1.0308958645948814e-05,
      "loss": 2.0365978240966798,
      "memory(GiB)": 77.56,
      "step": 92420,
      "token_acc": 0.5909090909090909,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 3.959770361167045,
      "grad_norm": 6.739819049835205,
      "learning_rate": 1.030486628253885e-05,
      "loss": 2.4002443313598634,
      "memory(GiB)": 77.56,
      "step": 92425,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 3.959984576496294,
      "grad_norm": 5.5007781982421875,
      "learning_rate": 1.0300774638243454e-05,
      "loss": 2.4650896072387694,
      "memory(GiB)": 77.56,
      "step": 92430,
      "token_acc": 0.501628664495114,
      "train_speed(iter/s)": 1.438004
    },
    {
      "epoch": 3.960198791825543,
      "grad_norm": 4.279637813568115,
      "learning_rate": 1.0296683713136779e-05,
      "loss": 1.9827146530151367,
      "memory(GiB)": 77.56,
      "step": 92435,
      "token_acc": 0.5673758865248227,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 3.960413007154792,
      "grad_norm": 7.978905200958252,
      "learning_rate": 1.0292593507292924e-05,
      "loss": 2.2765472412109373,
      "memory(GiB)": 77.56,
      "step": 92440,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.437978
    },
    {
      "epoch": 3.960627222484041,
      "grad_norm": 5.935574054718018,
      "learning_rate": 1.0288504020785983e-05,
      "loss": 2.584377670288086,
      "memory(GiB)": 77.56,
      "step": 92445,
      "token_acc": 0.4402332361516035,
      "train_speed(iter/s)": 1.437981
    },
    {
      "epoch": 3.96084143781329,
      "grad_norm": 5.890320777893066,
      "learning_rate": 1.0284415253690032e-05,
      "loss": 2.4728418350219727,
      "memory(GiB)": 77.56,
      "step": 92450,
      "token_acc": 0.5078864353312302,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 3.961055653142539,
      "grad_norm": 4.8769683837890625,
      "learning_rate": 1.0280327206079172e-05,
      "loss": 1.948649787902832,
      "memory(GiB)": 77.56,
      "step": 92455,
      "token_acc": 0.5642857142857143,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 3.961269868471788,
      "grad_norm": 7.111137390136719,
      "learning_rate": 1.0276239878027443e-05,
      "loss": 2.459162712097168,
      "memory(GiB)": 77.56,
      "step": 92460,
      "token_acc": 0.49122807017543857,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 3.961484083801037,
      "grad_norm": 5.310895919799805,
      "learning_rate": 1.0272153269608892e-05,
      "loss": 2.1758729934692385,
      "memory(GiB)": 77.56,
      "step": 92465,
      "token_acc": 0.5396341463414634,
      "train_speed(iter/s)": 1.438016
    },
    {
      "epoch": 3.961698299130286,
      "grad_norm": 6.717132091522217,
      "learning_rate": 1.0268067380897556e-05,
      "loss": 2.1248960494995117,
      "memory(GiB)": 77.56,
      "step": 92470,
      "token_acc": 0.5321428571428571,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 3.961912514459535,
      "grad_norm": 5.88025426864624,
      "learning_rate": 1.0263982211967433e-05,
      "loss": 2.3237077713012697,
      "memory(GiB)": 77.56,
      "step": 92475,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 3.9621267297887837,
      "grad_norm": 6.892989635467529,
      "learning_rate": 1.0259897762892562e-05,
      "loss": 2.2981101989746096,
      "memory(GiB)": 77.56,
      "step": 92480,
      "token_acc": 0.5353159851301115,
      "train_speed(iter/s)": 1.438019
    },
    {
      "epoch": 3.962340945118033,
      "grad_norm": 5.83104133605957,
      "learning_rate": 1.0255814033746925e-05,
      "loss": 2.2894113540649412,
      "memory(GiB)": 77.56,
      "step": 92485,
      "token_acc": 0.48179271708683474,
      "train_speed(iter/s)": 1.438026
    },
    {
      "epoch": 3.9625551604472817,
      "grad_norm": 6.483590602874756,
      "learning_rate": 1.0251731024604495e-05,
      "loss": 2.3141902923583983,
      "memory(GiB)": 77.56,
      "step": 92490,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 3.9627693757765305,
      "grad_norm": 5.813948631286621,
      "learning_rate": 1.0247648735539245e-05,
      "loss": 2.0924434661865234,
      "memory(GiB)": 77.56,
      "step": 92495,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 3.96298359110578,
      "grad_norm": 7.791220664978027,
      "learning_rate": 1.0243567166625117e-05,
      "loss": 2.3897720336914063,
      "memory(GiB)": 77.56,
      "step": 92500,
      "token_acc": 0.458041958041958,
      "train_speed(iter/s)": 1.43804
    },
    {
      "epoch": 3.96298359110578,
      "eval_loss": 1.9803563356399536,
      "eval_runtime": 14.5488,
      "eval_samples_per_second": 6.873,
      "eval_steps_per_second": 6.873,
      "eval_token_acc": 0.5121212121212121,
      "step": 92500
    },
    {
      "epoch": 3.9631978064350286,
      "grad_norm": 5.973801136016846,
      "learning_rate": 1.0239486317936081e-05,
      "loss": 2.1581859588623047,
      "memory(GiB)": 77.56,
      "step": 92505,
      "token_acc": 0.5138004246284501,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 3.9634120217642774,
      "grad_norm": 5.572224140167236,
      "learning_rate": 1.0235406189546031e-05,
      "loss": 2.3361495971679687,
      "memory(GiB)": 77.56,
      "step": 92510,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.437706
    },
    {
      "epoch": 3.9636262370935267,
      "grad_norm": 5.9734978675842285,
      "learning_rate": 1.0231326781528916e-05,
      "loss": 2.18347225189209,
      "memory(GiB)": 77.56,
      "step": 92515,
      "token_acc": 0.49828178694158076,
      "train_speed(iter/s)": 1.437718
    },
    {
      "epoch": 3.9638404524227755,
      "grad_norm": 8.143786430358887,
      "learning_rate": 1.022724809395862e-05,
      "loss": 2.1307796478271483,
      "memory(GiB)": 77.56,
      "step": 92520,
      "token_acc": 0.5337837837837838,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.9640546677520243,
      "grad_norm": 5.335572242736816,
      "learning_rate": 1.0223170126909038e-05,
      "loss": 2.262030601501465,
      "memory(GiB)": 77.56,
      "step": 92525,
      "token_acc": 0.5349544072948328,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 3.9642688830812736,
      "grad_norm": 6.198302268981934,
      "learning_rate": 1.021909288045404e-05,
      "loss": 2.2171550750732423,
      "memory(GiB)": 77.56,
      "step": 92530,
      "token_acc": 0.5402298850574713,
      "train_speed(iter/s)": 1.437723
    },
    {
      "epoch": 3.9644830984105224,
      "grad_norm": 4.79813814163208,
      "learning_rate": 1.0215016354667477e-05,
      "loss": 2.2162036895751953,
      "memory(GiB)": 77.56,
      "step": 92535,
      "token_acc": 0.546583850931677,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 3.964697313739771,
      "grad_norm": 6.9610514640808105,
      "learning_rate": 1.0210940549623227e-05,
      "loss": 2.520808792114258,
      "memory(GiB)": 77.56,
      "step": 92540,
      "token_acc": 0.4778156996587031,
      "train_speed(iter/s)": 1.437707
    },
    {
      "epoch": 3.9649115290690204,
      "grad_norm": 6.620671272277832,
      "learning_rate": 1.020686546539512e-05,
      "loss": 2.3940929412841796,
      "memory(GiB)": 77.56,
      "step": 92545,
      "token_acc": 0.4866666666666667,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 3.9651257443982693,
      "grad_norm": 6.23718786239624,
      "learning_rate": 1.0202791102056969e-05,
      "loss": 2.2938634872436525,
      "memory(GiB)": 77.56,
      "step": 92550,
      "token_acc": 0.4885057471264368,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 3.965339959727518,
      "grad_norm": 5.441056251525879,
      "learning_rate": 1.0198717459682594e-05,
      "loss": 2.1483461380004885,
      "memory(GiB)": 77.56,
      "step": 92555,
      "token_acc": 0.5405405405405406,
      "train_speed(iter/s)": 1.437734
    },
    {
      "epoch": 3.9655541750567673,
      "grad_norm": 6.096147537231445,
      "learning_rate": 1.0194644538345787e-05,
      "loss": 2.575356674194336,
      "memory(GiB)": 77.56,
      "step": 92560,
      "token_acc": 0.47865853658536583,
      "train_speed(iter/s)": 1.437737
    },
    {
      "epoch": 3.965768390386016,
      "grad_norm": 5.840301036834717,
      "learning_rate": 1.0190572338120318e-05,
      "loss": 2.2437034606933595,
      "memory(GiB)": 77.56,
      "step": 92565,
      "token_acc": 0.5088967971530249,
      "train_speed(iter/s)": 1.43775
    },
    {
      "epoch": 3.965982605715265,
      "grad_norm": 6.897130966186523,
      "learning_rate": 1.0186500859079995e-05,
      "loss": 2.32926082611084,
      "memory(GiB)": 77.56,
      "step": 92570,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.966196821044514,
      "grad_norm": 5.886148452758789,
      "learning_rate": 1.018243010129854e-05,
      "loss": 2.294659233093262,
      "memory(GiB)": 77.56,
      "step": 92575,
      "token_acc": 0.5052264808362369,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 3.966411036373763,
      "grad_norm": 5.895969867706299,
      "learning_rate": 1.017836006484973e-05,
      "loss": 2.4033576965332033,
      "memory(GiB)": 77.56,
      "step": 92580,
      "token_acc": 0.47634069400630913,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.966625251703012,
      "grad_norm": 7.525404930114746,
      "learning_rate": 1.0174290749807281e-05,
      "loss": 2.4287345886230467,
      "memory(GiB)": 77.56,
      "step": 92585,
      "token_acc": 0.44360902255639095,
      "train_speed(iter/s)": 1.437755
    },
    {
      "epoch": 3.966839467032261,
      "grad_norm": 5.77642297744751,
      "learning_rate": 1.0170222156244914e-05,
      "loss": 2.517479133605957,
      "memory(GiB)": 77.56,
      "step": 92590,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.437767
    },
    {
      "epoch": 3.96705368236151,
      "grad_norm": 7.495279788970947,
      "learning_rate": 1.0166154284236324e-05,
      "loss": 2.8124372482299806,
      "memory(GiB)": 77.56,
      "step": 92595,
      "token_acc": 0.4411764705882353,
      "train_speed(iter/s)": 1.437768
    },
    {
      "epoch": 3.9672678976907587,
      "grad_norm": 4.941127777099609,
      "learning_rate": 1.016208713385523e-05,
      "loss": 2.3981298446655273,
      "memory(GiB)": 77.56,
      "step": 92600,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 3.967482113020008,
      "grad_norm": 6.305639743804932,
      "learning_rate": 1.0158020705175298e-05,
      "loss": 2.4088714599609373,
      "memory(GiB)": 77.56,
      "step": 92605,
      "token_acc": 0.4876325088339223,
      "train_speed(iter/s)": 1.437787
    },
    {
      "epoch": 3.967696328349257,
      "grad_norm": 6.702791690826416,
      "learning_rate": 1.01539549982702e-05,
      "loss": 2.3219625473022463,
      "memory(GiB)": 77.56,
      "step": 92610,
      "token_acc": 0.5434083601286174,
      "train_speed(iter/s)": 1.437788
    },
    {
      "epoch": 3.9679105436785056,
      "grad_norm": 5.356317520141602,
      "learning_rate": 1.014989001321358e-05,
      "loss": 2.5029346466064455,
      "memory(GiB)": 77.56,
      "step": 92615,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.437779
    },
    {
      "epoch": 3.968124759007755,
      "grad_norm": 5.056755065917969,
      "learning_rate": 1.0145825750079075e-05,
      "loss": 2.2346929550170898,
      "memory(GiB)": 77.56,
      "step": 92620,
      "token_acc": 0.5064935064935064,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 3.9683389743370037,
      "grad_norm": 4.774623870849609,
      "learning_rate": 1.0141762208940337e-05,
      "loss": 2.6570507049560548,
      "memory(GiB)": 77.56,
      "step": 92625,
      "token_acc": 0.4778481012658228,
      "train_speed(iter/s)": 1.437789
    },
    {
      "epoch": 3.9685531896662525,
      "grad_norm": 6.32444429397583,
      "learning_rate": 1.0137699389870963e-05,
      "loss": 2.4473831176757814,
      "memory(GiB)": 77.56,
      "step": 92630,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.437799
    },
    {
      "epoch": 3.9687674049955017,
      "grad_norm": 7.622020721435547,
      "learning_rate": 1.013363729294456e-05,
      "loss": 2.42104434967041,
      "memory(GiB)": 77.56,
      "step": 92635,
      "token_acc": 0.49800796812749004,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 3.9689816203247505,
      "grad_norm": 6.257937908172607,
      "learning_rate": 1.0129575918234713e-05,
      "loss": 2.2202287673950196,
      "memory(GiB)": 77.56,
      "step": 92640,
      "token_acc": 0.5435435435435435,
      "train_speed(iter/s)": 1.437797
    },
    {
      "epoch": 3.9691958356539994,
      "grad_norm": 5.764307498931885,
      "learning_rate": 1.0125515265815005e-05,
      "loss": 2.311808395385742,
      "memory(GiB)": 77.56,
      "step": 92645,
      "token_acc": 0.5323741007194245,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 3.9694100509832486,
      "grad_norm": 7.226253032684326,
      "learning_rate": 1.0121455335758978e-05,
      "loss": 2.1958854675292967,
      "memory(GiB)": 77.56,
      "step": 92650,
      "token_acc": 0.506578947368421,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 3.9696242663124974,
      "grad_norm": 6.088156223297119,
      "learning_rate": 1.0117396128140194e-05,
      "loss": 2.3840669631958007,
      "memory(GiB)": 77.56,
      "step": 92655,
      "token_acc": 0.49814126394052044,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 3.9698384816417462,
      "grad_norm": 6.461009979248047,
      "learning_rate": 1.0113337643032206e-05,
      "loss": 2.5208641052246095,
      "memory(GiB)": 77.56,
      "step": 92660,
      "token_acc": 0.4838709677419355,
      "train_speed(iter/s)": 1.437806
    },
    {
      "epoch": 3.9700526969709955,
      "grad_norm": 5.879950046539307,
      "learning_rate": 1.0109279880508527e-05,
      "loss": 2.29910945892334,
      "memory(GiB)": 77.56,
      "step": 92665,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 3.9702669123002443,
      "grad_norm": 5.410061836242676,
      "learning_rate": 1.010522284064266e-05,
      "loss": 2.1926912307739257,
      "memory(GiB)": 77.56,
      "step": 92670,
      "token_acc": 0.5466666666666666,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 3.970481127629493,
      "grad_norm": 5.707328796386719,
      "learning_rate": 1.0101166523508104e-05,
      "loss": 2.2371553421020507,
      "memory(GiB)": 77.56,
      "step": 92675,
      "token_acc": 0.48582995951417,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 3.9706953429587424,
      "grad_norm": 5.6087846755981445,
      "learning_rate": 1.0097110929178332e-05,
      "loss": 2.560465431213379,
      "memory(GiB)": 77.56,
      "step": 92680,
      "token_acc": 0.4856115107913669,
      "train_speed(iter/s)": 1.437842
    },
    {
      "epoch": 3.970909558287991,
      "grad_norm": 6.74299430847168,
      "learning_rate": 1.009305605772684e-05,
      "loss": 2.0412059783935548,
      "memory(GiB)": 77.56,
      "step": 92685,
      "token_acc": 0.5539033457249071,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 3.97112377361724,
      "grad_norm": 5.704792499542236,
      "learning_rate": 1.008900190922708e-05,
      "loss": 2.361504554748535,
      "memory(GiB)": 77.56,
      "step": 92690,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 3.9713379889464893,
      "grad_norm": 5.117986679077148,
      "learning_rate": 1.0084948483752482e-05,
      "loss": 2.521312713623047,
      "memory(GiB)": 77.56,
      "step": 92695,
      "token_acc": 0.4791666666666667,
      "train_speed(iter/s)": 1.437825
    },
    {
      "epoch": 3.971552204275738,
      "grad_norm": 8.231768608093262,
      "learning_rate": 1.0080895781376493e-05,
      "loss": 2.7389270782470705,
      "memory(GiB)": 77.56,
      "step": 92700,
      "token_acc": 0.4358974358974359,
      "train_speed(iter/s)": 1.437851
    },
    {
      "epoch": 3.971766419604987,
      "grad_norm": 6.550184726715088,
      "learning_rate": 1.007684380217252e-05,
      "loss": 2.3109886169433596,
      "memory(GiB)": 77.56,
      "step": 92705,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 3.971980634934236,
      "grad_norm": 5.8178629875183105,
      "learning_rate": 1.007279254621396e-05,
      "loss": 2.345347785949707,
      "memory(GiB)": 77.56,
      "step": 92710,
      "token_acc": 0.5030120481927711,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 3.972194850263485,
      "grad_norm": 6.1667866706848145,
      "learning_rate": 1.0068742013574229e-05,
      "loss": 2.200144577026367,
      "memory(GiB)": 77.56,
      "step": 92715,
      "token_acc": 0.5234375,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 3.9724090655927338,
      "grad_norm": 5.679183006286621,
      "learning_rate": 1.0064692204326699e-05,
      "loss": 2.0072067260742186,
      "memory(GiB)": 77.56,
      "step": 92720,
      "token_acc": 0.5838926174496645,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 3.972623280921983,
      "grad_norm": 6.569579601287842,
      "learning_rate": 1.006064311854471e-05,
      "loss": 2.315715026855469,
      "memory(GiB)": 77.56,
      "step": 92725,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.437859
    },
    {
      "epoch": 3.972837496251232,
      "grad_norm": 5.709552764892578,
      "learning_rate": 1.0056594756301658e-05,
      "loss": 2.257454681396484,
      "memory(GiB)": 77.56,
      "step": 92730,
      "token_acc": 0.5214007782101168,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 3.9730517115804806,
      "grad_norm": 5.594107627868652,
      "learning_rate": 1.0052547117670863e-05,
      "loss": 2.382841110229492,
      "memory(GiB)": 77.56,
      "step": 92735,
      "token_acc": 0.4896142433234421,
      "train_speed(iter/s)": 1.437876
    },
    {
      "epoch": 3.97326592690973,
      "grad_norm": 5.784968852996826,
      "learning_rate": 1.004850020272563e-05,
      "loss": 2.4212446212768555,
      "memory(GiB)": 77.56,
      "step": 92740,
      "token_acc": 0.4785100286532951,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 3.9734801422389787,
      "grad_norm": 6.823126316070557,
      "learning_rate": 1.0044454011539317e-05,
      "loss": 2.487283706665039,
      "memory(GiB)": 77.56,
      "step": 92745,
      "token_acc": 0.46308724832214765,
      "train_speed(iter/s)": 1.437876
    },
    {
      "epoch": 3.9736943575682275,
      "grad_norm": 6.894071102142334,
      "learning_rate": 1.0040408544185192e-05,
      "loss": 2.0663173675537108,
      "memory(GiB)": 77.56,
      "step": 92750,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.437868
    },
    {
      "epoch": 3.973908572897477,
      "grad_norm": 5.159613132476807,
      "learning_rate": 1.0036363800736554e-05,
      "loss": 2.4883251190185547,
      "memory(GiB)": 77.56,
      "step": 92755,
      "token_acc": 0.5046728971962616,
      "train_speed(iter/s)": 1.437863
    },
    {
      "epoch": 3.9741227882267256,
      "grad_norm": 6.901095390319824,
      "learning_rate": 1.0032319781266674e-05,
      "loss": 2.4273557662963867,
      "memory(GiB)": 77.56,
      "step": 92760,
      "token_acc": 0.4912891986062718,
      "train_speed(iter/s)": 1.437861
    },
    {
      "epoch": 3.9743370035559744,
      "grad_norm": 4.95836877822876,
      "learning_rate": 1.0028276485848803e-05,
      "loss": 2.6295169830322265,
      "memory(GiB)": 77.56,
      "step": 92765,
      "token_acc": 0.4642857142857143,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 3.9745512188852237,
      "grad_norm": 9.252575874328613,
      "learning_rate": 1.0024233914556213e-05,
      "loss": 2.670606994628906,
      "memory(GiB)": 77.56,
      "step": 92770,
      "token_acc": 0.44054054054054054,
      "train_speed(iter/s)": 1.437874
    },
    {
      "epoch": 3.9747654342144725,
      "grad_norm": 5.035546779632568,
      "learning_rate": 1.0020192067462125e-05,
      "loss": 2.4046119689941405,
      "memory(GiB)": 77.56,
      "step": 92775,
      "token_acc": 0.4632768361581921,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 3.9749796495437213,
      "grad_norm": 5.519577503204346,
      "learning_rate": 1.0016150944639763e-05,
      "loss": 2.6053440093994142,
      "memory(GiB)": 77.56,
      "step": 92780,
      "token_acc": 0.45161290322580644,
      "train_speed(iter/s)": 1.437891
    },
    {
      "epoch": 3.9751938648729706,
      "grad_norm": 5.951094150543213,
      "learning_rate": 1.0012110546162334e-05,
      "loss": 2.1103424072265624,
      "memory(GiB)": 77.56,
      "step": 92785,
      "token_acc": 0.5580736543909348,
      "train_speed(iter/s)": 1.437887
    },
    {
      "epoch": 3.9754080802022194,
      "grad_norm": 5.941181659698486,
      "learning_rate": 1.0008070872103032e-05,
      "loss": 2.21679630279541,
      "memory(GiB)": 77.56,
      "step": 92790,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.437898
    },
    {
      "epoch": 3.975622295531468,
      "grad_norm": 5.627597332000732,
      "learning_rate": 1.0004031922535029e-05,
      "loss": 2.382151412963867,
      "memory(GiB)": 77.56,
      "step": 92795,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 3.9758365108607174,
      "grad_norm": 5.793582916259766,
      "learning_rate": 9.999993697531512e-06,
      "loss": 2.309271240234375,
      "memory(GiB)": 77.56,
      "step": 92800,
      "token_acc": 0.48909657320872274,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.9760507261899662,
      "grad_norm": 5.2836833000183105,
      "learning_rate": 9.99595619716564e-06,
      "loss": 2.372360038757324,
      "memory(GiB)": 77.56,
      "step": 92805,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.437915
    },
    {
      "epoch": 3.976264941519215,
      "grad_norm": 7.119791030883789,
      "learning_rate": 9.991919421510547e-06,
      "loss": 2.4161401748657227,
      "memory(GiB)": 77.56,
      "step": 92810,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.437893
    },
    {
      "epoch": 3.9764791568484643,
      "grad_norm": 5.952172756195068,
      "learning_rate": 9.987883370639368e-06,
      "loss": 1.9757646560668944,
      "memory(GiB)": 77.56,
      "step": 92815,
      "token_acc": 0.5891472868217055,
      "train_speed(iter/s)": 1.437897
    },
    {
      "epoch": 3.976693372177713,
      "grad_norm": 7.871514320373535,
      "learning_rate": 9.983848044625216e-06,
      "loss": 2.3188039779663088,
      "memory(GiB)": 77.56,
      "step": 92820,
      "token_acc": 0.49201277955271566,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 3.976907587506962,
      "grad_norm": 4.459116458892822,
      "learning_rate": 9.97981344354118e-06,
      "loss": 2.309185028076172,
      "memory(GiB)": 77.56,
      "step": 92825,
      "token_acc": 0.5210727969348659,
      "train_speed(iter/s)": 1.437908
    },
    {
      "epoch": 3.977121802836211,
      "grad_norm": 4.845858573913574,
      "learning_rate": 9.97577956746038e-06,
      "loss": 2.2914608001708983,
      "memory(GiB)": 77.56,
      "step": 92830,
      "token_acc": 0.4969512195121951,
      "train_speed(iter/s)": 1.437915
    },
    {
      "epoch": 3.97733601816546,
      "grad_norm": 6.601243495941162,
      "learning_rate": 9.971746416455879e-06,
      "loss": 2.3751863479614257,
      "memory(GiB)": 77.56,
      "step": 92835,
      "token_acc": 0.5308219178082192,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 3.977550233494709,
      "grad_norm": 7.530718803405762,
      "learning_rate": 9.96771399060074e-06,
      "loss": 2.3008981704711915,
      "memory(GiB)": 77.56,
      "step": 92840,
      "token_acc": 0.5055762081784386,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 3.977764448823958,
      "grad_norm": 7.323751449584961,
      "learning_rate": 9.963682289968018e-06,
      "loss": 2.1395172119140624,
      "memory(GiB)": 77.56,
      "step": 92845,
      "token_acc": 0.5019762845849802,
      "train_speed(iter/s)": 1.437928
    },
    {
      "epoch": 3.977978664153207,
      "grad_norm": 5.858083248138428,
      "learning_rate": 9.959651314630747e-06,
      "loss": 2.4996669769287108,
      "memory(GiB)": 77.56,
      "step": 92850,
      "token_acc": 0.4690909090909091,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 3.9781928794824557,
      "grad_norm": 5.75156307220459,
      "learning_rate": 9.955621064661935e-06,
      "loss": 2.37884578704834,
      "memory(GiB)": 77.56,
      "step": 92855,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.437977
    },
    {
      "epoch": 3.978407094811705,
      "grad_norm": 6.420896530151367,
      "learning_rate": 9.95159154013463e-06,
      "loss": 2.270208168029785,
      "memory(GiB)": 77.56,
      "step": 92860,
      "token_acc": 0.5034246575342466,
      "train_speed(iter/s)": 1.437959
    },
    {
      "epoch": 3.9786213101409538,
      "grad_norm": 5.730321407318115,
      "learning_rate": 9.947562741121813e-06,
      "loss": 2.3636148452758787,
      "memory(GiB)": 77.56,
      "step": 92865,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.437961
    },
    {
      "epoch": 3.9788355254702026,
      "grad_norm": 6.030406951904297,
      "learning_rate": 9.943534667696459e-06,
      "loss": 2.534542274475098,
      "memory(GiB)": 77.56,
      "step": 92870,
      "token_acc": 0.4576271186440678,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 3.979049740799452,
      "grad_norm": 4.293148994445801,
      "learning_rate": 9.939507319931557e-06,
      "loss": 2.248686981201172,
      "memory(GiB)": 77.56,
      "step": 92875,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 3.9792639561287007,
      "grad_norm": 6.239877223968506,
      "learning_rate": 9.935480697900063e-06,
      "loss": 2.353713607788086,
      "memory(GiB)": 77.56,
      "step": 92880,
      "token_acc": 0.5111821086261981,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 3.9794781714579495,
      "grad_norm": 6.3617682456970215,
      "learning_rate": 9.931454801674906e-06,
      "loss": 2.4965335845947267,
      "memory(GiB)": 77.56,
      "step": 92885,
      "token_acc": 0.4664429530201342,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 3.9796923867871987,
      "grad_norm": 7.433448314666748,
      "learning_rate": 9.92742963132905e-06,
      "loss": 2.303684616088867,
      "memory(GiB)": 77.56,
      "step": 92890,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 3.9799066021164475,
      "grad_norm": 6.065186977386475,
      "learning_rate": 9.923405186935392e-06,
      "loss": 2.179836654663086,
      "memory(GiB)": 77.56,
      "step": 92895,
      "token_acc": 0.5336134453781513,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 3.9801208174456963,
      "grad_norm": 5.21151876449585,
      "learning_rate": 9.919381468566846e-06,
      "loss": 2.4848758697509767,
      "memory(GiB)": 77.56,
      "step": 92900,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.437981
    },
    {
      "epoch": 3.9803350327749456,
      "grad_norm": 6.633542060852051,
      "learning_rate": 9.9153584762963e-06,
      "loss": 2.4606199264526367,
      "memory(GiB)": 77.56,
      "step": 92905,
      "token_acc": 0.4630225080385852,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.9805492481041944,
      "grad_norm": 5.363124847412109,
      "learning_rate": 9.911336210196626e-06,
      "loss": 2.0713830947875977,
      "memory(GiB)": 77.56,
      "step": 92910,
      "token_acc": 0.5266666666666666,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 3.9807634634334432,
      "grad_norm": 6.806240081787109,
      "learning_rate": 9.907314670340717e-06,
      "loss": 2.5236743927001952,
      "memory(GiB)": 77.56,
      "step": 92915,
      "token_acc": 0.4676923076923077,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 3.9809776787626925,
      "grad_norm": 5.791401386260986,
      "learning_rate": 9.903293856801415e-06,
      "loss": 2.592616081237793,
      "memory(GiB)": 77.56,
      "step": 92920,
      "token_acc": 0.444794952681388,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 3.9811918940919413,
      "grad_norm": 8.250869750976562,
      "learning_rate": 9.899273769651557e-06,
      "loss": 2.500656509399414,
      "memory(GiB)": 77.56,
      "step": 92925,
      "token_acc": 0.46863468634686345,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 3.98140610942119,
      "grad_norm": 6.224512577056885,
      "learning_rate": 9.895254408963972e-06,
      "loss": 2.297273063659668,
      "memory(GiB)": 77.56,
      "step": 92930,
      "token_acc": 0.5077881619937694,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 3.9816203247504394,
      "grad_norm": 5.6414923667907715,
      "learning_rate": 9.891235774811474e-06,
      "loss": 2.342025566101074,
      "memory(GiB)": 77.56,
      "step": 92935,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.437993
    },
    {
      "epoch": 3.981834540079688,
      "grad_norm": 6.357151985168457,
      "learning_rate": 9.88721786726685e-06,
      "loss": 2.497633934020996,
      "memory(GiB)": 77.56,
      "step": 92940,
      "token_acc": 0.5178571428571429,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 3.982048755408937,
      "grad_norm": 5.937796592712402,
      "learning_rate": 9.883200686402905e-06,
      "loss": 2.355463981628418,
      "memory(GiB)": 77.56,
      "step": 92945,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 3.9822629707381862,
      "grad_norm": 5.1598124504089355,
      "learning_rate": 9.879184232292426e-06,
      "loss": 2.229428291320801,
      "memory(GiB)": 77.56,
      "step": 92950,
      "token_acc": 0.49673202614379086,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 3.982477186067435,
      "grad_norm": 7.349733352661133,
      "learning_rate": 9.875168505008159e-06,
      "loss": 2.4955867767333983,
      "memory(GiB)": 77.56,
      "step": 92955,
      "token_acc": 0.5246478873239436,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 3.982691401396684,
      "grad_norm": 5.181321620941162,
      "learning_rate": 9.87115350462286e-06,
      "loss": 2.5198835372924804,
      "memory(GiB)": 77.56,
      "step": 92960,
      "token_acc": 0.4597315436241611,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.982905616725933,
      "grad_norm": 8.551186561584473,
      "learning_rate": 9.867139231209255e-06,
      "loss": 2.369637680053711,
      "memory(GiB)": 77.56,
      "step": 92965,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 3.983119832055182,
      "grad_norm": 5.37523078918457,
      "learning_rate": 9.863125684840059e-06,
      "loss": 2.2184019088745117,
      "memory(GiB)": 77.56,
      "step": 92970,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.9833340473844308,
      "grad_norm": 5.884568214416504,
      "learning_rate": 9.859112865588e-06,
      "loss": 2.433843994140625,
      "memory(GiB)": 77.56,
      "step": 92975,
      "token_acc": 0.5206349206349207,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 3.98354826271368,
      "grad_norm": 4.879148006439209,
      "learning_rate": 9.855100773525771e-06,
      "loss": 2.370548439025879,
      "memory(GiB)": 77.56,
      "step": 92980,
      "token_acc": 0.49441340782122906,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 3.983762478042929,
      "grad_norm": 5.4018330574035645,
      "learning_rate": 9.851089408726055e-06,
      "loss": 2.40179328918457,
      "memory(GiB)": 77.56,
      "step": 92985,
      "token_acc": 0.46688741721854304,
      "train_speed(iter/s)": 1.437993
    },
    {
      "epoch": 3.9839766933721776,
      "grad_norm": 5.380850315093994,
      "learning_rate": 9.847078771261509e-06,
      "loss": 2.7432432174682617,
      "memory(GiB)": 77.56,
      "step": 92990,
      "token_acc": 0.41317365269461076,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 3.984190908701427,
      "grad_norm": 7.40653133392334,
      "learning_rate": 9.843068861204801e-06,
      "loss": 2.4660552978515624,
      "memory(GiB)": 77.56,
      "step": 92995,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 3.9844051240306757,
      "grad_norm": 6.670955181121826,
      "learning_rate": 9.839059678628559e-06,
      "loss": 2.178139877319336,
      "memory(GiB)": 77.56,
      "step": 93000,
      "token_acc": 0.528169014084507,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 3.9844051240306757,
      "eval_loss": 2.020324945449829,
      "eval_runtime": 14.5715,
      "eval_samples_per_second": 6.863,
      "eval_steps_per_second": 6.863,
      "eval_token_acc": 0.510989010989011,
      "step": 93000
    },
    {
      "epoch": 3.9846193393599245,
      "grad_norm": 6.374236583709717,
      "learning_rate": 9.83505122360544e-06,
      "loss": 2.3164394378662108,
      "memory(GiB)": 77.56,
      "step": 93005,
      "token_acc": 0.5042735042735043,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 3.9848335546891738,
      "grad_norm": 5.763950347900391,
      "learning_rate": 9.831043496208042e-06,
      "loss": 2.2330081939697264,
      "memory(GiB)": 77.56,
      "step": 93010,
      "token_acc": 0.5268817204301075,
      "train_speed(iter/s)": 1.437643
    },
    {
      "epoch": 3.9850477700184226,
      "grad_norm": 5.373263359069824,
      "learning_rate": 9.827036496508973e-06,
      "loss": 2.189457321166992,
      "memory(GiB)": 77.56,
      "step": 93015,
      "token_acc": 0.5236486486486487,
      "train_speed(iter/s)": 1.437655
    },
    {
      "epoch": 3.9852619853476714,
      "grad_norm": 5.882124423980713,
      "learning_rate": 9.823030224580809e-06,
      "loss": 2.4792526245117186,
      "memory(GiB)": 77.56,
      "step": 93020,
      "token_acc": 0.48242811501597443,
      "train_speed(iter/s)": 1.437637
    },
    {
      "epoch": 3.9854762006769207,
      "grad_norm": 6.168560028076172,
      "learning_rate": 9.819024680496159e-06,
      "loss": 2.3294500350952148,
      "memory(GiB)": 77.56,
      "step": 93025,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.437635
    },
    {
      "epoch": 3.9856904160061695,
      "grad_norm": 4.794861793518066,
      "learning_rate": 9.815019864327551e-06,
      "loss": 2.4881208419799803,
      "memory(GiB)": 77.56,
      "step": 93030,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 3.9859046313354183,
      "grad_norm": 4.418397426605225,
      "learning_rate": 9.81101577614757e-06,
      "loss": 2.4073436737060545,
      "memory(GiB)": 77.56,
      "step": 93035,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.437645
    },
    {
      "epoch": 3.9861188466646675,
      "grad_norm": 5.007444381713867,
      "learning_rate": 9.807012416028739e-06,
      "loss": 2.1383548736572267,
      "memory(GiB)": 77.56,
      "step": 93040,
      "token_acc": 0.5189393939393939,
      "train_speed(iter/s)": 1.437648
    },
    {
      "epoch": 3.9863330619939163,
      "grad_norm": 6.416004657745361,
      "learning_rate": 9.803009784043581e-06,
      "loss": 2.5978683471679687,
      "memory(GiB)": 77.56,
      "step": 93045,
      "token_acc": 0.47337278106508873,
      "train_speed(iter/s)": 1.437653
    },
    {
      "epoch": 3.986547277323165,
      "grad_norm": 5.443553447723389,
      "learning_rate": 9.799007880264611e-06,
      "loss": 2.232929801940918,
      "memory(GiB)": 77.56,
      "step": 93050,
      "token_acc": 0.5195195195195195,
      "train_speed(iter/s)": 1.437666
    },
    {
      "epoch": 3.9867614926524144,
      "grad_norm": 7.6471381187438965,
      "learning_rate": 9.79500670476431e-06,
      "loss": 2.5438961029052733,
      "memory(GiB)": 77.56,
      "step": 93055,
      "token_acc": 0.48641304347826086,
      "train_speed(iter/s)": 1.437676
    },
    {
      "epoch": 3.9869757079816632,
      "grad_norm": 5.956918239593506,
      "learning_rate": 9.791006257615192e-06,
      "loss": 2.233340835571289,
      "memory(GiB)": 77.56,
      "step": 93060,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.437676
    },
    {
      "epoch": 3.987189923310912,
      "grad_norm": 7.244227886199951,
      "learning_rate": 9.787006538889714e-06,
      "loss": 2.315345573425293,
      "memory(GiB)": 77.56,
      "step": 93065,
      "token_acc": 0.5256410256410257,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 3.9874041386401613,
      "grad_norm": 5.253133296966553,
      "learning_rate": 9.783007548660338e-06,
      "loss": 2.043259048461914,
      "memory(GiB)": 77.56,
      "step": 93070,
      "token_acc": 0.5480427046263345,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 3.98761835396941,
      "grad_norm": 5.82855749130249,
      "learning_rate": 9.779009286999501e-06,
      "loss": 2.2444684982299803,
      "memory(GiB)": 77.56,
      "step": 93075,
      "token_acc": 0.5487804878048781,
      "train_speed(iter/s)": 1.437697
    },
    {
      "epoch": 3.987832569298659,
      "grad_norm": 4.557111740112305,
      "learning_rate": 9.775011753979646e-06,
      "loss": 2.2299251556396484,
      "memory(GiB)": 77.56,
      "step": 93080,
      "token_acc": 0.562962962962963,
      "train_speed(iter/s)": 1.437693
    },
    {
      "epoch": 3.988046784627908,
      "grad_norm": 6.188343524932861,
      "learning_rate": 9.771014949673174e-06,
      "loss": 2.198384666442871,
      "memory(GiB)": 77.56,
      "step": 93085,
      "token_acc": 0.47509578544061304,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 3.988260999957157,
      "grad_norm": 7.0401082038879395,
      "learning_rate": 9.767018874152517e-06,
      "loss": 2.136541748046875,
      "memory(GiB)": 77.56,
      "step": 93090,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437709
    },
    {
      "epoch": 3.988475215286406,
      "grad_norm": 5.74680757522583,
      "learning_rate": 9.76302352749004e-06,
      "loss": 2.3570327758789062,
      "memory(GiB)": 77.56,
      "step": 93095,
      "token_acc": 0.5213414634146342,
      "train_speed(iter/s)": 1.437723
    },
    {
      "epoch": 3.988689430615655,
      "grad_norm": 19.408462524414062,
      "learning_rate": 9.75902890975815e-06,
      "loss": 2.376866912841797,
      "memory(GiB)": 77.56,
      "step": 93100,
      "token_acc": 0.532319391634981,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 3.988903645944904,
      "grad_norm": 5.173044204711914,
      "learning_rate": 9.755035021029202e-06,
      "loss": 2.2135377883911134,
      "memory(GiB)": 77.56,
      "step": 93105,
      "token_acc": 0.525974025974026,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.9891178612741527,
      "grad_norm": 6.243752956390381,
      "learning_rate": 9.751041861375549e-06,
      "loss": 2.541492462158203,
      "memory(GiB)": 77.56,
      "step": 93110,
      "token_acc": 0.5160256410256411,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 3.989332076603402,
      "grad_norm": 8.36265754699707,
      "learning_rate": 9.74704943086951e-06,
      "loss": 2.5464977264404296,
      "memory(GiB)": 77.56,
      "step": 93115,
      "token_acc": 0.475177304964539,
      "train_speed(iter/s)": 1.437718
    },
    {
      "epoch": 3.9895462919326508,
      "grad_norm": 5.7020978927612305,
      "learning_rate": 9.74305772958345e-06,
      "loss": 2.017613983154297,
      "memory(GiB)": 77.56,
      "step": 93120,
      "token_acc": 0.5720338983050848,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 3.9897605072618996,
      "grad_norm": 6.958425998687744,
      "learning_rate": 9.73906675758966e-06,
      "loss": 2.670576477050781,
      "memory(GiB)": 77.56,
      "step": 93125,
      "token_acc": 0.4405594405594406,
      "train_speed(iter/s)": 1.437698
    },
    {
      "epoch": 3.989974722591149,
      "grad_norm": 6.194924354553223,
      "learning_rate": 9.735076514960445e-06,
      "loss": 2.3503650665283202,
      "memory(GiB)": 77.56,
      "step": 93130,
      "token_acc": 0.5136186770428015,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 3.9901889379203976,
      "grad_norm": 5.379251003265381,
      "learning_rate": 9.731087001768085e-06,
      "loss": 2.2071304321289062,
      "memory(GiB)": 77.56,
      "step": 93135,
      "token_acc": 0.53156146179402,
      "train_speed(iter/s)": 1.437718
    },
    {
      "epoch": 3.9904031532496465,
      "grad_norm": 7.103784561157227,
      "learning_rate": 9.727098218084868e-06,
      "loss": 2.3284521102905273,
      "memory(GiB)": 77.56,
      "step": 93140,
      "token_acc": 0.5121212121212121,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 3.9906173685788957,
      "grad_norm": 4.897048473358154,
      "learning_rate": 9.72311016398303e-06,
      "loss": 2.3074663162231444,
      "memory(GiB)": 77.56,
      "step": 93145,
      "token_acc": 0.514018691588785,
      "train_speed(iter/s)": 1.437746
    },
    {
      "epoch": 3.9908315839081445,
      "grad_norm": 6.873358249664307,
      "learning_rate": 9.719122839534844e-06,
      "loss": 2.3732961654663085,
      "memory(GiB)": 77.56,
      "step": 93150,
      "token_acc": 0.48985507246376814,
      "train_speed(iter/s)": 1.437747
    },
    {
      "epoch": 3.9910457992373933,
      "grad_norm": 5.830979824066162,
      "learning_rate": 9.715136244812533e-06,
      "loss": 2.592213821411133,
      "memory(GiB)": 77.56,
      "step": 93155,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.437744
    },
    {
      "epoch": 3.9912600145666426,
      "grad_norm": 5.25096321105957,
      "learning_rate": 9.711150379888328e-06,
      "loss": 2.093849945068359,
      "memory(GiB)": 77.56,
      "step": 93160,
      "token_acc": 0.5298245614035088,
      "train_speed(iter/s)": 1.437754
    },
    {
      "epoch": 3.9914742298958914,
      "grad_norm": 5.936777591705322,
      "learning_rate": 9.70716524483441e-06,
      "loss": 2.481352424621582,
      "memory(GiB)": 77.56,
      "step": 93165,
      "token_acc": 0.47126436781609193,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 3.99168844522514,
      "grad_norm": 9.324991226196289,
      "learning_rate": 9.703180839723003e-06,
      "loss": 2.2618932723999023,
      "memory(GiB)": 77.56,
      "step": 93170,
      "token_acc": 0.5049833887043189,
      "train_speed(iter/s)": 1.437752
    },
    {
      "epoch": 3.9919026605543895,
      "grad_norm": 5.5073041915893555,
      "learning_rate": 9.699197164626261e-06,
      "loss": 2.340204048156738,
      "memory(GiB)": 77.56,
      "step": 93175,
      "token_acc": 0.4723926380368098,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 3.9921168758836383,
      "grad_norm": 5.616020202636719,
      "learning_rate": 9.695214219616383e-06,
      "loss": 2.6657670974731444,
      "memory(GiB)": 77.56,
      "step": 93180,
      "token_acc": 0.4300341296928328,
      "train_speed(iter/s)": 1.43772
    },
    {
      "epoch": 3.992331091212887,
      "grad_norm": 4.7828369140625,
      "learning_rate": 9.69123200476551e-06,
      "loss": 2.3973552703857424,
      "memory(GiB)": 77.56,
      "step": 93185,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.437738
    },
    {
      "epoch": 3.9925453065421364,
      "grad_norm": 5.4492573738098145,
      "learning_rate": 9.687250520145785e-06,
      "loss": 2.4097145080566404,
      "memory(GiB)": 77.56,
      "step": 93190,
      "token_acc": 0.5092250922509225,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.992759521871385,
      "grad_norm": 4.831192493438721,
      "learning_rate": 9.683269765829328e-06,
      "loss": 2.4074705123901365,
      "memory(GiB)": 77.56,
      "step": 93195,
      "token_acc": 0.46417445482866043,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 3.992973737200634,
      "grad_norm": 5.701098918914795,
      "learning_rate": 9.679289741888248e-06,
      "loss": 2.3720630645751952,
      "memory(GiB)": 77.56,
      "step": 93200,
      "token_acc": 0.5162337662337663,
      "train_speed(iter/s)": 1.437715
    },
    {
      "epoch": 3.9931879525298832,
      "grad_norm": 5.549937725067139,
      "learning_rate": 9.67531044839467e-06,
      "loss": 2.468387794494629,
      "memory(GiB)": 77.56,
      "step": 93205,
      "token_acc": 0.4296028880866426,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.993402167859132,
      "grad_norm": 6.998640537261963,
      "learning_rate": 9.67133188542067e-06,
      "loss": 2.3003664016723633,
      "memory(GiB)": 77.56,
      "step": 93210,
      "token_acc": 0.5019157088122606,
      "train_speed(iter/s)": 1.43772
    },
    {
      "epoch": 3.993616383188381,
      "grad_norm": 6.147166728973389,
      "learning_rate": 9.667354053038329e-06,
      "loss": 2.1802204132080076,
      "memory(GiB)": 77.56,
      "step": 93215,
      "token_acc": 0.5493421052631579,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 3.99383059851763,
      "grad_norm": 5.608270645141602,
      "learning_rate": 9.663376951319702e-06,
      "loss": 1.9577875137329102,
      "memory(GiB)": 77.56,
      "step": 93220,
      "token_acc": 0.5537190082644629,
      "train_speed(iter/s)": 1.43774
    },
    {
      "epoch": 3.994044813846879,
      "grad_norm": 5.10335636138916,
      "learning_rate": 9.659400580336836e-06,
      "loss": 2.0693607330322266,
      "memory(GiB)": 77.56,
      "step": 93225,
      "token_acc": 0.5616438356164384,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.9942590291761277,
      "grad_norm": 5.3745436668396,
      "learning_rate": 9.655424940161761e-06,
      "loss": 2.2429805755615235,
      "memory(GiB)": 77.56,
      "step": 93230,
      "token_acc": 0.5016501650165016,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 3.994473244505377,
      "grad_norm": 4.842333793640137,
      "learning_rate": 9.651450030866527e-06,
      "loss": 2.092479705810547,
      "memory(GiB)": 77.56,
      "step": 93235,
      "token_acc": 0.5173501577287066,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 3.994687459834626,
      "grad_norm": 6.195989608764648,
      "learning_rate": 9.647475852523109e-06,
      "loss": 2.0949399948120115,
      "memory(GiB)": 77.56,
      "step": 93240,
      "token_acc": 0.5303030303030303,
      "train_speed(iter/s)": 1.437717
    },
    {
      "epoch": 3.9949016751638746,
      "grad_norm": 5.4378275871276855,
      "learning_rate": 9.64350240520353e-06,
      "loss": 2.0547292709350584,
      "memory(GiB)": 77.56,
      "step": 93245,
      "token_acc": 0.5273972602739726,
      "train_speed(iter/s)": 1.437716
    },
    {
      "epoch": 3.995115890493124,
      "grad_norm": 4.592390060424805,
      "learning_rate": 9.639529688979765e-06,
      "loss": 2.140955352783203,
      "memory(GiB)": 77.56,
      "step": 93250,
      "token_acc": 0.5402298850574713,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 3.9953301058223727,
      "grad_norm": 5.867456436157227,
      "learning_rate": 9.63555770392378e-06,
      "loss": 2.4671438217163084,
      "memory(GiB)": 77.56,
      "step": 93255,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.437727
    },
    {
      "epoch": 3.9955443211516215,
      "grad_norm": 4.547386169433594,
      "learning_rate": 9.63158645010752e-06,
      "loss": 2.209004592895508,
      "memory(GiB)": 77.56,
      "step": 93260,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.9957585364808708,
      "grad_norm": 4.398655414581299,
      "learning_rate": 9.627615927602957e-06,
      "loss": 2.3498098373413088,
      "memory(GiB)": 77.56,
      "step": 93265,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.437708
    },
    {
      "epoch": 3.9959727518101196,
      "grad_norm": 7.089205265045166,
      "learning_rate": 9.623646136482e-06,
      "loss": 2.0108156204223633,
      "memory(GiB)": 77.56,
      "step": 93270,
      "token_acc": 0.5541666666666667,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 3.9961869671393684,
      "grad_norm": 5.16269063949585,
      "learning_rate": 9.619677076816568e-06,
      "loss": 2.542580223083496,
      "memory(GiB)": 77.56,
      "step": 93275,
      "token_acc": 0.4822485207100592,
      "train_speed(iter/s)": 1.437707
    },
    {
      "epoch": 3.9964011824686176,
      "grad_norm": 6.381258964538574,
      "learning_rate": 9.615708748678565e-06,
      "loss": 2.168239974975586,
      "memory(GiB)": 77.56,
      "step": 93280,
      "token_acc": 0.5467625899280576,
      "train_speed(iter/s)": 1.437722
    },
    {
      "epoch": 3.9966153977978665,
      "grad_norm": 5.963865756988525,
      "learning_rate": 9.611741152139886e-06,
      "loss": 2.6363449096679688,
      "memory(GiB)": 77.56,
      "step": 93285,
      "token_acc": 0.4588235294117647,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 3.9968296131271153,
      "grad_norm": 7.664267539978027,
      "learning_rate": 9.607774287272386e-06,
      "loss": 2.522052001953125,
      "memory(GiB)": 77.56,
      "step": 93290,
      "token_acc": 0.48945147679324896,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 3.9970438284563645,
      "grad_norm": 4.96965217590332,
      "learning_rate": 9.603808154147958e-06,
      "loss": 2.44207763671875,
      "memory(GiB)": 77.56,
      "step": 93295,
      "token_acc": 0.4358974358974359,
      "train_speed(iter/s)": 1.437737
    },
    {
      "epoch": 3.9972580437856133,
      "grad_norm": 5.467095851898193,
      "learning_rate": 9.599842752838444e-06,
      "loss": 2.2445674896240235,
      "memory(GiB)": 77.56,
      "step": 93300,
      "token_acc": 0.5376044568245125,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 3.997472259114862,
      "grad_norm": 6.500300407409668,
      "learning_rate": 9.595878083415676e-06,
      "loss": 2.3761457443237304,
      "memory(GiB)": 77.56,
      "step": 93305,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 3.9976864744441114,
      "grad_norm": 5.187769889831543,
      "learning_rate": 9.59191414595148e-06,
      "loss": 2.103641700744629,
      "memory(GiB)": 77.56,
      "step": 93310,
      "token_acc": 0.5311475409836065,
      "train_speed(iter/s)": 1.437738
    },
    {
      "epoch": 3.99790068977336,
      "grad_norm": 6.747546195983887,
      "learning_rate": 9.587950940517648e-06,
      "loss": 2.561017608642578,
      "memory(GiB)": 77.56,
      "step": 93315,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.437749
    },
    {
      "epoch": 3.998114905102609,
      "grad_norm": 7.278005123138428,
      "learning_rate": 9.583988467185995e-06,
      "loss": 2.38443603515625,
      "memory(GiB)": 77.56,
      "step": 93320,
      "token_acc": 0.4730878186968839,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 3.9983291204318583,
      "grad_norm": 8.352123260498047,
      "learning_rate": 9.580026726028318e-06,
      "loss": 2.7073802947998047,
      "memory(GiB)": 77.56,
      "step": 93325,
      "token_acc": 0.4605678233438486,
      "train_speed(iter/s)": 1.437769
    },
    {
      "epoch": 3.998543335761107,
      "grad_norm": 6.051680088043213,
      "learning_rate": 9.576065717116378e-06,
      "loss": 2.3547239303588867,
      "memory(GiB)": 77.56,
      "step": 93330,
      "token_acc": 0.48299319727891155,
      "train_speed(iter/s)": 1.437758
    },
    {
      "epoch": 3.998757551090356,
      "grad_norm": 7.034042835235596,
      "learning_rate": 9.572105440521922e-06,
      "loss": 2.5420087814331054,
      "memory(GiB)": 77.56,
      "step": 93335,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 3.998971766419605,
      "grad_norm": 6.215058326721191,
      "learning_rate": 9.568145896316704e-06,
      "loss": 2.1047475814819334,
      "memory(GiB)": 77.56,
      "step": 93340,
      "token_acc": 0.5130718954248366,
      "train_speed(iter/s)": 1.43776
    },
    {
      "epoch": 3.999185981748854,
      "grad_norm": 7.614290714263916,
      "learning_rate": 9.56418708457244e-06,
      "loss": 2.460840034484863,
      "memory(GiB)": 77.56,
      "step": 93345,
      "token_acc": 0.5153846153846153,
      "train_speed(iter/s)": 1.43779
    },
    {
      "epoch": 3.999400197078103,
      "grad_norm": 6.198672771453857,
      "learning_rate": 9.560229005360866e-06,
      "loss": 2.296146583557129,
      "memory(GiB)": 77.56,
      "step": 93350,
      "token_acc": 0.47333333333333333,
      "train_speed(iter/s)": 1.437805
    },
    {
      "epoch": 3.999614412407352,
      "grad_norm": 5.386726379394531,
      "learning_rate": 9.556271658753685e-06,
      "loss": 2.3075128555297852,
      "memory(GiB)": 77.56,
      "step": 93355,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 3.999828627736601,
      "grad_norm": 5.895130634307861,
      "learning_rate": 9.552315044822579e-06,
      "loss": 2.198731231689453,
      "memory(GiB)": 77.56,
      "step": 93360,
      "token_acc": 0.49666666666666665,
      "train_speed(iter/s)": 1.437814
    },
    {
      "epoch": 4.00004284306585,
      "grad_norm": 5.6926727294921875,
      "learning_rate": 9.548359163639226e-06,
      "loss": 2.057406997680664,
      "memory(GiB)": 77.56,
      "step": 93365,
      "token_acc": 0.5617977528089888,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 4.000257058395099,
      "grad_norm": 6.77676248550415,
      "learning_rate": 9.5444040152753e-06,
      "loss": 2.1082324981689453,
      "memory(GiB)": 77.56,
      "step": 93370,
      "token_acc": 0.5057915057915058,
      "train_speed(iter/s)": 1.437811
    },
    {
      "epoch": 4.000471273724347,
      "grad_norm": 6.574366092681885,
      "learning_rate": 9.540449599802426e-06,
      "loss": 2.222467041015625,
      "memory(GiB)": 77.56,
      "step": 93375,
      "token_acc": 0.5288135593220339,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 4.0006854890535966,
      "grad_norm": 6.4058027267456055,
      "learning_rate": 9.536495917292283e-06,
      "loss": 2.3530227661132814,
      "memory(GiB)": 77.56,
      "step": 93380,
      "token_acc": 0.47333333333333333,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 4.000899704382846,
      "grad_norm": 4.862844467163086,
      "learning_rate": 9.532542967816466e-06,
      "loss": 2.409433937072754,
      "memory(GiB)": 77.56,
      "step": 93385,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 4.001113919712094,
      "grad_norm": 5.503967761993408,
      "learning_rate": 9.528590751446587e-06,
      "loss": 2.4542724609375,
      "memory(GiB)": 77.56,
      "step": 93390,
      "token_acc": 0.49557522123893805,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 4.001328135041343,
      "grad_norm": 7.940217971801758,
      "learning_rate": 9.524639268254259e-06,
      "loss": 2.7267017364501953,
      "memory(GiB)": 77.56,
      "step": 93395,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.437871
    },
    {
      "epoch": 4.001542350370593,
      "grad_norm": 5.918318271636963,
      "learning_rate": 9.520688518311065e-06,
      "loss": 2.1060964584350588,
      "memory(GiB)": 77.56,
      "step": 93400,
      "token_acc": 0.5583941605839416,
      "train_speed(iter/s)": 1.43788
    },
    {
      "epoch": 4.001756565699841,
      "grad_norm": 7.177742958068848,
      "learning_rate": 9.51673850168856e-06,
      "loss": 2.380371856689453,
      "memory(GiB)": 77.56,
      "step": 93405,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 4.00197078102909,
      "grad_norm": 4.978150844573975,
      "learning_rate": 9.512789218458323e-06,
      "loss": 2.395379638671875,
      "memory(GiB)": 77.56,
      "step": 93410,
      "token_acc": 0.4774774774774775,
      "train_speed(iter/s)": 1.437904
    },
    {
      "epoch": 4.00218499635834,
      "grad_norm": 7.750488758087158,
      "learning_rate": 9.508840668691887e-06,
      "loss": 2.1341812133789064,
      "memory(GiB)": 77.56,
      "step": 93415,
      "token_acc": 0.5531135531135531,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 4.002399211687588,
      "grad_norm": 5.484640121459961,
      "learning_rate": 9.50489285246079e-06,
      "loss": 2.158664512634277,
      "memory(GiB)": 77.56,
      "step": 93420,
      "token_acc": 0.6095890410958904,
      "train_speed(iter/s)": 1.43792
    },
    {
      "epoch": 4.002613427016837,
      "grad_norm": 5.87212610244751,
      "learning_rate": 9.500945769836545e-06,
      "loss": 2.2428680419921876,
      "memory(GiB)": 77.56,
      "step": 93425,
      "token_acc": 0.5257352941176471,
      "train_speed(iter/s)": 1.437921
    },
    {
      "epoch": 4.0028276423460865,
      "grad_norm": 5.944907188415527,
      "learning_rate": 9.496999420890646e-06,
      "loss": 1.8895322799682617,
      "memory(GiB)": 77.56,
      "step": 93430,
      "token_acc": 0.5977443609022557,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 4.003041857675335,
      "grad_norm": 4.8054633140563965,
      "learning_rate": 9.493053805694608e-06,
      "loss": 2.293653106689453,
      "memory(GiB)": 77.56,
      "step": 93435,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.437904
    },
    {
      "epoch": 4.003256073004584,
      "grad_norm": 8.297558784484863,
      "learning_rate": 9.4891089243199e-06,
      "loss": 2.5106374740600588,
      "memory(GiB)": 77.56,
      "step": 93440,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.437902
    },
    {
      "epoch": 4.003470288333833,
      "grad_norm": 6.737578868865967,
      "learning_rate": 9.48516477683799e-06,
      "loss": 2.484136199951172,
      "memory(GiB)": 77.56,
      "step": 93445,
      "token_acc": 0.49038461538461536,
      "train_speed(iter/s)": 1.437901
    },
    {
      "epoch": 4.003684503663082,
      "grad_norm": 7.1392107009887695,
      "learning_rate": 9.48122136332032e-06,
      "loss": 2.4458948135375977,
      "memory(GiB)": 77.56,
      "step": 93450,
      "token_acc": 0.5246478873239436,
      "train_speed(iter/s)": 1.437908
    },
    {
      "epoch": 4.003898718992331,
      "grad_norm": 7.190907955169678,
      "learning_rate": 9.477278683838341e-06,
      "loss": 2.170952796936035,
      "memory(GiB)": 77.56,
      "step": 93455,
      "token_acc": 0.5719844357976653,
      "train_speed(iter/s)": 1.437916
    },
    {
      "epoch": 4.00411293432158,
      "grad_norm": 4.597238063812256,
      "learning_rate": 9.473336738463457e-06,
      "loss": 2.47666015625,
      "memory(GiB)": 77.56,
      "step": 93460,
      "token_acc": 0.5159420289855072,
      "train_speed(iter/s)": 1.437924
    },
    {
      "epoch": 4.004327149650829,
      "grad_norm": 8.489082336425781,
      "learning_rate": 9.469395527267089e-06,
      "loss": 2.493121337890625,
      "memory(GiB)": 77.56,
      "step": 93465,
      "token_acc": 0.4940119760479042,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 4.004541364980078,
      "grad_norm": 6.523462295532227,
      "learning_rate": 9.46545505032066e-06,
      "loss": 2.1725048065185546,
      "memory(GiB)": 77.56,
      "step": 93470,
      "token_acc": 0.5358490566037736,
      "train_speed(iter/s)": 1.437932
    },
    {
      "epoch": 4.004755580309327,
      "grad_norm": 4.0298237800598145,
      "learning_rate": 9.46151530769554e-06,
      "loss": 2.107798194885254,
      "memory(GiB)": 77.56,
      "step": 93475,
      "token_acc": 0.5449275362318841,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 4.0049697956385755,
      "grad_norm": 4.946781158447266,
      "learning_rate": 9.457576299463094e-06,
      "loss": 2.421368217468262,
      "memory(GiB)": 77.56,
      "step": 93480,
      "token_acc": 0.47752808988764045,
      "train_speed(iter/s)": 1.437939
    },
    {
      "epoch": 4.005184010967825,
      "grad_norm": 5.446173191070557,
      "learning_rate": 9.45363802569469e-06,
      "loss": 2.1397424697875977,
      "memory(GiB)": 77.56,
      "step": 93485,
      "token_acc": 0.535593220338983,
      "train_speed(iter/s)": 1.437913
    },
    {
      "epoch": 4.005398226297074,
      "grad_norm": 6.718372344970703,
      "learning_rate": 9.449700486461649e-06,
      "loss": 2.100190353393555,
      "memory(GiB)": 77.56,
      "step": 93490,
      "token_acc": 0.5524475524475524,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 4.005612441626322,
      "grad_norm": 5.494898796081543,
      "learning_rate": 9.44576368183534e-06,
      "loss": 2.1900152206420898,
      "memory(GiB)": 77.56,
      "step": 93495,
      "token_acc": 0.5046728971962616,
      "train_speed(iter/s)": 1.437921
    },
    {
      "epoch": 4.005826656955572,
      "grad_norm": 6.694465637207031,
      "learning_rate": 9.44182761188706e-06,
      "loss": 2.4290599822998047,
      "memory(GiB)": 77.56,
      "step": 93500,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 4.005826656955572,
      "eval_loss": 2.255371332168579,
      "eval_runtime": 13.7707,
      "eval_samples_per_second": 7.262,
      "eval_steps_per_second": 7.262,
      "eval_token_acc": 0.4885386819484241,
      "step": 93500
    },
    {
      "epoch": 4.006040872284821,
      "grad_norm": 6.061179161071777,
      "learning_rate": 9.437892276688126e-06,
      "loss": 2.2897716522216798,
      "memory(GiB)": 77.56,
      "step": 93505,
      "token_acc": 0.5053974484789009,
      "train_speed(iter/s)": 1.437563
    },
    {
      "epoch": 4.006255087614069,
      "grad_norm": 6.244830131530762,
      "learning_rate": 9.433957676309823e-06,
      "loss": 2.6214229583740236,
      "memory(GiB)": 77.56,
      "step": 93510,
      "token_acc": 0.4588235294117647,
      "train_speed(iter/s)": 1.437578
    },
    {
      "epoch": 4.0064693029433185,
      "grad_norm": 6.47458553314209,
      "learning_rate": 9.43002381082343e-06,
      "loss": 2.2354877471923826,
      "memory(GiB)": 77.56,
      "step": 93515,
      "token_acc": 0.517799352750809,
      "train_speed(iter/s)": 1.4376
    },
    {
      "epoch": 4.006683518272568,
      "grad_norm": 5.488643169403076,
      "learning_rate": 9.426090680300197e-06,
      "loss": 2.1180740356445313,
      "memory(GiB)": 77.56,
      "step": 93520,
      "token_acc": 0.55078125,
      "train_speed(iter/s)": 1.437584
    },
    {
      "epoch": 4.006897733601816,
      "grad_norm": 6.219763278961182,
      "learning_rate": 9.422158284811406e-06,
      "loss": 2.352944183349609,
      "memory(GiB)": 77.56,
      "step": 93525,
      "token_acc": 0.5492063492063493,
      "train_speed(iter/s)": 1.437569
    },
    {
      "epoch": 4.007111948931065,
      "grad_norm": 5.979735851287842,
      "learning_rate": 9.41822662442829e-06,
      "loss": 2.187449264526367,
      "memory(GiB)": 77.56,
      "step": 93530,
      "token_acc": 0.5228758169934641,
      "train_speed(iter/s)": 1.437581
    },
    {
      "epoch": 4.007326164260315,
      "grad_norm": 5.848051071166992,
      "learning_rate": 9.414295699222048e-06,
      "loss": 2.336979103088379,
      "memory(GiB)": 77.56,
      "step": 93535,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 4.007540379589563,
      "grad_norm": 5.259042739868164,
      "learning_rate": 9.410365509263924e-06,
      "loss": 2.3839765548706056,
      "memory(GiB)": 77.56,
      "step": 93540,
      "token_acc": 0.5088235294117647,
      "train_speed(iter/s)": 1.437595
    },
    {
      "epoch": 4.007754594918812,
      "grad_norm": 7.334307670593262,
      "learning_rate": 9.406436054625106e-06,
      "loss": 2.255917549133301,
      "memory(GiB)": 77.56,
      "step": 93545,
      "token_acc": 0.49242424242424243,
      "train_speed(iter/s)": 1.437612
    },
    {
      "epoch": 4.0079688102480615,
      "grad_norm": 5.4149322509765625,
      "learning_rate": 9.40250733537677e-06,
      "loss": 2.18704891204834,
      "memory(GiB)": 77.56,
      "step": 93550,
      "token_acc": 0.5064516129032258,
      "train_speed(iter/s)": 1.437637
    },
    {
      "epoch": 4.00818302557731,
      "grad_norm": 5.233251094818115,
      "learning_rate": 9.398579351590103e-06,
      "loss": 2.2285476684570313,
      "memory(GiB)": 77.56,
      "step": 93555,
      "token_acc": 0.5016181229773463,
      "train_speed(iter/s)": 1.43764
    },
    {
      "epoch": 4.008397240906559,
      "grad_norm": 7.4715800285339355,
      "learning_rate": 9.394652103336265e-06,
      "loss": 2.255872344970703,
      "memory(GiB)": 77.56,
      "step": 93560,
      "token_acc": 0.5298804780876494,
      "train_speed(iter/s)": 1.437657
    },
    {
      "epoch": 4.008611456235808,
      "grad_norm": 7.020638465881348,
      "learning_rate": 9.39072559068639e-06,
      "loss": 2.1696699142456053,
      "memory(GiB)": 77.56,
      "step": 93565,
      "token_acc": 0.5379537953795379,
      "train_speed(iter/s)": 1.437671
    },
    {
      "epoch": 4.008825671565057,
      "grad_norm": 5.364024639129639,
      "learning_rate": 9.386799813711617e-06,
      "loss": 2.167232704162598,
      "memory(GiB)": 77.56,
      "step": 93570,
      "token_acc": 0.5307692307692308,
      "train_speed(iter/s)": 1.437667
    },
    {
      "epoch": 4.009039886894306,
      "grad_norm": 6.392496585845947,
      "learning_rate": 9.382874772483052e-06,
      "loss": 2.1624725341796873,
      "memory(GiB)": 77.56,
      "step": 93575,
      "token_acc": 0.5205479452054794,
      "train_speed(iter/s)": 1.437678
    },
    {
      "epoch": 4.009254102223555,
      "grad_norm": 4.925895690917969,
      "learning_rate": 9.378950467071829e-06,
      "loss": 1.9535303115844727,
      "memory(GiB)": 77.56,
      "step": 93580,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.437681
    },
    {
      "epoch": 4.009468317552804,
      "grad_norm": 6.108914852142334,
      "learning_rate": 9.375026897549017e-06,
      "loss": 2.4048938751220703,
      "memory(GiB)": 77.56,
      "step": 93585,
      "token_acc": 0.5203488372093024,
      "train_speed(iter/s)": 1.437692
    },
    {
      "epoch": 4.009682532882053,
      "grad_norm": 7.15532922744751,
      "learning_rate": 9.371104063985702e-06,
      "loss": 2.3725967407226562,
      "memory(GiB)": 77.56,
      "step": 93590,
      "token_acc": 0.5154929577464789,
      "train_speed(iter/s)": 1.437704
    },
    {
      "epoch": 4.009896748211302,
      "grad_norm": 5.323001861572266,
      "learning_rate": 9.367181966452953e-06,
      "loss": 2.1099403381347654,
      "memory(GiB)": 77.56,
      "step": 93595,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.437703
    },
    {
      "epoch": 4.0101109635405505,
      "grad_norm": 5.8606085777282715,
      "learning_rate": 9.363260605021824e-06,
      "loss": 2.096474838256836,
      "memory(GiB)": 77.56,
      "step": 93600,
      "token_acc": 0.5278810408921933,
      "train_speed(iter/s)": 1.437717
    },
    {
      "epoch": 4.0103251788698,
      "grad_norm": 5.25267219543457,
      "learning_rate": 9.359339979763326e-06,
      "loss": 2.1067861557006835,
      "memory(GiB)": 77.56,
      "step": 93605,
      "token_acc": 0.5127388535031847,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 4.010539394199049,
      "grad_norm": 4.698491096496582,
      "learning_rate": 9.355420090748518e-06,
      "loss": 2.2778825759887695,
      "memory(GiB)": 77.56,
      "step": 93610,
      "token_acc": 0.5337620578778135,
      "train_speed(iter/s)": 1.43772
    },
    {
      "epoch": 4.010753609528297,
      "grad_norm": 5.380602836608887,
      "learning_rate": 9.351500938048408e-06,
      "loss": 2.276262092590332,
      "memory(GiB)": 77.56,
      "step": 93615,
      "token_acc": 0.5287769784172662,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 4.010967824857547,
      "grad_norm": 5.973597526550293,
      "learning_rate": 9.347582521733994e-06,
      "loss": 2.2164852142333986,
      "memory(GiB)": 77.56,
      "step": 93620,
      "token_acc": 0.5174603174603175,
      "train_speed(iter/s)": 1.437714
    },
    {
      "epoch": 4.011182040186796,
      "grad_norm": 9.466315269470215,
      "learning_rate": 9.34366484187626e-06,
      "loss": 2.203280448913574,
      "memory(GiB)": 77.56,
      "step": 93625,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 4.011396255516045,
      "grad_norm": 7.253759384155273,
      "learning_rate": 9.339747898546169e-06,
      "loss": 2.2978134155273438,
      "memory(GiB)": 77.56,
      "step": 93630,
      "token_acc": 0.5181518151815182,
      "train_speed(iter/s)": 1.437718
    },
    {
      "epoch": 4.0116104708452935,
      "grad_norm": 5.4126410484313965,
      "learning_rate": 9.335831691814684e-06,
      "loss": 2.337779998779297,
      "memory(GiB)": 77.56,
      "step": 93635,
      "token_acc": 0.498220640569395,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 4.011824686174543,
      "grad_norm": 8.379061698913574,
      "learning_rate": 9.331916221752756e-06,
      "loss": 2.4290111541748045,
      "memory(GiB)": 77.56,
      "step": 93640,
      "token_acc": 0.48615384615384616,
      "train_speed(iter/s)": 1.437729
    },
    {
      "epoch": 4.012038901503792,
      "grad_norm": 5.679766654968262,
      "learning_rate": 9.328001488431326e-06,
      "loss": 2.4866424560546876,
      "memory(GiB)": 77.56,
      "step": 93645,
      "token_acc": 0.5047923322683706,
      "train_speed(iter/s)": 1.437711
    },
    {
      "epoch": 4.01225311683304,
      "grad_norm": 4.610808372497559,
      "learning_rate": 9.324087491921296e-06,
      "loss": 2.3665496826171877,
      "memory(GiB)": 77.56,
      "step": 93650,
      "token_acc": 0.4863013698630137,
      "train_speed(iter/s)": 1.437719
    },
    {
      "epoch": 4.01246733216229,
      "grad_norm": 6.19395637512207,
      "learning_rate": 9.320174232293576e-06,
      "loss": 2.4105527877807615,
      "memory(GiB)": 77.56,
      "step": 93655,
      "token_acc": 0.46959459459459457,
      "train_speed(iter/s)": 1.437728
    },
    {
      "epoch": 4.012681547491539,
      "grad_norm": 7.039161205291748,
      "learning_rate": 9.316261709619068e-06,
      "loss": 2.5016502380371093,
      "memory(GiB)": 77.56,
      "step": 93660,
      "token_acc": 0.4785714285714286,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 4.012895762820787,
      "grad_norm": 7.270907878875732,
      "learning_rate": 9.312349923968628e-06,
      "loss": 2.511243438720703,
      "memory(GiB)": 77.56,
      "step": 93665,
      "token_acc": 0.4697802197802198,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 4.013109978150037,
      "grad_norm": 6.870691299438477,
      "learning_rate": 9.308438875413145e-06,
      "loss": 2.474112701416016,
      "memory(GiB)": 77.56,
      "step": 93670,
      "token_acc": 0.4733893557422969,
      "train_speed(iter/s)": 1.437736
    },
    {
      "epoch": 4.013324193479286,
      "grad_norm": 5.674384593963623,
      "learning_rate": 9.304528564023468e-06,
      "loss": 2.4112449645996095,
      "memory(GiB)": 77.56,
      "step": 93675,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 4.013538408808534,
      "grad_norm": 5.993567943572998,
      "learning_rate": 9.30061898987043e-06,
      "loss": 2.24066162109375,
      "memory(GiB)": 77.56,
      "step": 93680,
      "token_acc": 0.5122699386503068,
      "train_speed(iter/s)": 1.437761
    },
    {
      "epoch": 4.013752624137783,
      "grad_norm": 6.70174503326416,
      "learning_rate": 9.296710153024846e-06,
      "loss": 2.271550941467285,
      "memory(GiB)": 77.56,
      "step": 93685,
      "token_acc": 0.5096525096525096,
      "train_speed(iter/s)": 1.437782
    },
    {
      "epoch": 4.013966839467033,
      "grad_norm": 5.078118324279785,
      "learning_rate": 9.292802053557553e-06,
      "loss": 2.5433586120605467,
      "memory(GiB)": 77.56,
      "step": 93690,
      "token_acc": 0.5234375,
      "train_speed(iter/s)": 1.437778
    },
    {
      "epoch": 4.014181054796281,
      "grad_norm": 6.76052188873291,
      "learning_rate": 9.288894691539323e-06,
      "loss": 2.1711614608764647,
      "memory(GiB)": 77.56,
      "step": 93695,
      "token_acc": 0.5687022900763359,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 4.01439527012553,
      "grad_norm": 5.271911144256592,
      "learning_rate": 9.284988067040973e-06,
      "loss": 2.108733367919922,
      "memory(GiB)": 77.56,
      "step": 93700,
      "token_acc": 0.5622775800711743,
      "train_speed(iter/s)": 1.437771
    },
    {
      "epoch": 4.01460948545478,
      "grad_norm": 7.146370887756348,
      "learning_rate": 9.281082180133254e-06,
      "loss": 2.4907913208007812,
      "memory(GiB)": 77.56,
      "step": 93705,
      "token_acc": 0.4721311475409836,
      "train_speed(iter/s)": 1.437772
    },
    {
      "epoch": 4.014823700784028,
      "grad_norm": 4.833681583404541,
      "learning_rate": 9.27717703088693e-06,
      "loss": 2.4081165313720705,
      "memory(GiB)": 77.56,
      "step": 93710,
      "token_acc": 0.5064102564102564,
      "train_speed(iter/s)": 1.437783
    },
    {
      "epoch": 4.015037916113277,
      "grad_norm": 5.994205951690674,
      "learning_rate": 9.273272619372748e-06,
      "loss": 2.372846984863281,
      "memory(GiB)": 77.56,
      "step": 93715,
      "token_acc": 0.5033333333333333,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 4.0152521314425265,
      "grad_norm": 8.07086181640625,
      "learning_rate": 9.269368945661422e-06,
      "loss": 2.5486459732055664,
      "memory(GiB)": 77.56,
      "step": 93720,
      "token_acc": 0.4468864468864469,
      "train_speed(iter/s)": 1.437797
    },
    {
      "epoch": 4.015466346771775,
      "grad_norm": 6.291993141174316,
      "learning_rate": 9.265466009823698e-06,
      "loss": 2.392086982727051,
      "memory(GiB)": 77.56,
      "step": 93725,
      "token_acc": 0.46418338108882523,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 4.015680562101024,
      "grad_norm": 6.247165679931641,
      "learning_rate": 9.261563811930269e-06,
      "loss": 2.0815719604492187,
      "memory(GiB)": 77.56,
      "step": 93730,
      "token_acc": 0.5604395604395604,
      "train_speed(iter/s)": 1.437791
    },
    {
      "epoch": 4.015894777430273,
      "grad_norm": 7.932577133178711,
      "learning_rate": 9.257662352051827e-06,
      "loss": 2.198335075378418,
      "memory(GiB)": 77.56,
      "step": 93735,
      "token_acc": 0.55859375,
      "train_speed(iter/s)": 1.437781
    },
    {
      "epoch": 4.016108992759522,
      "grad_norm": 5.848824501037598,
      "learning_rate": 9.25376163025905e-06,
      "loss": 2.5626031875610353,
      "memory(GiB)": 77.56,
      "step": 93740,
      "token_acc": 0.4887005649717514,
      "train_speed(iter/s)": 1.437787
    },
    {
      "epoch": 4.016323208088771,
      "grad_norm": 5.185723304748535,
      "learning_rate": 9.24986164662261e-06,
      "loss": 2.2850088119506835,
      "memory(GiB)": 77.56,
      "step": 93745,
      "token_acc": 0.5213414634146342,
      "train_speed(iter/s)": 1.437796
    },
    {
      "epoch": 4.01653742341802,
      "grad_norm": 7.207386493682861,
      "learning_rate": 9.245962401213132e-06,
      "loss": 2.306472969055176,
      "memory(GiB)": 77.56,
      "step": 93750,
      "token_acc": 0.5224489795918368,
      "train_speed(iter/s)": 1.437797
    },
    {
      "epoch": 4.016751638747269,
      "grad_norm": 6.612720012664795,
      "learning_rate": 9.242063894101294e-06,
      "loss": 2.2329856872558596,
      "memory(GiB)": 77.56,
      "step": 93755,
      "token_acc": 0.5225806451612903,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 4.016965854076518,
      "grad_norm": 7.3874192237854,
      "learning_rate": 9.238166125357689e-06,
      "loss": 2.7943016052246095,
      "memory(GiB)": 77.56,
      "step": 93760,
      "token_acc": 0.4260869565217391,
      "train_speed(iter/s)": 1.437808
    },
    {
      "epoch": 4.017180069405767,
      "grad_norm": 7.883175373077393,
      "learning_rate": 9.234269095052956e-06,
      "loss": 2.1302608489990233,
      "memory(GiB)": 77.56,
      "step": 93765,
      "token_acc": 0.5139442231075697,
      "train_speed(iter/s)": 1.437813
    },
    {
      "epoch": 4.0173942847350155,
      "grad_norm": 4.341423988342285,
      "learning_rate": 9.230372803257675e-06,
      "loss": 2.2670223236083986,
      "memory(GiB)": 77.56,
      "step": 93770,
      "token_acc": 0.5078125,
      "train_speed(iter/s)": 1.437814
    },
    {
      "epoch": 4.017608500064265,
      "grad_norm": 6.676019191741943,
      "learning_rate": 9.226477250042443e-06,
      "loss": 2.3452211380004884,
      "memory(GiB)": 77.56,
      "step": 93775,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.437821
    },
    {
      "epoch": 4.017822715393514,
      "grad_norm": 6.873841762542725,
      "learning_rate": 9.222582435477805e-06,
      "loss": 1.9828811645507813,
      "memory(GiB)": 77.56,
      "step": 93780,
      "token_acc": 0.5460526315789473,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 4.018036930722762,
      "grad_norm": 7.752004146575928,
      "learning_rate": 9.218688359634352e-06,
      "loss": 2.4788864135742186,
      "memory(GiB)": 77.56,
      "step": 93785,
      "token_acc": 0.4758364312267658,
      "train_speed(iter/s)": 1.437824
    },
    {
      "epoch": 4.018251146052012,
      "grad_norm": 5.326835632324219,
      "learning_rate": 9.214795022582617e-06,
      "loss": 2.3073713302612306,
      "memory(GiB)": 77.56,
      "step": 93790,
      "token_acc": 0.49848024316109424,
      "train_speed(iter/s)": 1.437815
    },
    {
      "epoch": 4.018465361381261,
      "grad_norm": 5.026467323303223,
      "learning_rate": 9.210902424393131e-06,
      "loss": 2.0833648681640624,
      "memory(GiB)": 77.56,
      "step": 93795,
      "token_acc": 0.5373134328358209,
      "train_speed(iter/s)": 1.437812
    },
    {
      "epoch": 4.018679576710509,
      "grad_norm": 6.776566505432129,
      "learning_rate": 9.207010565136409e-06,
      "loss": 1.9410991668701172,
      "memory(GiB)": 77.56,
      "step": 93800,
      "token_acc": 0.5875486381322957,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 4.0188937920397585,
      "grad_norm": 5.776382923126221,
      "learning_rate": 9.203119444882958e-06,
      "loss": 2.6139245986938477,
      "memory(GiB)": 77.56,
      "step": 93805,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 4.019108007369008,
      "grad_norm": 4.514953136444092,
      "learning_rate": 9.199229063703257e-06,
      "loss": 2.11966552734375,
      "memory(GiB)": 77.56,
      "step": 93810,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.437828
    },
    {
      "epoch": 4.019322222698256,
      "grad_norm": 5.997488975524902,
      "learning_rate": 9.195339421667804e-06,
      "loss": 2.1361522674560547,
      "memory(GiB)": 77.56,
      "step": 93815,
      "token_acc": 0.523943661971831,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 4.019536438027505,
      "grad_norm": 8.49476432800293,
      "learning_rate": 9.191450518847062e-06,
      "loss": 2.173126983642578,
      "memory(GiB)": 77.56,
      "step": 93820,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.437846
    },
    {
      "epoch": 4.019750653356755,
      "grad_norm": 6.412816524505615,
      "learning_rate": 9.187562355311469e-06,
      "loss": 2.045247268676758,
      "memory(GiB)": 77.56,
      "step": 93825,
      "token_acc": 0.5335968379446641,
      "train_speed(iter/s)": 1.437855
    },
    {
      "epoch": 4.019964868686003,
      "grad_norm": 6.843888282775879,
      "learning_rate": 9.183674931131464e-06,
      "loss": 2.1562389373779296,
      "memory(GiB)": 77.56,
      "step": 93830,
      "token_acc": 0.5461538461538461,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 4.020179084015252,
      "grad_norm": 7.188787460327148,
      "learning_rate": 9.17978824637748e-06,
      "loss": 2.0955738067626952,
      "memory(GiB)": 77.56,
      "step": 93835,
      "token_acc": 0.5481481481481482,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 4.0203932993445015,
      "grad_norm": 6.331221580505371,
      "learning_rate": 9.175902301119922e-06,
      "loss": 2.2976964950561523,
      "memory(GiB)": 77.56,
      "step": 93840,
      "token_acc": 0.565068493150685,
      "train_speed(iter/s)": 1.437905
    },
    {
      "epoch": 4.02060751467375,
      "grad_norm": 5.83984375,
      "learning_rate": 9.172017095429191e-06,
      "loss": 2.0562944412231445,
      "memory(GiB)": 77.56,
      "step": 93845,
      "token_acc": 0.5830258302583026,
      "train_speed(iter/s)": 1.437926
    },
    {
      "epoch": 4.020821730002999,
      "grad_norm": 4.872024059295654,
      "learning_rate": 9.16813262937568e-06,
      "loss": 2.4306571960449217,
      "memory(GiB)": 77.56,
      "step": 93850,
      "token_acc": 0.4879518072289157,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 4.021035945332248,
      "grad_norm": 11.080470085144043,
      "learning_rate": 9.164248903029742e-06,
      "loss": 2.336315155029297,
      "memory(GiB)": 77.56,
      "step": 93855,
      "token_acc": 0.49110320284697506,
      "train_speed(iter/s)": 1.437935
    },
    {
      "epoch": 4.021250160661497,
      "grad_norm": 7.394295692443848,
      "learning_rate": 9.16036591646175e-06,
      "loss": 2.073668670654297,
      "memory(GiB)": 77.56,
      "step": 93860,
      "token_acc": 0.547244094488189,
      "train_speed(iter/s)": 1.437947
    },
    {
      "epoch": 4.021464375990746,
      "grad_norm": 5.520383834838867,
      "learning_rate": 9.156483669742022e-06,
      "loss": 2.4991296768188476,
      "memory(GiB)": 77.56,
      "step": 93865,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 4.021678591319995,
      "grad_norm": 6.500822067260742,
      "learning_rate": 9.152602162940921e-06,
      "loss": 2.2651416778564455,
      "memory(GiB)": 77.56,
      "step": 93870,
      "token_acc": 0.5152542372881356,
      "train_speed(iter/s)": 1.437933
    },
    {
      "epoch": 4.021892806649244,
      "grad_norm": 5.594616889953613,
      "learning_rate": 9.148721396128751e-06,
      "loss": 2.369472885131836,
      "memory(GiB)": 77.56,
      "step": 93875,
      "token_acc": 0.48955223880597015,
      "train_speed(iter/s)": 1.437943
    },
    {
      "epoch": 4.022107021978493,
      "grad_norm": 6.3511223793029785,
      "learning_rate": 9.144841369375811e-06,
      "loss": 2.4668338775634764,
      "memory(GiB)": 77.56,
      "step": 93880,
      "token_acc": 0.48746518105849584,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 4.022321237307742,
      "grad_norm": 5.558446884155273,
      "learning_rate": 9.140962082752397e-06,
      "loss": 2.4822521209716797,
      "memory(GiB)": 77.56,
      "step": 93885,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.437964
    },
    {
      "epoch": 4.0225354526369905,
      "grad_norm": 8.363835334777832,
      "learning_rate": 9.137083536328783e-06,
      "loss": 2.3325389862060546,
      "memory(GiB)": 77.56,
      "step": 93890,
      "token_acc": 0.5117845117845118,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 4.02274966796624,
      "grad_norm": 6.519898891448975,
      "learning_rate": 9.133205730175221e-06,
      "loss": 2.4897390365600587,
      "memory(GiB)": 77.56,
      "step": 93895,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 4.022963883295489,
      "grad_norm": 4.96488094329834,
      "learning_rate": 9.129328664361985e-06,
      "loss": 2.0881967544555664,
      "memory(GiB)": 77.56,
      "step": 93900,
      "token_acc": 0.5475409836065573,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 4.023178098624737,
      "grad_norm": 6.539687633514404,
      "learning_rate": 9.12545233895929e-06,
      "loss": 2.14547061920166,
      "memory(GiB)": 77.56,
      "step": 93905,
      "token_acc": 0.5394190871369294,
      "train_speed(iter/s)": 1.437955
    },
    {
      "epoch": 4.023392313953987,
      "grad_norm": 6.46572208404541,
      "learning_rate": 9.121576754037381e-06,
      "loss": 2.5262537002563477,
      "memory(GiB)": 77.56,
      "step": 93910,
      "token_acc": 0.486013986013986,
      "train_speed(iter/s)": 1.437944
    },
    {
      "epoch": 4.023606529283236,
      "grad_norm": 5.656039714813232,
      "learning_rate": 9.117701909666453e-06,
      "loss": 1.9382194519042968,
      "memory(GiB)": 77.56,
      "step": 93915,
      "token_acc": 0.5793650793650794,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 4.023820744612484,
      "grad_norm": 5.459254264831543,
      "learning_rate": 9.113827805916714e-06,
      "loss": 2.266299247741699,
      "memory(GiB)": 77.56,
      "step": 93920,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.437903
    },
    {
      "epoch": 4.0240349599417335,
      "grad_norm": 6.356588363647461,
      "learning_rate": 9.109954442858315e-06,
      "loss": 2.159096336364746,
      "memory(GiB)": 77.56,
      "step": 93925,
      "token_acc": 0.5322033898305085,
      "train_speed(iter/s)": 1.43792
    },
    {
      "epoch": 4.024249175270983,
      "grad_norm": 6.883572578430176,
      "learning_rate": 9.106081820561469e-06,
      "loss": 2.130754852294922,
      "memory(GiB)": 77.56,
      "step": 93930,
      "token_acc": 0.48417721518987344,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 4.024463390600231,
      "grad_norm": 6.1585493087768555,
      "learning_rate": 9.102209939096311e-06,
      "loss": 2.3141834259033205,
      "memory(GiB)": 77.56,
      "step": 93935,
      "token_acc": 0.528125,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 4.02467760592948,
      "grad_norm": 4.8613691329956055,
      "learning_rate": 9.098338798532985e-06,
      "loss": 2.335850715637207,
      "memory(GiB)": 77.56,
      "step": 93940,
      "token_acc": 0.5034965034965035,
      "train_speed(iter/s)": 1.437926
    },
    {
      "epoch": 4.02489182125873,
      "grad_norm": 5.536412239074707,
      "learning_rate": 9.094468398941618e-06,
      "loss": 2.4390344619750977,
      "memory(GiB)": 77.56,
      "step": 93945,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 4.025106036587978,
      "grad_norm": 5.811814308166504,
      "learning_rate": 9.090598740392331e-06,
      "loss": 2.358005905151367,
      "memory(GiB)": 77.56,
      "step": 93950,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 4.025320251917227,
      "grad_norm": 6.258542537689209,
      "learning_rate": 9.086729822955214e-06,
      "loss": 2.34885311126709,
      "memory(GiB)": 77.56,
      "step": 93955,
      "token_acc": 0.5138339920948617,
      "train_speed(iter/s)": 1.43793
    },
    {
      "epoch": 4.025534467246477,
      "grad_norm": 4.842301845550537,
      "learning_rate": 9.082861646700374e-06,
      "loss": 2.214085006713867,
      "memory(GiB)": 77.56,
      "step": 93960,
      "token_acc": 0.5197568389057751,
      "train_speed(iter/s)": 1.437919
    },
    {
      "epoch": 4.025748682575725,
      "grad_norm": 4.775691032409668,
      "learning_rate": 9.078994211697883e-06,
      "loss": 2.193232536315918,
      "memory(GiB)": 77.56,
      "step": 93965,
      "token_acc": 0.5544554455445545,
      "train_speed(iter/s)": 1.437927
    },
    {
      "epoch": 4.025962897904974,
      "grad_norm": 5.557093143463135,
      "learning_rate": 9.075127518017796e-06,
      "loss": 2.130133628845215,
      "memory(GiB)": 77.56,
      "step": 93970,
      "token_acc": 0.535593220338983,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 4.0261771132342234,
      "grad_norm": 7.165945053100586,
      "learning_rate": 9.07126156573016e-06,
      "loss": 1.7786710739135743,
      "memory(GiB)": 77.56,
      "step": 93975,
      "token_acc": 0.596,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 4.026391328563472,
      "grad_norm": 6.821719169616699,
      "learning_rate": 9.06739635490501e-06,
      "loss": 2.408656120300293,
      "memory(GiB)": 77.56,
      "step": 93980,
      "token_acc": 0.5033783783783784,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 4.026605543892721,
      "grad_norm": 6.899848937988281,
      "learning_rate": 9.06353188561237e-06,
      "loss": 2.173081970214844,
      "memory(GiB)": 77.56,
      "step": 93985,
      "token_acc": 0.5335463258785943,
      "train_speed(iter/s)": 1.43795
    },
    {
      "epoch": 4.02681975922197,
      "grad_norm": 5.928413391113281,
      "learning_rate": 9.059668157922262e-06,
      "loss": 2.167915916442871,
      "memory(GiB)": 77.56,
      "step": 93990,
      "token_acc": 0.5222222222222223,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 4.027033974551219,
      "grad_norm": 5.2138190269470215,
      "learning_rate": 9.055805171904674e-06,
      "loss": 2.347804641723633,
      "memory(GiB)": 77.56,
      "step": 93995,
      "token_acc": 0.5224913494809689,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 4.027248189880468,
      "grad_norm": 6.200860500335693,
      "learning_rate": 9.051942927629586e-06,
      "loss": 2.168229675292969,
      "memory(GiB)": 77.56,
      "step": 94000,
      "token_acc": 0.5480427046263345,
      "train_speed(iter/s)": 1.437982
    },
    {
      "epoch": 4.027248189880468,
      "eval_loss": 2.3073084354400635,
      "eval_runtime": 13.6052,
      "eval_samples_per_second": 7.35,
      "eval_steps_per_second": 7.35,
      "eval_token_acc": 0.47229551451187335,
      "step": 94000
    },
    {
      "epoch": 4.027462405209717,
      "grad_norm": 6.543453693389893,
      "learning_rate": 9.04808142516696e-06,
      "loss": 2.0507335662841797,
      "memory(GiB)": 77.56,
      "step": 94005,
      "token_acc": 0.48221343873517786,
      "train_speed(iter/s)": 1.437684
    },
    {
      "epoch": 4.027676620538966,
      "grad_norm": 6.228626251220703,
      "learning_rate": 9.044220664586745e-06,
      "loss": 2.1440031051635744,
      "memory(GiB)": 77.56,
      "step": 94010,
      "token_acc": 0.5084175084175084,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 4.027890835868215,
      "grad_norm": 7.446531295776367,
      "learning_rate": 9.0403606459589e-06,
      "loss": 2.4961212158203123,
      "memory(GiB)": 77.56,
      "step": 94015,
      "token_acc": 0.47530864197530864,
      "train_speed(iter/s)": 1.437693
    },
    {
      "epoch": 4.028105051197464,
      "grad_norm": 6.428940296173096,
      "learning_rate": 9.03650136935335e-06,
      "loss": 2.323657989501953,
      "memory(GiB)": 77.56,
      "step": 94020,
      "token_acc": 0.5247524752475248,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 4.0283192665267125,
      "grad_norm": 5.986566066741943,
      "learning_rate": 9.03264283484e-06,
      "loss": 2.451040840148926,
      "memory(GiB)": 77.56,
      "step": 94025,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.437691
    },
    {
      "epoch": 4.028533481855962,
      "grad_norm": 5.285022735595703,
      "learning_rate": 9.028785042488758e-06,
      "loss": 2.0093597412109374,
      "memory(GiB)": 77.56,
      "step": 94030,
      "token_acc": 0.5622895622895623,
      "train_speed(iter/s)": 1.437685
    },
    {
      "epoch": 4.028747697185211,
      "grad_norm": 5.937748432159424,
      "learning_rate": 9.024927992369514e-06,
      "loss": 2.5239925384521484,
      "memory(GiB)": 77.56,
      "step": 94035,
      "token_acc": 0.5084175084175084,
      "train_speed(iter/s)": 1.437704
    },
    {
      "epoch": 4.028961912514459,
      "grad_norm": 9.210309982299805,
      "learning_rate": 9.021071684552117e-06,
      "loss": 2.227867317199707,
      "memory(GiB)": 77.56,
      "step": 94040,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.437711
    },
    {
      "epoch": 4.029176127843709,
      "grad_norm": 5.284056186676025,
      "learning_rate": 9.017216119106464e-06,
      "loss": 1.9757549285888671,
      "memory(GiB)": 77.56,
      "step": 94045,
      "token_acc": 0.5704225352112676,
      "train_speed(iter/s)": 1.437693
    },
    {
      "epoch": 4.029390343172958,
      "grad_norm": 8.236066818237305,
      "learning_rate": 9.013361296102385e-06,
      "loss": 2.0382003784179688,
      "memory(GiB)": 77.56,
      "step": 94050,
      "token_acc": 0.5403508771929825,
      "train_speed(iter/s)": 1.437692
    },
    {
      "epoch": 4.029604558502206,
      "grad_norm": 5.8400702476501465,
      "learning_rate": 9.009507215609702e-06,
      "loss": 2.4904153823852537,
      "memory(GiB)": 77.56,
      "step": 94055,
      "token_acc": 0.4738292011019284,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 4.0298187738314555,
      "grad_norm": 6.436703681945801,
      "learning_rate": 9.005653877698261e-06,
      "loss": 2.241098403930664,
      "memory(GiB)": 77.56,
      "step": 94060,
      "token_acc": 0.47987616099071206,
      "train_speed(iter/s)": 1.4377
    },
    {
      "epoch": 4.030032989160705,
      "grad_norm": 7.1326727867126465,
      "learning_rate": 9.001801282437854e-06,
      "loss": 2.3208244323730467,
      "memory(GiB)": 77.56,
      "step": 94065,
      "token_acc": 0.5141843971631206,
      "train_speed(iter/s)": 1.437725
    },
    {
      "epoch": 4.030247204489953,
      "grad_norm": 6.026688098907471,
      "learning_rate": 8.997949429898267e-06,
      "loss": 2.549468231201172,
      "memory(GiB)": 77.56,
      "step": 94070,
      "token_acc": 0.4784172661870504,
      "train_speed(iter/s)": 1.437724
    },
    {
      "epoch": 4.030461419819202,
      "grad_norm": 6.105508327484131,
      "learning_rate": 8.994098320149303e-06,
      "loss": 2.3366880416870117,
      "memory(GiB)": 77.56,
      "step": 94075,
      "token_acc": 0.5034965034965035,
      "train_speed(iter/s)": 1.437721
    },
    {
      "epoch": 4.030675635148452,
      "grad_norm": 5.545905113220215,
      "learning_rate": 8.99024795326071e-06,
      "loss": 2.3433456420898438,
      "memory(GiB)": 77.56,
      "step": 94080,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.437698
    },
    {
      "epoch": 4.0308898504777,
      "grad_norm": 5.896813869476318,
      "learning_rate": 8.986398329302248e-06,
      "loss": 2.2935924530029297,
      "memory(GiB)": 77.56,
      "step": 94085,
      "token_acc": 0.4906832298136646,
      "train_speed(iter/s)": 1.437692
    },
    {
      "epoch": 4.031104065806949,
      "grad_norm": 6.328481197357178,
      "learning_rate": 8.982549448343653e-06,
      "loss": 2.0394792556762695,
      "memory(GiB)": 77.56,
      "step": 94090,
      "token_acc": 0.55078125,
      "train_speed(iter/s)": 1.437695
    },
    {
      "epoch": 4.0313182811361985,
      "grad_norm": 6.472174167633057,
      "learning_rate": 8.978701310454652e-06,
      "loss": 2.1981388092041017,
      "memory(GiB)": 77.56,
      "step": 94095,
      "token_acc": 0.4812286689419795,
      "train_speed(iter/s)": 1.437712
    },
    {
      "epoch": 4.031532496465447,
      "grad_norm": 6.0419793128967285,
      "learning_rate": 8.974853915704945e-06,
      "loss": 2.242582893371582,
      "memory(GiB)": 77.56,
      "step": 94100,
      "token_acc": 0.5047318611987381,
      "train_speed(iter/s)": 1.437716
    },
    {
      "epoch": 4.031746711794696,
      "grad_norm": 6.998654842376709,
      "learning_rate": 8.97100726416426e-06,
      "loss": 2.5070701599121095,
      "memory(GiB)": 77.56,
      "step": 94105,
      "token_acc": 0.5048231511254019,
      "train_speed(iter/s)": 1.437699
    },
    {
      "epoch": 4.031960927123945,
      "grad_norm": 5.389399528503418,
      "learning_rate": 8.967161355902265e-06,
      "loss": 2.485128402709961,
      "memory(GiB)": 77.56,
      "step": 94110,
      "token_acc": 0.5278688524590164,
      "train_speed(iter/s)": 1.437689
    },
    {
      "epoch": 4.032175142453194,
      "grad_norm": 6.884333610534668,
      "learning_rate": 8.96331619098863e-06,
      "loss": 2.2493865966796873,
      "memory(GiB)": 77.56,
      "step": 94115,
      "token_acc": 0.5148514851485149,
      "train_speed(iter/s)": 1.437694
    },
    {
      "epoch": 4.032389357782443,
      "grad_norm": 5.905115604400635,
      "learning_rate": 8.959471769493022e-06,
      "loss": 2.235182189941406,
      "memory(GiB)": 77.56,
      "step": 94120,
      "token_acc": 0.4937888198757764,
      "train_speed(iter/s)": 1.437681
    },
    {
      "epoch": 4.032603573111692,
      "grad_norm": 6.828675746917725,
      "learning_rate": 8.955628091485068e-06,
      "loss": 1.9574359893798827,
      "memory(GiB)": 77.56,
      "step": 94125,
      "token_acc": 0.531986531986532,
      "train_speed(iter/s)": 1.437702
    },
    {
      "epoch": 4.032817788440941,
      "grad_norm": 5.532287120819092,
      "learning_rate": 8.951785157034415e-06,
      "loss": 1.9771404266357422,
      "memory(GiB)": 77.56,
      "step": 94130,
      "token_acc": 0.5559322033898305,
      "train_speed(iter/s)": 1.437713
    },
    {
      "epoch": 4.03303200377019,
      "grad_norm": 6.384617328643799,
      "learning_rate": 8.947942966210692e-06,
      "loss": 2.2690092086791993,
      "memory(GiB)": 77.56,
      "step": 94135,
      "token_acc": 0.49280575539568344,
      "train_speed(iter/s)": 1.437723
    },
    {
      "epoch": 4.033246219099439,
      "grad_norm": 5.13917350769043,
      "learning_rate": 8.944101519083492e-06,
      "loss": 2.5535118103027346,
      "memory(GiB)": 77.56,
      "step": 94140,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.437731
    },
    {
      "epoch": 4.0334604344286875,
      "grad_norm": 6.78342342376709,
      "learning_rate": 8.940260815722407e-06,
      "loss": 2.173914337158203,
      "memory(GiB)": 77.56,
      "step": 94145,
      "token_acc": 0.563573883161512,
      "train_speed(iter/s)": 1.437745
    },
    {
      "epoch": 4.033674649757937,
      "grad_norm": 7.350124359130859,
      "learning_rate": 8.936420856197015e-06,
      "loss": 2.1620296478271483,
      "memory(GiB)": 77.56,
      "step": 94150,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437743
    },
    {
      "epoch": 4.033888865087186,
      "grad_norm": 7.334506988525391,
      "learning_rate": 8.932581640576865e-06,
      "loss": 2.124253273010254,
      "memory(GiB)": 77.56,
      "step": 94155,
      "token_acc": 0.5351681957186545,
      "train_speed(iter/s)": 1.437739
    },
    {
      "epoch": 4.034103080416434,
      "grad_norm": 5.405720233917236,
      "learning_rate": 8.928743168931535e-06,
      "loss": 2.214800262451172,
      "memory(GiB)": 77.56,
      "step": 94160,
      "token_acc": 0.5426621160409556,
      "train_speed(iter/s)": 1.437762
    },
    {
      "epoch": 4.034317295745684,
      "grad_norm": 7.348898410797119,
      "learning_rate": 8.924905441330556e-06,
      "loss": 2.107699966430664,
      "memory(GiB)": 77.56,
      "step": 94165,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.437777
    },
    {
      "epoch": 4.034531511074933,
      "grad_norm": 6.633355617523193,
      "learning_rate": 8.921068457843445e-06,
      "loss": 2.515640640258789,
      "memory(GiB)": 77.56,
      "step": 94170,
      "token_acc": 0.46178343949044587,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 4.034745726404181,
      "grad_norm": 7.168416976928711,
      "learning_rate": 8.917232218539712e-06,
      "loss": 2.4323757171630858,
      "memory(GiB)": 77.56,
      "step": 94175,
      "token_acc": 0.47333333333333333,
      "train_speed(iter/s)": 1.437797
    },
    {
      "epoch": 4.0349599417334305,
      "grad_norm": 5.7977094650268555,
      "learning_rate": 8.913396723488854e-06,
      "loss": 2.229866600036621,
      "memory(GiB)": 77.56,
      "step": 94180,
      "token_acc": 0.525,
      "train_speed(iter/s)": 1.43781
    },
    {
      "epoch": 4.03517415706268,
      "grad_norm": 5.585672378540039,
      "learning_rate": 8.909561972760343e-06,
      "loss": 2.4388826370239256,
      "memory(GiB)": 77.56,
      "step": 94185,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.437827
    },
    {
      "epoch": 4.035388372391928,
      "grad_norm": 6.51699161529541,
      "learning_rate": 8.905727966423682e-06,
      "loss": 2.2358789443969727,
      "memory(GiB)": 77.56,
      "step": 94190,
      "token_acc": 0.5062111801242236,
      "train_speed(iter/s)": 1.437825
    },
    {
      "epoch": 4.035602587721177,
      "grad_norm": 8.528943061828613,
      "learning_rate": 8.901894704548308e-06,
      "loss": 2.3063486099243162,
      "memory(GiB)": 77.56,
      "step": 94195,
      "token_acc": 0.49615384615384617,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 4.035816803050427,
      "grad_norm": 5.621964931488037,
      "learning_rate": 8.89806218720365e-06,
      "loss": 2.2568063735961914,
      "memory(GiB)": 77.56,
      "step": 94200,
      "token_acc": 0.5015576323987538,
      "train_speed(iter/s)": 1.437817
    },
    {
      "epoch": 4.036031018379675,
      "grad_norm": 7.0758562088012695,
      "learning_rate": 8.89423041445916e-06,
      "loss": 2.210978698730469,
      "memory(GiB)": 77.56,
      "step": 94205,
      "token_acc": 0.5486381322957199,
      "train_speed(iter/s)": 1.437818
    },
    {
      "epoch": 4.036245233708924,
      "grad_norm": 6.545154571533203,
      "learning_rate": 8.890399386384251e-06,
      "loss": 2.154315948486328,
      "memory(GiB)": 77.56,
      "step": 94210,
      "token_acc": 0.5083056478405316,
      "train_speed(iter/s)": 1.437822
    },
    {
      "epoch": 4.0364594490381736,
      "grad_norm": 6.9491729736328125,
      "learning_rate": 8.88656910304831e-06,
      "loss": 2.5408910751342773,
      "memory(GiB)": 77.56,
      "step": 94215,
      "token_acc": 0.47735191637630664,
      "train_speed(iter/s)": 1.437829
    },
    {
      "epoch": 4.036673664367422,
      "grad_norm": 9.270106315612793,
      "learning_rate": 8.882739564520743e-06,
      "loss": 2.3467565536499024,
      "memory(GiB)": 77.56,
      "step": 94220,
      "token_acc": 0.5298804780876494,
      "train_speed(iter/s)": 1.437831
    },
    {
      "epoch": 4.036887879696671,
      "grad_norm": 5.729696273803711,
      "learning_rate": 8.878910770870924e-06,
      "loss": 2.2115558624267577,
      "memory(GiB)": 77.56,
      "step": 94225,
      "token_acc": 0.5102040816326531,
      "train_speed(iter/s)": 1.437833
    },
    {
      "epoch": 4.03710209502592,
      "grad_norm": 4.764277935028076,
      "learning_rate": 8.875082722168215e-06,
      "loss": 2.413976287841797,
      "memory(GiB)": 77.56,
      "step": 94230,
      "token_acc": 0.4691011235955056,
      "train_speed(iter/s)": 1.43784
    },
    {
      "epoch": 4.037316310355169,
      "grad_norm": 5.905101776123047,
      "learning_rate": 8.871255418481955e-06,
      "loss": 2.2787710189819337,
      "memory(GiB)": 77.56,
      "step": 94235,
      "token_acc": 0.5253623188405797,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 4.037530525684418,
      "grad_norm": 7.858553886413574,
      "learning_rate": 8.867428859881487e-06,
      "loss": 2.1349054336547852,
      "memory(GiB)": 77.56,
      "step": 94240,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.437854
    },
    {
      "epoch": 4.037744741013667,
      "grad_norm": 5.127308368682861,
      "learning_rate": 8.863603046436119e-06,
      "loss": 2.185748291015625,
      "memory(GiB)": 77.56,
      "step": 94245,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 4.037958956342916,
      "grad_norm": 6.771142482757568,
      "learning_rate": 8.859777978215184e-06,
      "loss": 2.623724365234375,
      "memory(GiB)": 77.56,
      "step": 94250,
      "token_acc": 0.46855345911949686,
      "train_speed(iter/s)": 1.437869
    },
    {
      "epoch": 4.038173171672165,
      "grad_norm": 6.034451484680176,
      "learning_rate": 8.855953655287962e-06,
      "loss": 2.1346485137939455,
      "memory(GiB)": 77.56,
      "step": 94255,
      "token_acc": 0.5110294117647058,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 4.038387387001414,
      "grad_norm": 7.054361343383789,
      "learning_rate": 8.852130077723736e-06,
      "loss": 2.2240100860595704,
      "memory(GiB)": 77.56,
      "step": 94260,
      "token_acc": 0.515625,
      "train_speed(iter/s)": 1.437889
    },
    {
      "epoch": 4.038601602330663,
      "grad_norm": 6.232862949371338,
      "learning_rate": 8.848307245591774e-06,
      "loss": 2.259897994995117,
      "memory(GiB)": 77.56,
      "step": 94265,
      "token_acc": 0.5390070921985816,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 4.038815817659912,
      "grad_norm": 5.6272149085998535,
      "learning_rate": 8.844485158961318e-06,
      "loss": 2.2962581634521486,
      "memory(GiB)": 77.56,
      "step": 94270,
      "token_acc": 0.49852507374631266,
      "train_speed(iter/s)": 1.437912
    },
    {
      "epoch": 4.039030032989161,
      "grad_norm": 7.337576866149902,
      "learning_rate": 8.840663817901618e-06,
      "loss": 2.1763942718505858,
      "memory(GiB)": 77.56,
      "step": 94275,
      "token_acc": 0.5251798561151079,
      "train_speed(iter/s)": 1.437912
    },
    {
      "epoch": 4.0392442483184094,
      "grad_norm": 5.830558776855469,
      "learning_rate": 8.836843222481917e-06,
      "loss": 2.210898208618164,
      "memory(GiB)": 77.56,
      "step": 94280,
      "token_acc": 0.5067567567567568,
      "train_speed(iter/s)": 1.437915
    },
    {
      "epoch": 4.039458463647659,
      "grad_norm": 6.63686990737915,
      "learning_rate": 8.833023372771415e-06,
      "loss": 2.1141281127929688,
      "memory(GiB)": 77.56,
      "step": 94285,
      "token_acc": 0.5627376425855514,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 4.039672678976908,
      "grad_norm": 6.3502020835876465,
      "learning_rate": 8.829204268839314e-06,
      "loss": 2.2010631561279297,
      "memory(GiB)": 77.56,
      "step": 94290,
      "token_acc": 0.5532646048109966,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 4.039886894306156,
      "grad_norm": 9.333070755004883,
      "learning_rate": 8.825385910754796e-06,
      "loss": 2.044743537902832,
      "memory(GiB)": 77.56,
      "step": 94295,
      "token_acc": 0.5625,
      "train_speed(iter/s)": 1.437934
    },
    {
      "epoch": 4.040101109635406,
      "grad_norm": 6.368394374847412,
      "learning_rate": 8.82156829858703e-06,
      "loss": 2.2585208892822264,
      "memory(GiB)": 77.56,
      "step": 94300,
      "token_acc": 0.50187265917603,
      "train_speed(iter/s)": 1.437931
    },
    {
      "epoch": 4.040315324964655,
      "grad_norm": 5.39070987701416,
      "learning_rate": 8.817751432405186e-06,
      "loss": 2.18239803314209,
      "memory(GiB)": 77.56,
      "step": 94305,
      "token_acc": 0.5197368421052632,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 4.040529540293903,
      "grad_norm": 7.199963092803955,
      "learning_rate": 8.813935312278414e-06,
      "loss": 2.1374454498291016,
      "memory(GiB)": 77.56,
      "step": 94310,
      "token_acc": 0.5467128027681661,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 4.0407437556231525,
      "grad_norm": 10.056708335876465,
      "learning_rate": 8.81011993827584e-06,
      "loss": 2.0617654800415037,
      "memory(GiB)": 77.56,
      "step": 94315,
      "token_acc": 0.5642201834862385,
      "train_speed(iter/s)": 1.437973
    },
    {
      "epoch": 4.040957970952402,
      "grad_norm": 5.356524467468262,
      "learning_rate": 8.806305310466579e-06,
      "loss": 2.0068901062011717,
      "memory(GiB)": 77.56,
      "step": 94320,
      "token_acc": 0.564,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 4.04117218628165,
      "grad_norm": 6.368059158325195,
      "learning_rate": 8.80249142891974e-06,
      "loss": 2.504183769226074,
      "memory(GiB)": 77.56,
      "step": 94325,
      "token_acc": 0.516,
      "train_speed(iter/s)": 1.43799
    },
    {
      "epoch": 4.041386401610899,
      "grad_norm": 7.350532054901123,
      "learning_rate": 8.798678293704405e-06,
      "loss": 2.2723342895507814,
      "memory(GiB)": 77.56,
      "step": 94330,
      "token_acc": 0.4961832061068702,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 4.041600616940149,
      "grad_norm": 4.41812801361084,
      "learning_rate": 8.794865904889677e-06,
      "loss": 1.8743091583251954,
      "memory(GiB)": 77.56,
      "step": 94335,
      "token_acc": 0.6118881118881119,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 4.041814832269397,
      "grad_norm": 7.434039115905762,
      "learning_rate": 8.791054262544601e-06,
      "loss": 2.0234777450561525,
      "memory(GiB)": 77.56,
      "step": 94340,
      "token_acc": 0.5508771929824562,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 4.042029047598646,
      "grad_norm": 8.150466918945312,
      "learning_rate": 8.78724336673824e-06,
      "loss": 2.347923469543457,
      "memory(GiB)": 77.56,
      "step": 94345,
      "token_acc": 0.5016181229773463,
      "train_speed(iter/s)": 1.438016
    },
    {
      "epoch": 4.0422432629278955,
      "grad_norm": 5.66418981552124,
      "learning_rate": 8.783433217539611e-06,
      "loss": 2.379938507080078,
      "memory(GiB)": 77.56,
      "step": 94350,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 4.042457478257144,
      "grad_norm": 7.125446796417236,
      "learning_rate": 8.779623815017763e-06,
      "loss": 2.1093027114868166,
      "memory(GiB)": 77.56,
      "step": 94355,
      "token_acc": 0.5347222222222222,
      "train_speed(iter/s)": 1.437997
    },
    {
      "epoch": 4.042671693586393,
      "grad_norm": 7.883285999298096,
      "learning_rate": 8.77581515924169e-06,
      "loss": 2.2761085510253904,
      "memory(GiB)": 77.56,
      "step": 94360,
      "token_acc": 0.4967105263157895,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 4.042885908915642,
      "grad_norm": 7.811742305755615,
      "learning_rate": 8.772007250280407e-06,
      "loss": 2.2774017333984373,
      "memory(GiB)": 77.56,
      "step": 94365,
      "token_acc": 0.5342019543973942,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 4.043100124244891,
      "grad_norm": 5.259608745574951,
      "learning_rate": 8.768200088202888e-06,
      "loss": 2.2066184997558596,
      "memory(GiB)": 77.56,
      "step": 94370,
      "token_acc": 0.5570934256055363,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 4.04331433957414,
      "grad_norm": 7.543445587158203,
      "learning_rate": 8.764393673078102e-06,
      "loss": 2.4556665420532227,
      "memory(GiB)": 77.56,
      "step": 94375,
      "token_acc": 0.45317220543806647,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 4.043528554903389,
      "grad_norm": 5.840571880340576,
      "learning_rate": 8.760588004975007e-06,
      "loss": 2.3214370727539064,
      "memory(GiB)": 77.56,
      "step": 94380,
      "token_acc": 0.5269121813031161,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 4.043742770232638,
      "grad_norm": 5.127938270568848,
      "learning_rate": 8.756783083962545e-06,
      "loss": 2.1085704803466796,
      "memory(GiB)": 77.56,
      "step": 94385,
      "token_acc": 0.5494880546075085,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 4.043956985561887,
      "grad_norm": 6.106111526489258,
      "learning_rate": 8.752978910109638e-06,
      "loss": 1.9957218170166016,
      "memory(GiB)": 77.56,
      "step": 94390,
      "token_acc": 0.5266903914590747,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 4.044171200891136,
      "grad_norm": 5.976374626159668,
      "learning_rate": 8.749175483485223e-06,
      "loss": 2.20544490814209,
      "memory(GiB)": 77.56,
      "step": 94395,
      "token_acc": 0.5342465753424658,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 4.0443854162203845,
      "grad_norm": 5.485307216644287,
      "learning_rate": 8.745372804158187e-06,
      "loss": 2.097411346435547,
      "memory(GiB)": 77.56,
      "step": 94400,
      "token_acc": 0.5338078291814946,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 4.044599631549634,
      "grad_norm": 5.8869147300720215,
      "learning_rate": 8.741570872197424e-06,
      "loss": 1.9910985946655273,
      "memory(GiB)": 77.56,
      "step": 94405,
      "token_acc": 0.5559440559440559,
      "train_speed(iter/s)": 1.437975
    },
    {
      "epoch": 4.044813846878883,
      "grad_norm": 5.397964000701904,
      "learning_rate": 8.73776968767181e-06,
      "loss": 2.1720096588134767,
      "memory(GiB)": 77.56,
      "step": 94410,
      "token_acc": 0.5754385964912281,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 4.045028062208131,
      "grad_norm": 7.197945594787598,
      "learning_rate": 8.733969250650187e-06,
      "loss": 2.190837860107422,
      "memory(GiB)": 77.56,
      "step": 94415,
      "token_acc": 0.5326797385620915,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 4.045242277537381,
      "grad_norm": 6.3045196533203125,
      "learning_rate": 8.73016956120144e-06,
      "loss": 2.2060741424560546,
      "memory(GiB)": 77.56,
      "step": 94420,
      "token_acc": 0.5245283018867924,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 4.04545649286663,
      "grad_norm": 5.721884727478027,
      "learning_rate": 8.72637061939437e-06,
      "loss": 2.2531305313110352,
      "memory(GiB)": 77.56,
      "step": 94425,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 4.045670708195878,
      "grad_norm": 6.655179023742676,
      "learning_rate": 8.722572425297832e-06,
      "loss": 2.2428781509399416,
      "memory(GiB)": 77.56,
      "step": 94430,
      "token_acc": 0.49842271293375395,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 4.0458849235251275,
      "grad_norm": 7.368620872497559,
      "learning_rate": 8.71877497898061e-06,
      "loss": 2.453813934326172,
      "memory(GiB)": 77.56,
      "step": 94435,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 4.046099138854377,
      "grad_norm": 9.210283279418945,
      "learning_rate": 8.714978280511505e-06,
      "loss": 2.456880569458008,
      "memory(GiB)": 77.56,
      "step": 94440,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 4.046313354183625,
      "grad_norm": 5.5849456787109375,
      "learning_rate": 8.711182329959284e-06,
      "loss": 2.386907196044922,
      "memory(GiB)": 77.56,
      "step": 94445,
      "token_acc": 0.4755700325732899,
      "train_speed(iter/s)": 1.438016
    },
    {
      "epoch": 4.046527569512874,
      "grad_norm": 5.408041954040527,
      "learning_rate": 8.70738712739274e-06,
      "loss": 2.325563430786133,
      "memory(GiB)": 77.56,
      "step": 94450,
      "token_acc": 0.4702549575070821,
      "train_speed(iter/s)": 1.438037
    },
    {
      "epoch": 4.046741784842124,
      "grad_norm": 5.731386661529541,
      "learning_rate": 8.703592672880617e-06,
      "loss": 2.483907699584961,
      "memory(GiB)": 77.56,
      "step": 94455,
      "token_acc": 0.4609164420485175,
      "train_speed(iter/s)": 1.43804
    },
    {
      "epoch": 4.046956000171372,
      "grad_norm": 7.373782634735107,
      "learning_rate": 8.699798966491651e-06,
      "loss": 2.0968038558959963,
      "memory(GiB)": 77.56,
      "step": 94460,
      "token_acc": 0.5520504731861199,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 4.047170215500621,
      "grad_norm": 7.003344535827637,
      "learning_rate": 8.696006008294566e-06,
      "loss": 2.1224441528320312,
      "memory(GiB)": 77.56,
      "step": 94465,
      "token_acc": 0.5379310344827586,
      "train_speed(iter/s)": 1.438032
    },
    {
      "epoch": 4.0473844308298705,
      "grad_norm": 6.923903942108154,
      "learning_rate": 8.692213798358084e-06,
      "loss": 2.5746427536010743,
      "memory(GiB)": 77.56,
      "step": 94470,
      "token_acc": 0.43731778425655976,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 4.047598646159119,
      "grad_norm": 5.497438907623291,
      "learning_rate": 8.688422336750879e-06,
      "loss": 2.2300783157348634,
      "memory(GiB)": 77.56,
      "step": 94475,
      "token_acc": 0.5408163265306123,
      "train_speed(iter/s)": 1.438061
    },
    {
      "epoch": 4.047812861488368,
      "grad_norm": 8.704301834106445,
      "learning_rate": 8.684631623541672e-06,
      "loss": 2.0712379455566405,
      "memory(GiB)": 77.56,
      "step": 94480,
      "token_acc": 0.5050847457627119,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 4.048027076817617,
      "grad_norm": 6.696622371673584,
      "learning_rate": 8.680841658799127e-06,
      "loss": 2.2620059967041017,
      "memory(GiB)": 77.56,
      "step": 94485,
      "token_acc": 0.5389408099688473,
      "train_speed(iter/s)": 1.438097
    },
    {
      "epoch": 4.048241292146866,
      "grad_norm": 6.093163013458252,
      "learning_rate": 8.677052442591888e-06,
      "loss": 2.1327383041381838,
      "memory(GiB)": 77.56,
      "step": 94490,
      "token_acc": 0.5069444444444444,
      "train_speed(iter/s)": 1.438107
    },
    {
      "epoch": 4.048455507476115,
      "grad_norm": 6.234643459320068,
      "learning_rate": 8.6732639749886e-06,
      "loss": 1.9161249160766602,
      "memory(GiB)": 77.56,
      "step": 94495,
      "token_acc": 0.5575757575757576,
      "train_speed(iter/s)": 1.438114
    },
    {
      "epoch": 4.048669722805364,
      "grad_norm": 6.737776756286621,
      "learning_rate": 8.669476256057913e-06,
      "loss": 2.289379119873047,
      "memory(GiB)": 77.56,
      "step": 94500,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.43813
    },
    {
      "epoch": 4.048669722805364,
      "eval_loss": 2.2524924278259277,
      "eval_runtime": 13.7648,
      "eval_samples_per_second": 7.265,
      "eval_steps_per_second": 7.265,
      "eval_token_acc": 0.45232273838630804,
      "step": 94500
    },
    {
      "epoch": 4.048883938134613,
      "grad_norm": 6.0029425621032715,
      "learning_rate": 8.665689285868416e-06,
      "loss": 2.143111801147461,
      "memory(GiB)": 77.56,
      "step": 94505,
      "token_acc": 0.47696476964769646,
      "train_speed(iter/s)": 1.437812
    },
    {
      "epoch": 4.049098153463862,
      "grad_norm": 7.486948490142822,
      "learning_rate": 8.661903064488753e-06,
      "loss": 2.164644241333008,
      "memory(GiB)": 77.56,
      "step": 94510,
      "token_acc": 0.5214521452145214,
      "train_speed(iter/s)": 1.43782
    },
    {
      "epoch": 4.049312368793111,
      "grad_norm": 5.806775093078613,
      "learning_rate": 8.65811759198749e-06,
      "loss": 2.0831167221069338,
      "memory(GiB)": 77.56,
      "step": 94515,
      "token_acc": 0.5513698630136986,
      "train_speed(iter/s)": 1.437837
    },
    {
      "epoch": 4.0495265841223596,
      "grad_norm": 6.2082343101501465,
      "learning_rate": 8.654332868433213e-06,
      "loss": 2.0956356048583986,
      "memory(GiB)": 77.56,
      "step": 94520,
      "token_acc": 0.5315985130111525,
      "train_speed(iter/s)": 1.437823
    },
    {
      "epoch": 4.049740799451609,
      "grad_norm": 4.95619535446167,
      "learning_rate": 8.650548893894484e-06,
      "loss": 2.246072006225586,
      "memory(GiB)": 77.56,
      "step": 94525,
      "token_acc": 0.5017301038062284,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 4.049955014780858,
      "grad_norm": 7.168889999389648,
      "learning_rate": 8.646765668439844e-06,
      "loss": 2.323457717895508,
      "memory(GiB)": 77.56,
      "step": 94530,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 4.050169230110106,
      "grad_norm": 5.401705741882324,
      "learning_rate": 8.642983192137827e-06,
      "loss": 2.381578063964844,
      "memory(GiB)": 77.56,
      "step": 94535,
      "token_acc": 0.5060606060606061,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 4.050383445439356,
      "grad_norm": 6.20339298248291,
      "learning_rate": 8.639201465056973e-06,
      "loss": 2.394364356994629,
      "memory(GiB)": 77.56,
      "step": 94540,
      "token_acc": 0.5125448028673835,
      "train_speed(iter/s)": 1.437798
    },
    {
      "epoch": 4.050597660768605,
      "grad_norm": 7.206087589263916,
      "learning_rate": 8.635420487265788e-06,
      "loss": 2.1705791473388674,
      "memory(GiB)": 77.56,
      "step": 94545,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.437805
    },
    {
      "epoch": 4.050811876097853,
      "grad_norm": 6.113037586212158,
      "learning_rate": 8.631640258832769e-06,
      "loss": 2.38806037902832,
      "memory(GiB)": 77.56,
      "step": 94550,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.437824
    },
    {
      "epoch": 4.051026091427103,
      "grad_norm": 7.499508857727051,
      "learning_rate": 8.627860779826385e-06,
      "loss": 2.2373558044433595,
      "memory(GiB)": 77.56,
      "step": 94555,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.437844
    },
    {
      "epoch": 4.051240306756352,
      "grad_norm": 6.184325695037842,
      "learning_rate": 8.624082050315107e-06,
      "loss": 2.2075767517089844,
      "memory(GiB)": 77.56,
      "step": 94560,
      "token_acc": 0.5404411764705882,
      "train_speed(iter/s)": 1.437845
    },
    {
      "epoch": 4.0514545220856,
      "grad_norm": 5.859940528869629,
      "learning_rate": 8.620304070367401e-06,
      "loss": 2.411662292480469,
      "memory(GiB)": 77.56,
      "step": 94565,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.437857
    },
    {
      "epoch": 4.0516687374148495,
      "grad_norm": 5.652826309204102,
      "learning_rate": 8.616526840051697e-06,
      "loss": 2.1072874069213867,
      "memory(GiB)": 77.56,
      "step": 94570,
      "token_acc": 0.519434628975265,
      "train_speed(iter/s)": 1.437864
    },
    {
      "epoch": 4.051882952744099,
      "grad_norm": 7.250621795654297,
      "learning_rate": 8.612750359436444e-06,
      "loss": 2.299967575073242,
      "memory(GiB)": 77.56,
      "step": 94575,
      "token_acc": 0.4903581267217631,
      "train_speed(iter/s)": 1.437865
    },
    {
      "epoch": 4.052097168073347,
      "grad_norm": 5.8685784339904785,
      "learning_rate": 8.608974628590044e-06,
      "loss": 2.617667388916016,
      "memory(GiB)": 77.56,
      "step": 94580,
      "token_acc": 0.47468354430379744,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 4.052311383402596,
      "grad_norm": 5.469445705413818,
      "learning_rate": 8.605199647580892e-06,
      "loss": 1.961815643310547,
      "memory(GiB)": 77.56,
      "step": 94585,
      "token_acc": 0.5443425076452599,
      "train_speed(iter/s)": 1.437894
    },
    {
      "epoch": 4.052525598731846,
      "grad_norm": 5.220660209655762,
      "learning_rate": 8.601425416477366e-06,
      "loss": 2.045301055908203,
      "memory(GiB)": 77.56,
      "step": 94590,
      "token_acc": 0.5451388888888888,
      "train_speed(iter/s)": 1.437881
    },
    {
      "epoch": 4.052739814061094,
      "grad_norm": 6.873932838439941,
      "learning_rate": 8.597651935347872e-06,
      "loss": 2.3389307022094727,
      "memory(GiB)": 77.56,
      "step": 94595,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.437891
    },
    {
      "epoch": 4.052954029390343,
      "grad_norm": 5.854475975036621,
      "learning_rate": 8.593879204260747e-06,
      "loss": 2.2878625869750975,
      "memory(GiB)": 77.56,
      "step": 94600,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.43789
    },
    {
      "epoch": 4.0531682447195925,
      "grad_norm": 5.454529285430908,
      "learning_rate": 8.590107223284344e-06,
      "loss": 2.299834442138672,
      "memory(GiB)": 77.56,
      "step": 94605,
      "token_acc": 0.48161764705882354,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 4.053382460048841,
      "grad_norm": 6.291232109069824,
      "learning_rate": 8.586335992486994e-06,
      "loss": 2.377471923828125,
      "memory(GiB)": 77.56,
      "step": 94610,
      "token_acc": 0.49411764705882355,
      "train_speed(iter/s)": 1.437895
    },
    {
      "epoch": 4.05359667537809,
      "grad_norm": 5.7743072509765625,
      "learning_rate": 8.582565511937019e-06,
      "loss": 2.4286264419555663,
      "memory(GiB)": 77.56,
      "step": 94615,
      "token_acc": 0.5376712328767124,
      "train_speed(iter/s)": 1.437903
    },
    {
      "epoch": 4.053810890707339,
      "grad_norm": 5.726933002471924,
      "learning_rate": 8.578795781702703e-06,
      "loss": 2.1536651611328126,
      "memory(GiB)": 77.56,
      "step": 94620,
      "token_acc": 0.4924812030075188,
      "train_speed(iter/s)": 1.43787
    },
    {
      "epoch": 4.054025106036588,
      "grad_norm": 5.126599311828613,
      "learning_rate": 8.575026801852375e-06,
      "loss": 2.1796525955200194,
      "memory(GiB)": 77.56,
      "step": 94625,
      "token_acc": 0.5085324232081911,
      "train_speed(iter/s)": 1.437866
    },
    {
      "epoch": 4.054239321365837,
      "grad_norm": 7.860447406768799,
      "learning_rate": 8.571258572454299e-06,
      "loss": 2.178411674499512,
      "memory(GiB)": 77.56,
      "step": 94630,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 4.054453536695086,
      "grad_norm": 5.048957824707031,
      "learning_rate": 8.567491093576735e-06,
      "loss": 2.1166738510131835,
      "memory(GiB)": 77.56,
      "step": 94635,
      "token_acc": 0.5180722891566265,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 4.054667752024335,
      "grad_norm": 4.927050590515137,
      "learning_rate": 8.563724365287934e-06,
      "loss": 2.2646278381347655,
      "memory(GiB)": 77.56,
      "step": 94640,
      "token_acc": 0.5075187969924813,
      "train_speed(iter/s)": 1.437896
    },
    {
      "epoch": 4.054881967353584,
      "grad_norm": 5.999354839324951,
      "learning_rate": 8.559958387656125e-06,
      "loss": 2.2153263092041016,
      "memory(GiB)": 77.56,
      "step": 94645,
      "token_acc": 0.5373665480427047,
      "train_speed(iter/s)": 1.437914
    },
    {
      "epoch": 4.055096182682833,
      "grad_norm": 7.3741374015808105,
      "learning_rate": 8.556193160749549e-06,
      "loss": 2.485906982421875,
      "memory(GiB)": 77.56,
      "step": 94650,
      "token_acc": 0.5091463414634146,
      "train_speed(iter/s)": 1.437926
    },
    {
      "epoch": 4.0553103980120815,
      "grad_norm": 4.946549892425537,
      "learning_rate": 8.552428684636416e-06,
      "loss": 2.4336305618286134,
      "memory(GiB)": 77.56,
      "step": 94655,
      "token_acc": 0.4907749077490775,
      "train_speed(iter/s)": 1.437949
    },
    {
      "epoch": 4.055524613341331,
      "grad_norm": 6.874972343444824,
      "learning_rate": 8.548664959384922e-06,
      "loss": 2.3196725845336914,
      "memory(GiB)": 77.56,
      "step": 94660,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.437943
    },
    {
      "epoch": 4.05573882867058,
      "grad_norm": 6.6315388679504395,
      "learning_rate": 8.544901985063247e-06,
      "loss": 2.160009765625,
      "memory(GiB)": 77.56,
      "step": 94665,
      "token_acc": 0.5318471337579618,
      "train_speed(iter/s)": 1.437963
    },
    {
      "epoch": 4.055953043999828,
      "grad_norm": 6.8424272537231445,
      "learning_rate": 8.541139761739559e-06,
      "loss": 2.456861877441406,
      "memory(GiB)": 77.56,
      "step": 94670,
      "token_acc": 0.49538461538461537,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 4.056167259329078,
      "grad_norm": 5.514702796936035,
      "learning_rate": 8.537378289482006e-06,
      "loss": 2.2611709594726563,
      "memory(GiB)": 77.56,
      "step": 94675,
      "token_acc": 0.56,
      "train_speed(iter/s)": 1.437997
    },
    {
      "epoch": 4.056381474658327,
      "grad_norm": 5.307720184326172,
      "learning_rate": 8.533617568358749e-06,
      "loss": 1.9555061340332032,
      "memory(GiB)": 77.56,
      "step": 94680,
      "token_acc": 0.5359712230215827,
      "train_speed(iter/s)": 1.438005
    },
    {
      "epoch": 4.056595689987575,
      "grad_norm": 6.721695423126221,
      "learning_rate": 8.529857598437912e-06,
      "loss": 2.4033859252929686,
      "memory(GiB)": 77.56,
      "step": 94685,
      "token_acc": 0.4473684210526316,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 4.0568099053168245,
      "grad_norm": 5.998652935028076,
      "learning_rate": 8.526098379787606e-06,
      "loss": 2.483084297180176,
      "memory(GiB)": 77.56,
      "step": 94690,
      "token_acc": 0.4688427299703264,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 4.057024120646074,
      "grad_norm": 5.612516403198242,
      "learning_rate": 8.522339912475936e-06,
      "loss": 2.3897886276245117,
      "memory(GiB)": 77.56,
      "step": 94695,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 4.057238335975322,
      "grad_norm": 6.510907173156738,
      "learning_rate": 8.51858219657098e-06,
      "loss": 2.3343429565429688,
      "memory(GiB)": 77.56,
      "step": 94700,
      "token_acc": 0.47191011235955055,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 4.057452551304571,
      "grad_norm": 5.609668254852295,
      "learning_rate": 8.514825232140816e-06,
      "loss": 2.215006637573242,
      "memory(GiB)": 77.56,
      "step": 94705,
      "token_acc": 0.5098684210526315,
      "train_speed(iter/s)": 1.438031
    },
    {
      "epoch": 4.057666766633821,
      "grad_norm": 5.9900078773498535,
      "learning_rate": 8.511069019253515e-06,
      "loss": 2.2272533416748046,
      "memory(GiB)": 77.56,
      "step": 94710,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 4.057880981963069,
      "grad_norm": 6.236502647399902,
      "learning_rate": 8.507313557977121e-06,
      "loss": 2.3736949920654298,
      "memory(GiB)": 77.56,
      "step": 94715,
      "token_acc": 0.49221183800623053,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 4.058095197292318,
      "grad_norm": 6.645170211791992,
      "learning_rate": 8.503558848379656e-06,
      "loss": 2.1365692138671877,
      "memory(GiB)": 77.56,
      "step": 94720,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 4.0583094126215675,
      "grad_norm": 6.495617389678955,
      "learning_rate": 8.499804890529157e-06,
      "loss": 2.329438400268555,
      "memory(GiB)": 77.56,
      "step": 94725,
      "token_acc": 0.47634069400630913,
      "train_speed(iter/s)": 1.437998
    },
    {
      "epoch": 4.058523627950816,
      "grad_norm": 6.9974164962768555,
      "learning_rate": 8.496051684493627e-06,
      "loss": 2.318832206726074,
      "memory(GiB)": 77.56,
      "step": 94730,
      "token_acc": 0.5236686390532544,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 4.058737843280065,
      "grad_norm": 5.69292688369751,
      "learning_rate": 8.492299230341039e-06,
      "loss": 2.136651611328125,
      "memory(GiB)": 77.56,
      "step": 94735,
      "token_acc": 0.5449101796407185,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 4.058952058609314,
      "grad_norm": 6.1332688331604,
      "learning_rate": 8.488547528139402e-06,
      "loss": 2.2839872360229494,
      "memory(GiB)": 77.56,
      "step": 94740,
      "token_acc": 0.5015197568389058,
      "train_speed(iter/s)": 1.43802
    },
    {
      "epoch": 4.059166273938563,
      "grad_norm": 6.985415458679199,
      "learning_rate": 8.484796577956666e-06,
      "loss": 1.9944267272949219,
      "memory(GiB)": 77.56,
      "step": 94745,
      "token_acc": 0.58203125,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 4.059380489267812,
      "grad_norm": 6.244665145874023,
      "learning_rate": 8.481046379860785e-06,
      "loss": 2.244589614868164,
      "memory(GiB)": 77.56,
      "step": 94750,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 4.059594704597061,
      "grad_norm": 5.623499870300293,
      "learning_rate": 8.477296933919693e-06,
      "loss": 2.019356918334961,
      "memory(GiB)": 77.56,
      "step": 94755,
      "token_acc": 0.5601503759398496,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 4.05980891992631,
      "grad_norm": 6.117869853973389,
      "learning_rate": 8.473548240201323e-06,
      "loss": 2.0456363677978517,
      "memory(GiB)": 77.56,
      "step": 94760,
      "token_acc": 0.532051282051282,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 4.060023135255559,
      "grad_norm": 7.197115898132324,
      "learning_rate": 8.469800298773567e-06,
      "loss": 2.1761476516723635,
      "memory(GiB)": 77.56,
      "step": 94765,
      "token_acc": 0.49145299145299143,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 4.060237350584808,
      "grad_norm": 5.124648094177246,
      "learning_rate": 8.466053109704353e-06,
      "loss": 2.4700817108154296,
      "memory(GiB)": 77.56,
      "step": 94770,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.438024
    },
    {
      "epoch": 4.0604515659140565,
      "grad_norm": 5.926012992858887,
      "learning_rate": 8.462306673061548e-06,
      "loss": 2.3996227264404295,
      "memory(GiB)": 77.56,
      "step": 94775,
      "token_acc": 0.5195195195195195,
      "train_speed(iter/s)": 1.438004
    },
    {
      "epoch": 4.060665781243306,
      "grad_norm": 6.4998369216918945,
      "learning_rate": 8.458560988913022e-06,
      "loss": 2.168090057373047,
      "memory(GiB)": 77.56,
      "step": 94780,
      "token_acc": 0.5298804780876494,
      "train_speed(iter/s)": 1.438004
    },
    {
      "epoch": 4.060879996572555,
      "grad_norm": 5.087442398071289,
      "learning_rate": 8.45481605732663e-06,
      "loss": 2.2478397369384764,
      "memory(GiB)": 77.56,
      "step": 94785,
      "token_acc": 0.5396825396825397,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 4.061094211901803,
      "grad_norm": 7.737916469573975,
      "learning_rate": 8.451071878370209e-06,
      "loss": 1.9614067077636719,
      "memory(GiB)": 77.56,
      "step": 94790,
      "token_acc": 0.5794392523364486,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 4.061308427231053,
      "grad_norm": 6.445511817932129,
      "learning_rate": 8.447328452111597e-06,
      "loss": 2.3380258560180662,
      "memory(GiB)": 77.56,
      "step": 94795,
      "token_acc": 0.519163763066202,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 4.061522642560302,
      "grad_norm": 6.61165714263916,
      "learning_rate": 8.443585778618618e-06,
      "loss": 2.4090763092041017,
      "memory(GiB)": 77.56,
      "step": 94800,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.438009
    },
    {
      "epoch": 4.06173685788955,
      "grad_norm": 4.864951133728027,
      "learning_rate": 8.439843857959074e-06,
      "loss": 2.5200191497802735,
      "memory(GiB)": 77.56,
      "step": 94805,
      "token_acc": 0.4803921568627451,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 4.0619510732188,
      "grad_norm": 8.043155670166016,
      "learning_rate": 8.436102690200737e-06,
      "loss": 2.285897445678711,
      "memory(GiB)": 77.56,
      "step": 94810,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 4.062165288548049,
      "grad_norm": 5.587471961975098,
      "learning_rate": 8.432362275411392e-06,
      "loss": 2.5362762451171874,
      "memory(GiB)": 77.56,
      "step": 94815,
      "token_acc": 0.46407185628742514,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 4.062379503877297,
      "grad_norm": 6.805018901824951,
      "learning_rate": 8.428622613658788e-06,
      "loss": 2.217032623291016,
      "memory(GiB)": 77.56,
      "step": 94820,
      "token_acc": 0.5048543689320388,
      "train_speed(iter/s)": 1.438021
    },
    {
      "epoch": 4.062593719206546,
      "grad_norm": 9.164910316467285,
      "learning_rate": 8.424883705010689e-06,
      "loss": 2.153001403808594,
      "memory(GiB)": 77.56,
      "step": 94825,
      "token_acc": 0.5323529411764706,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 4.062807934535796,
      "grad_norm": 5.749707221984863,
      "learning_rate": 8.421145549534826e-06,
      "loss": 2.01830997467041,
      "memory(GiB)": 77.56,
      "step": 94830,
      "token_acc": 0.5645161290322581,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 4.063022149865044,
      "grad_norm": 5.759202003479004,
      "learning_rate": 8.417408147298916e-06,
      "loss": 2.1515344619750976,
      "memory(GiB)": 77.56,
      "step": 94835,
      "token_acc": 0.5316455696202531,
      "train_speed(iter/s)": 1.438026
    },
    {
      "epoch": 4.063236365194293,
      "grad_norm": 6.5611772537231445,
      "learning_rate": 8.413671498370667e-06,
      "loss": 2.1368858337402346,
      "memory(GiB)": 77.56,
      "step": 94840,
      "token_acc": 0.5470383275261324,
      "train_speed(iter/s)": 1.438019
    },
    {
      "epoch": 4.063450580523543,
      "grad_norm": 4.8147292137146,
      "learning_rate": 8.409935602817764e-06,
      "loss": 2.245065689086914,
      "memory(GiB)": 77.56,
      "step": 94845,
      "token_acc": 0.5323076923076923,
      "train_speed(iter/s)": 1.438031
    },
    {
      "epoch": 4.063664795852791,
      "grad_norm": 6.519495964050293,
      "learning_rate": 8.406200460707885e-06,
      "loss": 2.3183284759521485,
      "memory(GiB)": 77.56,
      "step": 94850,
      "token_acc": 0.5242718446601942,
      "train_speed(iter/s)": 1.438026
    },
    {
      "epoch": 4.06387901118204,
      "grad_norm": 4.928180694580078,
      "learning_rate": 8.402466072108716e-06,
      "loss": 2.3318574905395506,
      "memory(GiB)": 77.56,
      "step": 94855,
      "token_acc": 0.4970588235294118,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 4.0640932265112895,
      "grad_norm": 6.480074405670166,
      "learning_rate": 8.398732437087892e-06,
      "loss": 2.319177436828613,
      "memory(GiB)": 77.56,
      "step": 94860,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.438027
    },
    {
      "epoch": 4.064307441840538,
      "grad_norm": 10.078136444091797,
      "learning_rate": 8.394999555713045e-06,
      "loss": 2.37449951171875,
      "memory(GiB)": 77.56,
      "step": 94865,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 4.064521657169787,
      "grad_norm": 5.725994110107422,
      "learning_rate": 8.391267428051819e-06,
      "loss": 2.318703842163086,
      "memory(GiB)": 77.56,
      "step": 94870,
      "token_acc": 0.5105105105105106,
      "train_speed(iter/s)": 1.438055
    },
    {
      "epoch": 4.064735872499036,
      "grad_norm": 8.054520606994629,
      "learning_rate": 8.387536054171818e-06,
      "loss": 2.2320878982543944,
      "memory(GiB)": 77.56,
      "step": 94875,
      "token_acc": 0.5066666666666667,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 4.064950087828285,
      "grad_norm": 8.072264671325684,
      "learning_rate": 8.383805434140623e-06,
      "loss": 2.4040828704833985,
      "memory(GiB)": 77.56,
      "step": 94880,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 4.065164303157534,
      "grad_norm": 5.176023960113525,
      "learning_rate": 8.380075568025847e-06,
      "loss": 2.440072250366211,
      "memory(GiB)": 77.56,
      "step": 94885,
      "token_acc": 0.46440677966101696,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.065378518486783,
      "grad_norm": 6.106320381164551,
      "learning_rate": 8.376346455895046e-06,
      "loss": 2.0221248626708985,
      "memory(GiB)": 77.56,
      "step": 94890,
      "token_acc": 0.5494505494505495,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 4.065592733816032,
      "grad_norm": 7.271626949310303,
      "learning_rate": 8.37261809781577e-06,
      "loss": 2.5189109802246095,
      "memory(GiB)": 77.56,
      "step": 94895,
      "token_acc": 0.46567164179104475,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 4.065806949145281,
      "grad_norm": 4.877325057983398,
      "learning_rate": 8.368890493855568e-06,
      "loss": 2.062717819213867,
      "memory(GiB)": 77.56,
      "step": 94900,
      "token_acc": 0.5578231292517006,
      "train_speed(iter/s)": 1.438055
    },
    {
      "epoch": 4.06602116447453,
      "grad_norm": 7.08648157119751,
      "learning_rate": 8.365163644081969e-06,
      "loss": 2.3857030868530273,
      "memory(GiB)": 77.56,
      "step": 94905,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 4.0662353798037785,
      "grad_norm": 8.799745559692383,
      "learning_rate": 8.361437548562478e-06,
      "loss": 2.10213623046875,
      "memory(GiB)": 77.56,
      "step": 94910,
      "token_acc": 0.5481481481481482,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 4.066449595133028,
      "grad_norm": 6.067258834838867,
      "learning_rate": 8.357712207364615e-06,
      "loss": 2.6216259002685547,
      "memory(GiB)": 77.56,
      "step": 94915,
      "token_acc": 0.4670846394984326,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 4.066663810462277,
      "grad_norm": 5.18054723739624,
      "learning_rate": 8.353987620555859e-06,
      "loss": 2.0616338729858397,
      "memory(GiB)": 77.56,
      "step": 94920,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 4.066878025791525,
      "grad_norm": 12.234222412109375,
      "learning_rate": 8.350263788203682e-06,
      "loss": 2.207724189758301,
      "memory(GiB)": 77.56,
      "step": 94925,
      "token_acc": 0.5517241379310345,
      "train_speed(iter/s)": 1.438092
    },
    {
      "epoch": 4.067092241120775,
      "grad_norm": 7.175808906555176,
      "learning_rate": 8.34654071037555e-06,
      "loss": 2.473559761047363,
      "memory(GiB)": 77.56,
      "step": 94930,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.438111
    },
    {
      "epoch": 4.067306456450024,
      "grad_norm": 7.485994338989258,
      "learning_rate": 8.342818387138895e-06,
      "loss": 2.137923240661621,
      "memory(GiB)": 77.56,
      "step": 94935,
      "token_acc": 0.5397350993377483,
      "train_speed(iter/s)": 1.438097
    },
    {
      "epoch": 4.067520671779272,
      "grad_norm": 5.164398670196533,
      "learning_rate": 8.339096818561165e-06,
      "loss": 2.4431785583496093,
      "memory(GiB)": 77.56,
      "step": 94940,
      "token_acc": 0.5162337662337663,
      "train_speed(iter/s)": 1.43809
    },
    {
      "epoch": 4.0677348871085215,
      "grad_norm": 5.50698184967041,
      "learning_rate": 8.335376004709784e-06,
      "loss": 2.2709802627563476,
      "memory(GiB)": 77.56,
      "step": 94945,
      "token_acc": 0.5227272727272727,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 4.067949102437771,
      "grad_norm": 8.534979820251465,
      "learning_rate": 8.331655945652155e-06,
      "loss": 2.074360466003418,
      "memory(GiB)": 77.56,
      "step": 94950,
      "token_acc": 0.52046783625731,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 4.068163317767019,
      "grad_norm": 8.147920608520508,
      "learning_rate": 8.327936641455663e-06,
      "loss": 2.441698455810547,
      "memory(GiB)": 77.56,
      "step": 94955,
      "token_acc": 0.5148936170212766,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 4.068377533096268,
      "grad_norm": 6.128410339355469,
      "learning_rate": 8.324218092187696e-06,
      "loss": 2.061117935180664,
      "memory(GiB)": 77.56,
      "step": 94960,
      "token_acc": 0.582995951417004,
      "train_speed(iter/s)": 1.438083
    },
    {
      "epoch": 4.068591748425518,
      "grad_norm": 5.585254669189453,
      "learning_rate": 8.320500297915595e-06,
      "loss": 2.482101821899414,
      "memory(GiB)": 77.56,
      "step": 94965,
      "token_acc": 0.4716312056737589,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 4.068805963754766,
      "grad_norm": 5.491774082183838,
      "learning_rate": 8.316783258706745e-06,
      "loss": 2.174637222290039,
      "memory(GiB)": 77.56,
      "step": 94970,
      "token_acc": 0.5292207792207793,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 4.069020179084015,
      "grad_norm": 7.608911991119385,
      "learning_rate": 8.31306697462847e-06,
      "loss": 2.291824150085449,
      "memory(GiB)": 77.56,
      "step": 94975,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.43807
    },
    {
      "epoch": 4.0692343944132645,
      "grad_norm": 6.652978420257568,
      "learning_rate": 8.309351445748086e-06,
      "loss": 2.495310592651367,
      "memory(GiB)": 77.56,
      "step": 94980,
      "token_acc": 0.5034013605442177,
      "train_speed(iter/s)": 1.438083
    },
    {
      "epoch": 4.069448609742513,
      "grad_norm": 4.828470706939697,
      "learning_rate": 8.305636672132916e-06,
      "loss": 2.0697349548339843,
      "memory(GiB)": 77.56,
      "step": 94985,
      "token_acc": 0.562962962962963,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 4.069662825071762,
      "grad_norm": 6.481742858886719,
      "learning_rate": 8.301922653850247e-06,
      "loss": 2.1012456893920897,
      "memory(GiB)": 77.56,
      "step": 94990,
      "token_acc": 0.5328467153284672,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 4.069877040401011,
      "grad_norm": 6.6270246505737305,
      "learning_rate": 8.298209390967355e-06,
      "loss": 2.248515319824219,
      "memory(GiB)": 77.56,
      "step": 94995,
      "token_acc": 0.5282258064516129,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 4.07009125573026,
      "grad_norm": 5.540470123291016,
      "learning_rate": 8.294496883551528e-06,
      "loss": 2.281123161315918,
      "memory(GiB)": 77.56,
      "step": 95000,
      "token_acc": 0.5102739726027398,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 4.07009125573026,
      "eval_loss": 2.3116610050201416,
      "eval_runtime": 14.0226,
      "eval_samples_per_second": 7.131,
      "eval_steps_per_second": 7.131,
      "eval_token_acc": 0.44823232323232326,
      "step": 95000
    },
    {
      "epoch": 4.070305471059509,
      "grad_norm": 5.085848331451416,
      "learning_rate": 8.290785131670014e-06,
      "loss": 2.211634635925293,
      "memory(GiB)": 77.56,
      "step": 95005,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 4.070519686388758,
      "grad_norm": 6.199611186981201,
      "learning_rate": 8.287074135390061e-06,
      "loss": 2.1248369216918945,
      "memory(GiB)": 77.56,
      "step": 95010,
      "token_acc": 0.5701357466063348,
      "train_speed(iter/s)": 1.437784
    },
    {
      "epoch": 4.070733901718007,
      "grad_norm": 7.530384063720703,
      "learning_rate": 8.283363894778868e-06,
      "loss": 2.291860580444336,
      "memory(GiB)": 77.56,
      "step": 95015,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.437786
    },
    {
      "epoch": 4.070948117047256,
      "grad_norm": 5.3922576904296875,
      "learning_rate": 8.279654409903687e-06,
      "loss": 2.3500675201416015,
      "memory(GiB)": 77.56,
      "step": 95020,
      "token_acc": 0.508833922261484,
      "train_speed(iter/s)": 1.437787
    },
    {
      "epoch": 4.071162332376505,
      "grad_norm": 4.7777557373046875,
      "learning_rate": 8.275945680831693e-06,
      "loss": 2.093250846862793,
      "memory(GiB)": 77.56,
      "step": 95025,
      "token_acc": 0.5708502024291497,
      "train_speed(iter/s)": 1.43778
    },
    {
      "epoch": 4.0713765477057535,
      "grad_norm": 5.861032485961914,
      "learning_rate": 8.272237707630092e-06,
      "loss": 2.6367023468017576,
      "memory(GiB)": 77.56,
      "step": 95030,
      "token_acc": 0.4919093851132686,
      "train_speed(iter/s)": 1.437769
    },
    {
      "epoch": 4.071590763035003,
      "grad_norm": 5.4793877601623535,
      "learning_rate": 8.26853049036605e-06,
      "loss": 2.145457458496094,
      "memory(GiB)": 77.56,
      "step": 95035,
      "token_acc": 0.5575539568345323,
      "train_speed(iter/s)": 1.43777
    },
    {
      "epoch": 4.071804978364252,
      "grad_norm": 8.698028564453125,
      "learning_rate": 8.264824029106727e-06,
      "loss": 2.360030746459961,
      "memory(GiB)": 77.56,
      "step": 95040,
      "token_acc": 0.4927536231884058,
      "train_speed(iter/s)": 1.437766
    },
    {
      "epoch": 4.0720191936935,
      "grad_norm": 6.546703815460205,
      "learning_rate": 8.261118323919265e-06,
      "loss": 2.737155532836914,
      "memory(GiB)": 77.56,
      "step": 95045,
      "token_acc": 0.4881889763779528,
      "train_speed(iter/s)": 1.437757
    },
    {
      "epoch": 4.07223340902275,
      "grad_norm": 8.735042572021484,
      "learning_rate": 8.257413374870804e-06,
      "loss": 2.4486717224121093,
      "memory(GiB)": 77.56,
      "step": 95050,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.437774
    },
    {
      "epoch": 4.072447624351999,
      "grad_norm": 7.635506629943848,
      "learning_rate": 8.25370918202844e-06,
      "loss": 2.271725654602051,
      "memory(GiB)": 77.56,
      "step": 95055,
      "token_acc": 0.5643939393939394,
      "train_speed(iter/s)": 1.437795
    },
    {
      "epoch": 4.072661839681247,
      "grad_norm": 6.147125244140625,
      "learning_rate": 8.25000574545931e-06,
      "loss": 2.3958587646484375,
      "memory(GiB)": 77.56,
      "step": 95060,
      "token_acc": 0.5033333333333333,
      "train_speed(iter/s)": 1.437809
    },
    {
      "epoch": 4.0728760550104965,
      "grad_norm": 7.892317295074463,
      "learning_rate": 8.246303065230488e-06,
      "loss": 2.2090301513671875,
      "memory(GiB)": 77.56,
      "step": 95065,
      "token_acc": 0.5358490566037736,
      "train_speed(iter/s)": 1.437815
    },
    {
      "epoch": 4.073090270339746,
      "grad_norm": 6.296213626861572,
      "learning_rate": 8.242601141409056e-06,
      "loss": 2.2790971755981446,
      "memory(GiB)": 77.56,
      "step": 95070,
      "token_acc": 0.532608695652174,
      "train_speed(iter/s)": 1.437835
    },
    {
      "epoch": 4.073304485668994,
      "grad_norm": 5.616121768951416,
      "learning_rate": 8.238899974062076e-06,
      "loss": 2.07140998840332,
      "memory(GiB)": 77.56,
      "step": 95075,
      "token_acc": 0.5268456375838926,
      "train_speed(iter/s)": 1.43783
    },
    {
      "epoch": 4.073518700998243,
      "grad_norm": 6.632471084594727,
      "learning_rate": 8.235199563256585e-06,
      "loss": 2.316982460021973,
      "memory(GiB)": 77.56,
      "step": 95080,
      "token_acc": 0.49836065573770494,
      "train_speed(iter/s)": 1.437819
    },
    {
      "epoch": 4.073732916327493,
      "grad_norm": 6.675464630126953,
      "learning_rate": 8.231499909059642e-06,
      "loss": 2.4138263702392577,
      "memory(GiB)": 77.56,
      "step": 95085,
      "token_acc": 0.47019867549668876,
      "train_speed(iter/s)": 1.437834
    },
    {
      "epoch": 4.073947131656741,
      "grad_norm": 5.686821460723877,
      "learning_rate": 8.227801011538256e-06,
      "loss": 2.0893932342529298,
      "memory(GiB)": 77.56,
      "step": 95090,
      "token_acc": 0.5178571428571429,
      "train_speed(iter/s)": 1.437832
    },
    {
      "epoch": 4.07416134698599,
      "grad_norm": 7.414421081542969,
      "learning_rate": 8.224102870759448e-06,
      "loss": 2.199736785888672,
      "memory(GiB)": 77.56,
      "step": 95095,
      "token_acc": 0.5283018867924528,
      "train_speed(iter/s)": 1.437835
    },
    {
      "epoch": 4.07437556231524,
      "grad_norm": 7.159857273101807,
      "learning_rate": 8.2204054867902e-06,
      "loss": 2.5327426910400392,
      "memory(GiB)": 77.56,
      "step": 95100,
      "token_acc": 0.48135593220338985,
      "train_speed(iter/s)": 1.437841
    },
    {
      "epoch": 4.074589777644488,
      "grad_norm": 7.560205459594727,
      "learning_rate": 8.216708859697508e-06,
      "loss": 2.262774658203125,
      "memory(GiB)": 77.56,
      "step": 95105,
      "token_acc": 0.49056603773584906,
      "train_speed(iter/s)": 1.437824
    },
    {
      "epoch": 4.074803992973737,
      "grad_norm": 4.729887962341309,
      "learning_rate": 8.213012989548313e-06,
      "loss": 2.0913330078125,
      "memory(GiB)": 77.56,
      "step": 95110,
      "token_acc": 0.5749128919860628,
      "train_speed(iter/s)": 1.437839
    },
    {
      "epoch": 4.0750182083029864,
      "grad_norm": 5.275120735168457,
      "learning_rate": 8.209317876409601e-06,
      "loss": 2.399083709716797,
      "memory(GiB)": 77.56,
      "step": 95115,
      "token_acc": 0.5071942446043165,
      "train_speed(iter/s)": 1.437854
    },
    {
      "epoch": 4.075232423632235,
      "grad_norm": 5.846557140350342,
      "learning_rate": 8.205623520348299e-06,
      "loss": 2.39022216796875,
      "memory(GiB)": 77.56,
      "step": 95120,
      "token_acc": 0.5051903114186851,
      "train_speed(iter/s)": 1.437867
    },
    {
      "epoch": 4.075446638961484,
      "grad_norm": 8.769886016845703,
      "learning_rate": 8.201929921431334e-06,
      "loss": 2.584098052978516,
      "memory(GiB)": 77.56,
      "step": 95125,
      "token_acc": 0.43956043956043955,
      "train_speed(iter/s)": 1.437883
    },
    {
      "epoch": 4.075660854290733,
      "grad_norm": 6.7576212882995605,
      "learning_rate": 8.198237079725613e-06,
      "loss": 2.396385383605957,
      "memory(GiB)": 77.56,
      "step": 95130,
      "token_acc": 0.5303030303030303,
      "train_speed(iter/s)": 1.437884
    },
    {
      "epoch": 4.075875069619982,
      "grad_norm": 5.972720146179199,
      "learning_rate": 8.194544995298048e-06,
      "loss": 2.2690196990966798,
      "memory(GiB)": 77.56,
      "step": 95135,
      "token_acc": 0.5469255663430421,
      "train_speed(iter/s)": 1.437879
    },
    {
      "epoch": 4.076089284949231,
      "grad_norm": 7.026678085327148,
      "learning_rate": 8.190853668215498e-06,
      "loss": 2.2759994506835937,
      "memory(GiB)": 77.56,
      "step": 95140,
      "token_acc": 0.5424354243542435,
      "train_speed(iter/s)": 1.437874
    },
    {
      "epoch": 4.07630350027848,
      "grad_norm": 7.014134883880615,
      "learning_rate": 8.18716309854487e-06,
      "loss": 2.3330326080322266,
      "memory(GiB)": 77.56,
      "step": 95145,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437891
    },
    {
      "epoch": 4.076517715607729,
      "grad_norm": 7.18118143081665,
      "learning_rate": 8.183473286353005e-06,
      "loss": 2.4699195861816405,
      "memory(GiB)": 77.56,
      "step": 95150,
      "token_acc": 0.46534653465346537,
      "train_speed(iter/s)": 1.43791
    },
    {
      "epoch": 4.076731930936978,
      "grad_norm": 6.186906814575195,
      "learning_rate": 8.179784231706749e-06,
      "loss": 2.1483314514160154,
      "memory(GiB)": 77.56,
      "step": 95155,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.437905
    },
    {
      "epoch": 4.076946146266227,
      "grad_norm": 5.541762828826904,
      "learning_rate": 8.176095934672933e-06,
      "loss": 2.1984859466552735,
      "memory(GiB)": 77.56,
      "step": 95160,
      "token_acc": 0.5270758122743683,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 4.0771603615954755,
      "grad_norm": 5.22767448425293,
      "learning_rate": 8.172408395318359e-06,
      "loss": 2.2991353988647463,
      "memory(GiB)": 77.56,
      "step": 95165,
      "token_acc": 0.5207547169811321,
      "train_speed(iter/s)": 1.437926
    },
    {
      "epoch": 4.077374576924725,
      "grad_norm": 6.797808647155762,
      "learning_rate": 8.168721613709846e-06,
      "loss": 2.3800064086914063,
      "memory(GiB)": 77.56,
      "step": 95170,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.437926
    },
    {
      "epoch": 4.077588792253974,
      "grad_norm": 6.5170578956604,
      "learning_rate": 8.165035589914193e-06,
      "loss": 2.2007110595703123,
      "memory(GiB)": 77.56,
      "step": 95175,
      "token_acc": 0.4939759036144578,
      "train_speed(iter/s)": 1.43794
    },
    {
      "epoch": 4.077803007583222,
      "grad_norm": 6.947213172912598,
      "learning_rate": 8.161350323998169e-06,
      "loss": 2.13214168548584,
      "memory(GiB)": 77.56,
      "step": 95180,
      "token_acc": 0.5136186770428015,
      "train_speed(iter/s)": 1.437945
    },
    {
      "epoch": 4.078017222912472,
      "grad_norm": 7.823688507080078,
      "learning_rate": 8.157665816028525e-06,
      "loss": 2.567250061035156,
      "memory(GiB)": 77.56,
      "step": 95185,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.437949
    },
    {
      "epoch": 4.078231438241721,
      "grad_norm": 5.183547019958496,
      "learning_rate": 8.153982066072018e-06,
      "loss": 2.184250068664551,
      "memory(GiB)": 77.56,
      "step": 95190,
      "token_acc": 0.5348837209302325,
      "train_speed(iter/s)": 1.437942
    },
    {
      "epoch": 4.078445653570969,
      "grad_norm": 6.618625164031982,
      "learning_rate": 8.150299074195383e-06,
      "loss": 2.26678352355957,
      "memory(GiB)": 77.56,
      "step": 95195,
      "token_acc": 0.5665399239543726,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 4.0786598689002185,
      "grad_norm": 5.132879734039307,
      "learning_rate": 8.146616840465326e-06,
      "loss": 2.4960214614868166,
      "memory(GiB)": 77.56,
      "step": 95200,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 4.078874084229468,
      "grad_norm": 7.099595069885254,
      "learning_rate": 8.142935364948579e-06,
      "loss": 2.251189041137695,
      "memory(GiB)": 77.56,
      "step": 95205,
      "token_acc": 0.5431034482758621,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 4.079088299558716,
      "grad_norm": 5.392854690551758,
      "learning_rate": 8.13925464771182e-06,
      "loss": 2.2247493743896483,
      "memory(GiB)": 77.56,
      "step": 95210,
      "token_acc": 0.51953125,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 4.079302514887965,
      "grad_norm": 6.9510297775268555,
      "learning_rate": 8.135574688821734e-06,
      "loss": 2.124391555786133,
      "memory(GiB)": 77.56,
      "step": 95215,
      "token_acc": 0.4936708860759494,
      "train_speed(iter/s)": 1.437965
    },
    {
      "epoch": 4.079516730217215,
      "grad_norm": 6.948002338409424,
      "learning_rate": 8.131895488344981e-06,
      "loss": 2.3184295654296876,
      "memory(GiB)": 77.56,
      "step": 95220,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437986
    },
    {
      "epoch": 4.079730945546463,
      "grad_norm": 6.519289493560791,
      "learning_rate": 8.128217046348207e-06,
      "loss": 2.0918804168701173,
      "memory(GiB)": 77.56,
      "step": 95225,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 4.079945160875712,
      "grad_norm": 5.949084758758545,
      "learning_rate": 8.124539362898076e-06,
      "loss": 2.291040802001953,
      "memory(GiB)": 77.56,
      "step": 95230,
      "token_acc": 0.5723905723905723,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 4.0801593762049615,
      "grad_norm": 6.39797306060791,
      "learning_rate": 8.120862438061183e-06,
      "loss": 2.545256996154785,
      "memory(GiB)": 77.56,
      "step": 95235,
      "token_acc": 0.4570446735395189,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 4.08037359153421,
      "grad_norm": 7.161879539489746,
      "learning_rate": 8.117186271904158e-06,
      "loss": 2.200242042541504,
      "memory(GiB)": 77.56,
      "step": 95240,
      "token_acc": 0.5627118644067797,
      "train_speed(iter/s)": 1.43803
    },
    {
      "epoch": 4.080587806863459,
      "grad_norm": 6.7964582443237305,
      "learning_rate": 8.113510864493602e-06,
      "loss": 2.7155387878417967,
      "memory(GiB)": 77.56,
      "step": 95245,
      "token_acc": 0.47474747474747475,
      "train_speed(iter/s)": 1.438041
    },
    {
      "epoch": 4.080802022192708,
      "grad_norm": 7.678805351257324,
      "learning_rate": 8.109836215896083e-06,
      "loss": 2.2545888900756834,
      "memory(GiB)": 77.56,
      "step": 95250,
      "token_acc": 0.5177865612648221,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 4.081016237521957,
      "grad_norm": 5.471217155456543,
      "learning_rate": 8.106162326178168e-06,
      "loss": 2.272845458984375,
      "memory(GiB)": 77.56,
      "step": 95255,
      "token_acc": 0.5429553264604811,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 4.081230452851206,
      "grad_norm": 6.315338611602783,
      "learning_rate": 8.102489195406432e-06,
      "loss": 2.491302490234375,
      "memory(GiB)": 77.56,
      "step": 95260,
      "token_acc": 0.5350553505535055,
      "train_speed(iter/s)": 1.438044
    },
    {
      "epoch": 4.081444668180455,
      "grad_norm": 5.879712104797363,
      "learning_rate": 8.09881682364741e-06,
      "loss": 2.1852466583251955,
      "memory(GiB)": 77.56,
      "step": 95265,
      "token_acc": 0.532319391634981,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 4.081658883509704,
      "grad_norm": 7.742741584777832,
      "learning_rate": 8.095145210967624e-06,
      "loss": 2.003104400634766,
      "memory(GiB)": 77.56,
      "step": 95270,
      "token_acc": 0.5421245421245421,
      "train_speed(iter/s)": 1.438035
    },
    {
      "epoch": 4.081873098838953,
      "grad_norm": 5.597198963165283,
      "learning_rate": 8.091474357433598e-06,
      "loss": 2.4399093627929687,
      "memory(GiB)": 77.56,
      "step": 95275,
      "token_acc": 0.5043103448275862,
      "train_speed(iter/s)": 1.438047
    },
    {
      "epoch": 4.082087314168202,
      "grad_norm": 5.738831996917725,
      "learning_rate": 8.087804263111825e-06,
      "loss": 1.9840042114257812,
      "memory(GiB)": 77.56,
      "step": 95280,
      "token_acc": 0.5518394648829431,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 4.0823015294974505,
      "grad_norm": 8.161035537719727,
      "learning_rate": 8.084134928068782e-06,
      "loss": 2.3944358825683594,
      "memory(GiB)": 77.56,
      "step": 95285,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 4.0825157448267,
      "grad_norm": 5.912302017211914,
      "learning_rate": 8.080466352370968e-06,
      "loss": 2.1009111404418945,
      "memory(GiB)": 77.56,
      "step": 95290,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 4.082729960155949,
      "grad_norm": 6.509933948516846,
      "learning_rate": 8.076798536084828e-06,
      "loss": 2.1947967529296877,
      "memory(GiB)": 77.56,
      "step": 95295,
      "token_acc": 0.5368098159509203,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 4.082944175485197,
      "grad_norm": 7.838466644287109,
      "learning_rate": 8.073131479276807e-06,
      "loss": 2.0913082122802735,
      "memory(GiB)": 77.56,
      "step": 95300,
      "token_acc": 0.5543478260869565,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 4.083158390814447,
      "grad_norm": 6.435009479522705,
      "learning_rate": 8.069465182013347e-06,
      "loss": 2.442860984802246,
      "memory(GiB)": 77.56,
      "step": 95305,
      "token_acc": 0.5273972602739726,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 4.083372606143696,
      "grad_norm": 5.454482555389404,
      "learning_rate": 8.065799644360844e-06,
      "loss": 2.46114501953125,
      "memory(GiB)": 77.56,
      "step": 95310,
      "token_acc": 0.48633879781420764,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 4.083586821472944,
      "grad_norm": 5.948737621307373,
      "learning_rate": 8.062134866385717e-06,
      "loss": 2.3383861541748048,
      "memory(GiB)": 77.56,
      "step": 95315,
      "token_acc": 0.5221843003412969,
      "train_speed(iter/s)": 1.438078
    },
    {
      "epoch": 4.0838010368021935,
      "grad_norm": 7.446720123291016,
      "learning_rate": 8.05847084815437e-06,
      "loss": 2.4039730072021483,
      "memory(GiB)": 77.56,
      "step": 95320,
      "token_acc": 0.4754601226993865,
      "train_speed(iter/s)": 1.43807
    },
    {
      "epoch": 4.084015252131443,
      "grad_norm": 8.093862533569336,
      "learning_rate": 8.054807589733166e-06,
      "loss": 2.3856197357177735,
      "memory(GiB)": 77.56,
      "step": 95325,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 4.084229467460691,
      "grad_norm": 4.595720291137695,
      "learning_rate": 8.051145091188472e-06,
      "loss": 2.3056467056274412,
      "memory(GiB)": 77.56,
      "step": 95330,
      "token_acc": 0.5381944444444444,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 4.08444368278994,
      "grad_norm": 6.777374267578125,
      "learning_rate": 8.047483352586638e-06,
      "loss": 2.122158432006836,
      "memory(GiB)": 77.56,
      "step": 95335,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 4.08465789811919,
      "grad_norm": 5.51237154006958,
      "learning_rate": 8.043822373993997e-06,
      "loss": 2.2516345977783203,
      "memory(GiB)": 77.56,
      "step": 95340,
      "token_acc": 0.5633802816901409,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 4.084872113448438,
      "grad_norm": 8.379335403442383,
      "learning_rate": 8.04016215547686e-06,
      "loss": 2.58554630279541,
      "memory(GiB)": 77.56,
      "step": 95345,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 4.085086328777687,
      "grad_norm": 6.422924041748047,
      "learning_rate": 8.036502697101555e-06,
      "loss": 2.3697158813476564,
      "memory(GiB)": 77.56,
      "step": 95350,
      "token_acc": 0.4897959183673469,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 4.0853005441069365,
      "grad_norm": 6.77065896987915,
      "learning_rate": 8.032843998934369e-06,
      "loss": 2.2962230682373046,
      "memory(GiB)": 77.56,
      "step": 95355,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 4.085514759436185,
      "grad_norm": 5.855922698974609,
      "learning_rate": 8.029186061041588e-06,
      "loss": 2.3217077255249023,
      "memory(GiB)": 77.56,
      "step": 95360,
      "token_acc": 0.49221183800623053,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 4.085728974765434,
      "grad_norm": 7.765806198120117,
      "learning_rate": 8.025528883489468e-06,
      "loss": 2.5206104278564454,
      "memory(GiB)": 77.56,
      "step": 95365,
      "token_acc": 0.44966442953020136,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.085943190094683,
      "grad_norm": 9.420412063598633,
      "learning_rate": 8.02187246634426e-06,
      "loss": 2.0320037841796874,
      "memory(GiB)": 77.56,
      "step": 95370,
      "token_acc": 0.5256410256410257,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 4.086157405423932,
      "grad_norm": 5.942191123962402,
      "learning_rate": 8.018216809672219e-06,
      "loss": 2.470602607727051,
      "memory(GiB)": 77.56,
      "step": 95375,
      "token_acc": 0.4811594202898551,
      "train_speed(iter/s)": 1.43809
    },
    {
      "epoch": 4.086371620753181,
      "grad_norm": 5.013623237609863,
      "learning_rate": 8.014561913539565e-06,
      "loss": 2.254461097717285,
      "memory(GiB)": 77.56,
      "step": 95380,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 4.08658583608243,
      "grad_norm": 5.06242036819458,
      "learning_rate": 8.010907778012494e-06,
      "loss": 2.2041616439819336,
      "memory(GiB)": 77.56,
      "step": 95385,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.438118
    },
    {
      "epoch": 4.086800051411679,
      "grad_norm": 4.987133502960205,
      "learning_rate": 8.007254403157233e-06,
      "loss": 2.111385726928711,
      "memory(GiB)": 77.56,
      "step": 95390,
      "token_acc": 0.5672131147540984,
      "train_speed(iter/s)": 1.438125
    },
    {
      "epoch": 4.087014266740928,
      "grad_norm": 6.158591270446777,
      "learning_rate": 8.003601789039944e-06,
      "loss": 2.337290573120117,
      "memory(GiB)": 77.56,
      "step": 95395,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.43814
    },
    {
      "epoch": 4.087228482070177,
      "grad_norm": 8.32453441619873,
      "learning_rate": 7.999949935726797e-06,
      "loss": 2.31163330078125,
      "memory(GiB)": 77.56,
      "step": 95400,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 4.087442697399426,
      "grad_norm": 7.872544765472412,
      "learning_rate": 7.996298843283967e-06,
      "loss": 2.3166744232177736,
      "memory(GiB)": 77.56,
      "step": 95405,
      "token_acc": 0.5276872964169381,
      "train_speed(iter/s)": 1.438148
    },
    {
      "epoch": 4.087656912728675,
      "grad_norm": 5.44281530380249,
      "learning_rate": 7.99264851177759e-06,
      "loss": 2.1454681396484374,
      "memory(GiB)": 77.56,
      "step": 95410,
      "token_acc": 0.5422535211267606,
      "train_speed(iter/s)": 1.438161
    },
    {
      "epoch": 4.087871128057924,
      "grad_norm": 5.412315845489502,
      "learning_rate": 7.988998941273784e-06,
      "loss": 2.164523124694824,
      "memory(GiB)": 77.56,
      "step": 95415,
      "token_acc": 0.5192307692307693,
      "train_speed(iter/s)": 1.438161
    },
    {
      "epoch": 4.088085343387172,
      "grad_norm": 5.925240993499756,
      "learning_rate": 7.985350131838676e-06,
      "loss": 2.5320085525512694,
      "memory(GiB)": 77.56,
      "step": 95420,
      "token_acc": 0.48056537102473496,
      "train_speed(iter/s)": 1.43817
    },
    {
      "epoch": 4.088299558716422,
      "grad_norm": 9.179483413696289,
      "learning_rate": 7.981702083538368e-06,
      "loss": 2.29078369140625,
      "memory(GiB)": 77.56,
      "step": 95425,
      "token_acc": 0.5140845070422535,
      "train_speed(iter/s)": 1.438184
    },
    {
      "epoch": 4.088513774045671,
      "grad_norm": 6.44094181060791,
      "learning_rate": 7.978054796438922e-06,
      "loss": 2.2283884048461915,
      "memory(GiB)": 77.56,
      "step": 95430,
      "token_acc": 0.5271317829457365,
      "train_speed(iter/s)": 1.438188
    },
    {
      "epoch": 4.088727989374919,
      "grad_norm": 5.382179260253906,
      "learning_rate": 7.974408270606448e-06,
      "loss": 2.0547115325927736,
      "memory(GiB)": 77.56,
      "step": 95435,
      "token_acc": 0.5709219858156028,
      "train_speed(iter/s)": 1.438188
    },
    {
      "epoch": 4.088942204704169,
      "grad_norm": 5.694599151611328,
      "learning_rate": 7.970762506106993e-06,
      "loss": 2.3554187774658204,
      "memory(GiB)": 77.56,
      "step": 95440,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.4382
    },
    {
      "epoch": 4.089156420033418,
      "grad_norm": 6.321893692016602,
      "learning_rate": 7.967117503006604e-06,
      "loss": 2.4103702545166015,
      "memory(GiB)": 77.56,
      "step": 95445,
      "token_acc": 0.4868913857677903,
      "train_speed(iter/s)": 1.438216
    },
    {
      "epoch": 4.089370635362666,
      "grad_norm": 6.720142364501953,
      "learning_rate": 7.963473261371307e-06,
      "loss": 2.3507190704345704,
      "memory(GiB)": 77.56,
      "step": 95450,
      "token_acc": 0.5314465408805031,
      "train_speed(iter/s)": 1.438202
    },
    {
      "epoch": 4.0895848506919155,
      "grad_norm": 7.978154182434082,
      "learning_rate": 7.959829781267114e-06,
      "loss": 2.2015880584716796,
      "memory(GiB)": 77.56,
      "step": 95455,
      "token_acc": 0.5072992700729927,
      "train_speed(iter/s)": 1.438195
    },
    {
      "epoch": 4.089799066021165,
      "grad_norm": 4.7595415115356445,
      "learning_rate": 7.956187062760045e-06,
      "loss": 2.0481498718261717,
      "memory(GiB)": 77.56,
      "step": 95460,
      "token_acc": 0.5399239543726235,
      "train_speed(iter/s)": 1.438206
    },
    {
      "epoch": 4.090013281350413,
      "grad_norm": 7.336482048034668,
      "learning_rate": 7.952545105916098e-06,
      "loss": 2.499127960205078,
      "memory(GiB)": 77.56,
      "step": 95465,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.438227
    },
    {
      "epoch": 4.090227496679662,
      "grad_norm": 6.1849365234375,
      "learning_rate": 7.948903910801236e-06,
      "loss": 2.1779363632202147,
      "memory(GiB)": 77.56,
      "step": 95470,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.438226
    },
    {
      "epoch": 4.090441712008912,
      "grad_norm": 5.000257968902588,
      "learning_rate": 7.945263477481434e-06,
      "loss": 2.1321340560913087,
      "memory(GiB)": 77.56,
      "step": 95475,
      "token_acc": 0.5507246376811594,
      "train_speed(iter/s)": 1.438216
    },
    {
      "epoch": 4.09065592733816,
      "grad_norm": 5.467508792877197,
      "learning_rate": 7.94162380602263e-06,
      "loss": 2.602534294128418,
      "memory(GiB)": 77.56,
      "step": 95480,
      "token_acc": 0.46464646464646464,
      "train_speed(iter/s)": 1.438223
    },
    {
      "epoch": 4.090870142667409,
      "grad_norm": 7.177868366241455,
      "learning_rate": 7.937984896490763e-06,
      "loss": 2.151432418823242,
      "memory(GiB)": 77.56,
      "step": 95485,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.438229
    },
    {
      "epoch": 4.0910843579966585,
      "grad_norm": 5.606530666351318,
      "learning_rate": 7.934346748951748e-06,
      "loss": 2.462385559082031,
      "memory(GiB)": 77.56,
      "step": 95490,
      "token_acc": 0.49843260188087773,
      "train_speed(iter/s)": 1.438235
    },
    {
      "epoch": 4.091298573325907,
      "grad_norm": 5.938510417938232,
      "learning_rate": 7.930709363471517e-06,
      "loss": 2.3911277770996096,
      "memory(GiB)": 77.56,
      "step": 95495,
      "token_acc": 0.49240121580547114,
      "train_speed(iter/s)": 1.438229
    },
    {
      "epoch": 4.091512788655156,
      "grad_norm": 5.3229217529296875,
      "learning_rate": 7.927072740115943e-06,
      "loss": 2.1502012252807616,
      "memory(GiB)": 77.56,
      "step": 95500,
      "token_acc": 0.5331230283911672,
      "train_speed(iter/s)": 1.438228
    },
    {
      "epoch": 4.091512788655156,
      "eval_loss": 2.348863124847412,
      "eval_runtime": 13.8788,
      "eval_samples_per_second": 7.205,
      "eval_steps_per_second": 7.205,
      "eval_token_acc": 0.4654498044328553,
      "step": 95500
    },
    {
      "epoch": 4.091727003984405,
      "grad_norm": 6.539514541625977,
      "learning_rate": 7.923436878950919e-06,
      "loss": 2.063608169555664,
      "memory(GiB)": 77.56,
      "step": 95505,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.437913
    },
    {
      "epoch": 4.091941219313654,
      "grad_norm": 6.420368194580078,
      "learning_rate": 7.919801780042307e-06,
      "loss": 2.0076530456542967,
      "memory(GiB)": 77.56,
      "step": 95510,
      "token_acc": 0.5300353356890459,
      "train_speed(iter/s)": 1.437926
    },
    {
      "epoch": 4.092155434642903,
      "grad_norm": 7.928882122039795,
      "learning_rate": 7.916167443455946e-06,
      "loss": 2.3196598052978517,
      "memory(GiB)": 77.56,
      "step": 95515,
      "token_acc": 0.5046153846153846,
      "train_speed(iter/s)": 1.437938
    },
    {
      "epoch": 4.092369649972152,
      "grad_norm": 5.978614807128906,
      "learning_rate": 7.912533869257704e-06,
      "loss": 2.494196128845215,
      "memory(GiB)": 77.56,
      "step": 95520,
      "token_acc": 0.4833948339483395,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 4.092583865301401,
      "grad_norm": 5.221742630004883,
      "learning_rate": 7.90890105751339e-06,
      "loss": 2.2820737838745115,
      "memory(GiB)": 77.56,
      "step": 95525,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 4.09279808063065,
      "grad_norm": 6.007336139678955,
      "learning_rate": 7.905269008288807e-06,
      "loss": 2.6127859115600587,
      "memory(GiB)": 77.56,
      "step": 95530,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.43797
    },
    {
      "epoch": 4.093012295959899,
      "grad_norm": 7.333731651306152,
      "learning_rate": 7.901637721649774e-06,
      "loss": 2.227700424194336,
      "memory(GiB)": 77.56,
      "step": 95535,
      "token_acc": 0.5087719298245614,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 4.0932265112891475,
      "grad_norm": 5.782073020935059,
      "learning_rate": 7.898007197662066e-06,
      "loss": 2.4427804946899414,
      "memory(GiB)": 77.56,
      "step": 95540,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.437967
    },
    {
      "epoch": 4.093440726618397,
      "grad_norm": 5.386437892913818,
      "learning_rate": 7.894377436391443e-06,
      "loss": 2.1094970703125,
      "memory(GiB)": 77.56,
      "step": 95545,
      "token_acc": 0.5207667731629393,
      "train_speed(iter/s)": 1.43796
    },
    {
      "epoch": 4.093654941947646,
      "grad_norm": 6.629706859588623,
      "learning_rate": 7.890748437903677e-06,
      "loss": 2.4512868881225587,
      "memory(GiB)": 77.56,
      "step": 95550,
      "token_acc": 0.5171102661596958,
      "train_speed(iter/s)": 1.437943
    },
    {
      "epoch": 4.093869157276894,
      "grad_norm": 6.183679580688477,
      "learning_rate": 7.887120202264514e-06,
      "loss": 2.1981393814086916,
      "memory(GiB)": 77.56,
      "step": 95555,
      "token_acc": 0.49640287769784175,
      "train_speed(iter/s)": 1.43792
    },
    {
      "epoch": 4.094083372606144,
      "grad_norm": 8.030348777770996,
      "learning_rate": 7.883492729539665e-06,
      "loss": 2.1019378662109376,
      "memory(GiB)": 77.56,
      "step": 95560,
      "token_acc": 0.5289256198347108,
      "train_speed(iter/s)": 1.437935
    },
    {
      "epoch": 4.094297587935393,
      "grad_norm": 5.193393707275391,
      "learning_rate": 7.879866019794858e-06,
      "loss": 2.3798805236816407,
      "memory(GiB)": 77.56,
      "step": 95565,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 4.094511803264641,
      "grad_norm": 8.052562713623047,
      "learning_rate": 7.876240073095793e-06,
      "loss": 2.3297948837280273,
      "memory(GiB)": 77.56,
      "step": 95570,
      "token_acc": 0.5060606060606061,
      "train_speed(iter/s)": 1.437971
    },
    {
      "epoch": 4.0947260185938905,
      "grad_norm": 6.5791401863098145,
      "learning_rate": 7.872614889508134e-06,
      "loss": 2.3634870529174803,
      "memory(GiB)": 77.56,
      "step": 95575,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.437967
    },
    {
      "epoch": 4.09494023392314,
      "grad_norm": 5.681127548217773,
      "learning_rate": 7.868990469097593e-06,
      "loss": 2.1815195083618164,
      "memory(GiB)": 77.56,
      "step": 95580,
      "token_acc": 0.5306859205776173,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 4.095154449252388,
      "grad_norm": 6.2353925704956055,
      "learning_rate": 7.86536681192981e-06,
      "loss": 2.1699281692504884,
      "memory(GiB)": 77.56,
      "step": 95585,
      "token_acc": 0.5482758620689655,
      "train_speed(iter/s)": 1.437968
    },
    {
      "epoch": 4.095368664581637,
      "grad_norm": 7.309560298919678,
      "learning_rate": 7.861743918070435e-06,
      "loss": 2.4608787536621093,
      "memory(GiB)": 77.56,
      "step": 95590,
      "token_acc": 0.49258160237388726,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 4.095582879910887,
      "grad_norm": 7.188143730163574,
      "learning_rate": 7.858121787585093e-06,
      "loss": 1.9250286102294922,
      "memory(GiB)": 77.56,
      "step": 95595,
      "token_acc": 0.588,
      "train_speed(iter/s)": 1.437992
    },
    {
      "epoch": 4.095797095240135,
      "grad_norm": 6.157402515411377,
      "learning_rate": 7.854500420539401e-06,
      "loss": 2.354156494140625,
      "memory(GiB)": 77.56,
      "step": 95600,
      "token_acc": 0.5194805194805194,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 4.096011310569384,
      "grad_norm": 7.745563507080078,
      "learning_rate": 7.850879816998962e-06,
      "loss": 2.1147571563720704,
      "memory(GiB)": 77.56,
      "step": 95605,
      "token_acc": 0.5211864406779662,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 4.0962255258986335,
      "grad_norm": 6.6154704093933105,
      "learning_rate": 7.847259977029392e-06,
      "loss": 2.508718490600586,
      "memory(GiB)": 77.56,
      "step": 95610,
      "token_acc": 0.4430769230769231,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 4.096439741227882,
      "grad_norm": 6.540365219116211,
      "learning_rate": 7.843640900696247e-06,
      "loss": 1.9651012420654297,
      "memory(GiB)": 77.56,
      "step": 95615,
      "token_acc": 0.581081081081081,
      "train_speed(iter/s)": 1.43798
    },
    {
      "epoch": 4.096653956557131,
      "grad_norm": 4.798427104949951,
      "learning_rate": 7.840022588065098e-06,
      "loss": 2.5056270599365233,
      "memory(GiB)": 77.56,
      "step": 95620,
      "token_acc": 0.47572815533980584,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 4.09686817188638,
      "grad_norm": 6.073946475982666,
      "learning_rate": 7.836405039201483e-06,
      "loss": 2.686945343017578,
      "memory(GiB)": 77.56,
      "step": 95625,
      "token_acc": 0.4906832298136646,
      "train_speed(iter/s)": 1.437987
    },
    {
      "epoch": 4.097082387215629,
      "grad_norm": 5.64340877532959,
      "learning_rate": 7.832788254170948e-06,
      "loss": 2.186273956298828,
      "memory(GiB)": 77.56,
      "step": 95630,
      "token_acc": 0.5478260869565217,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 4.097296602544878,
      "grad_norm": 9.057547569274902,
      "learning_rate": 7.829172233038996e-06,
      "loss": 2.23011474609375,
      "memory(GiB)": 77.56,
      "step": 95635,
      "token_acc": 0.5331230283911672,
      "train_speed(iter/s)": 1.438001
    },
    {
      "epoch": 4.097510817874127,
      "grad_norm": 4.938108444213867,
      "learning_rate": 7.825556975871156e-06,
      "loss": 2.4074224472045898,
      "memory(GiB)": 77.56,
      "step": 95640,
      "token_acc": 0.513677811550152,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 4.097725033203376,
      "grad_norm": 6.268854141235352,
      "learning_rate": 7.821942482732918e-06,
      "loss": 2.087177276611328,
      "memory(GiB)": 77.56,
      "step": 95645,
      "token_acc": 0.5434782608695652,
      "train_speed(iter/s)": 1.437996
    },
    {
      "epoch": 4.097939248532625,
      "grad_norm": 5.036936283111572,
      "learning_rate": 7.818328753689763e-06,
      "loss": 2.321675109863281,
      "memory(GiB)": 77.56,
      "step": 95650,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.437987
    },
    {
      "epoch": 4.098153463861874,
      "grad_norm": 7.010135173797607,
      "learning_rate": 7.814715788807148e-06,
      "loss": 2.2630308151245115,
      "memory(GiB)": 77.56,
      "step": 95655,
      "token_acc": 0.5158227848101266,
      "train_speed(iter/s)": 1.437983
    },
    {
      "epoch": 4.0983676791911225,
      "grad_norm": 7.799376964569092,
      "learning_rate": 7.811103588150514e-06,
      "loss": 2.4632816314697266,
      "memory(GiB)": 77.56,
      "step": 95660,
      "token_acc": 0.4613003095975232,
      "train_speed(iter/s)": 1.437979
    },
    {
      "epoch": 4.098581894520372,
      "grad_norm": 6.319108963012695,
      "learning_rate": 7.807492151785334e-06,
      "loss": 2.0347557067871094,
      "memory(GiB)": 77.56,
      "step": 95665,
      "token_acc": 0.5224489795918368,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 4.098796109849621,
      "grad_norm": 5.453060626983643,
      "learning_rate": 7.803881479777008e-06,
      "loss": 2.3898683547973634,
      "memory(GiB)": 77.56,
      "step": 95670,
      "token_acc": 0.495114006514658,
      "train_speed(iter/s)": 1.437994
    },
    {
      "epoch": 4.099010325178869,
      "grad_norm": 6.550273418426514,
      "learning_rate": 7.800271572190954e-06,
      "loss": 2.3743898391723635,
      "memory(GiB)": 77.56,
      "step": 95675,
      "token_acc": 0.5093167701863354,
      "train_speed(iter/s)": 1.437999
    },
    {
      "epoch": 4.099224540508119,
      "grad_norm": 6.600028991699219,
      "learning_rate": 7.796662429092555e-06,
      "loss": 2.088513946533203,
      "memory(GiB)": 77.56,
      "step": 95680,
      "token_acc": 0.5443037974683544,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 4.099438755837368,
      "grad_norm": 5.347016334533691,
      "learning_rate": 7.793054050547215e-06,
      "loss": 2.2152591705322267,
      "memory(GiB)": 77.56,
      "step": 95685,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.438031
    },
    {
      "epoch": 4.099652971166616,
      "grad_norm": 5.001875877380371,
      "learning_rate": 7.78944643662029e-06,
      "loss": 2.079903221130371,
      "memory(GiB)": 77.56,
      "step": 95690,
      "token_acc": 0.532258064516129,
      "train_speed(iter/s)": 1.438043
    },
    {
      "epoch": 4.099867186495866,
      "grad_norm": 6.1742682456970215,
      "learning_rate": 7.785839587377141e-06,
      "loss": 2.180209922790527,
      "memory(GiB)": 77.56,
      "step": 95695,
      "token_acc": 0.5306859205776173,
      "train_speed(iter/s)": 1.43804
    },
    {
      "epoch": 4.100081401825115,
      "grad_norm": 7.490435600280762,
      "learning_rate": 7.782233502883119e-06,
      "loss": 2.44027042388916,
      "memory(GiB)": 77.56,
      "step": 95700,
      "token_acc": 0.4416961130742049,
      "train_speed(iter/s)": 1.438058
    },
    {
      "epoch": 4.100295617154363,
      "grad_norm": 5.422607898712158,
      "learning_rate": 7.778628183203535e-06,
      "loss": 2.24871826171875,
      "memory(GiB)": 77.56,
      "step": 95705,
      "token_acc": 0.5113636363636364,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 4.1005098324836124,
      "grad_norm": 5.190501689910889,
      "learning_rate": 7.775023628403705e-06,
      "loss": 2.4379295349121093,
      "memory(GiB)": 77.56,
      "step": 95710,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.100724047812862,
      "grad_norm": 5.1958723068237305,
      "learning_rate": 7.771419838548938e-06,
      "loss": 2.0504478454589843,
      "memory(GiB)": 77.56,
      "step": 95715,
      "token_acc": 0.5514705882352942,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 4.10093826314211,
      "grad_norm": 5.958120346069336,
      "learning_rate": 7.767816813704499e-06,
      "loss": 2.1624988555908202,
      "memory(GiB)": 77.56,
      "step": 95720,
      "token_acc": 0.4961832061068702,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 4.101152478471359,
      "grad_norm": 5.108565807342529,
      "learning_rate": 7.76421455393569e-06,
      "loss": 2.1604690551757812,
      "memory(GiB)": 77.56,
      "step": 95725,
      "token_acc": 0.5089285714285714,
      "train_speed(iter/s)": 1.438108
    },
    {
      "epoch": 4.101366693800609,
      "grad_norm": 6.5323286056518555,
      "learning_rate": 7.760613059307748e-06,
      "loss": 2.244904708862305,
      "memory(GiB)": 77.56,
      "step": 95730,
      "token_acc": 0.5447761194029851,
      "train_speed(iter/s)": 1.438131
    },
    {
      "epoch": 4.101580909129857,
      "grad_norm": 5.427577018737793,
      "learning_rate": 7.757012329885933e-06,
      "loss": 2.2815195083618165,
      "memory(GiB)": 77.56,
      "step": 95735,
      "token_acc": 0.5239520958083832,
      "train_speed(iter/s)": 1.438148
    },
    {
      "epoch": 4.101795124459106,
      "grad_norm": 6.407330513000488,
      "learning_rate": 7.753412365735463e-06,
      "loss": 2.197022247314453,
      "memory(GiB)": 77.56,
      "step": 95740,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.438152
    },
    {
      "epoch": 4.1020093397883555,
      "grad_norm": 5.737429141998291,
      "learning_rate": 7.749813166921543e-06,
      "loss": 2.3980236053466797,
      "memory(GiB)": 77.56,
      "step": 95745,
      "token_acc": 0.4709480122324159,
      "train_speed(iter/s)": 1.43817
    },
    {
      "epoch": 4.102223555117604,
      "grad_norm": 5.706313610076904,
      "learning_rate": 7.746214733509411e-06,
      "loss": 2.078744125366211,
      "memory(GiB)": 77.56,
      "step": 95750,
      "token_acc": 0.534965034965035,
      "train_speed(iter/s)": 1.438184
    },
    {
      "epoch": 4.102437770446853,
      "grad_norm": 5.231309413909912,
      "learning_rate": 7.742617065564217e-06,
      "loss": 2.2042591094970705,
      "memory(GiB)": 77.56,
      "step": 95755,
      "token_acc": 0.5351351351351351,
      "train_speed(iter/s)": 1.438175
    },
    {
      "epoch": 4.102651985776102,
      "grad_norm": 5.8632378578186035,
      "learning_rate": 7.739020163151173e-06,
      "loss": 2.340458297729492,
      "memory(GiB)": 77.56,
      "step": 95760,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438177
    },
    {
      "epoch": 4.102866201105351,
      "grad_norm": 4.819559097290039,
      "learning_rate": 7.73542402633542e-06,
      "loss": 2.2115365982055666,
      "memory(GiB)": 77.56,
      "step": 95765,
      "token_acc": 0.5285714285714286,
      "train_speed(iter/s)": 1.4382
    },
    {
      "epoch": 4.1030804164346,
      "grad_norm": 5.6769537925720215,
      "learning_rate": 7.731828655182105e-06,
      "loss": 2.208267402648926,
      "memory(GiB)": 77.56,
      "step": 95770,
      "token_acc": 0.5418060200668896,
      "train_speed(iter/s)": 1.438197
    },
    {
      "epoch": 4.103294631763849,
      "grad_norm": 6.025320529937744,
      "learning_rate": 7.728234049756372e-06,
      "loss": 2.002666473388672,
      "memory(GiB)": 77.56,
      "step": 95775,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.43821
    },
    {
      "epoch": 4.103508847093098,
      "grad_norm": 6.932184219360352,
      "learning_rate": 7.724640210123312e-06,
      "loss": 2.4075464248657226,
      "memory(GiB)": 77.56,
      "step": 95780,
      "token_acc": 0.46417445482866043,
      "train_speed(iter/s)": 1.43822
    },
    {
      "epoch": 4.103723062422347,
      "grad_norm": 6.757833003997803,
      "learning_rate": 7.721047136348076e-06,
      "loss": 2.2383880615234375,
      "memory(GiB)": 77.56,
      "step": 95785,
      "token_acc": 0.5186440677966102,
      "train_speed(iter/s)": 1.438218
    },
    {
      "epoch": 4.103937277751596,
      "grad_norm": 5.736269950866699,
      "learning_rate": 7.717454828495724e-06,
      "loss": 2.28125114440918,
      "memory(GiB)": 77.56,
      "step": 95790,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.438219
    },
    {
      "epoch": 4.1041514930808445,
      "grad_norm": 5.992380619049072,
      "learning_rate": 7.71386328663134e-06,
      "loss": 2.301872444152832,
      "memory(GiB)": 77.56,
      "step": 95795,
      "token_acc": 0.5198776758409785,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 4.104365708410094,
      "grad_norm": 9.938249588012695,
      "learning_rate": 7.710272510819993e-06,
      "loss": 1.8846904754638671,
      "memory(GiB)": 77.56,
      "step": 95800,
      "token_acc": 0.5397489539748954,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 4.104579923739343,
      "grad_norm": 7.297255992889404,
      "learning_rate": 7.706682501126722e-06,
      "loss": 2.024399185180664,
      "memory(GiB)": 77.56,
      "step": 95805,
      "token_acc": 0.5242718446601942,
      "train_speed(iter/s)": 1.438224
    },
    {
      "epoch": 4.104794139068591,
      "grad_norm": 5.847188472747803,
      "learning_rate": 7.703093257616579e-06,
      "loss": 2.338427734375,
      "memory(GiB)": 77.56,
      "step": 95810,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438226
    },
    {
      "epoch": 4.105008354397841,
      "grad_norm": 5.563513278961182,
      "learning_rate": 7.699504780354582e-06,
      "loss": 2.284145736694336,
      "memory(GiB)": 77.56,
      "step": 95815,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.438242
    },
    {
      "epoch": 4.10522256972709,
      "grad_norm": 6.084682464599609,
      "learning_rate": 7.69591706940574e-06,
      "loss": 2.309156036376953,
      "memory(GiB)": 77.56,
      "step": 95820,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.438257
    },
    {
      "epoch": 4.105436785056338,
      "grad_norm": 6.657647609710693,
      "learning_rate": 7.69233012483504e-06,
      "loss": 2.151595115661621,
      "memory(GiB)": 77.56,
      "step": 95825,
      "token_acc": 0.5534351145038168,
      "train_speed(iter/s)": 1.438268
    },
    {
      "epoch": 4.1056510003855875,
      "grad_norm": 6.458507061004639,
      "learning_rate": 7.688743946707455e-06,
      "loss": 2.043415069580078,
      "memory(GiB)": 77.56,
      "step": 95830,
      "token_acc": 0.5304347826086957,
      "train_speed(iter/s)": 1.438267
    },
    {
      "epoch": 4.105865215714837,
      "grad_norm": 6.901049613952637,
      "learning_rate": 7.685158535087966e-06,
      "loss": 2.2042892456054686,
      "memory(GiB)": 77.56,
      "step": 95835,
      "token_acc": 0.5670103092783505,
      "train_speed(iter/s)": 1.438262
    },
    {
      "epoch": 4.106079431044085,
      "grad_norm": 5.083637714385986,
      "learning_rate": 7.681573890041539e-06,
      "loss": 2.165730094909668,
      "memory(GiB)": 77.56,
      "step": 95840,
      "token_acc": 0.5508196721311476,
      "train_speed(iter/s)": 1.438262
    },
    {
      "epoch": 4.106293646373334,
      "grad_norm": 8.376875877380371,
      "learning_rate": 7.677990011633096e-06,
      "loss": 2.294547271728516,
      "memory(GiB)": 77.56,
      "step": 95845,
      "token_acc": 0.483271375464684,
      "train_speed(iter/s)": 1.438274
    },
    {
      "epoch": 4.106507861702584,
      "grad_norm": 6.624063491821289,
      "learning_rate": 7.674406899927567e-06,
      "loss": 2.3447238922119142,
      "memory(GiB)": 77.56,
      "step": 95850,
      "token_acc": 0.5146443514644351,
      "train_speed(iter/s)": 1.438287
    },
    {
      "epoch": 4.106722077031832,
      "grad_norm": 8.687882423400879,
      "learning_rate": 7.670824554989858e-06,
      "loss": 2.3343679428100588,
      "memory(GiB)": 77.56,
      "step": 95855,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.438302
    },
    {
      "epoch": 4.106936292361081,
      "grad_norm": 5.211456298828125,
      "learning_rate": 7.667242976884875e-06,
      "loss": 2.2146095275878905,
      "memory(GiB)": 77.56,
      "step": 95860,
      "token_acc": 0.5668789808917197,
      "train_speed(iter/s)": 1.438301
    },
    {
      "epoch": 4.1071505076903305,
      "grad_norm": 6.3015217781066895,
      "learning_rate": 7.663662165677481e-06,
      "loss": 2.1318674087524414,
      "memory(GiB)": 77.56,
      "step": 95865,
      "token_acc": 0.5165165165165165,
      "train_speed(iter/s)": 1.438315
    },
    {
      "epoch": 4.107364723019579,
      "grad_norm": 5.2183027267456055,
      "learning_rate": 7.660082121432577e-06,
      "loss": 2.179535675048828,
      "memory(GiB)": 77.56,
      "step": 95870,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.438326
    },
    {
      "epoch": 4.107578938348828,
      "grad_norm": 6.747520923614502,
      "learning_rate": 7.656502844214996e-06,
      "loss": 2.480525588989258,
      "memory(GiB)": 77.56,
      "step": 95875,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.438313
    },
    {
      "epoch": 4.107793153678077,
      "grad_norm": 8.101838111877441,
      "learning_rate": 7.652924334089594e-06,
      "loss": 2.334914779663086,
      "memory(GiB)": 77.56,
      "step": 95880,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.438301
    },
    {
      "epoch": 4.108007369007326,
      "grad_norm": 6.330286026000977,
      "learning_rate": 7.649346591121193e-06,
      "loss": 2.2796926498413086,
      "memory(GiB)": 77.56,
      "step": 95885,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.438311
    },
    {
      "epoch": 4.108221584336575,
      "grad_norm": 6.005471706390381,
      "learning_rate": 7.645769615374592e-06,
      "loss": 2.2894432067871096,
      "memory(GiB)": 77.56,
      "step": 95890,
      "token_acc": 0.5425867507886435,
      "train_speed(iter/s)": 1.438319
    },
    {
      "epoch": 4.108435799665824,
      "grad_norm": 7.703975677490234,
      "learning_rate": 7.642193406914621e-06,
      "loss": 2.3478599548339845,
      "memory(GiB)": 77.56,
      "step": 95895,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 4.108650014995073,
      "grad_norm": 6.016992092132568,
      "learning_rate": 7.638617965806038e-06,
      "loss": 2.451349639892578,
      "memory(GiB)": 77.56,
      "step": 95900,
      "token_acc": 0.5111940298507462,
      "train_speed(iter/s)": 1.438318
    },
    {
      "epoch": 4.108864230324322,
      "grad_norm": 6.158310413360596,
      "learning_rate": 7.635043292113637e-06,
      "loss": 2.1001445770263674,
      "memory(GiB)": 77.56,
      "step": 95905,
      "token_acc": 0.5361216730038023,
      "train_speed(iter/s)": 1.438304
    },
    {
      "epoch": 4.109078445653571,
      "grad_norm": 6.671321392059326,
      "learning_rate": 7.631469385902173e-06,
      "loss": 2.239193344116211,
      "memory(GiB)": 77.56,
      "step": 95910,
      "token_acc": 0.5125448028673835,
      "train_speed(iter/s)": 1.438311
    },
    {
      "epoch": 4.1092926609828195,
      "grad_norm": 5.9768829345703125,
      "learning_rate": 7.627896247236382e-06,
      "loss": 2.190685844421387,
      "memory(GiB)": 77.56,
      "step": 95915,
      "token_acc": 0.5109034267912772,
      "train_speed(iter/s)": 1.438305
    },
    {
      "epoch": 4.109506876312069,
      "grad_norm": 7.609327793121338,
      "learning_rate": 7.624323876180989e-06,
      "loss": 2.5587928771972654,
      "memory(GiB)": 77.56,
      "step": 95920,
      "token_acc": 0.43490304709141275,
      "train_speed(iter/s)": 1.438326
    },
    {
      "epoch": 4.109721091641318,
      "grad_norm": 5.801846027374268,
      "learning_rate": 7.6207522728007354e-06,
      "loss": 2.462677764892578,
      "memory(GiB)": 77.56,
      "step": 95925,
      "token_acc": 0.4843205574912892,
      "train_speed(iter/s)": 1.438323
    },
    {
      "epoch": 4.109935306970566,
      "grad_norm": 6.480906963348389,
      "learning_rate": 7.6171814371603035e-06,
      "loss": 2.3157093048095705,
      "memory(GiB)": 77.56,
      "step": 95930,
      "token_acc": 0.5304659498207885,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 4.110149522299816,
      "grad_norm": 5.38148307800293,
      "learning_rate": 7.613611369324392e-06,
      "loss": 2.2307273864746096,
      "memory(GiB)": 77.56,
      "step": 95935,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.438307
    },
    {
      "epoch": 4.110363737629065,
      "grad_norm": 6.86534309387207,
      "learning_rate": 7.61004206935767e-06,
      "loss": 2.126034736633301,
      "memory(GiB)": 77.56,
      "step": 95940,
      "token_acc": 0.5255972696245734,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 4.110577952958313,
      "grad_norm": 7.398293495178223,
      "learning_rate": 7.606473537324804e-06,
      "loss": 2.361636734008789,
      "memory(GiB)": 77.56,
      "step": 95945,
      "token_acc": 0.5202702702702703,
      "train_speed(iter/s)": 1.438334
    },
    {
      "epoch": 4.1107921682875626,
      "grad_norm": 5.546005725860596,
      "learning_rate": 7.602905773290425e-06,
      "loss": 2.291697120666504,
      "memory(GiB)": 77.56,
      "step": 95950,
      "token_acc": 0.5035211267605634,
      "train_speed(iter/s)": 1.438326
    },
    {
      "epoch": 4.111006383616812,
      "grad_norm": 9.449408531188965,
      "learning_rate": 7.599338777319192e-06,
      "loss": 2.116642951965332,
      "memory(GiB)": 77.56,
      "step": 95955,
      "token_acc": 0.5528455284552846,
      "train_speed(iter/s)": 1.43834
    },
    {
      "epoch": 4.11122059894606,
      "grad_norm": 7.142069339752197,
      "learning_rate": 7.59577254947571e-06,
      "loss": 2.2153013229370115,
      "memory(GiB)": 77.56,
      "step": 95960,
      "token_acc": 0.5390625,
      "train_speed(iter/s)": 1.43835
    },
    {
      "epoch": 4.111434814275309,
      "grad_norm": 7.215594291687012,
      "learning_rate": 7.592207089824588e-06,
      "loss": 1.9601696014404297,
      "memory(GiB)": 77.56,
      "step": 95965,
      "token_acc": 0.564,
      "train_speed(iter/s)": 1.438337
    },
    {
      "epoch": 4.111649029604559,
      "grad_norm": 6.804347515106201,
      "learning_rate": 7.5886423984304125e-06,
      "loss": 2.282740592956543,
      "memory(GiB)": 77.56,
      "step": 95970,
      "token_acc": 0.5031446540880503,
      "train_speed(iter/s)": 1.43834
    },
    {
      "epoch": 4.111863244933807,
      "grad_norm": 5.42451286315918,
      "learning_rate": 7.58507847535776e-06,
      "loss": 2.1462989807128907,
      "memory(GiB)": 77.56,
      "step": 95975,
      "token_acc": 0.5443037974683544,
      "train_speed(iter/s)": 1.438356
    },
    {
      "epoch": 4.112077460263056,
      "grad_norm": 4.953402042388916,
      "learning_rate": 7.5815153206711955e-06,
      "loss": 1.9254533767700195,
      "memory(GiB)": 77.56,
      "step": 95980,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.438358
    },
    {
      "epoch": 4.112291675592306,
      "grad_norm": 7.234443187713623,
      "learning_rate": 7.577952934435284e-06,
      "loss": 2.3278614044189454,
      "memory(GiB)": 77.56,
      "step": 95985,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 4.112505890921554,
      "grad_norm": 6.257533073425293,
      "learning_rate": 7.574391316714552e-06,
      "loss": 2.1003753662109377,
      "memory(GiB)": 77.56,
      "step": 95990,
      "token_acc": 0.5436507936507936,
      "train_speed(iter/s)": 1.438397
    },
    {
      "epoch": 4.112720106250803,
      "grad_norm": 7.7184953689575195,
      "learning_rate": 7.57083046757352e-06,
      "loss": 2.2016576766967773,
      "memory(GiB)": 77.56,
      "step": 95995,
      "token_acc": 0.5159010600706714,
      "train_speed(iter/s)": 1.438376
    },
    {
      "epoch": 4.1129343215800525,
      "grad_norm": 7.000500202178955,
      "learning_rate": 7.567270387076692e-06,
      "loss": 2.1520095825195313,
      "memory(GiB)": 77.56,
      "step": 96000,
      "token_acc": 0.5143884892086331,
      "train_speed(iter/s)": 1.438385
    },
    {
      "epoch": 4.1129343215800525,
      "eval_loss": 2.0750722885131836,
      "eval_runtime": 14.6451,
      "eval_samples_per_second": 6.828,
      "eval_steps_per_second": 6.828,
      "eval_token_acc": 0.47690217391304346,
      "step": 96000
    },
    {
      "epoch": 4.113148536909301,
      "grad_norm": 5.765586853027344,
      "learning_rate": 7.563711075288571e-06,
      "loss": 2.3592966079711912,
      "memory(GiB)": 77.56,
      "step": 96005,
      "token_acc": 0.4831358249772106,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 4.11336275223855,
      "grad_norm": 8.4855375289917,
      "learning_rate": 7.560152532273618e-06,
      "loss": 2.289281463623047,
      "memory(GiB)": 77.56,
      "step": 96010,
      "token_acc": 0.5078125,
      "train_speed(iter/s)": 1.438074
    },
    {
      "epoch": 4.113576967567799,
      "grad_norm": 12.849916458129883,
      "learning_rate": 7.556594758096325e-06,
      "loss": 2.4361392974853517,
      "memory(GiB)": 77.56,
      "step": 96015,
      "token_acc": 0.476056338028169,
      "train_speed(iter/s)": 1.438061
    },
    {
      "epoch": 4.113791182897048,
      "grad_norm": 6.677891254425049,
      "learning_rate": 7.553037752821135e-06,
      "loss": 2.2970102310180662,
      "memory(GiB)": 77.56,
      "step": 96020,
      "token_acc": 0.5126353790613718,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 4.114005398226297,
      "grad_norm": 7.1309990882873535,
      "learning_rate": 7.549481516512485e-06,
      "loss": 2.2121131896972654,
      "memory(GiB)": 77.56,
      "step": 96025,
      "token_acc": 0.5389830508474577,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 4.114219613555546,
      "grad_norm": 4.925137042999268,
      "learning_rate": 7.5459260492347975e-06,
      "loss": 2.6382211685180663,
      "memory(GiB)": 77.56,
      "step": 96030,
      "token_acc": 0.4819277108433735,
      "train_speed(iter/s)": 1.438098
    },
    {
      "epoch": 4.114433828884795,
      "grad_norm": 5.852963447570801,
      "learning_rate": 7.5423713510524745e-06,
      "loss": 2.304315185546875,
      "memory(GiB)": 77.56,
      "step": 96035,
      "token_acc": 0.5285714285714286,
      "train_speed(iter/s)": 1.438099
    },
    {
      "epoch": 4.114648044214044,
      "grad_norm": 7.86866569519043,
      "learning_rate": 7.538817422029931e-06,
      "loss": 2.355795478820801,
      "memory(GiB)": 77.56,
      "step": 96040,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 4.114862259543293,
      "grad_norm": 6.939375400543213,
      "learning_rate": 7.535264262231545e-06,
      "loss": 1.9418508529663085,
      "memory(GiB)": 77.56,
      "step": 96045,
      "token_acc": 0.5752508361204013,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 4.1150764748725415,
      "grad_norm": 5.743912696838379,
      "learning_rate": 7.531711871721669e-06,
      "loss": 2.222469711303711,
      "memory(GiB)": 77.56,
      "step": 96050,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.438093
    },
    {
      "epoch": 4.115290690201791,
      "grad_norm": 5.764291763305664,
      "learning_rate": 7.528160250564681e-06,
      "loss": 2.367168998718262,
      "memory(GiB)": 77.56,
      "step": 96055,
      "token_acc": 0.47003154574132494,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 4.11550490553104,
      "grad_norm": 8.050865173339844,
      "learning_rate": 7.524609398824916e-06,
      "loss": 2.517007827758789,
      "memory(GiB)": 77.56,
      "step": 96060,
      "token_acc": 0.4984423676012461,
      "train_speed(iter/s)": 1.438095
    },
    {
      "epoch": 4.115719120860288,
      "grad_norm": 6.087296485900879,
      "learning_rate": 7.521059316566681e-06,
      "loss": 1.9852994918823241,
      "memory(GiB)": 77.56,
      "step": 96065,
      "token_acc": 0.5376344086021505,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 4.115933336189538,
      "grad_norm": 7.634925842285156,
      "learning_rate": 7.517510003854322e-06,
      "loss": 2.205913543701172,
      "memory(GiB)": 77.56,
      "step": 96070,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.438092
    },
    {
      "epoch": 4.116147551518787,
      "grad_norm": 6.784355163574219,
      "learning_rate": 7.513961460752117e-06,
      "loss": 2.393558883666992,
      "memory(GiB)": 77.56,
      "step": 96075,
      "token_acc": 0.4984423676012461,
      "train_speed(iter/s)": 1.438111
    },
    {
      "epoch": 4.116361766848035,
      "grad_norm": 6.767854690551758,
      "learning_rate": 7.510413687324358e-06,
      "loss": 2.4951629638671875,
      "memory(GiB)": 77.56,
      "step": 96080,
      "token_acc": 0.4797297297297297,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 4.1165759821772845,
      "grad_norm": 6.6120758056640625,
      "learning_rate": 7.506866683635311e-06,
      "loss": 2.4073081970214845,
      "memory(GiB)": 77.56,
      "step": 96085,
      "token_acc": 0.4900662251655629,
      "train_speed(iter/s)": 1.438106
    },
    {
      "epoch": 4.116790197506534,
      "grad_norm": 5.49633264541626,
      "learning_rate": 7.503320449749235e-06,
      "loss": 2.291116142272949,
      "memory(GiB)": 77.56,
      "step": 96090,
      "token_acc": 0.4963768115942029,
      "train_speed(iter/s)": 1.43811
    },
    {
      "epoch": 4.117004412835782,
      "grad_norm": 7.807966232299805,
      "learning_rate": 7.499774985730363e-06,
      "loss": 2.0695318222045898,
      "memory(GiB)": 77.56,
      "step": 96095,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 4.117218628165031,
      "grad_norm": 7.62118673324585,
      "learning_rate": 7.496230291642947e-06,
      "loss": 2.378623199462891,
      "memory(GiB)": 77.56,
      "step": 96100,
      "token_acc": 0.50390625,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 4.117432843494281,
      "grad_norm": 7.91900634765625,
      "learning_rate": 7.492686367551194e-06,
      "loss": 2.160654067993164,
      "memory(GiB)": 77.56,
      "step": 96105,
      "token_acc": 0.5568181818181818,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 4.117647058823529,
      "grad_norm": 6.8633503913879395,
      "learning_rate": 7.489143213519301e-06,
      "loss": 2.240801239013672,
      "memory(GiB)": 77.56,
      "step": 96110,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 4.117861274152778,
      "grad_norm": 5.698827266693115,
      "learning_rate": 7.485600829611456e-06,
      "loss": 2.2253496170043947,
      "memory(GiB)": 77.56,
      "step": 96115,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438083
    },
    {
      "epoch": 4.1180754894820275,
      "grad_norm": 5.872425079345703,
      "learning_rate": 7.482059215891823e-06,
      "loss": 2.0663137435913086,
      "memory(GiB)": 77.56,
      "step": 96120,
      "token_acc": 0.5353982300884956,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 4.118289704811276,
      "grad_norm": 6.3538055419921875,
      "learning_rate": 7.478518372424576e-06,
      "loss": 2.2023794174194338,
      "memory(GiB)": 77.56,
      "step": 96125,
      "token_acc": 0.5083612040133779,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 4.118503920140525,
      "grad_norm": 6.036527633666992,
      "learning_rate": 7.474978299273861e-06,
      "loss": 2.133736991882324,
      "memory(GiB)": 77.56,
      "step": 96130,
      "token_acc": 0.5282392026578073,
      "train_speed(iter/s)": 1.438077
    },
    {
      "epoch": 4.118718135469774,
      "grad_norm": 5.533919334411621,
      "learning_rate": 7.47143899650381e-06,
      "loss": 2.243769073486328,
      "memory(GiB)": 77.56,
      "step": 96135,
      "token_acc": 0.5213414634146342,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 4.118932350799023,
      "grad_norm": 8.145271301269531,
      "learning_rate": 7.467900464178534e-06,
      "loss": 2.3492404937744142,
      "memory(GiB)": 77.56,
      "step": 96140,
      "token_acc": 0.4934640522875817,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 4.119146566128272,
      "grad_norm": 5.510066032409668,
      "learning_rate": 7.4643627023621435e-06,
      "loss": 1.8543930053710938,
      "memory(GiB)": 77.56,
      "step": 96145,
      "token_acc": 0.5438066465256798,
      "train_speed(iter/s)": 1.438053
    },
    {
      "epoch": 4.119360781457521,
      "grad_norm": 8.078634262084961,
      "learning_rate": 7.460825711118724e-06,
      "loss": 2.5635005950927736,
      "memory(GiB)": 77.56,
      "step": 96150,
      "token_acc": 0.49477351916376305,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 4.11957499678677,
      "grad_norm": 6.9337239265441895,
      "learning_rate": 7.457289490512337e-06,
      "loss": 2.4758447647094726,
      "memory(GiB)": 77.56,
      "step": 96155,
      "token_acc": 0.46105919003115264,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 4.119789212116019,
      "grad_norm": 6.736894607543945,
      "learning_rate": 7.453754040607075e-06,
      "loss": 2.6031646728515625,
      "memory(GiB)": 77.56,
      "step": 96160,
      "token_acc": 0.48134328358208955,
      "train_speed(iter/s)": 1.438057
    },
    {
      "epoch": 4.120003427445268,
      "grad_norm": 6.146721363067627,
      "learning_rate": 7.450219361466965e-06,
      "loss": 2.0872114181518553,
      "memory(GiB)": 77.56,
      "step": 96165,
      "token_acc": 0.5278810408921933,
      "train_speed(iter/s)": 1.43806
    },
    {
      "epoch": 4.1202176427745165,
      "grad_norm": 6.551029682159424,
      "learning_rate": 7.44668545315605e-06,
      "loss": 2.1993005752563475,
      "memory(GiB)": 77.56,
      "step": 96170,
      "token_acc": 0.5533980582524272,
      "train_speed(iter/s)": 1.438061
    },
    {
      "epoch": 4.120431858103766,
      "grad_norm": 6.693967819213867,
      "learning_rate": 7.443152315738344e-06,
      "loss": 2.2617431640625,
      "memory(GiB)": 77.56,
      "step": 96175,
      "token_acc": 0.48760330578512395,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 4.120646073433015,
      "grad_norm": 5.57351016998291,
      "learning_rate": 7.439619949277848e-06,
      "loss": 2.294831466674805,
      "memory(GiB)": 77.56,
      "step": 96180,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43807
    },
    {
      "epoch": 4.120860288762263,
      "grad_norm": 11.081794738769531,
      "learning_rate": 7.436088353838566e-06,
      "loss": 2.3934621810913086,
      "memory(GiB)": 77.56,
      "step": 96185,
      "token_acc": 0.509090909090909,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 4.121074504091513,
      "grad_norm": 7.196897983551025,
      "learning_rate": 7.432557529484479e-06,
      "loss": 2.1505853652954103,
      "memory(GiB)": 77.56,
      "step": 96190,
      "token_acc": 0.5197568389057751,
      "train_speed(iter/s)": 1.438078
    },
    {
      "epoch": 4.121288719420762,
      "grad_norm": 9.193681716918945,
      "learning_rate": 7.4290274762795265e-06,
      "loss": 2.4362371444702147,
      "memory(GiB)": 77.56,
      "step": 96195,
      "token_acc": 0.4765625,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 4.12150293475001,
      "grad_norm": 4.986240863800049,
      "learning_rate": 7.425498194287689e-06,
      "loss": 2.081932258605957,
      "memory(GiB)": 77.56,
      "step": 96200,
      "token_acc": 0.5168195718654435,
      "train_speed(iter/s)": 1.438096
    },
    {
      "epoch": 4.1217171500792595,
      "grad_norm": 5.693068504333496,
      "learning_rate": 7.421969683572894e-06,
      "loss": 2.636934280395508,
      "memory(GiB)": 77.56,
      "step": 96205,
      "token_acc": 0.47079037800687284,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 4.121931365408509,
      "grad_norm": 5.989151954650879,
      "learning_rate": 7.418441944199045e-06,
      "loss": 2.4715244293212892,
      "memory(GiB)": 77.56,
      "step": 96210,
      "token_acc": 0.49591280653950953,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 4.122145580737757,
      "grad_norm": 5.698012828826904,
      "learning_rate": 7.414914976230075e-06,
      "loss": 2.150025177001953,
      "memory(GiB)": 77.56,
      "step": 96215,
      "token_acc": 0.5316455696202531,
      "train_speed(iter/s)": 1.438116
    },
    {
      "epoch": 4.122359796067006,
      "grad_norm": 6.786555290222168,
      "learning_rate": 7.411388779729872e-06,
      "loss": 2.3515079498291014,
      "memory(GiB)": 77.56,
      "step": 96220,
      "token_acc": 0.5035460992907801,
      "train_speed(iter/s)": 1.438104
    },
    {
      "epoch": 4.122574011396256,
      "grad_norm": 5.212714672088623,
      "learning_rate": 7.4078633547623065e-06,
      "loss": 2.3681928634643556,
      "memory(GiB)": 77.56,
      "step": 96225,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.438099
    },
    {
      "epoch": 4.122788226725504,
      "grad_norm": 6.217240810394287,
      "learning_rate": 7.404338701391256e-06,
      "loss": 2.2296422958374023,
      "memory(GiB)": 77.56,
      "step": 96230,
      "token_acc": 0.5311355311355311,
      "train_speed(iter/s)": 1.438124
    },
    {
      "epoch": 4.123002442054753,
      "grad_norm": 6.141689777374268,
      "learning_rate": 7.400814819680568e-06,
      "loss": 2.519417572021484,
      "memory(GiB)": 77.56,
      "step": 96235,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.438136
    },
    {
      "epoch": 4.123216657384003,
      "grad_norm": 6.020229339599609,
      "learning_rate": 7.397291709694065e-06,
      "loss": 2.364508628845215,
      "memory(GiB)": 77.56,
      "step": 96240,
      "token_acc": 0.4914772727272727,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 4.123430872713251,
      "grad_norm": 5.66908597946167,
      "learning_rate": 7.393769371495602e-06,
      "loss": 2.565126419067383,
      "memory(GiB)": 77.56,
      "step": 96245,
      "token_acc": 0.47107438016528924,
      "train_speed(iter/s)": 1.438136
    },
    {
      "epoch": 4.1236450880425,
      "grad_norm": 5.747981548309326,
      "learning_rate": 7.390247805148976e-06,
      "loss": 1.9654863357543946,
      "memory(GiB)": 77.56,
      "step": 96250,
      "token_acc": 0.5907172995780591,
      "train_speed(iter/s)": 1.438133
    },
    {
      "epoch": 4.123859303371749,
      "grad_norm": 5.34492301940918,
      "learning_rate": 7.386727010717981e-06,
      "loss": 2.502433013916016,
      "memory(GiB)": 77.56,
      "step": 96255,
      "token_acc": 0.5070921985815603,
      "train_speed(iter/s)": 1.43814
    },
    {
      "epoch": 4.124073518700998,
      "grad_norm": 5.8116607666015625,
      "learning_rate": 7.3832069882664e-06,
      "loss": 2.4343244552612306,
      "memory(GiB)": 77.56,
      "step": 96260,
      "token_acc": 0.49206349206349204,
      "train_speed(iter/s)": 1.43814
    },
    {
      "epoch": 4.124287734030247,
      "grad_norm": 7.909563064575195,
      "learning_rate": 7.379687737857987e-06,
      "loss": 2.525436210632324,
      "memory(GiB)": 77.56,
      "step": 96265,
      "token_acc": 0.5434782608695652,
      "train_speed(iter/s)": 1.438134
    },
    {
      "epoch": 4.124501949359496,
      "grad_norm": 6.475271701812744,
      "learning_rate": 7.37616925955652e-06,
      "loss": 2.166606903076172,
      "memory(GiB)": 77.56,
      "step": 96270,
      "token_acc": 0.5636363636363636,
      "train_speed(iter/s)": 1.438124
    },
    {
      "epoch": 4.124716164688745,
      "grad_norm": 4.453949928283691,
      "learning_rate": 7.372651553425735e-06,
      "loss": 2.191685676574707,
      "memory(GiB)": 77.56,
      "step": 96275,
      "token_acc": 0.5324675324675324,
      "train_speed(iter/s)": 1.438125
    },
    {
      "epoch": 4.124930380017994,
      "grad_norm": 6.0995306968688965,
      "learning_rate": 7.369134619529361e-06,
      "loss": 2.020674133300781,
      "memory(GiB)": 77.56,
      "step": 96280,
      "token_acc": 0.5579710144927537,
      "train_speed(iter/s)": 1.438135
    },
    {
      "epoch": 4.125144595347243,
      "grad_norm": 9.233741760253906,
      "learning_rate": 7.365618457931101e-06,
      "loss": 2.518416976928711,
      "memory(GiB)": 77.56,
      "step": 96285,
      "token_acc": 0.5167785234899329,
      "train_speed(iter/s)": 1.438151
    },
    {
      "epoch": 4.125358810676492,
      "grad_norm": 11.09237003326416,
      "learning_rate": 7.362103068694653e-06,
      "loss": 2.557550048828125,
      "memory(GiB)": 77.56,
      "step": 96290,
      "token_acc": 0.5231788079470199,
      "train_speed(iter/s)": 1.438159
    },
    {
      "epoch": 4.125573026005741,
      "grad_norm": 5.068997383117676,
      "learning_rate": 7.358588451883714e-06,
      "loss": 2.4432735443115234,
      "memory(GiB)": 77.56,
      "step": 96295,
      "token_acc": 0.4811320754716981,
      "train_speed(iter/s)": 1.438152
    },
    {
      "epoch": 4.12578724133499,
      "grad_norm": 6.230804920196533,
      "learning_rate": 7.355074607561929e-06,
      "loss": 2.265390396118164,
      "memory(GiB)": 77.56,
      "step": 96300,
      "token_acc": 0.5248447204968945,
      "train_speed(iter/s)": 1.438161
    },
    {
      "epoch": 4.126001456664239,
      "grad_norm": 11.891359329223633,
      "learning_rate": 7.351561535792984e-06,
      "loss": 2.3951742172241213,
      "memory(GiB)": 77.56,
      "step": 96305,
      "token_acc": 0.5065359477124183,
      "train_speed(iter/s)": 1.438163
    },
    {
      "epoch": 4.126215671993488,
      "grad_norm": 6.688968658447266,
      "learning_rate": 7.348049236640509e-06,
      "loss": 2.3005237579345703,
      "memory(GiB)": 77.56,
      "step": 96310,
      "token_acc": 0.4589041095890411,
      "train_speed(iter/s)": 1.438184
    },
    {
      "epoch": 4.126429887322737,
      "grad_norm": 7.9630279541015625,
      "learning_rate": 7.344537710168136e-06,
      "loss": 2.3335906982421877,
      "memory(GiB)": 77.56,
      "step": 96315,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.438193
    },
    {
      "epoch": 4.126644102651985,
      "grad_norm": 6.251548767089844,
      "learning_rate": 7.3410269564394725e-06,
      "loss": 2.216363525390625,
      "memory(GiB)": 77.56,
      "step": 96320,
      "token_acc": 0.538135593220339,
      "train_speed(iter/s)": 1.438198
    },
    {
      "epoch": 4.126858317981235,
      "grad_norm": 6.065381050109863,
      "learning_rate": 7.337516975518116e-06,
      "loss": 2.044463348388672,
      "memory(GiB)": 77.56,
      "step": 96325,
      "token_acc": 0.5126353790613718,
      "train_speed(iter/s)": 1.438206
    },
    {
      "epoch": 4.127072533310484,
      "grad_norm": 6.984500408172607,
      "learning_rate": 7.334007767467666e-06,
      "loss": 2.2201511383056642,
      "memory(GiB)": 77.56,
      "step": 96330,
      "token_acc": 0.49085365853658536,
      "train_speed(iter/s)": 1.438199
    },
    {
      "epoch": 4.127286748639733,
      "grad_norm": 7.817566871643066,
      "learning_rate": 7.330499332351692e-06,
      "loss": 2.153267669677734,
      "memory(GiB)": 77.56,
      "step": 96335,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.438213
    },
    {
      "epoch": 4.1275009639689815,
      "grad_norm": 7.226218223571777,
      "learning_rate": 7.326991670233751e-06,
      "loss": 2.1477020263671873,
      "memory(GiB)": 77.56,
      "step": 96340,
      "token_acc": 0.5301587301587302,
      "train_speed(iter/s)": 1.438211
    },
    {
      "epoch": 4.127715179298231,
      "grad_norm": 7.403777599334717,
      "learning_rate": 7.3234847811773755e-06,
      "loss": 2.1869073867797852,
      "memory(GiB)": 77.56,
      "step": 96345,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.438212
    },
    {
      "epoch": 4.127929394627479,
      "grad_norm": 8.69003963470459,
      "learning_rate": 7.319978665246113e-06,
      "loss": 2.5224687576293947,
      "memory(GiB)": 77.56,
      "step": 96350,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.438213
    },
    {
      "epoch": 4.128143609956728,
      "grad_norm": 5.238253116607666,
      "learning_rate": 7.31647332250347e-06,
      "loss": 2.2845026016235352,
      "memory(GiB)": 77.56,
      "step": 96355,
      "token_acc": 0.5364963503649635,
      "train_speed(iter/s)": 1.438218
    },
    {
      "epoch": 4.128357825285978,
      "grad_norm": 5.36712121963501,
      "learning_rate": 7.312968753012961e-06,
      "loss": 2.2127079010009765,
      "memory(GiB)": 77.56,
      "step": 96360,
      "token_acc": 0.5109034267912772,
      "train_speed(iter/s)": 1.43822
    },
    {
      "epoch": 4.128572040615227,
      "grad_norm": 5.671230792999268,
      "learning_rate": 7.309464956838063e-06,
      "loss": 2.3613243103027344,
      "memory(GiB)": 77.56,
      "step": 96365,
      "token_acc": 0.5079872204472844,
      "train_speed(iter/s)": 1.438233
    },
    {
      "epoch": 4.128786255944475,
      "grad_norm": 8.403008460998535,
      "learning_rate": 7.30596193404226e-06,
      "loss": 2.4431825637817384,
      "memory(GiB)": 77.56,
      "step": 96370,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438244
    },
    {
      "epoch": 4.1290004712737245,
      "grad_norm": 8.125631332397461,
      "learning_rate": 7.302459684689006e-06,
      "loss": 1.9763547897338867,
      "memory(GiB)": 77.56,
      "step": 96375,
      "token_acc": 0.5163636363636364,
      "train_speed(iter/s)": 1.43825
    },
    {
      "epoch": 4.129214686602973,
      "grad_norm": 8.112006187438965,
      "learning_rate": 7.298958208841744e-06,
      "loss": 2.1689355850219725,
      "memory(GiB)": 77.56,
      "step": 96380,
      "token_acc": 0.5598290598290598,
      "train_speed(iter/s)": 1.438259
    },
    {
      "epoch": 4.129428901932222,
      "grad_norm": 5.576782703399658,
      "learning_rate": 7.295457506563902e-06,
      "loss": 2.2749116897583006,
      "memory(GiB)": 77.56,
      "step": 96385,
      "token_acc": 0.5059288537549407,
      "train_speed(iter/s)": 1.438269
    },
    {
      "epoch": 4.129643117261471,
      "grad_norm": 7.517724514007568,
      "learning_rate": 7.291957577918923e-06,
      "loss": 2.2987342834472657,
      "memory(GiB)": 77.56,
      "step": 96390,
      "token_acc": 0.5234899328859061,
      "train_speed(iter/s)": 1.438272
    },
    {
      "epoch": 4.129857332590721,
      "grad_norm": 6.646448612213135,
      "learning_rate": 7.288458422970191e-06,
      "loss": 2.3035144805908203,
      "memory(GiB)": 77.56,
      "step": 96395,
      "token_acc": 0.5317725752508361,
      "train_speed(iter/s)": 1.438281
    },
    {
      "epoch": 4.130071547919969,
      "grad_norm": 5.043950080871582,
      "learning_rate": 7.284960041781102e-06,
      "loss": 2.1844324111938476,
      "memory(GiB)": 77.56,
      "step": 96400,
      "token_acc": 0.5355805243445693,
      "train_speed(iter/s)": 1.438276
    },
    {
      "epoch": 4.130285763249218,
      "grad_norm": 6.114025592803955,
      "learning_rate": 7.281462434415032e-06,
      "loss": 2.2071548461914063,
      "memory(GiB)": 77.56,
      "step": 96405,
      "token_acc": 0.5421686746987951,
      "train_speed(iter/s)": 1.438286
    },
    {
      "epoch": 4.130499978578467,
      "grad_norm": 5.625747203826904,
      "learning_rate": 7.277965600935333e-06,
      "loss": 2.4538200378417967,
      "memory(GiB)": 77.56,
      "step": 96410,
      "token_acc": 0.4783950617283951,
      "train_speed(iter/s)": 1.438285
    },
    {
      "epoch": 4.130714193907716,
      "grad_norm": 7.017910480499268,
      "learning_rate": 7.274469541405376e-06,
      "loss": 2.3580970764160156,
      "memory(GiB)": 77.56,
      "step": 96415,
      "token_acc": 0.5401929260450161,
      "train_speed(iter/s)": 1.438276
    },
    {
      "epoch": 4.130928409236965,
      "grad_norm": 5.083176612854004,
      "learning_rate": 7.270974255888469e-06,
      "loss": 2.1233327865600584,
      "memory(GiB)": 77.56,
      "step": 96420,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.438286
    },
    {
      "epoch": 4.131142624566214,
      "grad_norm": 7.452576637268066,
      "learning_rate": 7.267479744447958e-06,
      "loss": 2.0950721740722655,
      "memory(GiB)": 77.56,
      "step": 96425,
      "token_acc": 0.5398773006134969,
      "train_speed(iter/s)": 1.438276
    },
    {
      "epoch": 4.131356839895463,
      "grad_norm": 6.489446640014648,
      "learning_rate": 7.263986007147139e-06,
      "loss": 2.294108200073242,
      "memory(GiB)": 77.56,
      "step": 96430,
      "token_acc": 0.5363984674329502,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 4.131571055224712,
      "grad_norm": 5.427678108215332,
      "learning_rate": 7.260493044049299e-06,
      "loss": 2.200992774963379,
      "memory(GiB)": 77.56,
      "step": 96435,
      "token_acc": 0.5412844036697247,
      "train_speed(iter/s)": 1.438287
    },
    {
      "epoch": 4.13178527055396,
      "grad_norm": 6.177824974060059,
      "learning_rate": 7.257000855217727e-06,
      "loss": 2.490162467956543,
      "memory(GiB)": 77.56,
      "step": 96440,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.438294
    },
    {
      "epoch": 4.13199948588321,
      "grad_norm": 9.818201065063477,
      "learning_rate": 7.253509440715667e-06,
      "loss": 2.3024213790893553,
      "memory(GiB)": 77.56,
      "step": 96445,
      "token_acc": 0.4852941176470588,
      "train_speed(iter/s)": 1.438285
    },
    {
      "epoch": 4.132213701212459,
      "grad_norm": 6.740192413330078,
      "learning_rate": 7.25001880060639e-06,
      "loss": 2.364698028564453,
      "memory(GiB)": 77.56,
      "step": 96450,
      "token_acc": 0.4728434504792332,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 4.132427916541708,
      "grad_norm": 6.426738739013672,
      "learning_rate": 7.246528934953128e-06,
      "loss": 2.471163749694824,
      "memory(GiB)": 77.56,
      "step": 96455,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.438273
    },
    {
      "epoch": 4.1326421318709565,
      "grad_norm": 6.708327293395996,
      "learning_rate": 7.243039843819105e-06,
      "loss": 2.3847736358642577,
      "memory(GiB)": 77.56,
      "step": 96460,
      "token_acc": 0.503968253968254,
      "train_speed(iter/s)": 1.438269
    },
    {
      "epoch": 4.132856347200206,
      "grad_norm": 7.480318546295166,
      "learning_rate": 7.2395515272675186e-06,
      "loss": 2.181375503540039,
      "memory(GiB)": 77.56,
      "step": 96465,
      "token_acc": 0.547244094488189,
      "train_speed(iter/s)": 1.438272
    },
    {
      "epoch": 4.133070562529454,
      "grad_norm": 6.168023586273193,
      "learning_rate": 7.236063985361563e-06,
      "loss": 2.1938461303710937,
      "memory(GiB)": 77.56,
      "step": 96470,
      "token_acc": 0.5112540192926045,
      "train_speed(iter/s)": 1.438283
    },
    {
      "epoch": 4.133284777858703,
      "grad_norm": 4.6516313552856445,
      "learning_rate": 7.232577218164427e-06,
      "loss": 2.408511734008789,
      "memory(GiB)": 77.56,
      "step": 96475,
      "token_acc": 0.5015974440894568,
      "train_speed(iter/s)": 1.438278
    },
    {
      "epoch": 4.133498993187953,
      "grad_norm": 5.809952259063721,
      "learning_rate": 7.229091225739282e-06,
      "loss": 2.2135032653808593,
      "memory(GiB)": 77.56,
      "step": 96480,
      "token_acc": 0.5063291139240507,
      "train_speed(iter/s)": 1.438293
    },
    {
      "epoch": 4.133713208517202,
      "grad_norm": 6.514778137207031,
      "learning_rate": 7.225606008149266e-06,
      "loss": 2.145332908630371,
      "memory(GiB)": 77.56,
      "step": 96485,
      "token_acc": 0.528169014084507,
      "train_speed(iter/s)": 1.438294
    },
    {
      "epoch": 4.13392742384645,
      "grad_norm": 5.6674065589904785,
      "learning_rate": 7.222121565457529e-06,
      "loss": 2.0490707397460937,
      "memory(GiB)": 77.56,
      "step": 96490,
      "token_acc": 0.5305343511450382,
      "train_speed(iter/s)": 1.438278
    },
    {
      "epoch": 4.1341416391756995,
      "grad_norm": 4.809019565582275,
      "learning_rate": 7.2186378977271705e-06,
      "loss": 2.199611854553223,
      "memory(GiB)": 77.56,
      "step": 96495,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.438264
    },
    {
      "epoch": 4.134355854504949,
      "grad_norm": 5.550069808959961,
      "learning_rate": 7.215155005021323e-06,
      "loss": 2.290171432495117,
      "memory(GiB)": 77.56,
      "step": 96500,
      "token_acc": 0.5142857142857142,
      "train_speed(iter/s)": 1.438257
    },
    {
      "epoch": 4.134355854504949,
      "eval_loss": 2.1640238761901855,
      "eval_runtime": 14.0832,
      "eval_samples_per_second": 7.101,
      "eval_steps_per_second": 7.101,
      "eval_token_acc": 0.4731182795698925,
      "step": 96500
    },
    {
      "epoch": 4.134570069834197,
      "grad_norm": 6.027403354644775,
      "learning_rate": 7.211672887403087e-06,
      "loss": 2.3687294006347654,
      "memory(GiB)": 77.56,
      "step": 96505,
      "token_acc": 0.4844167408726625,
      "train_speed(iter/s)": 1.437937
    },
    {
      "epoch": 4.134784285163446,
      "grad_norm": 5.739585876464844,
      "learning_rate": 7.208191544935538e-06,
      "loss": 2.2002351760864256,
      "memory(GiB)": 77.56,
      "step": 96510,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.437952
    },
    {
      "epoch": 4.134998500492696,
      "grad_norm": 4.584177494049072,
      "learning_rate": 7.204710977681739e-06,
      "loss": 2.3370132446289062,
      "memory(GiB)": 77.56,
      "step": 96515,
      "token_acc": 0.546583850931677,
      "train_speed(iter/s)": 1.437925
    },
    {
      "epoch": 4.135212715821944,
      "grad_norm": 6.653926372528076,
      "learning_rate": 7.201231185704749e-06,
      "loss": 2.158795928955078,
      "memory(GiB)": 77.56,
      "step": 96520,
      "token_acc": 0.5636363636363636,
      "train_speed(iter/s)": 1.437918
    },
    {
      "epoch": 4.135426931151193,
      "grad_norm": 8.364823341369629,
      "learning_rate": 7.197752169067601e-06,
      "loss": 2.457591438293457,
      "memory(GiB)": 77.56,
      "step": 96525,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.437933
    },
    {
      "epoch": 4.135641146480443,
      "grad_norm": 5.409210681915283,
      "learning_rate": 7.194273927833312e-06,
      "loss": 2.1754566192626954,
      "memory(GiB)": 77.56,
      "step": 96530,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.437929
    },
    {
      "epoch": 4.135855361809691,
      "grad_norm": 5.88220739364624,
      "learning_rate": 7.190796462064919e-06,
      "loss": 2.0327999114990236,
      "memory(GiB)": 77.56,
      "step": 96535,
      "token_acc": 0.5018315018315018,
      "train_speed(iter/s)": 1.437908
    },
    {
      "epoch": 4.13606957713894,
      "grad_norm": 7.667297840118408,
      "learning_rate": 7.187319771825407e-06,
      "loss": 2.2790462493896486,
      "memory(GiB)": 77.56,
      "step": 96540,
      "token_acc": 0.5115384615384615,
      "train_speed(iter/s)": 1.437877
    },
    {
      "epoch": 4.1362837924681894,
      "grad_norm": 6.999119758605957,
      "learning_rate": 7.183843857177757e-06,
      "loss": 2.698093605041504,
      "memory(GiB)": 77.56,
      "step": 96545,
      "token_acc": 0.48424068767908307,
      "train_speed(iter/s)": 1.437885
    },
    {
      "epoch": 4.136498007797438,
      "grad_norm": 5.225920677185059,
      "learning_rate": 7.180368718184943e-06,
      "loss": 2.40865478515625,
      "memory(GiB)": 77.56,
      "step": 96550,
      "token_acc": 0.5017921146953405,
      "train_speed(iter/s)": 1.437899
    },
    {
      "epoch": 4.136712223126687,
      "grad_norm": 5.234836578369141,
      "learning_rate": 7.176894354909908e-06,
      "loss": 2.2354469299316406,
      "memory(GiB)": 77.56,
      "step": 96555,
      "token_acc": 0.5206611570247934,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 4.136926438455936,
      "grad_norm": 7.849437236785889,
      "learning_rate": 7.173420767415611e-06,
      "loss": 2.4241825103759767,
      "memory(GiB)": 77.56,
      "step": 96560,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.437909
    },
    {
      "epoch": 4.137140653785185,
      "grad_norm": 6.073554039001465,
      "learning_rate": 7.169947955764961e-06,
      "loss": 2.149165153503418,
      "memory(GiB)": 77.56,
      "step": 96565,
      "token_acc": 0.5327102803738317,
      "train_speed(iter/s)": 1.4379
    },
    {
      "epoch": 4.137354869114434,
      "grad_norm": 5.625076770782471,
      "learning_rate": 7.166475920020888e-06,
      "loss": 2.331562805175781,
      "memory(GiB)": 77.56,
      "step": 96570,
      "token_acc": 0.5208333333333334,
      "train_speed(iter/s)": 1.437911
    },
    {
      "epoch": 4.137569084443683,
      "grad_norm": 8.233735084533691,
      "learning_rate": 7.163004660246292e-06,
      "loss": 2.16846923828125,
      "memory(GiB)": 77.56,
      "step": 96575,
      "token_acc": 0.4883720930232558,
      "train_speed(iter/s)": 1.437924
    },
    {
      "epoch": 4.137783299772932,
      "grad_norm": 5.690011501312256,
      "learning_rate": 7.159534176504046e-06,
      "loss": 2.2551654815673827,
      "memory(GiB)": 77.56,
      "step": 96580,
      "token_acc": 0.547945205479452,
      "train_speed(iter/s)": 1.437924
    },
    {
      "epoch": 4.137997515102181,
      "grad_norm": 6.119157314300537,
      "learning_rate": 7.1560644688570324e-06,
      "loss": 2.2950843811035155,
      "memory(GiB)": 77.56,
      "step": 96585,
      "token_acc": 0.49848024316109424,
      "train_speed(iter/s)": 1.437946
    },
    {
      "epoch": 4.13821173043143,
      "grad_norm": 5.932070732116699,
      "learning_rate": 7.152595537368084e-06,
      "loss": 2.178785705566406,
      "memory(GiB)": 77.56,
      "step": 96590,
      "token_acc": 0.5406360424028268,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 4.1384259457606785,
      "grad_norm": 5.11445426940918,
      "learning_rate": 7.149127382100074e-06,
      "loss": 2.651362991333008,
      "memory(GiB)": 77.56,
      "step": 96595,
      "token_acc": 0.46397694524495675,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 4.138640161089928,
      "grad_norm": 8.318365097045898,
      "learning_rate": 7.145660003115822e-06,
      "loss": 1.9505754470825196,
      "memory(GiB)": 77.56,
      "step": 96600,
      "token_acc": 0.572992700729927,
      "train_speed(iter/s)": 1.437949
    },
    {
      "epoch": 4.138854376419177,
      "grad_norm": 9.093037605285645,
      "learning_rate": 7.142193400478136e-06,
      "loss": 2.238535499572754,
      "memory(GiB)": 77.56,
      "step": 96605,
      "token_acc": 0.5320754716981132,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 4.139068591748425,
      "grad_norm": 5.706610202789307,
      "learning_rate": 7.138727574249821e-06,
      "loss": 2.4575634002685547,
      "memory(GiB)": 77.56,
      "step": 96610,
      "token_acc": 0.4759036144578313,
      "train_speed(iter/s)": 1.437954
    },
    {
      "epoch": 4.139282807077675,
      "grad_norm": 5.96674919128418,
      "learning_rate": 7.135262524493652e-06,
      "loss": 2.4098133087158202,
      "memory(GiB)": 77.56,
      "step": 96615,
      "token_acc": 0.512987012987013,
      "train_speed(iter/s)": 1.437956
    },
    {
      "epoch": 4.139497022406924,
      "grad_norm": 5.251148223876953,
      "learning_rate": 7.131798251272426e-06,
      "loss": 2.6687137603759767,
      "memory(GiB)": 77.56,
      "step": 96620,
      "token_acc": 0.4594594594594595,
      "train_speed(iter/s)": 1.437946
    },
    {
      "epoch": 4.139711237736172,
      "grad_norm": 6.330445289611816,
      "learning_rate": 7.128334754648891e-06,
      "loss": 1.9899627685546875,
      "memory(GiB)": 77.56,
      "step": 96625,
      "token_acc": 0.5378486055776892,
      "train_speed(iter/s)": 1.437941
    },
    {
      "epoch": 4.1399254530654215,
      "grad_norm": 6.628699779510498,
      "learning_rate": 7.124872034685781e-06,
      "loss": 2.2640108108520507,
      "memory(GiB)": 77.56,
      "step": 96630,
      "token_acc": 0.5290519877675841,
      "train_speed(iter/s)": 1.437944
    },
    {
      "epoch": 4.140139668394671,
      "grad_norm": 6.793025016784668,
      "learning_rate": 7.1214100914458404e-06,
      "loss": 2.4432655334472657,
      "memory(GiB)": 77.56,
      "step": 96635,
      "token_acc": 0.5107033639143731,
      "train_speed(iter/s)": 1.437958
    },
    {
      "epoch": 4.140353883723919,
      "grad_norm": 7.4457855224609375,
      "learning_rate": 7.117948924991769e-06,
      "loss": 2.290312385559082,
      "memory(GiB)": 77.56,
      "step": 96640,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.437951
    },
    {
      "epoch": 4.140568099053168,
      "grad_norm": 6.8137359619140625,
      "learning_rate": 7.1144885353862714e-06,
      "loss": 2.3147441864013674,
      "memory(GiB)": 77.56,
      "step": 96645,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.437969
    },
    {
      "epoch": 4.140782314382418,
      "grad_norm": 7.0620245933532715,
      "learning_rate": 7.111028922692065e-06,
      "loss": 2.5058111190795898,
      "memory(GiB)": 77.56,
      "step": 96650,
      "token_acc": 0.4966442953020134,
      "train_speed(iter/s)": 1.437976
    },
    {
      "epoch": 4.140996529711666,
      "grad_norm": 5.347829818725586,
      "learning_rate": 7.107570086971793e-06,
      "loss": 2.2791021347045897,
      "memory(GiB)": 77.56,
      "step": 96655,
      "token_acc": 0.511326860841424,
      "train_speed(iter/s)": 1.437984
    },
    {
      "epoch": 4.141210745040915,
      "grad_norm": 6.416926860809326,
      "learning_rate": 7.104112028288135e-06,
      "loss": 2.041130256652832,
      "memory(GiB)": 77.56,
      "step": 96660,
      "token_acc": 0.5423728813559322,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 4.1414249603701645,
      "grad_norm": 6.40525484085083,
      "learning_rate": 7.100654746703722e-06,
      "loss": 2.4706687927246094,
      "memory(GiB)": 77.56,
      "step": 96665,
      "token_acc": 0.5189873417721519,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 4.141639175699413,
      "grad_norm": 5.112632751464844,
      "learning_rate": 7.097198242281189e-06,
      "loss": 2.180733299255371,
      "memory(GiB)": 77.56,
      "step": 96670,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 4.141853391028662,
      "grad_norm": 5.637341499328613,
      "learning_rate": 7.093742515083146e-06,
      "loss": 2.2847274780273437,
      "memory(GiB)": 77.56,
      "step": 96675,
      "token_acc": 0.47202797202797203,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 4.142067606357911,
      "grad_norm": 6.188724994659424,
      "learning_rate": 7.0902875651722215e-06,
      "loss": 2.177390480041504,
      "memory(GiB)": 77.56,
      "step": 96680,
      "token_acc": 0.5631399317406144,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 4.14228182168716,
      "grad_norm": 5.32813024520874,
      "learning_rate": 7.0868333926109865e-06,
      "loss": 2.1966196060180665,
      "memory(GiB)": 77.56,
      "step": 96685,
      "token_acc": 0.5379310344827586,
      "train_speed(iter/s)": 1.438015
    },
    {
      "epoch": 4.142496037016409,
      "grad_norm": 5.9052934646606445,
      "learning_rate": 7.08337999746202e-06,
      "loss": 2.3645824432373046,
      "memory(GiB)": 77.56,
      "step": 96690,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.438029
    },
    {
      "epoch": 4.142710252345658,
      "grad_norm": 5.724667072296143,
      "learning_rate": 7.079927379787887e-06,
      "loss": 2.331894111633301,
      "memory(GiB)": 77.56,
      "step": 96695,
      "token_acc": 0.4756756756756757,
      "train_speed(iter/s)": 1.43803
    },
    {
      "epoch": 4.142924467674907,
      "grad_norm": 5.6233134269714355,
      "learning_rate": 7.076475539651117e-06,
      "loss": 2.080643081665039,
      "memory(GiB)": 77.56,
      "step": 96700,
      "token_acc": 0.5288753799392097,
      "train_speed(iter/s)": 1.43805
    },
    {
      "epoch": 4.143138683004156,
      "grad_norm": 5.347681999206543,
      "learning_rate": 7.073024477114276e-06,
      "loss": 1.9573766708374023,
      "memory(GiB)": 77.56,
      "step": 96705,
      "token_acc": 0.5647058823529412,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 4.143352898333405,
      "grad_norm": 5.899239540100098,
      "learning_rate": 7.069574192239858e-06,
      "loss": 2.225295639038086,
      "memory(GiB)": 77.56,
      "step": 96710,
      "token_acc": 0.551094890510949,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 4.1435671136626535,
      "grad_norm": 6.060286045074463,
      "learning_rate": 7.06612468509037e-06,
      "loss": 2.174710273742676,
      "memory(GiB)": 77.56,
      "step": 96715,
      "token_acc": 0.5825242718446602,
      "train_speed(iter/s)": 1.438083
    },
    {
      "epoch": 4.143781328991903,
      "grad_norm": 6.017674922943115,
      "learning_rate": 7.062675955728315e-06,
      "loss": 2.4469844818115236,
      "memory(GiB)": 77.56,
      "step": 96720,
      "token_acc": 0.5124555160142349,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 4.143995544321152,
      "grad_norm": 4.882197856903076,
      "learning_rate": 7.0592280042161675e-06,
      "loss": 2.641741180419922,
      "memory(GiB)": 77.56,
      "step": 96725,
      "token_acc": 0.4620253164556962,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 4.1442097596504,
      "grad_norm": 5.449734687805176,
      "learning_rate": 7.055780830616382e-06,
      "loss": 2.0512874603271483,
      "memory(GiB)": 77.56,
      "step": 96730,
      "token_acc": 0.5447761194029851,
      "train_speed(iter/s)": 1.438078
    },
    {
      "epoch": 4.14442397497965,
      "grad_norm": 7.146668910980225,
      "learning_rate": 7.052334434991403e-06,
      "loss": 2.209654998779297,
      "memory(GiB)": 77.56,
      "step": 96735,
      "token_acc": 0.5503875968992248,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 4.144638190308899,
      "grad_norm": 7.1566057205200195,
      "learning_rate": 7.048888817403687e-06,
      "loss": 2.210976409912109,
      "memory(GiB)": 77.56,
      "step": 96740,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 4.144852405638147,
      "grad_norm": 7.269792556762695,
      "learning_rate": 7.045443977915639e-06,
      "loss": 2.459884834289551,
      "memory(GiB)": 77.56,
      "step": 96745,
      "token_acc": 0.46488294314381273,
      "train_speed(iter/s)": 1.438116
    },
    {
      "epoch": 4.1450666209673965,
      "grad_norm": 7.2651591300964355,
      "learning_rate": 7.04199991658967e-06,
      "loss": 2.309665489196777,
      "memory(GiB)": 77.56,
      "step": 96750,
      "token_acc": 0.5016611295681063,
      "train_speed(iter/s)": 1.438128
    },
    {
      "epoch": 4.145280836296646,
      "grad_norm": 7.6188225746154785,
      "learning_rate": 7.038556633488169e-06,
      "loss": 2.226517677307129,
      "memory(GiB)": 77.56,
      "step": 96755,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.438129
    },
    {
      "epoch": 4.145495051625894,
      "grad_norm": 6.360135078430176,
      "learning_rate": 7.035114128673503e-06,
      "loss": 2.3225950241088866,
      "memory(GiB)": 77.56,
      "step": 96760,
      "token_acc": 0.48363636363636364,
      "train_speed(iter/s)": 1.438132
    },
    {
      "epoch": 4.145709266955143,
      "grad_norm": 5.371589183807373,
      "learning_rate": 7.031672402208061e-06,
      "loss": 2.3779829025268553,
      "memory(GiB)": 77.56,
      "step": 96765,
      "token_acc": 0.4946236559139785,
      "train_speed(iter/s)": 1.438137
    },
    {
      "epoch": 4.145923482284393,
      "grad_norm": 5.895726203918457,
      "learning_rate": 7.028231454154183e-06,
      "loss": 2.240631103515625,
      "memory(GiB)": 77.56,
      "step": 96770,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.438147
    },
    {
      "epoch": 4.146137697613641,
      "grad_norm": 6.390429973602295,
      "learning_rate": 7.0247912845742005e-06,
      "loss": 2.4046411514282227,
      "memory(GiB)": 77.56,
      "step": 96775,
      "token_acc": 0.46474358974358976,
      "train_speed(iter/s)": 1.438166
    },
    {
      "epoch": 4.14635191294289,
      "grad_norm": 5.912595272064209,
      "learning_rate": 7.021351893530437e-06,
      "loss": 2.308026885986328,
      "memory(GiB)": 77.56,
      "step": 96780,
      "token_acc": 0.5457063711911357,
      "train_speed(iter/s)": 1.438171
    },
    {
      "epoch": 4.1465661282721396,
      "grad_norm": 5.500983238220215,
      "learning_rate": 7.0179132810851924e-06,
      "loss": 2.4201290130615236,
      "memory(GiB)": 77.56,
      "step": 96785,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.438191
    },
    {
      "epoch": 4.146780343601388,
      "grad_norm": 5.839399337768555,
      "learning_rate": 7.014475447300767e-06,
      "loss": 2.5619298934936525,
      "memory(GiB)": 77.56,
      "step": 96790,
      "token_acc": 0.46648793565683644,
      "train_speed(iter/s)": 1.43819
    },
    {
      "epoch": 4.146994558930637,
      "grad_norm": 6.6061201095581055,
      "learning_rate": 7.011038392239455e-06,
      "loss": 2.0807096481323244,
      "memory(GiB)": 77.56,
      "step": 96795,
      "token_acc": 0.5783582089552238,
      "train_speed(iter/s)": 1.438193
    },
    {
      "epoch": 4.147208774259886,
      "grad_norm": 7.838625431060791,
      "learning_rate": 7.007602115963513e-06,
      "loss": 2.33133487701416,
      "memory(GiB)": 77.56,
      "step": 96800,
      "token_acc": 0.4713804713804714,
      "train_speed(iter/s)": 1.438201
    },
    {
      "epoch": 4.147422989589135,
      "grad_norm": 6.6929168701171875,
      "learning_rate": 7.004166618535185e-06,
      "loss": 2.432495880126953,
      "memory(GiB)": 77.56,
      "step": 96805,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.438218
    },
    {
      "epoch": 4.147637204918384,
      "grad_norm": 6.055613994598389,
      "learning_rate": 7.00073190001671e-06,
      "loss": 2.3883293151855467,
      "memory(GiB)": 77.56,
      "step": 96810,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.438215
    },
    {
      "epoch": 4.147851420247633,
      "grad_norm": 8.714577674865723,
      "learning_rate": 6.997297960470317e-06,
      "loss": 2.3065279006958006,
      "memory(GiB)": 77.56,
      "step": 96815,
      "token_acc": 0.5066225165562914,
      "train_speed(iter/s)": 1.438213
    },
    {
      "epoch": 4.148065635576882,
      "grad_norm": 6.251741886138916,
      "learning_rate": 6.9938647999582e-06,
      "loss": 2.1662261962890623,
      "memory(GiB)": 77.56,
      "step": 96820,
      "token_acc": 0.5551601423487544,
      "train_speed(iter/s)": 1.438209
    },
    {
      "epoch": 4.148279850906131,
      "grad_norm": 8.280163764953613,
      "learning_rate": 6.990432418542575e-06,
      "loss": 2.1520479202270506,
      "memory(GiB)": 77.56,
      "step": 96825,
      "token_acc": 0.5163398692810458,
      "train_speed(iter/s)": 1.438209
    },
    {
      "epoch": 4.14849406623538,
      "grad_norm": 6.963266849517822,
      "learning_rate": 6.987000816285611e-06,
      "loss": 2.1842084884643556,
      "memory(GiB)": 77.56,
      "step": 96830,
      "token_acc": 0.5058823529411764,
      "train_speed(iter/s)": 1.43821
    },
    {
      "epoch": 4.148708281564629,
      "grad_norm": 8.484952926635742,
      "learning_rate": 6.983569993249478e-06,
      "loss": 2.486441421508789,
      "memory(GiB)": 77.56,
      "step": 96835,
      "token_acc": 0.4563106796116505,
      "train_speed(iter/s)": 1.438228
    },
    {
      "epoch": 4.148922496893878,
      "grad_norm": 6.30465841293335,
      "learning_rate": 6.9801399494963285e-06,
      "loss": 2.355697822570801,
      "memory(GiB)": 77.56,
      "step": 96840,
      "token_acc": 0.52046783625731,
      "train_speed(iter/s)": 1.438224
    },
    {
      "epoch": 4.149136712223127,
      "grad_norm": 6.986323833465576,
      "learning_rate": 6.976710685088289e-06,
      "loss": 2.2218799591064453,
      "memory(GiB)": 77.56,
      "step": 96845,
      "token_acc": 0.5446808510638298,
      "train_speed(iter/s)": 1.438228
    },
    {
      "epoch": 4.1493509275523754,
      "grad_norm": 7.946080207824707,
      "learning_rate": 6.973282200087506e-06,
      "loss": 2.2127965927124023,
      "memory(GiB)": 77.56,
      "step": 96850,
      "token_acc": 0.5133333333333333,
      "train_speed(iter/s)": 1.438242
    },
    {
      "epoch": 4.149565142881625,
      "grad_norm": 7.758993148803711,
      "learning_rate": 6.969854494556077e-06,
      "loss": 2.3757564544677736,
      "memory(GiB)": 77.56,
      "step": 96855,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.438243
    },
    {
      "epoch": 4.149779358210874,
      "grad_norm": 7.61237096786499,
      "learning_rate": 6.966427568556089e-06,
      "loss": 2.3386470794677736,
      "memory(GiB)": 77.56,
      "step": 96860,
      "token_acc": 0.4984126984126984,
      "train_speed(iter/s)": 1.438253
    },
    {
      "epoch": 4.149993573540122,
      "grad_norm": 5.416196823120117,
      "learning_rate": 6.963001422149646e-06,
      "loss": 2.200771522521973,
      "memory(GiB)": 77.56,
      "step": 96865,
      "token_acc": 0.5387205387205387,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 4.150207788869372,
      "grad_norm": 8.396271705627441,
      "learning_rate": 6.959576055398798e-06,
      "loss": 2.2988351821899413,
      "memory(GiB)": 77.56,
      "step": 96870,
      "token_acc": 0.49504950495049505,
      "train_speed(iter/s)": 1.43823
    },
    {
      "epoch": 4.150422004198621,
      "grad_norm": 7.2163825035095215,
      "learning_rate": 6.956151468365613e-06,
      "loss": 2.447060394287109,
      "memory(GiB)": 77.56,
      "step": 96875,
      "token_acc": 0.4603174603174603,
      "train_speed(iter/s)": 1.438237
    },
    {
      "epoch": 4.150636219527869,
      "grad_norm": 5.654470443725586,
      "learning_rate": 6.952727661112107e-06,
      "loss": 2.118244171142578,
      "memory(GiB)": 77.56,
      "step": 96880,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.438236
    },
    {
      "epoch": 4.1508504348571185,
      "grad_norm": 6.6399078369140625,
      "learning_rate": 6.949304633700332e-06,
      "loss": 2.211338424682617,
      "memory(GiB)": 77.56,
      "step": 96885,
      "token_acc": 0.5617529880478087,
      "train_speed(iter/s)": 1.438218
    },
    {
      "epoch": 4.151064650186368,
      "grad_norm": 8.192840576171875,
      "learning_rate": 6.945882386192293e-06,
      "loss": 2.1926952362060548,
      "memory(GiB)": 77.56,
      "step": 96890,
      "token_acc": 0.5179856115107914,
      "train_speed(iter/s)": 1.438221
    },
    {
      "epoch": 4.151278865515616,
      "grad_norm": 7.22297477722168,
      "learning_rate": 6.942460918649979e-06,
      "loss": 2.6257568359375,
      "memory(GiB)": 77.56,
      "step": 96895,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.43822
    },
    {
      "epoch": 4.151493080844865,
      "grad_norm": 5.9305243492126465,
      "learning_rate": 6.939040231135374e-06,
      "loss": 2.276951026916504,
      "memory(GiB)": 77.56,
      "step": 96900,
      "token_acc": 0.5076335877862596,
      "train_speed(iter/s)": 1.438207
    },
    {
      "epoch": 4.151707296174115,
      "grad_norm": 6.744251728057861,
      "learning_rate": 6.935620323710446e-06,
      "loss": 2.4633607864379883,
      "memory(GiB)": 77.56,
      "step": 96905,
      "token_acc": 0.46200607902735563,
      "train_speed(iter/s)": 1.438217
    },
    {
      "epoch": 4.151921511503363,
      "grad_norm": 6.1679534912109375,
      "learning_rate": 6.9322011964371605e-06,
      "loss": 2.0164342880249024,
      "memory(GiB)": 77.56,
      "step": 96910,
      "token_acc": 0.5588235294117647,
      "train_speed(iter/s)": 1.438231
    },
    {
      "epoch": 4.152135726832612,
      "grad_norm": 5.559601306915283,
      "learning_rate": 6.928782849377447e-06,
      "loss": 2.392054557800293,
      "memory(GiB)": 77.56,
      "step": 96915,
      "token_acc": 0.5168539325842697,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 4.1523499421618615,
      "grad_norm": 6.203959941864014,
      "learning_rate": 6.925365282593244e-06,
      "loss": 2.1733688354492187,
      "memory(GiB)": 77.56,
      "step": 96920,
      "token_acc": 0.5368421052631579,
      "train_speed(iter/s)": 1.438234
    },
    {
      "epoch": 4.15256415749111,
      "grad_norm": 9.566116333007812,
      "learning_rate": 6.921948496146452e-06,
      "loss": 2.2709945678710937,
      "memory(GiB)": 77.56,
      "step": 96925,
      "token_acc": 0.4880952380952381,
      "train_speed(iter/s)": 1.438238
    },
    {
      "epoch": 4.152778372820359,
      "grad_norm": 6.782430171966553,
      "learning_rate": 6.918532490098961e-06,
      "loss": 2.418376350402832,
      "memory(GiB)": 77.56,
      "step": 96930,
      "token_acc": 0.5072046109510087,
      "train_speed(iter/s)": 1.438242
    },
    {
      "epoch": 4.152992588149608,
      "grad_norm": 6.6341872215271,
      "learning_rate": 6.915117264512672e-06,
      "loss": 2.424847221374512,
      "memory(GiB)": 77.56,
      "step": 96935,
      "token_acc": 0.5045871559633027,
      "train_speed(iter/s)": 1.438256
    },
    {
      "epoch": 4.153206803478857,
      "grad_norm": 8.093936920166016,
      "learning_rate": 6.911702819449456e-06,
      "loss": 1.952655029296875,
      "memory(GiB)": 77.56,
      "step": 96940,
      "token_acc": 0.5662100456621004,
      "train_speed(iter/s)": 1.438273
    },
    {
      "epoch": 4.153421018808106,
      "grad_norm": 6.486035346984863,
      "learning_rate": 6.9082891549711705e-06,
      "loss": 2.4411787033081054,
      "memory(GiB)": 77.56,
      "step": 96945,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.438274
    },
    {
      "epoch": 4.153635234137355,
      "grad_norm": 7.692143440246582,
      "learning_rate": 6.904876271139643e-06,
      "loss": 2.8277603149414063,
      "memory(GiB)": 77.56,
      "step": 96950,
      "token_acc": 0.43042071197411,
      "train_speed(iter/s)": 1.438285
    },
    {
      "epoch": 4.153849449466604,
      "grad_norm": 7.961983680725098,
      "learning_rate": 6.901464168016713e-06,
      "loss": 2.3008981704711915,
      "memory(GiB)": 77.56,
      "step": 96955,
      "token_acc": 0.5390625,
      "train_speed(iter/s)": 1.438297
    },
    {
      "epoch": 4.154063664795853,
      "grad_norm": 5.8426594734191895,
      "learning_rate": 6.898052845664188e-06,
      "loss": 2.1396427154541016,
      "memory(GiB)": 77.56,
      "step": 96960,
      "token_acc": 0.5379537953795379,
      "train_speed(iter/s)": 1.438279
    },
    {
      "epoch": 4.154277880125102,
      "grad_norm": 7.845702648162842,
      "learning_rate": 6.894642304143856e-06,
      "loss": 2.1792814254760744,
      "memory(GiB)": 77.56,
      "step": 96965,
      "token_acc": 0.5265306122448979,
      "train_speed(iter/s)": 1.438293
    },
    {
      "epoch": 4.1544920954543505,
      "grad_norm": 7.331331729888916,
      "learning_rate": 6.891232543517529e-06,
      "loss": 1.9727485656738282,
      "memory(GiB)": 77.56,
      "step": 96970,
      "token_acc": 0.5252525252525253,
      "train_speed(iter/s)": 1.438295
    },
    {
      "epoch": 4.1547063107836,
      "grad_norm": 5.405606746673584,
      "learning_rate": 6.887823563846962e-06,
      "loss": 2.4060237884521483,
      "memory(GiB)": 77.56,
      "step": 96975,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.438295
    },
    {
      "epoch": 4.154920526112849,
      "grad_norm": 6.257388591766357,
      "learning_rate": 6.884415365193913e-06,
      "loss": 2.492660331726074,
      "memory(GiB)": 77.56,
      "step": 96980,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.438305
    },
    {
      "epoch": 4.155134741442097,
      "grad_norm": 7.470952987670898,
      "learning_rate": 6.881007947620127e-06,
      "loss": 2.0173486709594726,
      "memory(GiB)": 77.56,
      "step": 96985,
      "token_acc": 0.5942492012779552,
      "train_speed(iter/s)": 1.438313
    },
    {
      "epoch": 4.155348956771347,
      "grad_norm": 6.3645548820495605,
      "learning_rate": 6.877601311187321e-06,
      "loss": 2.078144073486328,
      "memory(GiB)": 77.56,
      "step": 96990,
      "token_acc": 0.5274390243902439,
      "train_speed(iter/s)": 1.43831
    },
    {
      "epoch": 4.155563172100596,
      "grad_norm": 5.185481071472168,
      "learning_rate": 6.874195455957227e-06,
      "loss": 2.1442182540893553,
      "memory(GiB)": 77.56,
      "step": 96995,
      "token_acc": 0.5112781954887218,
      "train_speed(iter/s)": 1.438319
    },
    {
      "epoch": 4.155777387429844,
      "grad_norm": 5.954684257507324,
      "learning_rate": 6.870790381991538e-06,
      "loss": 2.166065216064453,
      "memory(GiB)": 77.56,
      "step": 97000,
      "token_acc": 0.5612648221343873,
      "train_speed(iter/s)": 1.438328
    },
    {
      "epoch": 4.155777387429844,
      "eval_loss": 2.2515976428985596,
      "eval_runtime": 15.265,
      "eval_samples_per_second": 6.551,
      "eval_steps_per_second": 6.551,
      "eval_token_acc": 0.4630541871921182,
      "step": 97000
    },
    {
      "epoch": 4.1559916027590935,
      "grad_norm": 6.8432512283325195,
      "learning_rate": 6.86738608935194e-06,
      "loss": 2.371525192260742,
      "memory(GiB)": 77.56,
      "step": 97005,
      "token_acc": 0.46616541353383456,
      "train_speed(iter/s)": 1.437988
    },
    {
      "epoch": 4.156205818088343,
      "grad_norm": 6.098437309265137,
      "learning_rate": 6.863982578100098e-06,
      "loss": 2.3131662368774415,
      "memory(GiB)": 77.56,
      "step": 97010,
      "token_acc": 0.5130718954248366,
      "train_speed(iter/s)": 1.437991
    },
    {
      "epoch": 4.156420033417591,
      "grad_norm": 6.7221550941467285,
      "learning_rate": 6.860579848297683e-06,
      "loss": 2.4015438079833986,
      "memory(GiB)": 77.56,
      "step": 97015,
      "token_acc": 0.48231511254019294,
      "train_speed(iter/s)": 1.437995
    },
    {
      "epoch": 4.15663424874684,
      "grad_norm": 6.369801998138428,
      "learning_rate": 6.857177900006317e-06,
      "loss": 2.290873718261719,
      "memory(GiB)": 77.56,
      "step": 97020,
      "token_acc": 0.5244299674267101,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 4.15684846407609,
      "grad_norm": 7.332221984863281,
      "learning_rate": 6.853776733287665e-06,
      "loss": 2.112679862976074,
      "memory(GiB)": 77.56,
      "step": 97025,
      "token_acc": 0.4845360824742268,
      "train_speed(iter/s)": 1.438025
    },
    {
      "epoch": 4.157062679405338,
      "grad_norm": 7.508343696594238,
      "learning_rate": 6.850376348203313e-06,
      "loss": 2.262190818786621,
      "memory(GiB)": 77.56,
      "step": 97030,
      "token_acc": 0.5250836120401338,
      "train_speed(iter/s)": 1.438019
    },
    {
      "epoch": 4.157276894734587,
      "grad_norm": 6.953838348388672,
      "learning_rate": 6.8469767448148755e-06,
      "loss": 2.7071453094482423,
      "memory(GiB)": 77.56,
      "step": 97035,
      "token_acc": 0.4740484429065744,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 4.1574911100638365,
      "grad_norm": 6.497248649597168,
      "learning_rate": 6.843577923183936e-06,
      "loss": 1.9916418075561524,
      "memory(GiB)": 77.56,
      "step": 97040,
      "token_acc": 0.5538461538461539,
      "train_speed(iter/s)": 1.438013
    },
    {
      "epoch": 4.157705325393085,
      "grad_norm": 8.44781494140625,
      "learning_rate": 6.840179883372066e-06,
      "loss": 2.198017120361328,
      "memory(GiB)": 77.56,
      "step": 97045,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 4.157919540722334,
      "grad_norm": 6.003915309906006,
      "learning_rate": 6.836782625440813e-06,
      "loss": 2.099190521240234,
      "memory(GiB)": 77.56,
      "step": 97050,
      "token_acc": 0.5662251655629139,
      "train_speed(iter/s)": 1.438025
    },
    {
      "epoch": 4.158133756051583,
      "grad_norm": 5.657415866851807,
      "learning_rate": 6.833386149451748e-06,
      "loss": 2.389454650878906,
      "memory(GiB)": 77.56,
      "step": 97055,
      "token_acc": 0.4853801169590643,
      "train_speed(iter/s)": 1.438044
    },
    {
      "epoch": 4.158347971380832,
      "grad_norm": 5.5086140632629395,
      "learning_rate": 6.8299904554663785e-06,
      "loss": 2.06239013671875,
      "memory(GiB)": 77.56,
      "step": 97060,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.438035
    },
    {
      "epoch": 4.158562186710081,
      "grad_norm": 6.3214263916015625,
      "learning_rate": 6.8265955435462325e-06,
      "loss": 2.307545471191406,
      "memory(GiB)": 77.56,
      "step": 97065,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438028
    },
    {
      "epoch": 4.15877640203933,
      "grad_norm": 7.249273777008057,
      "learning_rate": 6.823201413752811e-06,
      "loss": 2.0276695251464845,
      "memory(GiB)": 77.56,
      "step": 97070,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.438037
    },
    {
      "epoch": 4.158990617368579,
      "grad_norm": 7.243667125701904,
      "learning_rate": 6.819808066147587e-06,
      "loss": 2.564301681518555,
      "memory(GiB)": 77.56,
      "step": 97075,
      "token_acc": 0.4750830564784053,
      "train_speed(iter/s)": 1.438021
    },
    {
      "epoch": 4.159204832697828,
      "grad_norm": 9.378226280212402,
      "learning_rate": 6.816415500792056e-06,
      "loss": 2.2360790252685545,
      "memory(GiB)": 77.56,
      "step": 97080,
      "token_acc": 0.5365853658536586,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 4.159419048027077,
      "grad_norm": 6.667816162109375,
      "learning_rate": 6.813023717747652e-06,
      "loss": 2.4280372619628907,
      "memory(GiB)": 77.56,
      "step": 97085,
      "token_acc": 0.532608695652174,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 4.1596332633563255,
      "grad_norm": 5.489818572998047,
      "learning_rate": 6.8096327170758535e-06,
      "loss": 2.134889030456543,
      "memory(GiB)": 77.56,
      "step": 97090,
      "token_acc": 0.5346153846153846,
      "train_speed(iter/s)": 1.438001
    },
    {
      "epoch": 4.159847478685575,
      "grad_norm": 5.1787109375,
      "learning_rate": 6.806242498838072e-06,
      "loss": 2.403506851196289,
      "memory(GiB)": 77.56,
      "step": 97095,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.438002
    },
    {
      "epoch": 4.160061694014824,
      "grad_norm": 5.666685104370117,
      "learning_rate": 6.802853063095727e-06,
      "loss": 1.9356639862060547,
      "memory(GiB)": 77.56,
      "step": 97100,
      "token_acc": 0.5779467680608364,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 4.160275909344072,
      "grad_norm": 6.130240440368652,
      "learning_rate": 6.799464409910222e-06,
      "loss": 2.1629528045654296,
      "memory(GiB)": 77.56,
      "step": 97105,
      "token_acc": 0.5436893203883495,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 4.160490124673322,
      "grad_norm": 5.2250447273254395,
      "learning_rate": 6.796076539342933e-06,
      "loss": 2.1709333419799806,
      "memory(GiB)": 77.56,
      "step": 97110,
      "token_acc": 0.5544871794871795,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 4.160704340002571,
      "grad_norm": 5.822183609008789,
      "learning_rate": 6.792689451455253e-06,
      "loss": 2.3373680114746094,
      "memory(GiB)": 77.56,
      "step": 97115,
      "token_acc": 0.5013850415512465,
      "train_speed(iter/s)": 1.438
    },
    {
      "epoch": 4.160918555331819,
      "grad_norm": 6.231523513793945,
      "learning_rate": 6.78930314630854e-06,
      "loss": 2.2892728805541993,
      "memory(GiB)": 77.56,
      "step": 97120,
      "token_acc": 0.49333333333333335,
      "train_speed(iter/s)": 1.438004
    },
    {
      "epoch": 4.161132770661069,
      "grad_norm": 5.831514358520508,
      "learning_rate": 6.785917623964133e-06,
      "loss": 2.402569580078125,
      "memory(GiB)": 77.56,
      "step": 97125,
      "token_acc": 0.5087719298245614,
      "train_speed(iter/s)": 1.437989
    },
    {
      "epoch": 4.161346985990318,
      "grad_norm": 7.194214344024658,
      "learning_rate": 6.782532884483367e-06,
      "loss": 2.309739112854004,
      "memory(GiB)": 77.56,
      "step": 97130,
      "token_acc": 0.521613832853026,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 4.161561201319566,
      "grad_norm": 6.924748420715332,
      "learning_rate": 6.779148927927548e-06,
      "loss": 2.4431758880615235,
      "memory(GiB)": 77.56,
      "step": 97135,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.438022
    },
    {
      "epoch": 4.1617754166488155,
      "grad_norm": 5.608720302581787,
      "learning_rate": 6.775765754358004e-06,
      "loss": 2.1150487899780273,
      "memory(GiB)": 77.56,
      "step": 97140,
      "token_acc": 0.540625,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 4.161989631978065,
      "grad_norm": 8.364985466003418,
      "learning_rate": 6.77238336383601e-06,
      "loss": 2.1854007720947264,
      "memory(GiB)": 77.56,
      "step": 97145,
      "token_acc": 0.5198412698412699,
      "train_speed(iter/s)": 1.438051
    },
    {
      "epoch": 4.162203847307313,
      "grad_norm": 6.343231201171875,
      "learning_rate": 6.769001756422838e-06,
      "loss": 2.1295557022094727,
      "memory(GiB)": 77.56,
      "step": 97150,
      "token_acc": 0.5414012738853503,
      "train_speed(iter/s)": 1.438045
    },
    {
      "epoch": 4.162418062636562,
      "grad_norm": 7.365429878234863,
      "learning_rate": 6.765620932179756e-06,
      "loss": 2.4055419921875,
      "memory(GiB)": 77.56,
      "step": 97155,
      "token_acc": 0.5076335877862596,
      "train_speed(iter/s)": 1.438032
    },
    {
      "epoch": 4.162632277965812,
      "grad_norm": 7.216892242431641,
      "learning_rate": 6.7622408911679976e-06,
      "loss": 2.376376724243164,
      "memory(GiB)": 77.56,
      "step": 97160,
      "token_acc": 0.5015384615384615,
      "train_speed(iter/s)": 1.438011
    },
    {
      "epoch": 4.16284649329506,
      "grad_norm": 5.5754594802856445,
      "learning_rate": 6.7588616334488045e-06,
      "loss": 2.0337640762329103,
      "memory(GiB)": 77.56,
      "step": 97165,
      "token_acc": 0.5785123966942148,
      "train_speed(iter/s)": 1.43801
    },
    {
      "epoch": 4.163060708624309,
      "grad_norm": 5.141740798950195,
      "learning_rate": 6.7554831590834086e-06,
      "loss": 2.223271942138672,
      "memory(GiB)": 77.56,
      "step": 97170,
      "token_acc": 0.512280701754386,
      "train_speed(iter/s)": 1.438006
    },
    {
      "epoch": 4.1632749239535585,
      "grad_norm": 7.582283973693848,
      "learning_rate": 6.752105468133002e-06,
      "loss": 2.301517105102539,
      "memory(GiB)": 77.56,
      "step": 97175,
      "token_acc": 0.5087108013937283,
      "train_speed(iter/s)": 1.438003
    },
    {
      "epoch": 4.163489139282807,
      "grad_norm": 5.5994391441345215,
      "learning_rate": 6.748728560658774e-06,
      "loss": 2.0906661987304687,
      "memory(GiB)": 77.56,
      "step": 97180,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.438012
    },
    {
      "epoch": 4.163703354612056,
      "grad_norm": 7.829991340637207,
      "learning_rate": 6.745352436721902e-06,
      "loss": 2.427323913574219,
      "memory(GiB)": 77.56,
      "step": 97185,
      "token_acc": 0.5152671755725191,
      "train_speed(iter/s)": 1.438017
    },
    {
      "epoch": 4.163917569941305,
      "grad_norm": 6.009195327758789,
      "learning_rate": 6.741977096383545e-06,
      "loss": 2.268332099914551,
      "memory(GiB)": 77.56,
      "step": 97190,
      "token_acc": 0.49843260188087773,
      "train_speed(iter/s)": 1.438031
    },
    {
      "epoch": 4.164131785270554,
      "grad_norm": 10.224699974060059,
      "learning_rate": 6.738602539704842e-06,
      "loss": 2.638759231567383,
      "memory(GiB)": 77.56,
      "step": 97195,
      "token_acc": 0.5117056856187291,
      "train_speed(iter/s)": 1.438054
    },
    {
      "epoch": 4.164346000599803,
      "grad_norm": 9.838624954223633,
      "learning_rate": 6.735228766746948e-06,
      "loss": 2.439218521118164,
      "memory(GiB)": 77.56,
      "step": 97200,
      "token_acc": 0.5076923076923077,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 4.164560215929052,
      "grad_norm": 6.8448710441589355,
      "learning_rate": 6.731855777570972e-06,
      "loss": 2.155548095703125,
      "memory(GiB)": 77.56,
      "step": 97205,
      "token_acc": 0.541958041958042,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.164774431258301,
      "grad_norm": 6.567218780517578,
      "learning_rate": 6.728483572238015e-06,
      "loss": 2.27388858795166,
      "memory(GiB)": 77.56,
      "step": 97210,
      "token_acc": 0.5175097276264592,
      "train_speed(iter/s)": 1.438095
    },
    {
      "epoch": 4.16498864658755,
      "grad_norm": 5.403478622436523,
      "learning_rate": 6.725112150809171e-06,
      "loss": 2.580466461181641,
      "memory(GiB)": 77.56,
      "step": 97215,
      "token_acc": 0.4702702702702703,
      "train_speed(iter/s)": 1.438118
    },
    {
      "epoch": 4.165202861916799,
      "grad_norm": 8.584498405456543,
      "learning_rate": 6.721741513345503e-06,
      "loss": 2.326335144042969,
      "memory(GiB)": 77.56,
      "step": 97220,
      "token_acc": 0.4560810810810811,
      "train_speed(iter/s)": 1.438127
    },
    {
      "epoch": 4.1654170772460475,
      "grad_norm": 8.195961952209473,
      "learning_rate": 6.718371659908101e-06,
      "loss": 2.3034795761108398,
      "memory(GiB)": 77.56,
      "step": 97225,
      "token_acc": 0.4792243767313019,
      "train_speed(iter/s)": 1.438129
    },
    {
      "epoch": 4.165631292575297,
      "grad_norm": 5.719882488250732,
      "learning_rate": 6.715002590557984e-06,
      "loss": 2.191481018066406,
      "memory(GiB)": 77.56,
      "step": 97230,
      "token_acc": 0.5443786982248521,
      "train_speed(iter/s)": 1.438123
    },
    {
      "epoch": 4.165845507904546,
      "grad_norm": 10.552340507507324,
      "learning_rate": 6.711634305356207e-06,
      "loss": 2.485573959350586,
      "memory(GiB)": 77.56,
      "step": 97235,
      "token_acc": 0.4753521126760563,
      "train_speed(iter/s)": 1.438118
    },
    {
      "epoch": 4.166059723233794,
      "grad_norm": 6.061198711395264,
      "learning_rate": 6.708266804363789e-06,
      "loss": 1.971318817138672,
      "memory(GiB)": 77.56,
      "step": 97240,
      "token_acc": 0.5469798657718121,
      "train_speed(iter/s)": 1.438113
    },
    {
      "epoch": 4.166273938563044,
      "grad_norm": 6.476930141448975,
      "learning_rate": 6.704900087641725e-06,
      "loss": 2.3940757751464843,
      "memory(GiB)": 77.56,
      "step": 97245,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.438116
    },
    {
      "epoch": 4.166488153892293,
      "grad_norm": 5.970035552978516,
      "learning_rate": 6.70153415525101e-06,
      "loss": 2.409836769104004,
      "memory(GiB)": 77.56,
      "step": 97250,
      "token_acc": 0.4773413897280967,
      "train_speed(iter/s)": 1.438113
    },
    {
      "epoch": 4.166702369221541,
      "grad_norm": 7.412076473236084,
      "learning_rate": 6.6981690072526085e-06,
      "loss": 2.2710716247558596,
      "memory(GiB)": 77.56,
      "step": 97255,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.438108
    },
    {
      "epoch": 4.1669165845507905,
      "grad_norm": 6.00469446182251,
      "learning_rate": 6.694804643707509e-06,
      "loss": 2.3177921295166017,
      "memory(GiB)": 77.56,
      "step": 97260,
      "token_acc": 0.5129032258064516,
      "train_speed(iter/s)": 1.438119
    },
    {
      "epoch": 4.16713079988004,
      "grad_norm": 7.901450157165527,
      "learning_rate": 6.691441064676651e-06,
      "loss": 2.195972442626953,
      "memory(GiB)": 77.56,
      "step": 97265,
      "token_acc": 0.5427631578947368,
      "train_speed(iter/s)": 1.438107
    },
    {
      "epoch": 4.167345015209288,
      "grad_norm": 6.463228702545166,
      "learning_rate": 6.68807827022096e-06,
      "loss": 2.4190383911132813,
      "memory(GiB)": 77.56,
      "step": 97270,
      "token_acc": 0.5163636363636364,
      "train_speed(iter/s)": 1.438111
    },
    {
      "epoch": 4.167559230538537,
      "grad_norm": 6.3011860847473145,
      "learning_rate": 6.684716260401358e-06,
      "loss": 1.8667572021484375,
      "memory(GiB)": 77.56,
      "step": 97275,
      "token_acc": 0.5923076923076923,
      "train_speed(iter/s)": 1.43811
    },
    {
      "epoch": 4.167773445867787,
      "grad_norm": 5.087489128112793,
      "learning_rate": 6.681355035278747e-06,
      "loss": 2.281972122192383,
      "memory(GiB)": 77.56,
      "step": 97280,
      "token_acc": 0.5055555555555555,
      "train_speed(iter/s)": 1.438112
    },
    {
      "epoch": 4.167987661197035,
      "grad_norm": 8.16202163696289,
      "learning_rate": 6.677994594914039e-06,
      "loss": 2.288997840881348,
      "memory(GiB)": 77.56,
      "step": 97285,
      "token_acc": 0.5121107266435986,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 4.168201876526284,
      "grad_norm": 8.033756256103516,
      "learning_rate": 6.674634939368096e-06,
      "loss": 2.1416812896728517,
      "memory(GiB)": 77.56,
      "step": 97290,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.438148
    },
    {
      "epoch": 4.1684160918555335,
      "grad_norm": 7.147014617919922,
      "learning_rate": 6.671276068701782e-06,
      "loss": 2.0187665939331056,
      "memory(GiB)": 77.56,
      "step": 97295,
      "token_acc": 0.5411255411255411,
      "train_speed(iter/s)": 1.438157
    },
    {
      "epoch": 4.168630307184782,
      "grad_norm": 6.444074630737305,
      "learning_rate": 6.667917982975947e-06,
      "loss": 2.212594223022461,
      "memory(GiB)": 77.56,
      "step": 97300,
      "token_acc": 0.5604395604395604,
      "train_speed(iter/s)": 1.438153
    },
    {
      "epoch": 4.168844522514031,
      "grad_norm": 5.972939968109131,
      "learning_rate": 6.66456068225142e-06,
      "loss": 2.571119689941406,
      "memory(GiB)": 77.56,
      "step": 97305,
      "token_acc": 0.43843843843843844,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 4.16905873784328,
      "grad_norm": 6.782871723175049,
      "learning_rate": 6.661204166589025e-06,
      "loss": 2.6636266708374023,
      "memory(GiB)": 77.56,
      "step": 97310,
      "token_acc": 0.4797047970479705,
      "train_speed(iter/s)": 1.438155
    },
    {
      "epoch": 4.169272953172529,
      "grad_norm": 5.814826965332031,
      "learning_rate": 6.657848436049585e-06,
      "loss": 2.2265905380249023,
      "memory(GiB)": 77.56,
      "step": 97315,
      "token_acc": 0.5512367491166078,
      "train_speed(iter/s)": 1.43817
    },
    {
      "epoch": 4.169487168501778,
      "grad_norm": 8.261284828186035,
      "learning_rate": 6.654493490693875e-06,
      "loss": 2.132475471496582,
      "memory(GiB)": 77.56,
      "step": 97320,
      "token_acc": 0.5515695067264574,
      "train_speed(iter/s)": 1.438193
    },
    {
      "epoch": 4.169701383831027,
      "grad_norm": 6.549203395843506,
      "learning_rate": 6.651139330582679e-06,
      "loss": 1.973699951171875,
      "memory(GiB)": 77.56,
      "step": 97325,
      "token_acc": 0.5362318840579711,
      "train_speed(iter/s)": 1.438184
    },
    {
      "epoch": 4.169915599160276,
      "grad_norm": 7.741466999053955,
      "learning_rate": 6.647785955776759e-06,
      "loss": 2.3276317596435545,
      "memory(GiB)": 77.56,
      "step": 97330,
      "token_acc": 0.46691176470588236,
      "train_speed(iter/s)": 1.438197
    },
    {
      "epoch": 4.170129814489525,
      "grad_norm": 5.569271087646484,
      "learning_rate": 6.644433366336861e-06,
      "loss": 2.323288345336914,
      "memory(GiB)": 77.56,
      "step": 97335,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.43821
    },
    {
      "epoch": 4.170344029818774,
      "grad_norm": 7.044909477233887,
      "learning_rate": 6.6410815623237145e-06,
      "loss": 2.2888736724853516,
      "memory(GiB)": 77.56,
      "step": 97340,
      "token_acc": 0.4880952380952381,
      "train_speed(iter/s)": 1.438202
    },
    {
      "epoch": 4.1705582451480225,
      "grad_norm": 6.836385250091553,
      "learning_rate": 6.6377305437980596e-06,
      "loss": 2.1167577743530273,
      "memory(GiB)": 77.56,
      "step": 97345,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.438211
    },
    {
      "epoch": 4.170772460477272,
      "grad_norm": 7.373457431793213,
      "learning_rate": 6.634380310820598e-06,
      "loss": 2.291274642944336,
      "memory(GiB)": 77.56,
      "step": 97350,
      "token_acc": 0.5441696113074205,
      "train_speed(iter/s)": 1.438214
    },
    {
      "epoch": 4.170986675806521,
      "grad_norm": 5.679426670074463,
      "learning_rate": 6.6310308634520144e-06,
      "loss": 2.4195085525512696,
      "memory(GiB)": 77.56,
      "step": 97355,
      "token_acc": 0.4930747922437673,
      "train_speed(iter/s)": 1.43821
    },
    {
      "epoch": 4.171200891135769,
      "grad_norm": 8.170836448669434,
      "learning_rate": 6.627682201752988e-06,
      "loss": 2.4505640029907227,
      "memory(GiB)": 77.56,
      "step": 97360,
      "token_acc": 0.45874587458745875,
      "train_speed(iter/s)": 1.438215
    },
    {
      "epoch": 4.171415106465019,
      "grad_norm": 5.908185958862305,
      "learning_rate": 6.6243343257841765e-06,
      "loss": 2.109127998352051,
      "memory(GiB)": 77.56,
      "step": 97365,
      "token_acc": 0.5523809523809524,
      "train_speed(iter/s)": 1.438178
    },
    {
      "epoch": 4.171629321794268,
      "grad_norm": 6.141332626342773,
      "learning_rate": 6.620987235606246e-06,
      "loss": 2.1970733642578124,
      "memory(GiB)": 77.56,
      "step": 97370,
      "token_acc": 0.5287356321839081,
      "train_speed(iter/s)": 1.438164
    },
    {
      "epoch": 4.171843537123516,
      "grad_norm": 7.779545307159424,
      "learning_rate": 6.617640931279828e-06,
      "loss": 2.199837303161621,
      "memory(GiB)": 77.56,
      "step": 97375,
      "token_acc": 0.5264797507788161,
      "train_speed(iter/s)": 1.438181
    },
    {
      "epoch": 4.1720577524527656,
      "grad_norm": 5.421979904174805,
      "learning_rate": 6.614295412865523e-06,
      "loss": 2.2061885833740233,
      "memory(GiB)": 77.56,
      "step": 97380,
      "token_acc": 0.5523465703971119,
      "train_speed(iter/s)": 1.438177
    },
    {
      "epoch": 4.172271967782015,
      "grad_norm": 7.4345479011535645,
      "learning_rate": 6.610950680423972e-06,
      "loss": 2.268479919433594,
      "memory(GiB)": 77.56,
      "step": 97385,
      "token_acc": 0.47151898734177217,
      "train_speed(iter/s)": 1.438193
    },
    {
      "epoch": 4.172486183111263,
      "grad_norm": 5.9489336013793945,
      "learning_rate": 6.607606734015753e-06,
      "loss": 2.3969257354736326,
      "memory(GiB)": 77.56,
      "step": 97390,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.438194
    },
    {
      "epoch": 4.172700398440512,
      "grad_norm": 6.3637919425964355,
      "learning_rate": 6.604263573701441e-06,
      "loss": 2.112289047241211,
      "memory(GiB)": 77.56,
      "step": 97395,
      "token_acc": 0.5512367491166078,
      "train_speed(iter/s)": 1.438205
    },
    {
      "epoch": 4.172914613769762,
      "grad_norm": 6.1700053215026855,
      "learning_rate": 6.600921199541593e-06,
      "loss": 2.129522705078125,
      "memory(GiB)": 77.56,
      "step": 97400,
      "token_acc": 0.515358361774744,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 4.17312882909901,
      "grad_norm": 4.967783451080322,
      "learning_rate": 6.597579611596782e-06,
      "loss": 2.297202491760254,
      "memory(GiB)": 77.56,
      "step": 97405,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438235
    },
    {
      "epoch": 4.173343044428259,
      "grad_norm": 6.289876937866211,
      "learning_rate": 6.594238809927527e-06,
      "loss": 2.1155399322509765,
      "memory(GiB)": 77.56,
      "step": 97410,
      "token_acc": 0.5232558139534884,
      "train_speed(iter/s)": 1.438236
    },
    {
      "epoch": 4.173557259757509,
      "grad_norm": 5.384018421173096,
      "learning_rate": 6.590898794594358e-06,
      "loss": 2.0759410858154297,
      "memory(GiB)": 77.56,
      "step": 97415,
      "token_acc": 0.5406360424028268,
      "train_speed(iter/s)": 1.438245
    },
    {
      "epoch": 4.173771475086757,
      "grad_norm": 6.808653354644775,
      "learning_rate": 6.587559565657775e-06,
      "loss": 1.897985076904297,
      "memory(GiB)": 77.56,
      "step": 97420,
      "token_acc": 0.5285714285714286,
      "train_speed(iter/s)": 1.438261
    },
    {
      "epoch": 4.173985690416006,
      "grad_norm": 5.335715293884277,
      "learning_rate": 6.584221123178263e-06,
      "loss": 2.418997383117676,
      "memory(GiB)": 77.56,
      "step": 97425,
      "token_acc": 0.4937888198757764,
      "train_speed(iter/s)": 1.438263
    },
    {
      "epoch": 4.1741999057452555,
      "grad_norm": 6.962917804718018,
      "learning_rate": 6.580883467216326e-06,
      "loss": 2.406814193725586,
      "memory(GiB)": 77.56,
      "step": 97430,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.438263
    },
    {
      "epoch": 4.174414121074504,
      "grad_norm": 6.551785945892334,
      "learning_rate": 6.577546597832413e-06,
      "loss": 2.2140954971313476,
      "memory(GiB)": 77.56,
      "step": 97435,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438246
    },
    {
      "epoch": 4.174628336403753,
      "grad_norm": 9.075545310974121,
      "learning_rate": 6.574210515086982e-06,
      "loss": 2.3686283111572264,
      "memory(GiB)": 77.56,
      "step": 97440,
      "token_acc": 0.5034722222222222,
      "train_speed(iter/s)": 1.438244
    },
    {
      "epoch": 4.174842551733002,
      "grad_norm": 5.752353191375732,
      "learning_rate": 6.570875219040457e-06,
      "loss": 2.3777570724487305,
      "memory(GiB)": 77.56,
      "step": 97445,
      "token_acc": 0.5054945054945055,
      "train_speed(iter/s)": 1.438259
    },
    {
      "epoch": 4.175056767062251,
      "grad_norm": 9.251683235168457,
      "learning_rate": 6.567540709753262e-06,
      "loss": 2.174790382385254,
      "memory(GiB)": 77.56,
      "step": 97450,
      "token_acc": 0.5531135531135531,
      "train_speed(iter/s)": 1.438251
    },
    {
      "epoch": 4.1752709823915,
      "grad_norm": 7.3459296226501465,
      "learning_rate": 6.564206987285809e-06,
      "loss": 2.243643379211426,
      "memory(GiB)": 77.56,
      "step": 97455,
      "token_acc": 0.5135135135135135,
      "train_speed(iter/s)": 1.43826
    },
    {
      "epoch": 4.175485197720749,
      "grad_norm": 5.697394371032715,
      "learning_rate": 6.560874051698502e-06,
      "loss": 2.3248973846435548,
      "memory(GiB)": 77.56,
      "step": 97460,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.438272
    },
    {
      "epoch": 4.175699413049998,
      "grad_norm": 7.8997673988342285,
      "learning_rate": 6.557541903051712e-06,
      "loss": 2.3239925384521483,
      "memory(GiB)": 77.56,
      "step": 97465,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.438279
    },
    {
      "epoch": 4.175913628379247,
      "grad_norm": 6.572592258453369,
      "learning_rate": 6.554210541405803e-06,
      "loss": 2.3544687271118163,
      "memory(GiB)": 77.56,
      "step": 97470,
      "token_acc": 0.5018450184501845,
      "train_speed(iter/s)": 1.438275
    },
    {
      "epoch": 4.176127843708496,
      "grad_norm": 9.373518943786621,
      "learning_rate": 6.550879966821122e-06,
      "loss": 2.298045349121094,
      "memory(GiB)": 77.56,
      "step": 97475,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438271
    },
    {
      "epoch": 4.1763420590377445,
      "grad_norm": 9.53109073638916,
      "learning_rate": 6.547550179358014e-06,
      "loss": 2.5055658340454103,
      "memory(GiB)": 77.56,
      "step": 97480,
      "token_acc": 0.501779359430605,
      "train_speed(iter/s)": 1.438271
    },
    {
      "epoch": 4.176556274366994,
      "grad_norm": 5.433248043060303,
      "learning_rate": 6.544221179076782e-06,
      "loss": 2.0234601974487303,
      "memory(GiB)": 77.56,
      "step": 97485,
      "token_acc": 0.49612403100775193,
      "train_speed(iter/s)": 1.438262
    },
    {
      "epoch": 4.176770489696243,
      "grad_norm": 7.148240566253662,
      "learning_rate": 6.5408929660377595e-06,
      "loss": 2.3061311721801756,
      "memory(GiB)": 77.56,
      "step": 97490,
      "token_acc": 0.5081967213114754,
      "train_speed(iter/s)": 1.438277
    },
    {
      "epoch": 4.176984705025491,
      "grad_norm": 5.337233543395996,
      "learning_rate": 6.537565540301227e-06,
      "loss": 2.241322708129883,
      "memory(GiB)": 77.56,
      "step": 97495,
      "token_acc": 0.5379061371841155,
      "train_speed(iter/s)": 1.438294
    },
    {
      "epoch": 4.177198920354741,
      "grad_norm": 13.385344505310059,
      "learning_rate": 6.534238901927469e-06,
      "loss": 2.5765384674072265,
      "memory(GiB)": 77.56,
      "step": 97500,
      "token_acc": 0.4503105590062112,
      "train_speed(iter/s)": 1.438301
    },
    {
      "epoch": 4.177198920354741,
      "eval_loss": 2.1568682193756104,
      "eval_runtime": 13.9265,
      "eval_samples_per_second": 7.181,
      "eval_steps_per_second": 7.181,
      "eval_token_acc": 0.48476454293628807,
      "step": 97500
    },
    {
      "epoch": 4.17741313568399,
      "grad_norm": 5.83855676651001,
      "learning_rate": 6.530913050976744e-06,
      "loss": 2.324435997009277,
      "memory(GiB)": 77.56,
      "step": 97505,
      "token_acc": 0.4935960591133005,
      "train_speed(iter/s)": 1.438004
    },
    {
      "epoch": 4.177627351013238,
      "grad_norm": 8.217616081237793,
      "learning_rate": 6.527587987509299e-06,
      "loss": 2.521082878112793,
      "memory(GiB)": 77.56,
      "step": 97510,
      "token_acc": 0.4769874476987448,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 4.1778415663424875,
      "grad_norm": 6.036564826965332,
      "learning_rate": 6.5242637115853876e-06,
      "loss": 2.2948129653930662,
      "memory(GiB)": 77.56,
      "step": 97515,
      "token_acc": 0.5035211267605634,
      "train_speed(iter/s)": 1.438018
    },
    {
      "epoch": 4.178055781671737,
      "grad_norm": 5.985663414001465,
      "learning_rate": 6.520940223265226e-06,
      "loss": 2.241323471069336,
      "memory(GiB)": 77.56,
      "step": 97520,
      "token_acc": 0.49498327759197325,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 4.178269997000985,
      "grad_norm": 6.278834342956543,
      "learning_rate": 6.517617522609015e-06,
      "loss": 2.413946533203125,
      "memory(GiB)": 77.56,
      "step": 97525,
      "token_acc": 0.5217391304347826,
      "train_speed(iter/s)": 1.438008
    },
    {
      "epoch": 4.178484212330234,
      "grad_norm": 9.926697731018066,
      "learning_rate": 6.51429560967694e-06,
      "loss": 2.7918434143066406,
      "memory(GiB)": 77.56,
      "step": 97530,
      "token_acc": 0.4568345323741007,
      "train_speed(iter/s)": 1.438033
    },
    {
      "epoch": 4.178698427659484,
      "grad_norm": 6.260964393615723,
      "learning_rate": 6.510974484529209e-06,
      "loss": 2.324381637573242,
      "memory(GiB)": 77.56,
      "step": 97535,
      "token_acc": 0.504885993485342,
      "train_speed(iter/s)": 1.43803
    },
    {
      "epoch": 4.178912642988732,
      "grad_norm": 4.834085941314697,
      "learning_rate": 6.5076541472259666e-06,
      "loss": 2.2774003982543944,
      "memory(GiB)": 77.56,
      "step": 97540,
      "token_acc": 0.5032679738562091,
      "train_speed(iter/s)": 1.438038
    },
    {
      "epoch": 4.179126858317981,
      "grad_norm": 6.412882328033447,
      "learning_rate": 6.50433459782736e-06,
      "loss": 2.52331600189209,
      "memory(GiB)": 77.56,
      "step": 97545,
      "token_acc": 0.4489795918367347,
      "train_speed(iter/s)": 1.438046
    },
    {
      "epoch": 4.1793410736472305,
      "grad_norm": 4.971535682678223,
      "learning_rate": 6.501015836393543e-06,
      "loss": 2.296042251586914,
      "memory(GiB)": 77.56,
      "step": 97550,
      "token_acc": 0.5026737967914439,
      "train_speed(iter/s)": 1.438053
    },
    {
      "epoch": 4.179555288976479,
      "grad_norm": 6.823814392089844,
      "learning_rate": 6.497697862984631e-06,
      "loss": 2.1310056686401366,
      "memory(GiB)": 77.56,
      "step": 97555,
      "token_acc": 0.5774647887323944,
      "train_speed(iter/s)": 1.438066
    },
    {
      "epoch": 4.179769504305728,
      "grad_norm": 13.579385757446289,
      "learning_rate": 6.494380677660733e-06,
      "loss": 2.1263118743896485,
      "memory(GiB)": 77.56,
      "step": 97560,
      "token_acc": 0.5327868852459017,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 4.179983719634977,
      "grad_norm": 6.955369472503662,
      "learning_rate": 6.491064280481934e-06,
      "loss": 2.122196006774902,
      "memory(GiB)": 77.56,
      "step": 97565,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 4.180197934964226,
      "grad_norm": 5.315395832061768,
      "learning_rate": 6.4877486715083144e-06,
      "loss": 2.205295753479004,
      "memory(GiB)": 77.56,
      "step": 97570,
      "token_acc": 0.5517241379310345,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 4.180412150293475,
      "grad_norm": 6.053256034851074,
      "learning_rate": 6.484433850799959e-06,
      "loss": 2.146514129638672,
      "memory(GiB)": 77.56,
      "step": 97575,
      "token_acc": 0.5387205387205387,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.180626365622724,
      "grad_norm": 6.181568622589111,
      "learning_rate": 6.481119818416898e-06,
      "loss": 2.221596527099609,
      "memory(GiB)": 77.56,
      "step": 97580,
      "token_acc": 0.5382165605095541,
      "train_speed(iter/s)": 1.438074
    },
    {
      "epoch": 4.180840580951973,
      "grad_norm": 5.924911022186279,
      "learning_rate": 6.477806574419182e-06,
      "loss": 2.24060173034668,
      "memory(GiB)": 77.56,
      "step": 97585,
      "token_acc": 0.49795918367346936,
      "train_speed(iter/s)": 1.438068
    },
    {
      "epoch": 4.181054796281222,
      "grad_norm": 6.827852725982666,
      "learning_rate": 6.474494118866825e-06,
      "loss": 1.9550445556640625,
      "memory(GiB)": 77.56,
      "step": 97590,
      "token_acc": 0.5991189427312775,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 4.181269011610471,
      "grad_norm": 8.712571144104004,
      "learning_rate": 6.471182451819824e-06,
      "loss": 2.5854272842407227,
      "memory(GiB)": 77.56,
      "step": 97595,
      "token_acc": 0.4777777777777778,
      "train_speed(iter/s)": 1.438049
    },
    {
      "epoch": 4.1814832269397195,
      "grad_norm": 10.523072242736816,
      "learning_rate": 6.467871573338186e-06,
      "loss": 2.2612483978271483,
      "memory(GiB)": 77.56,
      "step": 97600,
      "token_acc": 0.5347222222222222,
      "train_speed(iter/s)": 1.438035
    },
    {
      "epoch": 4.181697442268969,
      "grad_norm": 6.188942909240723,
      "learning_rate": 6.464561483481907e-06,
      "loss": 2.011410713195801,
      "memory(GiB)": 77.56,
      "step": 97605,
      "token_acc": 0.5547703180212014,
      "train_speed(iter/s)": 1.438048
    },
    {
      "epoch": 4.181911657598218,
      "grad_norm": 6.513525485992432,
      "learning_rate": 6.461252182310929e-06,
      "loss": 2.52322998046875,
      "memory(GiB)": 77.56,
      "step": 97610,
      "token_acc": 0.49838187702265374,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 4.182125872927466,
      "grad_norm": 5.843857288360596,
      "learning_rate": 6.457943669885219e-06,
      "loss": 1.9907096862792968,
      "memory(GiB)": 77.56,
      "step": 97615,
      "token_acc": 0.5400696864111498,
      "train_speed(iter/s)": 1.438071
    },
    {
      "epoch": 4.182340088256716,
      "grad_norm": 4.699541091918945,
      "learning_rate": 6.4546359462646975e-06,
      "loss": 2.001882553100586,
      "memory(GiB)": 77.56,
      "step": 97620,
      "token_acc": 0.5304659498207885,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.182554303585965,
      "grad_norm": 7.146362781524658,
      "learning_rate": 6.4513290115093e-06,
      "loss": 2.1803197860717773,
      "memory(GiB)": 77.56,
      "step": 97625,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 4.182768518915213,
      "grad_norm": 5.4522881507873535,
      "learning_rate": 6.448022865678916e-06,
      "loss": 2.0480764389038084,
      "memory(GiB)": 77.56,
      "step": 97630,
      "token_acc": 0.553030303030303,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 4.1829827342444625,
      "grad_norm": 7.047962665557861,
      "learning_rate": 6.444717508833464e-06,
      "loss": 2.206417465209961,
      "memory(GiB)": 77.56,
      "step": 97635,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 4.183196949573712,
      "grad_norm": 8.34561538696289,
      "learning_rate": 6.441412941032809e-06,
      "loss": 2.111255073547363,
      "memory(GiB)": 77.56,
      "step": 97640,
      "token_acc": 0.5482456140350878,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 4.18341116490296,
      "grad_norm": 7.390643119812012,
      "learning_rate": 6.4381091623368254e-06,
      "loss": 2.3486318588256836,
      "memory(GiB)": 77.56,
      "step": 97645,
      "token_acc": 0.5035971223021583,
      "train_speed(iter/s)": 1.438083
    },
    {
      "epoch": 4.183625380232209,
      "grad_norm": 7.53248405456543,
      "learning_rate": 6.434806172805358e-06,
      "loss": 2.304783248901367,
      "memory(GiB)": 77.56,
      "step": 97650,
      "token_acc": 0.4612676056338028,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 4.183839595561459,
      "grad_norm": 5.743891716003418,
      "learning_rate": 6.431503972498232e-06,
      "loss": 2.2595176696777344,
      "memory(GiB)": 77.56,
      "step": 97655,
      "token_acc": 0.5032258064516129,
      "train_speed(iter/s)": 1.438106
    },
    {
      "epoch": 4.184053810890707,
      "grad_norm": 6.216049671173096,
      "learning_rate": 6.428202561475288e-06,
      "loss": 2.3658281326293946,
      "memory(GiB)": 77.56,
      "step": 97660,
      "token_acc": 0.5301724137931034,
      "train_speed(iter/s)": 1.438122
    },
    {
      "epoch": 4.184268026219956,
      "grad_norm": 13.739350318908691,
      "learning_rate": 6.424901939796335e-06,
      "loss": 2.116403579711914,
      "memory(GiB)": 77.56,
      "step": 97665,
      "token_acc": 0.53,
      "train_speed(iter/s)": 1.438127
    },
    {
      "epoch": 4.184482241549206,
      "grad_norm": 8.467107772827148,
      "learning_rate": 6.421602107521157e-06,
      "loss": 2.258096694946289,
      "memory(GiB)": 77.56,
      "step": 97670,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.438125
    },
    {
      "epoch": 4.184696456878454,
      "grad_norm": 7.090216159820557,
      "learning_rate": 6.41830306470953e-06,
      "loss": 2.169106864929199,
      "memory(GiB)": 77.56,
      "step": 97675,
      "token_acc": 0.5591397849462365,
      "train_speed(iter/s)": 1.438141
    },
    {
      "epoch": 4.184910672207703,
      "grad_norm": 8.354065895080566,
      "learning_rate": 6.415004811421232e-06,
      "loss": 1.9704107284545898,
      "memory(GiB)": 77.56,
      "step": 97680,
      "token_acc": 0.56640625,
      "train_speed(iter/s)": 1.438139
    },
    {
      "epoch": 4.185124887536952,
      "grad_norm": 5.548110008239746,
      "learning_rate": 6.411707347716012e-06,
      "loss": 2.150985527038574,
      "memory(GiB)": 77.56,
      "step": 97685,
      "token_acc": 0.5508196721311476,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 4.185339102866201,
      "grad_norm": 6.08723783493042,
      "learning_rate": 6.408410673653586e-06,
      "loss": 2.199851417541504,
      "memory(GiB)": 77.56,
      "step": 97690,
      "token_acc": 0.5187713310580204,
      "train_speed(iter/s)": 1.43815
    },
    {
      "epoch": 4.18555331819545,
      "grad_norm": 6.072889804840088,
      "learning_rate": 6.405114789293709e-06,
      "loss": 2.3310268402099608,
      "memory(GiB)": 77.56,
      "step": 97695,
      "token_acc": 0.5204081632653061,
      "train_speed(iter/s)": 1.438156
    },
    {
      "epoch": 4.185767533524699,
      "grad_norm": 6.160730361938477,
      "learning_rate": 6.40181969469607e-06,
      "loss": 2.2882965087890623,
      "memory(GiB)": 77.56,
      "step": 97700,
      "token_acc": 0.4873417721518987,
      "train_speed(iter/s)": 1.438163
    },
    {
      "epoch": 4.185981748853948,
      "grad_norm": 7.00139045715332,
      "learning_rate": 6.398525389920368e-06,
      "loss": 2.1385366439819338,
      "memory(GiB)": 77.56,
      "step": 97705,
      "token_acc": 0.5390334572490706,
      "train_speed(iter/s)": 1.438176
    },
    {
      "epoch": 4.186195964183197,
      "grad_norm": 6.793406963348389,
      "learning_rate": 6.395231875026275e-06,
      "loss": 2.4661527633666993,
      "memory(GiB)": 77.56,
      "step": 97710,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.438172
    },
    {
      "epoch": 4.186410179512446,
      "grad_norm": 8.219132423400879,
      "learning_rate": 6.391939150073456e-06,
      "loss": 2.453174591064453,
      "memory(GiB)": 77.56,
      "step": 97715,
      "token_acc": 0.5115511551155115,
      "train_speed(iter/s)": 1.438166
    },
    {
      "epoch": 4.186624394841695,
      "grad_norm": 7.5466156005859375,
      "learning_rate": 6.388647215121579e-06,
      "loss": 2.2419387817382814,
      "memory(GiB)": 77.56,
      "step": 97720,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.438174
    },
    {
      "epoch": 4.186838610170944,
      "grad_norm": 7.3004326820373535,
      "learning_rate": 6.385356070230264e-06,
      "loss": 2.2508113861083983,
      "memory(GiB)": 77.56,
      "step": 97725,
      "token_acc": 0.47017543859649125,
      "train_speed(iter/s)": 1.438168
    },
    {
      "epoch": 4.187052825500193,
      "grad_norm": 6.568329811096191,
      "learning_rate": 6.38206571545914e-06,
      "loss": 2.0413450241088866,
      "memory(GiB)": 77.56,
      "step": 97730,
      "token_acc": 0.546875,
      "train_speed(iter/s)": 1.438171
    },
    {
      "epoch": 4.1872670408294415,
      "grad_norm": 10.363677978515625,
      "learning_rate": 6.37877615086781e-06,
      "loss": 2.472368621826172,
      "memory(GiB)": 77.56,
      "step": 97735,
      "token_acc": 0.4591549295774648,
      "train_speed(iter/s)": 1.438153
    },
    {
      "epoch": 4.187481256158691,
      "grad_norm": 8.492854118347168,
      "learning_rate": 6.375487376515859e-06,
      "loss": 2.3931474685668945,
      "memory(GiB)": 77.56,
      "step": 97740,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.438161
    },
    {
      "epoch": 4.18769547148794,
      "grad_norm": 5.212835311889648,
      "learning_rate": 6.372199392462891e-06,
      "loss": 1.8894960403442382,
      "memory(GiB)": 77.56,
      "step": 97745,
      "token_acc": 0.5466101694915254,
      "train_speed(iter/s)": 1.438182
    },
    {
      "epoch": 4.187909686817188,
      "grad_norm": 8.418830871582031,
      "learning_rate": 6.368912198768445e-06,
      "loss": 2.198039245605469,
      "memory(GiB)": 77.56,
      "step": 97750,
      "token_acc": 0.4773413897280967,
      "train_speed(iter/s)": 1.438171
    },
    {
      "epoch": 4.188123902146438,
      "grad_norm": 8.315765380859375,
      "learning_rate": 6.365625795492092e-06,
      "loss": 2.1980485916137695,
      "memory(GiB)": 77.56,
      "step": 97755,
      "token_acc": 0.5132075471698113,
      "train_speed(iter/s)": 1.43818
    },
    {
      "epoch": 4.188338117475687,
      "grad_norm": 9.58134651184082,
      "learning_rate": 6.362340182693366e-06,
      "loss": 2.1204824447631836,
      "memory(GiB)": 77.56,
      "step": 97760,
      "token_acc": 0.49264705882352944,
      "train_speed(iter/s)": 1.438172
    },
    {
      "epoch": 4.188552332804935,
      "grad_norm": 5.854559898376465,
      "learning_rate": 6.3590553604317756e-06,
      "loss": 2.447714996337891,
      "memory(GiB)": 77.56,
      "step": 97765,
      "token_acc": 0.5143769968051118,
      "train_speed(iter/s)": 1.438187
    },
    {
      "epoch": 4.1887665481341845,
      "grad_norm": 7.2104363441467285,
      "learning_rate": 6.3557713287668405e-06,
      "loss": 2.108713150024414,
      "memory(GiB)": 77.56,
      "step": 97770,
      "token_acc": 0.5732899022801303,
      "train_speed(iter/s)": 1.438199
    },
    {
      "epoch": 4.188980763463434,
      "grad_norm": 5.27855920791626,
      "learning_rate": 6.3524880877580405e-06,
      "loss": 1.8638744354248047,
      "memory(GiB)": 77.56,
      "step": 97775,
      "token_acc": 0.5415162454873647,
      "train_speed(iter/s)": 1.438193
    },
    {
      "epoch": 4.189194978792682,
      "grad_norm": 6.88019323348999,
      "learning_rate": 6.349205637464872e-06,
      "loss": 2.152066230773926,
      "memory(GiB)": 77.56,
      "step": 97780,
      "token_acc": 0.5234375,
      "train_speed(iter/s)": 1.438186
    },
    {
      "epoch": 4.189409194121931,
      "grad_norm": 6.190378665924072,
      "learning_rate": 6.345923977946789e-06,
      "loss": 2.0639495849609375,
      "memory(GiB)": 77.56,
      "step": 97785,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.438175
    },
    {
      "epoch": 4.189623409451181,
      "grad_norm": 6.240383625030518,
      "learning_rate": 6.3426431092632465e-06,
      "loss": 2.1495569229125975,
      "memory(GiB)": 77.56,
      "step": 97790,
      "token_acc": 0.5364238410596026,
      "train_speed(iter/s)": 1.438173
    },
    {
      "epoch": 4.189837624780429,
      "grad_norm": 7.886990070343018,
      "learning_rate": 6.339363031473677e-06,
      "loss": 2.1694637298583985,
      "memory(GiB)": 77.56,
      "step": 97795,
      "token_acc": 0.4899135446685879,
      "train_speed(iter/s)": 1.438139
    },
    {
      "epoch": 4.190051840109678,
      "grad_norm": 9.617639541625977,
      "learning_rate": 6.3360837446374945e-06,
      "loss": 2.3558122634887697,
      "memory(GiB)": 77.56,
      "step": 97800,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438146
    },
    {
      "epoch": 4.1902660554389275,
      "grad_norm": 5.966063976287842,
      "learning_rate": 6.3328052488141225e-06,
      "loss": 2.286698913574219,
      "memory(GiB)": 77.56,
      "step": 97805,
      "token_acc": 0.4981549815498155,
      "train_speed(iter/s)": 1.438145
    },
    {
      "epoch": 4.190480270768176,
      "grad_norm": 7.710631370544434,
      "learning_rate": 6.329527544062952e-06,
      "loss": 2.4361454010009767,
      "memory(GiB)": 77.56,
      "step": 97810,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.43816
    },
    {
      "epoch": 4.190694486097425,
      "grad_norm": 5.625353813171387,
      "learning_rate": 6.326250630443348e-06,
      "loss": 2.374558448791504,
      "memory(GiB)": 77.56,
      "step": 97815,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.438171
    },
    {
      "epoch": 4.190908701426674,
      "grad_norm": 5.857768535614014,
      "learning_rate": 6.32297450801469e-06,
      "loss": 2.1345306396484376,
      "memory(GiB)": 77.56,
      "step": 97820,
      "token_acc": 0.5030674846625767,
      "train_speed(iter/s)": 1.438184
    },
    {
      "epoch": 4.191122916755923,
      "grad_norm": 5.928982257843018,
      "learning_rate": 6.319699176836308e-06,
      "loss": 2.4232322692871096,
      "memory(GiB)": 77.56,
      "step": 97825,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.4382
    },
    {
      "epoch": 4.191337132085172,
      "grad_norm": 6.280635356903076,
      "learning_rate": 6.316424636967561e-06,
      "loss": 2.127361297607422,
      "memory(GiB)": 77.56,
      "step": 97830,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.438216
    },
    {
      "epoch": 4.191551347414421,
      "grad_norm": 6.911950588226318,
      "learning_rate": 6.313150888467751e-06,
      "loss": 2.2145158767700197,
      "memory(GiB)": 77.56,
      "step": 97835,
      "token_acc": 0.5469798657718121,
      "train_speed(iter/s)": 1.438214
    },
    {
      "epoch": 4.19176556274367,
      "grad_norm": 7.9430623054504395,
      "learning_rate": 6.309877931396202e-06,
      "loss": 2.287306213378906,
      "memory(GiB)": 77.56,
      "step": 97840,
      "token_acc": 0.49642857142857144,
      "train_speed(iter/s)": 1.438232
    },
    {
      "epoch": 4.191979778072919,
      "grad_norm": 7.1939697265625,
      "learning_rate": 6.306605765812202e-06,
      "loss": 2.0288459777832033,
      "memory(GiB)": 77.56,
      "step": 97845,
      "token_acc": 0.5517241379310345,
      "train_speed(iter/s)": 1.438224
    },
    {
      "epoch": 4.192193993402168,
      "grad_norm": 8.208932876586914,
      "learning_rate": 6.3033343917750235e-06,
      "loss": 1.899272346496582,
      "memory(GiB)": 77.56,
      "step": 97850,
      "token_acc": 0.581081081081081,
      "train_speed(iter/s)": 1.438229
    },
    {
      "epoch": 4.1924082087314165,
      "grad_norm": 6.498878479003906,
      "learning_rate": 6.300063809343936e-06,
      "loss": 2.156787109375,
      "memory(GiB)": 77.56,
      "step": 97855,
      "token_acc": 0.5234657039711191,
      "train_speed(iter/s)": 1.438243
    },
    {
      "epoch": 4.192622424060666,
      "grad_norm": 5.411504745483398,
      "learning_rate": 6.2967940185781785e-06,
      "loss": 2.2550283432006837,
      "memory(GiB)": 77.56,
      "step": 97860,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438255
    },
    {
      "epoch": 4.192836639389915,
      "grad_norm": 7.424903392791748,
      "learning_rate": 6.293525019537e-06,
      "loss": 2.3128299713134766,
      "memory(GiB)": 77.56,
      "step": 97865,
      "token_acc": 0.5077881619937694,
      "train_speed(iter/s)": 1.438269
    },
    {
      "epoch": 4.193050854719163,
      "grad_norm": 5.775293827056885,
      "learning_rate": 6.290256812279616e-06,
      "loss": 2.1359560012817385,
      "memory(GiB)": 77.56,
      "step": 97870,
      "token_acc": 0.4944237918215613,
      "train_speed(iter/s)": 1.438273
    },
    {
      "epoch": 4.193265070048413,
      "grad_norm": 6.167384624481201,
      "learning_rate": 6.286989396865234e-06,
      "loss": 2.305561065673828,
      "memory(GiB)": 77.56,
      "step": 97875,
      "token_acc": 0.5290102389078498,
      "train_speed(iter/s)": 1.438273
    },
    {
      "epoch": 4.193479285377662,
      "grad_norm": 4.649319648742676,
      "learning_rate": 6.283722773353046e-06,
      "loss": 2.201676368713379,
      "memory(GiB)": 77.56,
      "step": 97880,
      "token_acc": 0.5212121212121212,
      "train_speed(iter/s)": 1.438269
    },
    {
      "epoch": 4.19369350070691,
      "grad_norm": 7.607011795043945,
      "learning_rate": 6.280456941802215e-06,
      "loss": 2.393903923034668,
      "memory(GiB)": 77.56,
      "step": 97885,
      "token_acc": 0.5321100917431193,
      "train_speed(iter/s)": 1.438282
    },
    {
      "epoch": 4.1939077160361595,
      "grad_norm": 6.558903217315674,
      "learning_rate": 6.277191902271934e-06,
      "loss": 2.205317497253418,
      "memory(GiB)": 77.56,
      "step": 97890,
      "token_acc": 0.5387596899224806,
      "train_speed(iter/s)": 1.438305
    },
    {
      "epoch": 4.194121931365409,
      "grad_norm": 5.9045796394348145,
      "learning_rate": 6.273927654821321e-06,
      "loss": 2.454580307006836,
      "memory(GiB)": 77.56,
      "step": 97895,
      "token_acc": 0.44936708860759494,
      "train_speed(iter/s)": 1.43832
    },
    {
      "epoch": 4.194336146694657,
      "grad_norm": 6.074865341186523,
      "learning_rate": 6.2706641995095405e-06,
      "loss": 2.208053207397461,
      "memory(GiB)": 77.56,
      "step": 97900,
      "token_acc": 0.5136186770428015,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 4.194550362023906,
      "grad_norm": 6.615030288696289,
      "learning_rate": 6.267401536395701e-06,
      "loss": 2.2229873657226564,
      "memory(GiB)": 77.56,
      "step": 97905,
      "token_acc": 0.5093632958801498,
      "train_speed(iter/s)": 1.438318
    },
    {
      "epoch": 4.194764577353156,
      "grad_norm": 5.493546009063721,
      "learning_rate": 6.2641396655389095e-06,
      "loss": 2.2882667541503907,
      "memory(GiB)": 77.56,
      "step": 97910,
      "token_acc": 0.5252225519287834,
      "train_speed(iter/s)": 1.438315
    },
    {
      "epoch": 4.194978792682404,
      "grad_norm": 5.833301544189453,
      "learning_rate": 6.2608785869982475e-06,
      "loss": 2.1958927154541015,
      "memory(GiB)": 77.56,
      "step": 97915,
      "token_acc": 0.4910394265232975,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 4.195193008011653,
      "grad_norm": 5.743492126464844,
      "learning_rate": 6.257618300832796e-06,
      "loss": 2.340970993041992,
      "memory(GiB)": 77.56,
      "step": 97920,
      "token_acc": 0.5229681978798587,
      "train_speed(iter/s)": 1.438321
    },
    {
      "epoch": 4.1954072233409025,
      "grad_norm": 6.084545612335205,
      "learning_rate": 6.254358807101635e-06,
      "loss": 2.1390939712524415,
      "memory(GiB)": 77.56,
      "step": 97925,
      "token_acc": 0.545774647887324,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 4.195621438670151,
      "grad_norm": 6.5965352058410645,
      "learning_rate": 6.251100105863794e-06,
      "loss": 2.244105339050293,
      "memory(GiB)": 77.56,
      "step": 97930,
      "token_acc": 0.5547445255474452,
      "train_speed(iter/s)": 1.438319
    },
    {
      "epoch": 4.1958356539994,
      "grad_norm": 5.935375690460205,
      "learning_rate": 6.247842197178317e-06,
      "loss": 2.026915740966797,
      "memory(GiB)": 77.56,
      "step": 97935,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.438295
    },
    {
      "epoch": 4.196049869328649,
      "grad_norm": 5.711519241333008,
      "learning_rate": 6.2445850811042264e-06,
      "loss": 2.078433609008789,
      "memory(GiB)": 77.56,
      "step": 97940,
      "token_acc": 0.5261044176706827,
      "train_speed(iter/s)": 1.438309
    },
    {
      "epoch": 4.196264084657898,
      "grad_norm": 5.879454135894775,
      "learning_rate": 6.241328757700505e-06,
      "loss": 2.2151498794555664,
      "memory(GiB)": 77.56,
      "step": 97945,
      "token_acc": 0.5399239543726235,
      "train_speed(iter/s)": 1.438329
    },
    {
      "epoch": 4.196478299987147,
      "grad_norm": 5.962684154510498,
      "learning_rate": 6.238073227026176e-06,
      "loss": 2.297053909301758,
      "memory(GiB)": 77.56,
      "step": 97950,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438335
    },
    {
      "epoch": 4.196692515316396,
      "grad_norm": 6.07538366317749,
      "learning_rate": 6.234818489140204e-06,
      "loss": 2.4884803771972654,
      "memory(GiB)": 77.56,
      "step": 97955,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.438328
    },
    {
      "epoch": 4.196906730645645,
      "grad_norm": 7.4774932861328125,
      "learning_rate": 6.231564544101548e-06,
      "loss": 2.5031810760498048,
      "memory(GiB)": 77.56,
      "step": 97960,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438321
    },
    {
      "epoch": 4.197120945974894,
      "grad_norm": 5.270471096038818,
      "learning_rate": 6.22831139196916e-06,
      "loss": 2.3547115325927734,
      "memory(GiB)": 77.56,
      "step": 97965,
      "token_acc": 0.45751633986928103,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 4.197335161304143,
      "grad_norm": 7.197223663330078,
      "learning_rate": 6.225059032801961e-06,
      "loss": 2.2638641357421876,
      "memory(GiB)": 77.56,
      "step": 97970,
      "token_acc": 0.5364431486880467,
      "train_speed(iter/s)": 1.438324
    },
    {
      "epoch": 4.197549376633392,
      "grad_norm": 5.827846050262451,
      "learning_rate": 6.221807466658891e-06,
      "loss": 2.5818321228027346,
      "memory(GiB)": 77.56,
      "step": 97975,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.438332
    },
    {
      "epoch": 4.197763591962641,
      "grad_norm": 7.089547157287598,
      "learning_rate": 6.2185566935988375e-06,
      "loss": 1.9585094451904297,
      "memory(GiB)": 77.56,
      "step": 97980,
      "token_acc": 0.5866141732283464,
      "train_speed(iter/s)": 1.43833
    },
    {
      "epoch": 4.19797780729189,
      "grad_norm": 6.133625030517578,
      "learning_rate": 6.2153067136807076e-06,
      "loss": 2.40917854309082,
      "memory(GiB)": 77.56,
      "step": 97985,
      "token_acc": 0.4876543209876543,
      "train_speed(iter/s)": 1.438353
    },
    {
      "epoch": 4.198192022621138,
      "grad_norm": 5.699099063873291,
      "learning_rate": 6.212057526963372e-06,
      "loss": 2.2886009216308594,
      "memory(GiB)": 77.56,
      "step": 97990,
      "token_acc": 0.4887640449438202,
      "train_speed(iter/s)": 1.438359
    },
    {
      "epoch": 4.198406237950388,
      "grad_norm": 5.499150276184082,
      "learning_rate": 6.2088091335056945e-06,
      "loss": 2.3925298690795898,
      "memory(GiB)": 77.56,
      "step": 97995,
      "token_acc": 0.5226586102719033,
      "train_speed(iter/s)": 1.438378
    },
    {
      "epoch": 4.198620453279637,
      "grad_norm": 5.269199848175049,
      "learning_rate": 6.205561533366511e-06,
      "loss": 2.0267637252807615,
      "memory(GiB)": 77.56,
      "step": 98000,
      "token_acc": 0.551829268292683,
      "train_speed(iter/s)": 1.43839
    },
    {
      "epoch": 4.198620453279637,
      "eval_loss": 2.1447606086730957,
      "eval_runtime": 13.7184,
      "eval_samples_per_second": 7.289,
      "eval_steps_per_second": 7.289,
      "eval_token_acc": 0.5054945054945055,
      "step": 98000
    },
    {
      "epoch": 4.198834668608885,
      "grad_norm": 6.185956001281738,
      "learning_rate": 6.202314726604658e-06,
      "loss": 2.3849361419677733,
      "memory(GiB)": 77.56,
      "step": 98005,
      "token_acc": 0.5071907957813998,
      "train_speed(iter/s)": 1.438086
    },
    {
      "epoch": 4.199048883938135,
      "grad_norm": 7.049850940704346,
      "learning_rate": 6.199068713278966e-06,
      "loss": 2.6123046875,
      "memory(GiB)": 77.56,
      "step": 98010,
      "token_acc": 0.4548736462093863,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.199263099267384,
      "grad_norm": 6.718698024749756,
      "learning_rate": 6.1958234934482384e-06,
      "loss": 2.262205123901367,
      "memory(GiB)": 77.56,
      "step": 98015,
      "token_acc": 0.5111111111111111,
      "train_speed(iter/s)": 1.438081
    },
    {
      "epoch": 4.199477314596632,
      "grad_norm": 6.537539958953857,
      "learning_rate": 6.192579067171256e-06,
      "loss": 2.694133758544922,
      "memory(GiB)": 77.56,
      "step": 98020,
      "token_acc": 0.45733788395904434,
      "train_speed(iter/s)": 1.438085
    },
    {
      "epoch": 4.1996915299258815,
      "grad_norm": 5.617667198181152,
      "learning_rate": 6.189335434506799e-06,
      "loss": 2.2067539215087892,
      "memory(GiB)": 77.56,
      "step": 98025,
      "token_acc": 0.5698924731182796,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 4.199905745255131,
      "grad_norm": 9.685507774353027,
      "learning_rate": 6.186092595513615e-06,
      "loss": 2.10714054107666,
      "memory(GiB)": 77.56,
      "step": 98030,
      "token_acc": 0.5547445255474452,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 4.200119960584379,
      "grad_norm": 9.102299690246582,
      "learning_rate": 6.18285055025048e-06,
      "loss": 2.2809188842773436,
      "memory(GiB)": 77.56,
      "step": 98035,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.438114
    },
    {
      "epoch": 4.200334175913628,
      "grad_norm": 5.823385715484619,
      "learning_rate": 6.179609298776106e-06,
      "loss": 2.501713180541992,
      "memory(GiB)": 77.56,
      "step": 98040,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 4.200548391242878,
      "grad_norm": 6.058045864105225,
      "learning_rate": 6.176368841149211e-06,
      "loss": 2.2380088806152343,
      "memory(GiB)": 77.56,
      "step": 98045,
      "token_acc": 0.5564516129032258,
      "train_speed(iter/s)": 1.438108
    },
    {
      "epoch": 4.200762606572126,
      "grad_norm": 6.595065593719482,
      "learning_rate": 6.17312917742851e-06,
      "loss": 2.140004348754883,
      "memory(GiB)": 77.56,
      "step": 98050,
      "token_acc": 0.5681818181818182,
      "train_speed(iter/s)": 1.438093
    },
    {
      "epoch": 4.200976821901375,
      "grad_norm": 7.752303123474121,
      "learning_rate": 6.169890307672693e-06,
      "loss": 2.2860483169555663,
      "memory(GiB)": 77.56,
      "step": 98055,
      "token_acc": 0.4826254826254826,
      "train_speed(iter/s)": 1.438106
    },
    {
      "epoch": 4.2011910372306245,
      "grad_norm": 7.3059844970703125,
      "learning_rate": 6.166652231940423e-06,
      "loss": 2.3380111694335937,
      "memory(GiB)": 77.56,
      "step": 98060,
      "token_acc": 0.49836065573770494,
      "train_speed(iter/s)": 1.438126
    },
    {
      "epoch": 4.201405252559873,
      "grad_norm": 6.263556957244873,
      "learning_rate": 6.163414950290358e-06,
      "loss": 2.2195613861083983,
      "memory(GiB)": 77.56,
      "step": 98065,
      "token_acc": 0.5288135593220339,
      "train_speed(iter/s)": 1.438136
    },
    {
      "epoch": 4.201619467889122,
      "grad_norm": 7.176837921142578,
      "learning_rate": 6.1601784627811684e-06,
      "loss": 2.513922691345215,
      "memory(GiB)": 77.56,
      "step": 98070,
      "token_acc": 0.48175182481751827,
      "train_speed(iter/s)": 1.43815
    },
    {
      "epoch": 4.201833683218371,
      "grad_norm": 5.722093105316162,
      "learning_rate": 6.1569427694714635e-06,
      "loss": 2.1636272430419923,
      "memory(GiB)": 77.56,
      "step": 98075,
      "token_acc": 0.513986013986014,
      "train_speed(iter/s)": 1.438152
    },
    {
      "epoch": 4.20204789854762,
      "grad_norm": 5.712563991546631,
      "learning_rate": 6.1537078704198726e-06,
      "loss": 2.490376663208008,
      "memory(GiB)": 77.56,
      "step": 98080,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.43815
    },
    {
      "epoch": 4.202262113876869,
      "grad_norm": 6.3566575050354,
      "learning_rate": 6.150473765684994e-06,
      "loss": 2.4727546691894533,
      "memory(GiB)": 77.56,
      "step": 98085,
      "token_acc": 0.49175824175824173,
      "train_speed(iter/s)": 1.438159
    },
    {
      "epoch": 4.202476329206118,
      "grad_norm": 5.163562297821045,
      "learning_rate": 6.147240455325409e-06,
      "loss": 2.1134862899780273,
      "memory(GiB)": 77.56,
      "step": 98090,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.43818
    },
    {
      "epoch": 4.202690544535367,
      "grad_norm": 7.16831636428833,
      "learning_rate": 6.14400793939971e-06,
      "loss": 2.202798843383789,
      "memory(GiB)": 77.56,
      "step": 98095,
      "token_acc": 0.512,
      "train_speed(iter/s)": 1.438183
    },
    {
      "epoch": 4.202904759864616,
      "grad_norm": 7.538551330566406,
      "learning_rate": 6.140776217966443e-06,
      "loss": 2.4330799102783205,
      "memory(GiB)": 77.56,
      "step": 98100,
      "token_acc": 0.49044585987261147,
      "train_speed(iter/s)": 1.438191
    },
    {
      "epoch": 4.203118975193865,
      "grad_norm": 5.994205951690674,
      "learning_rate": 6.137545291084162e-06,
      "loss": 2.0087726593017576,
      "memory(GiB)": 77.56,
      "step": 98105,
      "token_acc": 0.5480769230769231,
      "train_speed(iter/s)": 1.438214
    },
    {
      "epoch": 4.2033331905231135,
      "grad_norm": 5.978413105010986,
      "learning_rate": 6.134315158811393e-06,
      "loss": 2.176012420654297,
      "memory(GiB)": 77.56,
      "step": 98110,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.438241
    },
    {
      "epoch": 4.203547405852363,
      "grad_norm": 6.383014678955078,
      "learning_rate": 6.1310858212066445e-06,
      "loss": 2.2739749908447267,
      "memory(GiB)": 77.56,
      "step": 98115,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 4.203761621181612,
      "grad_norm": 5.5930633544921875,
      "learning_rate": 6.127857278328442e-06,
      "loss": 2.281202697753906,
      "memory(GiB)": 77.56,
      "step": 98120,
      "token_acc": 0.5434083601286174,
      "train_speed(iter/s)": 1.43827
    },
    {
      "epoch": 4.20397583651086,
      "grad_norm": 5.5748772621154785,
      "learning_rate": 6.124629530235249e-06,
      "loss": 2.2819797515869142,
      "memory(GiB)": 77.56,
      "step": 98125,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.438267
    },
    {
      "epoch": 4.20419005184011,
      "grad_norm": 5.827408790588379,
      "learning_rate": 6.121402576985558e-06,
      "loss": 2.1849925994873045,
      "memory(GiB)": 77.56,
      "step": 98130,
      "token_acc": 0.5181818181818182,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 4.204404267169359,
      "grad_norm": 7.5553202629089355,
      "learning_rate": 6.118176418637822e-06,
      "loss": 2.213795471191406,
      "memory(GiB)": 77.56,
      "step": 98135,
      "token_acc": 0.5303643724696356,
      "train_speed(iter/s)": 1.438262
    },
    {
      "epoch": 4.204618482498607,
      "grad_norm": 5.586918354034424,
      "learning_rate": 6.114951055250484e-06,
      "loss": 2.4456783294677735,
      "memory(GiB)": 77.56,
      "step": 98140,
      "token_acc": 0.47580645161290325,
      "train_speed(iter/s)": 1.438272
    },
    {
      "epoch": 4.2048326978278565,
      "grad_norm": 5.2323899269104,
      "learning_rate": 6.111726486881975e-06,
      "loss": 2.0326816558837892,
      "memory(GiB)": 77.56,
      "step": 98145,
      "token_acc": 0.5570469798657718,
      "train_speed(iter/s)": 1.438278
    },
    {
      "epoch": 4.205046913157106,
      "grad_norm": 5.317123889923096,
      "learning_rate": 6.108502713590702e-06,
      "loss": 2.2312458038330076,
      "memory(GiB)": 77.56,
      "step": 98150,
      "token_acc": 0.4935483870967742,
      "train_speed(iter/s)": 1.438282
    },
    {
      "epoch": 4.205261128486354,
      "grad_norm": 6.194159507751465,
      "learning_rate": 6.105279735435082e-06,
      "loss": 2.4072303771972656,
      "memory(GiB)": 77.56,
      "step": 98155,
      "token_acc": 0.47115384615384615,
      "train_speed(iter/s)": 1.438296
    },
    {
      "epoch": 4.205475343815603,
      "grad_norm": 7.651185989379883,
      "learning_rate": 6.1020575524735005e-06,
      "loss": 2.192337989807129,
      "memory(GiB)": 77.56,
      "step": 98160,
      "token_acc": 0.5698529411764706,
      "train_speed(iter/s)": 1.438309
    },
    {
      "epoch": 4.205689559144853,
      "grad_norm": 6.0030717849731445,
      "learning_rate": 6.098836164764326e-06,
      "loss": 2.5083805084228517,
      "memory(GiB)": 77.56,
      "step": 98165,
      "token_acc": 0.5068027210884354,
      "train_speed(iter/s)": 1.438313
    },
    {
      "epoch": 4.205903774474101,
      "grad_norm": 5.139372825622559,
      "learning_rate": 6.095615572365921e-06,
      "loss": 2.4234233856201173,
      "memory(GiB)": 77.56,
      "step": 98170,
      "token_acc": 0.5299401197604791,
      "train_speed(iter/s)": 1.438324
    },
    {
      "epoch": 4.20611798980335,
      "grad_norm": 5.935065269470215,
      "learning_rate": 6.092395775336607e-06,
      "loss": 2.074169158935547,
      "memory(GiB)": 77.56,
      "step": 98175,
      "token_acc": 0.5476190476190477,
      "train_speed(iter/s)": 1.438337
    },
    {
      "epoch": 4.2063322051325995,
      "grad_norm": 6.742197036743164,
      "learning_rate": 6.089176773734745e-06,
      "loss": 2.3369895935058596,
      "memory(GiB)": 77.56,
      "step": 98180,
      "token_acc": 0.5152542372881356,
      "train_speed(iter/s)": 1.438345
    },
    {
      "epoch": 4.206546420461848,
      "grad_norm": 7.1738457679748535,
      "learning_rate": 6.085958567618638e-06,
      "loss": 2.0658105850219726,
      "memory(GiB)": 77.56,
      "step": 98185,
      "token_acc": 0.50199203187251,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 4.206760635791097,
      "grad_norm": 6.9448652267456055,
      "learning_rate": 6.08274115704659e-06,
      "loss": 2.035039520263672,
      "memory(GiB)": 77.56,
      "step": 98190,
      "token_acc": 0.5287356321839081,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 4.206974851120346,
      "grad_norm": 5.532359600067139,
      "learning_rate": 6.07952454207687e-06,
      "loss": 2.4549732208251953,
      "memory(GiB)": 77.56,
      "step": 98195,
      "token_acc": 0.5297297297297298,
      "train_speed(iter/s)": 1.438353
    },
    {
      "epoch": 4.207189066449595,
      "grad_norm": 8.086493492126465,
      "learning_rate": 6.076308722767776e-06,
      "loss": 2.3581161499023438,
      "memory(GiB)": 77.56,
      "step": 98200,
      "token_acc": 0.5197132616487455,
      "train_speed(iter/s)": 1.438359
    },
    {
      "epoch": 4.207403281778844,
      "grad_norm": 7.090773582458496,
      "learning_rate": 6.073093699177551e-06,
      "loss": 2.1348033905029298,
      "memory(GiB)": 77.56,
      "step": 98205,
      "token_acc": 0.5469798657718121,
      "train_speed(iter/s)": 1.438346
    },
    {
      "epoch": 4.207617497108093,
      "grad_norm": 6.109488487243652,
      "learning_rate": 6.069879471364426e-06,
      "loss": 2.171565818786621,
      "memory(GiB)": 77.56,
      "step": 98210,
      "token_acc": 0.583969465648855,
      "train_speed(iter/s)": 1.43834
    },
    {
      "epoch": 4.207831712437342,
      "grad_norm": 5.301544189453125,
      "learning_rate": 6.0666660393866596e-06,
      "loss": 2.2061752319335937,
      "memory(GiB)": 77.56,
      "step": 98215,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438338
    },
    {
      "epoch": 4.208045927766591,
      "grad_norm": 4.8507795333862305,
      "learning_rate": 6.063453403302449e-06,
      "loss": 2.079071044921875,
      "memory(GiB)": 77.56,
      "step": 98220,
      "token_acc": 0.5472972972972973,
      "train_speed(iter/s)": 1.438344
    },
    {
      "epoch": 4.20826014309584,
      "grad_norm": 4.961392879486084,
      "learning_rate": 6.06024156317e-06,
      "loss": 2.2174943923950194,
      "memory(GiB)": 77.56,
      "step": 98225,
      "token_acc": 0.5168918918918919,
      "train_speed(iter/s)": 1.438343
    },
    {
      "epoch": 4.2084743584250885,
      "grad_norm": 5.052983283996582,
      "learning_rate": 6.057030519047491e-06,
      "loss": 2.2127986907958985,
      "memory(GiB)": 77.56,
      "step": 98230,
      "token_acc": 0.5046439628482973,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 4.208688573754338,
      "grad_norm": 5.836075305938721,
      "learning_rate": 6.053820270993082e-06,
      "loss": 2.2800180435180666,
      "memory(GiB)": 77.56,
      "step": 98235,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.438351
    },
    {
      "epoch": 4.208902789083587,
      "grad_norm": 6.770490646362305,
      "learning_rate": 6.050610819064961e-06,
      "loss": 2.3001054763793944,
      "memory(GiB)": 77.56,
      "step": 98240,
      "token_acc": 0.5148936170212766,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 4.209117004412835,
      "grad_norm": 8.350386619567871,
      "learning_rate": 6.047402163321248e-06,
      "loss": 2.344517135620117,
      "memory(GiB)": 77.56,
      "step": 98245,
      "token_acc": 0.4746376811594203,
      "train_speed(iter/s)": 1.438367
    },
    {
      "epoch": 4.209331219742085,
      "grad_norm": 6.6530985832214355,
      "learning_rate": 6.044194303820078e-06,
      "loss": 2.306108856201172,
      "memory(GiB)": 77.56,
      "step": 98250,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.438371
    },
    {
      "epoch": 4.209545435071334,
      "grad_norm": 7.232362747192383,
      "learning_rate": 6.040987240619561e-06,
      "loss": 2.0119325637817385,
      "memory(GiB)": 77.56,
      "step": 98255,
      "token_acc": 0.5643153526970954,
      "train_speed(iter/s)": 1.438369
    },
    {
      "epoch": 4.209759650400582,
      "grad_norm": 5.363946437835693,
      "learning_rate": 6.037780973777785e-06,
      "loss": 2.3746849060058595,
      "memory(GiB)": 77.56,
      "step": 98260,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438365
    },
    {
      "epoch": 4.209973865729832,
      "grad_norm": 5.251760959625244,
      "learning_rate": 6.0345755033528514e-06,
      "loss": 1.975484848022461,
      "memory(GiB)": 77.56,
      "step": 98265,
      "token_acc": 0.6,
      "train_speed(iter/s)": 1.438374
    },
    {
      "epoch": 4.210188081059081,
      "grad_norm": 7.701329708099365,
      "learning_rate": 6.031370829402832e-06,
      "loss": 2.2395561218261717,
      "memory(GiB)": 77.56,
      "step": 98270,
      "token_acc": 0.5128205128205128,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 4.210402296388329,
      "grad_norm": 5.9888691902160645,
      "learning_rate": 6.0281669519857755e-06,
      "loss": 2.4060972213745115,
      "memory(GiB)": 77.56,
      "step": 98275,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.438378
    },
    {
      "epoch": 4.2106165117175784,
      "grad_norm": 5.402688980102539,
      "learning_rate": 6.024963871159722e-06,
      "loss": 2.3136241912841795,
      "memory(GiB)": 77.56,
      "step": 98280,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.438394
    },
    {
      "epoch": 4.210830727046828,
      "grad_norm": 7.225913047790527,
      "learning_rate": 6.021761586982705e-06,
      "loss": 2.4117456436157227,
      "memory(GiB)": 77.56,
      "step": 98285,
      "token_acc": 0.45627376425855515,
      "train_speed(iter/s)": 1.438403
    },
    {
      "epoch": 4.211044942376076,
      "grad_norm": 8.282779693603516,
      "learning_rate": 6.018560099512732e-06,
      "loss": 2.504245376586914,
      "memory(GiB)": 77.56,
      "step": 98290,
      "token_acc": 0.4865771812080537,
      "train_speed(iter/s)": 1.438405
    },
    {
      "epoch": 4.211259157705325,
      "grad_norm": 5.624424934387207,
      "learning_rate": 6.015359408807786e-06,
      "loss": 2.26424560546875,
      "memory(GiB)": 77.56,
      "step": 98295,
      "token_acc": 0.5101351351351351,
      "train_speed(iter/s)": 1.438421
    },
    {
      "epoch": 4.211473373034575,
      "grad_norm": 5.580746650695801,
      "learning_rate": 6.012159514925875e-06,
      "loss": 2.319450569152832,
      "memory(GiB)": 77.56,
      "step": 98300,
      "token_acc": 0.5431309904153354,
      "train_speed(iter/s)": 1.438429
    },
    {
      "epoch": 4.211687588363823,
      "grad_norm": 8.379209518432617,
      "learning_rate": 6.008960417924964e-06,
      "loss": 2.3237228393554688,
      "memory(GiB)": 77.56,
      "step": 98305,
      "token_acc": 0.466403162055336,
      "train_speed(iter/s)": 1.438434
    },
    {
      "epoch": 4.211901803693072,
      "grad_norm": 6.290983200073242,
      "learning_rate": 6.005762117862995e-06,
      "loss": 2.3621248245239257,
      "memory(GiB)": 77.56,
      "step": 98310,
      "token_acc": 0.4982078853046595,
      "train_speed(iter/s)": 1.438429
    },
    {
      "epoch": 4.2121160190223215,
      "grad_norm": 7.093401908874512,
      "learning_rate": 6.002564614797923e-06,
      "loss": 2.3215509414672852,
      "memory(GiB)": 77.56,
      "step": 98315,
      "token_acc": 0.483739837398374,
      "train_speed(iter/s)": 1.438416
    },
    {
      "epoch": 4.21233023435157,
      "grad_norm": 8.208555221557617,
      "learning_rate": 5.999367908787651e-06,
      "loss": 2.142250633239746,
      "memory(GiB)": 77.56,
      "step": 98320,
      "token_acc": 0.5368421052631579,
      "train_speed(iter/s)": 1.438417
    },
    {
      "epoch": 4.212544449680819,
      "grad_norm": 5.318323135375977,
      "learning_rate": 5.996171999890116e-06,
      "loss": 2.1883272171020507,
      "memory(GiB)": 77.56,
      "step": 98325,
      "token_acc": 0.55,
      "train_speed(iter/s)": 1.438418
    },
    {
      "epoch": 4.212758665010068,
      "grad_norm": 7.310027599334717,
      "learning_rate": 5.992976888163204e-06,
      "loss": 1.9704902648925782,
      "memory(GiB)": 77.56,
      "step": 98330,
      "token_acc": 0.6059322033898306,
      "train_speed(iter/s)": 1.438433
    },
    {
      "epoch": 4.212972880339317,
      "grad_norm": 6.041313171386719,
      "learning_rate": 5.989782573664798e-06,
      "loss": 2.436429977416992,
      "memory(GiB)": 77.56,
      "step": 98335,
      "token_acc": 0.47416413373860183,
      "train_speed(iter/s)": 1.438408
    },
    {
      "epoch": 4.213187095668566,
      "grad_norm": 5.936883926391602,
      "learning_rate": 5.9865890564527515e-06,
      "loss": 2.4405813217163086,
      "memory(GiB)": 77.56,
      "step": 98340,
      "token_acc": 0.5053003533568905,
      "train_speed(iter/s)": 1.4384
    },
    {
      "epoch": 4.213401310997815,
      "grad_norm": 6.204477787017822,
      "learning_rate": 5.983396336584945e-06,
      "loss": 2.279122734069824,
      "memory(GiB)": 77.56,
      "step": 98345,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.4384
    },
    {
      "epoch": 4.213615526327064,
      "grad_norm": 9.50565242767334,
      "learning_rate": 5.9802044141192056e-06,
      "loss": 2.4103214263916017,
      "memory(GiB)": 77.56,
      "step": 98350,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.438393
    },
    {
      "epoch": 4.213829741656313,
      "grad_norm": 6.365574359893799,
      "learning_rate": 5.977013289113348e-06,
      "loss": 2.070890426635742,
      "memory(GiB)": 77.56,
      "step": 98355,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.438387
    },
    {
      "epoch": 4.214043956985562,
      "grad_norm": 5.899397850036621,
      "learning_rate": 5.973822961625203e-06,
      "loss": 2.256831932067871,
      "memory(GiB)": 77.56,
      "step": 98360,
      "token_acc": 0.4742647058823529,
      "train_speed(iter/s)": 1.438397
    },
    {
      "epoch": 4.2142581723148105,
      "grad_norm": 7.603686809539795,
      "learning_rate": 5.970633431712552e-06,
      "loss": 2.135301971435547,
      "memory(GiB)": 77.56,
      "step": 98365,
      "token_acc": 0.5206611570247934,
      "train_speed(iter/s)": 1.438404
    },
    {
      "epoch": 4.21447238764406,
      "grad_norm": 7.04205846786499,
      "learning_rate": 5.967444699433183e-06,
      "loss": 2.2547868728637694,
      "memory(GiB)": 77.56,
      "step": 98370,
      "token_acc": 0.549618320610687,
      "train_speed(iter/s)": 1.438402
    },
    {
      "epoch": 4.214686602973309,
      "grad_norm": 9.290288925170898,
      "learning_rate": 5.964256764844855e-06,
      "loss": 2.0210039138793947,
      "memory(GiB)": 77.56,
      "step": 98375,
      "token_acc": 0.5750798722044729,
      "train_speed(iter/s)": 1.438395
    },
    {
      "epoch": 4.214900818302557,
      "grad_norm": 6.7043046951293945,
      "learning_rate": 5.961069628005317e-06,
      "loss": 2.5747379302978515,
      "memory(GiB)": 77.56,
      "step": 98380,
      "token_acc": 0.46078431372549017,
      "train_speed(iter/s)": 1.438374
    },
    {
      "epoch": 4.215115033631807,
      "grad_norm": 7.077681064605713,
      "learning_rate": 5.9578832889723215e-06,
      "loss": 2.150142860412598,
      "memory(GiB)": 77.56,
      "step": 98385,
      "token_acc": 0.5304054054054054,
      "train_speed(iter/s)": 1.438376
    },
    {
      "epoch": 4.215329248961056,
      "grad_norm": 10.056730270385742,
      "learning_rate": 5.954697747803584e-06,
      "loss": 2.591453742980957,
      "memory(GiB)": 77.56,
      "step": 98390,
      "token_acc": 0.47335423197492166,
      "train_speed(iter/s)": 1.438368
    },
    {
      "epoch": 4.215543464290304,
      "grad_norm": 8.434317588806152,
      "learning_rate": 5.951513004556819e-06,
      "loss": 2.3313249588012694,
      "memory(GiB)": 77.56,
      "step": 98395,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.438365
    },
    {
      "epoch": 4.2157576796195535,
      "grad_norm": 6.500161170959473,
      "learning_rate": 5.948329059289715e-06,
      "loss": 2.2964502334594727,
      "memory(GiB)": 77.56,
      "step": 98400,
      "token_acc": 0.5193798449612403,
      "train_speed(iter/s)": 1.438383
    },
    {
      "epoch": 4.215971894948803,
      "grad_norm": 5.685385227203369,
      "learning_rate": 5.945145912059946e-06,
      "loss": 2.4756072998046874,
      "memory(GiB)": 77.56,
      "step": 98405,
      "token_acc": 0.4525316455696203,
      "train_speed(iter/s)": 1.438395
    },
    {
      "epoch": 4.216186110278051,
      "grad_norm": 4.890557765960693,
      "learning_rate": 5.94196356292519e-06,
      "loss": 2.0542274475097657,
      "memory(GiB)": 77.56,
      "step": 98410,
      "token_acc": 0.5223880597014925,
      "train_speed(iter/s)": 1.438416
    },
    {
      "epoch": 4.2164003256073,
      "grad_norm": 6.49935245513916,
      "learning_rate": 5.938782011943089e-06,
      "loss": 2.075295066833496,
      "memory(GiB)": 77.56,
      "step": 98415,
      "token_acc": 0.5219123505976095,
      "train_speed(iter/s)": 1.43843
    },
    {
      "epoch": 4.21661454093655,
      "grad_norm": 7.217024803161621,
      "learning_rate": 5.935601259171292e-06,
      "loss": 2.615322303771973,
      "memory(GiB)": 77.56,
      "step": 98420,
      "token_acc": 0.4297082228116711,
      "train_speed(iter/s)": 1.438441
    },
    {
      "epoch": 4.216828756265798,
      "grad_norm": 5.824163913726807,
      "learning_rate": 5.932421304667418e-06,
      "loss": 2.466215896606445,
      "memory(GiB)": 77.56,
      "step": 98425,
      "token_acc": 0.459546925566343,
      "train_speed(iter/s)": 1.438423
    },
    {
      "epoch": 4.217042971595047,
      "grad_norm": 8.371705055236816,
      "learning_rate": 5.9292421484890674e-06,
      "loss": 2.3368967056274412,
      "memory(GiB)": 77.56,
      "step": 98430,
      "token_acc": 0.49404761904761907,
      "train_speed(iter/s)": 1.438412
    },
    {
      "epoch": 4.2172571869242965,
      "grad_norm": 5.895653247833252,
      "learning_rate": 5.926063790693837e-06,
      "loss": 2.2445981979370115,
      "memory(GiB)": 77.56,
      "step": 98435,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.438424
    },
    {
      "epoch": 4.217471402253545,
      "grad_norm": 8.02182388305664,
      "learning_rate": 5.922886231339297e-06,
      "loss": 1.728099250793457,
      "memory(GiB)": 77.56,
      "step": 98440,
      "token_acc": 0.6163265306122448,
      "train_speed(iter/s)": 1.438419
    },
    {
      "epoch": 4.217685617582794,
      "grad_norm": 5.7319135665893555,
      "learning_rate": 5.919709470483032e-06,
      "loss": 2.115361785888672,
      "memory(GiB)": 77.56,
      "step": 98445,
      "token_acc": 0.5150602409638554,
      "train_speed(iter/s)": 1.438422
    },
    {
      "epoch": 4.217899832912043,
      "grad_norm": 5.47056245803833,
      "learning_rate": 5.9165335081825754e-06,
      "loss": 2.3248260498046873,
      "memory(GiB)": 77.56,
      "step": 98450,
      "token_acc": 0.5045317220543807,
      "train_speed(iter/s)": 1.438393
    },
    {
      "epoch": 4.218114048241292,
      "grad_norm": 5.632694721221924,
      "learning_rate": 5.913358344495473e-06,
      "loss": 2.342509460449219,
      "memory(GiB)": 77.56,
      "step": 98455,
      "token_acc": 0.5056818181818182,
      "train_speed(iter/s)": 1.438398
    },
    {
      "epoch": 4.218328263570541,
      "grad_norm": 5.1728129386901855,
      "learning_rate": 5.910183979479239e-06,
      "loss": 1.9806270599365234,
      "memory(GiB)": 77.56,
      "step": 98460,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438412
    },
    {
      "epoch": 4.21854247889979,
      "grad_norm": 7.188817501068115,
      "learning_rate": 5.907010413191372e-06,
      "loss": 2.3121814727783203,
      "memory(GiB)": 77.56,
      "step": 98465,
      "token_acc": 0.47686832740213525,
      "train_speed(iter/s)": 1.43842
    },
    {
      "epoch": 4.218756694229039,
      "grad_norm": 6.093756675720215,
      "learning_rate": 5.903837645689381e-06,
      "loss": 2.184810447692871,
      "memory(GiB)": 77.56,
      "step": 98470,
      "token_acc": 0.4714285714285714,
      "train_speed(iter/s)": 1.438432
    },
    {
      "epoch": 4.218970909558288,
      "grad_norm": 5.1576337814331055,
      "learning_rate": 5.900665677030742e-06,
      "loss": 2.134235954284668,
      "memory(GiB)": 77.56,
      "step": 98475,
      "token_acc": 0.5548780487804879,
      "train_speed(iter/s)": 1.438437
    },
    {
      "epoch": 4.219185124887537,
      "grad_norm": 4.905875205993652,
      "learning_rate": 5.89749450727291e-06,
      "loss": 2.8184064865112304,
      "memory(GiB)": 77.56,
      "step": 98480,
      "token_acc": 0.44932432432432434,
      "train_speed(iter/s)": 1.438442
    },
    {
      "epoch": 4.2193993402167855,
      "grad_norm": 8.835637092590332,
      "learning_rate": 5.894324136473334e-06,
      "loss": 1.9978958129882813,
      "memory(GiB)": 77.56,
      "step": 98485,
      "token_acc": 0.5540983606557377,
      "train_speed(iter/s)": 1.438436
    },
    {
      "epoch": 4.219613555546035,
      "grad_norm": 7.472249984741211,
      "learning_rate": 5.891154564689444e-06,
      "loss": 2.3475620269775392,
      "memory(GiB)": 77.56,
      "step": 98490,
      "token_acc": 0.48172757475083056,
      "train_speed(iter/s)": 1.438432
    },
    {
      "epoch": 4.219827770875284,
      "grad_norm": 4.666188716888428,
      "learning_rate": 5.887985791978673e-06,
      "loss": 1.9814159393310546,
      "memory(GiB)": 77.56,
      "step": 98495,
      "token_acc": 0.5874587458745875,
      "train_speed(iter/s)": 1.438432
    },
    {
      "epoch": 4.220041986204532,
      "grad_norm": 9.00847053527832,
      "learning_rate": 5.884817818398414e-06,
      "loss": 2.3205339431762697,
      "memory(GiB)": 77.56,
      "step": 98500,
      "token_acc": 0.4868913857677903,
      "train_speed(iter/s)": 1.438444
    },
    {
      "epoch": 4.220041986204532,
      "eval_loss": 2.128352403640747,
      "eval_runtime": 14.862,
      "eval_samples_per_second": 6.729,
      "eval_steps_per_second": 6.729,
      "eval_token_acc": 0.47406340057636887,
      "step": 98500
    },
    {
      "epoch": 4.220256201533782,
      "grad_norm": 5.496734142303467,
      "learning_rate": 5.881650644006071e-06,
      "loss": 2.1217145919799805,
      "memory(GiB)": 77.56,
      "step": 98505,
      "token_acc": 0.5024485798237023,
      "train_speed(iter/s)": 1.438106
    },
    {
      "epoch": 4.220470416863031,
      "grad_norm": 5.40130090713501,
      "learning_rate": 5.878484268859013e-06,
      "loss": 2.4276966094970702,
      "memory(GiB)": 77.56,
      "step": 98510,
      "token_acc": 0.48493975903614456,
      "train_speed(iter/s)": 1.438105
    },
    {
      "epoch": 4.220684632192279,
      "grad_norm": 6.0814032554626465,
      "learning_rate": 5.875318693014603e-06,
      "loss": 2.6082441329956056,
      "memory(GiB)": 77.56,
      "step": 98515,
      "token_acc": 0.4849624060150376,
      "train_speed(iter/s)": 1.438112
    },
    {
      "epoch": 4.2208988475215286,
      "grad_norm": 5.476480484008789,
      "learning_rate": 5.872153916530187e-06,
      "loss": 2.203523635864258,
      "memory(GiB)": 77.56,
      "step": 98520,
      "token_acc": 0.5236220472440944,
      "train_speed(iter/s)": 1.438125
    },
    {
      "epoch": 4.221113062850778,
      "grad_norm": 6.536108493804932,
      "learning_rate": 5.868989939463087e-06,
      "loss": 1.9606212615966796,
      "memory(GiB)": 77.56,
      "step": 98525,
      "token_acc": 0.5457627118644067,
      "train_speed(iter/s)": 1.438128
    },
    {
      "epoch": 4.221327278180026,
      "grad_norm": 9.845624923706055,
      "learning_rate": 5.865826761870646e-06,
      "loss": 2.414236068725586,
      "memory(GiB)": 77.56,
      "step": 98530,
      "token_acc": 0.4852459016393443,
      "train_speed(iter/s)": 1.43812
    },
    {
      "epoch": 4.221541493509275,
      "grad_norm": 5.5067620277404785,
      "learning_rate": 5.862664383810151e-06,
      "loss": 2.319127655029297,
      "memory(GiB)": 77.56,
      "step": 98535,
      "token_acc": 0.5364238410596026,
      "train_speed(iter/s)": 1.438133
    },
    {
      "epoch": 4.221755708838525,
      "grad_norm": 8.395221710205078,
      "learning_rate": 5.859502805338896e-06,
      "loss": 2.334292984008789,
      "memory(GiB)": 77.56,
      "step": 98540,
      "token_acc": 0.5358649789029536,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 4.221969924167773,
      "grad_norm": 9.597164154052734,
      "learning_rate": 5.856342026514156e-06,
      "loss": 2.358758544921875,
      "memory(GiB)": 77.56,
      "step": 98545,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.438142
    },
    {
      "epoch": 4.222184139497022,
      "grad_norm": 5.924988746643066,
      "learning_rate": 5.85318204739318e-06,
      "loss": 2.1829099655151367,
      "memory(GiB)": 77.56,
      "step": 98550,
      "token_acc": 0.5510204081632653,
      "train_speed(iter/s)": 1.438136
    },
    {
      "epoch": 4.222398354826272,
      "grad_norm": 6.551088333129883,
      "learning_rate": 5.850022868033234e-06,
      "loss": 2.112716293334961,
      "memory(GiB)": 77.56,
      "step": 98555,
      "token_acc": 0.5375494071146245,
      "train_speed(iter/s)": 1.438142
    },
    {
      "epoch": 4.22261257015552,
      "grad_norm": 6.376165866851807,
      "learning_rate": 5.846864488491527e-06,
      "loss": 2.368951606750488,
      "memory(GiB)": 77.56,
      "step": 98560,
      "token_acc": 0.5100671140939598,
      "train_speed(iter/s)": 1.438148
    },
    {
      "epoch": 4.222826785484769,
      "grad_norm": 5.307037353515625,
      "learning_rate": 5.843706908825303e-06,
      "loss": 2.016937255859375,
      "memory(GiB)": 77.56,
      "step": 98565,
      "token_acc": 0.5357142857142857,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 4.2230410008140185,
      "grad_norm": 6.171824932098389,
      "learning_rate": 5.840550129091743e-06,
      "loss": 2.499873924255371,
      "memory(GiB)": 77.56,
      "step": 98570,
      "token_acc": 0.502906976744186,
      "train_speed(iter/s)": 1.43813
    },
    {
      "epoch": 4.223255216143267,
      "grad_norm": 6.762292861938477,
      "learning_rate": 5.837394149348052e-06,
      "loss": 2.5716716766357424,
      "memory(GiB)": 77.56,
      "step": 98575,
      "token_acc": 0.4393939393939394,
      "train_speed(iter/s)": 1.438143
    },
    {
      "epoch": 4.223469431472516,
      "grad_norm": 5.51644229888916,
      "learning_rate": 5.83423896965139e-06,
      "loss": 2.445097732543945,
      "memory(GiB)": 77.56,
      "step": 98580,
      "token_acc": 0.4763636363636364,
      "train_speed(iter/s)": 1.438148
    },
    {
      "epoch": 4.223683646801765,
      "grad_norm": 6.734059810638428,
      "learning_rate": 5.831084590058905e-06,
      "loss": 2.457259178161621,
      "memory(GiB)": 77.56,
      "step": 98585,
      "token_acc": 0.48255813953488375,
      "train_speed(iter/s)": 1.438157
    },
    {
      "epoch": 4.223897862131014,
      "grad_norm": 5.5849103927612305,
      "learning_rate": 5.827931010627774e-06,
      "loss": 2.0584590911865233,
      "memory(GiB)": 77.56,
      "step": 98590,
      "token_acc": 0.5528169014084507,
      "train_speed(iter/s)": 1.438164
    },
    {
      "epoch": 4.224112077460263,
      "grad_norm": 6.087093830108643,
      "learning_rate": 5.824778231415106e-06,
      "loss": 2.3791933059692383,
      "memory(GiB)": 77.56,
      "step": 98595,
      "token_acc": 0.4682274247491639,
      "train_speed(iter/s)": 1.438166
    },
    {
      "epoch": 4.224326292789512,
      "grad_norm": 5.346545696258545,
      "learning_rate": 5.821626252478018e-06,
      "loss": 2.2676155090332033,
      "memory(GiB)": 77.56,
      "step": 98600,
      "token_acc": 0.4566929133858268,
      "train_speed(iter/s)": 1.438172
    },
    {
      "epoch": 4.224540508118761,
      "grad_norm": 6.784097194671631,
      "learning_rate": 5.818475073873614e-06,
      "loss": 2.5720359802246096,
      "memory(GiB)": 77.56,
      "step": 98605,
      "token_acc": 0.45,
      "train_speed(iter/s)": 1.438171
    },
    {
      "epoch": 4.22475472344801,
      "grad_norm": 6.069586753845215,
      "learning_rate": 5.815324695658975e-06,
      "loss": 2.101156806945801,
      "memory(GiB)": 77.56,
      "step": 98610,
      "token_acc": 0.5467128027681661,
      "train_speed(iter/s)": 1.438168
    },
    {
      "epoch": 4.224968938777259,
      "grad_norm": 6.479527950286865,
      "learning_rate": 5.812175117891183e-06,
      "loss": 2.158317756652832,
      "memory(GiB)": 77.56,
      "step": 98615,
      "token_acc": 0.5425101214574899,
      "train_speed(iter/s)": 1.438174
    },
    {
      "epoch": 4.2251831541065075,
      "grad_norm": 5.19526481628418,
      "learning_rate": 5.809026340627288e-06,
      "loss": 2.595577621459961,
      "memory(GiB)": 77.56,
      "step": 98620,
      "token_acc": 0.4509090909090909,
      "train_speed(iter/s)": 1.438181
    },
    {
      "epoch": 4.225397369435757,
      "grad_norm": 6.5625481605529785,
      "learning_rate": 5.805878363924338e-06,
      "loss": 2.572738456726074,
      "memory(GiB)": 77.56,
      "step": 98625,
      "token_acc": 0.4875,
      "train_speed(iter/s)": 1.438194
    },
    {
      "epoch": 4.225611584765006,
      "grad_norm": 7.128811359405518,
      "learning_rate": 5.802731187839361e-06,
      "loss": 2.055022430419922,
      "memory(GiB)": 77.56,
      "step": 98630,
      "token_acc": 0.5895522388059702,
      "train_speed(iter/s)": 1.438193
    },
    {
      "epoch": 4.225825800094254,
      "grad_norm": 7.126205921173096,
      "learning_rate": 5.799584812429354e-06,
      "loss": 2.409199523925781,
      "memory(GiB)": 77.56,
      "step": 98635,
      "token_acc": 0.46,
      "train_speed(iter/s)": 1.438175
    },
    {
      "epoch": 4.226040015423504,
      "grad_norm": 7.448665618896484,
      "learning_rate": 5.7964392377513445e-06,
      "loss": 2.297069549560547,
      "memory(GiB)": 77.56,
      "step": 98640,
      "token_acc": 0.47619047619047616,
      "train_speed(iter/s)": 1.438181
    },
    {
      "epoch": 4.226254230752753,
      "grad_norm": 6.745507717132568,
      "learning_rate": 5.793294463862297e-06,
      "loss": 2.4201173782348633,
      "memory(GiB)": 77.56,
      "step": 98645,
      "token_acc": 0.5272108843537415,
      "train_speed(iter/s)": 1.438179
    },
    {
      "epoch": 4.226468446082001,
      "grad_norm": 7.347485542297363,
      "learning_rate": 5.790150490819196e-06,
      "loss": 2.2076059341430665,
      "memory(GiB)": 77.56,
      "step": 98650,
      "token_acc": 0.5212765957446809,
      "train_speed(iter/s)": 1.438188
    },
    {
      "epoch": 4.2266826614112505,
      "grad_norm": 8.259467124938965,
      "learning_rate": 5.787007318678994e-06,
      "loss": 2.267050552368164,
      "memory(GiB)": 77.56,
      "step": 98655,
      "token_acc": 0.5435540069686411,
      "train_speed(iter/s)": 1.438186
    },
    {
      "epoch": 4.2268968767405,
      "grad_norm": 5.506394386291504,
      "learning_rate": 5.7838649474986235e-06,
      "loss": 2.2759540557861326,
      "memory(GiB)": 77.56,
      "step": 98660,
      "token_acc": 0.5097402597402597,
      "train_speed(iter/s)": 1.438177
    },
    {
      "epoch": 4.227111092069748,
      "grad_norm": 6.584667205810547,
      "learning_rate": 5.7807233773350235e-06,
      "loss": 2.429344367980957,
      "memory(GiB)": 77.56,
      "step": 98665,
      "token_acc": 0.5272727272727272,
      "train_speed(iter/s)": 1.438176
    },
    {
      "epoch": 4.227325307398997,
      "grad_norm": 6.250676155090332,
      "learning_rate": 5.7775826082450915e-06,
      "loss": 2.323969268798828,
      "memory(GiB)": 77.56,
      "step": 98670,
      "token_acc": 0.49375,
      "train_speed(iter/s)": 1.438181
    },
    {
      "epoch": 4.227539522728247,
      "grad_norm": 5.680970191955566,
      "learning_rate": 5.774442640285738e-06,
      "loss": 1.863088035583496,
      "memory(GiB)": 77.56,
      "step": 98675,
      "token_acc": 0.532871972318339,
      "train_speed(iter/s)": 1.438194
    },
    {
      "epoch": 4.227753738057495,
      "grad_norm": 5.946530818939209,
      "learning_rate": 5.77130347351385e-06,
      "loss": 1.9548000335693358,
      "memory(GiB)": 77.56,
      "step": 98680,
      "token_acc": 0.5838150289017341,
      "train_speed(iter/s)": 1.438184
    },
    {
      "epoch": 4.227967953386744,
      "grad_norm": 6.0132155418396,
      "learning_rate": 5.768165107986284e-06,
      "loss": 2.1032190322875977,
      "memory(GiB)": 77.56,
      "step": 98685,
      "token_acc": 0.47692307692307695,
      "train_speed(iter/s)": 1.438202
    },
    {
      "epoch": 4.2281821687159935,
      "grad_norm": 6.270233631134033,
      "learning_rate": 5.765027543759904e-06,
      "loss": 2.336669158935547,
      "memory(GiB)": 77.56,
      "step": 98690,
      "token_acc": 0.46204620462046203,
      "train_speed(iter/s)": 1.438204
    },
    {
      "epoch": 4.228396384045242,
      "grad_norm": 6.963911533355713,
      "learning_rate": 5.7618907808915325e-06,
      "loss": 2.1510234832763673,
      "memory(GiB)": 77.56,
      "step": 98695,
      "token_acc": 0.5376344086021505,
      "train_speed(iter/s)": 1.438219
    },
    {
      "epoch": 4.228610599374491,
      "grad_norm": 6.034351348876953,
      "learning_rate": 5.758754819438017e-06,
      "loss": 2.3063426971435548,
      "memory(GiB)": 77.56,
      "step": 98700,
      "token_acc": 0.5249169435215947,
      "train_speed(iter/s)": 1.438227
    },
    {
      "epoch": 4.22882481470374,
      "grad_norm": 6.990081787109375,
      "learning_rate": 5.755619659456163e-06,
      "loss": 2.3793033599853515,
      "memory(GiB)": 77.56,
      "step": 98705,
      "token_acc": 0.47214076246334313,
      "train_speed(iter/s)": 1.438243
    },
    {
      "epoch": 4.229039030032989,
      "grad_norm": 5.555932998657227,
      "learning_rate": 5.752485301002752e-06,
      "loss": 2.240178871154785,
      "memory(GiB)": 77.56,
      "step": 98710,
      "token_acc": 0.4793103448275862,
      "train_speed(iter/s)": 1.438229
    },
    {
      "epoch": 4.229253245362238,
      "grad_norm": 6.410317897796631,
      "learning_rate": 5.74935174413459e-06,
      "loss": 2.4478342056274416,
      "memory(GiB)": 77.56,
      "step": 98715,
      "token_acc": 0.5328947368421053,
      "train_speed(iter/s)": 1.438243
    },
    {
      "epoch": 4.229467460691487,
      "grad_norm": 5.335734844207764,
      "learning_rate": 5.746218988908431e-06,
      "loss": 2.292744255065918,
      "memory(GiB)": 77.56,
      "step": 98720,
      "token_acc": 0.5223367697594502,
      "train_speed(iter/s)": 1.438243
    },
    {
      "epoch": 4.229681676020736,
      "grad_norm": 7.612647533416748,
      "learning_rate": 5.743087035381028e-06,
      "loss": 2.281413459777832,
      "memory(GiB)": 77.56,
      "step": 98725,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.438246
    },
    {
      "epoch": 4.229895891349985,
      "grad_norm": 6.659115314483643,
      "learning_rate": 5.739955883609105e-06,
      "loss": 2.3006193161010744,
      "memory(GiB)": 77.56,
      "step": 98730,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.438251
    },
    {
      "epoch": 4.230110106679234,
      "grad_norm": 7.927532196044922,
      "learning_rate": 5.736825533649415e-06,
      "loss": 2.4834529876708986,
      "memory(GiB)": 77.56,
      "step": 98735,
      "token_acc": 0.5052631578947369,
      "train_speed(iter/s)": 1.438254
    },
    {
      "epoch": 4.2303243220084825,
      "grad_norm": 5.332328796386719,
      "learning_rate": 5.733695985558651e-06,
      "loss": 2.2541717529296874,
      "memory(GiB)": 77.56,
      "step": 98740,
      "token_acc": 0.5382262996941896,
      "train_speed(iter/s)": 1.43824
    },
    {
      "epoch": 4.230538537337732,
      "grad_norm": 9.012012481689453,
      "learning_rate": 5.730567239393514e-06,
      "loss": 2.1209537506103517,
      "memory(GiB)": 77.56,
      "step": 98745,
      "token_acc": 0.5467625899280576,
      "train_speed(iter/s)": 1.438246
    },
    {
      "epoch": 4.230752752666981,
      "grad_norm": 5.358112812042236,
      "learning_rate": 5.7274392952106735e-06,
      "loss": 2.108470344543457,
      "memory(GiB)": 77.56,
      "step": 98750,
      "token_acc": 0.5492957746478874,
      "train_speed(iter/s)": 1.438253
    },
    {
      "epoch": 4.230966967996229,
      "grad_norm": 6.052133560180664,
      "learning_rate": 5.724312153066791e-06,
      "loss": 2.447452926635742,
      "memory(GiB)": 77.56,
      "step": 98755,
      "token_acc": 0.5016722408026756,
      "train_speed(iter/s)": 1.438258
    },
    {
      "epoch": 4.231181183325479,
      "grad_norm": 7.421567916870117,
      "learning_rate": 5.721185813018543e-06,
      "loss": 2.2603517532348634,
      "memory(GiB)": 77.56,
      "step": 98760,
      "token_acc": 0.5229357798165137,
      "train_speed(iter/s)": 1.438257
    },
    {
      "epoch": 4.231395398654728,
      "grad_norm": 5.925826549530029,
      "learning_rate": 5.71806027512255e-06,
      "loss": 2.478693389892578,
      "memory(GiB)": 77.56,
      "step": 98765,
      "token_acc": 0.47435897435897434,
      "train_speed(iter/s)": 1.438266
    },
    {
      "epoch": 4.231609613983976,
      "grad_norm": 5.559430122375488,
      "learning_rate": 5.714935539435429e-06,
      "loss": 2.3873729705810547,
      "memory(GiB)": 77.56,
      "step": 98770,
      "token_acc": 0.47151898734177217,
      "train_speed(iter/s)": 1.438274
    },
    {
      "epoch": 4.2318238293132255,
      "grad_norm": 5.392225742340088,
      "learning_rate": 5.711811606013801e-06,
      "loss": 1.9760448455810546,
      "memory(GiB)": 77.56,
      "step": 98775,
      "token_acc": 0.5381818181818182,
      "train_speed(iter/s)": 1.438269
    },
    {
      "epoch": 4.232038044642475,
      "grad_norm": 8.049981117248535,
      "learning_rate": 5.708688474914237e-06,
      "loss": 2.337434768676758,
      "memory(GiB)": 77.56,
      "step": 98780,
      "token_acc": 0.504424778761062,
      "train_speed(iter/s)": 1.43827
    },
    {
      "epoch": 4.232252259971723,
      "grad_norm": 5.579767227172852,
      "learning_rate": 5.705566146193342e-06,
      "loss": 2.1649669647216796,
      "memory(GiB)": 77.56,
      "step": 98785,
      "token_acc": 0.550561797752809,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 4.232466475300972,
      "grad_norm": 6.73727560043335,
      "learning_rate": 5.702444619907654e-06,
      "loss": 2.6488134384155275,
      "memory(GiB)": 77.56,
      "step": 98790,
      "token_acc": 0.4542124542124542,
      "train_speed(iter/s)": 1.438296
    },
    {
      "epoch": 4.232680690630222,
      "grad_norm": 8.497204780578613,
      "learning_rate": 5.699323896113751e-06,
      "loss": 2.3096485137939453,
      "memory(GiB)": 77.56,
      "step": 98795,
      "token_acc": 0.5138339920948617,
      "train_speed(iter/s)": 1.438298
    },
    {
      "epoch": 4.23289490595947,
      "grad_norm": 6.431375026702881,
      "learning_rate": 5.696203974868147e-06,
      "loss": 2.2419174194335936,
      "memory(GiB)": 77.56,
      "step": 98800,
      "token_acc": 0.5083612040133779,
      "train_speed(iter/s)": 1.438297
    },
    {
      "epoch": 4.233109121288719,
      "grad_norm": 6.082571983337402,
      "learning_rate": 5.6930848562273766e-06,
      "loss": 2.203379821777344,
      "memory(GiB)": 77.56,
      "step": 98805,
      "token_acc": 0.5638629283489096,
      "train_speed(iter/s)": 1.438305
    },
    {
      "epoch": 4.233323336617969,
      "grad_norm": 5.070413112640381,
      "learning_rate": 5.6899665402479315e-06,
      "loss": 2.2239059448242187,
      "memory(GiB)": 77.56,
      "step": 98810,
      "token_acc": 0.512,
      "train_speed(iter/s)": 1.438312
    },
    {
      "epoch": 4.233537551947217,
      "grad_norm": 7.9468536376953125,
      "learning_rate": 5.686849026986296e-06,
      "loss": 2.5571640014648436,
      "memory(GiB)": 77.56,
      "step": 98815,
      "token_acc": 0.49850746268656715,
      "train_speed(iter/s)": 1.438307
    },
    {
      "epoch": 4.233751767276466,
      "grad_norm": 7.240723133087158,
      "learning_rate": 5.683732316498974e-06,
      "loss": 2.2770036697387694,
      "memory(GiB)": 77.56,
      "step": 98820,
      "token_acc": 0.5231316725978647,
      "train_speed(iter/s)": 1.438304
    },
    {
      "epoch": 4.233965982605715,
      "grad_norm": 6.013421535491943,
      "learning_rate": 5.6806164088424154e-06,
      "loss": 2.408305549621582,
      "memory(GiB)": 77.56,
      "step": 98825,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438295
    },
    {
      "epoch": 4.234180197934964,
      "grad_norm": 8.179483413696289,
      "learning_rate": 5.677501304073063e-06,
      "loss": 2.2367530822753907,
      "memory(GiB)": 77.56,
      "step": 98830,
      "token_acc": 0.5178571428571429,
      "train_speed(iter/s)": 1.438265
    },
    {
      "epoch": 4.234394413264213,
      "grad_norm": 5.203120231628418,
      "learning_rate": 5.67438700224735e-06,
      "loss": 2.2220144271850586,
      "memory(GiB)": 77.56,
      "step": 98835,
      "token_acc": 0.5289256198347108,
      "train_speed(iter/s)": 1.438288
    },
    {
      "epoch": 4.234608628593462,
      "grad_norm": 7.3809919357299805,
      "learning_rate": 5.671273503421692e-06,
      "loss": 2.2949779510498045,
      "memory(GiB)": 77.56,
      "step": 98840,
      "token_acc": 0.5337837837837838,
      "train_speed(iter/s)": 1.438313
    },
    {
      "epoch": 4.234822843922711,
      "grad_norm": 5.515556335449219,
      "learning_rate": 5.668160807652506e-06,
      "loss": 2.5175912857055662,
      "memory(GiB)": 77.56,
      "step": 98845,
      "token_acc": 0.46107784431137727,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 4.23503705925196,
      "grad_norm": 6.509804725646973,
      "learning_rate": 5.6650489149961685e-06,
      "loss": 2.190166473388672,
      "memory(GiB)": 77.56,
      "step": 98850,
      "token_acc": 0.5270758122743683,
      "train_speed(iter/s)": 1.438328
    },
    {
      "epoch": 4.235251274581209,
      "grad_norm": 8.552698135375977,
      "learning_rate": 5.661937825509067e-06,
      "loss": 2.0742788314819336,
      "memory(GiB)": 77.56,
      "step": 98855,
      "token_acc": 0.5591836734693878,
      "train_speed(iter/s)": 1.438326
    },
    {
      "epoch": 4.235465489910458,
      "grad_norm": 5.004457950592041,
      "learning_rate": 5.658827539247541e-06,
      "loss": 2.5703901290893554,
      "memory(GiB)": 77.56,
      "step": 98860,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.438324
    },
    {
      "epoch": 4.235679705239707,
      "grad_norm": 7.116892337799072,
      "learning_rate": 5.655718056267962e-06,
      "loss": 2.416287422180176,
      "memory(GiB)": 77.56,
      "step": 98865,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.438343
    },
    {
      "epoch": 4.235893920568956,
      "grad_norm": 7.928796768188477,
      "learning_rate": 5.652609376626644e-06,
      "loss": 2.3415254592895507,
      "memory(GiB)": 77.56,
      "step": 98870,
      "token_acc": 0.5033557046979866,
      "train_speed(iter/s)": 1.438358
    },
    {
      "epoch": 4.2361081358982045,
      "grad_norm": 5.152106285095215,
      "learning_rate": 5.6495015003799e-06,
      "loss": 2.1879743576049804,
      "memory(GiB)": 77.56,
      "step": 98875,
      "token_acc": 0.5185185185185185,
      "train_speed(iter/s)": 1.438379
    },
    {
      "epoch": 4.236322351227454,
      "grad_norm": 5.262955665588379,
      "learning_rate": 5.6463944275840505e-06,
      "loss": 1.9860342025756836,
      "memory(GiB)": 77.56,
      "step": 98880,
      "token_acc": 0.5482625482625483,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 4.236536566556703,
      "grad_norm": 6.443000316619873,
      "learning_rate": 5.643288158295373e-06,
      "loss": 2.4924346923828127,
      "memory(GiB)": 77.56,
      "step": 98885,
      "token_acc": 0.45614035087719296,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 4.236750781885951,
      "grad_norm": 5.321374416351318,
      "learning_rate": 5.640182692570145e-06,
      "loss": 2.2751941680908203,
      "memory(GiB)": 77.56,
      "step": 98890,
      "token_acc": 0.5103626943005182,
      "train_speed(iter/s)": 1.438378
    },
    {
      "epoch": 4.236964997215201,
      "grad_norm": 5.902636528015137,
      "learning_rate": 5.637078030464615e-06,
      "loss": 2.33965950012207,
      "memory(GiB)": 77.56,
      "step": 98895,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438389
    },
    {
      "epoch": 4.23717921254445,
      "grad_norm": 6.948122978210449,
      "learning_rate": 5.633974172035023e-06,
      "loss": 2.2784000396728517,
      "memory(GiB)": 77.56,
      "step": 98900,
      "token_acc": 0.5305466237942122,
      "train_speed(iter/s)": 1.438372
    },
    {
      "epoch": 4.237393427873698,
      "grad_norm": 5.692466735839844,
      "learning_rate": 5.630871117337621e-06,
      "loss": 2.212904357910156,
      "memory(GiB)": 77.56,
      "step": 98905,
      "token_acc": 0.49842271293375395,
      "train_speed(iter/s)": 1.438363
    },
    {
      "epoch": 4.2376076432029475,
      "grad_norm": 6.55184268951416,
      "learning_rate": 5.627768866428606e-06,
      "loss": 2.5515163421630858,
      "memory(GiB)": 77.56,
      "step": 98910,
      "token_acc": 0.5160349854227405,
      "train_speed(iter/s)": 1.438366
    },
    {
      "epoch": 4.237821858532197,
      "grad_norm": 5.398247241973877,
      "learning_rate": 5.624667419364182e-06,
      "loss": 2.1727046966552734,
      "memory(GiB)": 77.56,
      "step": 98915,
      "token_acc": 0.551094890510949,
      "train_speed(iter/s)": 1.438361
    },
    {
      "epoch": 4.238036073861445,
      "grad_norm": 6.375877380371094,
      "learning_rate": 5.621566776200538e-06,
      "loss": 2.218283843994141,
      "memory(GiB)": 77.56,
      "step": 98920,
      "token_acc": 0.501577287066246,
      "train_speed(iter/s)": 1.438354
    },
    {
      "epoch": 4.238250289190694,
      "grad_norm": 4.74083948135376,
      "learning_rate": 5.618466936993832e-06,
      "loss": 2.5549644470214843,
      "memory(GiB)": 77.56,
      "step": 98925,
      "token_acc": 0.5013927576601671,
      "train_speed(iter/s)": 1.438321
    },
    {
      "epoch": 4.238464504519944,
      "grad_norm": 6.965949058532715,
      "learning_rate": 5.6153679018002394e-06,
      "loss": 2.1256568908691404,
      "memory(GiB)": 77.56,
      "step": 98930,
      "token_acc": 0.5222222222222223,
      "train_speed(iter/s)": 1.438319
    },
    {
      "epoch": 4.238678719849192,
      "grad_norm": 7.280891418457031,
      "learning_rate": 5.612269670675885e-06,
      "loss": 2.128206253051758,
      "memory(GiB)": 77.56,
      "step": 98935,
      "token_acc": 0.5138888888888888,
      "train_speed(iter/s)": 1.43832
    },
    {
      "epoch": 4.238892935178441,
      "grad_norm": 6.390178680419922,
      "learning_rate": 5.609172243676914e-06,
      "loss": 2.2373172760009767,
      "memory(GiB)": 77.56,
      "step": 98940,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.43834
    },
    {
      "epoch": 4.2391071505076905,
      "grad_norm": 5.835618019104004,
      "learning_rate": 5.60607562085943e-06,
      "loss": 2.502004623413086,
      "memory(GiB)": 77.56,
      "step": 98945,
      "token_acc": 0.4811594202898551,
      "train_speed(iter/s)": 1.438343
    },
    {
      "epoch": 4.23932136583694,
      "grad_norm": 6.4024176597595215,
      "learning_rate": 5.602979802279534e-06,
      "loss": 2.2551433563232424,
      "memory(GiB)": 77.56,
      "step": 98950,
      "token_acc": 0.5303514376996805,
      "train_speed(iter/s)": 1.438356
    },
    {
      "epoch": 4.239535581166188,
      "grad_norm": 5.6541666984558105,
      "learning_rate": 5.599884787993304e-06,
      "loss": 2.5244083404541016,
      "memory(GiB)": 77.56,
      "step": 98955,
      "token_acc": 0.45980707395498394,
      "train_speed(iter/s)": 1.438343
    },
    {
      "epoch": 4.239749796495437,
      "grad_norm": 6.16555118560791,
      "learning_rate": 5.5967905780567985e-06,
      "loss": 2.547613334655762,
      "memory(GiB)": 77.56,
      "step": 98960,
      "token_acc": 0.4722222222222222,
      "train_speed(iter/s)": 1.438342
    },
    {
      "epoch": 4.239964011824686,
      "grad_norm": 5.594049453735352,
      "learning_rate": 5.593697172526097e-06,
      "loss": 2.288450241088867,
      "memory(GiB)": 77.56,
      "step": 98965,
      "token_acc": 0.5497076023391813,
      "train_speed(iter/s)": 1.438352
    },
    {
      "epoch": 4.240178227153935,
      "grad_norm": 8.040581703186035,
      "learning_rate": 5.590604571457231e-06,
      "loss": 2.507685661315918,
      "memory(GiB)": 77.56,
      "step": 98970,
      "token_acc": 0.48322147651006714,
      "train_speed(iter/s)": 1.438341
    },
    {
      "epoch": 4.240392442483184,
      "grad_norm": 6.547095775604248,
      "learning_rate": 5.587512774906217e-06,
      "loss": 2.1984237670898437,
      "memory(GiB)": 77.56,
      "step": 98975,
      "token_acc": 0.5352112676056338,
      "train_speed(iter/s)": 1.43834
    },
    {
      "epoch": 4.2406066578124335,
      "grad_norm": 5.017219543457031,
      "learning_rate": 5.584421782929078e-06,
      "loss": 1.9073963165283203,
      "memory(GiB)": 77.56,
      "step": 98980,
      "token_acc": 0.547244094488189,
      "train_speed(iter/s)": 1.438345
    },
    {
      "epoch": 4.240820873141682,
      "grad_norm": 6.180419445037842,
      "learning_rate": 5.58133159558179e-06,
      "loss": 2.2009613037109377,
      "memory(GiB)": 77.56,
      "step": 98985,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.438353
    },
    {
      "epoch": 4.241035088470931,
      "grad_norm": 7.331000804901123,
      "learning_rate": 5.5782422129203534e-06,
      "loss": 2.3994380950927736,
      "memory(GiB)": 77.56,
      "step": 98990,
      "token_acc": 0.4717607973421927,
      "train_speed(iter/s)": 1.438374
    },
    {
      "epoch": 4.2412493038001795,
      "grad_norm": 6.821191310882568,
      "learning_rate": 5.575153635000735e-06,
      "loss": 2.139596939086914,
      "memory(GiB)": 77.56,
      "step": 98995,
      "token_acc": 0.5096525096525096,
      "train_speed(iter/s)": 1.438373
    },
    {
      "epoch": 4.241463519129429,
      "grad_norm": 5.0258917808532715,
      "learning_rate": 5.572065861878883e-06,
      "loss": 2.1131879806518556,
      "memory(GiB)": 77.56,
      "step": 99000,
      "token_acc": 0.5491525423728814,
      "train_speed(iter/s)": 1.438371
    },
    {
      "epoch": 4.241463519129429,
      "eval_loss": 2.3487865924835205,
      "eval_runtime": 14.4108,
      "eval_samples_per_second": 6.939,
      "eval_steps_per_second": 6.939,
      "eval_token_acc": 0.45481049562682213,
      "step": 99000
    },
    {
      "epoch": 4.241677734458678,
      "grad_norm": 5.596465110778809,
      "learning_rate": 5.568978893610727e-06,
      "loss": 2.135300064086914,
      "memory(GiB)": 77.56,
      "step": 99005,
      "token_acc": 0.46707193515704154,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 4.241891949787927,
      "grad_norm": 5.467395305633545,
      "learning_rate": 5.565892730252203e-06,
      "loss": 2.4273265838623046,
      "memory(GiB)": 77.56,
      "step": 99010,
      "token_acc": 0.5147540983606558,
      "train_speed(iter/s)": 1.438064
    },
    {
      "epoch": 4.242106165117176,
      "grad_norm": 7.064877986907959,
      "learning_rate": 5.5628073718592184e-06,
      "loss": 2.356730842590332,
      "memory(GiB)": 77.56,
      "step": 99015,
      "token_acc": 0.49166666666666664,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 4.242320380446425,
      "grad_norm": 5.671635627746582,
      "learning_rate": 5.55972281848765e-06,
      "loss": 2.390118217468262,
      "memory(GiB)": 77.56,
      "step": 99020,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 4.242534595775673,
      "grad_norm": 4.671056270599365,
      "learning_rate": 5.556639070193404e-06,
      "loss": 2.126407241821289,
      "memory(GiB)": 77.56,
      "step": 99025,
      "token_acc": 0.5173745173745173,
      "train_speed(iter/s)": 1.43806
    },
    {
      "epoch": 4.2427488111049225,
      "grad_norm": 5.497961044311523,
      "learning_rate": 5.553556127032333e-06,
      "loss": 2.2226787567138673,
      "memory(GiB)": 77.56,
      "step": 99030,
      "token_acc": 0.519434628975265,
      "train_speed(iter/s)": 1.438056
    },
    {
      "epoch": 4.242963026434172,
      "grad_norm": 8.50023078918457,
      "learning_rate": 5.550473989060284e-06,
      "loss": 2.1681802749633787,
      "memory(GiB)": 77.56,
      "step": 99035,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.438061
    },
    {
      "epoch": 4.243177241763421,
      "grad_norm": 6.356548309326172,
      "learning_rate": 5.5473926563331e-06,
      "loss": 2.2580083847045898,
      "memory(GiB)": 77.56,
      "step": 99040,
      "token_acc": 0.5054945054945055,
      "train_speed(iter/s)": 1.438069
    },
    {
      "epoch": 4.243391457092669,
      "grad_norm": 6.910303592681885,
      "learning_rate": 5.5443121289065814e-06,
      "loss": 2.2745216369628904,
      "memory(GiB)": 77.56,
      "step": 99045,
      "token_acc": 0.47318611987381703,
      "train_speed(iter/s)": 1.438052
    },
    {
      "epoch": 4.243605672421919,
      "grad_norm": 8.169788360595703,
      "learning_rate": 5.541232406836567e-06,
      "loss": 2.2359338760375977,
      "memory(GiB)": 77.56,
      "step": 99050,
      "token_acc": 0.4728682170542636,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 4.243819887751167,
      "grad_norm": 6.21835994720459,
      "learning_rate": 5.538153490178833e-06,
      "loss": 2.3249101638793945,
      "memory(GiB)": 77.56,
      "step": 99055,
      "token_acc": 0.49363057324840764,
      "train_speed(iter/s)": 1.438081
    },
    {
      "epoch": 4.244034103080416,
      "grad_norm": 5.761159896850586,
      "learning_rate": 5.5350753789891515e-06,
      "loss": 2.3328433990478517,
      "memory(GiB)": 77.56,
      "step": 99060,
      "token_acc": 0.5325443786982249,
      "train_speed(iter/s)": 1.438078
    },
    {
      "epoch": 4.2442483184096655,
      "grad_norm": 7.299875259399414,
      "learning_rate": 5.531998073323297e-06,
      "loss": 2.1044246673583986,
      "memory(GiB)": 77.56,
      "step": 99065,
      "token_acc": 0.5150375939849624,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 4.244462533738915,
      "grad_norm": 7.110866546630859,
      "learning_rate": 5.528921573237001e-06,
      "loss": 2.329795074462891,
      "memory(GiB)": 77.56,
      "step": 99070,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 4.244676749068163,
      "grad_norm": 7.9992804527282715,
      "learning_rate": 5.525845878786018e-06,
      "loss": 2.4182613372802733,
      "memory(GiB)": 77.56,
      "step": 99075,
      "token_acc": 0.5098039215686274,
      "train_speed(iter/s)": 1.438082
    },
    {
      "epoch": 4.244890964397412,
      "grad_norm": 7.7661519050598145,
      "learning_rate": 5.522770990026044e-06,
      "loss": 2.332794952392578,
      "memory(GiB)": 77.56,
      "step": 99080,
      "token_acc": 0.48788927335640137,
      "train_speed(iter/s)": 1.438091
    },
    {
      "epoch": 4.245105179726661,
      "grad_norm": 8.123416900634766,
      "learning_rate": 5.519696907012811e-06,
      "loss": 2.0453876495361327,
      "memory(GiB)": 77.56,
      "step": 99085,
      "token_acc": 0.5372168284789643,
      "train_speed(iter/s)": 1.438087
    },
    {
      "epoch": 4.24531939505591,
      "grad_norm": 7.143505096435547,
      "learning_rate": 5.516623629801987e-06,
      "loss": 2.131815719604492,
      "memory(GiB)": 77.56,
      "step": 99090,
      "token_acc": 0.5641025641025641,
      "train_speed(iter/s)": 1.43808
    },
    {
      "epoch": 4.245533610385159,
      "grad_norm": 7.471089839935303,
      "learning_rate": 5.51355115844926e-06,
      "loss": 2.3486806869506838,
      "memory(GiB)": 77.56,
      "step": 99095,
      "token_acc": 0.49258160237388726,
      "train_speed(iter/s)": 1.438084
    },
    {
      "epoch": 4.245747825714409,
      "grad_norm": 5.78510856628418,
      "learning_rate": 5.510479493010285e-06,
      "loss": 2.678004837036133,
      "memory(GiB)": 77.56,
      "step": 99100,
      "token_acc": 0.483695652173913,
      "train_speed(iter/s)": 1.438067
    },
    {
      "epoch": 4.245962041043657,
      "grad_norm": 5.872328281402588,
      "learning_rate": 5.5074086335407e-06,
      "loss": 2.4142339706420897,
      "memory(GiB)": 77.56,
      "step": 99105,
      "token_acc": 0.48826291079812206,
      "train_speed(iter/s)": 1.438059
    },
    {
      "epoch": 4.246176256372906,
      "grad_norm": 5.436133861541748,
      "learning_rate": 5.504338580096152e-06,
      "loss": 2.1725513458251955,
      "memory(GiB)": 77.56,
      "step": 99110,
      "token_acc": 0.5076452599388379,
      "train_speed(iter/s)": 1.438073
    },
    {
      "epoch": 4.246390471702155,
      "grad_norm": 7.735098361968994,
      "learning_rate": 5.501269332732256e-06,
      "loss": 2.75048828125,
      "memory(GiB)": 77.56,
      "step": 99115,
      "token_acc": 0.4528301886792453,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 4.246604687031404,
      "grad_norm": 5.928342342376709,
      "learning_rate": 5.498200891504606e-06,
      "loss": 2.1290210723876952,
      "memory(GiB)": 77.56,
      "step": 99120,
      "token_acc": 0.5490909090909091,
      "train_speed(iter/s)": 1.438072
    },
    {
      "epoch": 4.246818902360653,
      "grad_norm": 6.75995397567749,
      "learning_rate": 5.4951332564687926e-06,
      "loss": 2.1365535736083983,
      "memory(GiB)": 77.56,
      "step": 99125,
      "token_acc": 0.5247524752475248,
      "train_speed(iter/s)": 1.438076
    },
    {
      "epoch": 4.247033117689902,
      "grad_norm": 10.028285026550293,
      "learning_rate": 5.492066427680376e-06,
      "loss": 2.295850944519043,
      "memory(GiB)": 77.56,
      "step": 99130,
      "token_acc": 0.49454545454545457,
      "train_speed(iter/s)": 1.438083
    },
    {
      "epoch": 4.247247333019151,
      "grad_norm": 7.067670822143555,
      "learning_rate": 5.489000405194944e-06,
      "loss": 2.4321945190429686,
      "memory(GiB)": 77.56,
      "step": 99135,
      "token_acc": 0.46779661016949153,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.2474615483484,
      "grad_norm": 8.564013481140137,
      "learning_rate": 5.48593518906802e-06,
      "loss": 2.015664291381836,
      "memory(GiB)": 77.56,
      "step": 99140,
      "token_acc": 0.5782608695652174,
      "train_speed(iter/s)": 1.438062
    },
    {
      "epoch": 4.247675763677648,
      "grad_norm": 6.376132488250732,
      "learning_rate": 5.482870779355137e-06,
      "loss": 2.229803466796875,
      "memory(GiB)": 77.56,
      "step": 99145,
      "token_acc": 0.5241379310344828,
      "train_speed(iter/s)": 1.438075
    },
    {
      "epoch": 4.247889979006898,
      "grad_norm": 7.828505992889404,
      "learning_rate": 5.479807176111812e-06,
      "loss": 2.4596206665039064,
      "memory(GiB)": 77.56,
      "step": 99150,
      "token_acc": 0.4935064935064935,
      "train_speed(iter/s)": 1.438089
    },
    {
      "epoch": 4.248104194336147,
      "grad_norm": 6.324970245361328,
      "learning_rate": 5.476744379393534e-06,
      "loss": 2.4612213134765626,
      "memory(GiB)": 77.56,
      "step": 99155,
      "token_acc": 0.5148514851485149,
      "train_speed(iter/s)": 1.43809
    },
    {
      "epoch": 4.248318409665396,
      "grad_norm": 6.837876796722412,
      "learning_rate": 5.4736823892558075e-06,
      "loss": 1.9744007110595703,
      "memory(GiB)": 77.56,
      "step": 99160,
      "token_acc": 0.5381944444444444,
      "train_speed(iter/s)": 1.438102
    },
    {
      "epoch": 4.2485326249946445,
      "grad_norm": 7.05295467376709,
      "learning_rate": 5.470621205754084e-06,
      "loss": 1.9171564102172851,
      "memory(GiB)": 77.56,
      "step": 99165,
      "token_acc": 0.5826771653543307,
      "train_speed(iter/s)": 1.438111
    },
    {
      "epoch": 4.248746840323894,
      "grad_norm": 5.768963813781738,
      "learning_rate": 5.467560828943841e-06,
      "loss": 2.338684844970703,
      "memory(GiB)": 77.56,
      "step": 99170,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438106
    },
    {
      "epoch": 4.248961055653142,
      "grad_norm": 7.49630880355835,
      "learning_rate": 5.464501258880505e-06,
      "loss": 2.3309356689453127,
      "memory(GiB)": 77.56,
      "step": 99175,
      "token_acc": 0.527027027027027,
      "train_speed(iter/s)": 1.4381
    },
    {
      "epoch": 4.249175270982391,
      "grad_norm": 6.827689170837402,
      "learning_rate": 5.461442495619507e-06,
      "loss": 2.3259746551513674,
      "memory(GiB)": 77.56,
      "step": 99180,
      "token_acc": 0.4921875,
      "train_speed(iter/s)": 1.438103
    },
    {
      "epoch": 4.249389486311641,
      "grad_norm": 6.1214776039123535,
      "learning_rate": 5.45838453921626e-06,
      "loss": 2.5952426910400392,
      "memory(GiB)": 77.56,
      "step": 99185,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.438123
    },
    {
      "epoch": 4.24960370164089,
      "grad_norm": 6.791317939758301,
      "learning_rate": 5.455327389726151e-06,
      "loss": 2.275722694396973,
      "memory(GiB)": 77.56,
      "step": 99190,
      "token_acc": 0.5224913494809689,
      "train_speed(iter/s)": 1.438134
    },
    {
      "epoch": 4.249817916970138,
      "grad_norm": 5.986358165740967,
      "learning_rate": 5.452271047204582e-06,
      "loss": 2.1992698669433595,
      "memory(GiB)": 77.56,
      "step": 99195,
      "token_acc": 0.5631399317406144,
      "train_speed(iter/s)": 1.438152
    },
    {
      "epoch": 4.2500321322993875,
      "grad_norm": 6.667753219604492,
      "learning_rate": 5.449215511706907e-06,
      "loss": 2.2963939666748048,
      "memory(GiB)": 77.56,
      "step": 99200,
      "token_acc": 0.49169435215946844,
      "train_speed(iter/s)": 1.438165
    },
    {
      "epoch": 4.250246347628636,
      "grad_norm": 5.870570182800293,
      "learning_rate": 5.4461607832884895e-06,
      "loss": 2.5898883819580076,
      "memory(GiB)": 77.56,
      "step": 99205,
      "token_acc": 0.4813664596273292,
      "train_speed(iter/s)": 1.43815
    },
    {
      "epoch": 4.250460562957885,
      "grad_norm": 7.62021017074585,
      "learning_rate": 5.443106862004666e-06,
      "loss": 2.0108711242675783,
      "memory(GiB)": 77.56,
      "step": 99210,
      "token_acc": 0.5207667731629393,
      "train_speed(iter/s)": 1.438173
    },
    {
      "epoch": 4.250674778287134,
      "grad_norm": 5.2923808097839355,
      "learning_rate": 5.440053747910751e-06,
      "loss": 2.1698883056640623,
      "memory(GiB)": 77.56,
      "step": 99215,
      "token_acc": 0.5225563909774437,
      "train_speed(iter/s)": 1.438167
    },
    {
      "epoch": 4.250888993616384,
      "grad_norm": 5.813499927520752,
      "learning_rate": 5.4370014410620674e-06,
      "loss": 2.0973180770874023,
      "memory(GiB)": 77.56,
      "step": 99220,
      "token_acc": 0.5283018867924528,
      "train_speed(iter/s)": 1.438171
    },
    {
      "epoch": 4.251103208945632,
      "grad_norm": 6.895333290100098,
      "learning_rate": 5.4339499415139105e-06,
      "loss": 2.4307680130004883,
      "memory(GiB)": 77.56,
      "step": 99225,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438184
    },
    {
      "epoch": 4.251317424274881,
      "grad_norm": 6.923884391784668,
      "learning_rate": 5.4308992493215414e-06,
      "loss": 2.2147375106811524,
      "memory(GiB)": 77.56,
      "step": 99230,
      "token_acc": 0.5475285171102662,
      "train_speed(iter/s)": 1.438191
    },
    {
      "epoch": 4.25153163960413,
      "grad_norm": 7.689892768859863,
      "learning_rate": 5.427849364540255e-06,
      "loss": 2.25538330078125,
      "memory(GiB)": 77.56,
      "step": 99235,
      "token_acc": 0.511864406779661,
      "train_speed(iter/s)": 1.438182
    },
    {
      "epoch": 4.251745854933379,
      "grad_norm": 9.414280891418457,
      "learning_rate": 5.4248002872252904e-06,
      "loss": 2.115106773376465,
      "memory(GiB)": 77.56,
      "step": 99240,
      "token_acc": 0.5375494071146245,
      "train_speed(iter/s)": 1.438196
    },
    {
      "epoch": 4.251960070262628,
      "grad_norm": 6.857348918914795,
      "learning_rate": 5.42175201743188e-06,
      "loss": 2.075666809082031,
      "memory(GiB)": 77.56,
      "step": 99245,
      "token_acc": 0.5359477124183006,
      "train_speed(iter/s)": 1.438206
    },
    {
      "epoch": 4.252174285591877,
      "grad_norm": 5.502389430999756,
      "learning_rate": 5.418704555215243e-06,
      "loss": 2.500968360900879,
      "memory(GiB)": 77.56,
      "step": 99250,
      "token_acc": 0.48757763975155277,
      "train_speed(iter/s)": 1.43821
    },
    {
      "epoch": 4.252388500921126,
      "grad_norm": 6.886981010437012,
      "learning_rate": 5.415657900630605e-06,
      "loss": 2.1749717712402346,
      "memory(GiB)": 77.56,
      "step": 99255,
      "token_acc": 0.545774647887324,
      "train_speed(iter/s)": 1.438215
    },
    {
      "epoch": 4.252602716250375,
      "grad_norm": 6.770241737365723,
      "learning_rate": 5.412612053733146e-06,
      "loss": 2.2678466796875,
      "memory(GiB)": 77.56,
      "step": 99260,
      "token_acc": 0.49433962264150944,
      "train_speed(iter/s)": 1.438228
    },
    {
      "epoch": 4.252816931579623,
      "grad_norm": 8.066960334777832,
      "learning_rate": 5.409567014578043e-06,
      "loss": 2.125078582763672,
      "memory(GiB)": 77.56,
      "step": 99265,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.438237
    },
    {
      "epoch": 4.253031146908873,
      "grad_norm": 6.868201732635498,
      "learning_rate": 5.406522783220463e-06,
      "loss": 2.0965816497802736,
      "memory(GiB)": 77.56,
      "step": 99270,
      "token_acc": 0.5132743362831859,
      "train_speed(iter/s)": 1.43825
    },
    {
      "epoch": 4.253245362238122,
      "grad_norm": 6.25499963760376,
      "learning_rate": 5.403479359715552e-06,
      "loss": 2.4377157211303713,
      "memory(GiB)": 77.56,
      "step": 99275,
      "token_acc": 0.48554913294797686,
      "train_speed(iter/s)": 1.438259
    },
    {
      "epoch": 4.253459577567371,
      "grad_norm": 9.412491798400879,
      "learning_rate": 5.400436744118448e-06,
      "loss": 2.1824052810668944,
      "memory(GiB)": 77.56,
      "step": 99280,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.438257
    },
    {
      "epoch": 4.2536737928966195,
      "grad_norm": 5.684455394744873,
      "learning_rate": 5.397394936484274e-06,
      "loss": 2.156208801269531,
      "memory(GiB)": 77.56,
      "step": 99285,
      "token_acc": 0.5451388888888888,
      "train_speed(iter/s)": 1.438245
    },
    {
      "epoch": 4.253888008225869,
      "grad_norm": 5.2482476234436035,
      "learning_rate": 5.394353936868135e-06,
      "loss": 2.405049514770508,
      "memory(GiB)": 77.56,
      "step": 99290,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438257
    },
    {
      "epoch": 4.254102223555117,
      "grad_norm": 7.334107875823975,
      "learning_rate": 5.391313745325116e-06,
      "loss": 2.417220687866211,
      "memory(GiB)": 77.56,
      "step": 99295,
      "token_acc": 0.4619883040935672,
      "train_speed(iter/s)": 1.438267
    },
    {
      "epoch": 4.254316438884366,
      "grad_norm": 7.12433385848999,
      "learning_rate": 5.3882743619102786e-06,
      "loss": 2.387996482849121,
      "memory(GiB)": 77.56,
      "step": 99300,
      "token_acc": 0.5205479452054794,
      "train_speed(iter/s)": 1.438276
    },
    {
      "epoch": 4.254530654213616,
      "grad_norm": 7.295494556427002,
      "learning_rate": 5.385235786678716e-06,
      "loss": 2.5620138168334963,
      "memory(GiB)": 77.56,
      "step": 99305,
      "token_acc": 0.49466192170818507,
      "train_speed(iter/s)": 1.438286
    },
    {
      "epoch": 4.254744869542865,
      "grad_norm": 6.653997421264648,
      "learning_rate": 5.3821980196854475e-06,
      "loss": 2.3662322998046874,
      "memory(GiB)": 77.56,
      "step": 99310,
      "token_acc": 0.4907749077490775,
      "train_speed(iter/s)": 1.43828
    },
    {
      "epoch": 4.254959084872113,
      "grad_norm": 9.474124908447266,
      "learning_rate": 5.379161060985533e-06,
      "loss": 1.9860734939575195,
      "memory(GiB)": 77.56,
      "step": 99315,
      "token_acc": 0.5797665369649806,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 4.2551733002013625,
      "grad_norm": 5.496501922607422,
      "learning_rate": 5.376124910633967e-06,
      "loss": 1.9740488052368164,
      "memory(GiB)": 77.56,
      "step": 99320,
      "token_acc": 0.5180327868852459,
      "train_speed(iter/s)": 1.438286
    },
    {
      "epoch": 4.255387515530612,
      "grad_norm": 5.606962203979492,
      "learning_rate": 5.373089568685758e-06,
      "loss": 2.6144222259521483,
      "memory(GiB)": 77.56,
      "step": 99325,
      "token_acc": 0.44594594594594594,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 4.25560173085986,
      "grad_norm": 6.997704029083252,
      "learning_rate": 5.370055035195898e-06,
      "loss": 2.456332969665527,
      "memory(GiB)": 77.56,
      "step": 99330,
      "token_acc": 0.48830409356725146,
      "train_speed(iter/s)": 1.438308
    },
    {
      "epoch": 4.255815946189109,
      "grad_norm": 7.725266933441162,
      "learning_rate": 5.367021310219344e-06,
      "loss": 2.530974769592285,
      "memory(GiB)": 77.56,
      "step": 99335,
      "token_acc": 0.5074626865671642,
      "train_speed(iter/s)": 1.438306
    },
    {
      "epoch": 4.256030161518359,
      "grad_norm": 10.022805213928223,
      "learning_rate": 5.363988393811081e-06,
      "loss": 1.9978378295898438,
      "memory(GiB)": 77.56,
      "step": 99340,
      "token_acc": 0.5540540540540541,
      "train_speed(iter/s)": 1.438279
    },
    {
      "epoch": 4.256244376847607,
      "grad_norm": 7.094108581542969,
      "learning_rate": 5.360956286026042e-06,
      "loss": 2.327950859069824,
      "memory(GiB)": 77.56,
      "step": 99345,
      "token_acc": 0.540268456375839,
      "train_speed(iter/s)": 1.438294
    },
    {
      "epoch": 4.256458592176856,
      "grad_norm": 4.85783052444458,
      "learning_rate": 5.357924986919149e-06,
      "loss": 1.9933506011962892,
      "memory(GiB)": 77.56,
      "step": 99350,
      "token_acc": 0.5275590551181102,
      "train_speed(iter/s)": 1.438302
    },
    {
      "epoch": 4.2566728075061055,
      "grad_norm": 6.5165534019470215,
      "learning_rate": 5.354894496545326e-06,
      "loss": 2.5928401947021484,
      "memory(GiB)": 77.56,
      "step": 99355,
      "token_acc": 0.4441340782122905,
      "train_speed(iter/s)": 1.438301
    },
    {
      "epoch": 4.256887022835354,
      "grad_norm": 5.5560455322265625,
      "learning_rate": 5.351864814959462e-06,
      "loss": 2.4872867584228517,
      "memory(GiB)": 77.56,
      "step": 99360,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438303
    },
    {
      "epoch": 4.257101238164603,
      "grad_norm": 7.415276527404785,
      "learning_rate": 5.348835942216457e-06,
      "loss": 2.339769744873047,
      "memory(GiB)": 77.56,
      "step": 99365,
      "token_acc": 0.5576208178438662,
      "train_speed(iter/s)": 1.438312
    },
    {
      "epoch": 4.257315453493852,
      "grad_norm": 5.808889389038086,
      "learning_rate": 5.345807878371173e-06,
      "loss": 2.033720016479492,
      "memory(GiB)": 77.56,
      "step": 99370,
      "token_acc": 0.5274725274725275,
      "train_speed(iter/s)": 1.438325
    },
    {
      "epoch": 4.257529668823101,
      "grad_norm": 7.288916110992432,
      "learning_rate": 5.3427806234784606e-06,
      "loss": 2.2452007293701173,
      "memory(GiB)": 77.56,
      "step": 99375,
      "token_acc": 0.49393939393939396,
      "train_speed(iter/s)": 1.438321
    },
    {
      "epoch": 4.25774388415235,
      "grad_norm": 7.555299282073975,
      "learning_rate": 5.339754177593182e-06,
      "loss": 2.0741466522216796,
      "memory(GiB)": 77.56,
      "step": 99380,
      "token_acc": 0.5421686746987951,
      "train_speed(iter/s)": 1.438323
    },
    {
      "epoch": 4.257958099481599,
      "grad_norm": 7.6305670738220215,
      "learning_rate": 5.336728540770148e-06,
      "loss": 2.0681602478027346,
      "memory(GiB)": 77.56,
      "step": 99385,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 4.258172314810848,
      "grad_norm": 5.377715587615967,
      "learning_rate": 5.333703713064175e-06,
      "loss": 2.3758419036865233,
      "memory(GiB)": 77.56,
      "step": 99390,
      "token_acc": 0.5374592833876222,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 4.258386530140097,
      "grad_norm": 7.726016998291016,
      "learning_rate": 5.330679694530049e-06,
      "loss": 2.2782661437988283,
      "memory(GiB)": 77.56,
      "step": 99395,
      "token_acc": 0.4984025559105431,
      "train_speed(iter/s)": 1.438324
    },
    {
      "epoch": 4.258600745469346,
      "grad_norm": 6.461470127105713,
      "learning_rate": 5.327656485222576e-06,
      "loss": 2.293054389953613,
      "memory(GiB)": 77.56,
      "step": 99400,
      "token_acc": 0.5392857142857143,
      "train_speed(iter/s)": 1.43833
    },
    {
      "epoch": 4.258814960798595,
      "grad_norm": 5.0221076011657715,
      "learning_rate": 5.324634085196506e-06,
      "loss": 2.5675674438476563,
      "memory(GiB)": 77.56,
      "step": 99405,
      "token_acc": 0.4752475247524752,
      "train_speed(iter/s)": 1.438325
    },
    {
      "epoch": 4.259029176127844,
      "grad_norm": 7.002190589904785,
      "learning_rate": 5.321612494506606e-06,
      "loss": 2.286764144897461,
      "memory(GiB)": 77.56,
      "step": 99410,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.438327
    },
    {
      "epoch": 4.259243391457093,
      "grad_norm": 7.029500961303711,
      "learning_rate": 5.318591713207599e-06,
      "loss": 2.337006759643555,
      "memory(GiB)": 77.56,
      "step": 99415,
      "token_acc": 0.5364963503649635,
      "train_speed(iter/s)": 1.438332
    },
    {
      "epoch": 4.259457606786341,
      "grad_norm": 5.700656414031982,
      "learning_rate": 5.315571741354214e-06,
      "loss": 2.400377655029297,
      "memory(GiB)": 77.56,
      "step": 99420,
      "token_acc": 0.45714285714285713,
      "train_speed(iter/s)": 1.438333
    },
    {
      "epoch": 4.259671822115591,
      "grad_norm": 5.522071838378906,
      "learning_rate": 5.312552579001173e-06,
      "loss": 2.3048778533935548,
      "memory(GiB)": 77.56,
      "step": 99425,
      "token_acc": 0.5310077519379846,
      "train_speed(iter/s)": 1.438343
    },
    {
      "epoch": 4.25988603744484,
      "grad_norm": 6.354565143585205,
      "learning_rate": 5.309534226203161e-06,
      "loss": 2.4640254974365234,
      "memory(GiB)": 77.56,
      "step": 99430,
      "token_acc": 0.5075987841945289,
      "train_speed(iter/s)": 1.438337
    },
    {
      "epoch": 4.260100252774088,
      "grad_norm": 5.910831451416016,
      "learning_rate": 5.306516683014862e-06,
      "loss": 2.3290300369262695,
      "memory(GiB)": 77.56,
      "step": 99435,
      "token_acc": 0.4899135446685879,
      "train_speed(iter/s)": 1.438323
    },
    {
      "epoch": 4.260314468103338,
      "grad_norm": 6.572525978088379,
      "learning_rate": 5.303499949490937e-06,
      "loss": 2.138357925415039,
      "memory(GiB)": 77.56,
      "step": 99440,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.438321
    },
    {
      "epoch": 4.260528683432587,
      "grad_norm": 6.1764235496521,
      "learning_rate": 5.300484025686037e-06,
      "loss": 2.2882753372192384,
      "memory(GiB)": 77.56,
      "step": 99445,
      "token_acc": 0.5261194029850746,
      "train_speed(iter/s)": 1.438326
    },
    {
      "epoch": 4.260742898761835,
      "grad_norm": 6.314894199371338,
      "learning_rate": 5.297468911654796e-06,
      "loss": 2.247754669189453,
      "memory(GiB)": 77.56,
      "step": 99450,
      "token_acc": 0.5164473684210527,
      "train_speed(iter/s)": 1.438326
    },
    {
      "epoch": 4.2609571140910845,
      "grad_norm": 6.027066707611084,
      "learning_rate": 5.294454607451838e-06,
      "loss": 2.119722366333008,
      "memory(GiB)": 77.56,
      "step": 99455,
      "token_acc": 0.523972602739726,
      "train_speed(iter/s)": 1.438303
    },
    {
      "epoch": 4.261171329420334,
      "grad_norm": 6.138664245605469,
      "learning_rate": 5.291441113131779e-06,
      "loss": 1.921840286254883,
      "memory(GiB)": 77.56,
      "step": 99460,
      "token_acc": 0.575187969924812,
      "train_speed(iter/s)": 1.438305
    },
    {
      "epoch": 4.261385544749582,
      "grad_norm": 6.930879592895508,
      "learning_rate": 5.2884284287492034e-06,
      "loss": 2.1499303817749023,
      "memory(GiB)": 77.56,
      "step": 99465,
      "token_acc": 0.5457627118644067,
      "train_speed(iter/s)": 1.438311
    },
    {
      "epoch": 4.261599760078831,
      "grad_norm": 5.446767330169678,
      "learning_rate": 5.285416554358696e-06,
      "loss": 2.422323799133301,
      "memory(GiB)": 77.56,
      "step": 99470,
      "token_acc": 0.5078125,
      "train_speed(iter/s)": 1.438318
    },
    {
      "epoch": 4.261813975408081,
      "grad_norm": 7.983272075653076,
      "learning_rate": 5.282405490014808e-06,
      "loss": 2.199492645263672,
      "memory(GiB)": 77.56,
      "step": 99475,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438315
    },
    {
      "epoch": 4.262028190737329,
      "grad_norm": 5.806427001953125,
      "learning_rate": 5.279395235772084e-06,
      "loss": 2.2700372695922852,
      "memory(GiB)": 77.56,
      "step": 99480,
      "token_acc": 0.5454545454545454,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 4.262242406066578,
      "grad_norm": 5.768808364868164,
      "learning_rate": 5.276385791685079e-06,
      "loss": 2.4191593170166015,
      "memory(GiB)": 77.56,
      "step": 99485,
      "token_acc": 0.4808259587020649,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 4.2624566213958275,
      "grad_norm": 5.224371433258057,
      "learning_rate": 5.273377157808296e-06,
      "loss": 2.337211799621582,
      "memory(GiB)": 77.56,
      "step": 99490,
      "token_acc": 0.5056179775280899,
      "train_speed(iter/s)": 1.438322
    },
    {
      "epoch": 4.262670836725076,
      "grad_norm": 5.661655426025391,
      "learning_rate": 5.270369334196246e-06,
      "loss": 2.1746469497680665,
      "memory(GiB)": 77.56,
      "step": 99495,
      "token_acc": 0.5447154471544715,
      "train_speed(iter/s)": 1.438317
    },
    {
      "epoch": 4.262885052054325,
      "grad_norm": 6.167550086975098,
      "learning_rate": 5.267362320903413e-06,
      "loss": 2.273846244812012,
      "memory(GiB)": 77.56,
      "step": 99500,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.438327
    },
    {
      "epoch": 4.262885052054325,
      "eval_loss": 2.259369373321533,
      "eval_runtime": 14.0799,
      "eval_samples_per_second": 7.102,
      "eval_steps_per_second": 7.102,
      "eval_token_acc": 0.4695290858725762,
      "step": 99500
    },
    {
      "epoch": 4.263099267383574,
      "grad_norm": 5.5677571296691895,
      "learning_rate": 5.264356117984265e-06,
      "loss": 2.3016624450683594,
      "memory(GiB)": 77.56,
      "step": 99505,
      "token_acc": 0.49193548387096775,
      "train_speed(iter/s)": 1.43803
    },
    {
      "epoch": 4.263313482712823,
      "grad_norm": 6.390848159790039,
      "learning_rate": 5.261350725493286e-06,
      "loss": 2.1338474273681642,
      "memory(GiB)": 77.56,
      "step": 99510,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.43802
    },
    {
      "epoch": 4.263527698042072,
      "grad_norm": 7.957448959350586,
      "learning_rate": 5.258346143484899e-06,
      "loss": 2.2904277801513673,
      "memory(GiB)": 77.56,
      "step": 99515,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.438023
    },
    {
      "epoch": 4.263741913371321,
      "grad_norm": 6.106991291046143,
      "learning_rate": 5.2553423720135494e-06,
      "loss": 2.507122611999512,
      "memory(GiB)": 77.56,
      "step": 99520,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.438021
    },
    {
      "epoch": 4.26395612870057,
      "grad_norm": 5.650023460388184,
      "learning_rate": 5.2523394111336334e-06,
      "loss": 2.479850006103516,
      "memory(GiB)": 77.56,
      "step": 99525,
      "token_acc": 0.47601476014760147,
      "train_speed(iter/s)": 1.438014
    },
    {
      "epoch": 4.264170344029819,
      "grad_norm": 5.992207050323486,
      "learning_rate": 5.249337260899573e-06,
      "loss": 2.4205209732055666,
      "memory(GiB)": 77.56,
      "step": 99530,
      "token_acc": 0.494949494949495,
      "train_speed(iter/s)": 1.438026
    },
    {
      "epoch": 4.264384559359068,
      "grad_norm": 6.949053764343262,
      "learning_rate": 5.2463359213657525e-06,
      "loss": 2.3510799407958984,
      "memory(GiB)": 77.56,
      "step": 99535,
      "token_acc": 0.506993006993007,
      "train_speed(iter/s)": 1.438037
    },
    {
      "epoch": 4.2645987746883165,
      "grad_norm": 5.720202445983887,
      "learning_rate": 5.243335392586524e-06,
      "loss": 2.0583702087402345,
      "memory(GiB)": 77.56,
      "step": 99540,
      "token_acc": 0.5466666666666666,
      "train_speed(iter/s)": 1.438042
    },
    {
      "epoch": 4.264812990017566,
      "grad_norm": 6.283769607543945,
      "learning_rate": 5.240335674616276e-06,
      "loss": 2.2413457870483398,
      "memory(GiB)": 77.56,
      "step": 99545,
      "token_acc": 0.5437262357414449,
      "train_speed(iter/s)": 1.438047
    },
    {
      "epoch": 4.265027205346815,
      "grad_norm": 6.344459056854248,
      "learning_rate": 5.237336767509332e-06,
      "loss": 2.4375762939453125,
      "memory(GiB)": 77.56,
      "step": 99550,
      "token_acc": 0.4772727272727273,
      "train_speed(iter/s)": 1.438053
    },
    {
      "epoch": 4.265241420676063,
      "grad_norm": 5.999893665313721,
      "learning_rate": 5.234338671320021e-06,
      "loss": 2.261827087402344,
      "memory(GiB)": 77.56,
      "step": 99555,
      "token_acc": 0.5031847133757962,
      "train_speed(iter/s)": 1.438065
    },
    {
      "epoch": 4.265455636005313,
      "grad_norm": 6.464510440826416,
      "learning_rate": 5.2313413861026614e-06,
      "loss": 2.314762496948242,
      "memory(GiB)": 77.56,
      "step": 99560,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.438079
    },
    {
      "epoch": 4.265669851334562,
      "grad_norm": 6.643327236175537,
      "learning_rate": 5.2283449119115366e-06,
      "loss": 2.312142753601074,
      "memory(GiB)": 77.56,
      "step": 99565,
      "token_acc": 0.5225225225225225,
      "train_speed(iter/s)": 1.438094
    },
    {
      "epoch": 4.26588406666381,
      "grad_norm": 8.836644172668457,
      "learning_rate": 5.225349248800954e-06,
      "loss": 2.2041854858398438,
      "memory(GiB)": 77.56,
      "step": 99570,
      "token_acc": 0.5218855218855218,
      "train_speed(iter/s)": 1.438101
    },
    {
      "epoch": 4.2660982819930595,
      "grad_norm": 5.8553242683410645,
      "learning_rate": 5.222354396825174e-06,
      "loss": 2.2258806228637695,
      "memory(GiB)": 77.56,
      "step": 99575,
      "token_acc": 0.48942598187311176,
      "train_speed(iter/s)": 1.438115
    },
    {
      "epoch": 4.266312497322309,
      "grad_norm": 5.980251312255859,
      "learning_rate": 5.2193603560384495e-06,
      "loss": 2.4428539276123047,
      "memory(GiB)": 77.56,
      "step": 99580,
      "token_acc": 0.5227272727272727,
      "train_speed(iter/s)": 1.438121
    },
    {
      "epoch": 4.266526712651557,
      "grad_norm": 6.075314044952393,
      "learning_rate": 5.216367126495015e-06,
      "loss": 2.2006053924560547,
      "memory(GiB)": 77.56,
      "step": 99585,
      "token_acc": 0.5625,
      "train_speed(iter/s)": 1.438121
    },
    {
      "epoch": 4.266740927980806,
      "grad_norm": 5.85908842086792,
      "learning_rate": 5.213374708249097e-06,
      "loss": 2.2895736694335938,
      "memory(GiB)": 77.56,
      "step": 99590,
      "token_acc": 0.4984520123839009,
      "train_speed(iter/s)": 1.43813
    },
    {
      "epoch": 4.266955143310056,
      "grad_norm": 7.5717854499816895,
      "learning_rate": 5.210383101354915e-06,
      "loss": 2.5901809692382813,
      "memory(GiB)": 77.56,
      "step": 99595,
      "token_acc": 0.4503105590062112,
      "train_speed(iter/s)": 1.438129
    },
    {
      "epoch": 4.267169358639304,
      "grad_norm": 7.153294563293457,
      "learning_rate": 5.207392305866648e-06,
      "loss": 2.409659194946289,
      "memory(GiB)": 77.56,
      "step": 99600,
      "token_acc": 0.4935897435897436,
      "train_speed(iter/s)": 1.438133
    },
    {
      "epoch": 4.267383573968553,
      "grad_norm": 8.248741149902344,
      "learning_rate": 5.204402321838503e-06,
      "loss": 2.3266618728637694,
      "memory(GiB)": 77.56,
      "step": 99605,
      "token_acc": 0.5140845070422535,
      "train_speed(iter/s)": 1.438144
    },
    {
      "epoch": 4.2675977892978025,
      "grad_norm": 6.237697124481201,
      "learning_rate": 5.2014131493246246e-06,
      "loss": 2.298316764831543,
      "memory(GiB)": 77.56,
      "step": 99610,
      "token_acc": 0.48201438848920863,
      "train_speed(iter/s)": 1.438138
    },
    {
      "epoch": 4.267812004627051,
      "grad_norm": 6.186761856079102,
      "learning_rate": 5.198424788379175e-06,
      "loss": 2.2674753189086916,
      "memory(GiB)": 77.56,
      "step": 99615,
      "token_acc": 0.49842271293375395,
      "train_speed(iter/s)": 1.43814
    },
    {
      "epoch": 4.2680262199563,
      "grad_norm": 5.699726104736328,
      "learning_rate": 5.195437239056289e-06,
      "loss": 2.4532527923583984,
      "memory(GiB)": 77.56,
      "step": 99620,
      "token_acc": 0.46814404432132967,
      "train_speed(iter/s)": 1.438153
    },
    {
      "epoch": 4.268240435285549,
      "grad_norm": 8.543010711669922,
      "learning_rate": 5.192450501410079e-06,
      "loss": 2.053589630126953,
      "memory(GiB)": 77.56,
      "step": 99625,
      "token_acc": 0.5792880258899676,
      "train_speed(iter/s)": 1.438157
    },
    {
      "epoch": 4.268454650614798,
      "grad_norm": 6.113498210906982,
      "learning_rate": 5.189464575494668e-06,
      "loss": 2.100913429260254,
      "memory(GiB)": 77.56,
      "step": 99630,
      "token_acc": 0.5437956204379562,
      "train_speed(iter/s)": 1.438163
    },
    {
      "epoch": 4.268668865944047,
      "grad_norm": 6.139715671539307,
      "learning_rate": 5.186479461364141e-06,
      "loss": 2.4362077713012695,
      "memory(GiB)": 77.56,
      "step": 99635,
      "token_acc": 0.48951048951048953,
      "train_speed(iter/s)": 1.438176
    },
    {
      "epoch": 4.268883081273296,
      "grad_norm": 6.840907573699951,
      "learning_rate": 5.183495159072583e-06,
      "loss": 2.3649318695068358,
      "memory(GiB)": 77.56,
      "step": 99640,
      "token_acc": 0.5551839464882943,
      "train_speed(iter/s)": 1.438193
    },
    {
      "epoch": 4.269097296602545,
      "grad_norm": 7.1368560791015625,
      "learning_rate": 5.180511668674043e-06,
      "loss": 2.3247711181640627,
      "memory(GiB)": 77.56,
      "step": 99645,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.438179
    },
    {
      "epoch": 4.269311511931794,
      "grad_norm": 10.226576805114746,
      "learning_rate": 5.177528990222574e-06,
      "loss": 2.3273765563964846,
      "memory(GiB)": 77.56,
      "step": 99650,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438191
    },
    {
      "epoch": 4.269525727261043,
      "grad_norm": 6.319547176361084,
      "learning_rate": 5.174547123772222e-06,
      "loss": 2.251133346557617,
      "memory(GiB)": 77.56,
      "step": 99655,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438192
    },
    {
      "epoch": 4.2697399425902915,
      "grad_norm": 7.835252285003662,
      "learning_rate": 5.1715660693769985e-06,
      "loss": 2.324872589111328,
      "memory(GiB)": 77.56,
      "step": 99660,
      "token_acc": 0.5252100840336135,
      "train_speed(iter/s)": 1.438204
    },
    {
      "epoch": 4.269954157919541,
      "grad_norm": 5.754541873931885,
      "learning_rate": 5.168585827090911e-06,
      "loss": 2.4149904251098633,
      "memory(GiB)": 77.56,
      "step": 99665,
      "token_acc": 0.47151898734177217,
      "train_speed(iter/s)": 1.438215
    },
    {
      "epoch": 4.27016837324879,
      "grad_norm": 10.338605880737305,
      "learning_rate": 5.165606396967932e-06,
      "loss": 2.2170360565185545,
      "memory(GiB)": 77.56,
      "step": 99670,
      "token_acc": 0.5296296296296297,
      "train_speed(iter/s)": 1.438225
    },
    {
      "epoch": 4.270382588578038,
      "grad_norm": 8.407870292663574,
      "learning_rate": 5.1626277790620625e-06,
      "loss": 2.248361015319824,
      "memory(GiB)": 77.56,
      "step": 99675,
      "token_acc": 0.5071428571428571,
      "train_speed(iter/s)": 1.438236
    },
    {
      "epoch": 4.270596803907288,
      "grad_norm": 7.160726070404053,
      "learning_rate": 5.159649973427255e-06,
      "loss": 2.2546363830566407,
      "memory(GiB)": 77.56,
      "step": 99680,
      "token_acc": 0.5387596899224806,
      "train_speed(iter/s)": 1.438224
    },
    {
      "epoch": 4.270811019236537,
      "grad_norm": 5.566939830780029,
      "learning_rate": 5.1566729801174385e-06,
      "loss": 2.351669120788574,
      "memory(GiB)": 77.56,
      "step": 99685,
      "token_acc": 0.5227272727272727,
      "train_speed(iter/s)": 1.438222
    },
    {
      "epoch": 4.271025234565785,
      "grad_norm": 6.521866798400879,
      "learning_rate": 5.153696799186569e-06,
      "loss": 2.3032325744628905,
      "memory(GiB)": 77.56,
      "step": 99690,
      "token_acc": 0.49328859060402686,
      "train_speed(iter/s)": 1.438222
    },
    {
      "epoch": 4.271239449895035,
      "grad_norm": 5.597241401672363,
      "learning_rate": 5.1507214306885485e-06,
      "loss": 2.2157379150390626,
      "memory(GiB)": 77.56,
      "step": 99695,
      "token_acc": 0.5017064846416383,
      "train_speed(iter/s)": 1.438209
    },
    {
      "epoch": 4.271453665224284,
      "grad_norm": 8.706413269042969,
      "learning_rate": 5.147746874677284e-06,
      "loss": 2.3960054397583006,
      "memory(GiB)": 77.56,
      "step": 99700,
      "token_acc": 0.5019011406844106,
      "train_speed(iter/s)": 1.438216
    },
    {
      "epoch": 4.271667880553532,
      "grad_norm": 6.453300952911377,
      "learning_rate": 5.144773131206659e-06,
      "loss": 2.070814514160156,
      "memory(GiB)": 77.56,
      "step": 99705,
      "token_acc": 0.5404411764705882,
      "train_speed(iter/s)": 1.438211
    },
    {
      "epoch": 4.2718820958827814,
      "grad_norm": 6.497753143310547,
      "learning_rate": 5.141800200330538e-06,
      "loss": 2.2009424209594726,
      "memory(GiB)": 77.56,
      "step": 99710,
      "token_acc": 0.5041322314049587,
      "train_speed(iter/s)": 1.438226
    },
    {
      "epoch": 4.272096311212031,
      "grad_norm": 5.68049430847168,
      "learning_rate": 5.138828082102792e-06,
      "loss": 2.321919822692871,
      "memory(GiB)": 77.56,
      "step": 99715,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.438223
    },
    {
      "epoch": 4.272310526541279,
      "grad_norm": 5.51497220993042,
      "learning_rate": 5.135856776577263e-06,
      "loss": 1.954019546508789,
      "memory(GiB)": 77.56,
      "step": 99720,
      "token_acc": 0.5350553505535055,
      "train_speed(iter/s)": 1.438235
    },
    {
      "epoch": 4.272524741870528,
      "grad_norm": 5.677099227905273,
      "learning_rate": 5.1328862838077754e-06,
      "loss": 2.156208801269531,
      "memory(GiB)": 77.56,
      "step": 99725,
      "token_acc": 0.5297619047619048,
      "train_speed(iter/s)": 1.438239
    },
    {
      "epoch": 4.272738957199778,
      "grad_norm": 7.19480037689209,
      "learning_rate": 5.129916603848139e-06,
      "loss": 2.5902456283569335,
      "memory(GiB)": 77.56,
      "step": 99730,
      "token_acc": 0.48854961832061067,
      "train_speed(iter/s)": 1.438242
    },
    {
      "epoch": 4.272953172529026,
      "grad_norm": 8.099814414978027,
      "learning_rate": 5.126947736752141e-06,
      "loss": 2.340462303161621,
      "memory(GiB)": 77.56,
      "step": 99735,
      "token_acc": 0.5370370370370371,
      "train_speed(iter/s)": 1.438259
    },
    {
      "epoch": 4.273167387858275,
      "grad_norm": 4.437596321105957,
      "learning_rate": 5.123979682573599e-06,
      "loss": 2.2372098922729493,
      "memory(GiB)": 77.56,
      "step": 99740,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.438264
    },
    {
      "epoch": 4.2733816031875245,
      "grad_norm": 5.398176193237305,
      "learning_rate": 5.12101244136624e-06,
      "loss": 2.361555290222168,
      "memory(GiB)": 77.56,
      "step": 99745,
      "token_acc": 0.5288461538461539,
      "train_speed(iter/s)": 1.43828
    },
    {
      "epoch": 4.273595818516773,
      "grad_norm": 6.621888637542725,
      "learning_rate": 5.118046013183858e-06,
      "loss": 2.6222583770751955,
      "memory(GiB)": 77.56,
      "step": 99750,
      "token_acc": 0.4788732394366197,
      "train_speed(iter/s)": 1.438284
    },
    {
      "epoch": 4.273810033846022,
      "grad_norm": 6.522944927215576,
      "learning_rate": 5.115080398080174e-06,
      "loss": 2.205311584472656,
      "memory(GiB)": 77.56,
      "step": 99755,
      "token_acc": 0.5215827338129496,
      "train_speed(iter/s)": 1.438284
    },
    {
      "epoch": 4.274024249175271,
      "grad_norm": 5.84501838684082,
      "learning_rate": 5.112115596108918e-06,
      "loss": 2.0059930801391603,
      "memory(GiB)": 77.56,
      "step": 99760,
      "token_acc": 0.5326460481099656,
      "train_speed(iter/s)": 1.438284
    },
    {
      "epoch": 4.27423846450452,
      "grad_norm": 5.14306116104126,
      "learning_rate": 5.109151607323792e-06,
      "loss": 1.9309453964233398,
      "memory(GiB)": 77.56,
      "step": 99765,
      "token_acc": 0.588,
      "train_speed(iter/s)": 1.438288
    },
    {
      "epoch": 4.274452679833769,
      "grad_norm": 6.364627361297607,
      "learning_rate": 5.1061884317784855e-06,
      "loss": 2.3918811798095705,
      "memory(GiB)": 77.56,
      "step": 99770,
      "token_acc": 0.49158249158249157,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 4.274666895163018,
      "grad_norm": 7.238813877105713,
      "learning_rate": 5.103226069526701e-06,
      "loss": 2.0740615844726564,
      "memory(GiB)": 77.56,
      "step": 99775,
      "token_acc": 0.5627705627705628,
      "train_speed(iter/s)": 1.438284
    },
    {
      "epoch": 4.274881110492267,
      "grad_norm": 6.310179233551025,
      "learning_rate": 5.100264520622089e-06,
      "loss": 2.251984405517578,
      "memory(GiB)": 77.56,
      "step": 99780,
      "token_acc": 0.5387205387205387,
      "train_speed(iter/s)": 1.438312
    },
    {
      "epoch": 4.275095325821516,
      "grad_norm": 5.881538391113281,
      "learning_rate": 5.097303785118307e-06,
      "loss": 2.276052474975586,
      "memory(GiB)": 77.56,
      "step": 99785,
      "token_acc": 0.49019607843137253,
      "train_speed(iter/s)": 1.43831
    },
    {
      "epoch": 4.275309541150765,
      "grad_norm": 6.447880268096924,
      "learning_rate": 5.09434386306899e-06,
      "loss": 2.420423889160156,
      "memory(GiB)": 77.56,
      "step": 99790,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.438302
    },
    {
      "epoch": 4.2755237564800135,
      "grad_norm": 4.215697765350342,
      "learning_rate": 5.09138475452775e-06,
      "loss": 2.231801223754883,
      "memory(GiB)": 77.56,
      "step": 99795,
      "token_acc": 0.5253164556962026,
      "train_speed(iter/s)": 1.438266
    },
    {
      "epoch": 4.275737971809263,
      "grad_norm": 7.529350757598877,
      "learning_rate": 5.088426459548207e-06,
      "loss": 2.0099105834960938,
      "memory(GiB)": 77.56,
      "step": 99800,
      "token_acc": 0.49224806201550386,
      "train_speed(iter/s)": 1.438277
    },
    {
      "epoch": 4.275952187138512,
      "grad_norm": 11.920438766479492,
      "learning_rate": 5.085468978183955e-06,
      "loss": 2.4065725326538088,
      "memory(GiB)": 77.56,
      "step": 99805,
      "token_acc": 0.46808510638297873,
      "train_speed(iter/s)": 1.438287
    },
    {
      "epoch": 4.27616640246776,
      "grad_norm": 4.699000835418701,
      "learning_rate": 5.082512310488563e-06,
      "loss": 1.7878015518188477,
      "memory(GiB)": 77.56,
      "step": 99810,
      "token_acc": 0.5902777777777778,
      "train_speed(iter/s)": 1.438268
    },
    {
      "epoch": 4.27638061779701,
      "grad_norm": 6.252408027648926,
      "learning_rate": 5.079556456515599e-06,
      "loss": 2.4323404312133787,
      "memory(GiB)": 77.56,
      "step": 99815,
      "token_acc": 0.4633333333333333,
      "train_speed(iter/s)": 1.438267
    },
    {
      "epoch": 4.276594833126259,
      "grad_norm": 4.815014362335205,
      "learning_rate": 5.0766014163185935e-06,
      "loss": 2.3686960220336912,
      "memory(GiB)": 77.56,
      "step": 99820,
      "token_acc": 0.5086505190311419,
      "train_speed(iter/s)": 1.438279
    },
    {
      "epoch": 4.276809048455507,
      "grad_norm": 8.460524559020996,
      "learning_rate": 5.073647189951109e-06,
      "loss": 2.3323667526245115,
      "memory(GiB)": 77.56,
      "step": 99825,
      "token_acc": 0.496,
      "train_speed(iter/s)": 1.438276
    },
    {
      "epoch": 4.2770232637847565,
      "grad_norm": 6.189655303955078,
      "learning_rate": 5.07069377746664e-06,
      "loss": 2.2044851303100588,
      "memory(GiB)": 77.56,
      "step": 99830,
      "token_acc": 0.55859375,
      "train_speed(iter/s)": 1.438289
    },
    {
      "epoch": 4.277237479114006,
      "grad_norm": 7.330394268035889,
      "learning_rate": 5.067741178918711e-06,
      "loss": 2.4030878067016603,
      "memory(GiB)": 77.56,
      "step": 99835,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.4383
    },
    {
      "epoch": 4.277451694443254,
      "grad_norm": 8.211701393127441,
      "learning_rate": 5.0647893943607975e-06,
      "loss": 1.9339447021484375,
      "memory(GiB)": 77.56,
      "step": 99840,
      "token_acc": 0.5652173913043478,
      "train_speed(iter/s)": 1.438315
    },
    {
      "epoch": 4.277665909772503,
      "grad_norm": 5.901163578033447,
      "learning_rate": 5.061838423846377e-06,
      "loss": 2.2432674407958983,
      "memory(GiB)": 77.56,
      "step": 99845,
      "token_acc": 0.5164835164835165,
      "train_speed(iter/s)": 1.438325
    },
    {
      "epoch": 4.277880125101753,
      "grad_norm": 6.125593662261963,
      "learning_rate": 5.058888267428913e-06,
      "loss": 2.3655691146850586,
      "memory(GiB)": 77.56,
      "step": 99850,
      "token_acc": 0.5080385852090032,
      "train_speed(iter/s)": 1.438347
    },
    {
      "epoch": 4.278094340431001,
      "grad_norm": 7.210329532623291,
      "learning_rate": 5.055938925161835e-06,
      "loss": 2.3708343505859375,
      "memory(GiB)": 77.56,
      "step": 99855,
      "token_acc": 0.4660493827160494,
      "train_speed(iter/s)": 1.438351
    },
    {
      "epoch": 4.27830855576025,
      "grad_norm": 6.790513515472412,
      "learning_rate": 5.052990397098589e-06,
      "loss": 2.2558460235595703,
      "memory(GiB)": 77.56,
      "step": 99860,
      "token_acc": 0.47648902821316613,
      "train_speed(iter/s)": 1.43836
    },
    {
      "epoch": 4.2785227710894995,
      "grad_norm": 6.5188798904418945,
      "learning_rate": 5.050042683292589e-06,
      "loss": 2.4967369079589843,
      "memory(GiB)": 77.56,
      "step": 99865,
      "token_acc": 0.49480968858131485,
      "train_speed(iter/s)": 1.43837
    },
    {
      "epoch": 4.278736986418748,
      "grad_norm": 7.214264869689941,
      "learning_rate": 5.047095783797234e-06,
      "loss": 2.3858945846557615,
      "memory(GiB)": 77.56,
      "step": 99870,
      "token_acc": 0.4662379421221865,
      "train_speed(iter/s)": 1.438385
    },
    {
      "epoch": 4.278951201747997,
      "grad_norm": 7.626358509063721,
      "learning_rate": 5.044149698665906e-06,
      "loss": 2.491080474853516,
      "memory(GiB)": 77.56,
      "step": 99875,
      "token_acc": 0.4726027397260274,
      "train_speed(iter/s)": 1.438398
    },
    {
      "epoch": 4.279165417077246,
      "grad_norm": 7.813310146331787,
      "learning_rate": 5.041204427951968e-06,
      "loss": 2.2463138580322264,
      "memory(GiB)": 77.56,
      "step": 99880,
      "token_acc": 0.5541666666666667,
      "train_speed(iter/s)": 1.438414
    },
    {
      "epoch": 4.279379632406495,
      "grad_norm": 7.381002902984619,
      "learning_rate": 5.0382599717087974e-06,
      "loss": 2.4186691284179687,
      "memory(GiB)": 77.56,
      "step": 99885,
      "token_acc": 0.47865853658536583,
      "train_speed(iter/s)": 1.43842
    },
    {
      "epoch": 4.279593847735744,
      "grad_norm": 6.85996675491333,
      "learning_rate": 5.035316329989725e-06,
      "loss": 2.375261688232422,
      "memory(GiB)": 77.56,
      "step": 99890,
      "token_acc": 0.5086505190311419,
      "train_speed(iter/s)": 1.438439
    },
    {
      "epoch": 4.279808063064993,
      "grad_norm": 6.606855392456055,
      "learning_rate": 5.03237350284807e-06,
      "loss": 2.1992351531982424,
      "memory(GiB)": 77.56,
      "step": 99895,
      "token_acc": 0.5296167247386759,
      "train_speed(iter/s)": 1.438452
    },
    {
      "epoch": 4.280022278394242,
      "grad_norm": 8.225568771362305,
      "learning_rate": 5.029431490337156e-06,
      "loss": 2.110148048400879,
      "memory(GiB)": 77.56,
      "step": 99900,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.438458
    },
    {
      "epoch": 4.280236493723491,
      "grad_norm": 6.616367816925049,
      "learning_rate": 5.026490292510283e-06,
      "loss": 2.4189706802368165,
      "memory(GiB)": 77.56,
      "step": 99905,
      "token_acc": 0.4870848708487085,
      "train_speed(iter/s)": 1.438453
    },
    {
      "epoch": 4.28045070905274,
      "grad_norm": 5.795795917510986,
      "learning_rate": 5.023549909420722e-06,
      "loss": 2.2419610977172852,
      "memory(GiB)": 77.56,
      "step": 99910,
      "token_acc": 0.5198863636363636,
      "train_speed(iter/s)": 1.438447
    },
    {
      "epoch": 4.2806649243819885,
      "grad_norm": 7.112375736236572,
      "learning_rate": 5.02061034112174e-06,
      "loss": 2.3884592056274414,
      "memory(GiB)": 77.56,
      "step": 99915,
      "token_acc": 0.49851632047477745,
      "train_speed(iter/s)": 1.438437
    },
    {
      "epoch": 4.280879139711238,
      "grad_norm": 6.334680080413818,
      "learning_rate": 5.0176715876666e-06,
      "loss": 1.8271564483642577,
      "memory(GiB)": 77.56,
      "step": 99920,
      "token_acc": 0.6336206896551724,
      "train_speed(iter/s)": 1.438451
    },
    {
      "epoch": 4.281093355040487,
      "grad_norm": 6.141162872314453,
      "learning_rate": 5.014733649108538e-06,
      "loss": 2.3352020263671873,
      "memory(GiB)": 77.56,
      "step": 99925,
      "token_acc": 0.4827586206896552,
      "train_speed(iter/s)": 1.438447
    },
    {
      "epoch": 4.281307570369735,
      "grad_norm": 7.723939418792725,
      "learning_rate": 5.011796525500778e-06,
      "loss": 2.5333572387695313,
      "memory(GiB)": 77.56,
      "step": 99930,
      "token_acc": 0.46355685131195334,
      "train_speed(iter/s)": 1.438454
    },
    {
      "epoch": 4.281521785698985,
      "grad_norm": 5.576572895050049,
      "learning_rate": 5.008860216896527e-06,
      "loss": 2.1228477478027346,
      "memory(GiB)": 77.56,
      "step": 99935,
      "token_acc": 0.5338078291814946,
      "train_speed(iter/s)": 1.438457
    },
    {
      "epoch": 4.281736001028234,
      "grad_norm": 7.131587505340576,
      "learning_rate": 5.005924723348965e-06,
      "loss": 1.948977279663086,
      "memory(GiB)": 77.56,
      "step": 99940,
      "token_acc": 0.5807692307692308,
      "train_speed(iter/s)": 1.438464
    },
    {
      "epoch": 4.281950216357482,
      "grad_norm": 6.809156894683838,
      "learning_rate": 5.002990044911299e-06,
      "loss": 2.086215019226074,
      "memory(GiB)": 77.56,
      "step": 99945,
      "token_acc": 0.5305466237942122,
      "train_speed(iter/s)": 1.438466
    },
    {
      "epoch": 4.2821644316867316,
      "grad_norm": 5.828101634979248,
      "learning_rate": 5.000056181636676e-06,
      "loss": 2.0575798034667967,
      "memory(GiB)": 77.56,
      "step": 99950,
      "token_acc": 0.5777777777777777,
      "train_speed(iter/s)": 1.438452
    },
    {
      "epoch": 4.282378647015981,
      "grad_norm": 9.026351928710938,
      "learning_rate": 4.9971231335782465e-06,
      "loss": 2.1061494827270506,
      "memory(GiB)": 77.56,
      "step": 99955,
      "token_acc": 0.5261044176706827,
      "train_speed(iter/s)": 1.438473
    },
    {
      "epoch": 4.282592862345229,
      "grad_norm": 6.032985210418701,
      "learning_rate": 4.9941909007891476e-06,
      "loss": 2.1595590591430662,
      "memory(GiB)": 77.56,
      "step": 99960,
      "token_acc": 0.5440613026819924,
      "train_speed(iter/s)": 1.438473
    },
    {
      "epoch": 4.282807077674478,
      "grad_norm": 9.163704872131348,
      "learning_rate": 4.991259483322491e-06,
      "loss": 2.091373825073242,
      "memory(GiB)": 77.56,
      "step": 99965,
      "token_acc": 0.5257731958762887,
      "train_speed(iter/s)": 1.438489
    },
    {
      "epoch": 4.283021293003728,
      "grad_norm": 11.740559577941895,
      "learning_rate": 4.988328881231402e-06,
      "loss": 2.576679992675781,
      "memory(GiB)": 77.56,
      "step": 99970,
      "token_acc": 0.5202312138728323,
      "train_speed(iter/s)": 1.438502
    },
    {
      "epoch": 4.283235508332976,
      "grad_norm": 6.126219749450684,
      "learning_rate": 4.985399094568949e-06,
      "loss": 2.186813735961914,
      "memory(GiB)": 77.56,
      "step": 99975,
      "token_acc": 0.4968944099378882,
      "train_speed(iter/s)": 1.438499
    },
    {
      "epoch": 4.283449723662225,
      "grad_norm": 6.322460174560547,
      "learning_rate": 4.982470123388228e-06,
      "loss": 2.2531089782714844,
      "memory(GiB)": 77.56,
      "step": 99980,
      "token_acc": 0.5354330708661418,
      "train_speed(iter/s)": 1.438503
    },
    {
      "epoch": 4.283663938991475,
      "grad_norm": 6.5808820724487305,
      "learning_rate": 4.97954196774229e-06,
      "loss": 2.277579116821289,
      "memory(GiB)": 77.56,
      "step": 99985,
      "token_acc": 0.47703180212014135,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 4.283878154320723,
      "grad_norm": 7.900747776031494,
      "learning_rate": 4.976614627684184e-06,
      "loss": 2.1548343658447267,
      "memory(GiB)": 77.56,
      "step": 99990,
      "token_acc": 0.532520325203252,
      "train_speed(iter/s)": 1.43854
    },
    {
      "epoch": 4.284092369649972,
      "grad_norm": 5.826950550079346,
      "learning_rate": 4.973688103266938e-06,
      "loss": 2.300174522399902,
      "memory(GiB)": 77.56,
      "step": 99995,
      "token_acc": 0.5017182130584192,
      "train_speed(iter/s)": 1.438535
    },
    {
      "epoch": 4.2843065849792215,
      "grad_norm": 5.82090425491333,
      "learning_rate": 4.97076239454356e-06,
      "loss": 2.033364486694336,
      "memory(GiB)": 77.56,
      "step": 100000,
      "token_acc": 0.5442622950819672,
      "train_speed(iter/s)": 1.438538
    },
    {
      "epoch": 4.2843065849792215,
      "eval_loss": 2.166041612625122,
      "eval_runtime": 14.5673,
      "eval_samples_per_second": 6.865,
      "eval_steps_per_second": 6.865,
      "eval_token_acc": 0.4884393063583815,
      "step": 100000
    },
    {
      "epoch": 4.28452080030847,
      "grad_norm": 6.366593837738037,
      "learning_rate": 4.967837501567069e-06,
      "loss": 2.4353309631347657,
      "memory(GiB)": 77.56,
      "step": 100005,
      "token_acc": 0.49745158002038736,
      "train_speed(iter/s)": 1.43823
    },
    {
      "epoch": 4.284735015637719,
      "grad_norm": 8.043924331665039,
      "learning_rate": 4.964913424390449e-06,
      "loss": 2.4790821075439453,
      "memory(GiB)": 77.56,
      "step": 100010,
      "token_acc": 0.47115384615384615,
      "train_speed(iter/s)": 1.438242
    },
    {
      "epoch": 4.284949230966968,
      "grad_norm": 6.464168071746826,
      "learning_rate": 4.961990163066671e-06,
      "loss": 2.2506282806396483,
      "memory(GiB)": 77.56,
      "step": 100015,
      "token_acc": 0.5129151291512916,
      "train_speed(iter/s)": 1.438234
    },
    {
      "epoch": 4.285163446296217,
      "grad_norm": 6.957046031951904,
      "learning_rate": 4.959067717648685e-06,
      "loss": 2.345835494995117,
      "memory(GiB)": 77.56,
      "step": 100020,
      "token_acc": 0.4944649446494465,
      "train_speed(iter/s)": 1.438209
    },
    {
      "epoch": 4.285377661625466,
      "grad_norm": 8.82361125946045,
      "learning_rate": 4.956146088189434e-06,
      "loss": 2.1141170501708983,
      "memory(GiB)": 77.56,
      "step": 100025,
      "token_acc": 0.5229357798165137,
      "train_speed(iter/s)": 1.438228
    },
    {
      "epoch": 4.285591876954715,
      "grad_norm": 4.408382892608643,
      "learning_rate": 4.953225274741857e-06,
      "loss": 2.1523441314697265,
      "memory(GiB)": 77.56,
      "step": 100030,
      "token_acc": 0.5242165242165242,
      "train_speed(iter/s)": 1.43824
    },
    {
      "epoch": 4.285806092283964,
      "grad_norm": 8.105433464050293,
      "learning_rate": 4.9503052773588635e-06,
      "loss": 2.459971046447754,
      "memory(GiB)": 77.56,
      "step": 100035,
      "token_acc": 0.5089820359281437,
      "train_speed(iter/s)": 1.438239
    },
    {
      "epoch": 4.286020307613213,
      "grad_norm": 6.042153835296631,
      "learning_rate": 4.947386096093337e-06,
      "loss": 2.1405755996704103,
      "memory(GiB)": 77.56,
      "step": 100040,
      "token_acc": 0.5802469135802469,
      "train_speed(iter/s)": 1.438237
    },
    {
      "epoch": 4.286234522942462,
      "grad_norm": 5.343184947967529,
      "learning_rate": 4.944467730998187e-06,
      "loss": 2.4361061096191405,
      "memory(GiB)": 77.56,
      "step": 100045,
      "token_acc": 0.5462555066079295,
      "train_speed(iter/s)": 1.438252
    },
    {
      "epoch": 4.2864487382717105,
      "grad_norm": 5.78253698348999,
      "learning_rate": 4.94155018212627e-06,
      "loss": 2.562644195556641,
      "memory(GiB)": 77.56,
      "step": 100050,
      "token_acc": 0.47796610169491527,
      "train_speed(iter/s)": 1.438264
    },
    {
      "epoch": 4.28666295360096,
      "grad_norm": 6.665125846862793,
      "learning_rate": 4.9386334495304384e-06,
      "loss": 2.2552669525146483,
      "memory(GiB)": 77.56,
      "step": 100055,
      "token_acc": 0.5592105263157895,
      "train_speed(iter/s)": 1.438284
    },
    {
      "epoch": 4.286877168930209,
      "grad_norm": 7.766232490539551,
      "learning_rate": 4.935717533263523e-06,
      "loss": 2.300341987609863,
      "memory(GiB)": 77.56,
      "step": 100060,
      "token_acc": 0.4982698961937716,
      "train_speed(iter/s)": 1.438295
    },
    {
      "epoch": 4.287091384259457,
      "grad_norm": 5.08329963684082,
      "learning_rate": 4.93280243337837e-06,
      "loss": 2.548065185546875,
      "memory(GiB)": 77.56,
      "step": 100065,
      "token_acc": 0.5151515151515151,
      "train_speed(iter/s)": 1.438307
    },
    {
      "epoch": 4.287305599588707,
      "grad_norm": 7.803770542144775,
      "learning_rate": 4.929888149927775e-06,
      "loss": 2.387367820739746,
      "memory(GiB)": 77.56,
      "step": 100070,
      "token_acc": 0.4913494809688581,
      "train_speed(iter/s)": 1.438299
    },
    {
      "epoch": 4.287519814917956,
      "grad_norm": 5.919239044189453,
      "learning_rate": 4.926974682964536e-06,
      "loss": 2.143289566040039,
      "memory(GiB)": 77.56,
      "step": 100075,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438321
    },
    {
      "epoch": 4.287734030247204,
      "grad_norm": 6.287697792053223,
      "learning_rate": 4.924062032541432e-06,
      "loss": 2.140834426879883,
      "memory(GiB)": 77.56,
      "step": 100080,
      "token_acc": 0.5063291139240507,
      "train_speed(iter/s)": 1.43833
    },
    {
      "epoch": 4.2879482455764535,
      "grad_norm": 5.852848052978516,
      "learning_rate": 4.921150198711217e-06,
      "loss": 2.340757369995117,
      "memory(GiB)": 77.56,
      "step": 100085,
      "token_acc": 0.5469387755102041,
      "train_speed(iter/s)": 1.438334
    },
    {
      "epoch": 4.288162460905703,
      "grad_norm": 6.5657782554626465,
      "learning_rate": 4.9182391815266685e-06,
      "loss": 2.3492603302001953,
      "memory(GiB)": 77.56,
      "step": 100090,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.438349
    },
    {
      "epoch": 4.288376676234951,
      "grad_norm": 7.167140960693359,
      "learning_rate": 4.915328981040501e-06,
      "loss": 2.291988563537598,
      "memory(GiB)": 77.56,
      "step": 100095,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438348
    },
    {
      "epoch": 4.2885908915642,
      "grad_norm": 6.007768630981445,
      "learning_rate": 4.912419597305446e-06,
      "loss": 2.480708122253418,
      "memory(GiB)": 77.56,
      "step": 100100,
      "token_acc": 0.46853146853146854,
      "train_speed(iter/s)": 1.438346
    },
    {
      "epoch": 4.28880510689345,
      "grad_norm": 5.0034565925598145,
      "learning_rate": 4.909511030374209e-06,
      "loss": 2.386271858215332,
      "memory(GiB)": 77.56,
      "step": 100105,
      "token_acc": 0.4861111111111111,
      "train_speed(iter/s)": 1.438368
    },
    {
      "epoch": 4.289019322222698,
      "grad_norm": 7.46071195602417,
      "learning_rate": 4.906603280299471e-06,
      "loss": 2.638846588134766,
      "memory(GiB)": 77.56,
      "step": 100110,
      "token_acc": 0.4553846153846154,
      "train_speed(iter/s)": 1.438382
    },
    {
      "epoch": 4.289233537551947,
      "grad_norm": 8.779317855834961,
      "learning_rate": 4.903696347133912e-06,
      "loss": 2.289180946350098,
      "memory(GiB)": 77.56,
      "step": 100115,
      "token_acc": 0.5339805825242718,
      "train_speed(iter/s)": 1.438385
    },
    {
      "epoch": 4.2894477528811965,
      "grad_norm": 7.734889030456543,
      "learning_rate": 4.900790230930191e-06,
      "loss": 2.256235694885254,
      "memory(GiB)": 77.56,
      "step": 100120,
      "token_acc": 0.541958041958042,
      "train_speed(iter/s)": 1.43839
    },
    {
      "epoch": 4.289661968210445,
      "grad_norm": 6.054869651794434,
      "learning_rate": 4.897884931740976e-06,
      "loss": 2.2363828659057616,
      "memory(GiB)": 77.56,
      "step": 100125,
      "token_acc": 0.5494505494505495,
      "train_speed(iter/s)": 1.438407
    },
    {
      "epoch": 4.289876183539694,
      "grad_norm": 7.103828430175781,
      "learning_rate": 4.894980449618886e-06,
      "loss": 2.4280017852783202,
      "memory(GiB)": 77.56,
      "step": 100130,
      "token_acc": 0.5049180327868853,
      "train_speed(iter/s)": 1.438403
    },
    {
      "epoch": 4.290090398868943,
      "grad_norm": 4.5956711769104,
      "learning_rate": 4.892076784616534e-06,
      "loss": 2.622050476074219,
      "memory(GiB)": 77.56,
      "step": 100135,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.438415
    },
    {
      "epoch": 4.290304614198192,
      "grad_norm": 6.303820610046387,
      "learning_rate": 4.889173936786523e-06,
      "loss": 2.3801748275756838,
      "memory(GiB)": 77.56,
      "step": 100140,
      "token_acc": 0.4896551724137931,
      "train_speed(iter/s)": 1.438387
    },
    {
      "epoch": 4.290518829527441,
      "grad_norm": 6.977520942687988,
      "learning_rate": 4.886271906181439e-06,
      "loss": 2.4256723403930662,
      "memory(GiB)": 77.56,
      "step": 100145,
      "token_acc": 0.4748427672955975,
      "train_speed(iter/s)": 1.438392
    },
    {
      "epoch": 4.29073304485669,
      "grad_norm": 7.3222575187683105,
      "learning_rate": 4.883370692853867e-06,
      "loss": 2.3569026947021485,
      "memory(GiB)": 77.56,
      "step": 100150,
      "token_acc": 0.5034965034965035,
      "train_speed(iter/s)": 1.438397
    },
    {
      "epoch": 4.290947260185939,
      "grad_norm": 5.459466934204102,
      "learning_rate": 4.8804702968563596e-06,
      "loss": 2.3904865264892576,
      "memory(GiB)": 77.56,
      "step": 100155,
      "token_acc": 0.5116279069767442,
      "train_speed(iter/s)": 1.438407
    },
    {
      "epoch": 4.291161475515188,
      "grad_norm": 5.739074230194092,
      "learning_rate": 4.877570718241453e-06,
      "loss": 2.444879913330078,
      "memory(GiB)": 77.56,
      "step": 100160,
      "token_acc": 0.4815950920245399,
      "train_speed(iter/s)": 1.438404
    },
    {
      "epoch": 4.291375690844437,
      "grad_norm": 4.854170799255371,
      "learning_rate": 4.8746719570616885e-06,
      "loss": 1.8629764556884765,
      "memory(GiB)": 77.56,
      "step": 100165,
      "token_acc": 0.5481481481481482,
      "train_speed(iter/s)": 1.438423
    },
    {
      "epoch": 4.2915899061736855,
      "grad_norm": 5.591575622558594,
      "learning_rate": 4.871774013369556e-06,
      "loss": 2.2301511764526367,
      "memory(GiB)": 77.56,
      "step": 100170,
      "token_acc": 0.5173745173745173,
      "train_speed(iter/s)": 1.438429
    },
    {
      "epoch": 4.291804121502935,
      "grad_norm": 6.782240867614746,
      "learning_rate": 4.868876887217583e-06,
      "loss": 2.4777896881103514,
      "memory(GiB)": 77.56,
      "step": 100175,
      "token_acc": 0.4808362369337979,
      "train_speed(iter/s)": 1.438436
    },
    {
      "epoch": 4.292018336832184,
      "grad_norm": 6.378512859344482,
      "learning_rate": 4.865980578658241e-06,
      "loss": 2.316455268859863,
      "memory(GiB)": 77.56,
      "step": 100180,
      "token_acc": 0.468944099378882,
      "train_speed(iter/s)": 1.438451
    },
    {
      "epoch": 4.292232552161432,
      "grad_norm": 6.03173828125,
      "learning_rate": 4.8630850877440014e-06,
      "loss": 2.268305778503418,
      "memory(GiB)": 77.56,
      "step": 100185,
      "token_acc": 0.5370919881305638,
      "train_speed(iter/s)": 1.438449
    },
    {
      "epoch": 4.292446767490682,
      "grad_norm": 5.537770748138428,
      "learning_rate": 4.86019041452731e-06,
      "loss": 1.9870378494262695,
      "memory(GiB)": 77.56,
      "step": 100190,
      "token_acc": 0.5748987854251012,
      "train_speed(iter/s)": 1.43845
    },
    {
      "epoch": 4.292660982819931,
      "grad_norm": 5.093726634979248,
      "learning_rate": 4.8572965590606204e-06,
      "loss": 2.1830625534057617,
      "memory(GiB)": 77.56,
      "step": 100195,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.438463
    },
    {
      "epoch": 4.292875198149179,
      "grad_norm": 8.003413200378418,
      "learning_rate": 4.8544035213963505e-06,
      "loss": 2.4027179718017577,
      "memory(GiB)": 77.56,
      "step": 100200,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.438477
    },
    {
      "epoch": 4.2930894134784285,
      "grad_norm": 7.073882102966309,
      "learning_rate": 4.851511301586903e-06,
      "loss": 2.0694869995117187,
      "memory(GiB)": 77.56,
      "step": 100205,
      "token_acc": 0.5134099616858238,
      "train_speed(iter/s)": 1.438485
    },
    {
      "epoch": 4.293303628807678,
      "grad_norm": 8.690638542175293,
      "learning_rate": 4.848619899684686e-06,
      "loss": 2.4516393661499025,
      "memory(GiB)": 77.56,
      "step": 100210,
      "token_acc": 0.503125,
      "train_speed(iter/s)": 1.438481
    },
    {
      "epoch": 4.293517844136926,
      "grad_norm": 8.586088180541992,
      "learning_rate": 4.845729315742081e-06,
      "loss": 2.0516942977905273,
      "memory(GiB)": 77.56,
      "step": 100215,
      "token_acc": 0.5545851528384279,
      "train_speed(iter/s)": 1.438497
    },
    {
      "epoch": 4.293732059466175,
      "grad_norm": 5.63650369644165,
      "learning_rate": 4.8428395498114455e-06,
      "loss": 2.5344547271728515,
      "memory(GiB)": 77.56,
      "step": 100220,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438481
    },
    {
      "epoch": 4.293946274795425,
      "grad_norm": 6.443324565887451,
      "learning_rate": 4.839950601945131e-06,
      "loss": 2.1202959060668944,
      "memory(GiB)": 77.56,
      "step": 100225,
      "token_acc": 0.5425531914893617,
      "train_speed(iter/s)": 1.43848
    },
    {
      "epoch": 4.294160490124673,
      "grad_norm": 6.104748725891113,
      "learning_rate": 4.837062472195469e-06,
      "loss": 2.277557373046875,
      "memory(GiB)": 77.56,
      "step": 100230,
      "token_acc": 0.5344129554655871,
      "train_speed(iter/s)": 1.438481
    },
    {
      "epoch": 4.294374705453922,
      "grad_norm": 5.983887195587158,
      "learning_rate": 4.83417516061479e-06,
      "loss": 2.1089056015014647,
      "memory(GiB)": 77.56,
      "step": 100235,
      "token_acc": 0.5063291139240507,
      "train_speed(iter/s)": 1.438491
    },
    {
      "epoch": 4.294588920783172,
      "grad_norm": 6.209619998931885,
      "learning_rate": 4.831288667255401e-06,
      "loss": 2.2133438110351564,
      "memory(GiB)": 77.56,
      "step": 100240,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.438501
    },
    {
      "epoch": 4.29480313611242,
      "grad_norm": 6.0841851234436035,
      "learning_rate": 4.82840299216959e-06,
      "loss": 2.4208400726318358,
      "memory(GiB)": 77.56,
      "step": 100245,
      "token_acc": 0.4966887417218543,
      "train_speed(iter/s)": 1.438505
    },
    {
      "epoch": 4.295017351441669,
      "grad_norm": 5.9809980392456055,
      "learning_rate": 4.8255181354096345e-06,
      "loss": 2.4528121948242188,
      "memory(GiB)": 77.56,
      "step": 100250,
      "token_acc": 0.477124183006536,
      "train_speed(iter/s)": 1.438512
    },
    {
      "epoch": 4.295231566770918,
      "grad_norm": 5.294595718383789,
      "learning_rate": 4.822634097027789e-06,
      "loss": 2.408539581298828,
      "memory(GiB)": 77.56,
      "step": 100255,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.438519
    },
    {
      "epoch": 4.295445782100167,
      "grad_norm": 5.184760570526123,
      "learning_rate": 4.819750877076301e-06,
      "loss": 2.137157440185547,
      "memory(GiB)": 77.56,
      "step": 100260,
      "token_acc": 0.5254901960784314,
      "train_speed(iter/s)": 1.438515
    },
    {
      "epoch": 4.295659997429416,
      "grad_norm": 5.924389362335205,
      "learning_rate": 4.816868475607411e-06,
      "loss": 2.3872756958007812,
      "memory(GiB)": 77.56,
      "step": 100265,
      "token_acc": 0.4843205574912892,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 4.295874212758665,
      "grad_norm": 5.923495292663574,
      "learning_rate": 4.813986892673339e-06,
      "loss": 2.3210147857666015,
      "memory(GiB)": 77.56,
      "step": 100270,
      "token_acc": 0.4724137931034483,
      "train_speed(iter/s)": 1.43852
    },
    {
      "epoch": 4.296088428087914,
      "grad_norm": 6.01570463180542,
      "learning_rate": 4.811106128326281e-06,
      "loss": 2.3359352111816407,
      "memory(GiB)": 77.56,
      "step": 100275,
      "token_acc": 0.5047619047619047,
      "train_speed(iter/s)": 1.438513
    },
    {
      "epoch": 4.296302643417163,
      "grad_norm": 6.0657525062561035,
      "learning_rate": 4.808226182618431e-06,
      "loss": 1.992770004272461,
      "memory(GiB)": 77.56,
      "step": 100280,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.438514
    },
    {
      "epoch": 4.296516858746412,
      "grad_norm": 4.778995037078857,
      "learning_rate": 4.8053470556019555e-06,
      "loss": 2.1875919342041015,
      "memory(GiB)": 77.56,
      "step": 100285,
      "token_acc": 0.5047021943573667,
      "train_speed(iter/s)": 1.438521
    },
    {
      "epoch": 4.296731074075661,
      "grad_norm": 7.304758071899414,
      "learning_rate": 4.802468747329003e-06,
      "loss": 2.5573429107666015,
      "memory(GiB)": 77.56,
      "step": 100290,
      "token_acc": 0.47440273037542663,
      "train_speed(iter/s)": 1.438547
    },
    {
      "epoch": 4.29694528940491,
      "grad_norm": 10.679129600524902,
      "learning_rate": 4.799591257851738e-06,
      "loss": 2.335013198852539,
      "memory(GiB)": 77.56,
      "step": 100295,
      "token_acc": 0.48013245033112584,
      "train_speed(iter/s)": 1.438551
    },
    {
      "epoch": 4.297159504734159,
      "grad_norm": 6.13103723526001,
      "learning_rate": 4.796714587222278e-06,
      "loss": 2.4073814392089843,
      "memory(GiB)": 77.56,
      "step": 100300,
      "token_acc": 0.531986531986532,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 4.2973737200634075,
      "grad_norm": 5.2924675941467285,
      "learning_rate": 4.7938387354927396e-06,
      "loss": 2.375481605529785,
      "memory(GiB)": 77.56,
      "step": 100305,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.438582
    },
    {
      "epoch": 4.297587935392657,
      "grad_norm": 6.4611124992370605,
      "learning_rate": 4.790963702715218e-06,
      "loss": 2.6847564697265627,
      "memory(GiB)": 77.56,
      "step": 100310,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.438577
    },
    {
      "epoch": 4.297802150721906,
      "grad_norm": 6.762071132659912,
      "learning_rate": 4.788089488941788e-06,
      "loss": 2.3069812774658205,
      "memory(GiB)": 77.56,
      "step": 100315,
      "token_acc": 0.5241379310344828,
      "train_speed(iter/s)": 1.438588
    },
    {
      "epoch": 4.298016366051154,
      "grad_norm": 7.382822036743164,
      "learning_rate": 4.785216094224543e-06,
      "loss": 2.3777690887451173,
      "memory(GiB)": 77.56,
      "step": 100320,
      "token_acc": 0.4850498338870432,
      "train_speed(iter/s)": 1.438596
    },
    {
      "epoch": 4.298230581380404,
      "grad_norm": 6.4481048583984375,
      "learning_rate": 4.782343518615517e-06,
      "loss": 2.204959487915039,
      "memory(GiB)": 77.56,
      "step": 100325,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 4.298444796709653,
      "grad_norm": 6.603975296020508,
      "learning_rate": 4.779471762166759e-06,
      "loss": 2.197288703918457,
      "memory(GiB)": 77.56,
      "step": 100330,
      "token_acc": 0.550185873605948,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 4.298659012038901,
      "grad_norm": 7.149195194244385,
      "learning_rate": 4.776600824930283e-06,
      "loss": 2.354528045654297,
      "memory(GiB)": 77.56,
      "step": 100335,
      "token_acc": 0.5156794425087108,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 4.2988732273681505,
      "grad_norm": 6.460785865783691,
      "learning_rate": 4.773730706958113e-06,
      "loss": 2.0070682525634767,
      "memory(GiB)": 77.56,
      "step": 100340,
      "token_acc": 0.5463576158940397,
      "train_speed(iter/s)": 1.43858
    },
    {
      "epoch": 4.2990874426974,
      "grad_norm": 4.652647495269775,
      "learning_rate": 4.770861408302235e-06,
      "loss": 2.4208192825317383,
      "memory(GiB)": 77.56,
      "step": 100345,
      "token_acc": 0.5165562913907285,
      "train_speed(iter/s)": 1.438566
    },
    {
      "epoch": 4.299301658026648,
      "grad_norm": 6.717465877532959,
      "learning_rate": 4.767992929014625e-06,
      "loss": 2.0254501342773437,
      "memory(GiB)": 77.56,
      "step": 100350,
      "token_acc": 0.5609756097560976,
      "train_speed(iter/s)": 1.43856
    },
    {
      "epoch": 4.299515873355897,
      "grad_norm": 6.3007683753967285,
      "learning_rate": 4.765125269147264e-06,
      "loss": 2.3138801574707033,
      "memory(GiB)": 77.56,
      "step": 100355,
      "token_acc": 0.4931972789115646,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 4.299730088685147,
      "grad_norm": 5.693936824798584,
      "learning_rate": 4.762258428752092e-06,
      "loss": 2.073953628540039,
      "memory(GiB)": 77.56,
      "step": 100360,
      "token_acc": 0.5366795366795367,
      "train_speed(iter/s)": 1.438572
    },
    {
      "epoch": 4.299944304014395,
      "grad_norm": 6.544804096221924,
      "learning_rate": 4.759392407881041e-06,
      "loss": 2.211111068725586,
      "memory(GiB)": 77.56,
      "step": 100365,
      "token_acc": 0.5318471337579618,
      "train_speed(iter/s)": 1.438561
    },
    {
      "epoch": 4.300158519343644,
      "grad_norm": 6.404198169708252,
      "learning_rate": 4.756527206586037e-06,
      "loss": 2.1870901107788088,
      "memory(GiB)": 77.56,
      "step": 100370,
      "token_acc": 0.5134228187919463,
      "train_speed(iter/s)": 1.438564
    },
    {
      "epoch": 4.3003727346728935,
      "grad_norm": 7.706130504608154,
      "learning_rate": 4.7536628249189805e-06,
      "loss": 2.2626474380493162,
      "memory(GiB)": 77.56,
      "step": 100375,
      "token_acc": 0.510548523206751,
      "train_speed(iter/s)": 1.438566
    },
    {
      "epoch": 4.300586950002142,
      "grad_norm": 6.014760494232178,
      "learning_rate": 4.750799262931771e-06,
      "loss": 2.1988319396972655,
      "memory(GiB)": 77.56,
      "step": 100380,
      "token_acc": 0.5226480836236934,
      "train_speed(iter/s)": 1.438571
    },
    {
      "epoch": 4.300801165331391,
      "grad_norm": 6.171198844909668,
      "learning_rate": 4.747936520676277e-06,
      "loss": 2.14705753326416,
      "memory(GiB)": 77.56,
      "step": 100385,
      "token_acc": 0.5104895104895105,
      "train_speed(iter/s)": 1.438571
    },
    {
      "epoch": 4.30101538066064,
      "grad_norm": 6.147973537445068,
      "learning_rate": 4.745074598204369e-06,
      "loss": 2.2761959075927733,
      "memory(GiB)": 77.56,
      "step": 100390,
      "token_acc": 0.5198863636363636,
      "train_speed(iter/s)": 1.438563
    },
    {
      "epoch": 4.301229595989889,
      "grad_norm": 6.9807515144348145,
      "learning_rate": 4.742213495567882e-06,
      "loss": 2.3244550704956053,
      "memory(GiB)": 77.56,
      "step": 100395,
      "token_acc": 0.4889589905362776,
      "train_speed(iter/s)": 1.438567
    },
    {
      "epoch": 4.301443811319138,
      "grad_norm": 4.6973042488098145,
      "learning_rate": 4.739353212818659e-06,
      "loss": 2.414212226867676,
      "memory(GiB)": 77.56,
      "step": 100400,
      "token_acc": 0.5088235294117647,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 4.301658026648387,
      "grad_norm": 6.312798976898193,
      "learning_rate": 4.736493750008497e-06,
      "loss": 2.2111274719238283,
      "memory(GiB)": 77.56,
      "step": 100405,
      "token_acc": 0.5231788079470199,
      "train_speed(iter/s)": 1.438588
    },
    {
      "epoch": 4.301872241977636,
      "grad_norm": 6.337025165557861,
      "learning_rate": 4.7336351071892105e-06,
      "loss": 2.433379554748535,
      "memory(GiB)": 77.56,
      "step": 100410,
      "token_acc": 0.5163636363636364,
      "train_speed(iter/s)": 1.438607
    },
    {
      "epoch": 4.302086457306885,
      "grad_norm": 6.201536178588867,
      "learning_rate": 4.730777284412596e-06,
      "loss": 2.057038497924805,
      "memory(GiB)": 77.56,
      "step": 100415,
      "token_acc": 0.5830258302583026,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 4.302300672636134,
      "grad_norm": 5.896797180175781,
      "learning_rate": 4.727920281730425e-06,
      "loss": 2.0697708129882812,
      "memory(GiB)": 77.56,
      "step": 100420,
      "token_acc": 0.5719844357976653,
      "train_speed(iter/s)": 1.43862
    },
    {
      "epoch": 4.3025148879653825,
      "grad_norm": 5.5664873123168945,
      "learning_rate": 4.7250640991944375e-06,
      "loss": 2.4261806488037108,
      "memory(GiB)": 77.56,
      "step": 100425,
      "token_acc": 0.48059701492537316,
      "train_speed(iter/s)": 1.43862
    },
    {
      "epoch": 4.302729103294632,
      "grad_norm": 5.498995780944824,
      "learning_rate": 4.722208736856387e-06,
      "loss": 2.5309362411499023,
      "memory(GiB)": 77.56,
      "step": 100430,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.438622
    },
    {
      "epoch": 4.302943318623881,
      "grad_norm": 6.2231526374816895,
      "learning_rate": 4.719354194767994e-06,
      "loss": 2.106704330444336,
      "memory(GiB)": 77.56,
      "step": 100435,
      "token_acc": 0.5170068027210885,
      "train_speed(iter/s)": 1.438638
    },
    {
      "epoch": 4.303157533953129,
      "grad_norm": 7.580305099487305,
      "learning_rate": 4.716500472980983e-06,
      "loss": 2.345419692993164,
      "memory(GiB)": 77.56,
      "step": 100440,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.438651
    },
    {
      "epoch": 4.303371749282379,
      "grad_norm": 7.637293338775635,
      "learning_rate": 4.713647571547048e-06,
      "loss": 2.1894128799438475,
      "memory(GiB)": 77.56,
      "step": 100445,
      "token_acc": 0.5037593984962406,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 4.303585964611628,
      "grad_norm": 5.935568809509277,
      "learning_rate": 4.710795490517861e-06,
      "loss": 2.150985908508301,
      "memory(GiB)": 77.56,
      "step": 100450,
      "token_acc": 0.5195729537366548,
      "train_speed(iter/s)": 1.438666
    },
    {
      "epoch": 4.303800179940876,
      "grad_norm": 7.975590705871582,
      "learning_rate": 4.707944229945105e-06,
      "loss": 2.370107650756836,
      "memory(GiB)": 77.56,
      "step": 100455,
      "token_acc": 0.524822695035461,
      "train_speed(iter/s)": 1.438675
    },
    {
      "epoch": 4.3040143952701255,
      "grad_norm": 6.179610252380371,
      "learning_rate": 4.705093789880416e-06,
      "loss": 2.187881851196289,
      "memory(GiB)": 77.56,
      "step": 100460,
      "token_acc": 0.5478927203065134,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 4.304228610599375,
      "grad_norm": 5.767256259918213,
      "learning_rate": 4.702244170375453e-06,
      "loss": 1.9687423706054688,
      "memory(GiB)": 77.56,
      "step": 100465,
      "token_acc": 0.5649122807017544,
      "train_speed(iter/s)": 1.438679
    },
    {
      "epoch": 4.304442825928623,
      "grad_norm": 6.775595188140869,
      "learning_rate": 4.699395371481829e-06,
      "loss": 1.9267642974853516,
      "memory(GiB)": 77.56,
      "step": 100470,
      "token_acc": 0.5648854961832062,
      "train_speed(iter/s)": 1.438684
    },
    {
      "epoch": 4.304657041257872,
      "grad_norm": 6.519343376159668,
      "learning_rate": 4.696547393251155e-06,
      "loss": 2.1873184204101563,
      "memory(GiB)": 77.56,
      "step": 100475,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.438678
    },
    {
      "epoch": 4.304871256587122,
      "grad_norm": 6.8834075927734375,
      "learning_rate": 4.693700235735021e-06,
      "loss": 2.5468477249145507,
      "memory(GiB)": 77.56,
      "step": 100480,
      "token_acc": 0.47491638795986624,
      "train_speed(iter/s)": 1.43867
    },
    {
      "epoch": 4.30508547191637,
      "grad_norm": 7.11431884765625,
      "learning_rate": 4.690853898985004e-06,
      "loss": 2.3558788299560547,
      "memory(GiB)": 77.56,
      "step": 100485,
      "token_acc": 0.541958041958042,
      "train_speed(iter/s)": 1.438675
    },
    {
      "epoch": 4.305299687245619,
      "grad_norm": 6.505257606506348,
      "learning_rate": 4.688008383052672e-06,
      "loss": 2.3135259628295897,
      "memory(GiB)": 77.56,
      "step": 100490,
      "token_acc": 0.5161290322580645,
      "train_speed(iter/s)": 1.438689
    },
    {
      "epoch": 4.3055139025748685,
      "grad_norm": 6.5288920402526855,
      "learning_rate": 4.685163687989569e-06,
      "loss": 2.2993209838867186,
      "memory(GiB)": 77.56,
      "step": 100495,
      "token_acc": 0.503125,
      "train_speed(iter/s)": 1.438669
    },
    {
      "epoch": 4.305728117904117,
      "grad_norm": 5.279067039489746,
      "learning_rate": 4.682319813847247e-06,
      "loss": 2.2329715728759765,
      "memory(GiB)": 77.56,
      "step": 100500,
      "token_acc": 0.47904191616766467,
      "train_speed(iter/s)": 1.438674
    },
    {
      "epoch": 4.305728117904117,
      "eval_loss": 2.1699624061584473,
      "eval_runtime": 15.0777,
      "eval_samples_per_second": 6.632,
      "eval_steps_per_second": 6.632,
      "eval_token_acc": 0.4699300699300699,
      "step": 100500
    },
    {
      "epoch": 4.305942333233366,
      "grad_norm": 5.510031223297119,
      "learning_rate": 4.679476760677209e-06,
      "loss": 2.283426284790039,
      "memory(GiB)": 77.56,
      "step": 100505,
      "token_acc": 0.4893410852713178,
      "train_speed(iter/s)": 1.438351
    },
    {
      "epoch": 4.306156548562615,
      "grad_norm": 7.931199550628662,
      "learning_rate": 4.676634528530965e-06,
      "loss": 2.383808898925781,
      "memory(GiB)": 77.56,
      "step": 100510,
      "token_acc": 0.5145985401459854,
      "train_speed(iter/s)": 1.438348
    },
    {
      "epoch": 4.306370763891864,
      "grad_norm": 5.487090110778809,
      "learning_rate": 4.673793117460007e-06,
      "loss": 2.330915641784668,
      "memory(GiB)": 77.56,
      "step": 100515,
      "token_acc": 0.46706586826347307,
      "train_speed(iter/s)": 1.43836
    },
    {
      "epoch": 4.306584979221113,
      "grad_norm": 5.396183490753174,
      "learning_rate": 4.670952527515793e-06,
      "loss": 1.9868366241455078,
      "memory(GiB)": 77.56,
      "step": 100520,
      "token_acc": 0.5668016194331984,
      "train_speed(iter/s)": 1.438372
    },
    {
      "epoch": 4.306799194550362,
      "grad_norm": 7.76336145401001,
      "learning_rate": 4.668112758749804e-06,
      "loss": 2.415496826171875,
      "memory(GiB)": 77.56,
      "step": 100525,
      "token_acc": 0.49640287769784175,
      "train_speed(iter/s)": 1.438386
    },
    {
      "epoch": 4.307013409879611,
      "grad_norm": 5.985770225524902,
      "learning_rate": 4.665273811213478e-06,
      "loss": 2.116036224365234,
      "memory(GiB)": 77.56,
      "step": 100530,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438396
    },
    {
      "epoch": 4.30722762520886,
      "grad_norm": 6.645129680633545,
      "learning_rate": 4.662435684958244e-06,
      "loss": 2.400859260559082,
      "memory(GiB)": 77.56,
      "step": 100535,
      "token_acc": 0.5085910652920962,
      "train_speed(iter/s)": 1.438401
    },
    {
      "epoch": 4.307441840538109,
      "grad_norm": 9.147539138793945,
      "learning_rate": 4.659598380035518e-06,
      "loss": 2.2356204986572266,
      "memory(GiB)": 77.56,
      "step": 100540,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.438412
    },
    {
      "epoch": 4.307656055867358,
      "grad_norm": 8.270538330078125,
      "learning_rate": 4.656761896496703e-06,
      "loss": 2.107296371459961,
      "memory(GiB)": 77.56,
      "step": 100545,
      "token_acc": 0.5245901639344263,
      "train_speed(iter/s)": 1.438416
    },
    {
      "epoch": 4.307870271196607,
      "grad_norm": 6.911714553833008,
      "learning_rate": 4.653926234393169e-06,
      "loss": 1.792007827758789,
      "memory(GiB)": 77.56,
      "step": 100550,
      "token_acc": 0.5497835497835498,
      "train_speed(iter/s)": 1.438426
    },
    {
      "epoch": 4.308084486525856,
      "grad_norm": 5.8084940910339355,
      "learning_rate": 4.651091393776308e-06,
      "loss": 2.2597923278808594,
      "memory(GiB)": 77.56,
      "step": 100555,
      "token_acc": 0.5229681978798587,
      "train_speed(iter/s)": 1.438438
    },
    {
      "epoch": 4.308298701855104,
      "grad_norm": 7.472497940063477,
      "learning_rate": 4.648257374697462e-06,
      "loss": 2.2256282806396483,
      "memory(GiB)": 77.56,
      "step": 100560,
      "token_acc": 0.5048231511254019,
      "train_speed(iter/s)": 1.438452
    },
    {
      "epoch": 4.308512917184354,
      "grad_norm": 6.116946697235107,
      "learning_rate": 4.645424177207985e-06,
      "loss": 2.4894184112548827,
      "memory(GiB)": 77.56,
      "step": 100565,
      "token_acc": 0.5123456790123457,
      "train_speed(iter/s)": 1.438463
    },
    {
      "epoch": 4.308727132513603,
      "grad_norm": 10.386627197265625,
      "learning_rate": 4.6425918013591916e-06,
      "loss": 2.465712547302246,
      "memory(GiB)": 77.56,
      "step": 100570,
      "token_acc": 0.4756554307116105,
      "train_speed(iter/s)": 1.43847
    },
    {
      "epoch": 4.308941347842851,
      "grad_norm": 5.519584655761719,
      "learning_rate": 4.639760247202396e-06,
      "loss": 2.4339309692382813,
      "memory(GiB)": 77.56,
      "step": 100575,
      "token_acc": 0.5269709543568465,
      "train_speed(iter/s)": 1.438484
    },
    {
      "epoch": 4.309155563172101,
      "grad_norm": 5.392373561859131,
      "learning_rate": 4.636929514788891e-06,
      "loss": 2.471701431274414,
      "memory(GiB)": 77.56,
      "step": 100580,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.43849
    },
    {
      "epoch": 4.30936977850135,
      "grad_norm": 6.107878684997559,
      "learning_rate": 4.6340996041699665e-06,
      "loss": 2.0578975677490234,
      "memory(GiB)": 77.56,
      "step": 100585,
      "token_acc": 0.5387323943661971,
      "train_speed(iter/s)": 1.438492
    },
    {
      "epoch": 4.309583993830598,
      "grad_norm": 6.998330116271973,
      "learning_rate": 4.631270515396891e-06,
      "loss": 2.5220548629760744,
      "memory(GiB)": 77.56,
      "step": 100590,
      "token_acc": 0.48880597014925375,
      "train_speed(iter/s)": 1.438485
    },
    {
      "epoch": 4.3097982091598475,
      "grad_norm": 7.377869606018066,
      "learning_rate": 4.628442248520904e-06,
      "loss": 2.4531581878662108,
      "memory(GiB)": 77.56,
      "step": 100595,
      "token_acc": 0.4702194357366771,
      "train_speed(iter/s)": 1.438482
    },
    {
      "epoch": 4.310012424489097,
      "grad_norm": 6.266183376312256,
      "learning_rate": 4.625614803593248e-06,
      "loss": 2.3262874603271486,
      "memory(GiB)": 77.56,
      "step": 100600,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.438479
    },
    {
      "epoch": 4.310226639818345,
      "grad_norm": 8.184282302856445,
      "learning_rate": 4.622788180665133e-06,
      "loss": 2.115907669067383,
      "memory(GiB)": 77.56,
      "step": 100605,
      "token_acc": 0.527972027972028,
      "train_speed(iter/s)": 1.43849
    },
    {
      "epoch": 4.310440855147594,
      "grad_norm": 7.2758469581604,
      "learning_rate": 4.61996237978779e-06,
      "loss": 2.3972925186157226,
      "memory(GiB)": 77.56,
      "step": 100610,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.438486
    },
    {
      "epoch": 4.310655070476844,
      "grad_norm": 5.9413957595825195,
      "learning_rate": 4.6171374010123945e-06,
      "loss": 2.192464828491211,
      "memory(GiB)": 77.56,
      "step": 100615,
      "token_acc": 0.5015105740181269,
      "train_speed(iter/s)": 1.438507
    },
    {
      "epoch": 4.310869285806092,
      "grad_norm": 8.079716682434082,
      "learning_rate": 4.614313244390133e-06,
      "loss": 2.2437488555908205,
      "memory(GiB)": 77.56,
      "step": 100620,
      "token_acc": 0.531986531986532,
      "train_speed(iter/s)": 1.438519
    },
    {
      "epoch": 4.311083501135341,
      "grad_norm": 10.538756370544434,
      "learning_rate": 4.611489909972161e-06,
      "loss": 2.578044319152832,
      "memory(GiB)": 77.56,
      "step": 100625,
      "token_acc": 0.4440789473684211,
      "train_speed(iter/s)": 1.438522
    },
    {
      "epoch": 4.3112977164645905,
      "grad_norm": 8.221933364868164,
      "learning_rate": 4.6086673978096125e-06,
      "loss": 2.3713968276977537,
      "memory(GiB)": 77.56,
      "step": 100630,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.438533
    },
    {
      "epoch": 4.311511931793839,
      "grad_norm": 5.455938339233398,
      "learning_rate": 4.605845707953649e-06,
      "loss": 2.565324401855469,
      "memory(GiB)": 77.56,
      "step": 100635,
      "token_acc": 0.528169014084507,
      "train_speed(iter/s)": 1.438541
    },
    {
      "epoch": 4.311726147123088,
      "grad_norm": 6.34124755859375,
      "learning_rate": 4.603024840455367e-06,
      "loss": 2.448826789855957,
      "memory(GiB)": 77.56,
      "step": 100640,
      "token_acc": 0.5108695652173914,
      "train_speed(iter/s)": 1.438558
    },
    {
      "epoch": 4.311940362452337,
      "grad_norm": 5.32847785949707,
      "learning_rate": 4.600204795365881e-06,
      "loss": 2.1911535263061523,
      "memory(GiB)": 77.56,
      "step": 100645,
      "token_acc": 0.5086705202312138,
      "train_speed(iter/s)": 1.43856
    },
    {
      "epoch": 4.312154577781586,
      "grad_norm": 5.827273845672607,
      "learning_rate": 4.597385572736273e-06,
      "loss": 2.0310623168945314,
      "memory(GiB)": 77.56,
      "step": 100650,
      "token_acc": 0.5263157894736842,
      "train_speed(iter/s)": 1.438557
    },
    {
      "epoch": 4.312368793110835,
      "grad_norm": 6.4080610275268555,
      "learning_rate": 4.59456717261762e-06,
      "loss": 2.2133161544799806,
      "memory(GiB)": 77.56,
      "step": 100655,
      "token_acc": 0.5350877192982456,
      "train_speed(iter/s)": 1.438562
    },
    {
      "epoch": 4.312583008440084,
      "grad_norm": 4.702672958374023,
      "learning_rate": 4.591749595060974e-06,
      "loss": 1.986974334716797,
      "memory(GiB)": 77.56,
      "step": 100660,
      "token_acc": 0.5964285714285714,
      "train_speed(iter/s)": 1.438555
    },
    {
      "epoch": 4.312797223769333,
      "grad_norm": 6.637474536895752,
      "learning_rate": 4.588932840117366e-06,
      "loss": 2.2598472595214845,
      "memory(GiB)": 77.56,
      "step": 100665,
      "token_acc": 0.5318471337579618,
      "train_speed(iter/s)": 1.438548
    },
    {
      "epoch": 4.313011439098582,
      "grad_norm": 7.995211124420166,
      "learning_rate": 4.586116907837857e-06,
      "loss": 2.6341548919677735,
      "memory(GiB)": 77.56,
      "step": 100670,
      "token_acc": 0.48623853211009177,
      "train_speed(iter/s)": 1.438552
    },
    {
      "epoch": 4.313225654427831,
      "grad_norm": 5.640625476837158,
      "learning_rate": 4.583301798273437e-06,
      "loss": 2.3597785949707033,
      "memory(GiB)": 77.56,
      "step": 100675,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 4.3134398697570795,
      "grad_norm": 5.814727783203125,
      "learning_rate": 4.580487511475112e-06,
      "loss": 2.283009147644043,
      "memory(GiB)": 77.56,
      "step": 100680,
      "token_acc": 0.5444839857651246,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 4.313654085086329,
      "grad_norm": 7.987022876739502,
      "learning_rate": 4.577674047493857e-06,
      "loss": 2.1850864410400392,
      "memory(GiB)": 77.56,
      "step": 100685,
      "token_acc": 0.531496062992126,
      "train_speed(iter/s)": 1.438571
    },
    {
      "epoch": 4.313868300415578,
      "grad_norm": 6.397132396697998,
      "learning_rate": 4.574861406380654e-06,
      "loss": 2.5650733947753905,
      "memory(GiB)": 77.56,
      "step": 100690,
      "token_acc": 0.4542857142857143,
      "train_speed(iter/s)": 1.438585
    },
    {
      "epoch": 4.314082515744826,
      "grad_norm": 4.941383361816406,
      "learning_rate": 4.572049588186433e-06,
      "loss": 2.2035526275634765,
      "memory(GiB)": 77.56,
      "step": 100695,
      "token_acc": 0.5666666666666667,
      "train_speed(iter/s)": 1.4386
    },
    {
      "epoch": 4.314296731074076,
      "grad_norm": 7.093830108642578,
      "learning_rate": 4.5692385929621604e-06,
      "loss": 2.2035017013549805,
      "memory(GiB)": 77.56,
      "step": 100700,
      "token_acc": 0.4981949458483754,
      "train_speed(iter/s)": 1.438623
    },
    {
      "epoch": 4.314510946403325,
      "grad_norm": 5.65403938293457,
      "learning_rate": 4.566428420758739e-06,
      "loss": 2.0075403213500977,
      "memory(GiB)": 77.56,
      "step": 100705,
      "token_acc": 0.5738255033557047,
      "train_speed(iter/s)": 1.43863
    },
    {
      "epoch": 4.314725161732573,
      "grad_norm": 5.659056186676025,
      "learning_rate": 4.563619071627096e-06,
      "loss": 2.4824567794799806,
      "memory(GiB)": 77.56,
      "step": 100710,
      "token_acc": 0.45901639344262296,
      "train_speed(iter/s)": 1.438643
    },
    {
      "epoch": 4.3149393770618225,
      "grad_norm": 6.608780860900879,
      "learning_rate": 4.560810545618116e-06,
      "loss": 2.5676361083984376,
      "memory(GiB)": 77.56,
      "step": 100715,
      "token_acc": 0.42424242424242425,
      "train_speed(iter/s)": 1.438649
    },
    {
      "epoch": 4.315153592391072,
      "grad_norm": 7.0502543449401855,
      "learning_rate": 4.558002842782677e-06,
      "loss": 2.3460012435913087,
      "memory(GiB)": 77.56,
      "step": 100720,
      "token_acc": 0.48028673835125446,
      "train_speed(iter/s)": 1.438654
    },
    {
      "epoch": 4.31536780772032,
      "grad_norm": 5.865872859954834,
      "learning_rate": 4.5551959631716376e-06,
      "loss": 2.1148401260375977,
      "memory(GiB)": 77.56,
      "step": 100725,
      "token_acc": 0.47985347985347987,
      "train_speed(iter/s)": 1.43866
    },
    {
      "epoch": 4.315582023049569,
      "grad_norm": 5.7145538330078125,
      "learning_rate": 4.552389906835863e-06,
      "loss": 2.3243362426757814,
      "memory(GiB)": 77.56,
      "step": 100730,
      "token_acc": 0.4694533762057878,
      "train_speed(iter/s)": 1.438674
    },
    {
      "epoch": 4.315796238378819,
      "grad_norm": 7.404398441314697,
      "learning_rate": 4.549584673826179e-06,
      "loss": 2.294243049621582,
      "memory(GiB)": 77.56,
      "step": 100735,
      "token_acc": 0.5384615384615384,
      "train_speed(iter/s)": 1.438694
    },
    {
      "epoch": 4.316010453708067,
      "grad_norm": 5.528452396392822,
      "learning_rate": 4.546780264193406e-06,
      "loss": 2.5592575073242188,
      "memory(GiB)": 77.56,
      "step": 100740,
      "token_acc": 0.5051546391752577,
      "train_speed(iter/s)": 1.438674
    },
    {
      "epoch": 4.316224669037316,
      "grad_norm": 7.086595058441162,
      "learning_rate": 4.543976677988343e-06,
      "loss": 2.309468460083008,
      "memory(GiB)": 77.56,
      "step": 100745,
      "token_acc": 0.5255813953488372,
      "train_speed(iter/s)": 1.43867
    },
    {
      "epoch": 4.3164388843665655,
      "grad_norm": 6.76436710357666,
      "learning_rate": 4.541173915261776e-06,
      "loss": 2.2618900299072267,
      "memory(GiB)": 77.56,
      "step": 100750,
      "token_acc": 0.5277777777777778,
      "train_speed(iter/s)": 1.438672
    },
    {
      "epoch": 4.316653099695814,
      "grad_norm": 5.549356460571289,
      "learning_rate": 4.538371976064499e-06,
      "loss": 2.2458831787109377,
      "memory(GiB)": 77.56,
      "step": 100755,
      "token_acc": 0.5054545454545455,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 4.316867315025063,
      "grad_norm": 6.185845851898193,
      "learning_rate": 4.535570860447258e-06,
      "loss": 2.2630590438842773,
      "memory(GiB)": 77.56,
      "step": 100760,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.43869
    },
    {
      "epoch": 4.317081530354312,
      "grad_norm": 5.358894348144531,
      "learning_rate": 4.532770568460798e-06,
      "loss": 2.2653152465820314,
      "memory(GiB)": 77.56,
      "step": 100765,
      "token_acc": 0.5121951219512195,
      "train_speed(iter/s)": 1.438674
    },
    {
      "epoch": 4.317295745683561,
      "grad_norm": 6.094043731689453,
      "learning_rate": 4.529971100155855e-06,
      "loss": 2.092005729675293,
      "memory(GiB)": 77.56,
      "step": 100770,
      "token_acc": 0.5031055900621118,
      "train_speed(iter/s)": 1.438673
    },
    {
      "epoch": 4.31750996101281,
      "grad_norm": 5.736149787902832,
      "learning_rate": 4.52717245558314e-06,
      "loss": 2.1980871200561523,
      "memory(GiB)": 77.56,
      "step": 100775,
      "token_acc": 0.5379061371841155,
      "train_speed(iter/s)": 1.438664
    },
    {
      "epoch": 4.317724176342059,
      "grad_norm": 5.013285160064697,
      "learning_rate": 4.524374634793338e-06,
      "loss": 2.191897392272949,
      "memory(GiB)": 77.56,
      "step": 100780,
      "token_acc": 0.5111940298507462,
      "train_speed(iter/s)": 1.438662
    },
    {
      "epoch": 4.317938391671308,
      "grad_norm": 5.689235687255859,
      "learning_rate": 4.521577637837154e-06,
      "loss": 2.2575002670288087,
      "memory(GiB)": 77.56,
      "step": 100785,
      "token_acc": 0.5149253731343284,
      "train_speed(iter/s)": 1.438694
    },
    {
      "epoch": 4.318152607000557,
      "grad_norm": 6.8830695152282715,
      "learning_rate": 4.518781464765259e-06,
      "loss": 2.3067911148071287,
      "memory(GiB)": 77.56,
      "step": 100790,
      "token_acc": 0.5547945205479452,
      "train_speed(iter/s)": 1.438686
    },
    {
      "epoch": 4.318366822329806,
      "grad_norm": 6.526504993438721,
      "learning_rate": 4.515986115628306e-06,
      "loss": 2.084463882446289,
      "memory(GiB)": 77.56,
      "step": 100795,
      "token_acc": 0.5362903225806451,
      "train_speed(iter/s)": 1.438692
    },
    {
      "epoch": 4.3185810376590545,
      "grad_norm": 5.212656497955322,
      "learning_rate": 4.513191590476934e-06,
      "loss": 2.4213384628295898,
      "memory(GiB)": 77.56,
      "step": 100800,
      "token_acc": 0.5096774193548387,
      "train_speed(iter/s)": 1.438692
    },
    {
      "epoch": 4.318795252988304,
      "grad_norm": 6.196159362792969,
      "learning_rate": 4.510397889361761e-06,
      "loss": 2.316198539733887,
      "memory(GiB)": 77.56,
      "step": 100805,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.438701
    },
    {
      "epoch": 4.319009468317553,
      "grad_norm": 7.193774700164795,
      "learning_rate": 4.507605012333394e-06,
      "loss": 2.364968109130859,
      "memory(GiB)": 77.56,
      "step": 100810,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.438687
    },
    {
      "epoch": 4.319223683646801,
      "grad_norm": 5.920011520385742,
      "learning_rate": 4.504812959442451e-06,
      "loss": 2.2006067276000976,
      "memory(GiB)": 77.56,
      "step": 100815,
      "token_acc": 0.5238095238095238,
      "train_speed(iter/s)": 1.438692
    },
    {
      "epoch": 4.319437898976051,
      "grad_norm": 6.084973335266113,
      "learning_rate": 4.502021730739497e-06,
      "loss": 2.400521087646484,
      "memory(GiB)": 77.56,
      "step": 100820,
      "token_acc": 0.465625,
      "train_speed(iter/s)": 1.438706
    },
    {
      "epoch": 4.3196521143053,
      "grad_norm": 6.604528427124023,
      "learning_rate": 4.4992313262751e-06,
      "loss": 2.133163642883301,
      "memory(GiB)": 77.56,
      "step": 100825,
      "token_acc": 0.5413533834586466,
      "train_speed(iter/s)": 1.438702
    },
    {
      "epoch": 4.319866329634548,
      "grad_norm": 6.3021955490112305,
      "learning_rate": 4.496441746099811e-06,
      "loss": 2.4390275955200194,
      "memory(GiB)": 77.56,
      "step": 100830,
      "token_acc": 0.49709302325581395,
      "train_speed(iter/s)": 1.43871
    },
    {
      "epoch": 4.320080544963798,
      "grad_norm": 11.538179397583008,
      "learning_rate": 4.493652990264152e-06,
      "loss": 2.1963125228881837,
      "memory(GiB)": 77.56,
      "step": 100835,
      "token_acc": 0.5358361774744027,
      "train_speed(iter/s)": 1.438729
    },
    {
      "epoch": 4.320294760293047,
      "grad_norm": 6.0271220207214355,
      "learning_rate": 4.4908650588186715e-06,
      "loss": 2.2506488800048827,
      "memory(GiB)": 77.56,
      "step": 100840,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.438735
    },
    {
      "epoch": 4.320508975622295,
      "grad_norm": 6.2990617752075195,
      "learning_rate": 4.488077951813863e-06,
      "loss": 2.4979904174804686,
      "memory(GiB)": 77.56,
      "step": 100845,
      "token_acc": 0.4774011299435028,
      "train_speed(iter/s)": 1.438761
    },
    {
      "epoch": 4.3207231909515444,
      "grad_norm": 6.008528709411621,
      "learning_rate": 4.4852916693002124e-06,
      "loss": 2.040965270996094,
      "memory(GiB)": 77.56,
      "step": 100850,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.438757
    },
    {
      "epoch": 4.320937406280794,
      "grad_norm": 10.73942756652832,
      "learning_rate": 4.482506211328191e-06,
      "loss": 2.5719877243041993,
      "memory(GiB)": 77.56,
      "step": 100855,
      "token_acc": 0.4878048780487805,
      "train_speed(iter/s)": 1.438757
    },
    {
      "epoch": 4.321151621610042,
      "grad_norm": 6.381577491760254,
      "learning_rate": 4.479721577948276e-06,
      "loss": 2.053926467895508,
      "memory(GiB)": 77.56,
      "step": 100860,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.438775
    },
    {
      "epoch": 4.321365836939291,
      "grad_norm": 5.6428961753845215,
      "learning_rate": 4.476937769210909e-06,
      "loss": 2.0882631301879884,
      "memory(GiB)": 77.56,
      "step": 100865,
      "token_acc": 0.5440613026819924,
      "train_speed(iter/s)": 1.43879
    },
    {
      "epoch": 4.321580052268541,
      "grad_norm": 6.470463275909424,
      "learning_rate": 4.474154785166506e-06,
      "loss": 2.0445018768310548,
      "memory(GiB)": 77.56,
      "step": 100870,
      "token_acc": 0.5030487804878049,
      "train_speed(iter/s)": 1.438798
    },
    {
      "epoch": 4.321794267597789,
      "grad_norm": 6.739324569702148,
      "learning_rate": 4.471372625865511e-06,
      "loss": 2.3265600204467773,
      "memory(GiB)": 77.56,
      "step": 100875,
      "token_acc": 0.5482625482625483,
      "train_speed(iter/s)": 1.438785
    },
    {
      "epoch": 4.322008482927038,
      "grad_norm": 5.522739410400391,
      "learning_rate": 4.468591291358304e-06,
      "loss": 2.0621692657470705,
      "memory(GiB)": 77.56,
      "step": 100880,
      "token_acc": 0.546875,
      "train_speed(iter/s)": 1.438787
    },
    {
      "epoch": 4.3222226982562875,
      "grad_norm": 5.56320333480835,
      "learning_rate": 4.465810781695279e-06,
      "loss": 2.2735784530639647,
      "memory(GiB)": 77.56,
      "step": 100885,
      "token_acc": 0.5173611111111112,
      "train_speed(iter/s)": 1.438801
    },
    {
      "epoch": 4.322436913585536,
      "grad_norm": 6.924131393432617,
      "learning_rate": 4.463031096926806e-06,
      "loss": 2.1173870086669924,
      "memory(GiB)": 77.56,
      "step": 100890,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.438804
    },
    {
      "epoch": 4.322651128914785,
      "grad_norm": 9.428475379943848,
      "learning_rate": 4.460252237103235e-06,
      "loss": 2.134607696533203,
      "memory(GiB)": 77.56,
      "step": 100895,
      "token_acc": 0.5547169811320755,
      "train_speed(iter/s)": 1.438816
    },
    {
      "epoch": 4.322865344244034,
      "grad_norm": 5.395373344421387,
      "learning_rate": 4.457474202274925e-06,
      "loss": 2.0755924224853515,
      "memory(GiB)": 77.56,
      "step": 100900,
      "token_acc": 0.5684647302904564,
      "train_speed(iter/s)": 1.438822
    },
    {
      "epoch": 4.323079559573283,
      "grad_norm": 7.781178951263428,
      "learning_rate": 4.454696992492185e-06,
      "loss": 2.2073802947998047,
      "memory(GiB)": 77.56,
      "step": 100905,
      "token_acc": 0.5282258064516129,
      "train_speed(iter/s)": 1.438836
    },
    {
      "epoch": 4.323293774902532,
      "grad_norm": 5.264420032501221,
      "learning_rate": 4.451920607805343e-06,
      "loss": 2.2307435989379885,
      "memory(GiB)": 77.56,
      "step": 100910,
      "token_acc": 0.5306859205776173,
      "train_speed(iter/s)": 1.438828
    },
    {
      "epoch": 4.323507990231781,
      "grad_norm": 5.754848957061768,
      "learning_rate": 4.44914504826468e-06,
      "loss": 2.3392467498779297,
      "memory(GiB)": 77.56,
      "step": 100915,
      "token_acc": 0.4942528735632184,
      "train_speed(iter/s)": 1.438813
    },
    {
      "epoch": 4.32372220556103,
      "grad_norm": 6.3487653732299805,
      "learning_rate": 4.446370313920489e-06,
      "loss": 2.3639129638671874,
      "memory(GiB)": 77.56,
      "step": 100920,
      "token_acc": 0.503125,
      "train_speed(iter/s)": 1.438844
    },
    {
      "epoch": 4.323936420890279,
      "grad_norm": 7.507499694824219,
      "learning_rate": 4.44359640482302e-06,
      "loss": 2.2884601593017577,
      "memory(GiB)": 77.56,
      "step": 100925,
      "token_acc": 0.4959349593495935,
      "train_speed(iter/s)": 1.438819
    },
    {
      "epoch": 4.324150636219528,
      "grad_norm": 5.1254143714904785,
      "learning_rate": 4.440823321022541e-06,
      "loss": 2.084254837036133,
      "memory(GiB)": 77.56,
      "step": 100930,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.438812
    },
    {
      "epoch": 4.3243648515487765,
      "grad_norm": 6.045876979827881,
      "learning_rate": 4.438051062569293e-06,
      "loss": 2.577326202392578,
      "memory(GiB)": 77.56,
      "step": 100935,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.438821
    },
    {
      "epoch": 4.324579066878026,
      "grad_norm": 6.5383806228637695,
      "learning_rate": 4.435279629513489e-06,
      "loss": 2.339593505859375,
      "memory(GiB)": 77.56,
      "step": 100940,
      "token_acc": 0.43573667711598746,
      "train_speed(iter/s)": 1.438824
    },
    {
      "epoch": 4.324793282207275,
      "grad_norm": 5.796730041503906,
      "learning_rate": 4.432509021905334e-06,
      "loss": 2.5958990097045898,
      "memory(GiB)": 77.56,
      "step": 100945,
      "token_acc": 0.44,
      "train_speed(iter/s)": 1.438822
    },
    {
      "epoch": 4.325007497536523,
      "grad_norm": 5.288985729217529,
      "learning_rate": 4.4297392397950334e-06,
      "loss": 2.131447601318359,
      "memory(GiB)": 77.56,
      "step": 100950,
      "token_acc": 0.550561797752809,
      "train_speed(iter/s)": 1.438842
    },
    {
      "epoch": 4.325221712865773,
      "grad_norm": 6.5406904220581055,
      "learning_rate": 4.42697028323274e-06,
      "loss": 2.53708553314209,
      "memory(GiB)": 77.56,
      "step": 100955,
      "token_acc": 0.5172413793103449,
      "train_speed(iter/s)": 1.438847
    },
    {
      "epoch": 4.325435928195022,
      "grad_norm": 7.722628116607666,
      "learning_rate": 4.424202152268642e-06,
      "loss": 2.323473358154297,
      "memory(GiB)": 77.56,
      "step": 100960,
      "token_acc": 0.4787234042553192,
      "train_speed(iter/s)": 1.438837
    },
    {
      "epoch": 4.32565014352427,
      "grad_norm": 7.154002666473389,
      "learning_rate": 4.421434846952871e-06,
      "loss": 2.1355976104736327,
      "memory(GiB)": 77.56,
      "step": 100965,
      "token_acc": 0.4952978056426332,
      "train_speed(iter/s)": 1.438837
    },
    {
      "epoch": 4.3258643588535195,
      "grad_norm": 4.168364524841309,
      "learning_rate": 4.4186683673355714e-06,
      "loss": 2.0892311096191407,
      "memory(GiB)": 77.56,
      "step": 100970,
      "token_acc": 0.5490196078431373,
      "train_speed(iter/s)": 1.438834
    },
    {
      "epoch": 4.326078574182769,
      "grad_norm": 5.667438507080078,
      "learning_rate": 4.415902713466846e-06,
      "loss": 2.48551025390625,
      "memory(GiB)": 77.56,
      "step": 100975,
      "token_acc": 0.45098039215686275,
      "train_speed(iter/s)": 1.438838
    },
    {
      "epoch": 4.326292789512017,
      "grad_norm": 5.668957233428955,
      "learning_rate": 4.413137885396795e-06,
      "loss": 2.481974220275879,
      "memory(GiB)": 77.56,
      "step": 100980,
      "token_acc": 0.49624060150375937,
      "train_speed(iter/s)": 1.438847
    },
    {
      "epoch": 4.326507004841266,
      "grad_norm": 5.5176568031311035,
      "learning_rate": 4.4103738831755285e-06,
      "loss": 2.5879791259765623,
      "memory(GiB)": 77.56,
      "step": 100985,
      "token_acc": 0.48014440433212996,
      "train_speed(iter/s)": 1.438841
    },
    {
      "epoch": 4.326721220170516,
      "grad_norm": 5.67827033996582,
      "learning_rate": 4.407610706853105e-06,
      "loss": 2.2679656982421874,
      "memory(GiB)": 77.56,
      "step": 100990,
      "token_acc": 0.4952076677316294,
      "train_speed(iter/s)": 1.43886
    },
    {
      "epoch": 4.326935435499764,
      "grad_norm": 5.687675476074219,
      "learning_rate": 4.404848356479585e-06,
      "loss": 2.174545669555664,
      "memory(GiB)": 77.56,
      "step": 100995,
      "token_acc": 0.5147058823529411,
      "train_speed(iter/s)": 1.438869
    },
    {
      "epoch": 4.327149650829013,
      "grad_norm": 6.245078086853027,
      "learning_rate": 4.402086832104996e-06,
      "loss": 2.2083816528320312,
      "memory(GiB)": 77.56,
      "step": 101000,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.438879
    },
    {
      "epoch": 4.327149650829013,
      "eval_loss": 2.0820162296295166,
      "eval_runtime": 14.3835,
      "eval_samples_per_second": 6.952,
      "eval_steps_per_second": 6.952,
      "eval_token_acc": 0.48262032085561496,
      "step": 101000
    },
    {
      "epoch": 4.3273638661582625,
      "grad_norm": 8.662749290466309,
      "learning_rate": 4.399326133779386e-06,
      "loss": 2.3326534271240233,
      "memory(GiB)": 77.56,
      "step": 101005,
      "token_acc": 0.4814453125,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 4.327578081487511,
      "grad_norm": 6.474173069000244,
      "learning_rate": 4.396566261552765e-06,
      "loss": 2.2989614486694334,
      "memory(GiB)": 77.56,
      "step": 101010,
      "token_acc": 0.5028901734104047,
      "train_speed(iter/s)": 1.438565
    },
    {
      "epoch": 4.32779229681676,
      "grad_norm": 6.23170280456543,
      "learning_rate": 4.393807215475115e-06,
      "loss": 2.3394250869750977,
      "memory(GiB)": 77.56,
      "step": 101015,
      "token_acc": 0.506896551724138,
      "train_speed(iter/s)": 1.438564
    },
    {
      "epoch": 4.328006512146009,
      "grad_norm": 7.6872687339782715,
      "learning_rate": 4.3910489955964365e-06,
      "loss": 1.8481668472290038,
      "memory(GiB)": 77.56,
      "step": 101020,
      "token_acc": 0.602112676056338,
      "train_speed(iter/s)": 1.438575
    },
    {
      "epoch": 4.328220727475258,
      "grad_norm": 6.086094856262207,
      "learning_rate": 4.388291601966688e-06,
      "loss": 2.4521799087524414,
      "memory(GiB)": 77.56,
      "step": 101025,
      "token_acc": 0.49295774647887325,
      "train_speed(iter/s)": 1.438588
    },
    {
      "epoch": 4.328434942804507,
      "grad_norm": 8.996162414550781,
      "learning_rate": 4.385535034635829e-06,
      "loss": 2.3587310791015623,
      "memory(GiB)": 77.56,
      "step": 101030,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.438591
    },
    {
      "epoch": 4.328649158133756,
      "grad_norm": 6.893521785736084,
      "learning_rate": 4.382779293653788e-06,
      "loss": 2.069912910461426,
      "memory(GiB)": 77.56,
      "step": 101035,
      "token_acc": 0.5331125827814569,
      "train_speed(iter/s)": 1.438614
    },
    {
      "epoch": 4.328863373463005,
      "grad_norm": 6.17027473449707,
      "learning_rate": 4.380024379070491e-06,
      "loss": 2.389052391052246,
      "memory(GiB)": 77.56,
      "step": 101040,
      "token_acc": 0.48502994011976047,
      "train_speed(iter/s)": 1.438622
    },
    {
      "epoch": 4.329077588792254,
      "grad_norm": 6.016664981842041,
      "learning_rate": 4.377270290935848e-06,
      "loss": 2.3577083587646483,
      "memory(GiB)": 77.56,
      "step": 101045,
      "token_acc": 0.46546546546546547,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 4.329291804121503,
      "grad_norm": 8.608914375305176,
      "learning_rate": 4.374517029299757e-06,
      "loss": 2.3155344009399412,
      "memory(GiB)": 77.56,
      "step": 101050,
      "token_acc": 0.5061728395061729,
      "train_speed(iter/s)": 1.438644
    },
    {
      "epoch": 4.3295060194507515,
      "grad_norm": 6.313039779663086,
      "learning_rate": 4.3717645942120855e-06,
      "loss": 2.2667036056518555,
      "memory(GiB)": 77.56,
      "step": 101055,
      "token_acc": 0.5322033898305085,
      "train_speed(iter/s)": 1.438648
    },
    {
      "epoch": 4.329720234780001,
      "grad_norm": 4.852303981781006,
      "learning_rate": 4.369012985722704e-06,
      "loss": 1.8334367752075196,
      "memory(GiB)": 77.56,
      "step": 101060,
      "token_acc": 0.6038461538461538,
      "train_speed(iter/s)": 1.438659
    },
    {
      "epoch": 4.32993445010925,
      "grad_norm": 5.725183486938477,
      "learning_rate": 4.366262203881461e-06,
      "loss": 2.0838666915893556,
      "memory(GiB)": 77.56,
      "step": 101065,
      "token_acc": 0.5844155844155844,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 4.330148665438498,
      "grad_norm": 5.849481582641602,
      "learning_rate": 4.363512248738172e-06,
      "loss": 2.5036239624023438,
      "memory(GiB)": 77.56,
      "step": 101070,
      "token_acc": 0.5157232704402516,
      "train_speed(iter/s)": 1.438666
    },
    {
      "epoch": 4.330362880767748,
      "grad_norm": 6.329677104949951,
      "learning_rate": 4.3607631203426704e-06,
      "loss": 2.133966064453125,
      "memory(GiB)": 77.56,
      "step": 101075,
      "token_acc": 0.5471014492753623,
      "train_speed(iter/s)": 1.438674
    },
    {
      "epoch": 4.330577096096997,
      "grad_norm": 9.487719535827637,
      "learning_rate": 4.358014818744765e-06,
      "loss": 2.397746467590332,
      "memory(GiB)": 77.56,
      "step": 101080,
      "token_acc": 0.49635036496350365,
      "train_speed(iter/s)": 1.438684
    },
    {
      "epoch": 4.330791311426245,
      "grad_norm": 6.407149314880371,
      "learning_rate": 4.355267343994241e-06,
      "loss": 2.0790287017822267,
      "memory(GiB)": 77.56,
      "step": 101085,
      "token_acc": 0.549800796812749,
      "train_speed(iter/s)": 1.438693
    },
    {
      "epoch": 4.3310055267554946,
      "grad_norm": 5.770221710205078,
      "learning_rate": 4.352520696140866e-06,
      "loss": 2.3384258270263674,
      "memory(GiB)": 77.56,
      "step": 101090,
      "token_acc": 0.4965277777777778,
      "train_speed(iter/s)": 1.438699
    },
    {
      "epoch": 4.331219742084744,
      "grad_norm": 7.723062038421631,
      "learning_rate": 4.349774875234397e-06,
      "loss": 2.349536895751953,
      "memory(GiB)": 77.56,
      "step": 101095,
      "token_acc": 0.5048231511254019,
      "train_speed(iter/s)": 1.438695
    },
    {
      "epoch": 4.331433957413992,
      "grad_norm": 5.800660133361816,
      "learning_rate": 4.3470298813245715e-06,
      "loss": 2.1011037826538086,
      "memory(GiB)": 77.56,
      "step": 101100,
      "token_acc": 0.5457627118644067,
      "train_speed(iter/s)": 1.438695
    },
    {
      "epoch": 4.331648172743241,
      "grad_norm": 6.468334197998047,
      "learning_rate": 4.344285714461138e-06,
      "loss": 2.0974365234375,
      "memory(GiB)": 77.56,
      "step": 101105,
      "token_acc": 0.545774647887324,
      "train_speed(iter/s)": 1.438705
    },
    {
      "epoch": 4.331862388072491,
      "grad_norm": 7.865931510925293,
      "learning_rate": 4.341542374693791e-06,
      "loss": 2.2582515716552733,
      "memory(GiB)": 77.56,
      "step": 101110,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.438698
    },
    {
      "epoch": 4.332076603401739,
      "grad_norm": 4.625708103179932,
      "learning_rate": 4.3387998620722405e-06,
      "loss": 2.579315757751465,
      "memory(GiB)": 77.56,
      "step": 101115,
      "token_acc": 0.493006993006993,
      "train_speed(iter/s)": 1.438709
    },
    {
      "epoch": 4.332290818730988,
      "grad_norm": 5.9434356689453125,
      "learning_rate": 4.336058176646163e-06,
      "loss": 2.4340280532836913,
      "memory(GiB)": 77.56,
      "step": 101120,
      "token_acc": 0.49808429118773945,
      "train_speed(iter/s)": 1.438704
    },
    {
      "epoch": 4.332505034060238,
      "grad_norm": 4.637039661407471,
      "learning_rate": 4.3333173184652185e-06,
      "loss": 2.443637466430664,
      "memory(GiB)": 77.56,
      "step": 101125,
      "token_acc": 0.5109489051094891,
      "train_speed(iter/s)": 1.4387
    },
    {
      "epoch": 4.332719249389486,
      "grad_norm": 5.87703275680542,
      "learning_rate": 4.330577287579079e-06,
      "loss": 2.3531347274780274,
      "memory(GiB)": 77.56,
      "step": 101130,
      "token_acc": 0.5105633802816901,
      "train_speed(iter/s)": 1.438698
    },
    {
      "epoch": 4.332933464718735,
      "grad_norm": 4.111064434051514,
      "learning_rate": 4.327838084037372e-06,
      "loss": 2.38293514251709,
      "memory(GiB)": 77.56,
      "step": 101135,
      "token_acc": 0.4930555555555556,
      "train_speed(iter/s)": 1.438711
    },
    {
      "epoch": 4.3331476800479845,
      "grad_norm": 4.8027663230896,
      "learning_rate": 4.325099707889724e-06,
      "loss": 2.373774528503418,
      "memory(GiB)": 77.56,
      "step": 101140,
      "token_acc": 0.5059880239520959,
      "train_speed(iter/s)": 1.438719
    },
    {
      "epoch": 4.333361895377233,
      "grad_norm": 6.752740383148193,
      "learning_rate": 4.322362159185739e-06,
      "loss": 2.351709175109863,
      "memory(GiB)": 77.56,
      "step": 101145,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438718
    },
    {
      "epoch": 4.333576110706482,
      "grad_norm": 7.650457859039307,
      "learning_rate": 4.319625437975006e-06,
      "loss": 2.3520641326904297,
      "memory(GiB)": 77.56,
      "step": 101150,
      "token_acc": 0.5127272727272727,
      "train_speed(iter/s)": 1.438725
    },
    {
      "epoch": 4.333790326035731,
      "grad_norm": 6.833845138549805,
      "learning_rate": 4.3168895443071124e-06,
      "loss": 2.153400993347168,
      "memory(GiB)": 77.56,
      "step": 101155,
      "token_acc": 0.5193548387096775,
      "train_speed(iter/s)": 1.438744
    },
    {
      "epoch": 4.33400454136498,
      "grad_norm": 6.676098346710205,
      "learning_rate": 4.314154478231619e-06,
      "loss": 2.3579586029052733,
      "memory(GiB)": 77.56,
      "step": 101160,
      "token_acc": 0.5202492211838006,
      "train_speed(iter/s)": 1.438749
    },
    {
      "epoch": 4.334218756694229,
      "grad_norm": 6.239243507385254,
      "learning_rate": 4.3114202397980755e-06,
      "loss": 2.4325647354125977,
      "memory(GiB)": 77.56,
      "step": 101165,
      "token_acc": 0.45454545454545453,
      "train_speed(iter/s)": 1.438739
    },
    {
      "epoch": 4.334432972023478,
      "grad_norm": 5.6411309242248535,
      "learning_rate": 4.308686829056019e-06,
      "loss": 2.430034637451172,
      "memory(GiB)": 77.56,
      "step": 101170,
      "token_acc": 0.45964912280701753,
      "train_speed(iter/s)": 1.438736
    },
    {
      "epoch": 4.334647187352727,
      "grad_norm": 6.036357402801514,
      "learning_rate": 4.305954246054961e-06,
      "loss": 2.224743461608887,
      "memory(GiB)": 77.56,
      "step": 101175,
      "token_acc": 0.5119453924914675,
      "train_speed(iter/s)": 1.438735
    },
    {
      "epoch": 4.334861402681976,
      "grad_norm": 6.004115104675293,
      "learning_rate": 4.303222490844405e-06,
      "loss": 2.494205093383789,
      "memory(GiB)": 77.56,
      "step": 101180,
      "token_acc": 0.4515235457063712,
      "train_speed(iter/s)": 1.438745
    },
    {
      "epoch": 4.335075618011225,
      "grad_norm": 7.252013206481934,
      "learning_rate": 4.30049156347383e-06,
      "loss": 2.3625343322753904,
      "memory(GiB)": 77.56,
      "step": 101185,
      "token_acc": 0.49473684210526314,
      "train_speed(iter/s)": 1.438754
    },
    {
      "epoch": 4.3352898333404735,
      "grad_norm": 10.650192260742188,
      "learning_rate": 4.297761463992733e-06,
      "loss": 2.4016536712646483,
      "memory(GiB)": 77.56,
      "step": 101190,
      "token_acc": 0.5319148936170213,
      "train_speed(iter/s)": 1.438771
    },
    {
      "epoch": 4.335504048669723,
      "grad_norm": 7.285583972930908,
      "learning_rate": 4.29503219245056e-06,
      "loss": 2.0923715591430665,
      "memory(GiB)": 77.56,
      "step": 101195,
      "token_acc": 0.5442622950819672,
      "train_speed(iter/s)": 1.438759
    },
    {
      "epoch": 4.335718263998972,
      "grad_norm": 7.679899215698242,
      "learning_rate": 4.292303748896748e-06,
      "loss": 2.3973468780517577,
      "memory(GiB)": 77.56,
      "step": 101200,
      "token_acc": 0.5154639175257731,
      "train_speed(iter/s)": 1.438773
    },
    {
      "epoch": 4.33593247932822,
      "grad_norm": 4.87914514541626,
      "learning_rate": 4.289576133380735e-06,
      "loss": 2.195694160461426,
      "memory(GiB)": 77.56,
      "step": 101205,
      "token_acc": 0.4986376021798365,
      "train_speed(iter/s)": 1.438754
    },
    {
      "epoch": 4.33614669465747,
      "grad_norm": 7.900181293487549,
      "learning_rate": 4.286849345951932e-06,
      "loss": 2.3426822662353515,
      "memory(GiB)": 77.56,
      "step": 101210,
      "token_acc": 0.48366013071895425,
      "train_speed(iter/s)": 1.438747
    },
    {
      "epoch": 4.336360909986719,
      "grad_norm": 7.688501834869385,
      "learning_rate": 4.284123386659722e-06,
      "loss": 2.1559396743774415,
      "memory(GiB)": 77.56,
      "step": 101215,
      "token_acc": 0.5458715596330275,
      "train_speed(iter/s)": 1.438751
    },
    {
      "epoch": 4.336575125315967,
      "grad_norm": 5.1097002029418945,
      "learning_rate": 4.281398255553515e-06,
      "loss": 2.0712360382080077,
      "memory(GiB)": 77.56,
      "step": 101220,
      "token_acc": 0.5292307692307693,
      "train_speed(iter/s)": 1.438759
    },
    {
      "epoch": 4.3367893406452165,
      "grad_norm": 6.921679973602295,
      "learning_rate": 4.278673952682655e-06,
      "loss": 2.407971954345703,
      "memory(GiB)": 77.56,
      "step": 101225,
      "token_acc": 0.49050632911392406,
      "train_speed(iter/s)": 1.438766
    },
    {
      "epoch": 4.337003555974466,
      "grad_norm": 7.617849349975586,
      "learning_rate": 4.275950478096513e-06,
      "loss": 2.2217613220214845,
      "memory(GiB)": 77.56,
      "step": 101230,
      "token_acc": 0.527027027027027,
      "train_speed(iter/s)": 1.43877
    },
    {
      "epoch": 4.337217771303714,
      "grad_norm": 7.503749847412109,
      "learning_rate": 4.273227831844423e-06,
      "loss": 2.209715461730957,
      "memory(GiB)": 77.56,
      "step": 101235,
      "token_acc": 0.4968944099378882,
      "train_speed(iter/s)": 1.438782
    },
    {
      "epoch": 4.337431986632963,
      "grad_norm": 5.194855213165283,
      "learning_rate": 4.2705060139757054e-06,
      "loss": 2.44457893371582,
      "memory(GiB)": 77.56,
      "step": 101240,
      "token_acc": 0.5123287671232877,
      "train_speed(iter/s)": 1.438781
    },
    {
      "epoch": 4.337646201962213,
      "grad_norm": 5.237576484680176,
      "learning_rate": 4.267785024539661e-06,
      "loss": 2.204781913757324,
      "memory(GiB)": 77.56,
      "step": 101245,
      "token_acc": 0.5330882352941176,
      "train_speed(iter/s)": 1.43879
    },
    {
      "epoch": 4.337860417291461,
      "grad_norm": 6.590160846710205,
      "learning_rate": 4.2650648635856e-06,
      "loss": 2.4221874237060548,
      "memory(GiB)": 77.56,
      "step": 101250,
      "token_acc": 0.4657039711191336,
      "train_speed(iter/s)": 1.438804
    },
    {
      "epoch": 4.33807463262071,
      "grad_norm": 7.082057952880859,
      "learning_rate": 4.262345531162792e-06,
      "loss": 2.2764734268188476,
      "memory(GiB)": 77.56,
      "step": 101255,
      "token_acc": 0.49097472924187724,
      "train_speed(iter/s)": 1.438804
    },
    {
      "epoch": 4.3382888479499595,
      "grad_norm": 8.069806098937988,
      "learning_rate": 4.259627027320501e-06,
      "loss": 2.6750423431396486,
      "memory(GiB)": 77.56,
      "step": 101260,
      "token_acc": 0.47527472527472525,
      "train_speed(iter/s)": 1.438807
    },
    {
      "epoch": 4.338503063279208,
      "grad_norm": 5.843471050262451,
      "learning_rate": 4.256909352107969e-06,
      "loss": 2.217774200439453,
      "memory(GiB)": 77.56,
      "step": 101265,
      "token_acc": 0.5335689045936396,
      "train_speed(iter/s)": 1.438779
    },
    {
      "epoch": 4.338717278608457,
      "grad_norm": 6.1159491539001465,
      "learning_rate": 4.254192505574433e-06,
      "loss": 2.247700500488281,
      "memory(GiB)": 77.56,
      "step": 101270,
      "token_acc": 0.5396825396825397,
      "train_speed(iter/s)": 1.438781
    },
    {
      "epoch": 4.338931493937706,
      "grad_norm": 6.537121295928955,
      "learning_rate": 4.2514764877691184e-06,
      "loss": 2.1320571899414062,
      "memory(GiB)": 77.56,
      "step": 101275,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.438781
    },
    {
      "epoch": 4.339145709266955,
      "grad_norm": 5.182265758514404,
      "learning_rate": 4.2487612987412215e-06,
      "loss": 2.166476821899414,
      "memory(GiB)": 77.56,
      "step": 101280,
      "token_acc": 0.5013404825737265,
      "train_speed(iter/s)": 1.438781
    },
    {
      "epoch": 4.339359924596204,
      "grad_norm": 6.607207775115967,
      "learning_rate": 4.246046938539933e-06,
      "loss": 2.3562170028686524,
      "memory(GiB)": 77.56,
      "step": 101285,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.438781
    },
    {
      "epoch": 4.339574139925453,
      "grad_norm": 6.849672317504883,
      "learning_rate": 4.243333407214422e-06,
      "loss": 2.2435211181640624,
      "memory(GiB)": 77.56,
      "step": 101290,
      "token_acc": 0.5174825174825175,
      "train_speed(iter/s)": 1.438782
    },
    {
      "epoch": 4.339788355254702,
      "grad_norm": 5.224052906036377,
      "learning_rate": 4.240620704813847e-06,
      "loss": 2.089272117614746,
      "memory(GiB)": 77.56,
      "step": 101295,
      "token_acc": 0.5416666666666666,
      "train_speed(iter/s)": 1.438787
    },
    {
      "epoch": 4.340002570583951,
      "grad_norm": 6.374125003814697,
      "learning_rate": 4.23790883138736e-06,
      "loss": 2.4972208023071287,
      "memory(GiB)": 77.56,
      "step": 101300,
      "token_acc": 0.4880239520958084,
      "train_speed(iter/s)": 1.438792
    },
    {
      "epoch": 4.3402167859132,
      "grad_norm": 5.530206203460693,
      "learning_rate": 4.235197786984074e-06,
      "loss": 2.1429861068725584,
      "memory(GiB)": 77.56,
      "step": 101305,
      "token_acc": 0.5565217391304348,
      "train_speed(iter/s)": 1.438789
    },
    {
      "epoch": 4.3404310012424485,
      "grad_norm": 8.494129180908203,
      "learning_rate": 4.232487571653115e-06,
      "loss": 1.9244573593139649,
      "memory(GiB)": 77.56,
      "step": 101310,
      "token_acc": 0.5615384615384615,
      "train_speed(iter/s)": 1.438804
    },
    {
      "epoch": 4.340645216571698,
      "grad_norm": 5.509167194366455,
      "learning_rate": 4.229778185443583e-06,
      "loss": 2.167266845703125,
      "memory(GiB)": 77.56,
      "step": 101315,
      "token_acc": 0.5387453874538746,
      "train_speed(iter/s)": 1.438791
    },
    {
      "epoch": 4.340859431900947,
      "grad_norm": 6.201045513153076,
      "learning_rate": 4.227069628404551e-06,
      "loss": 2.067679023742676,
      "memory(GiB)": 77.56,
      "step": 101320,
      "token_acc": 0.5020242914979757,
      "train_speed(iter/s)": 1.438786
    },
    {
      "epoch": 4.341073647230195,
      "grad_norm": 9.918316841125488,
      "learning_rate": 4.224361900585094e-06,
      "loss": 1.9724859237670898,
      "memory(GiB)": 77.56,
      "step": 101325,
      "token_acc": 0.5564516129032258,
      "train_speed(iter/s)": 1.438791
    },
    {
      "epoch": 4.341287862559445,
      "grad_norm": 6.729365348815918,
      "learning_rate": 4.221655002034253e-06,
      "loss": 2.2674991607666017,
      "memory(GiB)": 77.56,
      "step": 101330,
      "token_acc": 0.4864864864864865,
      "train_speed(iter/s)": 1.438787
    },
    {
      "epoch": 4.341502077888694,
      "grad_norm": 6.806403636932373,
      "learning_rate": 4.218948932801087e-06,
      "loss": 2.418392372131348,
      "memory(GiB)": 77.56,
      "step": 101335,
      "token_acc": 0.4940239043824701,
      "train_speed(iter/s)": 1.438794
    },
    {
      "epoch": 4.341716293217942,
      "grad_norm": 6.020773887634277,
      "learning_rate": 4.216243692934602e-06,
      "loss": 2.2918224334716797,
      "memory(GiB)": 77.56,
      "step": 101340,
      "token_acc": 0.5094339622641509,
      "train_speed(iter/s)": 1.438798
    },
    {
      "epoch": 4.3419305085471915,
      "grad_norm": 6.5184221267700195,
      "learning_rate": 4.213539282483814e-06,
      "loss": 2.542782020568848,
      "memory(GiB)": 77.56,
      "step": 101345,
      "token_acc": 0.4965753424657534,
      "train_speed(iter/s)": 1.438802
    },
    {
      "epoch": 4.342144723876441,
      "grad_norm": 7.2213568687438965,
      "learning_rate": 4.210835701497712e-06,
      "loss": 2.598005485534668,
      "memory(GiB)": 77.56,
      "step": 101350,
      "token_acc": 0.4755244755244755,
      "train_speed(iter/s)": 1.438799
    },
    {
      "epoch": 4.342358939205689,
      "grad_norm": 7.181949615478516,
      "learning_rate": 4.208132950025273e-06,
      "loss": 2.2167888641357423,
      "memory(GiB)": 77.56,
      "step": 101355,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.4388
    },
    {
      "epoch": 4.342573154534938,
      "grad_norm": 7.149319648742676,
      "learning_rate": 4.205431028115458e-06,
      "loss": 2.608908462524414,
      "memory(GiB)": 77.56,
      "step": 101360,
      "token_acc": 0.49236641221374045,
      "train_speed(iter/s)": 1.438818
    },
    {
      "epoch": 4.342787369864188,
      "grad_norm": 6.05332088470459,
      "learning_rate": 4.202729935817224e-06,
      "loss": 2.5681474685668944,
      "memory(GiB)": 77.56,
      "step": 101365,
      "token_acc": 0.4732824427480916,
      "train_speed(iter/s)": 1.438832
    },
    {
      "epoch": 4.343001585193436,
      "grad_norm": 5.106203556060791,
      "learning_rate": 4.20002967317949e-06,
      "loss": 2.19384822845459,
      "memory(GiB)": 77.56,
      "step": 101370,
      "token_acc": 0.5227963525835866,
      "train_speed(iter/s)": 1.438829
    },
    {
      "epoch": 4.343215800522685,
      "grad_norm": 6.929333209991455,
      "learning_rate": 4.19733024025119e-06,
      "loss": 2.2576053619384764,
      "memory(GiB)": 77.56,
      "step": 101375,
      "token_acc": 0.4723127035830619,
      "train_speed(iter/s)": 1.438824
    },
    {
      "epoch": 4.3434300158519346,
      "grad_norm": 5.270244598388672,
      "learning_rate": 4.19463163708122e-06,
      "loss": 1.8078964233398438,
      "memory(GiB)": 77.56,
      "step": 101380,
      "token_acc": 0.5830721003134797,
      "train_speed(iter/s)": 1.43883
    },
    {
      "epoch": 4.343644231181183,
      "grad_norm": 5.989297389984131,
      "learning_rate": 4.191933863718467e-06,
      "loss": 2.2562576293945313,
      "memory(GiB)": 77.56,
      "step": 101385,
      "token_acc": 0.5247148288973384,
      "train_speed(iter/s)": 1.438844
    },
    {
      "epoch": 4.343858446510432,
      "grad_norm": 7.0800065994262695,
      "learning_rate": 4.189236920211798e-06,
      "loss": 2.221025848388672,
      "memory(GiB)": 77.56,
      "step": 101390,
      "token_acc": 0.5183823529411765,
      "train_speed(iter/s)": 1.438855
    },
    {
      "epoch": 4.344072661839681,
      "grad_norm": 6.620867729187012,
      "learning_rate": 4.1865408066100785e-06,
      "loss": 2.088029479980469,
      "memory(GiB)": 77.56,
      "step": 101395,
      "token_acc": 0.5067114093959731,
      "train_speed(iter/s)": 1.438868
    },
    {
      "epoch": 4.34428687716893,
      "grad_norm": 5.606696128845215,
      "learning_rate": 4.183845522962154e-06,
      "loss": 2.7019441604614256,
      "memory(GiB)": 77.56,
      "step": 101400,
      "token_acc": 0.4634146341463415,
      "train_speed(iter/s)": 1.438864
    },
    {
      "epoch": 4.344501092498179,
      "grad_norm": 9.510915756225586,
      "learning_rate": 4.181151069316846e-06,
      "loss": 2.212398719787598,
      "memory(GiB)": 77.56,
      "step": 101405,
      "token_acc": 0.5494505494505495,
      "train_speed(iter/s)": 1.438869
    },
    {
      "epoch": 4.344715307827428,
      "grad_norm": 5.622706413269043,
      "learning_rate": 4.178457445722967e-06,
      "loss": 2.5402738571166994,
      "memory(GiB)": 77.56,
      "step": 101410,
      "token_acc": 0.446064139941691,
      "train_speed(iter/s)": 1.438842
    },
    {
      "epoch": 4.344929523156677,
      "grad_norm": 5.042716979980469,
      "learning_rate": 4.175764652229308e-06,
      "loss": 2.2626226425170897,
      "memory(GiB)": 77.56,
      "step": 101415,
      "token_acc": 0.5049180327868853,
      "train_speed(iter/s)": 1.438852
    },
    {
      "epoch": 4.345143738485926,
      "grad_norm": 5.740492820739746,
      "learning_rate": 4.173072688884666e-06,
      "loss": 2.135538864135742,
      "memory(GiB)": 77.56,
      "step": 101420,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43885
    },
    {
      "epoch": 4.345357953815175,
      "grad_norm": 5.865013599395752,
      "learning_rate": 4.170381555737801e-06,
      "loss": 2.191912078857422,
      "memory(GiB)": 77.56,
      "step": 101425,
      "token_acc": 0.4659498207885305,
      "train_speed(iter/s)": 1.438855
    },
    {
      "epoch": 4.345572169144424,
      "grad_norm": 5.078368186950684,
      "learning_rate": 4.167691252837463e-06,
      "loss": 2.3704860687255858,
      "memory(GiB)": 77.56,
      "step": 101430,
      "token_acc": 0.4868804664723032,
      "train_speed(iter/s)": 1.438866
    },
    {
      "epoch": 4.345786384473673,
      "grad_norm": 4.758178234100342,
      "learning_rate": 4.165001780232397e-06,
      "loss": 1.9802011489868163,
      "memory(GiB)": 77.56,
      "step": 101435,
      "token_acc": 0.5574912891986062,
      "train_speed(iter/s)": 1.43888
    },
    {
      "epoch": 4.346000599802922,
      "grad_norm": 7.279247283935547,
      "learning_rate": 4.1623131379713186e-06,
      "loss": 2.258874702453613,
      "memory(GiB)": 77.56,
      "step": 101440,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43887
    },
    {
      "epoch": 4.3462148151321705,
      "grad_norm": 6.9657511711120605,
      "learning_rate": 4.159625326102934e-06,
      "loss": 2.529164123535156,
      "memory(GiB)": 77.56,
      "step": 101445,
      "token_acc": 0.4652567975830816,
      "train_speed(iter/s)": 1.438873
    },
    {
      "epoch": 4.34642903046142,
      "grad_norm": 6.005016803741455,
      "learning_rate": 4.156938344675932e-06,
      "loss": 2.33060302734375,
      "memory(GiB)": 77.56,
      "step": 101450,
      "token_acc": 0.4945054945054945,
      "train_speed(iter/s)": 1.438882
    },
    {
      "epoch": 4.346643245790669,
      "grad_norm": 5.210018634796143,
      "learning_rate": 4.154252193739006e-06,
      "loss": 2.1816497802734376,
      "memory(GiB)": 77.56,
      "step": 101455,
      "token_acc": 0.5291970802919708,
      "train_speed(iter/s)": 1.438884
    },
    {
      "epoch": 4.346857461119917,
      "grad_norm": 6.272582054138184,
      "learning_rate": 4.151566873340806e-06,
      "loss": 2.108576011657715,
      "memory(GiB)": 77.56,
      "step": 101460,
      "token_acc": 0.5241379310344828,
      "train_speed(iter/s)": 1.438877
    },
    {
      "epoch": 4.347071676449167,
      "grad_norm": 5.846337795257568,
      "learning_rate": 4.148882383529984e-06,
      "loss": 2.2197166442871095,
      "memory(GiB)": 77.56,
      "step": 101465,
      "token_acc": 0.5049180327868853,
      "train_speed(iter/s)": 1.438862
    },
    {
      "epoch": 4.347285891778416,
      "grad_norm": 5.538933753967285,
      "learning_rate": 4.146198724355171e-06,
      "loss": 2.0510019302368163,
      "memory(GiB)": 77.56,
      "step": 101470,
      "token_acc": 0.54,
      "train_speed(iter/s)": 1.438848
    },
    {
      "epoch": 4.347500107107664,
      "grad_norm": 5.194798469543457,
      "learning_rate": 4.143515895864969e-06,
      "loss": 2.1033184051513674,
      "memory(GiB)": 77.56,
      "step": 101475,
      "token_acc": 0.5537974683544303,
      "train_speed(iter/s)": 1.438858
    },
    {
      "epoch": 4.3477143224369135,
      "grad_norm": 8.422808647155762,
      "learning_rate": 4.1408338981080096e-06,
      "loss": 2.1998289108276365,
      "memory(GiB)": 77.56,
      "step": 101480,
      "token_acc": 0.5441696113074205,
      "train_speed(iter/s)": 1.438859
    },
    {
      "epoch": 4.347928537766163,
      "grad_norm": 5.203928470611572,
      "learning_rate": 4.138152731132855e-06,
      "loss": 2.3962709426879885,
      "memory(GiB)": 77.56,
      "step": 101485,
      "token_acc": 0.5393258426966292,
      "train_speed(iter/s)": 1.438862
    },
    {
      "epoch": 4.348142753095411,
      "grad_norm": 6.964390277862549,
      "learning_rate": 4.135472394988094e-06,
      "loss": 2.3716407775878907,
      "memory(GiB)": 77.56,
      "step": 101490,
      "token_acc": 0.5592105263157895,
      "train_speed(iter/s)": 1.43887
    },
    {
      "epoch": 4.34835696842466,
      "grad_norm": 5.814328670501709,
      "learning_rate": 4.1327928897222705e-06,
      "loss": 2.2129642486572267,
      "memory(GiB)": 77.56,
      "step": 101495,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.438872
    },
    {
      "epoch": 4.34857118375391,
      "grad_norm": 6.519754409790039,
      "learning_rate": 4.130114215383929e-06,
      "loss": 2.336849594116211,
      "memory(GiB)": 77.56,
      "step": 101500,
      "token_acc": 0.5322128851540616,
      "train_speed(iter/s)": 1.43889
    },
    {
      "epoch": 4.34857118375391,
      "eval_loss": 2.228541851043701,
      "eval_runtime": 13.7904,
      "eval_samples_per_second": 7.251,
      "eval_steps_per_second": 7.251,
      "eval_token_acc": 0.44966442953020136,
      "step": 101500
    },
    {
      "epoch": 4.348785399083158,
      "grad_norm": 6.421586036682129,
      "learning_rate": 4.1274363720215924e-06,
      "loss": 2.311581039428711,
      "memory(GiB)": 77.56,
      "step": 101505,
      "token_acc": 0.4560229445506692,
      "train_speed(iter/s)": 1.438596
    },
    {
      "epoch": 4.348999614412407,
      "grad_norm": 6.308603286743164,
      "learning_rate": 4.124759359683783e-06,
      "loss": 1.9527198791503906,
      "memory(GiB)": 77.56,
      "step": 101510,
      "token_acc": 0.5738396624472574,
      "train_speed(iter/s)": 1.438586
    },
    {
      "epoch": 4.3492138297416565,
      "grad_norm": 12.110864639282227,
      "learning_rate": 4.122083178418995e-06,
      "loss": 2.024698257446289,
      "memory(GiB)": 77.56,
      "step": 101515,
      "token_acc": 0.5038167938931297,
      "train_speed(iter/s)": 1.438599
    },
    {
      "epoch": 4.349428045070905,
      "grad_norm": 5.848987579345703,
      "learning_rate": 4.119407828275696e-06,
      "loss": 2.2689754486083986,
      "memory(GiB)": 77.56,
      "step": 101520,
      "token_acc": 0.5266903914590747,
      "train_speed(iter/s)": 1.438603
    },
    {
      "epoch": 4.349642260400154,
      "grad_norm": 6.438787460327148,
      "learning_rate": 4.116733309302373e-06,
      "loss": 2.233139991760254,
      "memory(GiB)": 77.56,
      "step": 101525,
      "token_acc": 0.5539033457249071,
      "train_speed(iter/s)": 1.43861
    },
    {
      "epoch": 4.349856475729403,
      "grad_norm": 5.045784950256348,
      "learning_rate": 4.114059621547472e-06,
      "loss": 2.1629079818725585,
      "memory(GiB)": 77.56,
      "step": 101530,
      "token_acc": 0.5317725752508361,
      "train_speed(iter/s)": 1.43862
    },
    {
      "epoch": 4.350070691058653,
      "grad_norm": 5.2980265617370605,
      "learning_rate": 4.111386765059416e-06,
      "loss": 2.513823318481445,
      "memory(GiB)": 77.56,
      "step": 101535,
      "token_acc": 0.49244712990936557,
      "train_speed(iter/s)": 1.438613
    },
    {
      "epoch": 4.350284906387901,
      "grad_norm": 5.629308700561523,
      "learning_rate": 4.108714739886638e-06,
      "loss": 2.1469287872314453,
      "memory(GiB)": 77.56,
      "step": 101540,
      "token_acc": 0.5270758122743683,
      "train_speed(iter/s)": 1.438624
    },
    {
      "epoch": 4.35049912171715,
      "grad_norm": 5.630993843078613,
      "learning_rate": 4.106043546077548e-06,
      "loss": 2.0595304489135744,
      "memory(GiB)": 77.56,
      "step": 101545,
      "token_acc": 0.5429553264604811,
      "train_speed(iter/s)": 1.438622
    },
    {
      "epoch": 4.350713337046399,
      "grad_norm": 5.7489471435546875,
      "learning_rate": 4.103373183680531e-06,
      "loss": 2.268590545654297,
      "memory(GiB)": 77.56,
      "step": 101550,
      "token_acc": 0.52,
      "train_speed(iter/s)": 1.438623
    },
    {
      "epoch": 4.350927552375648,
      "grad_norm": 5.641597270965576,
      "learning_rate": 4.100703652743959e-06,
      "loss": 2.2816795349121093,
      "memory(GiB)": 77.56,
      "step": 101555,
      "token_acc": 0.5214723926380368,
      "train_speed(iter/s)": 1.438596
    },
    {
      "epoch": 4.351141767704897,
      "grad_norm": 6.621870517730713,
      "learning_rate": 4.098034953316194e-06,
      "loss": 2.422886848449707,
      "memory(GiB)": 77.56,
      "step": 101560,
      "token_acc": 0.5123674911660777,
      "train_speed(iter/s)": 1.438595
    },
    {
      "epoch": 4.351355983034146,
      "grad_norm": 6.43054723739624,
      "learning_rate": 4.09536708544559e-06,
      "loss": 2.2933895111083986,
      "memory(GiB)": 77.56,
      "step": 101565,
      "token_acc": 0.511400651465798,
      "train_speed(iter/s)": 1.438597
    },
    {
      "epoch": 4.351570198363395,
      "grad_norm": 6.566086769104004,
      "learning_rate": 4.092700049180476e-06,
      "loss": 2.3913063049316405,
      "memory(GiB)": 77.56,
      "step": 101570,
      "token_acc": 0.46551724137931033,
      "train_speed(iter/s)": 1.438609
    },
    {
      "epoch": 4.351784413692644,
      "grad_norm": 5.274600028991699,
      "learning_rate": 4.090033844569163e-06,
      "loss": 2.3955715179443358,
      "memory(GiB)": 77.56,
      "step": 101575,
      "token_acc": 0.4691358024691358,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 4.351998629021892,
      "grad_norm": 6.710582256317139,
      "learning_rate": 4.087368471659958e-06,
      "loss": 2.0774911880493163,
      "memory(GiB)": 77.56,
      "step": 101580,
      "token_acc": 0.5167785234899329,
      "train_speed(iter/s)": 1.438601
    },
    {
      "epoch": 4.352212844351142,
      "grad_norm": 7.038486480712891,
      "learning_rate": 4.084703930501138e-06,
      "loss": 2.4165790557861326,
      "memory(GiB)": 77.56,
      "step": 101585,
      "token_acc": 0.45695364238410596,
      "train_speed(iter/s)": 1.438597
    },
    {
      "epoch": 4.352427059680391,
      "grad_norm": 4.987838268280029,
      "learning_rate": 4.08204022114097e-06,
      "loss": 2.3569786071777346,
      "memory(GiB)": 77.56,
      "step": 101590,
      "token_acc": 0.48923076923076925,
      "train_speed(iter/s)": 1.438615
    },
    {
      "epoch": 4.35264127500964,
      "grad_norm": 5.530329704284668,
      "learning_rate": 4.079377343627722e-06,
      "loss": 2.36212272644043,
      "memory(GiB)": 77.56,
      "step": 101595,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.438629
    },
    {
      "epoch": 4.3528554903388885,
      "grad_norm": 7.183777809143066,
      "learning_rate": 4.076715298009637e-06,
      "loss": 2.0948135375976564,
      "memory(GiB)": 77.56,
      "step": 101600,
      "token_acc": 0.5285714285714286,
      "train_speed(iter/s)": 1.438643
    },
    {
      "epoch": 4.353069705668138,
      "grad_norm": 6.435431003570557,
      "learning_rate": 4.0740540843349325e-06,
      "loss": 2.3647430419921873,
      "memory(GiB)": 77.56,
      "step": 101605,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438664
    },
    {
      "epoch": 4.353283920997386,
      "grad_norm": 6.683962345123291,
      "learning_rate": 4.071393702651821e-06,
      "loss": 2.464950942993164,
      "memory(GiB)": 77.56,
      "step": 101610,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43866
    },
    {
      "epoch": 4.353498136326635,
      "grad_norm": 7.889510631561279,
      "learning_rate": 4.068734153008497e-06,
      "loss": 2.216324234008789,
      "memory(GiB)": 77.56,
      "step": 101615,
      "token_acc": 0.5441696113074205,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 4.353712351655885,
      "grad_norm": 6.581700801849365,
      "learning_rate": 4.066075435453132e-06,
      "loss": 2.064121055603027,
      "memory(GiB)": 77.56,
      "step": 101620,
      "token_acc": 0.5521885521885522,
      "train_speed(iter/s)": 1.438636
    },
    {
      "epoch": 4.353926566985134,
      "grad_norm": 6.234287261962891,
      "learning_rate": 4.06341755003391e-06,
      "loss": 2.140406036376953,
      "memory(GiB)": 77.56,
      "step": 101625,
      "token_acc": 0.5369127516778524,
      "train_speed(iter/s)": 1.438624
    },
    {
      "epoch": 4.354140782314382,
      "grad_norm": 7.736416339874268,
      "learning_rate": 4.060760496798971e-06,
      "loss": 2.075102615356445,
      "memory(GiB)": 77.56,
      "step": 101630,
      "token_acc": 0.5727699530516432,
      "train_speed(iter/s)": 1.438637
    },
    {
      "epoch": 4.3543549976436315,
      "grad_norm": 5.36174201965332,
      "learning_rate": 4.058104275796449e-06,
      "loss": 2.3465129852294924,
      "memory(GiB)": 77.56,
      "step": 101635,
      "token_acc": 0.5179153094462541,
      "train_speed(iter/s)": 1.438633
    },
    {
      "epoch": 4.35456921297288,
      "grad_norm": 5.539736270904541,
      "learning_rate": 4.0554488870744645e-06,
      "loss": 2.255598449707031,
      "memory(GiB)": 77.56,
      "step": 101640,
      "token_acc": 0.5353535353535354,
      "train_speed(iter/s)": 1.438621
    },
    {
      "epoch": 4.354783428302129,
      "grad_norm": 6.58910608291626,
      "learning_rate": 4.052794330681125e-06,
      "loss": 2.3917633056640626,
      "memory(GiB)": 77.56,
      "step": 101645,
      "token_acc": 0.5044776119402985,
      "train_speed(iter/s)": 1.438622
    },
    {
      "epoch": 4.354997643631378,
      "grad_norm": 5.834932327270508,
      "learning_rate": 4.050140606664505e-06,
      "loss": 2.3838035583496096,
      "memory(GiB)": 77.56,
      "step": 101650,
      "token_acc": 0.4967948717948718,
      "train_speed(iter/s)": 1.438621
    },
    {
      "epoch": 4.355211858960628,
      "grad_norm": 5.4448933601379395,
      "learning_rate": 4.047487715072706e-06,
      "loss": 2.1299243927001954,
      "memory(GiB)": 77.56,
      "step": 101655,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.438628
    },
    {
      "epoch": 4.355426074289876,
      "grad_norm": 6.422670841217041,
      "learning_rate": 4.044835655953772e-06,
      "loss": 2.3987396240234373,
      "memory(GiB)": 77.56,
      "step": 101660,
      "token_acc": 0.4803921568627451,
      "train_speed(iter/s)": 1.438624
    },
    {
      "epoch": 4.355640289619125,
      "grad_norm": 6.349356651306152,
      "learning_rate": 4.042184429355739e-06,
      "loss": 2.0001426696777345,
      "memory(GiB)": 77.56,
      "step": 101665,
      "token_acc": 0.5591397849462365,
      "train_speed(iter/s)": 1.438618
    },
    {
      "epoch": 4.355854504948374,
      "grad_norm": 5.946943759918213,
      "learning_rate": 4.039534035326659e-06,
      "loss": 2.1214082717895506,
      "memory(GiB)": 77.56,
      "step": 101670,
      "token_acc": 0.5420875420875421,
      "train_speed(iter/s)": 1.438596
    },
    {
      "epoch": 4.356068720277623,
      "grad_norm": 7.886771202087402,
      "learning_rate": 4.0368844739145315e-06,
      "loss": 2.4991241455078126,
      "memory(GiB)": 77.56,
      "step": 101675,
      "token_acc": 0.4817073170731707,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 4.356282935606872,
      "grad_norm": 7.021752834320068,
      "learning_rate": 4.034235745167353e-06,
      "loss": 2.363290023803711,
      "memory(GiB)": 77.56,
      "step": 101680,
      "token_acc": 0.47987616099071206,
      "train_speed(iter/s)": 1.43862
    },
    {
      "epoch": 4.356497150936121,
      "grad_norm": 7.677444934844971,
      "learning_rate": 4.031587849133123e-06,
      "loss": 2.320631408691406,
      "memory(GiB)": 77.56,
      "step": 101685,
      "token_acc": 0.5136186770428015,
      "train_speed(iter/s)": 1.438621
    },
    {
      "epoch": 4.35671136626537,
      "grad_norm": 7.164308071136475,
      "learning_rate": 4.0289407858597974e-06,
      "loss": 2.4267263412475586,
      "memory(GiB)": 77.56,
      "step": 101690,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438621
    },
    {
      "epoch": 4.356925581594619,
      "grad_norm": 7.085773468017578,
      "learning_rate": 4.026294555395332e-06,
      "loss": 2.3393688201904297,
      "memory(GiB)": 77.56,
      "step": 101695,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.43862
    },
    {
      "epoch": 4.357139796923867,
      "grad_norm": 7.105220317840576,
      "learning_rate": 4.023649157787673e-06,
      "loss": 2.223957061767578,
      "memory(GiB)": 77.56,
      "step": 101700,
      "token_acc": 0.5267605633802817,
      "train_speed(iter/s)": 1.438626
    },
    {
      "epoch": 4.357354012253117,
      "grad_norm": 5.647937297821045,
      "learning_rate": 4.021004593084726e-06,
      "loss": 2.199010467529297,
      "memory(GiB)": 77.56,
      "step": 101705,
      "token_acc": 0.5323741007194245,
      "train_speed(iter/s)": 1.438619
    },
    {
      "epoch": 4.357568227582366,
      "grad_norm": 6.133718490600586,
      "learning_rate": 4.0183608613344244e-06,
      "loss": 2.3756628036499023,
      "memory(GiB)": 77.56,
      "step": 101710,
      "token_acc": 0.4897260273972603,
      "train_speed(iter/s)": 1.438602
    },
    {
      "epoch": 4.357782442911615,
      "grad_norm": 6.701653480529785,
      "learning_rate": 4.015717962584648e-06,
      "loss": 2.4356565475463867,
      "memory(GiB)": 77.56,
      "step": 101715,
      "token_acc": 0.4923547400611621,
      "train_speed(iter/s)": 1.438594
    },
    {
      "epoch": 4.357996658240864,
      "grad_norm": 6.717548847198486,
      "learning_rate": 4.013075896883278e-06,
      "loss": 2.23958740234375,
      "memory(GiB)": 77.56,
      "step": 101720,
      "token_acc": 0.4885245901639344,
      "train_speed(iter/s)": 1.4386
    },
    {
      "epoch": 4.358210873570113,
      "grad_norm": 4.824134826660156,
      "learning_rate": 4.0104346642781785e-06,
      "loss": 2.1966087341308596,
      "memory(GiB)": 77.56,
      "step": 101725,
      "token_acc": 0.5040983606557377,
      "train_speed(iter/s)": 1.438594
    },
    {
      "epoch": 4.358425088899361,
      "grad_norm": 6.535068035125732,
      "learning_rate": 4.007794264817194e-06,
      "loss": 2.3015857696533204,
      "memory(GiB)": 77.56,
      "step": 101730,
      "token_acc": 0.47477744807121663,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 4.3586393042286105,
      "grad_norm": 6.709981441497803,
      "learning_rate": 4.005154698548152e-06,
      "loss": 2.2980566024780273,
      "memory(GiB)": 77.56,
      "step": 101735,
      "token_acc": 0.5331412103746398,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 4.35885351955786,
      "grad_norm": 5.353514671325684,
      "learning_rate": 4.002515965518883e-06,
      "loss": 2.2750146865844725,
      "memory(GiB)": 77.56,
      "step": 101740,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.438606
    },
    {
      "epoch": 4.359067734887109,
      "grad_norm": 8.38457202911377,
      "learning_rate": 3.999878065777191e-06,
      "loss": 2.6203729629516603,
      "memory(GiB)": 77.56,
      "step": 101745,
      "token_acc": 0.44694533762057875,
      "train_speed(iter/s)": 1.438619
    },
    {
      "epoch": 4.359281950216357,
      "grad_norm": 6.5823493003845215,
      "learning_rate": 3.9972409993708605e-06,
      "loss": 2.3757259368896486,
      "memory(GiB)": 77.56,
      "step": 101750,
      "token_acc": 0.5067114093959731,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 4.359496165545607,
      "grad_norm": 11.166370391845703,
      "learning_rate": 3.994604766347665e-06,
      "loss": 2.050105667114258,
      "memory(GiB)": 77.56,
      "step": 101755,
      "token_acc": 0.5276752767527675,
      "train_speed(iter/s)": 1.438627
    },
    {
      "epoch": 4.359710380874855,
      "grad_norm": 4.874205112457275,
      "learning_rate": 3.99196936675536e-06,
      "loss": 2.4722108840942383,
      "memory(GiB)": 77.56,
      "step": 101760,
      "token_acc": 0.4896755162241888,
      "train_speed(iter/s)": 1.438639
    },
    {
      "epoch": 4.359924596204104,
      "grad_norm": 5.018775939941406,
      "learning_rate": 3.989334800641681e-06,
      "loss": 1.9645587921142578,
      "memory(GiB)": 77.56,
      "step": 101765,
      "token_acc": 0.5522875816993464,
      "train_speed(iter/s)": 1.438644
    },
    {
      "epoch": 4.3601388115333535,
      "grad_norm": 4.494512557983398,
      "learning_rate": 3.986701068054371e-06,
      "loss": 2.1807619094848634,
      "memory(GiB)": 77.56,
      "step": 101770,
      "token_acc": 0.5564516129032258,
      "train_speed(iter/s)": 1.438631
    },
    {
      "epoch": 4.360353026862603,
      "grad_norm": 6.0012359619140625,
      "learning_rate": 3.9840681690411384e-06,
      "loss": 2.3009567260742188,
      "memory(GiB)": 77.56,
      "step": 101775,
      "token_acc": 0.51440329218107,
      "train_speed(iter/s)": 1.43862
    },
    {
      "epoch": 4.360567242191851,
      "grad_norm": 6.165236473083496,
      "learning_rate": 3.981436103649672e-06,
      "loss": 2.0071544647216797,
      "memory(GiB)": 77.56,
      "step": 101780,
      "token_acc": 0.5934959349593496,
      "train_speed(iter/s)": 1.438632
    },
    {
      "epoch": 4.3607814575211,
      "grad_norm": 5.125621318817139,
      "learning_rate": 3.978804871927661e-06,
      "loss": 2.307464599609375,
      "memory(GiB)": 77.56,
      "step": 101785,
      "token_acc": 0.5119453924914675,
      "train_speed(iter/s)": 1.438634
    },
    {
      "epoch": 4.360995672850349,
      "grad_norm": 6.096005916595459,
      "learning_rate": 3.976174473922772e-06,
      "loss": 2.285843276977539,
      "memory(GiB)": 77.56,
      "step": 101790,
      "token_acc": 0.5118110236220472,
      "train_speed(iter/s)": 1.438644
    },
    {
      "epoch": 4.361209888179598,
      "grad_norm": 5.178644180297852,
      "learning_rate": 3.973544909682647e-06,
      "loss": 2.1128358840942383,
      "memory(GiB)": 77.56,
      "step": 101795,
      "token_acc": 0.5150375939849624,
      "train_speed(iter/s)": 1.438645
    },
    {
      "epoch": 4.361424103508847,
      "grad_norm": 6.157804489135742,
      "learning_rate": 3.970916179254941e-06,
      "loss": 2.3324485778808595,
      "memory(GiB)": 77.56,
      "step": 101800,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438658
    },
    {
      "epoch": 4.3616383188380965,
      "grad_norm": 6.665396213531494,
      "learning_rate": 3.968288282687272e-06,
      "loss": 2.3250059127807616,
      "memory(GiB)": 77.56,
      "step": 101805,
      "token_acc": 0.5232198142414861,
      "train_speed(iter/s)": 1.438664
    },
    {
      "epoch": 4.361852534167345,
      "grad_norm": 6.363424301147461,
      "learning_rate": 3.965661220027233e-06,
      "loss": 2.289792060852051,
      "memory(GiB)": 77.56,
      "step": 101810,
      "token_acc": 0.5324232081911263,
      "train_speed(iter/s)": 1.438675
    },
    {
      "epoch": 4.362066749496594,
      "grad_norm": 6.050394058227539,
      "learning_rate": 3.963034991322423e-06,
      "loss": 2.2415313720703125,
      "memory(GiB)": 77.56,
      "step": 101815,
      "token_acc": 0.5114503816793893,
      "train_speed(iter/s)": 1.438702
    },
    {
      "epoch": 4.3622809648258425,
      "grad_norm": 7.389623641967773,
      "learning_rate": 3.9604095966204295e-06,
      "loss": 2.287662887573242,
      "memory(GiB)": 77.56,
      "step": 101820,
      "token_acc": 0.5364238410596026,
      "train_speed(iter/s)": 1.438705
    },
    {
      "epoch": 4.362495180155092,
      "grad_norm": 6.708871841430664,
      "learning_rate": 3.957785035968792e-06,
      "loss": 2.300059509277344,
      "memory(GiB)": 77.56,
      "step": 101825,
      "token_acc": 0.5661764705882353,
      "train_speed(iter/s)": 1.438701
    },
    {
      "epoch": 4.362709395484341,
      "grad_norm": 6.89017915725708,
      "learning_rate": 3.955161309415078e-06,
      "loss": 2.5353761672973634,
      "memory(GiB)": 77.56,
      "step": 101830,
      "token_acc": 0.4414715719063545,
      "train_speed(iter/s)": 1.438716
    },
    {
      "epoch": 4.36292361081359,
      "grad_norm": 5.644868850708008,
      "learning_rate": 3.9525384170068125e-06,
      "loss": 2.173910140991211,
      "memory(GiB)": 77.56,
      "step": 101835,
      "token_acc": 0.524904214559387,
      "train_speed(iter/s)": 1.438723
    },
    {
      "epoch": 4.363137826142839,
      "grad_norm": 7.357850074768066,
      "learning_rate": 3.949916358791511e-06,
      "loss": 2.234305572509766,
      "memory(GiB)": 77.56,
      "step": 101840,
      "token_acc": 0.5340136054421769,
      "train_speed(iter/s)": 1.438726
    },
    {
      "epoch": 4.363352041472088,
      "grad_norm": 6.908636093139648,
      "learning_rate": 3.94729513481667e-06,
      "loss": 2.748577880859375,
      "memory(GiB)": 77.56,
      "step": 101845,
      "token_acc": 0.4460431654676259,
      "train_speed(iter/s)": 1.438723
    },
    {
      "epoch": 4.363566256801336,
      "grad_norm": 7.517205715179443,
      "learning_rate": 3.944674745129773e-06,
      "loss": 2.2894689559936525,
      "memory(GiB)": 77.56,
      "step": 101850,
      "token_acc": 0.4823943661971831,
      "train_speed(iter/s)": 1.438732
    },
    {
      "epoch": 4.3637804721305855,
      "grad_norm": 5.923609256744385,
      "learning_rate": 3.942055189778305e-06,
      "loss": 2.2906402587890624,
      "memory(GiB)": 77.56,
      "step": 101855,
      "token_acc": 0.5177514792899408,
      "train_speed(iter/s)": 1.438738
    },
    {
      "epoch": 4.363994687459835,
      "grad_norm": 6.85636043548584,
      "learning_rate": 3.93943646880971e-06,
      "loss": 1.9102947235107421,
      "memory(GiB)": 77.56,
      "step": 101860,
      "token_acc": 0.5371900826446281,
      "train_speed(iter/s)": 1.438741
    },
    {
      "epoch": 4.364208902789084,
      "grad_norm": 5.837848663330078,
      "learning_rate": 3.936818582271428e-06,
      "loss": 1.9517528533935546,
      "memory(GiB)": 77.56,
      "step": 101865,
      "token_acc": 0.535031847133758,
      "train_speed(iter/s)": 1.438729
    },
    {
      "epoch": 4.364423118118332,
      "grad_norm": 6.050806522369385,
      "learning_rate": 3.934201530210896e-06,
      "loss": 2.189637565612793,
      "memory(GiB)": 77.56,
      "step": 101870,
      "token_acc": 0.5250836120401338,
      "train_speed(iter/s)": 1.438734
    },
    {
      "epoch": 4.364637333447582,
      "grad_norm": 7.866569519042969,
      "learning_rate": 3.931585312675512e-06,
      "loss": 2.72440242767334,
      "memory(GiB)": 77.56,
      "step": 101875,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.438745
    },
    {
      "epoch": 4.36485154877683,
      "grad_norm": 7.117494583129883,
      "learning_rate": 3.928969929712673e-06,
      "loss": 2.551620292663574,
      "memory(GiB)": 77.56,
      "step": 101880,
      "token_acc": 0.4583333333333333,
      "train_speed(iter/s)": 1.438731
    },
    {
      "epoch": 4.365065764106079,
      "grad_norm": 7.407101154327393,
      "learning_rate": 3.926355381369762e-06,
      "loss": 2.266368865966797,
      "memory(GiB)": 77.56,
      "step": 101885,
      "token_acc": 0.47719298245614034,
      "train_speed(iter/s)": 1.438732
    },
    {
      "epoch": 4.3652799794353285,
      "grad_norm": 8.699575424194336,
      "learning_rate": 3.923741667694142e-06,
      "loss": 2.057978057861328,
      "memory(GiB)": 77.56,
      "step": 101890,
      "token_acc": 0.49586776859504134,
      "train_speed(iter/s)": 1.438738
    },
    {
      "epoch": 4.365494194764578,
      "grad_norm": 5.718353271484375,
      "learning_rate": 3.921128788733169e-06,
      "loss": 2.2229318618774414,
      "memory(GiB)": 77.56,
      "step": 101895,
      "token_acc": 0.5104166666666666,
      "train_speed(iter/s)": 1.438749
    },
    {
      "epoch": 4.365708410093826,
      "grad_norm": 6.512753963470459,
      "learning_rate": 3.918516744534178e-06,
      "loss": 2.528089141845703,
      "memory(GiB)": 77.56,
      "step": 101900,
      "token_acc": 0.4919614147909968,
      "train_speed(iter/s)": 1.438768
    },
    {
      "epoch": 4.365922625423075,
      "grad_norm": 6.803748607635498,
      "learning_rate": 3.915905535144482e-06,
      "loss": 2.234787178039551,
      "memory(GiB)": 77.56,
      "step": 101905,
      "token_acc": 0.5070921985815603,
      "train_speed(iter/s)": 1.438779
    },
    {
      "epoch": 4.366136840752324,
      "grad_norm": 6.593135833740234,
      "learning_rate": 3.9132951606113745e-06,
      "loss": 2.3033109664916993,
      "memory(GiB)": 77.56,
      "step": 101910,
      "token_acc": 0.5231788079470199,
      "train_speed(iter/s)": 1.43879
    },
    {
      "epoch": 4.366351056081573,
      "grad_norm": 8.200020790100098,
      "learning_rate": 3.910685620982174e-06,
      "loss": 2.102082061767578,
      "memory(GiB)": 77.56,
      "step": 101915,
      "token_acc": 0.5261194029850746,
      "train_speed(iter/s)": 1.438796
    },
    {
      "epoch": 4.366565271410822,
      "grad_norm": 6.162394046783447,
      "learning_rate": 3.908076916304132e-06,
      "loss": 2.1550668716430663,
      "memory(GiB)": 77.56,
      "step": 101920,
      "token_acc": 0.5176848874598071,
      "train_speed(iter/s)": 1.4388
    },
    {
      "epoch": 4.3667794867400715,
      "grad_norm": 5.4941511154174805,
      "learning_rate": 3.9054690466245215e-06,
      "loss": 2.4001319885253904,
      "memory(GiB)": 77.56,
      "step": 101925,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.438814
    },
    {
      "epoch": 4.36699370206932,
      "grad_norm": 7.408539295196533,
      "learning_rate": 3.902862011990571e-06,
      "loss": 2.372382926940918,
      "memory(GiB)": 77.56,
      "step": 101930,
      "token_acc": 0.5079872204472844,
      "train_speed(iter/s)": 1.438816
    },
    {
      "epoch": 4.367207917398569,
      "grad_norm": 5.480215072631836,
      "learning_rate": 3.900255812449527e-06,
      "loss": 1.9784221649169922,
      "memory(GiB)": 77.56,
      "step": 101935,
      "token_acc": 0.5536480686695279,
      "train_speed(iter/s)": 1.438819
    },
    {
      "epoch": 4.3674221327278175,
      "grad_norm": 6.481391906738281,
      "learning_rate": 3.897650448048579e-06,
      "loss": 2.184911346435547,
      "memory(GiB)": 77.56,
      "step": 101940,
      "token_acc": 0.5221843003412969,
      "train_speed(iter/s)": 1.438829
    },
    {
      "epoch": 4.367636348057067,
      "grad_norm": 8.044012069702148,
      "learning_rate": 3.89504591883495e-06,
      "loss": 2.596766471862793,
      "memory(GiB)": 77.56,
      "step": 101945,
      "token_acc": 0.4857142857142857,
      "train_speed(iter/s)": 1.438852
    },
    {
      "epoch": 4.367850563386316,
      "grad_norm": 7.4052300453186035,
      "learning_rate": 3.892442224855813e-06,
      "loss": 2.018967056274414,
      "memory(GiB)": 77.56,
      "step": 101950,
      "token_acc": 0.5510204081632653,
      "train_speed(iter/s)": 1.438858
    },
    {
      "epoch": 4.368064778715565,
      "grad_norm": 5.536005973815918,
      "learning_rate": 3.889839366158343e-06,
      "loss": 2.1921310424804688,
      "memory(GiB)": 77.56,
      "step": 101955,
      "token_acc": 0.475,
      "train_speed(iter/s)": 1.438861
    },
    {
      "epoch": 4.368278994044814,
      "grad_norm": 6.33701753616333,
      "learning_rate": 3.8872373427896735e-06,
      "loss": 2.0811885833740233,
      "memory(GiB)": 77.56,
      "step": 101960,
      "token_acc": 0.5306122448979592,
      "train_speed(iter/s)": 1.43887
    },
    {
      "epoch": 4.368493209374063,
      "grad_norm": 7.837864398956299,
      "learning_rate": 3.884636154796967e-06,
      "loss": 2.385394287109375,
      "memory(GiB)": 77.56,
      "step": 101965,
      "token_acc": 0.48484848484848486,
      "train_speed(iter/s)": 1.438865
    },
    {
      "epoch": 4.368707424703311,
      "grad_norm": 5.585497856140137,
      "learning_rate": 3.88203580222733e-06,
      "loss": 2.2544960021972655,
      "memory(GiB)": 77.56,
      "step": 101970,
      "token_acc": 0.5222222222222223,
      "train_speed(iter/s)": 1.43886
    },
    {
      "epoch": 4.368921640032561,
      "grad_norm": 5.798569679260254,
      "learning_rate": 3.879436285127886e-06,
      "loss": 2.688420295715332,
      "memory(GiB)": 77.56,
      "step": 101975,
      "token_acc": 0.43769968051118213,
      "train_speed(iter/s)": 1.438862
    },
    {
      "epoch": 4.36913585536181,
      "grad_norm": 6.69862174987793,
      "learning_rate": 3.876837603545713e-06,
      "loss": 2.441334533691406,
      "memory(GiB)": 77.56,
      "step": 101980,
      "token_acc": 0.48606811145510836,
      "train_speed(iter/s)": 1.438852
    },
    {
      "epoch": 4.369350070691059,
      "grad_norm": 7.233700752258301,
      "learning_rate": 3.874239757527897e-06,
      "loss": 2.4201616287231444,
      "memory(GiB)": 77.56,
      "step": 101985,
      "token_acc": 0.48717948717948717,
      "train_speed(iter/s)": 1.438855
    },
    {
      "epoch": 4.369564286020307,
      "grad_norm": 5.579944610595703,
      "learning_rate": 3.8716427471214955e-06,
      "loss": 2.0034767150878907,
      "memory(GiB)": 77.56,
      "step": 101990,
      "token_acc": 0.5595667870036101,
      "train_speed(iter/s)": 1.438849
    },
    {
      "epoch": 4.369778501349557,
      "grad_norm": 7.078153610229492,
      "learning_rate": 3.869046572373552e-06,
      "loss": 2.261469841003418,
      "memory(GiB)": 77.56,
      "step": 101995,
      "token_acc": 0.4964788732394366,
      "train_speed(iter/s)": 1.438849
    },
    {
      "epoch": 4.369992716678805,
      "grad_norm": 7.603664875030518,
      "learning_rate": 3.866451233331108e-06,
      "loss": 2.156891632080078,
      "memory(GiB)": 77.56,
      "step": 102000,
      "token_acc": 0.5174825174825175,
      "train_speed(iter/s)": 1.438855
    },
    {
      "epoch": 4.369992716678805,
      "eval_loss": 2.340041160583496,
      "eval_runtime": 14.1504,
      "eval_samples_per_second": 7.067,
      "eval_steps_per_second": 7.067,
      "eval_token_acc": 0.4856020942408377,
      "step": 102000
    },
    {
      "epoch": 4.370206932008054,
      "grad_norm": 5.848587512969971,
      "learning_rate": 3.863856730041182e-06,
      "loss": 2.3252744674682617,
      "memory(GiB)": 77.56,
      "step": 102005,
      "token_acc": 0.4905123339658444,
      "train_speed(iter/s)": 1.438554
    },
    {
      "epoch": 4.370421147337304,
      "grad_norm": 5.226724147796631,
      "learning_rate": 3.861263062550769e-06,
      "loss": 2.389395523071289,
      "memory(GiB)": 77.56,
      "step": 102010,
      "token_acc": 0.4746666666666667,
      "train_speed(iter/s)": 1.438556
    },
    {
      "epoch": 4.370635362666553,
      "grad_norm": 8.791924476623535,
      "learning_rate": 3.858670230906852e-06,
      "loss": 2.389272117614746,
      "memory(GiB)": 77.56,
      "step": 102015,
      "token_acc": 0.5029585798816568,
      "train_speed(iter/s)": 1.438563
    },
    {
      "epoch": 4.370849577995801,
      "grad_norm": 5.486352443695068,
      "learning_rate": 3.856078235156413e-06,
      "loss": 2.3030513763427733,
      "memory(GiB)": 77.56,
      "step": 102020,
      "token_acc": 0.4563758389261745,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 4.3710637933250505,
      "grad_norm": 5.798378944396973,
      "learning_rate": 3.853487075346396e-06,
      "loss": 2.3755891799926756,
      "memory(GiB)": 77.56,
      "step": 102025,
      "token_acc": 0.5255474452554745,
      "train_speed(iter/s)": 1.438557
    },
    {
      "epoch": 4.371278008654299,
      "grad_norm": 6.607595443725586,
      "learning_rate": 3.850896751523758e-06,
      "loss": 2.451091766357422,
      "memory(GiB)": 77.56,
      "step": 102030,
      "token_acc": 0.4888888888888889,
      "train_speed(iter/s)": 1.438574
    },
    {
      "epoch": 4.371492223983548,
      "grad_norm": 5.402934551239014,
      "learning_rate": 3.8483072637354065e-06,
      "loss": 2.4583719253540037,
      "memory(GiB)": 77.56,
      "step": 102035,
      "token_acc": 0.48427672955974843,
      "train_speed(iter/s)": 1.438562
    },
    {
      "epoch": 4.371706439312797,
      "grad_norm": 5.248274803161621,
      "learning_rate": 3.845718612028271e-06,
      "loss": 2.4814741134643556,
      "memory(GiB)": 77.56,
      "step": 102040,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438561
    },
    {
      "epoch": 4.371920654642047,
      "grad_norm": 6.561869144439697,
      "learning_rate": 3.843130796449235e-06,
      "loss": 2.2307239532470704,
      "memory(GiB)": 77.56,
      "step": 102045,
      "token_acc": 0.49390243902439024,
      "train_speed(iter/s)": 1.438575
    },
    {
      "epoch": 4.372134869971295,
      "grad_norm": 5.527139663696289,
      "learning_rate": 3.840543817045189e-06,
      "loss": 2.2846038818359373,
      "memory(GiB)": 77.56,
      "step": 102050,
      "token_acc": 0.5017543859649123,
      "train_speed(iter/s)": 1.43857
    },
    {
      "epoch": 4.372349085300544,
      "grad_norm": 5.8063812255859375,
      "learning_rate": 3.8379576738629856e-06,
      "loss": 2.3370399475097656,
      "memory(GiB)": 77.56,
      "step": 102055,
      "token_acc": 0.5147540983606558,
      "train_speed(iter/s)": 1.438553
    },
    {
      "epoch": 4.3725633006297935,
      "grad_norm": 6.326207637786865,
      "learning_rate": 3.8353723669494855e-06,
      "loss": 2.1681535720825194,
      "memory(GiB)": 77.56,
      "step": 102060,
      "token_acc": 0.5203761755485894,
      "train_speed(iter/s)": 1.438543
    },
    {
      "epoch": 4.372777515959042,
      "grad_norm": 6.0897417068481445,
      "learning_rate": 3.8327878963515256e-06,
      "loss": 2.502007484436035,
      "memory(GiB)": 77.56,
      "step": 102065,
      "token_acc": 0.48333333333333334,
      "train_speed(iter/s)": 1.438546
    },
    {
      "epoch": 4.372991731288291,
      "grad_norm": 7.496880054473877,
      "learning_rate": 3.830204262115916e-06,
      "loss": 2.9834030151367186,
      "memory(GiB)": 77.56,
      "step": 102070,
      "token_acc": 0.4358974358974359,
      "train_speed(iter/s)": 1.438553
    },
    {
      "epoch": 4.37320594661754,
      "grad_norm": 5.736611843109131,
      "learning_rate": 3.827621464289471e-06,
      "loss": 2.283222961425781,
      "memory(GiB)": 77.56,
      "step": 102075,
      "token_acc": 0.5428571428571428,
      "train_speed(iter/s)": 1.438566
    },
    {
      "epoch": 4.373420161946789,
      "grad_norm": 7.284671306610107,
      "learning_rate": 3.825039502918965e-06,
      "loss": 2.4526615142822266,
      "memory(GiB)": 77.56,
      "step": 102080,
      "token_acc": 0.5220125786163522,
      "train_speed(iter/s)": 1.438559
    },
    {
      "epoch": 4.373634377276038,
      "grad_norm": 5.200911045074463,
      "learning_rate": 3.822458378051197e-06,
      "loss": 2.2039566040039062,
      "memory(GiB)": 77.56,
      "step": 102085,
      "token_acc": 0.5105633802816901,
      "train_speed(iter/s)": 1.438561
    },
    {
      "epoch": 4.373848592605287,
      "grad_norm": 5.566181659698486,
      "learning_rate": 3.819878089732909e-06,
      "loss": 2.448895072937012,
      "memory(GiB)": 77.56,
      "step": 102090,
      "token_acc": 0.5016181229773463,
      "train_speed(iter/s)": 1.438564
    },
    {
      "epoch": 4.374062807934536,
      "grad_norm": 6.476151466369629,
      "learning_rate": 3.817298638010852e-06,
      "loss": 2.356591796875,
      "memory(GiB)": 77.56,
      "step": 102095,
      "token_acc": 0.4879032258064516,
      "train_speed(iter/s)": 1.438573
    },
    {
      "epoch": 4.374277023263785,
      "grad_norm": 5.828040599822998,
      "learning_rate": 3.814720022931756e-06,
      "loss": 2.1494466781616213,
      "memory(GiB)": 77.56,
      "step": 102100,
      "token_acc": 0.524822695035461,
      "train_speed(iter/s)": 1.438574
    },
    {
      "epoch": 4.374491238593034,
      "grad_norm": 6.093513488769531,
      "learning_rate": 3.812142244542327e-06,
      "loss": 2.407297134399414,
      "memory(GiB)": 77.56,
      "step": 102105,
      "token_acc": 0.48109965635738833,
      "train_speed(iter/s)": 1.438578
    },
    {
      "epoch": 4.3747054539222825,
      "grad_norm": 7.090908050537109,
      "learning_rate": 3.809565302889262e-06,
      "loss": 2.3766469955444336,
      "memory(GiB)": 77.56,
      "step": 102110,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438595
    },
    {
      "epoch": 4.374919669251532,
      "grad_norm": 6.150954723358154,
      "learning_rate": 3.8069891980192507e-06,
      "loss": 2.2138465881347655,
      "memory(GiB)": 77.56,
      "step": 102115,
      "token_acc": 0.5566343042071198,
      "train_speed(iter/s)": 1.438603
    },
    {
      "epoch": 4.375133884580781,
      "grad_norm": 5.788748264312744,
      "learning_rate": 3.8044139299789727e-06,
      "loss": 2.3531898498535155,
      "memory(GiB)": 77.56,
      "step": 102120,
      "token_acc": 0.4879518072289157,
      "train_speed(iter/s)": 1.438619
    },
    {
      "epoch": 4.375348099910029,
      "grad_norm": 7.598631381988525,
      "learning_rate": 3.801839498815074e-06,
      "loss": 2.2558507919311523,
      "memory(GiB)": 77.56,
      "step": 102125,
      "token_acc": 0.5335820895522388,
      "train_speed(iter/s)": 1.438629
    },
    {
      "epoch": 4.375562315239279,
      "grad_norm": 5.6159467697143555,
      "learning_rate": 3.799265904574184e-06,
      "loss": 2.3174352645874023,
      "memory(GiB)": 77.56,
      "step": 102130,
      "token_acc": 0.521594684385382,
      "train_speed(iter/s)": 1.43865
    },
    {
      "epoch": 4.375776530568528,
      "grad_norm": 6.232875823974609,
      "learning_rate": 3.7966931473029378e-06,
      "loss": 2.404511642456055,
      "memory(GiB)": 77.56,
      "step": 102135,
      "token_acc": 0.5055350553505535,
      "train_speed(iter/s)": 1.43866
    },
    {
      "epoch": 4.375990745897776,
      "grad_norm": 7.546611309051514,
      "learning_rate": 3.7941212270479254e-06,
      "loss": 2.143113708496094,
      "memory(GiB)": 77.56,
      "step": 102140,
      "token_acc": 0.5462184873949579,
      "train_speed(iter/s)": 1.438673
    },
    {
      "epoch": 4.3762049612270255,
      "grad_norm": 7.115991592407227,
      "learning_rate": 3.79155014385576e-06,
      "loss": 2.464577293395996,
      "memory(GiB)": 77.56,
      "step": 102145,
      "token_acc": 0.4666666666666667,
      "train_speed(iter/s)": 1.438675
    },
    {
      "epoch": 4.376419176556275,
      "grad_norm": 7.377319812774658,
      "learning_rate": 3.7889798977730153e-06,
      "loss": 2.2669668197631836,
      "memory(GiB)": 77.56,
      "step": 102150,
      "token_acc": 0.5251798561151079,
      "train_speed(iter/s)": 1.438667
    },
    {
      "epoch": 4.376633391885523,
      "grad_norm": 5.948807716369629,
      "learning_rate": 3.7864104888462425e-06,
      "loss": 2.1526660919189453,
      "memory(GiB)": 77.56,
      "step": 102155,
      "token_acc": 0.5144927536231884,
      "train_speed(iter/s)": 1.438643
    },
    {
      "epoch": 4.376847607214772,
      "grad_norm": 5.095167636871338,
      "learning_rate": 3.783841917122e-06,
      "loss": 2.2469051361083983,
      "memory(GiB)": 77.56,
      "step": 102160,
      "token_acc": 0.554858934169279,
      "train_speed(iter/s)": 1.438652
    },
    {
      "epoch": 4.377061822544022,
      "grad_norm": 5.583045959472656,
      "learning_rate": 3.781274182646816e-06,
      "loss": 2.276612663269043,
      "memory(GiB)": 77.56,
      "step": 102165,
      "token_acc": 0.532520325203252,
      "train_speed(iter/s)": 1.438657
    },
    {
      "epoch": 4.37727603787327,
      "grad_norm": 5.439349174499512,
      "learning_rate": 3.778707285467198e-06,
      "loss": 2.2353775024414064,
      "memory(GiB)": 77.56,
      "step": 102170,
      "token_acc": 0.5279503105590062,
      "train_speed(iter/s)": 1.438663
    },
    {
      "epoch": 4.377490253202519,
      "grad_norm": 5.987916469573975,
      "learning_rate": 3.776141225629659e-06,
      "loss": 2.150084686279297,
      "memory(GiB)": 77.56,
      "step": 102175,
      "token_acc": 0.5375,
      "train_speed(iter/s)": 1.438653
    },
    {
      "epoch": 4.3777044685317685,
      "grad_norm": 6.209805488586426,
      "learning_rate": 3.77357600318069e-06,
      "loss": 2.354393196105957,
      "memory(GiB)": 77.56,
      "step": 102180,
      "token_acc": 0.484375,
      "train_speed(iter/s)": 1.438647
    },
    {
      "epoch": 4.377918683861017,
      "grad_norm": 7.4460649490356445,
      "learning_rate": 3.7710116181667422e-06,
      "loss": 2.400659942626953,
      "memory(GiB)": 77.56,
      "step": 102185,
      "token_acc": 0.47770700636942676,
      "train_speed(iter/s)": 1.438632
    },
    {
      "epoch": 4.378132899190266,
      "grad_norm": 9.189557075500488,
      "learning_rate": 3.7684480706343005e-06,
      "loss": 2.434188461303711,
      "memory(GiB)": 77.56,
      "step": 102190,
      "token_acc": 0.4775641025641026,
      "train_speed(iter/s)": 1.438635
    },
    {
      "epoch": 4.378347114519515,
      "grad_norm": 5.722678184509277,
      "learning_rate": 3.765885360629784e-06,
      "loss": 2.1983808517456054,
      "memory(GiB)": 77.56,
      "step": 102195,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.438633
    },
    {
      "epoch": 4.378561329848764,
      "grad_norm": 6.487076759338379,
      "learning_rate": 3.7633234881996216e-06,
      "loss": 2.265846061706543,
      "memory(GiB)": 77.56,
      "step": 102200,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.438646
    },
    {
      "epoch": 4.378775545178013,
      "grad_norm": 6.377604007720947,
      "learning_rate": 3.760762453390232e-06,
      "loss": 2.0761249542236326,
      "memory(GiB)": 77.56,
      "step": 102205,
      "token_acc": 0.5091575091575091,
      "train_speed(iter/s)": 1.43866
    },
    {
      "epoch": 4.378989760507262,
      "grad_norm": 6.834473133087158,
      "learning_rate": 3.7582022562480112e-06,
      "loss": 1.9617494583129882,
      "memory(GiB)": 77.56,
      "step": 102210,
      "token_acc": 0.5702479338842975,
      "train_speed(iter/s)": 1.438667
    },
    {
      "epoch": 4.379203975836511,
      "grad_norm": 6.393004417419434,
      "learning_rate": 3.7556428968193335e-06,
      "loss": 1.9575017929077148,
      "memory(GiB)": 77.56,
      "step": 102215,
      "token_acc": 0.5806451612903226,
      "train_speed(iter/s)": 1.438675
    },
    {
      "epoch": 4.37941819116576,
      "grad_norm": 7.924510478973389,
      "learning_rate": 3.7530843751505617e-06,
      "loss": 2.2697858810424805,
      "memory(GiB)": 77.56,
      "step": 102220,
      "token_acc": 0.5330578512396694,
      "train_speed(iter/s)": 1.43868
    },
    {
      "epoch": 4.379632406495009,
      "grad_norm": 6.417081356048584,
      "learning_rate": 3.7505266912880423e-06,
      "loss": 1.9659889221191407,
      "memory(GiB)": 77.56,
      "step": 102225,
      "token_acc": 0.5542168674698795,
      "train_speed(iter/s)": 1.438684
    },
    {
      "epoch": 4.3798466218242575,
      "grad_norm": 5.71816349029541,
      "learning_rate": 3.7479698452781264e-06,
      "loss": 2.333475112915039,
      "memory(GiB)": 77.56,
      "step": 102230,
      "token_acc": 0.5338078291814946,
      "train_speed(iter/s)": 1.438668
    },
    {
      "epoch": 4.380060837153507,
      "grad_norm": 5.253359317779541,
      "learning_rate": 3.7454138371671275e-06,
      "loss": 2.054024314880371,
      "memory(GiB)": 77.56,
      "step": 102235,
      "token_acc": 0.5498392282958199,
      "train_speed(iter/s)": 1.438679
    },
    {
      "epoch": 4.380275052482756,
      "grad_norm": 5.206414699554443,
      "learning_rate": 3.742858667001342e-06,
      "loss": 2.145540237426758,
      "memory(GiB)": 77.56,
      "step": 102240,
      "token_acc": 0.5130718954248366,
      "train_speed(iter/s)": 1.438676
    },
    {
      "epoch": 4.380489267812004,
      "grad_norm": 7.803227424621582,
      "learning_rate": 3.7403043348270716e-06,
      "loss": 2.354961395263672,
      "memory(GiB)": 77.56,
      "step": 102245,
      "token_acc": 0.4647887323943662,
      "train_speed(iter/s)": 1.438677
    },
    {
      "epoch": 4.380703483141254,
      "grad_norm": 6.29659366607666,
      "learning_rate": 3.737750840690579e-06,
      "loss": 2.321546173095703,
      "memory(GiB)": 77.56,
      "step": 102250,
      "token_acc": 0.5518672199170125,
      "train_speed(iter/s)": 1.43869
    },
    {
      "epoch": 4.380917698470503,
      "grad_norm": 6.761232852935791,
      "learning_rate": 3.735198184638117e-06,
      "loss": 2.027303123474121,
      "memory(GiB)": 77.56,
      "step": 102255,
      "token_acc": 0.5261324041811847,
      "train_speed(iter/s)": 1.438702
    },
    {
      "epoch": 4.381131913799751,
      "grad_norm": 6.1805524826049805,
      "learning_rate": 3.732646366715942e-06,
      "loss": 2.292635917663574,
      "memory(GiB)": 77.56,
      "step": 102260,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.438691
    },
    {
      "epoch": 4.381346129129001,
      "grad_norm": 7.5055975914001465,
      "learning_rate": 3.7300953869702894e-06,
      "loss": 1.9432130813598634,
      "memory(GiB)": 77.56,
      "step": 102265,
      "token_acc": 0.572992700729927,
      "train_speed(iter/s)": 1.438707
    },
    {
      "epoch": 4.38156034445825,
      "grad_norm": 5.628097057342529,
      "learning_rate": 3.7275452454473613e-06,
      "loss": 2.564580535888672,
      "memory(GiB)": 77.56,
      "step": 102270,
      "token_acc": 0.4925373134328358,
      "train_speed(iter/s)": 1.438709
    },
    {
      "epoch": 4.381774559787498,
      "grad_norm": 6.682338237762451,
      "learning_rate": 3.72499594219336e-06,
      "loss": 2.2786733627319338,
      "memory(GiB)": 77.56,
      "step": 102275,
      "token_acc": 0.5155709342560554,
      "train_speed(iter/s)": 1.438721
    },
    {
      "epoch": 4.3819887751167474,
      "grad_norm": 6.442184925079346,
      "learning_rate": 3.7224474772544706e-06,
      "loss": 2.3629243850708006,
      "memory(GiB)": 77.56,
      "step": 102280,
      "token_acc": 0.5136054421768708,
      "train_speed(iter/s)": 1.438701
    },
    {
      "epoch": 4.382202990445997,
      "grad_norm": 5.647762298583984,
      "learning_rate": 3.7198998506768444e-06,
      "loss": 2.2429290771484376,
      "memory(GiB)": 77.56,
      "step": 102285,
      "token_acc": 0.5249169435215947,
      "train_speed(iter/s)": 1.438715
    },
    {
      "epoch": 4.382417205775245,
      "grad_norm": 5.52142858505249,
      "learning_rate": 3.717353062506651e-06,
      "loss": 2.2926504135131838,
      "memory(GiB)": 77.56,
      "step": 102290,
      "token_acc": 0.4810606060606061,
      "train_speed(iter/s)": 1.438706
    },
    {
      "epoch": 4.382631421104494,
      "grad_norm": 6.8619537353515625,
      "learning_rate": 3.714807112790031e-06,
      "loss": 2.184850311279297,
      "memory(GiB)": 77.56,
      "step": 102295,
      "token_acc": 0.5617529880478087,
      "train_speed(iter/s)": 1.438721
    },
    {
      "epoch": 4.382845636433744,
      "grad_norm": 5.306491851806641,
      "learning_rate": 3.7122620015730967e-06,
      "loss": 2.162005615234375,
      "memory(GiB)": 77.56,
      "step": 102300,
      "token_acc": 0.525974025974026,
      "train_speed(iter/s)": 1.438727
    },
    {
      "epoch": 4.383059851762992,
      "grad_norm": 5.403891086578369,
      "learning_rate": 3.709717728901957e-06,
      "loss": 1.9421588897705078,
      "memory(GiB)": 77.56,
      "step": 102305,
      "token_acc": 0.553030303030303,
      "train_speed(iter/s)": 1.438741
    },
    {
      "epoch": 4.383274067092241,
      "grad_norm": 6.866203308105469,
      "learning_rate": 3.707174294822702e-06,
      "loss": 2.7477653503417967,
      "memory(GiB)": 77.56,
      "step": 102310,
      "token_acc": 0.46325878594249204,
      "train_speed(iter/s)": 1.438757
    },
    {
      "epoch": 4.3834882824214905,
      "grad_norm": 6.528674125671387,
      "learning_rate": 3.7046316993814058e-06,
      "loss": 2.1985727310180665,
      "memory(GiB)": 77.56,
      "step": 102315,
      "token_acc": 0.5447470817120622,
      "train_speed(iter/s)": 1.438736
    },
    {
      "epoch": 4.383702497750739,
      "grad_norm": 5.5168867111206055,
      "learning_rate": 3.702089942624143e-06,
      "loss": 2.0589027404785156,
      "memory(GiB)": 77.56,
      "step": 102320,
      "token_acc": 0.5611285266457681,
      "train_speed(iter/s)": 1.438728
    },
    {
      "epoch": 4.383916713079988,
      "grad_norm": 5.517563343048096,
      "learning_rate": 3.699549024596949e-06,
      "loss": 2.136216735839844,
      "memory(GiB)": 77.56,
      "step": 102325,
      "token_acc": 0.5387205387205387,
      "train_speed(iter/s)": 1.438728
    },
    {
      "epoch": 4.384130928409237,
      "grad_norm": 5.827166557312012,
      "learning_rate": 3.6970089453458534e-06,
      "loss": 2.275006103515625,
      "memory(GiB)": 77.56,
      "step": 102330,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438743
    },
    {
      "epoch": 4.384345143738486,
      "grad_norm": 5.796579360961914,
      "learning_rate": 3.6944697049168807e-06,
      "loss": 2.2470897674560546,
      "memory(GiB)": 77.56,
      "step": 102335,
      "token_acc": 0.5337837837837838,
      "train_speed(iter/s)": 1.438738
    },
    {
      "epoch": 4.384559359067735,
      "grad_norm": 7.104578018188477,
      "learning_rate": 3.6919313033560278e-06,
      "loss": 2.325518035888672,
      "memory(GiB)": 77.56,
      "step": 102340,
      "token_acc": 0.5147540983606558,
      "train_speed(iter/s)": 1.438753
    },
    {
      "epoch": 4.384773574396984,
      "grad_norm": 5.3219075202941895,
      "learning_rate": 3.689393740709274e-06,
      "loss": 2.276840591430664,
      "memory(GiB)": 77.56,
      "step": 102345,
      "token_acc": 0.5555555555555556,
      "train_speed(iter/s)": 1.438754
    },
    {
      "epoch": 4.384987789726233,
      "grad_norm": 7.892182350158691,
      "learning_rate": 3.686857017022605e-06,
      "loss": 2.577449417114258,
      "memory(GiB)": 77.56,
      "step": 102350,
      "token_acc": 0.47368421052631576,
      "train_speed(iter/s)": 1.438761
    },
    {
      "epoch": 4.385202005055482,
      "grad_norm": 7.663252830505371,
      "learning_rate": 3.684321132341961e-06,
      "loss": 2.270546722412109,
      "memory(GiB)": 77.56,
      "step": 102355,
      "token_acc": 0.5243445692883895,
      "train_speed(iter/s)": 1.438747
    },
    {
      "epoch": 4.385416220384731,
      "grad_norm": 6.864215850830078,
      "learning_rate": 3.681786086713296e-06,
      "loss": 2.064653205871582,
      "memory(GiB)": 77.56,
      "step": 102360,
      "token_acc": 0.5594855305466238,
      "train_speed(iter/s)": 1.438756
    },
    {
      "epoch": 4.3856304357139795,
      "grad_norm": 5.660896301269531,
      "learning_rate": 3.679251880182516e-06,
      "loss": 2.261896324157715,
      "memory(GiB)": 77.56,
      "step": 102365,
      "token_acc": 0.49570200573065903,
      "train_speed(iter/s)": 1.438753
    },
    {
      "epoch": 4.385844651043229,
      "grad_norm": 6.448981761932373,
      "learning_rate": 3.6767185127955407e-06,
      "loss": 2.3489830017089846,
      "memory(GiB)": 77.56,
      "step": 102370,
      "token_acc": 0.5365079365079365,
      "train_speed(iter/s)": 1.438765
    },
    {
      "epoch": 4.386058866372478,
      "grad_norm": 5.482466220855713,
      "learning_rate": 3.674185984598266e-06,
      "loss": 2.160551071166992,
      "memory(GiB)": 77.56,
      "step": 102375,
      "token_acc": 0.5189003436426117,
      "train_speed(iter/s)": 1.438774
    },
    {
      "epoch": 4.386273081701726,
      "grad_norm": 6.697409629821777,
      "learning_rate": 3.671654295636573e-06,
      "loss": 2.5021800994873047,
      "memory(GiB)": 77.56,
      "step": 102380,
      "token_acc": 0.49310344827586206,
      "train_speed(iter/s)": 1.438769
    },
    {
      "epoch": 4.386487297030976,
      "grad_norm": 6.35969877243042,
      "learning_rate": 3.669123445956324e-06,
      "loss": 2.221974182128906,
      "memory(GiB)": 77.56,
      "step": 102385,
      "token_acc": 0.5191082802547771,
      "train_speed(iter/s)": 1.438766
    },
    {
      "epoch": 4.386701512360225,
      "grad_norm": 7.7098612785339355,
      "learning_rate": 3.6665934356033604e-06,
      "loss": 2.6257568359375,
      "memory(GiB)": 77.56,
      "step": 102390,
      "token_acc": 0.4819672131147541,
      "train_speed(iter/s)": 1.438783
    },
    {
      "epoch": 4.386915727689473,
      "grad_norm": 6.42761754989624,
      "learning_rate": 3.6640642646235236e-06,
      "loss": 2.2071720123291017,
      "memory(GiB)": 77.56,
      "step": 102395,
      "token_acc": 0.5340501792114696,
      "train_speed(iter/s)": 1.438795
    },
    {
      "epoch": 4.3871299430187225,
      "grad_norm": 5.812616348266602,
      "learning_rate": 3.661535933062621e-06,
      "loss": 2.221953010559082,
      "memory(GiB)": 77.56,
      "step": 102400,
      "token_acc": 0.4889705882352941,
      "train_speed(iter/s)": 1.438794
    },
    {
      "epoch": 4.387344158347972,
      "grad_norm": 7.074045181274414,
      "learning_rate": 3.6590084409664605e-06,
      "loss": 2.0088850021362306,
      "memory(GiB)": 77.56,
      "step": 102405,
      "token_acc": 0.532608695652174,
      "train_speed(iter/s)": 1.438804
    },
    {
      "epoch": 4.38755837367722,
      "grad_norm": 5.440163612365723,
      "learning_rate": 3.65648178838085e-06,
      "loss": 2.287728691101074,
      "memory(GiB)": 77.56,
      "step": 102410,
      "token_acc": 0.5341246290801187,
      "train_speed(iter/s)": 1.438801
    },
    {
      "epoch": 4.387772589006469,
      "grad_norm": 5.775484085083008,
      "learning_rate": 3.653955975351536e-06,
      "loss": 2.1079334259033202,
      "memory(GiB)": 77.56,
      "step": 102415,
      "token_acc": 0.5373134328358209,
      "train_speed(iter/s)": 1.438803
    },
    {
      "epoch": 4.387986804335719,
      "grad_norm": 5.511375904083252,
      "learning_rate": 3.651431001924294e-06,
      "loss": 2.3807945251464844,
      "memory(GiB)": 77.56,
      "step": 102420,
      "token_acc": 0.5325670498084292,
      "train_speed(iter/s)": 1.43882
    },
    {
      "epoch": 4.388201019664967,
      "grad_norm": 6.314801216125488,
      "learning_rate": 3.6489068681448522e-06,
      "loss": 2.421588134765625,
      "memory(GiB)": 77.56,
      "step": 102425,
      "token_acc": 0.4770992366412214,
      "train_speed(iter/s)": 1.438836
    },
    {
      "epoch": 4.388415234994216,
      "grad_norm": 5.989437103271484,
      "learning_rate": 3.646383574058937e-06,
      "loss": 2.293304443359375,
      "memory(GiB)": 77.56,
      "step": 102430,
      "token_acc": 0.49829351535836175,
      "train_speed(iter/s)": 1.43883
    },
    {
      "epoch": 4.3886294503234655,
      "grad_norm": 6.0139312744140625,
      "learning_rate": 3.6438611197122773e-06,
      "loss": 2.3289308547973633,
      "memory(GiB)": 77.56,
      "step": 102435,
      "token_acc": 0.4886731391585761,
      "train_speed(iter/s)": 1.438849
    },
    {
      "epoch": 4.388843665652714,
      "grad_norm": 5.903653144836426,
      "learning_rate": 3.641339505150554e-06,
      "loss": 2.2890724182128905,
      "memory(GiB)": 77.56,
      "step": 102440,
      "token_acc": 0.4748201438848921,
      "train_speed(iter/s)": 1.438867
    },
    {
      "epoch": 4.389057880981963,
      "grad_norm": 7.067310810089111,
      "learning_rate": 3.6388187304194577e-06,
      "loss": 2.2559886932373048,
      "memory(GiB)": 77.56,
      "step": 102445,
      "token_acc": 0.4868421052631579,
      "train_speed(iter/s)": 1.438856
    },
    {
      "epoch": 4.389272096311212,
      "grad_norm": 5.744080066680908,
      "learning_rate": 3.6362987955646468e-06,
      "loss": 2.105891799926758,
      "memory(GiB)": 77.56,
      "step": 102450,
      "token_acc": 0.53125,
      "train_speed(iter/s)": 1.438868
    },
    {
      "epoch": 4.389486311640461,
      "grad_norm": 7.300531387329102,
      "learning_rate": 3.6337797006317785e-06,
      "loss": 2.1764875411987306,
      "memory(GiB)": 77.56,
      "step": 102455,
      "token_acc": 0.551948051948052,
      "train_speed(iter/s)": 1.438872
    },
    {
      "epoch": 4.38970052696971,
      "grad_norm": 5.576843738555908,
      "learning_rate": 3.6312614456664782e-06,
      "loss": 2.132879638671875,
      "memory(GiB)": 77.56,
      "step": 102460,
      "token_acc": 0.5182481751824818,
      "train_speed(iter/s)": 1.438879
    },
    {
      "epoch": 4.389914742298959,
      "grad_norm": 8.703091621398926,
      "learning_rate": 3.6287440307143816e-06,
      "loss": 1.9866539001464845,
      "memory(GiB)": 77.56,
      "step": 102465,
      "token_acc": 0.5665399239543726,
      "train_speed(iter/s)": 1.43889
    },
    {
      "epoch": 4.390128957628208,
      "grad_norm": 8.888564109802246,
      "learning_rate": 3.6262274558210852e-06,
      "loss": 2.4895147323608398,
      "memory(GiB)": 77.56,
      "step": 102470,
      "token_acc": 0.4968152866242038,
      "train_speed(iter/s)": 1.438902
    },
    {
      "epoch": 4.390343172957457,
      "grad_norm": 5.700866222381592,
      "learning_rate": 3.6237117210321802e-06,
      "loss": 2.2201826095581056,
      "memory(GiB)": 77.56,
      "step": 102475,
      "token_acc": 0.5571955719557196,
      "train_speed(iter/s)": 1.438914
    },
    {
      "epoch": 4.390557388286706,
      "grad_norm": 8.399176597595215,
      "learning_rate": 3.621196826393236e-06,
      "loss": 2.436996269226074,
      "memory(GiB)": 77.56,
      "step": 102480,
      "token_acc": 0.48355263157894735,
      "train_speed(iter/s)": 1.438921
    },
    {
      "epoch": 4.3907716036159545,
      "grad_norm": 6.0412373542785645,
      "learning_rate": 3.6186827719498272e-06,
      "loss": 2.242756462097168,
      "memory(GiB)": 77.56,
      "step": 102485,
      "token_acc": 0.5272108843537415,
      "train_speed(iter/s)": 1.438926
    },
    {
      "epoch": 4.390985818945204,
      "grad_norm": 5.355155944824219,
      "learning_rate": 3.6161695577474784e-06,
      "loss": 2.1781797409057617,
      "memory(GiB)": 77.56,
      "step": 102490,
      "token_acc": 0.5345454545454545,
      "train_speed(iter/s)": 1.438928
    },
    {
      "epoch": 4.391200034274453,
      "grad_norm": 5.779280662536621,
      "learning_rate": 3.613657183831737e-06,
      "loss": 2.3542625427246096,
      "memory(GiB)": 77.56,
      "step": 102495,
      "token_acc": 0.49825783972125437,
      "train_speed(iter/s)": 1.438938
    },
    {
      "epoch": 4.391414249603701,
      "grad_norm": 8.196554183959961,
      "learning_rate": 3.61114565024811e-06,
      "loss": 1.898379135131836,
      "memory(GiB)": 77.56,
      "step": 102500,
      "token_acc": 0.5422535211267606,
      "train_speed(iter/s)": 1.438942
    },
    {
      "epoch": 4.391414249603701,
      "eval_loss": 2.239689350128174,
      "eval_runtime": 14.3867,
      "eval_samples_per_second": 6.951,
      "eval_steps_per_second": 6.951,
      "eval_token_acc": 0.4548919949174079,
      "step": 102500
    },
    {
      "epoch": 4.391628464932951,
      "grad_norm": 5.600312232971191,
      "learning_rate": 3.608634957042095e-06,
      "loss": 2.354815673828125,
      "memory(GiB)": 77.56,
      "step": 102505,
      "token_acc": 0.47314814814814815,
      "train_speed(iter/s)": 1.438618
    },
    {
      "epoch": 4.3918426802622,
      "grad_norm": 5.942442417144775,
      "learning_rate": 3.6061251042591837e-06,
      "loss": 2.652938461303711,
      "memory(GiB)": 77.56,
      "step": 102510,
      "token_acc": 0.4782608695652174,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 4.392056895591448,
      "grad_norm": 5.3099589347839355,
      "learning_rate": 3.6036160919448226e-06,
      "loss": 2.3568065643310545,
      "memory(GiB)": 77.56,
      "step": 102515,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.4386
    },
    {
      "epoch": 4.3922711109206976,
      "grad_norm": 8.654638290405273,
      "learning_rate": 3.6011079201444865e-06,
      "loss": 2.4184709548950196,
      "memory(GiB)": 77.56,
      "step": 102520,
      "token_acc": 0.5171232876712328,
      "train_speed(iter/s)": 1.438594
    },
    {
      "epoch": 4.392485326249947,
      "grad_norm": 8.568259239196777,
      "learning_rate": 3.598600588903611e-06,
      "loss": 2.0518417358398438,
      "memory(GiB)": 77.56,
      "step": 102525,
      "token_acc": 0.5579710144927537,
      "train_speed(iter/s)": 1.438581
    },
    {
      "epoch": 4.392699541579195,
      "grad_norm": 7.474752902984619,
      "learning_rate": 3.5960940982676105e-06,
      "loss": 2.4256092071533204,
      "memory(GiB)": 77.56,
      "step": 102530,
      "token_acc": 0.48859934853420195,
      "train_speed(iter/s)": 1.438567
    },
    {
      "epoch": 4.392913756908444,
      "grad_norm": 5.823642253875732,
      "learning_rate": 3.593588448281893e-06,
      "loss": 2.354109191894531,
      "memory(GiB)": 77.56,
      "step": 102535,
      "token_acc": 0.5107913669064749,
      "train_speed(iter/s)": 1.438576
    },
    {
      "epoch": 4.393127972237694,
      "grad_norm": 6.928681373596191,
      "learning_rate": 3.59108363899186e-06,
      "loss": 2.4611879348754884,
      "memory(GiB)": 77.56,
      "step": 102540,
      "token_acc": 0.5154320987654321,
      "train_speed(iter/s)": 1.438585
    },
    {
      "epoch": 4.393342187566942,
      "grad_norm": 6.965407371520996,
      "learning_rate": 3.5885796704428707e-06,
      "loss": 2.399798583984375,
      "memory(GiB)": 77.56,
      "step": 102545,
      "token_acc": 0.5177514792899408,
      "train_speed(iter/s)": 1.438598
    },
    {
      "epoch": 4.393556402896191,
      "grad_norm": 5.064464569091797,
      "learning_rate": 3.5860765426803056e-06,
      "loss": 2.3432443618774412,
      "memory(GiB)": 77.56,
      "step": 102550,
      "token_acc": 0.5016949152542373,
      "train_speed(iter/s)": 1.438605
    },
    {
      "epoch": 4.393770618225441,
      "grad_norm": 5.201739311218262,
      "learning_rate": 3.5835742557494943e-06,
      "loss": 2.1250043869018556,
      "memory(GiB)": 77.56,
      "step": 102555,
      "token_acc": 0.5448275862068965,
      "train_speed(iter/s)": 1.438611
    },
    {
      "epoch": 4.393984833554689,
      "grad_norm": 7.4527788162231445,
      "learning_rate": 3.58107280969579e-06,
      "loss": 2.435356330871582,
      "memory(GiB)": 77.56,
      "step": 102560,
      "token_acc": 0.46830985915492956,
      "train_speed(iter/s)": 1.4386
    },
    {
      "epoch": 4.394199048883938,
      "grad_norm": 7.7264227867126465,
      "learning_rate": 3.5785722045644898e-06,
      "loss": 2.2883148193359375,
      "memory(GiB)": 77.56,
      "step": 102565,
      "token_acc": 0.528169014084507,
      "train_speed(iter/s)": 1.438586
    },
    {
      "epoch": 4.3944132642131875,
      "grad_norm": 7.861260414123535,
      "learning_rate": 3.576072440400907e-06,
      "loss": 2.727778434753418,
      "memory(GiB)": 77.56,
      "step": 102570,
      "token_acc": 0.5084745762711864,
      "train_speed(iter/s)": 1.438589
    },
    {
      "epoch": 4.394627479542436,
      "grad_norm": 5.563624382019043,
      "learning_rate": 3.5735735172503117e-06,
      "loss": 2.060395622253418,
      "memory(GiB)": 77.56,
      "step": 102575,
      "token_acc": 0.5494505494505495,
      "train_speed(iter/s)": 1.438593
    },
    {
      "epoch": 4.394841694871685,
      "grad_norm": 6.749576568603516,
      "learning_rate": 3.571075435157989e-06,
      "loss": 2.6023681640625,
      "memory(GiB)": 77.56,
      "step": 102580,
      "token_acc": 0.49834983498349833,
      "train_speed(iter/s)": 1.438602
    },
    {
      "epoch": 4.395055910200934,
      "grad_norm": 6.05499792098999,
      "learning_rate": 3.568578194169192e-06,
      "loss": 2.258266639709473,
      "memory(GiB)": 77.56,
      "step": 102585,
      "token_acc": 0.5401459854014599,
      "train_speed(iter/s)": 1.438594
    },
    {
      "epoch": 4.395270125530183,
      "grad_norm": 6.144804954528809,
      "learning_rate": 3.566081794329157e-06,
      "loss": 2.0408275604248045,
      "memory(GiB)": 77.56,
      "step": 102590,
      "token_acc": 0.4981684981684982,
      "train_speed(iter/s)": 1.438602
    },
    {
      "epoch": 4.395484340859432,
      "grad_norm": 6.501466274261475,
      "learning_rate": 3.563586235683103e-06,
      "loss": 2.487798309326172,
      "memory(GiB)": 77.56,
      "step": 102595,
      "token_acc": 0.4882943143812709,
      "train_speed(iter/s)": 1.438604
    },
    {
      "epoch": 4.395698556188681,
      "grad_norm": 5.503562927246094,
      "learning_rate": 3.5610915182762492e-06,
      "loss": 2.207347297668457,
      "memory(GiB)": 77.56,
      "step": 102600,
      "token_acc": 0.543918918918919,
      "train_speed(iter/s)": 1.438598
    },
    {
      "epoch": 4.39591277151793,
      "grad_norm": 6.674918174743652,
      "learning_rate": 3.5585976421537825e-06,
      "loss": 2.199715232849121,
      "memory(GiB)": 77.56,
      "step": 102605,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.438614
    },
    {
      "epoch": 4.396126986847179,
      "grad_norm": 6.238319396972656,
      "learning_rate": 3.556104607360888e-06,
      "loss": 2.393204689025879,
      "memory(GiB)": 77.56,
      "step": 102610,
      "token_acc": 0.5439330543933054,
      "train_speed(iter/s)": 1.438619
    },
    {
      "epoch": 4.396341202176428,
      "grad_norm": 7.066052436828613,
      "learning_rate": 3.553612413942725e-06,
      "loss": 2.345548629760742,
      "memory(GiB)": 77.56,
      "step": 102615,
      "token_acc": 0.4807692307692308,
      "train_speed(iter/s)": 1.4386
    },
    {
      "epoch": 4.3965554175056765,
      "grad_norm": 7.078065872192383,
      "learning_rate": 3.551121061944446e-06,
      "loss": 2.510301208496094,
      "memory(GiB)": 77.56,
      "step": 102620,
      "token_acc": 0.5308219178082192,
      "train_speed(iter/s)": 1.438595
    },
    {
      "epoch": 4.396769632834926,
      "grad_norm": 6.518613338470459,
      "learning_rate": 3.5486305514111805e-06,
      "loss": 2.0175621032714846,
      "memory(GiB)": 77.56,
      "step": 102625,
      "token_acc": 0.508695652173913,
      "train_speed(iter/s)": 1.438608
    },
    {
      "epoch": 4.396983848164175,
      "grad_norm": 8.077980995178223,
      "learning_rate": 3.5461408823880437e-06,
      "loss": 2.7288206100463865,
      "memory(GiB)": 77.56,
      "step": 102630,
      "token_acc": 0.4749262536873156,
      "train_speed(iter/s)": 1.438634
    },
    {
      "epoch": 4.397198063493423,
      "grad_norm": 7.078575611114502,
      "learning_rate": 3.543652054920138e-06,
      "loss": 2.1935190200805663,
      "memory(GiB)": 77.56,
      "step": 102635,
      "token_acc": 0.5304659498207885,
      "train_speed(iter/s)": 1.438636
    },
    {
      "epoch": 4.397412278822673,
      "grad_norm": 6.157913684844971,
      "learning_rate": 3.5411640690525606e-06,
      "loss": 2.056357574462891,
      "memory(GiB)": 77.56,
      "step": 102640,
      "token_acc": 0.5543859649122806,
      "train_speed(iter/s)": 1.438657
    },
    {
      "epoch": 4.397626494151922,
      "grad_norm": 8.185633659362793,
      "learning_rate": 3.538676924830381e-06,
      "loss": 2.2879461288452148,
      "memory(GiB)": 77.56,
      "step": 102645,
      "token_acc": 0.5294117647058824,
      "train_speed(iter/s)": 1.438674
    },
    {
      "epoch": 4.39784070948117,
      "grad_norm": 7.19227409362793,
      "learning_rate": 3.5361906222986517e-06,
      "loss": 2.1820980072021485,
      "memory(GiB)": 77.56,
      "step": 102650,
      "token_acc": 0.5333333333333333,
      "train_speed(iter/s)": 1.438682
    },
    {
      "epoch": 4.3980549248104195,
      "grad_norm": 5.405209064483643,
      "learning_rate": 3.5337051615024207e-06,
      "loss": 2.4287492752075197,
      "memory(GiB)": 77.56,
      "step": 102655,
      "token_acc": 0.4986149584487535,
      "train_speed(iter/s)": 1.438705
    },
    {
      "epoch": 4.398269140139669,
      "grad_norm": 10.108190536499023,
      "learning_rate": 3.531220542486696e-06,
      "loss": 2.46765251159668,
      "memory(GiB)": 77.56,
      "step": 102660,
      "token_acc": 0.5056603773584906,
      "train_speed(iter/s)": 1.43871
    },
    {
      "epoch": 4.398483355468917,
      "grad_norm": 7.863072395324707,
      "learning_rate": 3.528736765296514e-06,
      "loss": 2.4931325912475586,
      "memory(GiB)": 77.56,
      "step": 102665,
      "token_acc": 0.5017667844522968,
      "train_speed(iter/s)": 1.438728
    },
    {
      "epoch": 4.398697570798166,
      "grad_norm": 7.470780372619629,
      "learning_rate": 3.5262538299768553e-06,
      "loss": 2.3910215377807615,
      "memory(GiB)": 77.56,
      "step": 102670,
      "token_acc": 0.48518518518518516,
      "train_speed(iter/s)": 1.438713
    },
    {
      "epoch": 4.398911786127416,
      "grad_norm": 7.713631629943848,
      "learning_rate": 3.5237717365727065e-06,
      "loss": 2.2994256973266602,
      "memory(GiB)": 77.56,
      "step": 102675,
      "token_acc": 0.553030303030303,
      "train_speed(iter/s)": 1.438708
    },
    {
      "epoch": 4.399126001456664,
      "grad_norm": 6.646881103515625,
      "learning_rate": 3.521290485129031e-06,
      "loss": 2.212492752075195,
      "memory(GiB)": 77.56,
      "step": 102680,
      "token_acc": 0.5166051660516605,
      "train_speed(iter/s)": 1.438698
    },
    {
      "epoch": 4.399340216785913,
      "grad_norm": 4.960170269012451,
      "learning_rate": 3.5188100756907815e-06,
      "loss": 2.3690673828125,
      "memory(GiB)": 77.56,
      "step": 102685,
      "token_acc": 0.5079787234042553,
      "train_speed(iter/s)": 1.438704
    },
    {
      "epoch": 4.3995544321151625,
      "grad_norm": 5.805353164672852,
      "learning_rate": 3.5163305083028785e-06,
      "loss": 2.434545707702637,
      "memory(GiB)": 77.56,
      "step": 102690,
      "token_acc": 0.526813880126183,
      "train_speed(iter/s)": 1.438705
    },
    {
      "epoch": 4.399768647444411,
      "grad_norm": 6.253900527954102,
      "learning_rate": 3.5138517830102635e-06,
      "loss": 2.1208045959472654,
      "memory(GiB)": 77.56,
      "step": 102695,
      "token_acc": 0.5684210526315789,
      "train_speed(iter/s)": 1.43871
    },
    {
      "epoch": 4.39998286277366,
      "grad_norm": 7.896162509918213,
      "learning_rate": 3.5113738998578227e-06,
      "loss": 2.377148246765137,
      "memory(GiB)": 77.56,
      "step": 102700,
      "token_acc": 0.496551724137931,
      "train_speed(iter/s)": 1.438703
    },
    {
      "epoch": 4.400197078102909,
      "grad_norm": 5.874502658843994,
      "learning_rate": 3.5088968588904593e-06,
      "loss": 2.0496097564697267,
      "memory(GiB)": 77.56,
      "step": 102705,
      "token_acc": 0.5114754098360655,
      "train_speed(iter/s)": 1.438715
    },
    {
      "epoch": 4.400411293432158,
      "grad_norm": 7.194603443145752,
      "learning_rate": 3.5064206601530423e-06,
      "loss": 2.3831321716308596,
      "memory(GiB)": 77.56,
      "step": 102710,
      "token_acc": 0.4963235294117647,
      "train_speed(iter/s)": 1.438732
    },
    {
      "epoch": 4.400625508761407,
      "grad_norm": 8.089038848876953,
      "learning_rate": 3.503945303690431e-06,
      "loss": 2.2116674423217773,
      "memory(GiB)": 77.56,
      "step": 102715,
      "token_acc": 0.5373134328358209,
      "train_speed(iter/s)": 1.438744
    },
    {
      "epoch": 4.400839724090656,
      "grad_norm": 5.585841655731201,
      "learning_rate": 3.501470789547462e-06,
      "loss": 2.2736770629882814,
      "memory(GiB)": 77.56,
      "step": 102720,
      "token_acc": 0.5110410094637224,
      "train_speed(iter/s)": 1.438748
    },
    {
      "epoch": 4.401053939419905,
      "grad_norm": 5.96068000793457,
      "learning_rate": 3.4989971177689707e-06,
      "loss": 2.405515670776367,
      "memory(GiB)": 77.56,
      "step": 102725,
      "token_acc": 0.5115511551155115,
      "train_speed(iter/s)": 1.43873
    },
    {
      "epoch": 4.401268154749154,
      "grad_norm": 7.212528228759766,
      "learning_rate": 3.496524288399772e-06,
      "loss": 2.602706718444824,
      "memory(GiB)": 77.56,
      "step": 102730,
      "token_acc": 0.4377358490566038,
      "train_speed(iter/s)": 1.438727
    },
    {
      "epoch": 4.401482370078403,
      "grad_norm": 6.397942066192627,
      "learning_rate": 3.4940523014846572e-06,
      "loss": 2.353199577331543,
      "memory(GiB)": 77.56,
      "step": 102735,
      "token_acc": 0.5437262357414449,
      "train_speed(iter/s)": 1.438732
    },
    {
      "epoch": 4.4016965854076515,
      "grad_norm": 6.335452079772949,
      "learning_rate": 3.491581157068413e-06,
      "loss": 2.5333284378051757,
      "memory(GiB)": 77.56,
      "step": 102740,
      "token_acc": 0.45307443365695793,
      "train_speed(iter/s)": 1.438732
    },
    {
      "epoch": 4.401910800736901,
      "grad_norm": 5.486098766326904,
      "learning_rate": 3.4891108551958042e-06,
      "loss": 2.1428770065307616,
      "memory(GiB)": 77.56,
      "step": 102745,
      "token_acc": 0.5115384615384615,
      "train_speed(iter/s)": 1.438739
    },
    {
      "epoch": 4.40212501606615,
      "grad_norm": 8.776297569274902,
      "learning_rate": 3.4866413959115774e-06,
      "loss": 2.274058151245117,
      "memory(GiB)": 77.56,
      "step": 102750,
      "token_acc": 0.5376344086021505,
      "train_speed(iter/s)": 1.438712
    },
    {
      "epoch": 4.402339231395398,
      "grad_norm": 7.713545322418213,
      "learning_rate": 3.4841727792604805e-06,
      "loss": 2.3287942886352537,
      "memory(GiB)": 77.56,
      "step": 102755,
      "token_acc": 0.49851632047477745,
      "train_speed(iter/s)": 1.438731
    },
    {
      "epoch": 4.402553446724648,
      "grad_norm": 8.110536575317383,
      "learning_rate": 3.481705005287228e-06,
      "loss": 2.162745475769043,
      "memory(GiB)": 77.56,
      "step": 102760,
      "token_acc": 0.548951048951049,
      "train_speed(iter/s)": 1.438728
    },
    {
      "epoch": 4.402767662053897,
      "grad_norm": 6.409671306610107,
      "learning_rate": 3.479238074036528e-06,
      "loss": 2.3971099853515625,
      "memory(GiB)": 77.56,
      "step": 102765,
      "token_acc": 0.5057471264367817,
      "train_speed(iter/s)": 1.43873
    },
    {
      "epoch": 4.402981877383145,
      "grad_norm": 8.387430191040039,
      "learning_rate": 3.476771985553068e-06,
      "loss": 2.7748668670654295,
      "memory(GiB)": 77.56,
      "step": 102770,
      "token_acc": 0.4749262536873156,
      "train_speed(iter/s)": 1.43875
    },
    {
      "epoch": 4.4031960927123945,
      "grad_norm": 7.343550205230713,
      "learning_rate": 3.474306739881522e-06,
      "loss": 2.3631175994873046,
      "memory(GiB)": 77.56,
      "step": 102775,
      "token_acc": 0.5017421602787456,
      "train_speed(iter/s)": 1.438751
    },
    {
      "epoch": 4.403410308041644,
      "grad_norm": 6.574812889099121,
      "learning_rate": 3.47184233706655e-06,
      "loss": 2.47608642578125,
      "memory(GiB)": 77.56,
      "step": 102780,
      "token_acc": 0.4939759036144578,
      "train_speed(iter/s)": 1.438751
    },
    {
      "epoch": 4.403624523370892,
      "grad_norm": 4.98598051071167,
      "learning_rate": 3.469378777152815e-06,
      "loss": 2.6040485382080076,
      "memory(GiB)": 77.56,
      "step": 102785,
      "token_acc": 0.46111111111111114,
      "train_speed(iter/s)": 1.438734
    },
    {
      "epoch": 4.403838738700141,
      "grad_norm": 7.161421298980713,
      "learning_rate": 3.4669160601849272e-06,
      "loss": 2.1930047988891603,
      "memory(GiB)": 77.56,
      "step": 102790,
      "token_acc": 0.5504587155963303,
      "train_speed(iter/s)": 1.43876
    },
    {
      "epoch": 4.404052954029391,
      "grad_norm": 7.1895623207092285,
      "learning_rate": 3.4644541862075054e-06,
      "loss": 2.4721134185791014,
      "memory(GiB)": 77.56,
      "step": 102795,
      "token_acc": 0.47750865051903113,
      "train_speed(iter/s)": 1.438757
    },
    {
      "epoch": 4.404267169358639,
      "grad_norm": 7.5333781242370605,
      "learning_rate": 3.461993155265153e-06,
      "loss": 2.3281063079833983,
      "memory(GiB)": 77.56,
      "step": 102800,
      "token_acc": 0.5133333333333333,
      "train_speed(iter/s)": 1.438762
    },
    {
      "epoch": 4.404481384687888,
      "grad_norm": 8.899251937866211,
      "learning_rate": 3.4595329674024456e-06,
      "loss": 2.4926557540893555,
      "memory(GiB)": 77.56,
      "step": 102805,
      "token_acc": 0.44516129032258067,
      "train_speed(iter/s)": 1.438755
    },
    {
      "epoch": 4.404695600017138,
      "grad_norm": 6.462435245513916,
      "learning_rate": 3.4570736226639588e-06,
      "loss": 2.4888607025146485,
      "memory(GiB)": 77.56,
      "step": 102810,
      "token_acc": 0.47474747474747475,
      "train_speed(iter/s)": 1.438761
    },
    {
      "epoch": 4.404909815346386,
      "grad_norm": 6.181706428527832,
      "learning_rate": 3.454615121094251e-06,
      "loss": 2.405670166015625,
      "memory(GiB)": 77.56,
      "step": 102815,
      "token_acc": 0.49645390070921985,
      "train_speed(iter/s)": 1.438755
    },
    {
      "epoch": 4.405124030675635,
      "grad_norm": 6.386962890625,
      "learning_rate": 3.4521574627378485e-06,
      "loss": 2.4266725540161134,
      "memory(GiB)": 77.56,
      "step": 102820,
      "token_acc": 0.5103448275862069,
      "train_speed(iter/s)": 1.438766
    },
    {
      "epoch": 4.405338246004884,
      "grad_norm": 7.187178134918213,
      "learning_rate": 3.4497006476392813e-06,
      "loss": 2.229561996459961,
      "memory(GiB)": 77.56,
      "step": 102825,
      "token_acc": 0.4965986394557823,
      "train_speed(iter/s)": 1.438753
    },
    {
      "epoch": 4.405552461334133,
      "grad_norm": 4.833610534667969,
      "learning_rate": 3.4472446758430533e-06,
      "loss": 2.310893249511719,
      "memory(GiB)": 77.56,
      "step": 102830,
      "token_acc": 0.5296052631578947,
      "train_speed(iter/s)": 1.43875
    },
    {
      "epoch": 4.405766676663382,
      "grad_norm": 5.882937431335449,
      "learning_rate": 3.4447895473936453e-06,
      "loss": 2.151133346557617,
      "memory(GiB)": 77.56,
      "step": 102835,
      "token_acc": 0.5286195286195287,
      "train_speed(iter/s)": 1.438752
    },
    {
      "epoch": 4.405980891992631,
      "grad_norm": 7.544086933135986,
      "learning_rate": 3.442335262335561e-06,
      "loss": 2.4032720565795898,
      "memory(GiB)": 77.56,
      "step": 102840,
      "token_acc": 0.5310344827586206,
      "train_speed(iter/s)": 1.438748
    },
    {
      "epoch": 4.40619510732188,
      "grad_norm": 5.912472248077393,
      "learning_rate": 3.439881820713248e-06,
      "loss": 2.4105043411254883,
      "memory(GiB)": 77.56,
      "step": 102845,
      "token_acc": 0.4718498659517426,
      "train_speed(iter/s)": 1.438752
    },
    {
      "epoch": 4.406409322651129,
      "grad_norm": 6.072861671447754,
      "learning_rate": 3.4374292225711423e-06,
      "loss": 2.415452575683594,
      "memory(GiB)": 77.56,
      "step": 102850,
      "token_acc": 0.5098591549295775,
      "train_speed(iter/s)": 1.438743
    },
    {
      "epoch": 4.406623537980378,
      "grad_norm": 6.437143325805664,
      "learning_rate": 3.434977467953693e-06,
      "loss": 2.346902275085449,
      "memory(GiB)": 77.56,
      "step": 102855,
      "token_acc": 0.5,
      "train_speed(iter/s)": 1.438762
    },
    {
      "epoch": 4.406837753309627,
      "grad_norm": 8.530142784118652,
      "learning_rate": 3.4325265569053134e-06,
      "loss": 2.1180702209472657,
      "memory(GiB)": 77.56,
      "step": 102860,
      "token_acc": 0.5441696113074205,
      "train_speed(iter/s)": 1.438751
    },
    {
      "epoch": 4.407051968638876,
      "grad_norm": 6.524062156677246,
      "learning_rate": 3.430076489470385e-06,
      "loss": 2.5981245040893555,
      "memory(GiB)": 77.56,
      "step": 102865,
      "token_acc": 0.4542857142857143,
      "train_speed(iter/s)": 1.438745
    },
    {
      "epoch": 4.407266183968125,
      "grad_norm": 5.981161594390869,
      "learning_rate": 3.4276272656933174e-06,
      "loss": 2.2366401672363283,
      "memory(GiB)": 77.56,
      "step": 102870,
      "token_acc": 0.48863636363636365,
      "train_speed(iter/s)": 1.438755
    },
    {
      "epoch": 4.4074803992973735,
      "grad_norm": 5.634190559387207,
      "learning_rate": 3.425178885618474e-06,
      "loss": 2.318357467651367,
      "memory(GiB)": 77.56,
      "step": 102875,
      "token_acc": 0.48148148148148145,
      "train_speed(iter/s)": 1.438759
    },
    {
      "epoch": 4.407694614626623,
      "grad_norm": 6.3136162757873535,
      "learning_rate": 3.422731349290198e-06,
      "loss": 2.2633522033691404,
      "memory(GiB)": 77.56,
      "step": 102880,
      "token_acc": 0.5413533834586466,
      "train_speed(iter/s)": 1.438763
    },
    {
      "epoch": 4.407908829955872,
      "grad_norm": 6.478562355041504,
      "learning_rate": 3.4202846567528424e-06,
      "loss": 2.2929569244384767,
      "memory(GiB)": 77.56,
      "step": 102885,
      "token_acc": 0.5254237288135594,
      "train_speed(iter/s)": 1.438769
    },
    {
      "epoch": 4.40812304528512,
      "grad_norm": 4.679701805114746,
      "learning_rate": 3.4178388080507274e-06,
      "loss": 2.3439220428466796,
      "memory(GiB)": 77.56,
      "step": 102890,
      "token_acc": 0.4779874213836478,
      "train_speed(iter/s)": 1.43878
    },
    {
      "epoch": 4.40833726061437,
      "grad_norm": 5.395583629608154,
      "learning_rate": 3.4153938032281454e-06,
      "loss": 2.462462615966797,
      "memory(GiB)": 77.56,
      "step": 102895,
      "token_acc": 0.48284960422163586,
      "train_speed(iter/s)": 1.43879
    },
    {
      "epoch": 4.408551475943619,
      "grad_norm": 7.386928081512451,
      "learning_rate": 3.412949642329416e-06,
      "loss": 2.326919364929199,
      "memory(GiB)": 77.56,
      "step": 102900,
      "token_acc": 0.48736462093862815,
      "train_speed(iter/s)": 1.438804
    },
    {
      "epoch": 4.408765691272867,
      "grad_norm": 10.29039478302002,
      "learning_rate": 3.4105063253988045e-06,
      "loss": 2.6700963973999023,
      "memory(GiB)": 77.56,
      "step": 102905,
      "token_acc": 0.4552238805970149,
      "train_speed(iter/s)": 1.438803
    },
    {
      "epoch": 4.4089799066021165,
      "grad_norm": 6.046471118927002,
      "learning_rate": 3.408063852480581e-06,
      "loss": 2.424648094177246,
      "memory(GiB)": 77.56,
      "step": 102910,
      "token_acc": 0.4934210526315789,
      "train_speed(iter/s)": 1.438803
    },
    {
      "epoch": 4.409194121931366,
      "grad_norm": 6.12212610244751,
      "learning_rate": 3.4056222236189815e-06,
      "loss": 2.226657485961914,
      "memory(GiB)": 77.56,
      "step": 102915,
      "token_acc": 0.4879725085910653,
      "train_speed(iter/s)": 1.438799
    },
    {
      "epoch": 4.409408337260614,
      "grad_norm": 6.161651134490967,
      "learning_rate": 3.4031814388582384e-06,
      "loss": 2.162016677856445,
      "memory(GiB)": 77.56,
      "step": 102920,
      "token_acc": 0.5346153846153846,
      "train_speed(iter/s)": 1.438802
    },
    {
      "epoch": 4.409622552589863,
      "grad_norm": 8.177957534790039,
      "learning_rate": 3.4007414982425766e-06,
      "loss": 2.0994197845458986,
      "memory(GiB)": 77.56,
      "step": 102925,
      "token_acc": 0.5490909090909091,
      "train_speed(iter/s)": 1.438809
    },
    {
      "epoch": 4.409836767919113,
      "grad_norm": 6.875726222991943,
      "learning_rate": 3.3983024018162004e-06,
      "loss": 2.3808807373046874,
      "memory(GiB)": 77.56,
      "step": 102930,
      "token_acc": 0.5229007633587787,
      "train_speed(iter/s)": 1.438812
    },
    {
      "epoch": 4.410050983248361,
      "grad_norm": 5.5511627197265625,
      "learning_rate": 3.395864149623296e-06,
      "loss": 1.9341264724731446,
      "memory(GiB)": 77.56,
      "step": 102935,
      "token_acc": 0.5606557377049181,
      "train_speed(iter/s)": 1.438827
    },
    {
      "epoch": 4.41026519857761,
      "grad_norm": 5.994646072387695,
      "learning_rate": 3.393426741708028e-06,
      "loss": 1.996798324584961,
      "memory(GiB)": 77.56,
      "step": 102940,
      "token_acc": 0.5318471337579618,
      "train_speed(iter/s)": 1.438837
    },
    {
      "epoch": 4.4104794139068595,
      "grad_norm": 5.379752159118652,
      "learning_rate": 3.390990178114556e-06,
      "loss": 2.398251533508301,
      "memory(GiB)": 77.56,
      "step": 102945,
      "token_acc": 0.5015873015873016,
      "train_speed(iter/s)": 1.438839
    },
    {
      "epoch": 4.410693629236108,
      "grad_norm": 6.1705546379089355,
      "learning_rate": 3.388554458887011e-06,
      "loss": 2.2360561370849608,
      "memory(GiB)": 77.56,
      "step": 102950,
      "token_acc": 0.5434782608695652,
      "train_speed(iter/s)": 1.438836
    },
    {
      "epoch": 4.410907844565357,
      "grad_norm": 8.667627334594727,
      "learning_rate": 3.386119584069536e-06,
      "loss": 2.3674911499023437,
      "memory(GiB)": 77.56,
      "step": 102955,
      "token_acc": 0.502127659574468,
      "train_speed(iter/s)": 1.438829
    },
    {
      "epoch": 4.411122059894606,
      "grad_norm": 6.950751304626465,
      "learning_rate": 3.383685553706234e-06,
      "loss": 2.710087776184082,
      "memory(GiB)": 77.56,
      "step": 102960,
      "token_acc": 0.4533333333333333,
      "train_speed(iter/s)": 1.438818
    },
    {
      "epoch": 4.411336275223855,
      "grad_norm": 5.777164936065674,
      "learning_rate": 3.381252367841192e-06,
      "loss": 2.175288772583008,
      "memory(GiB)": 77.56,
      "step": 102965,
      "token_acc": 0.5572755417956656,
      "train_speed(iter/s)": 1.438833
    },
    {
      "epoch": 4.411550490553104,
      "grad_norm": 6.37148904800415,
      "learning_rate": 3.3788200265185033e-06,
      "loss": 2.2468969345092775,
      "memory(GiB)": 77.56,
      "step": 102970,
      "token_acc": 0.5583941605839416,
      "train_speed(iter/s)": 1.438831
    },
    {
      "epoch": 4.411764705882353,
      "grad_norm": 7.066985130310059,
      "learning_rate": 3.376388529782215e-06,
      "loss": 2.5912525177001955,
      "memory(GiB)": 77.56,
      "step": 102975,
      "token_acc": 0.46598639455782315,
      "train_speed(iter/s)": 1.438843
    },
    {
      "epoch": 4.411978921211602,
      "grad_norm": 6.8794331550598145,
      "learning_rate": 3.3739578776763814e-06,
      "loss": 2.218465042114258,
      "memory(GiB)": 77.56,
      "step": 102980,
      "token_acc": 0.519298245614035,
      "train_speed(iter/s)": 1.438834
    },
    {
      "epoch": 4.412193136540851,
      "grad_norm": 7.623315811157227,
      "learning_rate": 3.3715280702450502e-06,
      "loss": 2.2979253768920898,
      "memory(GiB)": 77.56,
      "step": 102985,
      "token_acc": 0.5209003215434084,
      "train_speed(iter/s)": 1.438825
    },
    {
      "epoch": 4.4124073518701,
      "grad_norm": 6.517643928527832,
      "learning_rate": 3.3690991075322254e-06,
      "loss": 2.3695747375488283,
      "memory(GiB)": 77.56,
      "step": 102990,
      "token_acc": 0.5095057034220533,
      "train_speed(iter/s)": 1.438826
    },
    {
      "epoch": 4.4126215671993485,
      "grad_norm": 7.054264545440674,
      "learning_rate": 3.3666709895819104e-06,
      "loss": 2.424116325378418,
      "memory(GiB)": 77.56,
      "step": 102995,
      "token_acc": 0.4950166112956811,
      "train_speed(iter/s)": 1.438823
    },
    {
      "epoch": 4.412835782528598,
      "grad_norm": 6.089029788970947,
      "learning_rate": 3.364243716438098e-06,
      "loss": 2.2391313552856444,
      "memory(GiB)": 77.56,
      "step": 103000,
      "token_acc": 0.5394321766561514,
      "train_speed(iter/s)": 1.438825
    },
    {
      "epoch": 4.412835782528598,
      "eval_loss": 1.950676679611206,
      "eval_runtime": 14.8525,
      "eval_samples_per_second": 6.733,
      "eval_steps_per_second": 6.733,
      "eval_token_acc": 0.509641873278237,
      "step": 103000
    }
  ],
  "logging_steps": 5,
  "max_steps": 116705,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.717583870903251e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}